EC, March 14, 2006, No M.3868

THE COMMISSION OF THE EUROPEAN COMMUNITIES,

Having regard to the Treaty establishing the European Community, Having regard to the Agreement on the European Economic Area, and in particular Article 57 thereof, Having regard to Council Regulation (EC) No 139-2004 of 20 January 2004 on the control of concentrations between undertakings (1), and in particular Article 8(2) thereof, Having regard to the Commission's decision of 18 October 2005 to initiate proceedings in this case, Having given the undertakings concerned the opportunity to make known their views on the objections raised by the Commission, Having regard to the opinion of the Advisory Committee on Concentrations (2), Having regard to the final report of the Hearing Officer in this case (3), Whereas:

(1) On 13 September 2005, the Commission received a notification of a proposed concentration pursuant to Article 4 of Regulation (EC) No 139-2004 ("the Merger Regulation") by which the undertaking DONG A/S ("DONG", Denmark, notifying party) acquires within the meaning of Article 3(1)(b) of the Merger Regulation sole control of the undertakings Elsam A/S ("Elsam", Denmark), Energi E2 ("E2", Denmark), Københavns Energi Holding A/S ("KE", Denmark) and Frederiksberg Elnet A/S ("FE", Denmark) by way of purchase of shares and assets. (DONG, Elsam, E2, KE and FE are henceforth collectively referred to as "the parties".)

(2) After examination of the notification, the Commission has concluded that the notified operation falls within the scope of the Merger Regulation and raises concerns as to its compatibility with the common market.

I. THE PARTIES

(3) DONG is the Danish state-owned gas incumbent active in exploration, production, offshore transport and sale of oil and natural gas, as well as storage and distribution of natural gas. It also has minor activities related to wind electricity generation and supply of electricity and heat.

(4) Elsam and E2 are the Danish electricity generation incumbents in West Denmark (Elsam) and East Denmark (E2), respectively. They are both active in production and trading of electricity (financial and physical) on the wholesale market and in production of district heating. Elsam, since its acquisition of the (East Danish) electricity retailer NESA in 2004, also has substantial activities in electricity retailing to household and business customers. Elsam and E2 have, on the one hand, a core ownership of local authorities and, on the other hand, substantial shareholdings by DONG and Vattenfall (in Elsam) and by NESA and KE (in E2).

(5) KE and FE supply household and business customers with electricity in the Copenhagen area. They are currently owned by the City of Copenhagen and the City of Frederiksberg, respectively.

II. THE OPERATION AND THE CONCENTRATION

(6) In January 2005, Vattenfall AB ("Vattenfall", Sweden) acquired 35.3% of Elsam's shares. Between 4 and 9 February 2005, DONG, which had previously owned 24.1% of Elsam's shares, entered into option agreements with several local utilities for the acquisition of an additional 40.6% of Elsam's shares. In order to resolve the deadlock situation in Elsam, on 31 May 2005 DONG and Vattenfall entered into an agreement under which Vattenfall would transfer its 35.3% shareholding in Elsam to DONG and receive, as a counterpart, assets (mainly power plants) of Elsam and E2 corresponding to the value of the transferred shareholding. As a result of this agreement, and the preceding transactions, DONG acquires sole control of Elsam and E2, less the assets of these companies that will be attributed to Vattenfall according to that agreement.

(7) The acquisition of altogether 64% of E2's shares by DONG, the remaining 36% being held by Elsam's subsidiary NESA, is therefore a necessary precondition for the completion of the DONG/Vattenfall agreement. On 7 February 2005, DONG acquired KE from the City of Copenhagen, including KE's 34% share in E2. On 10 February 2005, DONG acquired FE from the City of Frederiksberg, including FE's 2.3% share in E2. In addition, DONG entered into parallel agreements with other E2 shareholders4 to increase its shareholding in E2 by a further 28% to approximately 64%.

(8) Both the acquisition of FE and the agreements with the other E2 shareholders are legally conditional upon the successful acquisition of KE (and the 34% of E2's shares held by KE) by DONG. The shares transfer agreement between the City of Frederiksberg and DONG contains the explicit condition that the "the main agreement (on purchase of shares in Copenhagen Energy Holding A/S) of 7 February 2005 between the Municipality of Copenhagen and the buyer (DONG) becomes final". (5) Further, DONG's agreements with the Cities of Helsingør, Roskilde, Hillerød and Slagelse (SK-EL) and with the cooperative NVE-SEAS relating to the transfer of altogether 28% of shares in E2 contain the explicit condition that "the agreement between the Municipality of Copenhagen and DONG concerning DONG's take-over of Copenhagen Energy Holding A/S is finalised". (6)

(9) In view of Elsam's indirect (through NESA) 36% shareholding in E2, the full acquisition of Elsam by DONG is also a legal precondition for the acquisition of full control over E2. Without NESA's shares in E2, DONG would not be in a position to divest important assets of E2 to Vattenfall. In turn, without the divestiture of some of E2's assets to Vattenfall, DONG would not be able to acquire Vattenfall's 35.3% shareholding in Elsam. All these transactions by DONG are thus closely connected with each other as well as with the DONG/Vattenfall agreement. The present concentration within the meaning of Article 3(1) of the EC Merger Regulation thus comprises the acquisition of control by DONG of Elsam, E2, KE and FE.

(10) The asset acquisition by Vattenfall constitutes a separate concentration. (7) Through this concentration Vattenfall acquires electricity generation assets currently owned by Elsam and E2, and predominantly located in Western and Eastern Denmark. (8) These assets have a combined electricity generation capacity of 2452 MW (9). Most acquired power plants are co-generation plants, producing both electricity and heat, and are fuelled by coal, biomass or natural gas; the remainder is wind power, mainly in West Denmark.

III. COMMUNITY DIMENSION

(11) The undertakings concerned have a combined aggregate world-wide turnover of more than EUR 2.5 billion (10). In each of at least three Member States the combined aggregate turnover of all the undertakings concerned is more than EUR 100 million, and in three of these Member States the aggregate turnover of each of at least two of the undertakings concerned is more than EUR 25 million. In addition, the aggregate Community-wide turnover of each of at least two of the undertakings concerned is more than EUR 100 million. Furthermore, DONG does not achieve more than two thirds of its aggregate Community-wide turnover within one and the same Member State. The notified operation therefore has a Community dimension pursuant to Article 1 (3) of the EC Merger Regulation.

IV. PROCEDURE

(12) By decision of 18 October 2005, the Commission found that the notified operation raised serious doubts as to its compatibility with the Common Market and the functioning of the EEA Agreement. The Commission accordingly initiated proceedings pursuant to Article 6 (1) (c) of Regulation (EC) No 139-2004.

(13) On 09 November 2005, DONG provided a response to the Commission's decision pursuant to Article 6 (1) (c) of Regulation (EC) No 139-2004 ("Response to the 6(1)(c) decision").

(14) On 19 December 2005, a Statement of Objections ("SO") was sent to DONG. In the following days, DONG was granted access to the Commission's file. In its reply to the SO ("the Reply") dated 05 January 2006, DONG made use of the opportunity to comment on the preliminary findings set out in the SO.

(15) DONG and the other involved parties did not request to develop their arguments in a formal oral hearing.

(16) Upon application, the Hearing Officer, by letter of 22 December 2005, admitted Naturgas Fyn as an interested third party pursuant to Article 18 (4) of the Merger Regulation.

(17) On 30 January 2006, DONG formally submitted commitments in view of remedying the Commission's competition concerns. It amended these commitments on 01 March 2006.

(18) The Advisory Committee on Concentrations discussed the draft decision on 01 March 2006.

V. LEGAL, REGULATORY AND STRUCTURAL FRAMEWORK FOR THE NATURAL GAS AND ELECTRICITY SECTORS IN DENMARK

(19) The notified concentration primarily concerns the natural gas sector and the electricity sector in Denmark. For these sectors, the consideration of the legal and regulatory framework is a necessary prerequisite for defining markets and assessing the impact of the concentration on them.

A. NATURAL GAS SECTOR

1. Legal and regulatory framework

(20) The Danish natural gas market was first opened to competition for the largest customers with the passing of the Danish Natural Gas Supply Act on 1 July 2000, according to the requirements of the European Union's first Gas Directive (11). The Natural Gas Supply Act regulates most aspects of natural gas activities, except the production of natural gas, which is regulated by the Subsoil Act (12). The Danish Natural Gas Supply Act applies to the transmission, distribution, supply and storage of natural gas, including liquefied natural gas (LNG). Pursuant to the Act, all customers, including household customers, are since 2004 entitled to choose their gas supplier at their own discretion. Moreover, all network operations and gas trading activities have been unbundled (Cf. paragraphs (23) and (37) below). Storages and the offshore pipelines connecting the fields in the Danish part of the North Sea with mainland Denmark are both owned by DONG:

(21) The consumer can switch suppliers within a minimum 1 month/maximum 2 months timeframe. The customer only has to make an agreement with a new supplier. During the first year of full liberalisation of the natural gas market in 2004, 20% of volume has switched. Switching is free of charge.

(22) The Danish Energy Regulatory Authority ("DERA") is an independent supervisory body in the energy sector. The authority was set up as part of the liberalisation of the Danish energy sector and began its activities on 1 January 2000. It supervises the Danish energy sector, including natural gas, electricity and district heating. It regulates the prices and terms of supply fixed by the actors in this sector, including the terms applicable to access to transmission and distribution networks. DERA's decisions can be appealed to the Energy Board of Appeal.

(23) In 2004 the gas transmission network and the transmission system operator were fully unbundled when DONG Transmission was separated from the DONG Group and incorporated into a separate, independent, fully state-owned company, Gastra A/S. On 1 January 2005, the two Danish electricity system operators Elkraft System (Denmark East), Eltra (Denmark West) and Gastra were merged to form the new Transmission System Operator (TSO) Energinet.dk which is responsible for the transmission in both the electricity and gas sector (13). Energinet.dk is an independent public corporation owned by the Danish state under the Ministry of Transport and Energy and has its own Supervisory Board. Energinet.dk, which owns the transmission system, offers capacity by entering into agreements on balancing services and emergency supplies. In addition, it is responsible for transmission, security of supply, market facilitation and general physical balance under a licence granted pursuant to the Danish Gas Supply Act. It also offers balancing services (hourly flexibility), interruptible transportation and longer term transportation agreements.

(24) According to the Natural Gas Supply Act, Energinet.dk is responsible for security of supply. In this context Energinet.dk is responsible for ensuring that there is sufficient gas stored together with other measures to handle an emergency situation. Gas suppliers are required to have a certain percentage of gas in storage during the winter months due to security of supply. The latest version of the Rules for Gas Transport (14) entered into force on 1 October 2005 and regulates the Danish natural gas system in accordance with the Danish Natural Gas Supply Act.

2. Natural Gas Demand, Production and international trade

(25) In 2004 Danish gas consumption was about 4.2 bcm (15). However, yearly gas consumption varies from year to year, in particular, on account of winter temperatures.

(26) Denmark is a net exporter of natural gas, due to the production activity on the Danish Continental Shelf. In 2004, in total 3.5 (16) bcm natural gas were transported through the Danish transmission system to the Swedish and German markets, 0.9 bcm of which was to Sweden and 2.6 bcm of which was to Germany (17).

(27) The gas is produced on the basis of a concession, which is assigned for each production field. The DUC consortium (18) sells most of the gas it produces to DONG Naturgas in accordance with several long term take-or-pay agreements from 1979, 1990 and 1993. The Syd Arne Group (19) and the Lulita Parties (20) sell the remaining volume, which corresponds to less than 10%. Syd Arne sells gas exclusively to DONG on a depletion type of contract.

(28) In 2004, the Danish gas production in North Sea fields was 10.93 bcm, which constitutes an increase of 0.72 bcm in comparison to 2003. Natural gas sales increased from [...]* bcm in 2003 to [...]* bcm in 2004, of which [...]* bcm were sold to DONG (21). The increase in gas sales is due to a new pipeline for gas export, connecting the Tyra field to the Dutch NOGAT (22) pipeline. The pipeline was commissioned on 18 July 2004 and has a capacity of 16.5 million m3 ("mcm"), (23) per day. In 2004 about 10% of all gas sold was exported through the NOGAT pipeline (24). Based on the ongoing contracts it is expected that sale of gas will be about 8.5 bcm for the years 2005-08 (25). The forecast includes natural gas production resulting from new contracts for the export of gas through the pipeline from Tyra via the NOGAT pipeline to the Netherlands. According to the Subsoil Act, in May 2005 the Minister for Transport and Energy launched an invitation to parties wishing to apply for the concession for the exploration and production of Hydrocarbons in an area of the North Sea. The DUC consortium are likely to win an extension of their concession from 2012 until 2042 (26).

3. Infrastructure

(29) The system consists onshore of one transmission and several distribution sub-systems. Unlike the case of electricity, there is a pipeline connection between Jutland and Zealand, which also belongs to the overall transmission system and which does not experience congestion. The transmission system (high pressure network) acts as the transport system's highway and is connected to two storage facilities and to three other entry/exit points (one toward the Danish offshore pipelines at Nybro, one toward Germany at Ellund and one toward Sweden at Dragør). It is also connected to the distribution systems. The distribution systems (low pressure network) transport the natural gas from the transmission network to end users. (27)

(30) The North Sea production areas are connected to the Danish mainland through the Tyra-Nybro offshore pipeline which has a capacity of 26 mcm per day (28). A second upstream pipeline runs from the South Arne fields to Nybro. Both pipelines are owned and operated by DONG Naturgas A/S. The Tyra-Nybro pipeline has a negotiated third-party access (TPA) regime with published tariffs.

(31) The Danish market is connected to Germany and the continental gas market via the DEUDAN onshore pipeline which runs from the Danish-German border at Ellund to Quarnstedt in Germany. This connection is owned by DONG through Dangas GmbH (49%), E.ON Ruhrgas Transportation GmbH & Co. KG (24.99%) and BEB Erdgas und Erdöl GmbH (26.01%). The DEUDAN has a capacity of approximately 3 bcm per year in the north to south direction (29). Since July 2005 the DEUDAN has been operated as a pipe-in-pipe system whereby all three owners are operating as separate TSOs with respect to their own capacity allotted on the basis of their ownership. Prior to the establishment of the new system, BEB operated the pipeline and the owners had capacity reservation rights according to their ownership rights. At present DONG is using the German/Danish border connection to export [...]* (30) a year, of which ca. [...]* physically leave Denmark. Furthermore, Denmark is connected to Sweden between Dragør and Malmö. This connection is operated by Svenska Kraftnät, the Swedish TSO.

(32) The Danish transmission system is designed as an entry-exit system with three transit/entry/exit points (Nybro, Ellund, Dragør) and a national exit zone (31). Energinet.dk offers capacity contracts to the incumbent DONG as well as to all other shippers on non-discriminatory and regulated terms. The regulated tariffs are published on Energinet.dk's website. The tariffs comprise capacity charges (entry and exit fees), which are applicable when transporting through the system, and a commodity element, which is only applicable once, upon exit (32). Furthermore, there is separately a firm and an interruptible emergency supply tariff (33) that shippers must pay when transporting to the national exit zones but not when transporting to one of the three transit/entry/exit points. The split between the capacity charge and the commodity component is 75%/25% respectively. In 2004, Energinet.dk's revenues from gas transportation were EUR while 7.7 bcm of natural gas were transported.

(33) Energinet.dk facilitates trading and operates a virtual Gas Transfer Facility (GTF) - opened in May 2004 - in which it is possible for shippers to transfer gas bilaterally free of charge prior to and on the day of gas delivery. The service is suitable for shippers who have either a surplus or a shortfall of gas available in the transmission system. Gas transfer through GTF does not require reservation of both entry and exit capacity. Therefore, it is also intended for those who mainly want to deliver natural gas into the Danish gas transmission system, and thus have only booked entry capacity or a very small exit capacity. It is also meant for those who have mainly bought exit capacity, if for instance they engage in delivery of gas to the distribution areas or delivery for transit. In parallel to the GTF, Energinet.dk operates a Capacity Transfer Facility (CTF) and a Balance Transfer Facility (BTF) for secondary market trading. These facilities enable shippers to transfer excess capacity or balance margin, respectively, to another shipper in the system.

(34) Currently, there is no congestion in the Danish transmission system, either nationally, or on cross-border links. If, in the future, contractual congestion occurs, Energinet.dk will make the unused capacity available to the market on an interruptible basis. The balancing area corresponds to Energinet.dk's transmission system.

(35) Energinet.dk uses several tools to ensure system balance. These include the use of "line pack" (34) and the use of both of the two Danish storages. The system-needs reflect the seasonal fluctuations in consumption (due to the temperature differences between summer and winter). System-users are given incentives from the balancing charges. For daily balancing, no hourly restrictions or constraints apply. Shippers can pool imbalances in their portfolios. It is possible to trade imbalances between individual network users/shippers via the Gas Transfer Facility. The gas day begins and ends at 6.00 a.m. Imbalances are settled thereafter, and shippers are informed of their gas balance at 11.00 a.m. on each day at the latest (35).

(36) With respect to utilisation of capacity, the TSO Energinet.dk publishes both data and information on its website. This information and data includes aggregate hourly, daily flows and aggregate booked capacity on each entry and each exit point. Moreover, total available capacity is published, as is information regarding planned investments and maintenance.

(37) The total length of the Danish distribution networks amounts to around 17,000 km of distribution pipeline and is owned by DONG Distribution A/S (36) in South Jutland and Southwest Zealand, and three regional distribution companies, namely Naturgas Fyn A/S, Hovedstadens Naturgas I/S (HNG), and Naturgas Midt-Nord I/S (MN) (37). While DONG Distribution A/S is an integrated part of DONG Naturgas A/S, the other distribution companies are owned by municipalities. All distribution companies have been legally unbundled since 1 January 2003. The distribution companies have been regulated by an income-cap regulation since 1 January 2005.

(38) DONG Lager A/S ("DONG Storage") owns the storage facilities and offers storage of natural gas. Storage injection capacity is the maximum hourly capacity agreed by a storage customer with DONG Storage with a view to injection of natural gas into the storage facility in accordance with a storage agreement. The hourly capacity agreed in the opposite direction is called withdrawal capacity. The conditions for storage injections and withdrawals are governed by DONG Storage's rules for gas storage. DONG Storage is responsible for the storage function under a licence granted pursuant to the Danish Natural Gas Supply Act.

(39) There are two gas storage facilities in Denmark. The storage in Stenlille (from 1994) has a working capacity of 300 mcm, while the storage in Lille Torup (from 1987) has a working capacity of 400 mcm (38).

(40) The total storage capacity for withdrawable gas is approximately 700 mcm (39), which is unlikely to be increased in the near future. In 2004/2005, DONG Lager sold [150- 200]* mcm of storage capacity to the transmission company Energinet.dk (40) for emergency supplies. According to the regulation DONG Storage is able to turn down a request for storage if there is no available capacity. However in 2001 DONG Storage has committed to a remedy that a minimum of 50 million mcm of storage capacity would be available on a yearly basis to alternative suppliers (41). Currently, the companies [Competitors]* and Energi E2 together buy approximately [75-125]* mcm of storage capacity (42).

(41) Access to storage facilities is obtained by negotiated third-party access ("TPA"). However, DONG Storage has published tariffs for two standard storage packages with respectively low and high injection and withdrawal capacities. By purchasing package 1 (low flexibility) the storage customer can inject and withdraw the reserved storage volume of minimum 200 and 100 days, respectively. By purchasing package 2 (high flexibility) the storage customer can inject and withdraw the reserved storage volume of minimum 40 and 20 days, respectively. In addition to the capacity payment for the storage package the customer pays a variable payment for the injected gas volume.

(42) The gas storages have three major purposes. First, the storages are used for emergency supplies. DONG Storage is under an obligation to provide storage capacity to the transmission company Energinet.dk. Second, the gas storages allow suppliers to accommodate the seasonal variation in the use of natural gas. The exploration of gas in the North Sea takes place continuously, while the end usage is highly dependent on seasonal fluctuations. Third, access to gas storage is necessary to gas suppliers since storage capacity allows a supplier to buy large quantities of gas when the price is favourable.

B. ELECTRICITY SECTOR

1. Legal and regulatory framework

(43) The Danish generation and wholesale market was gradually opened to competition from 1998 and is now fully liberalised (43). By 1 January 2003 the end-user market was also fully opened. Since then all customers have a free choice of electricity supplier in Denmark. The Danish Electricity Supply Act entered into force on 1 January 2000 and, including the latest amendments in June 2004, constitutes the regulatory basis of the electricity sector in Denmark. After liberalization, customers not wishing to choose freely among suppliers continue to be supplied with electricity from a universal service obligation ("USO") provider operating under a licence as a service provider.

(44) The Danish electricity sector is regulated by the Danish Ministry of Transport and Energy. The ministry takes decisions related to licenses. The electricity sector is overseen by the same independent regulator as the natural gas sector, DERA. The regulatory tasks of Directive 2003/54 and Regulation 1228-2003 are shared between DERA and the Danish Energy Authority ("DEA"), a directorate in the Ministry of Transport and Energy. The Danish Competition Authority ("DCA") applies the Competition Act to the electricity sector and monitors the competitive conditions of production and trading of electricity.

(45) Environmental policy objectives are achieved by requiring utilities to purchase a relatively high share of "green" electricity generated from renewable sources such as wind power and biomass and to pay a share of the costs involved. The CO2 emission of the generation sector is regulated through a CO2 quota system (tradable emission permits).

2. Infrastructure: Transmission and distribution

(46) Energinet.dk is the Danish state-owned transmission system operator (TSO). It owns the 400 kilovolt electricity transmission grid and also has the 132 and 150 kilovolt electricity grids at its disposal. Furthermore, it is co-owner of the electrical cable connections ("interconnectors") to Norway, Sweden and Germany. The day-to-day administration of the Nordic interconnectors has been delegated to the Nordic electricity exchange Nord Pool (jointly owned by the Nordic TSOs). Nord Pool allocates the interconnector capacity through implicit auctions whereby it is guaranteed that the physical flow of electricity always goes from the area with a high price to the area with a low price (44). The principle of regulated third party access has been chosen for the access to both the transmission and the distribution network within Denmark. The non-distance related postage stamp principle is applied for use of both the transmission and the distribution grid.

(47) The prices of the network services must be equitable, non-discriminatory and reasonable and must not contain terms restricting competition. The prices have to be published. DERA has the opportunity to intervene - ex-officio or on request - and require adjustments to ensure compliance with the electricity legislation. The tariffs of the transmission system operator are regulated by an ex ante approval procedure. (45)

(48) The structure of ownership in the Danish electricity distribution sector is quite fragmented. There are about 100 grid companies (owned, via co-operatives, by the consumers directly or by municipalities). Of the distribution companies, nine companies have a customer base exceeding 100,000. (46) Generation must be legally separated from transmission/distribution, and each activity has to be carried out in separate companies with different management and board representation.

VI. THE RELEVANT MARKETS

(49) The transaction affects the gas and electricity sectors. In both sectors various relevant product markets can be delineated.

PART A - NATURAL GAS

1. Relevant product markets

a) Market for gas storage and gas flexibility

(50) Gas demand varies both seasonally and within a day or week (peak and off-peak times). To balance supply and demand, companies active in gas markets need a considerable amount of flexibility. In addition, the transmission system operators need access to flexibility in order to balance imbalances of supply and demand and to provide for emergency situations.

(51) The most important flexibility tool is physical storage, which allows for storing gas when supplies exceed demand (e.g. during the summer or at the week-end) and to withdraw gas when demand exceeds supply (e.g. during the winter or at peak-times).

(52) In previous decisions the Commission found that gas storage constitutes a separate product market. (47) In the Total/GdF decision the Commission concluded that other flexibility tools, such as flexible supply contracts or sales to interruptible customers, cannot sufficiently fulfil the same functions as physical storage.

(53) However, in the notification, DONG has proposed a market for gas flexibility including physical storage, changes in production, international trade, line-pack, switches to other fuels, interruptible contracts and market hubs with forward markets. The notification focuses on physical storage while noting that there are significant constraints from other sources of flexibility, particularly changes in production and/or international trade. The notification acknowledges, however, that such sources of flexibility are often bundled with the supply of gas, so that a separate price cannot be calculated. A source of flexibility not explicitly mentioned in the notification is adhoc gas trading including ad hoc swaps. The notifying party's list does not further explicitly mention flexible consumption by central CHPs.

(54) The notified operation combines DONG's storage facilities with Elsam's and E2's flexible consumption. Therefore, for the purpose of defining the relevant product market in this decision it is necessary to examine the possible substitutability of these two flexibility tools and with other sources of flexibility. This must be done in the light of the different flexibility needs whereby a distinction can be made between: (i) seasonal flexibility; (ii) short-term flexibility; and (iii) emergency flexibility. While these situations reflect quite different customer needs, it is, however, not necessary to distinguish between these types of demand in terms of product market definition for the purpose of the current decision. This is because the two sources of flexibility are suitable for all these needs. It can be said, however, that it is uncontested by DONG that the other sources of flexibility mentioned in the notification are not all equally good alternatives for each of the three types of demand. (48) :

(55) The Commission considers that the flexibility tools mentioned above may be divided into 5 groups, namely:

(1) Storage in dedicated storage facilities;

(2) Interruptibility or other modulation of customers' demand, e.g. of central CHPs;

(3) Flexible supply contracts (irrespective of whether based on imports, domestic production or secondary domestic supply contracts);

(4) Flexible trading in gas on hubs or bilaterally (irrespective of whether forward looking or ad hoc);

(5) Line-pack, i.e. storage in transport pipelines by increasing or decreasing the gas pressure in those pipelines.

(56) The Commission's market investigation has clearly shown that line-pack in Energinet.dk's Danish on-shore transmission system is an insufficient flexibility tool even for short-term demand variations as soon as these exceed the duration of one day. (49) This is because prohibitive payments to the transmission system operator are due if the daily imbalances exceed a certain tolerable amount (Cf. recital (35)). (50) Energinet.dk's transmission system therefore does not constitute a relevant alternative to storage in physical storage facilities.

(57) There are two further types of line pack which are of potential relevance. First, it has been suggested in the market investigation that line pack in DONG's offshore pipelines (for which DONG obviously does not incur penalty payments) is an important additional flexibility tool for DONG (but not for its competitors). However, it does not seem necessary, for the purpose of this case, to separate this flexibility from the one DONG achieves with its upstream supply. Second, line pack in regional distribution grids, owned by DONG, HNG-MN and NGF, is an additional flexibility tool. In practice, this tool is of very limited importance in Denmark and can best be considered part of the demand flexibility of these customers.

(58) Regarding flexible trading at gas hubs or bilaterally, the market investigation shows that this is not a viable flexibility alternative for Danish (or for Swedish) customers as there is currently very little flexible trading in Danish hubs. In addition, Danish customers generally consider other hubs - such as Emden/Bunde - too remote for flexibility purposes. Whether domestic trading in the future could form part of a wider storage/flexibility market can remain open for the purpose of this decision.

(59) Regarding flexible supply contracts the market investigation has confirmed that they may be considered as fairly good substitutes for physical storage. However, in order to play that role, supply contracts must be available and offer a high degree of flexibility. It appears from the market investigation that not all customers have access to flexible supply contracts. It is also apparent that supply contracts frequently only allow for a limited daily upward flexibility, i.e. for limited extra off-takes in peak periods. It also appears, even from DONG's own submissions that flexible supply contracts cannot provide a full substitute for physical storage needs, in particular as regards the seasonal flexibility that is required by higher winter demand. (51) This was confirmed by other market participants. One such player remarked that flexible import contracts were "inferior to storage" (HNG-MN). Another stated that winter flexibility was not sufficient to satisfy the market demand (SGA-Naturgas Fyn). A further market player expressed some reservations concerning "the commercial terms and the available capacity" (Dansk Shell). These market players' scepticism was even greater regarding the availability and pricing of available domestic Danish wholesale contracts.

(60) Therefore whilst in principle flexible supply contracts constitute fairly good substitutes for physical storage, under current market conditions they are insufficient even for the best placed market player to fully cover its physical storage needs. They are even less readily available for other market participants. A price comparison between both forms of flexibility is impossible on the basis of current market conditions. The Commission's market investigation has therefore not confirmed that such flexible supply contracts form part of the same relevant market as physical storage. However, the question can remain open for the purpose of this decision as the concentration would on either basis lead to competition problems of a similar size and nature.

(61) With respect to modulation of customers' demand, the market investigation has shown that there are two entirely different needs that are addressed by this type of flexibility.

(i) Energinet.dk using modulation as an emergency tool. Customers subscribing to such interruptible contracts are remunerated by lower transmission tariffs payable to Energinet.dk. The existence of this service thus clearly indicates that there is a value attached (and a price paid) for flexibility services provided by customers. Beyond that, the market investigation has shown that this interruptibility is generally not used by Energinet.dk to cover its need for "normal"/non-emergency short-term or seasonal modulation, and that customers subscribing to this service are only interrupted in the rare cases of emergency. Therefore, this kind of emergency flexibility only has limited impact on commercial supply and its short-term and seasonal balancing needs.

(ii) The possibility of modulation of customers' demand in the interest of commercial suppliers. This possibility, in particular the modulation of demand by the largest Danish customers, the CHPs of Elsam and E2, is clearly seen as a distinct option for both short-term and seasonal balancing (52) by both suppliers and customers for which a price would be paid. (53) It is this flexibility which is therefore of much greater relevance for the purposes of this decision and which will be further considered in the following recitals.

(62) Regarding the question which customers other than central CHPs in Denmark are sufficiently flexible for the provision of such variations, it can be stated, first, that normal industrial customers appear to be rather limited in their ability to provide such services, in particular on a seasonal basis, as they have to sustain their industrial production. This is true even if some industrial customers may have retained dual fuel sources (e.g. oil and gas) for their energy needs and could therefore offer such service more easily with respect to gas. However, the market investigation has shown that such possibilities are rather limited. Second, decentralised CHPs' flexibility is only rarely dual fuel flexibility. (54) Rather, it is flexibility provided by e.g. hot water storage tanks or (in the opposite sense) by cooling towers, and is such (economically and, partly, technically) unsuitable for seasonal flexibility. Consequently, it cannot be used to any relevant degree as a substitute for physical storage.

(63) By contrast, by far the most relevant source of flexible demand available to satisfy flexible supply needs in Denmark, is flexible consumption by central CHPs. Central CHPs account for a very large part of overall consumption (approximately 20-25%) in Denmark and any demand shift is therefore significant in absolute terms of gas volumes. Central CHPs in Denmark can provide flexibility in three ways. Firstly, they can provide flexibility by means of output variations with corresponding variations in their gas consumption. Though local electricity demand is rather inelastic, output from central CHPs can be varied due to the flexibility of supply of other sources of electricity on Nord Pool. Secondly, they can provide flexibility by switching their fuel on any given production unit. Elsam and E2 currently own five blocks in four central CHPs, which can be fired by gas and at least one alternative fuel (coal, biomass, oil). Thirdly, flexibility can be obtained by an "indirect fuel switch" among power plants, namely by switching on/off gas-fired units and the simultaneous switching off/on of other production units.

(64) This potential of large gas-fired power plants as providers of "virtual storages" (55) does not appear so far to have been marketed as such by central CHPs for short-term flexibility, (56) whereas for seasonal flexibility there is evidence of counter-seasonal purchases of gas by E2, i.e. purchases of gas in summer and sales in winter. (57) However, both for short-term and seasonal flexibility, the modulation capability of central CHPs is clearly perceived to be an important factor by market participants. [Contains information on internal DONG documents]* (58). As regards short-term flexibility, it can be noted that DONG's Reply to the SO acknowledges that central CHPs are suitable to provide such services.

(65) The provision of flexibility services by central CHPs or (marginally) by other customers is thus an important actual and potential means of achieving flexibility, which is a close substitute for physical storage as it is very well suited to address both short-term and seasonal imbalances.

(66) Regarding both seasonal and short-term flexibility most market players consider that physical storage is the most appropriate tool and that the two closest substitutes are flexible consumption of central CHPs and flexible supply contracts (import contracts and domestic contracts). The market investigation has thus confirmed that flexible consumption by central CHPs constitutes a close substitute for storage in dedicated storage facilities, both for seasonal and short-term flexibility and emergency supply.

(67) However, a note of caution was struck by the Danish TSO, which stated with regard to short-term balancing (i.e. not with regard to provision of emergency supply as above) that "There is none other existing viable alternatives to DONG's two storages for the means of daily balancing/operation of the Danish transmission system - except that interruption of customers could to some extent work as an alternative." (59) This is in line with the fact that other market participants rated the storage in physical storage facilities (in Denmark) highest. In these circumstances it seems prudent not to discard the possibility of a narrow relevant product market for storage services in dedicated storage facilities.

(68) The Commission thus concludes that it is clear that for the different types of flexibility needs there are differences as to the adequate degree of substitutability of the different means of obtaining flexibility.

(69) As is explained further below, it is the Commission's assessment that the removal of E2 and Elsam - the only two customers with significant demand flexibility - is likely to lead to an increased dependence on DONG as provider of flexibility, inter alia, through likely increases in storage tariffs and through elimination of alternative sources of flexibility. This dependence will, in turn, diminish the likely competitive pressure exercised by any competitor to DONG as supplier of gas. This finding, including the establishment of the causal link between the transaction and its likely consequences, does not depend on whether the market is defined broadly as one of flexibility or more narrowly as one of storage only.

(70) Consequently, the exact product market definition of such a market (whether taken narrowly as a storage market or wider as a gas flexibility market comprising in particular (i) physical storage, (ii) flexibility of central CHPs and (iii) flexible supply contracts) can remain open.

b) Markets for wholesale gas supplies for Denmark and for Sweden

- Market for wholesale gas supplies for Denmark

(71) The Commission's market investigation has provided strong indications of the existence of a separate market for wholesale supplies of natural gas. In its response to the decision adopted pursuant to Article 6(1) (c) of the Merger Regulation, DONG recognised "the existence of gas sales and trading at the wholesale level" and has not contested this or the below product market definition, given below, relating to the wholesale market in their reply to the Statement of Objections. The term wholesale is, for the purpose of this decision, in general and with the exception of transactions as discussed below with regard to E2 (see recital (82)), understood as covering transactions between traders/resellers and not between a trader/supplier and an end customer.

(72) More particularly, the market investigation has shown that there is - on the demand side - a need for gas supply at the wholesale level in Denmark. That need arises in the Danish onshore area and cannot, according to the results of the market investigation, be satisfied with wholesale sales taking place elsewhere with a view to other geographic areas. The market investigation has also shown that, on the supply side, there is an offer of such wholesale gas in or for the Danish onshore area and by consequence that there are commercial transactions in which wholesale gas is supplied and purchased in or for this area. Consequently, there exists a market for wholesale gas supplies for Denmark.

(73) Supply side substitution from wholesale sales for other geographic areas may be taken into account when defining the geographic scope of this market and will be discussed below in the section on geographic market definition.

(74) It is notable that the wholesale of gas for Danish consumption has a number of structural characteristics such as the sole reliance, as regards physical wholesale gas, on gas from the Danish off-shore fields.

(75) In Denmark, this market comprises at least the following categories of transactions:

(1) sales of physically transported gas (i.e. of gas "imported" (60) from the Danish offshore fields) net of quantities exported to Germany or Sweden;

(2) sales of contractually imported/obtained gas (as opposed to physically imported gas, this gas is only contractually imported whereas the corresponding physical gas never leaves Denmark; this is because physical imports are only possible in the flow direction, which is from the Danish North Sea offshore fields to Denmark and on from Denmark southwards to Germany and eastwards to Sweden);

(3) sales of gas between traders within Denmark (e.g. through secondary wholesale spot exchanges at the GTF).

(76) DONG is the only company accessing physical gas in Denmark. It has or had swap agreements with [competitor]* and [competitor]*, captive sales to its two distribution companies and sales to the regional supplier and distribution company [...]* (61) (ca. [...]* mcm) to the two generators E2 and Elsam, to other independent decentral CHPs, and to industrial customers.

(77) [...]*, the swap partners of DONG, are selling gas on the Danish wholesale market to their affiliated companies; [...]*.

(78) [...]* (62) [...]*. Its retail subsidiary E.ON Sverige (formerly Sydkraft) is operating predominantly on the Swedish and (to some extent) on the Danish market. [...]*.

(79) [...]*. There are as yet only low volumes of genuine trading on the GTF [...]* (63).

(80) Contractually, wholesale gas for Denmark can be imported/obtained in three ways: (i) through a bilateral swap arrangement with a party having access to physical gas in Denmark (the beneficiary thereby obtains the title to that physical gas in Denmark and gives the title to physical gas somewhere else) or alternatively through a bilateral or unilateral purchase of gas; the only primary source for this gas being DONG; (ii) through so-called "turn-around" of gas that is to be physically exported with delivery at the Danish exit point at Ellund. The beneficiary of this gas can choose, subject to observation of the Danish TSO's entry rules (such as booking of entry capacity and payment of entry fees) to contractually re-import this gas (which physically never leaves the Danish onshore grid) into Denmark; (iii) through purchase of rights to gas at a point outside the Danish grid (typically at Emden); and arranging with all TSOs on the way from the point of purchase to the Danish border (Ellund) the transfer of that right to gas to the Danish exit/entry point (again, transfer "north" against the flow direction can only constitute a contractual import, which nevertheless must be paid for to each TSO along the way (64) unless swaps can be arranged), from where the same entry mechanism applies as the one described in option (ii). These transactions are analyzed further in the relevant geographic market section below.

(81) Customers of wholesale gas in Denmark can be traders without any (retail) supply obligations (who would sell on the gas to other traders) (65) or traders with supply obligations (e.g. regional distribution/trading companies - RDCs).

(82) [Contains information on E2's position as a consumer and E2's access to the wholesle market]*. Large power plants (in Denmark equivalent to central CHPs (66)) are the largest gas consumers in the country. Their operators, E2 and Elsam, purchase some [20-30%]* of Danish gas consumption. E2 has, due to market liberalisation, been able to purchase gas under conditions bearing all the characteristics of wholesale sales, with the exception that it has so far not (or only to a very minor extent) resold that gas, but used it for its own 'captive' consumption. Purchases by central CHPs from traders or from abroad can be considered to fall within the scope of the gas wholesale market if they take over more functions or risks than a regular supply customer. Such additional functions may include the responsibility for balancing, storage and transport of the acquired gas volumes. By contrast, if the supply contract provides for gas delivery at the gate of the power plant without the customer's intention of reselling such gas, the central CHP does not usually behave like a trader on the wholesale market but rather like a mere retail customer. Such purchases would thus be part of the (retail) market for supply of gas to large CHPs (see further discussion of this issue in the section on supply to central CHPs below).

(83) For the purposes of this case and in the light of the above, the market for wholesale supplies of gas for Denmark therefore comprises, * all sales - irrespective of whether via the GTF, supply contracts or other agreements - ??by physical or contractual importers67, re-importers (in case of turn-around gas), producers (if applicable in the future) and traders ??to other traders (such as supply companies) or to central CHPs (the latter only to the extent that they take over at least some of the services regularly provided by a supplier for delivery at the site or intend to resell the gas), ??which satisfy the needs of these customers to have access to wholesale gas in Denmark.

*Markets for wholesale gas supplies for Germany and for Sweden

(84) It is obvious that such a need for wholesale gas supplies also exists in neighbouring areas such as Northern Germany and Sweden.

(85) Apart from Denmark, the notified operation also has an important impact on Sweden. The Swedish wholesale supply situation is different from the one in Denmark. This is partly due to the later and hitherto much smaller development of the Swedish natural gas market. Unlike Denmark, Sweden does not have any domestic gas production. Sweden's only route of physical gas import runs through Denmark and enters the Swedish grid at Dragør. None of the Swedish wholesale gas customers are identical to Danish wholesale gas customers, which shows that the structure of demand is very different in Sweden in terms of market players. This is also true in terms of wholesale customers' supply needs which are determined by the supply needs of their own customers. The demand of industrial end customers has a higher relative importance in Sweden, while the demand by central power plants has so far been very small. Also regulatory conditions and supervision in Sweden are different. For these reasons, this decision will consider the demand for wholesale gas in Sweden separately but only with regard to it being an ancillary market without competitive constraints for the reasons outlined above. The market for wholesale gas supplies for Sweden thereby comprises all imports and wholesales (i.e. sales to on-sellers) of natural gas with a view to consumption in Sweden.

(86) The operation has a much lesser impact on competition in Germany. It is therefore not necessary to examine German demand for wholesales except to the extent that this influences the demand and supply situation in Denmark or Sweden. (68)

(87) Whether wholesale sales for each of these areas (Denmark, Sweden, Germany), from the supply side perspective, can sufficiently by substituted by wholesale sales in other areas will be discussed in the section on geographic market definition below.

c) Markets for the supply of natural gas

(88) Concerning the supply of natural gas, in addition to considering and advocating the possibility of an "all gas supply market" (69), the notifying party has proposed the following market segmentation:

- supply of natural gas to power producers operating centralised CHP plants;

- supply of natural gas to other business customers and decentralised CHP plants;

- supply of natural gas to household customers.

d) Market for supply of natural gas to central CHP plants

The supply of natural gas to central CHP plants constitutes a market separate from wholesale supply of natural gas and separate from other natural gas supply markets, in particular from supply to decentral CHP plants.

Supply to central CHPs is distinct from the wholesale market, but with low mutual entry barriers

(89) Whereas participants in the wholesale market need to be traders/shippers, this is not necessary (but not excluded either) for central CHP plant customers. To the extent that central CHP plant customers may choose to be supplied "at the gate" of their power plant, with a gas contract that includes all deliveries and all their service needs (e.g. with regard to storage, balancing and flexibility) these customers do not constitute traders/shippers nor wholesale (as opposed to retail) customers but simply very large and sophisticated retail customers.

(90) This is notwithstanding the fact that, as argued above, central CHP plant customers have a very strong commercial interest in commencing activities in the wholesale market thereby actually becoming wholesale sellers of surplus quantities and wholesale customers (and at the same time retail suppliers "supplying themselves" through management of a supply portfolio). In line with the arguments set out in the sections on the wholesale market, it also can be concluded that, in the process of the opening of natural gas markets, all retail central CHP customers may at some point expand their activities into the wholesale market (at least as wholesale purchasers), at which point a separate market for the retail supply to central CHPs would disappear. However, this decision also needs to consider the possibility that such a separate market may continue to exist in Denmark and/or Sweden in the foreseeable future.

(91) The different approaches that can, at least currently, be taken by central CHPs are reflected in the distinct commercial approaches taken by Elsam and Energi E2. Whereas Elsam has so far been supplied only through a single supply contract (with DONG) and beyond that only has some limited exchanges with E2 - which may be seen as wholesale activities - Energi E2 has had a whole range of supply contracts and has even acted as an importer of natural gas from Germany, thereby being to a much higher degree active on the wholesale market. (See also the discussion of E2's and Elsam's role on the wholesale market above).

(92) In general, the fact that an undertaking, perhaps even without encountering many difficulties, may choose to become active in a vertically related upstream activity (such as purchasing, and selling, on a wholesale market) is in itself not an argument against the existence of a separate downstream market. However, it can in such circumstances be difficult to attribute the sales to one or the other market. In the case of the markets at stake, in order to do so, it is necessary to consider that the added value between wholesale purchase and retail delivery consists of a number of elements, namely the: (i) management of a supply portfolio; (ii) booking and management of storage capacity and balancing requirements; and (iii) booking of transport "to the power plant gate" in order to meet a power plant's exact demand on time, including, for each of (i)-(iii) the management of the associated risks. The clearest distinctive criterion for attributing a particular transaction is "at the gate delivery", as this presupposes no involvement on the part of the customer in other activities linked to wholesale. For the purpose of this decision, only "at the gate deliveries" will qualify as (retail) supply to central CHP power producers. (This is also consistent with the delineations of the other relevant retail markets, cf. below.)

(93) In addition, the existence of entry barriers between both markets needs to be considered. As discussed above, a central CHP power producer (viewed as a sophisticated large retail customer) can enter the wholesale market as a buyer or a seller.

(94) Similarly, entry barriers for wholesalers wanting to supply some segments of retail supply are surmountable. There may, for example, be a situation in which a 1-month short-term contract is concluded between a wholesale supplier and a power producer with delivery at the Danish entry point in Ellund. If the power producer (acting as a wholesale purchaser taking over retail supply to its plant) subsequently merely books transmission capacity to its power plant with a firm intention to burn all the gas that is delivered to Ellund instantly, it is obvious that for this type of need also the wholesale supplier itself could easily (in the absence of capacity booking obstacles) enter this "segment" of the demand as a supplier itself by providing the extra service of booking the transmission capacity up to the power plant's gate itself. The entry barrier from the wholesale market to this type of delivery (which can be viewed as a "segment" of the market) is therefore, rather low.

(95) However, there are other kinds of deliveries at the wholesale level where such retail entry is more complex and requires higher long-term commitment and commercial risk between wholesale and retail level. This will for example be the case whenever a supply contract to central CHPs gives those CHPs a high degree of flexibility in respect of their demand. In an extreme case the customer will only need to inform its supplier the day before delivery whether it needs 100% or 0% of its maximum capacity. Entry barriers into such kind of delivery are obviously high.

(96) An important distinction between a pure wholesale market and the retail market for supplies to central CHPs is also (at least as regards Denmark) that gas sales to central CHPs are closely linked to coal prices [...]* as opposed to prices on the general wholesale market which appear to be closely tied to the development of oil prices. (70)

(97) In the Commission's view therefore, at the current stage of development of the Danish natural gas sector, it is appropriate not to regard wholesale supply and retail supply to power producers as constituting one single product market. However, purchases and sales made by owners of central CHPs can take place both on the wholesale market and on the retail market.

(98) However, as will be shown in the analysis of the competitive assessment of the notified operation below, even if supply to CHPs and wholesale supply were considered to constitute just one joint relevant market, the competition problems created by the proposed operation would not change.

Supply to central CHPs is distinct from other retail supply markets

- Different product markets for supply to central and decentralised CHPs

(99) The notifying party argues that a distinction must be drawn between gas supply to larger centralised CHP plants and supply to smaller decentralised CHP plants. The results of the market investigation support the view that centralised and decentralised CHP plants are not part of the same relevant product market. The Commission therefore agrees with the approach proposed by the notifying party.

(100) In Denmark, there are currently no power plants that use natural gas for electricity production only. Instead, natural-gas-fired "central" or "decentralised" CHP plants produce both electricity and heat. However, there are important differences between both groups of CHPs.

* Decentralised power plants have only a limited degree of flexibility between heat- and electricity production as they in general cannot "switch off" heat production while producing electricity. Central power plants, by contrast, are considerably more flexible in regard to their electricity/heat production ratio as they can switch off heat production while producing electricity. This type of production mode is called condensing mode.

* Central CHPs, individually, consume much larger amounts of gas. Whereas the annual consumption of central CHPs can go far beyond a value of 100 million cubic metres ("mcm"), the consumption of the largest decentral CHPs will always remain, usually far, below 100 mcm. These consumption differences are linked to capacity differences. The capacity of central CHPs is generally far above 100 MW, whereas the capacity of decentralised CHPs is clearly below 100 MW.

* Due to their different heat and electricity production modes, the seasonal consumption patterns of central and decentralised CHPs also vary significantly. Decentralised CHPs' demand is mainly temperature driven (due to their district heating output as their primary function). They therefore, consume gas mainly in the winter months. By contrast, temperature is only one of the factors determining the gas use of central CHPs because the primary function of central CHPs is electricity production, which means that these plants will tend to use natural gas for electricity production whenever they can achieve a sufficient margin for their electricity output, whatever the season.

* Decentralised CHPs have traditionally been under strict regulation, inter alia, in guaranteeing their output prices. This regulation is currently being rolled back, and more and more decentralised CHPs will have to sell their production on market terms. It is clear, however, that decentralised CHPs are not yet fully used to (or, for some - especially the smaller ones - not even interested in) operating on the market. This is also supported by some of the answers to the market investigation. [Contains information on decentralised CHPs' relations vis-à-vis DONG]*.

* The owners of central CHPs have the possibility and incentive (due to their large gas consumption and flexibility thereof) to participate directly in the natural gas wholesale market. Evidence for this is that they can have an interest in booking storage capacity, in importing natural gas on their own account and in managing a supply portfolio composed of various short and long-term contracts. [...]*. Decentral CHPs do not have this possibility or incentive, which would expose them to excessive commercial risks and costs which would outweigh the potential benefits.

* These differences are also expressed in the very different types of contracts presently available. Decentral CHPs usually have one or two-year contracts indexed to oil products. Danish central CHPs, by contrast, have traditionally had the bulk of their gas needs covered by multi-annual basic contracts, indexed to coal prices and, possibly, a number of shorter-term supplementary contracts of varying durations that can be as short as one month, that account for demand peaks based on shifts in electricity prices.

* It is clear, therefore, that prices and price developments between central and decentralised CHPs are also very different. (71)

(101) For these reasons, the Commission regards gas supplies to central CHPs as constituting a market distinct from gas supplies to decentral CHPs.

- Different product markets for supply to central CHPs and other retail supply

(102) The distinction between gas supply to central CHPs and gas supply to other customer groups is even more obvious. The Commission agrees with the notifying party that these belong to separate product markets.

* Demand patterns are very different. Large industrial customers have a rather flat and stable demand which is subject to little seasonal variation, except as far as holiday seasons are concerned. Small business customers (usually commercial customers), like household customers, have a demand that mostly follows heating requirements.

* Prices are very different as well. In particular, when taking storage requirements into account (which is much smaller with industrial customers), the Commission's market investigation has clearly shown that prices paid by central power plants are significantly below prices paid by large industrial customers. Prices paid by small business customers and households are obviously even higher. (72)

* Also the above statements on different contract types, price indexation clauses and lack of incentive and ability to access wholesale sources of supply directly are fully applicable to these other customer groups. For example, the Commission's market investigation has not found evidence that any Danish industrial customer has booked storage capacity on its own account.

(103) The Commission notes, that in its reply to the SO, the notifying party did not question the existence of a separate market for supply of natural gas to central CHPs.

(104) Therefore, for the purposes of this decision, the Commission considers that a separate relevant market for supply of natural gas to central CHPs exists.

e) Market(s) for supply of gas to decentral CHPs and to large industrial customers

The supply of natural gas to decentral CHPs and the supply of natural gas to large industrial customers constitute one or two different product markets.

(105) According to DONG, (73) supply to decentralised CHP plants forms part of the same relevant product market as supplies to "other industrial customers".

(106) On the basis of the results of its market investigation (74) the Commission agrees only insofar as decentral CHPs and large industrial customers may or may not form part of the same relevant product market, being either two quite distinct segments of the same market or two separate markets. The Commission disagrees that small business customers belong to such a market.

Small business customers do not belong to the same product market as large business customers and decentralised CHPs

(107) Small business customers do not belong to the same product market as large business customers and decentralised CHPs. This is for the following reasons:

* Prices are very different. (75)

* Margins (76) and margin developments are very different.

* Marketing and distribution channels and costs are different. Whereas large industrial customers benefit from key account management this is not the case with small businesses. [...]* (77)

* Storage costs and prices are different. (78)

* Market structure is different. [Contains observations on competitors' market positions]* (79).

* Switching rates are substantially different as is price sensitivity of customers, both of which are low to medium with small business customers and higher with large industrial customers. (80)

* Metering requirements only exist for customers with an annual consumption above 0.3 mcm. The customer groups below, households and small business customers do not have such requirements.

* [Contains a reference to DONG's internal papers]* (81)

* [Contains observations on one of DONG's agreements]*.

Supply to large industrial customers and decentralised CHPs has a number of similarities

(108) A number of elements can be stated pointing towards similarities between large industrial customers and decentralised CHPs:

* The annual gas consumption of decentralised CHP plants is similar to that of larger industrial customers. In both customer groups the annual demand of the largest companies is between 50 and 100 mcm.

* Gas contracts of both customer groups appear to be similar, both groups being supplied on the basis of "large customer contracts", usually with oil price indexation clauses.

* Some suppliers, in line with DONG, have also stated that they view these two groups as one relevant customer group for their marketing purposes.

* In the Commission's market investigation one group of customers, comprising both decentralised CHPs and industrial customers ("Gasgruppen") presented a common point of view and did not distinguish between different needs for both groups. "Gasgruppen" has on several occasions pooled its consumption of gas and put it out to tender, which also points in the direction of limited differences in consumption and demand patterns at least between these decentralised CHPs and industrial customers.

But supply to large industrial customers and supply to decentralised CHPs are also characterised by important differences

(109) The demand of industrial customers is much more regular and less exposed to seasonal variations than the demand of decentralised CHPs. This results in significantly higher load factors and lower storage costs and supply flexibility requirements for industrial customers whose consumption is much more foreseeable. While the gas demand of decentralised CHPs is essentially driven by the heating requirements of Danish households and commerce, the demand of industrial customers is driven by the needs and rhythm of industrial production. (82)

(110) Ownership tends to be different. Whereas industrial customers are privately owned, decentral CHPs are often municipality owned or are organised as cooperatives owned by their customers. Together with a different primary use of natural gas (for industrial processes rather than for heating purposes) this leads to differences in elasticity of demand. Whereas municipal/cooperative heating CHPs may have low heat prices for their customers as their primary objective while positive margins are only of secondary importance, the benchmarking of industrial customers is done vis-à- vis the energy costs of their competitors, which opens possibilities for price discrimination between these customer groups.

(111) The Commission's market investigation has yielded evidence of some differences in future demand development. Whereas no or very little positive growth is expected for large industrial customers, negative growth is expected for decentralised power plants. This expected negative growth can be explained by the deregulation referred to above, which makes gas-based electricity production less competitive for operators that have to source their gas supply on the market.

(112) Customer loyalty and switching rates appear to be different between industrial customers and decentral CHPs: somewhat higher loyalty is seen with decentralised CHPs. (83)

(113) [...]* (84) It must be noted that already prior to the proposed concentration DONG had a specific strategic interest in decentral CHPs which also produce electricity. (85) [...]*.

Conclusion

(114) For the reasons set out above and for the purposes of this decision, the Commission concludes that there exists a separate market for supply of gas to decentral CHPs and large industrial customers. The question whether these two types of customers form a joint relevant product market, as claimed by the notifying party, or two separate markets, can, however, remain open as the assessment for both groups leads to the same result.

f) Market(s) for supplies of gas to households and small businesses

(115) The notifying party submits that there exist separate retail markets for the sale of natural gas to businesses and for sales to households respectively. The notifying party observes that supplies to households are subject to different competitive conditions than supplies to businesses. Business customers to a large degree demand individually adapted products. Prices and other terms are negotiated commercially.

(116) The Commission agrees with the notifying party that supplies to household customers do not belong to the same product market as supplies to central CHPs, to decentralised CHPs and to large business customers with an annual consumption above 300,000 m3. (86) Competitive conditions applying to sales to household customers are essentially different from the ones applying to sales to large business customers and CHPs.

(117) As outlined above, small business customers do not belong to the same product market as large business customers or decentral CHPs. The question arises therefore, whether sales to small business customers and household customers belong to the same or to different product markets.

(118) On the one hand, demand and supply conditions between these two types of customers in Denmark show many similarities.?

* Neither of these customer groups has individual metering requirements; both groups of customers are therefore "non-metered" customers. (87)

* Non-metered customers have different contract types as compared to metered customers. In the contracts for supplies to non-metered customers, yearly consumption is based on an assumed load profile, whereas flexibility provisions are directly included in the contracts for metered customers. Furthermore the non-metered customers conclude contracts for an indefinite time period which may be terminated at a short notice, while the contracts of metered customers usually run for 1-2 years.

* Gas prices are higher for non-metered customers and they are published, whereas the prices for metered customers are negotiated.

* Non-metered customers have a much lower propensity to switch supplier than metered customers. Figures received from the Danish Competition Authority (88) suggest that while approximately 30 % of metered customers have switched supplier during 2004, only 0.24 % of non-metered customers have switched.

* [...]* (89).

* From a supplier's perspective, the large number of locations and customers requires customer portfolio management tools.

* Suppliers develop standardized general offers for these customers.

* Brand image plays a more important role for both small industrial and commercial customers and household customers.

* The market structure for both Danish customer groups is similar: the main suppliers to both groups are the remaining three regional distribution companies, DONG (having purchased two such companies), HNG/MN (two formerly independent regional distribution companies having merged most of their activities) or Naturgas Fyn.

* Based upon DONG's figures, average storage costs and average storage prices for both customer groups are the same. This suggests similar consumption patterns (heating- and cooking-driven).

* Entry barriers to both groups are very high.

(119) On the other hand, demand and supply conditions between these two types of customers in Denmark are also characterised by a number of relevant differences.

* Average consumption is different. Households only consume on average 2000 m3 per year whereas the average consumption of small business customers is a multiple thereof.

* Prices, costs and margins are different.

<emplacement tableau>

??DONG's marketing strategies to small business customers and to households differ. For example, DONG uses a number of sales agents [...]* for the small business customer segment.

??At least one competitor (Shell) is present, albeit to a very marginal degree, in sales to small business customers but not to household customers.

(120) In the reply to the SO, the parties maintain that the relevant market comprises supply to households only, and that small business customers should be included in a market for all business customers including decentral CHPs.

(121) The Commission agrees that there could excist a separate market for supply of gas to households. However, as will be shown in the competitive assessment below, this assessment does not differ significantly between these two customer groups. The question of whether or not supplies to households and supplies to small business customers form different segments of the same relevant market or two different relevant markets can therefore be left open for the purpose of this decision.

(122) For the purpose of this decision, the relevant markets defined are therefore the following ones:

(6) the market for storage or, alternatively, the market for flexibility of natural gas;

(7) the markets for wholesale of natural gas (a) for Denmark and (b) for Sweden;

(8) the market for the supply of natural gas to central CHPs;

(9) the market or markets for the supply of natural gas to large industrial customers and to decentralized CHPs;

(10) the market or markets for the supply of natural gas to small business customers and to households.

2. Relevant geographic markets

a) Market for gas storage or gas flexibility

(123) In the notification DONG submitted that the relevant geographic market is wider than Denmark and also encompasses Sweden, northern Germany and the Netherlands.

(124) However, the market investigation has provided evidence that the geographic market for storage (or for flexibility services), irrespective of its precise product market definition, is confined to Denmark. It appears that most flexibility tools are only available on a national level. Cross-border flexibility by means of contracts seems to be extremely difficult. The only theoretical option of cross-border flexibility may thus be physical storage abroad.

(125) In Sweden there is only very limited physical storage capacity. (91) "Virtual storage" capacity, i.e. flexibility capacity by large gas-fired power plants, currently only exists to a minor degree. (92) However, one new plant is coming on-stream in 2007 (Göteborg Energi) and a further one in 2009 [...]*, both of which could - in principle - also provide flexibility services like those of Elsam's and E2's power plants. As regards [...]* planned central CHP, it should be noted that the timing of this project is uncertain and construction has not yet started. Currently, therefore, Sweden depends almost entirely on Danish storage/flexibility capacity. The extent to which this will change in the foreseeable future cannot be stated with certainty. However, in any event it is rather unlikely that this "virtual storage" capacity in Sweden will be entirely sufficient to meet Swedish storage needs. Therefore, it is even more unlikely that this new "virtual storage" capacity can constrain a monopolist in flexibility/storage services in Denmark.

(126) In Germany, storage tariffs for small injection and withdrawal capacity amount to 242% of the Danish tariffs, and for large injection and withdrawal capacity German tariffs are still 56% higher than Danish tariffs. (93) In addition, any physical transport from Denmark to the storage in Germany would cause additional transport costs, including entry and exit fees in the different transmission networks. A further difficulty of storage abroad results from the complex synchronisation of the reserved storage and transport capacities with the actual gas needs. Capacity reservations providing for the necessary flexibility would be difficult to obtain (94) and even if they were available, they would be very expensive.

(127) The Commission notes that in previous Commission decisions, (95) it was found that, for technical reasons, the economic radius for pore storage is less than 200 km, and less than 50 km for cavern storage. However, the closest German cavern storage is located at Kiel, thus more than 50 km from the Danish border. Similarly, the closest pore storage is located south of Hamburg and thus approximately 200 km from the Danish border. This could indicate that physical storage could also encounter technical difficulties at least in emergency situations.

(128) Respondents to the Commission's market investigation generally considered it more efficient to store near the area where the gas is used and expressed considerable doubts as to the possibility of storage abroad.

(129) DONG's own behaviour also runs counter to the existence of a combined Danish and Northern German flexibility/storage market: for its activities in (Northern) Germany and the Netherlands DONG books storage capacities in these countries and [...]* (96).

(130) DONG has submitted that [Competitors]* have their own storage capacities in Germany and one should thus not compare Danish tariffs with German tariffs but rather with the internal costs of these companies. In addition, DONG argues that the availability of such alternatives for some customers would put a downward pressure also on prices to other customers due to the non-discrimination obligation imposed on DONG's prices.

(131) The Commission notes in this regard, that available alternatives, even if they were available to some customers to DONG Storage, are unlikely to put a downward pressure on its prices generally. Firstly, because none of the three customers mentioned are significant customers today. Secondly, because DONG does not have an incentive to attract storage customers, because that would lead to increased competitive pressure on DONG's other gas activities in Denmark. This is the very reason for the necessity of regulating access to storage.

(132) DONG also submits that [Competitors]* already exercise significant pressure due to the fact that they have significant volumes available at Ellund. Instead of using Danish storage facilities, a relatively high volume could be sent to Germany during the summer (and put in their own storages), while higher volumes could then be turned around during the winter.

(133) The Commission also notes that this opportunity is only available for the three firms receiving DONG gas at Ellund. From this it follows that even if the behaviour mentioned by DONG were to be taken into account, it would not lead to the conclusion that all of the storage capacity in Northern Germany were to be included in the relevant market, but only that some of the flexibility needs of [Competitors]* in relation to the gas supplied by DONG need not be provided by DONG Storage.

(134) In order to assess the actual degree of flexibility obtained by seasonal variation in turn-arounds, the Commission has verified import figures into Denmark in the summer of 2004, winter of 2004/2005 and again the summer of 2005. As illustrated in Figure 1 below, these figures do not indicate any relevant increase in import volumes into Denmark in the winter of 2004/2005, which indicate that such (re-)import volumes currently only have a very limited, if any, impact on the provision of flexibility/storage services in Denmark.

<emplacement tableau>

(135) The Commission has also assessed the seasonal variation in the net exports to Germany. As can be seen from the general trend of exports from the Danish Monthly Natural Gas Statistics outlined in the Figure 2 below. The pattern shows that exports to Germany are higher in the winter months than during the rest of the year. (97)

<emplacement tableau>

(136) On the basis of a clear pattern of seasonality in the net exports combined with a lack of similar seasonality in turnaround, the Commission concludes that neither the flexibility provisions in [Competitors]* respective long-term supply agreements with DONG, nor the use of German storages are currently being used for the Danish market. Usage of this flexibility for the purpose of the Danish market would thus result in less seasonal flexibility available in Germany.

(137) In addition, the fact that both [...]* and [...]* have reserved and used storage capacities in Denmark indicates that turn-around flexibility is not a viable commercial option for these companies (99), at least not for their entire flexibility needs in Denmark, and as far as [...]* is concerned also in Sweden (100). The Commission agrees with DONG that the purchases of storage by [...]* and [...]* currently are of limited volumes. This fact combined with the lack of seasonality in turn-around volumes (101) as shown above, indicates that these firms currently have limited their Danish activities to customers with limited seasonal flexibility needs.

(138) DONG has argued that the evidence is also consistent with the contention that the mere threat of additional imports exerts sufficient competitive pressure. In this context, in its reply to the SO, DONG refers to the fact that DONG's physical storage facilities equals 760 mcm while the total volume exported is [...]*.

(139) As is explained further below, DONG's behaviour as provider of storage is guided by other considerations than implicit threats of alternatives by customers. This is due to the fact that these customers are mainly seen as competitors to DONG.

(140) Furthermore, the Commission notes that for the purpose of assessing flexibility, it is less relevant to compare the volumes of flow in the DEUDAN pipeline, but rather the injection and withdrawal capacities. In this context it should be noted that while the maximum total entry capacity at Ellund is 200.000 m3/hours the storages allow an hourly withdrawal capacity of 950,000 m3/hour (102). The Commission notes in this regard that DONG has stated that though there has been some spare withdrawal capacity available for sale in recent years, the withdrawal capacity is occasionally fully used on very cold days. (103)

(141) This, combined with the additional costs relating to booking seasonal re-entry capacity, means that competitive pressure from turnarounds, even if they were to increase in the future, is unlikely to significantly constrain the DONG's storage activities, nor remove the need for such storage.

(142) It is also convenient to draw attention to the statement by the Danish TSO that "[i]t is not possible to source flexibility services needed by Energinet.dk outside Denmark for normal daily operation" (104).

(143) In conclusion, the physical storage in Germany cannot be considered as part of the relevant geographic market. The same applies, a fortiori, to the Netherlands, which at least for high withdrawal capacity has even higher storage tariffs than Germany (105) and where transport costs would exceed those to Germany. The degree of substitution available to [...]* and [...]* at Ellund appears not to have been relevant to any significant degree, and is unlikely to become so in the future to an extent to which it would bring into question the finding of a Danish market for storage (or, alternatively, the market for flexibility).

(144) The Commission also concludes that, since the only entry point into Sweden is via Dragør, the market for storage (or the market for flexibility) in Sweden is either Swedish in scope or, due to a certain dependency of Sweden on the provision of Danish storage facilities, Danish-Swedish in scope.

b) Market for wholesale gas supplies for Denmark

- Overview

(145) In its response to the decision adopted pursuant to Article 6(1)(c) of the Merger Regulation as well as in its reply to the SO, the notifying party submitted that the relevant geographic market was wider than Denmark (or Denmark and Sweden)106 and included at least Germany. (107)

(146) The Commission does not share this view.

(147) The market investigation provided strong indications that the market for wholesale supplies of gas for Denmark is confined to Denmark. Due to its specific situation, the German side of the Danish entry/exit point at Ellund may or may not be included in the geographic market (to the extent that this gas is actually used for wholesale supplies for Denmark). Swaps with delivery points in Denmark [...]* part of the geographic market.

(148) Regarding Sweden, no physical or contractual delivery has ever been made from Sweden to Denmark which makes any consideration of an inclusion of Sweden into the geographic market rather hypothetical as far as wholesale sales for Denmark are concerned. This may be different if the effect of an operation for Swedish wholesale customers were to be considered (108). The current situation, which is unlikely to change in the foreseeable future, is therefore at most one of asymmetric constraint (Danish transactions might constrain Swedish ones but not the other way round). (109)

- Germany, the Benelux and the United Kingdom are not part of the relevant geographic market

(149) As indicated, the notifying party submits that Germany is part of the geographic market. The Commission's analysis of the arguments put forward by DONG, however, has not confirmed such a large scope of the relevant geographic market.

(150) First, physically, all gas consumed in Denmark is Danish gas. Denmark only has gas production on the Danish Continental Shelf which is connected to the Danish west shore via two pipelines, and a pipeline southwards connecting to the NOGAT system for onwards connection to the Dutch market. Onshore the system is connected to Germany and to Sweden but the exclusive flow direction, set to continue unchanged, (110) is towards Germany and Sweden.

(151) In its reply to the SO, disputes the relevance of the physical origin of gas. In its view, what is relevant is whether price rises would result in gas that otherwise would have been sold in Germany being sold in Denmark. Irrespective of what economic comparison to apply to the valuation of the gas stemming from Denmark (on this see also further below), it is clear that (i) contractual gas sales do not have the same security (and often not the same flexibility) (111) of supply as physical gas, (ii) there are constraints as to the possible shifting of physical gas from Denmark to other markets, not least through capacity constraints of export pipelines,112 whereas swaps, a means to overcome physical capacity constraints, cannot exert any competitive constraint on DONG as they need the consent of DONG.

Figure 3:German and Danish import/upstream price comparison

Comparison of the development of Danish (DONG) import prices with German import prices

[Confidential chart bearing the following timeline: monthly, form January 2003 to July 2005]*

(152) Second, in 2003-2005, all (commercial) imports (also including turnaround gas at Ellund) merely represented a very small amount (< 12%) of the total Danish consumption. (113)

(153) In its reply to the SO DONG claims that the Commission's approach is based on an insufficient application of the SSNIP (114) principle. Focusing on actual volumes of commercial imports can only provide a partial picture and is not conclusive of what would happen in the event of a SSNIP price increase.

(154) In the Commission's view, however, [Contains considerations on prices in Germany and Denmark and on the SSNIP test]*.

(155) Further, DONG argues that the presence of commercial imports demonstrates the feasibility thereof. With price rises further imports will come, especially since gas is relatively available at German gas hubs, referring to the Commission's Issue Paper on the Energy Sector Inquiry.

The Commission, while not discounting the existence of contractual imports, maintains that they do not exert sufficient competitive constraint. Furthermore the Energy Sector Inquiry Issues Paper of the Commission quoted by DONG underlines that hub trading (among others at Emden/Eurohub) by incumbents is very low115. Furthermore, it states that swapping only constitutes 5% of total supply volume, and the findings stated that Germany is the only country where there is significant volume swapped between network points does not justify the conclusion that gas is readily available as the swapped volume is low in a relative sense and could be due to the multiple TSO areas in place in the German system. Consequently, contrary to DONG's analysis, gas does not seem to be so readily available to be transported to Denmark.

(156) Third, prices in transactions generally not affecting wholesale gas needs in Denmark, e.g. on the Benelux or United Kingdom or Emden hubs, appear to have little recognizable (and in any event insufficient) impact on the wholesale price situation in Denmark driven by other demand factors. (116)

(157) In its reply to the SO, DONG has underlined the difficulty of price comparison (117) between contracts. While the Commission accepts the argument regarding the difficulty of comparing prices of contracts, this does not alter the ultimate conclusion, as the fact that foreign hub prices are not used as a basis for price calculation in Denmark - as shown by the market investigation - is sufficiently evident [...]*.

(158) Fourth, the market structure in the different countries shows very considerable differences in the market shares of the different players. Whereas DONG reaches market shares largely above [60-70%]* in Denmark and Sweden, its market position in Germany and the Netherlands is below [0-10%]*, and even smaller in Belgium and the United Kingdom. In turn, E.ON Ruhrgas, one of the largest gas wholesalers in Germany, holds a market share of less than [0-10%]* in Denmark. Other important German gas companies such as RWE Trading and Wingas have only small activities in Denmark, or are not present at all. No supplies from the Netherlands appear to have ever been made to Denmark.

(159) In DONG's view, expressed in its reply to the SO, the market share differential is an inadequate indicator of substitutability and competitive constraint, citing the Commission's Relevant Market Notice. DONG notes that the differences stem from market structure and recent liberalization and refers to the inconsistency of DONG having an over [60-70%]* market share in both Denmark and Sweden and yet the two countries are not assumed to be in same market and furthermore, that it is easier for a competitor from Germany to access the Danish market than vice versa.

(160) In the Commission's view, market share is a relevant indicator and provides indications for market strategy of participants. In the assessment of the relevance of diverging market shares the Commission has taken account of evidence relating to the continuing process of market integration in the Community. However this process, in regard to the gas markets is not at such a stage nor can be expected to be at such a stage within the near future or indeed the foreseeable future with sufficient certainty, as to render current very important market share differences of mere historic relevance. (118)

(161) The fact that DONG has very strong positions in the wholesale market(s) in both Denmark and Sweden does not mean that Sweden can exert a competitive constraint on Denmark (as will also be discussed further below). Whether, due to differences in the regulatory set-up and unbundling (119), it is easier for a competitor from Germany to access the Danish market than vice versa, is subject to discussion, but is not supported by the fact that DONG exports much larger quantities to Germany than are imported from Germany to Denmark and the fact that DONG has also established itself in Northern Germany. (120)

(162) Fifth, in the Commission's market investigation Danish wholesale gas customers have had little knowledge about the wholesale price level in Germany. Importing gas (e.g. from a gas hub in Benelux) was generally considered a weak substitute to the availability of wholesale gas in Denmark. (121)

(163) DONG states in its reply to the SO that the argument focuses purely on alleged limitations of customers' own knowledge and does not account for supply-side substitutability. In addition, DONG states that as contractual imports from Germany take place, some customers must have some knowledge (122).

(164) The Commission maintains that demand side considerations are of high importance for any finding concerning the relevant market. (123) In addition, customers' level of awareness is an important indicator both of the demand side substitutability and of supply side substitutability as it can be based on both the customer "pulling" the information and on suppliers "pushing" the information to the customer. The Commission also maintains that such information level is low in Denmark (as is the level of imports).

(165) Sixth, wholesale gas suppliers which could import gas from Germany or further from the Netherlands face significant transport costs (124) through pan-caking,(125) capacity constraints (e.g. in the BEB network or regarding non-interruptible gas transport northward in the DEUDAN pipeline which can currently not be contracted) and important administrative obstacles (such as inquiries on capacity availability, negotiation on cost and terms of reservation and of possibility of cancellation of reservation, with a number of network operators, e.g. for the MIDAL, NETRA, RGH and DEUDAN pipelines and the entry into and transport through the BEB network; all procedures significantly complicated by other-than-regulated access regimes).

(166) According to DONG, as spelled out in its reply to the SO, the fact that commercial imports take place currently is itself evidence that potential obstacles noted by the Commission (at pre-SSNIP levels) are insufficient. They furthermore note that pancaking is exaggerated in the Commission's analysis - likely competitors have gas at BEB-VEP or Emden. Furthermore they note the existence of the BEB entry-exit regime, where Ellund is an exit point, thus from the BEB-VEP no entry and only an exit fee has to be paid for transport which as they state is lower than shipping to Lübeck.

(167) The Commission notes that DONG's assertion that pan-caking is not very relevant is based on the hypothesis that a shipper has gas at Emden, which nevertheless still creates extra cost to any transaction. For those participants having gas at other network points in Germany except the BEB-VEP the pan-caking effect applies even more. As to the BEB VEP, the Commission has not received any indication that there is sufficient liquidity at this point126 and for the companies receiving gas at Ellund the obstacles for transferring this gas to Denmark (see separate section below) apply a fortiori to gas they may have at BEB VEP. The comparison of the transport fees between Emden and Ellund and Emden and Lübeck, respectively, "compares apples and pears" (127) and, if anything, can explain why DONG was, commercially, in a good position for its participation in and its supplies to some "Stadtwerke" in Northern Germany.

(168) Therefore Germany as a whole, or even Northern Germany (with the possible exception of Ellund as argued in the following paragraphs), does not belong to the geographic scope of the Danish market for wholesale supplies for Denmark.

- The specific case of Ellund gas

(169) In its response to decision pursuant to Article 6(1)(c) of the Merger Regulation and the SO, the notifying party strongly argued with gas delivered by DONG, on the basis of contracts with originally [...]* to these third parties at the German side of the Danish entry/exit point at Ellund. The contracts for these deliveries can be terminated as of [...]*. DONG emphasised that [...]* together obtain approximately [...]* bcm of natural gas from DONG under long-term contracts with delivery point at Ellund. DONG submits that, in the case of a price increase in Denmark, this gas, which otherwise would have been sold in Germany, could be supplied in Denmark. DONG claims that [...]* mcm are currently being turned around back into Denmark by either these companies themselves or through their swap or sales agreements with other suppliers selling gas in Denmark.

(170) By way of a preliminary remark, it can be stated that the figure of [...]* mcm is exaggerated as regards turn around gas or swaps of gas delivered by DONG to [...]*, [...]* in Ellund. This follows from the mere fact that imports by E2 [...]* are not such turn-around gas, are included in this figure. Similarly there may be (and indeed are) other parties which contractually import gas to and into Denmark from south of the DEUDAN pipeline. Even leaving aside this possibility, the gas turned around at Ellund (or potentially the gas swaps into/for Denmark concluded with turn-around gas) could not have accounted for more than [5-10%]* of Danish consumption in 2004. (128) As stated, this percentage is actually even lower.

(171) The Commission acknowledges that the costs and obstacles of shifting this gas back into Denmark are not the same as the costs and obstacles for gas that comes from places south of the DEUDAN pipeline such as the Emden hub. For the potential turn- around gas, pan-caking costs are limited to cost of re-entry into the Danish system which are about [...]* of the purchasing price of that gas by [...]* and the opportunity cost of not using the capacity previously reserved in the DEUDAN pipeline. However, depending on the arrangements with the wholesale customer, the Danish exit costs and commodity charges may have to be paid (adding approximately [...]* of cost to the purchase price). The technical/administrative obstacles of transport into Denmark are also lower for Ellund turn-around gas. A re-entry nomination to Energinet.dk has to be made into a network not currently facing capacity constraints.

(172) However, in spite of this, nearly all of this potential turn-around gas is currently destined for southward transport. In this context, it needs to be considered that [...]* all have considerable downstream operations in Germany and that this gas is an integral and essential part of their German gas portfolio. They cannot, therefore, easily (i.e. without considerable alternative procurement and transport costs) replace the Ellund gas in their German portfolio. Consequently, nearly all the Ellund gas is transported south and therefore, even on the assumption that it would form part of the geographic market, would be "exported" from the geographic market. Consequently, such southward-earmarked gas, even if it were still a part of the geographic market, only exerts a limited competitive pressure on DONG as it is only to a very limited extent available for Danish consumption. The Commission notes that this is in line with statements of market participants that they did not succeed in obtaining natural gas from the recipients of potential turn-around gas. (129)

(173) DONG has stated in its reply to the SO that the volume of gas currently being turned around is inconsequential because at the margin, sales in Germany and in Denmark are equally profitable. Therefore in the event of a Danish price rise, greater volumes would be turned around. DONG has also stated that long-term reservations can be resold as there is a general scarcity of available transport capacity, thus, it should be easy to resell previously booked capacity in the DEUDAN pipeline. DONG also contends that companies purchasing gas at Ellund are able to plan ahead their capacity reservations in line with price development and that there is free capacity at Ellund to make this possible. [...]* all have much larger reserves, sourcing contract portfolios, and supplies than DONG has and can thus replace German-destined gas within their supply portfolio. In particular, Ellund gas purchased by [...]* is a very minor part of its portfolio, and [...]* The Commission does not provide evidence that [...]* cannot replace Ellund gas with alternative sources. Prices of Ellund gas are in line with market prices in northern Germany implying existence of alternatives. HNG's inability to obtain gas from Ellund is not relevant, as it was pre-liberalization, and therefore the gas availability was not an accurate reflection of current market conditions.

(174) In the Commission's view, based on the results of the market investigation, it is not evident that sales on the margin are equally profitable in Germany and Denmark. Reference can be made to the evidence in : Figure 3 above on German and Danish import prices and on Ellund prices and to the evidence on significant differences in downstream prices (130) indicating, as proxies, significantly different wholesale prices and margins.

(175) While long-term reservations can be resold, there seems to be no other counterparty in the case of the DEUDAN pipeline than DONG, who, on the one hand has no interest in facilitating turn-around that could endanger its market position in Denmark and on the other hand, with extra-southward capacity, would pose a clear competitive threat for each of the three companies in Northern Germany.

(176) The market investigation has demonstrated that all primary capacity reservations in the DEUDAN are made on a long-term basis, so it is unrealistic that these could be done on the short term. The fact that there is free entry capacity at Ellund is not, and has not been, the driver for any decision for turning around gas.

(177) It is correct that [...]* on a European scale all have larger reserves or contracted supply volumes than DONG, which limits their dependence on the DONG gas. However, the gas they are currently purchasing from DONG [...]* at Ellund and shipping south is an integral part of their supply portfolio. Taking that part out means having to replace it. Even if this were possible, the economic rationale of doing that may still be absent. The fact that up to now no significant volumes have been turned around is obviously due to the fact that it would not have been economically viable for them to turn around larger volumes. The evidence in Figure 3 above on the Ellund prices as compared to general German import prices proves that it is doubtful whether a 5% price increase in Denmark would induce them to profitably shift sales back to Denmark and replace them with other wholesale purchases (which would very probably exceed the BAFA prices, which will be influenced downward through more favourable long-term contracts, whereby it is likely that the flexibility in the most favourable contracts is already used to the full). Up to the end of 2005, there also does not seem to be any evidence supporting [...]* (131).

(178) The Commission furthermore maintains that HNG's unsuccessful request for Ellund gas (contrary to DONG's assertion only possible due to liberalisation of wholesale markets) is important evidence of the (lack of) economic incentives to sell that gas on the Danish wholesale market. DONG's long-term contracts with the three companies at the border have not changed since then and nor is there any evidence that changes have taken place concerning the overall economic rationale concerning this gas.

(179) On balance, the Commission therefore finds that, despite strong indications that this is not the case, the question of whether these sales belong to the geographic market or not can actually be left open. What is important, though, is that Ellund gas does not exert any strong competitive constraint upon DONG. (This is further reasoned in the discussion on the competitive assessment of the effects of the merger on this market below.)

- DUC sales to the Netherlands do not widen the market whereas DUC sales to Denmark would be in the market

(180) In its response to the decision pursuant to Article 6(1)(c) of the Merger Regulation and also in its reply to the SO, DONG submitted further that, following a commitment to the European Commission, the DUC partners, i.e. Shell, Chevron and A.P. Møller Maersk, are obliged to sell a total quantity of 7 bcm over a period of five years starting on 1 January 2005 to parties other than DONG. DONG stated that this gas could be sold in Denmark by transport via its offshore pipeline between Tyra and Nybro. The Commission finds that sales by the DUC partners to Denmark would not widen the relevant geographic market as such sales would be considered part of the Danish gas wholesale market anyhow. DUC sales through the NOGAT pipeline to the Netherlands, however, are outside the geographic market as this gas is not destined for wholesales for Denmark.

(181) In DONG's view the fact that none of this 7 bcm is currently sold into Denmark, does not mean that none of the DUC sales should be counted as part of market. In its view, the relevant question is what would happen in the vent of a price rise. DONG also claims that Energinet.dk implicitly confirms that DUC gas can go to the Netherlands or to Denmark (132).

(182) The Commission maintains its view that the additional gas available to the DUC parties, that they committed to sell to parties other than DONG as part of the agreement with the Commission in the DUC-DONG case, has not placed and is unlikely to place a significant competitive constraint on DONG, as it has so far been sold to the Netherlands and is likely to continue to be sold to the Netherlands. Norsk Hydro's press release confirms that Maersk has committed its non-DONG gas to the Latter (133) and Shell has also reported long-term commitments to the Netherlands. Thus it seems that it is not just the current state, but a likely scenario for the remaining duration of the entire relevant five-year period, that the DUC volume is destined for the Dutch market.

(183) The fact that Energinet implicitly confirms the legal (and regulatory) possibility of DUC gas going to Denmark does not change this assessment. Once a contractual arrangement provides for the delivery point to be e.g. at Den Helder, there are very significant financial/commercial and, according to the results of the Commission's market investigation, also technical obstacles to arranging for deliveries of this gas at Tyra. (134)

- Sweden is not part of the relevant geographic market for Denmark; it forms a separate market which is Swedish or Swedish-Danish in scope

(184) On the basis of the market investigation it appears that the gas wholesale market in Sweden does not exert strong competitive constraints on the gas wholesale market in Denmark and that the relevant geographic market is thus confined to Denmark. The Commission acknowledges that there are certain similarities between the gas wholesale markets in Denmark and Sweden. In particular, DONG holds a very strong position on the wholesale level in both countries.

(185) However, there are also important differences between the two countries. Whilst the Danish overall gas consumption is approximately 4 to 4.5 bcm p.a., the total Swedish gas consumption amounted to 0.9 to 1.0 bcm in 2004. Whereas the Danish gas consumption is considered as saturated, the Swedish demand is expected to grow considerably in the course of the next years.

(186) More importantly, as Sweden is not a gas producer, there is no gas available in Sweden to be supplied to Denmark, neither via physical nor via commercial flow. Any gas supplied from Sweden would be either a re-import of gas previously supplied from Denmark to Sweden, or gas from Germany which would have been previously shipped through Denmark since there is no direct pipeline between Germany and Sweden.

(187) DONG confirms [Contains considerations on Swedish and Danish terms of wholesale deliveries]* (135).

(188) Swedish gas wholesale currently does not exert any appreciable significant direct competitive constraint on the Danish wholesale market and there are no indications that this is likely to change. This is true also for the possibility of turn-around Danishoriginated wholesale gas at Dragør (136) into Denmark, although technically possible.

(189) DONG considers that market size and saturation are irrelevant to the consideration of whether Sweden exerts competitive constraint on Denmark. The relevant analysis is whether a SSNIP results in gas which would have been sold in Sweden being sold in Denmark. There is insufficient evidence and analysis on the possibility of turning around gas at Dragør. Conclusions about general market structure in Sweden should not be based on three contracts. (137)

(190) The Commission considers that the level of market growth in Sweden is likely to prompt market players in Sweden to focus on their home market, rather than on exercising competitive constraint on other markets. Any gas sold to Denmark from Sweden would have to be replaced by gas again coming from Denmark, and the fact that this has not happened in the past is a clear indication of the doubtful economics of such transactions. The fact that the Dragør deliveries have been taking place to Nova previously and have not exerted competitive pressure on Denmark substantiates this assumption which equally holds true for the deliveries [...]* and cannot afford to endanger the servicing of its customers in Sweden. The difficulty of comparing contracts of different backgrounds does not mean that Sweden can pose a competitive constraint on the Danish market.

(191) There are thus strong indications that Sweden is not part of the relevant market for wholesale sales to and for Denmark and will not become part of this market in the foreseeable future.

(192) However, the question of whether the wholesale market for Sweden should be defined as a Swedish one in scope or one that would be, in view of the asymmetry of constraints, a "Danish-Swedish market for Sweden" (138), can remain open for the purpose of the current decision. Any negative effect that the proposed operation may have on the market for Sweden would be an effect derived from the effect of the proposed operation on Denmark. If, therefore, as will be shown below, the competition problems on such a Danish market are solved by commitments submitted by DONG, then there cannot be any remaining negative effects on the market for Sweden. It is therefore not necessary to examine this market for Sweden in the assessment.

c) Markets for the supply of natural gas

(193) The parties submit that the identified product markets for supply of natural gas are wider than Denmark, also including Sweden, Germany, the Netherlands and the United Kingdom, except for the market of supplies to households which is defined as national. The Commission's preliminary findings do not confirm this view. On the contrary, markets are national, and for household customers may even be narrower than national, whereas for the market for supply to central CHPs the possibility of a market that encompasses both Denmark and Sweden is not excluded.

d) Market for the supply of natural gas to central CHPs

(194) The market for the supply of natural gas to central CHP plants is likely to be national in scope and may at most include Denmark and Sweden.

(195) General reference can be made to the discussion on the geographic scope of the wholesale market above. More specific reasons are given below.

The geographic market for the supply of central CHPs is probably not wider than Denmark

(196) This is for the following reasons:

* The geographic scope of the market for the supply of natural gas to central CHPs cannot be wider than the geographic scope of the wholesale market, as activity on the wholesale market is a prerequisite for any activity on the downstream market for supply of natural gas to centralised CHPs. Suppliers to central CHPs need a high degree of supply flexibility, at least for the basic multi-annual contracts which constitute a large part of the demand of central CHPs. This supply flexibility in turn can only be achieved with wholesale operations and storage facilities (or equivalent flexibility tools) within the same grid area.

* Due to administrative and technical obstacles (procurement, supply chain management such as reservation and booking of transmission capacity including the risk of unavailability of required amount of capacity), only parts of the supply needed by central CHPs can be sourced from neighbouring countries [...]*, but this can never be a substitute for the entire demand, or even a significant part thereof, partly because there is considerable uncertainty as to firm capacity bookings in the DEUDAN pipeline in the northerly direction.

* Wholesale supply management needs to be conducted in the context of supply within a specific grid, with specific in-depth knowledge of the exact operation of the regulatory system and the transmission system (and the day-to-day interaction with the system operator). This allows for some scale/scope effects (and therefore constitutes a barrier to entry for companies not able to realise these) if more than one plant is supplied or if other gas customers (in a neighbouring product market) are supplied.

For Sweden, the market for the supply of central CHPs is nascent - Sweden is likely to be a separate geographic market.

(197) The supply situation in Sweden is specific for a number of reasons outlined below.

(198) First, there is currently only one (fairly small) central CHP plant in Sweden (Heleneholmsverket in Malmö, owned by E.ON), which is atypical in the sense that it cannot, in contrast to the Danish central CHPs, operate in condensing mode. However, the entry into operation of one such central CHP plant (owned by Göteborg Energi) is foreseen in the near future, and another E.ON plant may be built and start operating in 2009.

(199) Second, Sweden has next to no storage capacity. This may however be changing, as Sweden has least planned for a significant increase in 'virtual storages', i.e. central CHPs providing if not storage then at least flexibility.

(200) For the pipeline linking Denmark and Sweden there are currently no capacity constraints and it appears that non-interruptible additional capacity can be booked at short notice without undue commercial risk as to non-availability. Pipeline capacity can be increased and seems to be sufficient at least until the entry into operation of a third Swedish central CHP (see above) perhaps in 2009.

(201) Entry barriers from the Danish into the Swedish market could therefore be low. Entry barriers in the opposite direction, i.e. for a hypothetical future supplier of Swedish central CHP plants supplying Danish central CHP plants, could be slightly higher, as there might be a certain risk that this supplier faces more difficulties in "counter-flow/ flow-back" access to the pipeline from Sweden to Denmark if it wants to divert quantities (back) to Denmark. (139) However, until the Baltic Gas Interconnector ("BGI") pipeline is completed (cf. below), such a hypothetical supplier is either likely to be established in Denmark and have access to wholesale gas in Denmark or, alternatively, in order to be competitive, it may need special arrangements with DONG e.g. as regards swapping of gas from the German-Danish border to the Danish Swedish border. (140) Such benefit-sharing arrangements would have the effect of "DONG facilitated market access" of such a supplier in Sweden, while at the same time keeping this player outside the Danish wholesale market, and under such circumstances any flow-back would be uneconomical.

(202) In view of the somewhat different market structure in Sweden, characterised by current and future differences in ownership of central CHPs, the retail strength of E.ON in Sweden, [...]*, and the market expansion in Sweden, the market structure seems to be evolving in a different way in Sweden than in Denmark, which may make the separation of a separate Swedish market for the supply of natural gas in Sweden, the more likely option. The supply situation of central CHPs in Sweden will, in any event, not have a decisive influence on the supply situation in Denmark, largely due to Sweden's current and future pre-BGI "cul-de-sac" position, which is unlikely to have any important repercussions for wholesale and supply to central CHPs in Denmark.

(203) In their reply to the SO, the notifying party did not question the existence of separate markets for supply of gas to central CHPs in Denmark and Sweden respectively.

(204) Therefore, for the purpose of this decision, and in line with the geographic market definition in the wholesale market, the existence of (i) a Danish market for Denmark and (ii) a Swedish market for Sweden is therefore assumed. However, as stated in the SO, the assessment of the operation on a "Danish-Swedish market for Sweden" would lead to the same result. (It can be added that this would even be true for an assessment of a joint market for both countries, although such a market, due to the asymmetry of competitive constraint exercised from Denmark to Sweden but not from Sweden to Denmark, is only a hypothetical possibility.)

e) Market(s) for supply of gas to decentralised CHPs and large industrial customers

(205) The market or markets for the supply of natural gas to decentralised CHPs and large industrial customers is/are national in scope.

(206) The parties in the Form CO and in their reply to the SO submit that the market is wider than national.

(207) For the following reasons, the Commission disagrees with the view of the parties.

(208) First, there are no direct imports. All suppliers to these customer groups are established in Denmark. (141) This need for a Danish subsidiary (or at least a dedicated Danish business unit) (142) is due to (i) dedicated marketing requirements for which the command of the Danish language also has an undeniable importance, (ii) the need to carry out nomination requirements on a national basis as the Danish grid is national, (iii) the fact that national legislation/regulation has an important influence on the conduct of a supply business in Denmark. There is furthermore, the need for national storage, unless the supplier has a very flexible supply contract in Denmark or at most at a Danish entry point.

(209) Second, the market structures are clearly different. Most of the companies active in supplies to these customer groups in Germany are not active in Denmark: Wingas, EnBW, ExxonMobil, RWE. Most of the companies active in supplies to these customer groups in Denmark are not active in supplies to these customer groups in Germany: HNG-MN, SGA, DONG. Companies active in supplies to these customer groups in Sweden are not active in such supplies in Denmark: Göteborg Energi, Lund Energi. DONG itself supplies in Sweden through a Swedish subsidiary. Companies active in supply to this customer group in Denmark are not active in Sweden: e.g. HNG-MN, SGA.

(210) Third, there are substantial price differences between Denmark, Germany, Sweden and the Netherlands as shown by the Eurostat statistics in the following table. (143)

<emplacement tableau>

(211) Fourth the timetables in the different countries for the opening of the market for these customer groups, although now completed, were different which may still have a nonsynchronized effect further accentuating national differences.

(212) There are, by contrast, no indications that the relevant geographic scope of the market(s) for supply of natural gas to decentralised CHPs and large industrial customers is narrower than national for Denmark and Sweden either. East Denmark and West Denmark are well connected in terms of high pressure gas pipelines. There is a uniform regulatory scheme and tariff for access to the transmission network inside Denmark. The same holds true for the Swedish regulatory scheme (and transmission system).

f) Market for supplies of gas to households and small businesses

(213) The notifying party submits that the market for supply to businesses is not narrower than Denmark, Sweden, Germany, the Netherlands, Belgium and the United Kingdom.

The market for supplies to households is considered by DONG to be national since customers have the right to choose their gas supplier, since the regulatory regime is the same throughout Denmark and since the suppliers are sophisticated companies which are well placed to compete nationally.

(214) The Commission agrees that the market for supplies to household customers in Denmark is not wider than national. As regards the geographic scope of the market for supplies to small business customers (on the assumption that this could be a separate market), reference can be made to the arguments set out above for large industrial customers and decentralised CHPs, which a fortiori militate against wider than national markets for this customer group.

(215) Sales on this market require a national sales and service office. The current suppliers to non-metered customers in Denmark are all Danish and Danish companies are not supplying non-metered customers in Sweden. It would therefore not be possible for customers on this market to source their supply with companies abroad, e.g. in Sweden. The same holds true with regard to Germany. This clear focus on no wider than national activities is motivated by a number of factors such as language requirements of the marketing and customer service functions. Furthermore, regulation is national, in particular as regards TPA to the transmission and distribution networks and to storage facilities. Danish supply companies to small business and household customers source their supplies on the wholesale markets which, as noted above, are no wider than Denmark (and Sweden). Furthermore the timing of the opening of markets to competition is different: all Danish customers were eligible as per 1 January 2004, while this is only the case for Swedish households as per 1 January 2007. The differences between these national markets are also borne out by the price differences as indicated in the table below.

<emplacement tableau>

(216) With regard to small business customers DONG has argued in its Reply to the SO that Eurostat's figures are not correct and that prices are indeed lower than indicated in Denmark. The Commission notes that DONG does not adduce clear evidence for its contention as to the incorrectness of Eurostat figures and in any event does not argue that prices are the same in Denmark as in its neighbouring countries, which makes it unnecessary to arrive at a definite conclusion as to the correctness of Eurostat prices for small Danish business customers.

(217) For these reasons, it can be concluded that the market or markets for supply of gas to non-metered customers (small business customers and household customers) is or are no wider than Denmark.

(218) In previous decisions, too, the Commission has considered gas supply to households and small customers to be no wider than national. However, the question of whether markets are even smaller than national and should be defined along the boundaries of the distribution grids/area has been raised. Whereas all Danish customers, including non-metered customers were eligible as of 1 January 2004, only a very few customers have so far decided to change supplier. The vast majority of customers have stayed with their respective regional distribution company. In 2004 less than 1% of nonmetered customers (depending on the distribution area between 0.08% and 0.67% of customers) changed supplier.

(219) It follows that the regional distribution companies still have very large market shares (>98%) in their respective areas. This is also consistent with experience from other countries in the first years after market opening (e.g. Germany, United Kingdom and Norway).

(220) Furthermore, prices to end consumers (households) are different in the three areas, as can be seen from the table below, where prices deviate by up to three per cent.

<emplacement tableau>

(221) The existence of large differences in market share as well as price differences between regions would normally be indications that regional markets are not integrated.

(222) Furthermore, access to the regional distribution grids, which are operated by the regional distribution companies is not fully harmonised. Naturgas Fyn, the operator of the distribution grid on the island of Funen and one of the parent companies of Statoil Gazelle, has made its own standard terms for access to the distribution network, while DONG and HNG/MN coordinate their terms and conditions with the TSO.

(223) In the reply to the SO, the parties maintain that the relevant market for supply of gas to households is national rather than regional since both HNG/MN and Statoil Gazelle are actively competing for customers outside their traditional areas, and provide as an example an advertisement by Statoil Gazelle aimed at household customers in the Zealand area.

(224) The Commission acknowledges, that Statoil Gazelle - at least on one occasion - has advertised for customers outside its own area. However, the fact that Statoil Gazelle tries to win customers in other areas may be seen as an attempt to enter these areas rather than evidence that the market has become national. Furthermore, the parties do not present any evidence indicating that DONG is marketing to small customers in other areas, e.g. the HNG/MN areas, nor that HNG/MN is marketing in other areas, e.g. the DONG areas. The fact that one competitor, perhaps on a single occasion, invests in an advertisement aimed at households in another area, can therefore not in itself be seen as evidence that the market is national especially not in the light of the evidence spelled out above, pointing in the direction of regional markets.

(225) It is therefore concluded that there still exist regional markets for retail supply to nonmetered customers and that these markets coincide with the distribution areas owned by DONG, Naturgas Fyn and HNG/MN.

(226) However, the Commission notes that the geographic scope of the market for retail supply of gas to non-metered customers may become wider, as the process of the opening of markets to competition reaches a higher level of maturity. This question can ultimately be left open for the purpose of this decision.

3. Conclusion on relevant markets in the natural gas sector

(227) For the purpose of this decision, the relevant markets defined are therefore the following ones:

(1) the Danish market for storage or alternatively the Danish market for flexibility of natural gas; the Swedish market for storage or alternatively the Swedish market for flexibility of natural gas (Swedish or alternatively Swedish-Danish in scope);

(2) the market for wholesale of natural gas for Denmark (Danish in scope); the market for wholesale of natural gas for Sweden (Swedish or alternatively Swedish-Danish in scope);

(3) the Danish market for the supply of natural gas to central CHPs; the Swedish market for the supply of natural gas to central CHPs (Swedish in scope or Swedish-Danish scope);

(4) the market or markets for the supply of natural gas to large industrial customers and to decentralized CHPs (Danish in scope);

(5) the market or markets for the supply of natural gas to small business customers and to households (Danish or regional in scope).

VI. PART B - ELECTRICITY

1. Relevant product markets

(228) The Commission has in the past distinguished separate product markets for the generation and wholesale supply of electricity (i.e. production of electricity in power plants and physical import of electricity through inter-connectors and its sale on the wholesale market to traders, distribution companies, electricity exchanges or large industrial end-users); regulating/balancing services; transmission of electricity (via high-voltage grids); distribution of electricity (via medium and low-voltage grids) and retail supply of electricity. On the retail level, the Commission has distinguished between large (industrial) customers and small (small business and household) customers (147).

(229) The parties have presented information on the following markets:

- generation and wholesale of physical electricity (including system services);

- financial derivatives of electricity;

- electricity retailing to business customers;

- electricity retailing to households.

Generation and wholesale of physical electricity; Ancillary services

(230) The notifying party submits that the relevant market includes sales on the Nordic electricity exchange Nord Pool, bilateral contracts, as well as reserve capacity and regulating/balancing power.

(231) The notifying party's assumption of a separate relevant market for generation and wholesale of physical electricity (henceforth "electricity wholesale market") is in line with previous Commission decisions. However, the scope of this market proposed by the notifying party deviates from previous Commission decisions, as the notifying party includes ancillary services (also termed as balancing services and regulating power or system services) within the wholesale market (see further below).

- Electricity wholesale

(232) An important part of the sales and purchases of wholesale electricity in the Nordic area is conducted via the Nordic electricity power exchange Nord Pool Spot ASA ("Nord Pool"), which covers wholesale transactions of physical electricity in Finland, Sweden, Norway and Denmark.148 Nord Pool is a joint venture between the Nordic electricity TSOs and operates a day-ahead market matching supply and demand of wholesale electricity in the entire Nordic area 12 to 36 hours before delivery. Nord Pool complements this day-ahead spot exchange called 'Elspot' by a trading option termed 'Elbas', which permits trading of electricity even closer to its physical delivery and consumption.

(233) In addition to these Nord Pool-based forms of trading of physical electricity at wholesale level, producers/traders and customers also engage in bilateral contracts of physical wholesale electricity. (149)

(234) For a certain group of wholesale customers (usually small retail suppliers), direct access to Nord Pool seems difficult. While Nord Pool trades standard products (e.g. delivery/purchase of 1 GWh (150) of electricity in flat supply/consumption profile during a certain hour of the following day), bilateral contracts can be more individualised. (151) However, it appears that prices of bilateral contracts are usually tied to Nord Pool prices, which points to a very close link between both forms of wholesale electricity trading. For the purpose of this decision, the question whether the supply of electricity to retailers that are not active on Nord Pool (152) could constitute a separate (sub-) market can be left open, as the proposed operation will not lead to competition problems on either alternative.

- Ancillary Services

(235) The Commission has on several occasions considered the provision of regulating/balancing power and reserve capacity to constitute one (or more) separate market(s) (153). For this market (these markets), the terms "balancing power", or "ancillary services" have partly been used in these decisions. (154) In the Commission's market investigation the terms "regulating power" and "system services" have also been mentioned for either the same or closely related services. For the purpose of this decision it is not necessary to define these terms exactly.

(236) The immediate customers of systems services are TSOs, who are responsible for maintaining the balance in the grid and for securing supply in emergency situations. TSOs need reserve capacity, i.e. capacity (e.g. a power plant unit) kept on stand-by mode for major emergency situations. They also need regulating/balancing power which is used continuously to maintain the balance between supply and demand of electricity in order to maintain the correct grid tension. Such regulating/balancing power can itself be divided into a capacity stand-by function (i.e. the willingness to provide such services up to an agreed limit, measured in MW (155)) (156) and the actual use of this service whereby the TSO asks service providers to regulate downwards or upwards the tension on the grid, measured in MWh (157), (158). The TSO secures the means to perform its tasks by concluding agreements with producers to maintain (positive or negative) capacity on stand-by in case of unexpected shifts in demand or supply. The TSO has to pay for these services but passes on its costs to other market participants.

(237) The Commission's market investigation has led to clear indications that ancillary services /system services are not easily substitutable with other electricity supply at the wholesale level. This is not disputed by the notifying party as far as the demand-side is concerned. Regulating/balancing power is not traded on Nord Pool and Nord Pool does not in the foreseeable future expect regulating and balancing power to be traded on the exchange. The TSOs are the sole buyers of these services and cannot readily substitute regulating power/balancing power with trades on Nord Pool.

(238) With regard to the supply-side the notifying party submits that producers can decide to switch their offer between selling electricity on Nord Pool and selling system services to the TSO. However, the need to commit capacity to system services well in advance of the Nord Pool spot market (and usually for much longer periods) and the need to be able to launch extra capacity at very short notice distinguishes both types of electricity/services and restricts the possibility of supply-side substitutability. The substitutability is further limited by the fact that the pricing of regulating/balancing power is different from the rest of the wholesale market. Furthermore, the business area of sales of regulating/balancing power to the Danish TSO is characterised by a very small number of suppliers compared with the number of electricity producers in Denmark, with suppliers on Nord Pool and even compared with the suppliers on the basis of bilateral contracts.

(239) The Commission also notes that an operator's share of regulating power may vary significantly over time without any apparent relation to its position on the market for electricity production. According to data submitted by the notifying parties, while Elsam's share of regulating power in Western Denmark declined from [...]* in 2002 to [...]* in 2004, its market share in generation in Western Denmark increased from [...]* in 2002 to [...]* in 2004.

- Conclusion on electricity wholesale and ancillary services

(240) For the purpose of this decision, the precise delineation of this market/these markets of wholesale electricity and system services/ancillary services can be left open. In any event, there is at least one separate market for wholesale electricity, whether or not it encompasses (i) supplies to customers without a direct access to Nord Pool and (ii) system services/ancillary services. Bilateral supplies to customers without a direct access to Nord Pool may or may not constitute a separate product market and the same can be said for ancillary services.

Financial derivatives of electricity

(241) The notifying party has submitted that there exists a relevant product market for financial derivatives of electricity trading, i.e. a financial wholesale electricity market separate from the physical electricity wholesale market (or markets).

(242) At present, the contract types traded on Nord Pool's Financial Market, called Eltermin, comprise (i) electricity derivatives, (ii) CO2 quotas and (iii) certificates for renewable electricity (Swedish certificates only). The latter is relevant for Sweden only and does not need to be considered for the purpose of the current decision. CO2 emission rights trading, which serves a purpose entirely different from financial derivatives of electricity trading, will be discussed in a separate section below. Electricity derivatives as products constituting, possibly, one or more separate product market(s) for financial derivatives of electricity require further consideration.

(243) The electricity derivatives traded on Eltermin are base load futures, forwards, options, and so-called contracts for differences ("CfDs"), hedging against the risk of Nord Pool area price differentials. The reference price for these contracts is Nord Pool's System Price. The maximum trading time horizon is currently four years. There is no physical delivery of financial market electricity contracts. Cash settlement is made throughout the delivery period, starting at the due date of each contract. In addition to Nord Pool, there are also bilateral financial contracts (i.e. contracts concluded between two operators directly or through a trading platform but without the activity of Nord Pool as a counterparty guaranteeing financial settlement) which, however, seem to be closely linked to Eltermin. (159)

(244) Financial derivatives of electricity contracts have been designed to satisfy the needs of various players: ??generators, retailers and end-users that use the products as risk management tools;

??traders who profit from volatility in the power market and contribute to high liquidity and trading activity and indirectly provide a financial insurance service to generators, retailers and end-users.

(245) Fundamentally, therefore, the financial market is about trading risk while the physical market is about trading electricity for consumption. Although to some extent the two forms of electricity trading are related (and mutually influence each other's prices), (160) the Commission's market investigation has indicated that there exists at least one separate market for financial derivatives of electricity as distinct from the market (or markets) for physical contracts. First, physical electricity and financial derivatives of electricity are not completely interchangeable as regards settlement and time horizon. Financial derivatives of electricity always have a cash settlement and are not sold on a spot (day-ahead) basis, whereas physical electricity from Elspot or bilateral contracts are delivered physically and are partly contracted on a spot basis. Thus, the financial market cannot provide a "perfect hedge" for the players on the physical market as, e.g., the financial market uses base load as its reference value and takes monthly averages as reference values, not the hourly prices actually payable by market participants. Secondly, a considerable market volume on Eltermin is due to market participants, such as trading platforms, banks and other financial service operators which are not, or to a much lesser degree, active on Nord Pool spot or on other physical electricity markets.

(246) Within financial derivatives of electricity a special role is played by CfDs (161). Companies using the CfDs - other than for pure trading purposes - can "lock" the Nord Pool area prices to the Nord Pool system price by buying (or selling) CfDs in addition to the financial forward and the physical spot positions they have. This means that the different CfD products managed by Eltermin (162) have their relevance when hedging the respective area prices to the system price, i.e. when those particular markets form separate price areas for certain periods. As the various area CfDs are not substitutable with each other or with the other financial products hedging electricity (except for purely speculative financial players), and since CfDs appear to be mainly supplied from market participants within the relevant price area, the CfDs could be regarded as separate sub-markets. For the purpose of this decision, this question can, however, remain open.

Electricity retail

(247) The notifying party has submitted that there exists a relevant market for the sale of electricity to retail customers, which consists of the physical sale of electricity to end consumers. This market can, in the view of the notifying party, be further split into the retail supply to business customers on the one hand, and to private households on the other.

(248) The Danish Competition Authority has in a recent decision (on the Elsam/NESA merger) divided electricity retail markets into two separate markets: one for customers with hourly metering requirements, and one for customers without metering requirements (i.e. having a standard load profile). (163)

(249) The market investigation has clearly confirmed the relevance of this distinction between metered and non-metered customers. It is easy to identify which of these groups a customer belongs to and there is no arbitrage possibility between them. These customer groups pay different prices, consume different products (metered/nonmetered) and purchase in different ways (negotiation vs. standard purchases).

(250) The question whether it is appropriate to distinguish smaller business customers, e.g. commercial customers, as a separate customer group is of no relevance to the assessment of this case and therefore does not need to be answered. This is because the market investigation has yielded no indication (i) that the assessment of either a large customer market or a small customer market would change if smaller business customers were excluded from such a market and (ii) that the assessment of the impact of the concentration on such a hypothetical smaller business customer market would lead to any different result from the assessments of the large customer market and of the small customer market, both of which, as will be shown below, do not give rise to competition concerns.

(251) For the purpose of this decision, the retail supply of electricity to metered business customers and the supply to non-metered (predominantly household) customers therefore constitute the two relevant retail electricity markets.

Summary of electricity product markets

(252) Therefore, the following electricity product markets are deemed to constitute relevant product markets for the purpose of this decision:

- wholesale of electricity

- whether or not including bilateral wholesales to customers not having access to Nord Pool (which may or may not constitute a separate market);

- whether or not including ancillary services (which are likely to constitute one or more separate markets);

- financial derivatives of electricity,

- whether or not including CfDs (which may or may not constitute a separate market);

- electricity sales to metered (business) customers;

- electricity sales to non-metered (predominantly household) customers.

2. Relevant geographic markets

Wholesale of physical electricity

(253) The notifying party considers the market for electricity wholesale to be wider than Denmark, at least pan-Nordic, essentially as an effect of the Nord Pool spot market which covers wholesale of electricity in Denmark, Norway, Sweden, and Finland. The notifying party argues that an additional integrating effect stems from the Nord Pool financial derivatives of electricity market which allow producers, traders and wholesale customers to hedge their production and consumption against price volatilities of the Nord Pool spot market. Moreover, the notifying party claims that congestion levels at the inter connectors between the various Nord Pool areas are fairly low. (Low congestion levels result in a limited number of hours in which the wholesale prices between neighbouring Nordic areas differ.) The notifying party mentions average yearly deviations in wholesale prices of less than 6 % between the Nordic areas as further substantiation of a wider than Danish market.

(254) In previous Commission decisions, the relevant geographic market for wholesale supply of electricity has normally been considered as national (164). However, the Commission has occasionally left open the possibility of wider than national markets considering that following market opening, the relevant geographic market may be broader than national if there are no longer barriers - in particular physical or regulatory barriers - to electricity exchange. (165) The Danish Competition Authority, in its decision on the Elsam/NESA merger in 2004, defined the relevant geographic wholesale markets as Denmark East and Denmark West respectively, inter alia on the basis of significant price differences and congestion between these entirely separate Nord Pool areas and their neighbouring areas within or outside the Nord Pool zone.

(255) The Commission notes initially that according to data made available to the Commission by the notifying party, Denmark West constituted a separate price area at Nord Pool in 39% of the hours in 2005 (up until August 13). In previous years this number varied from 19.3% in 2001 to 48.9 in 2003. According to the same figures, Denmark East was a separate price area only in 13.7% of the hours in 2005. In the two previous years the number was 6.1% and 5.0%.

(256) From the numbers regarding West Denmark alone it is apparent that it would be inappropriate for the Commission to assess the market on the basis of a Nordic-wide market, since in a very substantial fraction of the hours producers in West Denmark are not constrained - and hence not in direct competition with - producers in the rest of the Nordic region.

(257) The market investigation in this case has led to the following results:

(1) As there is at present no direct interconnection between the two Danish price areas, Denmark East and Denmark West, it would seem inappropriate to define the relevant geographic market as all-Danish in scope. The electricity wholesale market can therefore currently only be either wider than Denmark or narrower than Denmark (i.e. Denmark East and Denmark West) but not all-Danish.

(2) Between Sweden and East Denmark the congestion of the interconnector is not perceived as being high by many market participants. It is generally below 15%, though reaching values of around or above 20% in individual months. (166)

(3) Between West Denmark and Sweden/Norway congestion levels are significantly higher, and are perceived as being high by a large number of market participants. Congestion reached a value of above 56% for the first nine months of the year 2005. (167)

(4) Deviations in annual wholesale prices of up to 6% between Nordic areas are not insignificant and, in addition, are based on much larger fluctuations and deviations over shorter time horizons. When asked whether diverging wholesale prices between East Denmark and West Denmark have an important effect on differences in retail pricing in East and West Denmark, respectively, many market participants answered in the affirmative.

(5) Financial derivatives of electricity may have a certain integrating effect by smoothing the risk of diverging wholesale prices. However, belonging to a separate product market, financial derivatives of electricity cannot determine the geographic scope of the physical wholesale electricity market. (168)

(6) Moreover, respondents to the Commission's market investigation have expressed the view, based on past observations of market behaviour, that Elsam and, to a somewhat lesser extent, E2 can have some influence on when Denmark West and East respectively will be part of a wider Nordic market and when they will be isolated price areas. This would mean that Elsam and E2 could wield market power in their respective price areas even in periods of non-congestion, thereby weakening the evidence provided by average figures on periods of noncongestion. The Commission takes note of these views but cannot, in the context of this procedure, establish whether they are well-founded or not.

(7) Even though interconnector capacities are relatively high in West Denmark (67% of peak demand) and in East Denmark (74% of peak demand), there is still a considerable amount of demand that can only be met by supplies from within the price area. Moreover, these nominal capacities overestimate the actual capacities available for market disciplining. (169) The figures of interconnection capacity as a percentage of peak demand therefore need to be interpreted with some caution.

(8) If sales to customers not having access to Nord Pool Spot were, as submitted by DONG, to form part of a general wholesale electricity market, the fact that a number of (wholesale) customers may need to rely on sourcing electricity within the respective Danish areas can be taken as an indication of the relevance of an assessment at individual Nord Pool wholesale price level. This argument obviously does not apply if sales to these customers do not form part of the general wholesale market.

(258) Therefore, an analysis of the competitive effects of this concentration needs to take into account the possibility that, at least partly and possibly limited to some (congested) periods, the relevant framework for analysis is separate (a) East-Danish and (b) West Danish wholesale sub-markets.

(259) However, the analysis of the competitive effects of this operation likewise needs to take into account situations in which Nord Pool price zones are larger than individual East Danish or West Danish price zones. At one extreme all Nord Pool areas could, at times, be joined price areas. (170) In addition, any other combination linking East Denmark and/or West Denmark with one or more other Nord Pool price area may be relevant as well.

(260) Considering that the parties' Nordic activities are almost exclusively limited to Denmark and bearing in mind the current absence of a physical connection between East and West Denmark, the maximum (hypothetical) negative impact of the merger (in terms of maximum market share) outside the narrow East Danish and West Danish markets would thus arise (as long as no physical connection between East Denmark and West Denmark is built) in the case of a price area combination of the following Nord Pool areas: East Denmark plus West Denmark plus Sweden. (171) If the concentration does not lead to negative effects on competition in such a combination of areas then it cannot lead to negative effects in any wider price area. (172)

(261) Furthermore, one specific combination needs to be analysed with regard to a situation after 2010, at which point the Great Belt Interconnector cable, which is currently in project stage but likely to come into operation as of 2010, is likely to have an integrating effect on the hitherto not directly connected East Danish and West Danish wholesale areas. However, both the remote time-frame and the uncertainty as to how strong that integrating effect will actually be limit the weight that the Commission can give to those effects in comparison with other much more immediate effects.

(262) All other geographic market definition issues for the wholesale electricity market with regard to the current operation can, for the purpose of this decision, remain open.

Ancillary services

(263) The notifying party points to tendencies to source on a wider (Nordic) geographic level if ancillary services/system services were defined as constituting one or more separate product markets. Whether or not such services are distinct from the wholesale market can be left open for the purpose of this decision (cf. recital (240)). Thus, the four Nordic TSOs have recently established a co-operation for regulating power.

(264) However, ancillary services are dependent on immediate and reliable availability within a certain price area. The existence of congestion of interconnectors thus effectively impedes cross-border trade in these services in a considerable number of hours. This is also supported by the market investigation which has shown that ancillary services traditionally have been sourced from within each price area, with Elsam being the main supplier in Denmark West and Energi E2 being the main supplier in Denmark East.

(265) Moreover, the cooperation between the Nordic TSOs is still at a nascent stage. The likelihood of the cooperation between the Nordic TSOs actually leading to a wider geographic market in the foreseeable future is at the very least questionable.

(266) Accordingly, the possible market(s) for ancillary services is/are, for the purpose of this decision, considered to be limited to each of the two Danish areas.

Financial derivatives of electricity

(267) As regards financial derivatives of electricity, the parties submit that the market is wider than national and at least pan-Nordic, because financial derivatives of electricity trading take place on Nord Pool's Eltermin with the different financial products priced against the Nordic system price. On a general level, the market investigation has not resulted in a contrary assessment.

(268) However, as regards a specific group of products - the CfDs, (a special financial product which hedges against the risk of area prices e.g. Denmark East or Denmark West being different from the Nord Pool system price) - the relevant market could comprise the particular price area only. Companies trading in these products will mainly be the wholesale suppliers and customers in these very price areas.

Electricity retail

(269) In electricity retail markets, the parties have submitted that the geographic market for the household and business retail electricity markets is at least Denmark but - looking ahead to the future - likely to extend to the Nordic countries. (173)

(270) The results of the market investigation indicate that the geographic market for metered customers is not broader than Denmark. This is supported by the fact that the largest industrial companies also source electricity from Danish electricity supply companies and there is no direct entry by foreign supply companies. (174)

(271) There is no indication that the geographic scope of the market for business customers is narrower than Denmark. This is even though the competitive strength of market players (such as NESA, or KE) still appears to be different in East Denmark and West Denmark. However, customers both the east and the west regard suppliers that, historically, originated in the other area as viable alternatives and these suppliers sell in both areas. The historic pre-liberalisation separation of intra-Danish markets has therefore lost its importance and no longer constitutes the relevant framework for the assessment of this concentration. This is in line with the findings of the Danish Competition Authority in the Elsam/NESA merger.

(272) As regards the market for small non-metered customers there is still considerable unevenness of competitive strength of market players within their previous monopoly areas and outside these areas. Many customers have remained with their local universal service obligation (USO) supplier, which is in clear contrast to the much more price-sensitive behaviour of customers in the metered market. Switching rates in the non-metered customer area are low. This is an indication that this market could still be regional in scope. However, the emergence of new suppliers (such as OK and Scanenergi) which do not have a geographical imbalance of sales, and the fact that some competitors (e.g. NESA, KE, Nordjysk, NOE) are acquiring customers in areas outside their traditional territory can also be seen as indications of an, albeit slow, transformation of the Danish small customer electricity market(s) into a national market. There is, by contrast, no indication of the existence or emergence of separate East Danish and West Danish small customer markets. For the purpose of this decision, the market or markets for supply of electricity to non-metered customers (household customers and small business customers) is/are therefore either regional (i.e. coinciding with previous monopoly distribution areas) or national in scope.

Summary of relevant electricity markets

(273) Therefore, and with reference to the findings on the relevant product markets (cf. recital (252) above), the following electricity markets are deemed to constitute relevant markets for the purpose of this decision:

(1) wholesale sale of electricity (being East Danish and West Danish respectively or wider)

(2) possibly, bilateral wholesales to customers not having access to Nord Pool (being East Danish and West Danish respectively);

(3) possibly, ancillary services (being East Danish and West Danish respectively);

(4) financial derivatives of electricity (encompassing the Nord Pool area if excluding CfDs)

(5) possibly CfDs (being East Danish and West Danish respectively);

(6) electricity sales to metered (business) customers (being Danishnational);

(7) electricity sales to non-metered (predominantly household) customers (being Danish-national or Danish-regional).

VI. PART C - OTHER MARKETS

(274) A number of other markets are affected by the proposed concentration.

1. Product markets

District heating

(275) The parties submit that a separate relevant product market exists for district heating (i.e. heat in the form of steam or hot water used for district heating purposes), which is produced both in connection with electricity generation and in dedicated heating plants. The results of the market investigation have confirmed the parties' view.

Fly ash production

(276) The notifying party submits that fly ash production is affected by the transaction and that the relevant product market should be defined as one for aggregates and substitutes for cement, clinker, calcium silicate and sand. The parties' activities are limited to fly ash production (as a by-product of coal fired power plants). As the concentration does not raise any competition concerns even on this narrowest possible product market, fly ash, the question of the correct scope of the product market can be left open.

CO2 trading

(277) The notifying party has submitted that there exists a relevant product market for CO2 emission rights trading. As mentioned above, CO2 emission rights are currently one of the contract types traded on Nord Pool's Financial Market Eltermin. The Commission's market investigation has confirmed that a separate relevant product market for CO2 emission rights trading exists.

2. Geographic markets

District heating

(278) The notifying party submits that the geographic scope of the market for district heating is local since the sale of district heating is geographically limited to the specific heating supply grids, which are not interconnected in Denmark. The results of the market investigation have supported the parties' suggestion.

Fly ash production

(279) The notifying party submits that the geographical scope of the market for fly ash production is Northern Europe and the North Atlantic Basin, due to the low costs of shipping fly ash in this area. However, the notifying party also acknowledges that "transport by road is rather expensive". It is therefore not excluded that the geographic scope of the market could be national or even local. For the purpose of this decision it is not necessary to define the relevant geographic market.

CO2 trading

(280) As regards CO2 trading, the notifying party has submitted that the market is at least EEA-wide. The market investigation has confirmed that the market is, on the basis of the existing legal provisions, (175) likely to be EU-wide.

3. Absence of competition concerns in these other markets

(281) On none of these markets can the operation have any negative impact on competition. This is for the following reasons:

- The parties' market shares in CO2 trading rights in the EU are clearly below 10%.

- In fly ash production, the activities of Elsam and Energi E2 have already been combined in a joint venture (176) pre-merger so that post-merger there will be no addition of market shares.

- In district heating, there are no geographical overlaps in activities. (177)

(282) It is therefore not necessary to examine the effect of the concentration on such markets any further.

VII. COMPETITIVE ASSESSMENT

VII./1 PRELIMINARY REMARK ON MINORITY SHAREHOLDINGS

(283) In its reply to the SO, DONG argues that the Commission's analysis, to the extent that its conclusions are based upon E2 and/or Elsam being actual or potential competitors of DONG, is flawed by the failure to acknowledge DONG's pre-merger influence over both E2 and Elsam. (178) According to DONG this influence could be sufficient, in the absence of the merger, to at least inhibit either E2 or Elsam from entering any gas market in competition with DONG, or from remaining as a competitor of DONG to any appreciable extent.

(284) According to DONG's reply to the SO, the following shareholding links existed among the different undertakings concerned existed immediately prior to the notification:

(a) DONG owned 24% of the shares in Elsam and Vattenfall owned 35% of the shares in Elsam (neither of these acquisitions were notified to the Commission or the DCA);

(b) Elsam owned 86% (this acquisition was cleared by the DCA in 2004) and DONG owned 13% of the shares in NESA;

(c) NESA owned 36% of the shares in E2;

(d) local authorities, consumer owned supply companies, or undertakings controlled by them, held the remaining 64% of the shares in E2, including KE (with 34%) and FE (with 2%).

(285) The reply to the SO infers from these shareholdings that DONG is able to influence Elsam and, partially through Elsam, NESA. It also suggests that DONG is, through Elsam and NESA, able to influence E2. Furthermore, it suggests KE is able to influence E2, and Vattenfall is able to influence Elsam.

(286) As DONG itself acknowledges in its reply to the SO, the various shareholdings (except Elsam's shareholding in NESA) fall short of control for the purposes of the Merger Regulation or the equivalent provisions of the Danish merger legislation. However, control within the meaning of Article 3 of the Merger Regulation is defined as the possibility of exercising decisive influence on an undertaking. DONG thus agrees that the various shareholdings do not confer such a possibility of exercising decisive influence, since the acquisition of these shareholdings would otherwise have required notification to the Commission or the DCA. This applies in particular, to the recent acquisitions of shares in Elsam by DONG and Vattenfall.

(287) Prior to its acquisition of 24% in Elsam, DONG only had a 13% shareholding in NESA which in turn held 36% of E2's shares. (179) This means that DONG had no influence at all in Elsam, and virtually no influence in E2, far less control of either these companies.

(288) The proposed concentration therefore leads, in any event, to a change from no control to sole control by DONG. This change of control constitutes a significant alteration of the competitive structure of the affected markets. As it is undisputed that the preexisting shareholdings did not confer control to DONG, i.e. even not the possibility to exercise decisive influence, DONG is, prior to the merger, not in a position to prevent Elsam or E2 from entering gas markets or from remaining in these markets.

(289) This analysis is not altered by DONG's reference to Commission decisions in which pre-existing minority shareholdings were taken into consideration in the competitive assessment. The Commission does not deny that minority shareholdings can, in conjunction with other elements, play a role in the analysis of the competitive situation and structure on certain markets and the assessment of the competitive effects of proposed operations.

(290) However, the decisions invoked by DONG do not support its conclusions. Inferences made by DONG in its Reply to the SO are incorrect as the analysis made in the decisions to which it refers concern a very different issue, namely the role of shareholdings in third companies within the competitive assessment. Against this background, the Commission found in those decisions that numerous shareholdings (many of which were majority shareholdings) of an undertaking concerned in third companies which were not involved in the concentration, would, in addition to several other factors, strengthen the competitive position of the parties. This analysis of the competitive strength of the merged entity in the market differs manifestly from the analysis of pre-existing minority shareholdings between the merging parties.

(291) In any event, the analysis made in those cases cannot be transposed to this case, particularly since the factual circumstances differ considerably. (180) For instance, the differences in size and in dependency which played a role in the competitive assessment in the decisions referred to by DONG, do not exist in this case: neither Elsam nor E2 needed DONG's know-how or access to generation and transmission capacity in order to run their business and, in terms of size, Elsam has a higher turnover in Denmark than DONG.

(292) For these reasons, the Commission disagrees with DONG's arguments concerning DONG's alleged pre-merger influence over Elsam and E2 and maintains that the notified transaction brings about a significant change in the affected markets as Elsam and E2 pass from not being controlled by DONG to being solely controlled by DONG.

VII./2 PART A - NATURAL GAS MARKETS

(293) The transaction will result in a significant impediment of effective competition in several natural gas markets. (181)

1. Market for gas storage or flexibility

Preliminary observations

(294) The Commission's investigation has led it to conclude that DONG is already dominant in the Danish market for storage/flexibility.

(295) The market for storage/flexibility is adjacent to the other gas markets in the sense that flexibility is necessary for a gas supplier in order to be able to deliver gas with the necessary degree of flexibility and seasonal profile that their customers demand. The storage facilities are operated by DONG Storage - a legal entity, which is separate from DONG Trade.

(296) A customer of DONG Storage is thus also a competitor of DONG Trade. The Commission has assessed the likely future behaviour of DONG Storage in this context, on the assumption that both the storage and trade activities are guided by the joint aim of promoting the profitability of the DONG group as a whole.

(297) On this basis the Commission has found that the acquisition of the main sources of flexible demand in Denmark will give DONG the ability and incentive to simultaneously reduce its own storage needs and the storage available to its competitors. Due to the regulatory regime in Denmark, this will, all other things being equal, result in higher storage tariffs and reduce other gas wholesalers' ability to compete with DONG Trade.

DONG is dominant (on either market definition)

(298) Irrespective of whether it constitutes a separate market, storage constitutes the largest and most important source of flexibility. Many respondents to the Commission's market investigation have indicated that DONG's overall dominance in the Danish gas sector is partly due to its storage monopoly in Denmark. As mentioned in recital (38), DONG controls (182) both Danish dedicated physical storage facilities with a total working capacity of some 700 mcm.

(299) In addition, DONG controls the entire supply flexibility of gas entering the Danish onshore area, due to the flexibility of its contracts with the DUC consortium and with the Syd Arne/Lulita consortia.

(300) Furthermore, DONG controls the upstream pipelines, with the line pack capacity of which allows DONG extra short-term flexibility which it can also sell on the market. (183)

(301) A further, though limited, factor of DONG's flexibility strength on the Danish market is its control over two regional networks in the south of Denmark, providing DONG, some extra flexibility on the downstream/demand-side through the achievable line pack in this network.

(302) Moreover, DONG has developed close business relations with the customer group of decentral CHPs. (184) The limited demand flexibility of this customer group can only exercise a weak constraint on a dominant company in storage/flexibility. However, the fact that DONG has a privileged position in regard to this customer group further limits their potential for exercising constraint. For decentral CHPs and the other large customer customer group (large business customers) with equally limited potential for exercising flexibility constraint the fact that DONG has a high market share on this retail market/these retail markets limits their accessibility of these customer groups by competitors.

(303) DONG is insufficiently constrained by flexible wholesale supply contracts of its competitors. Within Denmark the only current relevant flexible wholesale supply contracts are those that have their ultimate source in DONG, which has therefore determined the terms of these contracts.

(304) On the basis of such a supply contract HNG/MN, for industrial customers, has to purchase the flexibility with DONG. (185) The swap contracts with [...]* and with [...]* had comparatively generous flexibility terms. (186) However, given the limited duration of these swap contracts, DONG could not be constrained by the counterparties of these swaps in Denmark. As to the flexibility that is enshrined in DONG's long-term contracts with Elsam and E2, see below. Imports and turn-arounds, as argued in the wholesale market definition section above, exert little competitive constraint on DONG. (187) This constraint is further lessened by DONG's commercial arrangements with [...]* concerning the delivery/diversion of gas to Dragør.

(305) DONG submits that trade and import, including swaps, play an important role on the flexibility market. The Commission considers that swaps can provide a certain, although only minor, degree of flexibility. However, in comparison with flexible consumption by central CHPs, swaps are less appropriate substitutes to storage as they are only interesting for flexibility seeking companies with gas needs or reserves abroad. Swaps require coordination of (and an agreement on) volumes, timing, places of delivery, values and prices between the swap partners. Against this complex background, no potential swap partner appears to offer the same degree of flexibility as Elsam and E2, which are the largest gas consumers in Denmark and have the possibility to switch fuels and vary the output of their gas-fired central CHPs.

(306) The most important constraint on DONG is the possibility for Elsam and E2 to offer flexibility services. As neither of these companies is currently actively marketing this flexibility as regards short term flexibility (188), however, this constraint cannot, even on the assumption of a wider storage/flexibility market, be sufficient to put into question the finding that DONG is dominant on such a market.

Regulatory and competition law constraints

(307) In regard to storage, DONG's market power is limited by the boundaries of competition law and by the regulatory supervision exercised by the Danish authorities. Just as is the case with the essential parts of the gas infrastructure, the fact that storage facilities are subject to regulatory supervision, is in itself a reflection of the fact that exercise of market power with respect to storage has the potential of hindering competition.

(308) However, the Commission's investigation has shown that the current legal framework and regulatory regime does not prevent all exercise of market power by DONG, nor would it prevent increased storage costs for competitors to DONG after the merger.

(309) In accordance with the Danish Natural Gas Supply Act (Naturgasloven) DONG is obliged to give third parties access to the storage it operates under the negotiated third party access (TPA) regime, which applies to the two DONG-owned storage facilities. DONG has access to the storage facilities on the same terms and conditions as other users. DONG also offers two types of standard storage agreements on terms and conditions that are published on DONG's website. (189) The standard storage agreements provide for a "package concept", which includes volume, injection and withdrawal. They have to be purchased for a period of one year, running from 1 May until 30 April. According to DONG (190), the standard storage agreements have to be signed before the storage year begins, i.e., before 1 May. (191)

(310) General and basic principles governing individually negotiated access are also published on DONG's website. All individual storage agreements concluded under the negotiated TPA regime must be provided to the Danish Energy Regulatory Authority, Energitilsynet, which monitors prices and other contractual conditions.

(311) According to DONG, the storage costs reflect the long-term total costs for the establishment and efficient operation of the storage facilities (192). Tariff setting also takes into account that the tariffs should not significantly exceed prices of comparable storage products in other countries. Storage tariffs for standard packages are public. According to DONG, prices for storage under individually negotiated contracts are in general higher than the tariffs for the standard storage packages because deviations from the terms in the standard storage agreements will usually impair an efficient use of the storage facilities (193).

(312) The market investigation has shown that DONG's ownership of the storage facilities confers upon it competitive advantages over third party storage customers and thus essentially competitors. Even if DONG's claim were to be accepted that current storage tariffs and other terms and conditions for purchase of storage facilities are cost based, objective and non-discriminatory, it seems that their application may result in competitive disadvantages for competitors to DONG.

(313) In both the standard storage agreements and the individually negotiated contracts concluded between DONG Storage and its customers, provision is made for potential amendments to the contracted tariffs. The reservation applies to amendments following directives from the Danish Energy Regulator, as well as 'general tariff amendments' by DONG Storage. The contracts further state that the storage customer is to be notified in writing about any tariff amendment and also that tariff amendments do not entitle the storage customer to re-negotiate the storage agreement. This provision seems to allow DONG Storage to arbitrarily raise the storage tariffs, after the conclusion of the agreement/contract with the customer and its approval by the regulator. (194) However, competitors that are not satisfied with DONG's terms always have the possibility to complain to the regulator.

(314) The established contract period for standard storage agreements (i.e., from 1 May to 30 April), is rather inflexible and could disadvantage companies that want to enter the Danish gas supply markets, or companies that currently are present on these markets but do not purchase storage on a permanent/continuous basis and would not want to enter into a one year storage contract for the period 1 May until 30 April May, but from, for example, 1 October until 30 September, which is the period corresponding to the gas year. In that situation, the customer is forced to conclude a standard storage agreement for a period of 2 years, an individually negotiated contract for a period of one year at a higher tariff or a combination of standard storage agreement for a period of one year and an individually negotiated contract for the remaining months. The three hypotheses lead to higher costs for competitors in comparison with the conclusion of a standard storage agreement, and the period fixed for standard storage agreements make it more challenging to compete for flexible gas supplies, without reserving storage volume prior to concluding the supply contracts.

(315) Another condition that limits the usage of the standard packages is that the customer must sign the standard storage agreement prior to the beginning of the storage year, i.e., before 1 May. If the customer could sign such an agreement at a later stage during the storage year (i.e., if the customer would not have to pay also for the period before it starts to use the storage facilities), the usability of standard storage agreements would certainly increase. (195)

(316) Standard storage agreements correspond to [...]* of the total storage volume reserved for the period 2005 to 2006, and individually negotiated contracts correspond to [...]* of the total storage volume reserved for the period 2005 to 2006 (196). According to the storage contracts for the period 2005-2006, [...]* of the total contracted volume is used by third party competitors and [...]* is used by Energinet.dk. The rest of the contracted storage volume, i.e., [...]*, is occupied by DONG. (197) Of the total volume purchased under standard storage agreements for the period 2005-2006, [...]* is purchased by DONG, and [...]* by third party competitors, and the rest by the TSO Energinet.dk [...]* (198) Of the total volume reserved under individually negotiated contracts, [80- 85%]* are purchased by third party competitors and Energinet.dk, and only [...]* by DONG. [...]* of the total storage volume reserved by third parties is purchased on the basis of individually negotiated storage contracts and only [...]* on the basis of standard packages whereas DONG's own demand is satisfied, as is apparent from the above figures, overwhelmingly through standard storage contracts. From these figures it can be concluded that the terms and conditions of the standard storage agreements seem to suit DONG's storage needs and usage profile in the best way, but that they are less suitable for the needs of DONG's competitors, which to a much greater extent on individually negotiated storage contracts and the higher tariffs provided for by these.

(317) The fact that the established terms and conditions, and in particular the contractual period, of standard storage agreement greatly benefit the business of DONG, therefore seems to constitute a competitive advantage to the company.

(318) A further example of the imperfect deterrent effect of competition law is the fact that a number of customers have stated that DONG does not charge any storage costs to them, although it is obvious that these customers do not have a completely flat demand and thereby are likely to create some storage cost. This in transparent storage pricing makes it very difficult for any competitors to compete with such offers and has in addition, probably constituted a deterrent to the entry of any competitors (e.g. E2 and Elsam) as actual providers on this storage/flexibility market.

(319) The parties argue in the reply to the SO that "[t]he absence of complaints, in relation to the standard contract period as well as gas storage as a whole, strongly suggests that DONG Lager's arrangements are regarded as satisfactory by customers."

(320) However, the parties also refer to a Decision dated September 2003 by DERA regarding complaints launched by Naturgas Fyn and Energi Viborg, who found the standard conditions regarding inter alia injection capacities and compulsory storage during the winter made their operations more costly. They also complained about the lack of flexibility in access to DONG's storage in that the standard period could not be used in connection with transport of gas from the German system (199). DERA did not object to DONG's regime because it found that the conditions were not unfair insofar as the restrictions were justified by the physical and technical restrictions that were imposed by the system. The decision contains no assessment as to whether the system actually had the negative effects alleged by the complainant.

(321) In order to substantiate it claim that it is subject to effective cost-based regulation, DONG submitted, after its reply to the SO, two documents relating to DONG's storage. One document is a DONG Memo dated 17 January 2006 regarding the current economic performance of DONG Storage. The document contains a reference to the other document which is dated 15 February 2002 from DERA entitled "Memo regarding emergency supply on the Danish market for natural gas and DERA's assessment of the fairness of DONG's transmission prices". In the latter document it is concluded that DONG's proposed tariffs for storage in the context of emergency supply (and hence in general) were assessed not to be fair. The memo calculates prices which, according to DERA, would be fair and which would allow DONG Transmission (the legal entity which at that time held the storage facilities) to obtain revenue from its storage activities of 382m DKK in 2001. As shown in the table 7, DERA arrive at the figure 382m DKK on the basis of three components.

Table 7: DERA calculation of 15 February 2002

Component m DKK

- cost of operating the storage 85

- depreciation (1/40 of the investment cost 2.86bn DKK) 72

- return (7.4% of 3.0 bn DKK) 225

382

(322) As is clear from the memo, the cost of building the storage amounted to 2.86bn DKK, and its residual value in 2001, when the assessment was undertaken, was 2.17bn DKK. Yet as shown in table 7 above, the return component is calculated on a basis of 3.0bn DKK of capital for the storage facilities.

(323) The provision in the law regulating the access to all gas infrastructure is not solely based on considerations relating to the promotion of competition, but also allows DONG Storage to obtain a "reasonable return on the capital invested." (200)

(324) On the same day as the memo was written, DERA also published a Decision entitled "Prices for usage of DONG's transmission grid". That Decision contains the method whereby DERA reaches the conclusion that DONG Transmission is allowed to generate a return on an adjusted level of capital which is 40% above the current value of the storage facilities and even above the value of building the new storage. In the years prior to 2001, DONG Naturgas had not paid its owner, the Government, sufficient dividends to provide the latter with an adequate return on its investment. In broad terms the Decision takes as its basis that the Net Present Value ("NPV") of the investment by the Danish government exceeds the NPV of the dividends paid out by DONG to the Government in the same period by 7.1 bn DKK (201). This "deficit" (shortfall of dividends) is then allocated to the different components of DONG's infrastructure (transmission, storage and off-shore pipes) in the sense that the future return on each type of infrastructure should be sufficient to not only generate a return on their current value, but in addition generate a return to compensate for the historical lack of dividends.

(325) The consequence of this method was that the three types of infrastructure: transmission, storage and off-shore pipelines were allowed to generate revenues that in themselves would ensure that all the invested capital in DONG Naturgas would achieve a reasonable return. As is clear from DONG's annual reports and as stated in the report by Rothschild prepared for the Ministry of Finance, "the overwhelmingly largest contributors to EBITDA in 2003 were E&P and Naturgas (gas wholesale and offshore pipelines)". These statements relate to the period when the storage and transmission activities had already been separated from the wholesale activities. So while the burden of generating a reasonable return on all of the historic investments in DONG Naturgas had been put on the infrastructure, significant subsequent revenues were achieved with the wholesale activities. This raises some doubts as to whether cost based regulation solely based on a view to promote competition would also have reached the conclusion that it was incumbent on future customers of DONG Storage to generate a return on capital significantly in excess of the actual value of the storage facility.

(326) The legality of DERA's approach was subsequently tested because DONG appealed the Decision taken on that day with regard to the transmission system. The Energy Board of Appeal (Energiklagenævnet) upheld that Article 38 of the Natural Gas Supply Act had been correctly applied by DERA when it allowed the different components of the infrastructure to generate a return on the basis of the adjusted level of capital. But it also held that DERA's approach, as also set out in the memo presented by DONG to the Commission, was too intrusive compared to the provisions in the law whereby it is incumbent on DERA only to supervise the fairness of the prices, not to directly regulate them. In order to induce a change in the prices it was thus incumbent upon the DERA to substantiate that DONG's prices fell outside the scope of what could be considered fair, and that had not been done. The Energy Board of Appeal thus sent the Decision back to DERA for a renewed assessment.

(327) Due to subsequent changes to the law, there is now a distinction between access to the storage, which is still regulated according to Article 38 of the Natural Gas Supply Act and access to distribution and transmission networks, which are now regulated by a new Article 37. This new Article 37 requires DERA to carry out ex ante control of the methods applied for cost based revenue ceilings from the regulated activities.

(328) Apart from the memo of 15 February 2002, which rests on principles that were subsequent deemed unlawful, DONG has not provided any evidence that DERA is actively assessing the fairness of the prices for storage. On 29 September 2003, DERA issued a decision regarding a number of complaints raised regarding access to DONG's storage facilities, in which it states that for the purpose of the decision it "does not opine on the fairness of the tariffs. This subject will be dealt with separately" (202) According to information provided by DERA to the Commission in the context of this investigation, no such assessment has so far been carried out (203).

(329) In the DONG memo submitted to the Commission dated 17 January 2006, it is stated that "[i]n 2002 DERA evaluated the tariffs of DONG Lager as part of a decision on the transmission tariffs in Denmark.... In the decision DERA found that a fair annual revenue of DONG Lager amounts to DKK 382 m." According to the memo, DONG Storage's revenue was broadly below that level in the years 2003-2005.

(330) According to the method applied by DERA in the mentioned memo (cf. recital (321)), the allowed revenue was calculated in part on the basis of allowing a fair return on the remaining capital (forrentningsgrundlag). It is apparent from the method applied that if it were to be repeated for subsequent years it would lead to steadily decreasing annual revenues as the remaining capital would shrink. Yet, DONG in its letter maintains that what is relevant is the annual amount of 380m DKK as such.

(331) This should also be compared with what happened following the decision by the Energy Board of Appeal to send the decision regarding the transmission system back to DERA. DERA's subsequent decision relates to the transmission system prices, for the entire period of 2001-2004 (204). As requested by the Energy Board of Appeal, DERA generated an interval of allowed rates of return including a maximum rate based on a more generous assumption than in the initial decision. In addition, one issue of particular importance is how to take account of the revenues generated by the DONG group since 2000. In contrast to the early years of DONG's existence the late period had generated significant surpluses as well as dividend payments to the government. Since the initial calculation from the 2002 Decision allowed the infrastructure to generate return on an extended capital basis (called the regulatory equity) to allow for lack of historic dividends, it was natural to subsequently reduce that basis, once additional dividends accrued. In the context of the transmission grid, DERA found that this would reduce the regulatory equity from 3.348m DKK in 2001 to 1.904m DKK in 2004- a reduction of 43%. (205) On the basis of such a reduction DERA's calculations showed that even though the revenue had been declining in the period, DONG had still generated a return on this (reduced) capital base which exceeded even the maximum allowed return. Since the transmission network had been unbundled from DONG and transferred to Gastra (now energinet.dk) in the period in question, Gastra disputed that future changes in tariffs in this way should depend on decisions by DONG to pay out dividends to its shareholder. Inter alia on this basis DERA decided not to conclude that the prices had been unfair.

(332) Furthermore, it is apparent from the memo, that the basis for a fair return was a risk free interest rate of 5.6% This interest rate was in part calculated on the basis of the 10 year bond rate which at the time of the decision was 5.1% (206). When DERA issued its decision regarding revenue caps for the distribution networks, that interest rate had fallen to 3.36% (207) This decline in isolation would result in a reduction of the revenue ceiling by approximately 53m DKK. (208)

(333) Finally, with regards to the costs related to operating the distribution networks which are under ex ante regulation (cf. recital (327)) DERA imposes on DONG distribution an annual increase in productivity in 2006 of 1.5% (209) Were similar requirements imposed on DONG Storage the revenue ceilings would further decline each year. It should be noted, though, that a correction in the opposite direction would be appropriate if input prices were to rise.

(334) On the basis of the above analysis it is concluded that:

a) the regulatory regime with respect to storage is significantly weaker than that pertaining to DONG's distribution network which is subject to ex ante control;

b) DONG thus enjoys a certain degree of discretion as to the revenues it generates from its storage facilities. In particular, it has not found it necessary to decrease the revenue even though the interest rate has fallen significantly;

c) Notwithstanding the relatively weak level of legal scrutiny regarding storage tariffs, the law does allow DONG to obtain a revenue from its storage activities which covers more than what is necessary to generate a fair return on the historic costs related to building the storage. While past financial difficulties of DONG prior to 2000 in accordance with the law to justify increases in future storage revenue ceilings, the subsequent positive economic performance of DONG (inter alia due to higher oil prices) does not appear to have resulted in a similar decrease of that ceiling.

(335) In view of these facts, it is concluded that DONG is dominant on the Danish market for storage or, alternatively, the market for flexibility.

The Swedish storage or flexibility market

(336) In Sweden, DONG has a dominant position on this market. As mentioned above there is currently hardly any effective physical (210) storage capacity or virtual storage capacity in Sweden. Only Swedish demand modulation of industrial customers and - in, the future, demand modulation of Swedish central CHPs - may provide a flexibility capacity.

(337) The other relevant source of flexibility for Sweden is therefore currently DONG's Danish storage capacity and flexibility provisions in the wholesale supply contracts. However, as shown below, DONG has a market share of [45-55%]* or above on a Swedish wholesale market. Even if some of DONG's competitors' wholesale contracts are also quite flexible (e.g. [...]* current supply contract with DONG ex Dragør), this is unlikely as such to put in question the existence of DONG's current dominance also in Sweden.

Raising storage (or flexibility) costs and raising rivals' costs

(338) One of the competition problems raised by the proposed operation is that the concentration eliminates the most important independent source of alternative flexibility to DONG's storage facilities. On the possibility of a narrow storage-only market the alternative flexibility provided by Elsam and E2 would not belong to the same product market. However, even in the case of a narrow market the operation eliminates the closest independent source of flexibility to DONG's storage facilities. By consequence, customers will be more dependent on using DONG's storage, thereby strengthening DONG's dominance on such a market.

(339) In addition, and irrespective of the exact market definition, the acquisition of Elsam and E2 will lead to higher prices on storage thereby strengthening DONG's dominance because DONG will have the ability and incentive to raise rivals' storage costs. This is because the Commission's economic assessment has shown that the merger will give the merged entity an increased ability to raise rivals' costs, by making it more difficult and expensive to obtain adequate flexibility to fully compete with DONG on the Danish market. In particular since DONG is a dominant incumbent with the main part of its profit accruing from wholesale activities, such a strategy is likely to be profitable. From this it follows that DONG is also likely to have the incentive to raise rivals' costs, so that effective competition will be significantly impeded. This effect will be further elaborated in the following sections.

Ability to use E2 and Elsam flexibility

(340) As stated above in the section dealing with the relevant product market for gas storage or gas flexibility, the flexible consumption by central CHPs constitutes a very important flexibility source. The Commission's investigation has clearly established that the transaction will give DONG the ability to use the central CHP plants owned by Elsam and Energi E2 for flexibility purposes and thus (at least partly) replace its use of the storage facilities.

(341) [Contains information on internal documents concerning E2's ability to provide flexibility]* (211).

(342) [Contains information on internal documents concerning E2's ability to provide flexibility]*:

(a) [Contains information on internal documents concerning E2's ability to provide flexibility]* (212).

(b) [Contains information on internal documents concerning E2's ability to provide flexibility]*.

(c) [Contains information on internal documents concerning E2's ability to provide flexibility]*.

(343) [Contains information on internal documents concerning E2's ability to provide flexibility]* (213).

(344) [Contains information on internal documents concerning E2's ability to provide flexibility]*.

(345) [Contains information on internal documents concerning E2's ability to provide flexibility]*.

(346) Though one competitor stated that the use of power production as a provider of flexibility "would work" [...]*, DONG maintains in its reply to the SO that Elsam and E2 would offer the "wrong" type of flexibility, since they have limited ability to give up gas during the winter when demand for electricity and heat consumption is particularly high. It thus states that E2 or Elsam, if any, could only offer short-term flexibility.

(347) The Commission cannot accept this argument, not only because it clearly contradicts the findings [relating to the internal documents concerning E2's ability to provide flexibility]* because of the Commission's findings below (Cf. recital (354) in regard to E2's historic behaviour.

(348) A visual interpretation of the fuel consumption of Avedøreværket and HC Ørstedsværket outlined in Figure 4 and Figure 5 below also shows that while the latter does not use gas countercyclically, that is not true for Avedøreværket. The latter has indeed a consumption pattern whereby it has relied relatively more on gas during the summer and relatively more on oil during the winter.

Figure 4:[...]*

Figure 5: [...]*

Ability to raise prices in the light of the regulatory regime

(349) The fact that DONG could benefit from integrating with Elsam and E2, by making use of their flexibility is not in itself a competition concern. The harm to competition arises because the decrease in its own needs is likely to allow DONG to increase storage tariffs to third parties.

(350) As mentioned above in recital (330), according to DONG's own submission, it is entitled to an annual revenue of approximately 380m DKK to cover its costs from its storage operations. As is not disputed by DONG in its reply to the SO, a decrease in demand for storage by DONG would leave a smaller base of demand over which such cost could be spread thus leading to increased tariffs without DERA having any legal basis for objecting.

(351) DONG argues in its reply to the SO that the Commission has failed to take into account that DONG's storage prices "may not, both pre-merger and post-merger, significantly exceed the price level in comparable markets" (214). The Commission recalls in this respect that DONG's annual report for 2004 states that "Tariffs for access to the storages are among the lowest in North- and Central Europe" (215) From the same publication it is apparent that while the prices in Denmark are 0.36/0.80 DKK/m3 for large and small withdrawal capacity respectively, in other countries similar prices are as high as 0.87/1.59 (216) DKK/m3. It is thus apparent that even if prices were to double after the merger, they would not significantly exceed those in comparable markets.

(352) Furthermore, the effect is not limited to increased tariffs but will be further aggravated by the fact that the transaction also removes Elsam and E2 as a potential alternative source of flexibility to third parties.

(353) The fact that Elsam and E2 could provide an alternative source of flexibility to DONG as demonstrated above (Cf. paragraphs (340)-(348)), also means that they could provide such an alternative to third parties. This is, inter alia, borne out by the statement of SGA, which noted that having E2 or Elsam as a customer could be "very useful because the costs for storage will be reduced dramatically if power plants use gas in the summer and are interrupted in winter" (SGA).

(354) The Commission's evaluation of historic contracts also reveals that E2 were already, prior to the merger, providing this kind of seasonal flexibility to third parties. [Contains internal information on E2's ability to provide seasonal flexibility]*.

(355) Even when the off-take is not directly counter-cyclical, it is apparent from an examination of the entire E2 portfolio of gas contracts in the period 2002-2005(1H) that the flexibility of its central CHPs has allowed E2 to acquire gas even with very limited flexibility provisions, which predominantly has benefited the suppliers to E2 other than DONG. All E2 contracts can essentially be grouped into three categories:

(a) E2 has full flexibility: These contracts have no minimum daily or annual take-off nor are there any take-or-pay obligations. Apart from one contract with [...]*, the supplier is always DONG. Delivery is always at the gate of the non-flexible CHPs such as Ringsted and Hundested.

(b) E2 has some flexibility: These contracts have annual minimum (take or pay) quantities of 80-90%, but no daily minimum quantity. The two contracts in this category are the long term contracts supplied by DONG with delivery foreseen at the gate of the large central CHPs (H.C. Ørstedsværket, Svanemølleværket and Avedøreværket).

(c) E2 has (almost) no flexibility: These contracts have annual minimum quantities (take or pay quantities) of at least [...]*, but in addition daily minimum quantities of [...]* and often [...]* or more. Except for the contract with [...]* mentioned above, all contracts with other suppliers than DONG fall into this category. They are usually delivered at either Emden, Ellund or GTF.

(356) The reason why E2's, and in particular Elsam's, activities as suppliers of flexibility have not reached their full potential historically is likely due to the nature of the long term contracts with DONG. The flexibility available to Elsam and E2 in these contracts has reduced the incentive to actively optimise the gas consumption with a view to the general flexibility needs in the market.

(357) In view of the expiry of E2's and Elsam's long-term supply agreements with DONG in [...]* and the likely development of a liquid Danish trading market (in the absence of the proposed transaction), it is expected that, in the future, E2 and Elsam would conclude more similar transactions. Therefore the proposed concentration is not only likely to remove the already existing potential competitive constraint on DONG's storage facilities but it will have an even stronger effect in the future, as E2 is likely to expand its balancing activities and Elsam has the potential to act in a similar way as E2.

(358) Even prior to the expiry of these contracts, it is not impossible that the increased liberalisation would lead in particular E2 to seek to optimise its own consumption of gas and then effectively resell some of the flexibility available in its DONG contracts. DONG has brought to the Commission's attention that the contractual delivery point for this gas is at the gate of the power plants. DONG argues that even though the long term contracts do not contain any resell restrictions, it would not be technically possible to resell the gas because the power plants do not constitute entry-points at the GTF.

(359) In this context the Commission notes that in the light of DONG's dominant position it might be unlawful if DONG were to reject a request from E2 to deliver the gas at the GTF rather than at the gate. This is due to the fact that such an arrangement might simply save DONG the exit fee. A rejection could further lack justification in the light of the fact that DONG is already granting Elsam and E2 significant flexibility to swap the gas among them and among the individual production sites. However, with regard to the necessity for speed which characterises the general scheme of the Merger Regulation, the question whether any future refusal, by DONG, to change contractually agreed delivery points would be compatible with Community law, cannot be decided in this decision. Such findings in this case could not be made without undertaking a detailed investigation into the matter.

(360) Even if DONG were to reject such a request, an alternative method to get the gas back into the GTF would be for E2 to persuade Energinet.dk to reconfigure GTF so as to make additional entry points at the gate of the power plants. Energinet.dk has confirmed to the Commission that, although such a change would require some (IT related) investments, a request could have been accommodated in the absence of the merger had it been made a priority and that if E2 had agreed to become a market maker at GTF this would have further increased the likelihood that it would have been given high priority (217).

Incentive to raise rivals' cost

(361) The Commission also concludes that DONG will have the incentive to raise rivals' cost of storage (or of storage/flexibility) post merger. The incentive to raise rivals cost is the inherent assumption underlying the need to regulate prices of storage in the first place. To the extent that the current regulatory regime has any downward effect on DONG's tariffs, it follows that DONG has an incentive to seek to raise prices.

(362) If DONG lowered its own demand for storage after the merger and in turn increased its tariffs in order to maintain the same overall revenue for DONG Storage from the remaining sales, it would have a direct positive effect on the overall revenues of the DONG group (because it would replace in-house payment for storage with outside payment of equivalent amounts). This would also be the case even if the strategy would reduce the overall demand for storage from third parties. This follows from the fact that less flexibility available to competitors would diminish the competitive pressure on DONG's gas sales activities thereby allowing DONG to recoup losses in storage revenues through higher margins in gas sales.

(363) It is relevant in this regard to recall the statement by Rothschild mentioned above (Cf. recital (325)) that "the overwhelmingly largest contributors to EBITDA in 2003 were E&P and Naturgas (gas wholesale and offshore pipelines)". While the revenues from storage contribute only 5% of the DONG groups EBITDA, its gas-trading activities contribute 39% (E&P contributes 36%) (218).

(364) On this basis, the Commission concludes that DONG does indeed have the incentive to raise rivals' costs.

Effect on the market

(365) The decrease in internal needs for storage is likely to have a significant effect on the storage prices. [Contains information on possible decrease of storage need of DONG due to the acquisition of gas-fired power plant assets]*.

(366) The higher storage tariffs will affect the storage customers' input costs, and thus their supply prices downstream since they would have to pass through the higher storage costs to the end customer. This would lead to a situation of input foreclosure for DONG's competitors on such markets, having an effect on the wholesale markets and on the gas supply markets by benefiting DONG's wholesale and downstream positions. The market investigation shows that storage costs are not an insignificant part of the gas supply price; for some customers, with a very flat consumption profile, they can be rather low. However, the market investigation has brought indications that they are often around 5% and can go up to 10%, of the gas supply price to the end customer. Also, storage costs can be an important a cost factor on the wholesale level.

(367) The Commission notes in this regard that the divestiture to Vattenfall of certain power plants of Elsam and Energi E2 will not replace the flexibility that is removed from the market. The acquired plants are less suitable for flexible gas consumption than the central ones that will remain with Elsam and Energi E2, which will be controlled by DONG following the notified transaction.

(368) DONG argues that all the current customers of DONG Storage would be able to turn to alternative sources of flexibility. It refers, inter alia, to the fact that one significant customer, Energinet.dk does not need seasonal flexibility and that another main customer, [...]*, could satisfy its flexibility needs through the swap agreement with DONG.

(369) With respect to Energinet.dk the price increase is unlikely to be limited to a certain group of customers. It is thus not relevant whether the customer needs storage for one or the other purpose. It is relevant in this regard to note Energinet.dk's own assessment of the merger:

"Seen from the gas sector alone - the merger results in one group of companies controlling all commercial flexibility tools available on the market: i) production, ii) transmission options for Netherlands and/or Denmark, iii) line pack in offshore transmission, iv) two relatively large storages and v) the largest endusers of gas being able to substitute to/from gas or simply shut down"

(370) With respect to [...]*, it suffices to note, that entering into swap agreements with DONG would not remove the dependence on DONG as supplier of flexibility. The flexibility available in this context is equally controlled by DONG.

(371) With respect to [...]*, DONG argues that they could replace the small volumes of flexibility they purchase with flexibility from the long-term supply agreements with DONG in Ellund. In this context reference is made to the detailed analysis in the market definition section of this decision, regarding the difficulties and limits to this alternative. In addition, it is very unlikely that competitors would buy storage from their competitor unless they deemed such purchases necessary. The fact that the current volumes purchased by [...]* are small, combined with the fact that turn around volumes do not appear to show a seasonal pattern, illustrates in that these competitors currently are not fully competing for customers with high flexibility needs. After the merger such a strategy would be further impeded by the increased storage prices.

(372) Finally, the issue is not only the effect on current storage customers, but also the effect on potential future customers. The increased difficulty of obtaining flexibility independently of DONG makes it less likely that new entrants would find it worthwhile to enter and compete with DONG.

Conclusion on storage (or flexibility) for Denmark

(373) The transaction is thus likely to confer on DONG the ability and incentive to increase storage tariffs in Denmark and thereby lead to the raising of rivals' costs. It will also lead to the elimination of the most important independent source of flexibility. For these reasons, the operation would lead to a significant impediment to effective competition on the possible Danish market for storage or on the possible Danish market for storage/flexibility in particular through the strengthening of DONG's dominant position on such markets.

(374) In addition, the notified transaction will further contribute to the significant impediment to effective competition, in particular through the strengthening of DONG's dominance, on the wholesale markets for Denmark and for Sweden (see recitals (453)-(532)). Furthermore, and for the same reasons, the conclusions arrived at in other parts of this decision as to the creation of a significant impediment to effective competition, in particular through the strengthening of DONG's dominance, on the markets for the supply of natural gas to large industrial customers and decentralised CHPs (see recitals (544)-(605)) as well as on the market or markets for supply to small business customers and for supply to households (see recitals (630)- (662)) are reinforced.

Sweden

(375) Since the need for flexibility for suppliers on the Swedish market must be provided from Denmark, the conclusion above (Cf. recital (373)-(374) applies independently of whether the market is Danish (as is the case for Denmark) or (as is possibly the case for Sweden) is either Swedish or Danish-Swedish in scope.

(376) However, as mentioned above in recital (198), new central power plant capacity is coming on-stream in 2007 and, possibly in 2009. The owners of these plants, Göteborg Energi and E.ON will likely be able to exercise some competitive constraint on DONG's flexibility/storage capacity. Consequently it may be argued that some flexibility may be available in Sweden within a reasonable time-horizon. However, it is not likely that this will suffice to remove the need for Danish flexibility even for the purpose of serving the Swedish market.

(377) The parties have argued that the Commission's analysis in the SO of the Swedish market has not presented any arguments in detail and is thus insufficient to satisfy the necessary burden of proof.

(378) Due to the fact that Sweden sources exclusively from Denmark, the finding of potential harm to competition in Denmark, may in principle be extended to also include Sweden. Nevertheless, a detailed assessment of the particular features of the Swedish market is not necessary, since the overall assessment of the transaction is independent of this conclusion. The remedies proposed by the parties are necessary to solve the competition problem identified in the Danish storage/flexibility market and sufficient to remove any harm which could arise in Sweden as a consequence of the notified transaction.

2. Markets for wholesale supply of gas for Denmark and for Sweden The relevant temporal framework for the analysis

(379) In its reply to the SO DONG has questioned the relevance of looking beyond a [...]* year horizon and has stated that there has not been any thorough examination in the Commission's statement of objections of the post-[...]* competitive landscape in connection with the termination of the long-term contracts DONG has with E2 and Elsam. Such an analysis should include the production decline after [...]* in the Danish continental shelf and the expected effects from the Northern European Gas Pipeline ("NEGP") (219) and the Baltic Gas Interconnector.

(380) The specific nature of natural gas markets justifies an analysis that includes a time horizon beyond the year [...]*. It is inherent to merger control that the analysis has to take full account of the particularities of the markets concerned. (220) The specificities of the gas sector are founded in the market characteristics, inter alia, of long-term investments in infrastructure and long-term upstream (and downstream) supply contracts. (221) It can be acknowledged, however, that the more remote in time future events are, the more important it is to examine their likelihood.

(381) In this context it is, first, important to note that the termination of DONG's large long-term supply contracts with Elsam and E2 in [...]* is the most likely assumption.

[...]*.

(382) Second, it is relevant to examine the alleged production decline of the Danish North Sea fields. According to the Danish Energy Authority Danish gas reserves amount to approximately 132 bcm. [...]* (222). According to a report commissioned by the Danish government (223), Denmark will be, self-sufficient with respect to gas supplies at least until 2015 (on the basis of a constant or slightly increasing demand of 4-4.5 bcm). (224) According to its own submission, DONG is entitled to get the full production of the Danish gas fields, minus the gas used on the platforms. In the light of these facts, the Commission concludes that the competitive landscape relating to DONG's DUC gas is unlikely to change significantly before 2012-2015 and therefore should not influence an analysis of the competitive effects that examines the period until 2012.

(383) Third, the NEGP, if realised, will mean a new channel of Russian gas into the Community, and a way of satisfying growing demand in the coming years. According to current plans (225), construction is expected to be completed in December 2010 and the planned capacity is 27.5 bcm. According to the German Government the main purpose of NEGP is to satisfy increased gas demand in Germany, among others stemming from the construction of several gas-fired power plants close to the pipeline landing point. The fact that the project is financed by Gazprom, E.ON and Wingas, makes it plausible that they will focus sales where they have their main activities. However, assuming that DONG's assumption of a significant amount of "liquid" wholesale gas in Northern Germany as from 2011 is correct, such gas (i) would not influence the Danish import situation between 2009 and 2011 (nor of course earlier), (ii) would make it, as stated in the SO, more likely that the BGI is built (see below recital (384)) and (iii) would make it more likely that potential entrants into the Danish wholesale and retail markets (such as Elsam, E2 and KE) would have independent access to wholesale gas. It is therefore highly likely that the consideration of the NEGP, if assumed to be built according to plan, would not constitute a factor alleviating any concerns raised by the notified transaction, as DONG's acquisition of E2 (including its partnership rights in the BGI) would preempt the challenge that would be posed to DONG's dominance by the BGI itself and by Elsam's, E2's and KE's increased possibilities to source wholesale gas independently of DONG. These effects would be merger-specific and would be likely to outweigh any (non-merger specific) beneficial effect that the access to gas could have on DONG's other competitors (for which consideration also the possibility that DONG itself obtains access to this gas should not be discounted).

(384) Fourth, the construction of the Baltic Gas Interconnector (BGI) linking Germany and Denmark/Sweden is possible before 2009, while the likelihood increases when considering a later construction date. E.ON and E2 are the main shareholders of the project and permits have been or will soon be obtained. No decision has yet been taken as to the start of construction but construction could take place within short timeframe. The mere fact of the pipeline project has already had some disciplining effect on the dominant player on the Danish market, but this effect diminishes significantly as E2 is eliminated. Furthermore E.ON's (the other major pipeline consortium partner's) focus is clearly more on the Swedish than on the Danish market. Development of the branch of the pipeline connecting to Denmark (226) or even, as voiced in the Commission's market investigation, construction of the pipeline as such may be dependent on whether the merger goes through, which means a clear difference between the scenario of an independent E2 as driving project partner or the alternative scenario of DONG taking over this role. As to the BGI, therefore, the likelihood that it would exert competitive pressure on DONG (even before 2009 but more likely after 2009) is next to eliminated by the proposed merger.

(385) In any event, as will be argued further below and as has partly already become apparent in the above discussion, even in the absence of any consideration of a time horizon post [...]* the merger will have significant anti-competitive effects on the Danish wholesale market, inter alia, through elimination of the constraining effect emanating from the BGI project and through the fact that negotiation of contracts for supply of Elsam and E2's CHPs will need to begin in [...]* at the latest and have an immediate impact on entry decisions of potential competitors to DONG (and on the commercial rationale of existing competitors to remain in the market).

DONG is dominant on the Danish market

(386) There are strong indications that DONG is currently dominant on the Danish natural gas wholesale market with a market share that was stated in the SO as [80-90%]* in 2004 (and that would by the same token be [80-90%]* in the first half of 2005) (227). For reasons of confidentiality other companies' market shares cannot be stated but in 2004 other market players were DONG's swap partners [...]* and [...]* and [...]* and [...]*.

(387) In Table 8 to Table 12, compiled by the Commission on the basis of figures largely provided by the parties but complemented by some aggregated third-party figures based on Energinet.dk data, the Commission has varied a number of input parameter, to account for different possibilities of calculating a wholesale market share. The result in all of these approaches is that DONG's wholesale market share in Denmark (but also in Denmark and Sweden) is very high. It is in any event above [75-85%]* and in some calculations above [80-90%]*. To make the aggregated third party data less recognisable to DONG, they have been stated as "less than" figures. In this approach market shares were calculated taking into consideration the highest possible figure (obviously somewhat overstating third party market shares).

<emplacement tableau>

(388) When considering the total gas supply - both physical and contractual, and defining swaps as wholesale transactions attributable to DONG as they need the consent of DONG - to Sweden and Denmark together (see Table 8), the market share of DONG is [85-95%]* in 2004. (228) Only considering the Danish market under the same assumptions (Table 9), the figure shows [85-95%]* for the same period. [...]*.

(389) Even if these figures were subjected to some further adjustment, for example, assuming that long-term sales to Elsam and E2 at the gate are not to be counted as wholesales (but rather as retail sales to central CHPs) (Table 11), DONG's wholesale share in Denmark would still amount to [75-85%]*.

(390) Finally also on the hypothetical basis (which is in this case not considered appropriate by the Commission) of a limitation of the perspective on that segment of the wholesale market that does not constitute captive intra-group sales (Table 12), DONG's market/segment share is high [85-95%]*.

(391) Thus, it is obvious that DONG has a very high market share in Denmark irrespective of the angle from which the wholesale sector/market is regarded. This market share has remained on this very high level since the first opening of the Danish wholesale and large customer market in 2000 from which point on DONG could be challenged by competitors. This, in itself is a clear indication of DONG's (very entrenched) dominance on the Danish wholesale market.

(392) The foundation of such dominance, apart from DONG's incumbency advantage as historically dominant quasi monopolistic company, is that all gas sold in Denmark is physically imported by DONG. This upstream control of gas gives DONG a clear price advantage over all other market players. Due to DONG's position as the first level/originator supplier of gas which - regarding particular volumes - only DONG can purchase and has so far purchased (229) from the Danish gas production fields. All competitors to DONG therefore have a structural competitive disadvantage compared to DONG with regard to price. In the Commission's market investigation, a very large majority of respondents have therefore strongly expressed their view that DONG is dominant in the Danish gas sector as a whole. As further elements, DONG's dominant position on Danish retail markets (discussed further below as from recital (533)) and its storage dominance also further strengthen its wholesale position.

(393) In its reply to the SO, DONG argues that the price advantage that DONG has over other market players in accounting terms is irrelevant, but instead the concept of opportunity cost should be used. This argument has already been addressed in the discussion of the geographic market. The Commission agrees that the opportunity cost principle would be applicable if no limitations as to transport, customer acquisition, etc. were present, (i.e. if gas flowed without obstacles to the point with higher prices). However, the market investigation and DONG's own submissions have shown that this is not at all the case and the alternative of selling to Germany (or to the Netherlands) is only present to a limited degree (230). Under these circumstances, the Commission therefore maintains its view that DONG has an overall advantage for the sale of gas in Denmark.

(394) In spite of DONG's strong prima facie indications of dominance (231), the weight of actual and potential competitive constraints upon DONG (which may exist even in a situation of dominance) will be examined and assessed whether any of these constraints is already currently so strong, or will be so strong in the foreseeable future, that DONG's dominance could be put in question (Cf. recitals (395)-(447)).

(395) Potentially, competitive constraints for DONG could come from five sources:

(a) from the Danish offshore area,

(b) from gas imported from Germany south of the DEUDAN pipeline,

(c) from gas turned around at Ellund,

(d) from a liquid Danish wholesale market,

(e) from new pipeline capacity or other import facilities.

(396) As will be shown below these competitive constraints are weak (constraints (a) to

(d)) or not sufficiently certain in the short term to constitute an effective constraint (constraint (e)).

- Weak constraints from Danish offshore area

(397) The notifying party has argued that it is constrained by the possibility of sales by the members of the Danish offshore consortium DUC (A. P. Møller Maersk, Shell and Chevron). Other sources of potential constraint are be the non-DONG part of the Syd Arne consortium (Amerada Hess, DONG, Denerco and Danoil) and imports from the Netherlands via the NOGAT pipeline, the pipeline linking NOGAT to Tyra and DONG's pipeline from Tyra to the Danish on-shore area at Nybro.

(398) The DUC consortium is the source of the largest volume of gas production in the Danish part of the North Sea. The three companies Shell, A.P. Møller - Maersk and Chevron sell gas to DONG via three long-term take-or-pay agreements dating from 1979, 1990 and 1993. After the settlement between the Commission and the DUC parties as well as DONG in 2003 (232), the DUC Parties have agreed to market the gas at their offshore platform Tyra independently to DONG and to also sell 7 bcm of additional gas to companies other than DONG over a five year period. The companies of the DUC and DONG have together constructed and, in July 2004, put into service the pipeline segment connecting the Tyra platform with the NOGAT system that enables all parties to export the gas they have available at Tyra to Den Helder in the Netherlands.

(399) Total Danish production in 2004 was 8.3 bcm and 7.6 bcm was sold to DONG. DONG has stated that no reservations have been made to transport gas from Tyra to Nybro via its offshore pipeline consequently no gas has been transported by any other party than DONG (233). This means that in 2004 and 2005 DUC parties have sold all their additional gas volumes to the Netherlands. The Commission has no indication that there is a reasonable possibility for that to change during the remaining period of the commitment given by the DUC parties to the Commission. The market investigation also supports the argument of the free DUC gas being sold to the Netherlands in the response of Shell: "As part of the settlement with DG COMP in 2003 Shell has contracted out a total volume of [...] from the DUC operations to The Netherlands", "SOGU BV (234) has booked long-term transmission capacity in the NOGAT pipeline equivalent to its current and expected supply commitments to the Dutch market". (235)

(400) The notifying party in its reply to the SO argues that the Commission "attributes disproportionate weight to the fact that currently all of those (DUC) sales are going to the Netherlands", and that the relevant question is what would happen in the event of a price rise in Denmark. The Parties also provide information about an approach by [...] (236) concerning capacity reservations in the upstream Tyra-Nybro pipeline.

(401) The Commission contends that even in the application of the SSNIP test concerning the future price developments, it is unlikely that the Danish producers will mean anything more than a weak constraint to DONG's dominant position, as they have invested substantial amounts in the pipeline to the NOGAT platform, the usage of which they would be unlikely to forfeit, and on a similar logic, they have made their reservations in the NOGAT pipeline, which they are by consequence likely to make use of. In the highly unlikely event of contractual northward flows on the NOGAT, which faces a number of technical, administrative and cost obstacles, or in the equally unlikely event of a change in the delivery point for the contracts in place, it is likely that the terms of such a contract would have to factor in any opportunity cost on the pipelines, potentially resulting in a significant price increase for the gas.

(402) [Contains information on third party]*.

(403) DONG has an exclusive life-of-field contract with the Syd Arne group (Amerada Hess, DONG and others) which provide it with around [...]* of the gas portfolio or [...]* (237) per year. The Syd Arne production site is an associated gas production site so gas is essentially only the by-product of the primary oil production. According to the contractual arrangements, the production agreement cannot be terminated, so DONG is indeed the only company that has had, has and will have access to the gas from those fields as long as the production persists. [...]*.

(404) As mentioned earlier (Cf. recital (30)), DONG is the sole owner of the Tyra-Nybro pipeline as well as the pipeline connection between the Syd Arne and the Tyra production platforms. There is a negotiated TPA regime on the pipeline but the pipeline is directly operated by DONG, thus, if there were any reservations made on it, DONG's supply operation would know about these along with the nominations made by any potential competitor in shipping gas from the production area to the shore at Nybro.

(405) In its reply to the SO, DONG expresses its belief that DONG's simple knowledge of reservation in the pipeline does not provide it with the ability to prevent such reservations. DONG also underlines the published tariffs and third-party access regime in place.

(406) In the Commission's view, simple knowledge - which does not necessarily mean the direct prevention of third-party reservations - can confer sufficient competitive advantage to DONG and may also lead to indirect prevention of access if DONG raises its own reservations. Furthermore the mere fact that the pipeline is governed by negotiated TPA - and not legally unbundled from DONG's commercial operations -, does not discount the possibility of DONG taking advantage of being the owner of the pipeline and thereby indirectly preventing access even if there is some regulatory supervision. Third party access has not been tested yet and has not proven to function as there has not been one single reservation since initiation of the regime.

(407) Furthermore, it must be noted that the market investigation has shown that it is currently - for at least technical reasons - not possible to export gas from the Dutch end of the pipeline via the offshore networks towards Denmark (238). The reason is that the NOGAT is a two phase solid and liquid system while the Tyra-F3, and the Tyra- Nybro pipelines are single phase pipelines. Therefore investments would need to be made in order to accommodate a flow in the reverse direction of the current north south flow from Tyra to Den Helder. (239) As the offshore pipelines towards the Netherlands have been designed for a north-south flow and additionally the NOGAT system and the Tyra systems are not compatible if not flowing in a southward direction, no investment with the opposite aim seems likely, also taking into account the additional transportation cost incurred.

(408) In its reply to the SO, DONG points out that in order to create a constraint, there is no need for any physical flow of gas to take place, instead this can be done on a contractual basis.

(409) As already explained recital (401), the contractual flow would create significant opportunity costs that are unlikely to make the sale of gas viable even in the case of a SSNIP. Additionally, contractual flow is overwhelmingly interruptible flow, which would result in significantly lower security of supply and, as in this case, the counterflow even involves two pipelines (NOGAT-Den Helder; Tyra-F3), it results in significant additional commercial risks, likely necessitating the back-up of any such operations with significant additional storage within Denmark.

(410) Moreover, there is only 1 mcm/day capacity on the Tyra-Nybro pipeline240 and in addition, the negotiated access for third parties is twice as expensive as the Energinet.dk system (241).

(411) In its reply to the SO, DONG does not contest the capacity constraint brought about by the 1 mcm/day limit but, with reference to the pricing, provides comparative information to underline that the access price to the pipeline is competitive in comparison with other offshore pipelines, which are in its view the relevant comparison.

(412) The Commission notes that even if DONG's argument that the access prices are indeed within the range of those of other transport pipelines is correct, this does not change the fact that access prices are high for third parties planning to transport gas for the Danish gas fields onto shore.

(413) There has not been any concrete planning on the construction of a connection between the Tyra platform and the Europipe I pipeline, which may be a technically and commercially viable solution to source gas into Denmark when production starts to decline in the Danish continental shelf. However as such a connection would have to link Europipe I to DONG's Danish offshore pipelines, DONG has full control of whether such linking, which might lead to Norwegian supplies to Denmark, will occur or not. It therefore does not place any potential constraint on DONG.

(414) On the upstream Danish offshore segment DONG's dominance is therefore characterized by contractual or de facto exclusivity in off-take of gas production in the North Sea as well as the ownership of the upstream pipelines from the platforms to the shore. There are commercial incentives for the DUC members to sell the gas they committed to sell to parties other than DONG not to Denmark but to the Netherlands and it is currently impossible to back-haul gas from the Netherlands towards Denmark. Any link with Europipe I would require DONG's consent.

(415) In its reply to the SO, DONG notes that the fact that there are commercial incentives for the DUC members to sell gas to the Netherlands is in itself not evidential.

(416) However, DONG itself is selling gas to the Netherlands and has increased this volume significantly from 2004 to 2005 (242). Furthermore, as already explained, it is clear that that fact that the DUC parties, along with DONG, have invested into the construction of a pipeline to facilitate exports to the Netherlands and booked capacity in the NOGAT pipeline, are clear signals that they have committed themselves to selling gas to the Netherlands, whereas they have not made any attempt at reserving capacity on the only pipeline via which they can access the Danish market, namely that of DONG.

- Gas imports from Germany south of the DEUDAN pipeline face many obstacles and extra costs

(417) In its response to the decision pursuant to Article 6(1)(c) of the Merger Regulation and also to the SO, DONG submitted that gas could be physically imported from Germany through the DEUDAN pipeline. However, the physical import via DEUDAN does not exert any significant competitive constraint on the Danish wholesale market. (243) First of all, the physical flow in DEUDAN since the upswing of production in the North Sea has always been in North-South direction. Second, "commercial/contractual" transport in South-North direction is possible but only on an interruptible basis. Only very small volumes, [...]* have so far been contractually transported from Germany to Denmark via DEUDAN.

(418) In its reply to the SO, DONG stresses the irrelevance of the flow direction in assessing any competitive constraints on DONG and cites a GTE report (244) claiming that the report makes a statement on the possibility of uninterruptible capacity reservations in the south-north direction.

(419) On this, it needs to be noted that the statement by BEB was hypothetical (245) and in addition contrasts with the statement BEB has made in its answer to the Commission's questionnaire as part of the market investigation. Irrespective of whether there could actually be firm capacity available from south to north, until the present time the imports into Denmark through the DEUDAN have been sporadic and of low volume considering the significant price differences between the two markets that can be inferred by comparing the prices to the three companies receiving gas from DONG at Ellund and the German border prices as published by BAFA (246).

(420) Moreover, it needs to be borne in mind that the transport via DEUDAN is even more expensive than turn-around at Ellund by [Competitors]* (247). For companies importing gas via the contractual counter-flow (248) the additional cost for reserving capacity on the DEUDAN is 0.3 EUR cent/m3/h (249) on a yearly booking and additional (possible pancaking of, i.e. paying several) transmission charges apply when transporting the gas through Germany. Furthermore only interruptible capacity can be booked on the northerly direction on the pipeline (250). In the light of these facts, imports from Germany through DEUDAN cannot be considered as strong competitive constraints to DONG. From today's point of view a reverse physical flow on the DEUDAN pipeline is a completely unrealistic scenario and cannot therefore exert any constraint on DONG.

(421) This also makes DONG's argument, which, while conceding the relatively high prices for transport from Emden through the DEUDAN to Ellund, compares those to the transport costs between Emden and Lübeck (and concludes that the latter are higher), irrelevant. Regarding DONG's inference from a statement in the Commission's Issues Paper that this means that gas could be easily sourced outside of Denmark, it can be emphasized that contrary to the Parties inference, the Issues Paper concluded overall on the difficulty in gaining access to gas on European trading points. It must also be borne in mind that the most significant volumes that have entered Denmark from Germany have been through E2's participation in the E.ON Ruhrgas Gas Release Programme at Emden.

- Gas that could potentially be turned around at Ellund is transported south

(422) The market investigation has also shown that the competitive constraint exerted by the possibility of re-entry of gas sold on the German side of the border is much weaker than suggested by DONG. (251) First of all, the gas volumes acquired by [customers]* (252), are already committed to their respective gas portfolios for the German market. Therefore, any gas turned-around to Denmark would need to be replaced by alternative sources at comparable prices, otherwise the turn-around would not be interesting for any of these three companies.

(423) DONG claims that the fact that the gas purchased at Ellund is already committed for sale in Germany is exaggerated as the three companies buying the gas would be able to find alternative sources to supply their customers in Germany. As already concluded in the section on relevant geographic markets (253), it is unrealistic to assume that - given the illiquidity of most of the European hubs as also highlighted in the Commission's Issues Paper - large volumes of gas could be replaced at ease, especially taking into consideration the economic effects. Using the economic concept of opportunity cost, as suggested by DONG in its reply, and considering the substantial price differences between the two markets as outlined above (254), sales of substantially increased turnaround volumes, to an extent that would constitute an effective constraint on DONG, in the future do not seem likely. (255)

(424) In addition, turn-around of gas at Ellund triggers costs for the re-entry into the gas network of the Danish TSO Energinet.dk. On the basis of the long-term agreements, the gas is delivered to [...]* on the German side of the border, i.e. outside the Energinet.dk network. (256)

(425) These additional costs for re-entry and transport in the Danish system in case of a turn-around of gas at Ellund are not simply compensated by transport cost savings in Germany. On the contrary, costs for capacity reservation for southward transport of the gas in the DEUDAN pipeline to Quarnstedt and further to the destination points become futile. All three companies have made long-term reservations on DEUDAN in North-South direction, and also in other German pipeline networks which are characterised by a general scarcity of available transport capacities. These reservations would be futile in case of a turn-around. Volumes not sold onwards to Germany or the Netherlands thus entail a cost equal to the value of the capacity reservation made but not used in the DEUDAN pipeline and the ensuing network tranches. Therefore the sales price achieved in Denmark would have to compensate also for these reservation costs incurred in order for a turn-around to be profitable as compared to the price which could have been achieved by the initially planned supply via the German gas portfolio.

(426) In view of these additional transport costs triggered by a turn-around, it is already doubtful whether it would be profitable for [...]* to turn-around gas at Ellund in reaction to a price increase at the wholesale level in Denmark. Reference can be made in this context to the evidence submitted by DONG on [...]*. Moreover, it appears from the Eurostat statistics that the price level for supplies to large industrial customers is generally considerably higher in Germany than in Denmark. According to the Eurostat "statistics in focus" publication "Gas prices for EU industry on 1 January 2005", large industrial customers with an annual consumption of 41,860 GJ (approximately 1.1 mcm) paid EUR 6.01 (257) in Denmark whereas the same industrial customer profile paid EUR 8.19 in Hamburg, the Eurostat reference city which is closest to the Danish border and part of the E.ON area. For very large industrial customers with an annual consumption of 418,600 GJ (approximately 11 mcm) the corresponding prices were EUR 5.08 for Denmark and EUR 6.94 for Hamburg. Similar observations on higher German retail prices can be made with regard to other cities such as Hannover, which is situated in the North German BEB area. The quarterly Eurostat statistics confirm such a considerably higher price level in Germany as compared to Denmark also over time. (258) Suppliers such as [...]* which all have strong sales to industrial customers in Northern Germany, have only a limited interest to shift their sales of Ellund gas, which is an essential part of their German supply portfolio to the Danish wholesale market, thereby foregoing (high margin) sales to North German industrial customers. (259)

(427) In its reply to the SO, DONG has acknowledged the differences in retail prices between Germany and Denmark and has furnished further data on Danish retail prices for the smallest industrial customer segments, underlining that the Danish prices are lower than those of Germany even in the latter segment. DONG claims that the reason for the lower prices in Denmark is the significant competitive constraint on DONG stemming from the gas available to [...]* at Ellund.

(428) However, DONG fails to consider other pricing constraints which it has been faced with, notably scarcity of export capacity and corresponding need to encourage consumption in Denmark and DONG's constraint on its pricing policy following from its role as fully state-owned company. [...]* (260).

(429) A further disincentive for [...]* to turn around gas results from the fact that the gas they could re-import into the Danish market would be gas which they have previously bought from DONG, their main competitor on the wholesale market in Denmark. DONG already has a margin on the sales to [...]*, and thus lower costs than these [...]* competitors. It could therefore easily outperform any bid by these companies for important customers. [...]* (261), [...]* (262).

(430) With regard to DONG's comment - identical to the one spelled out in the market definition section -, that accounting costs are not a relevant comparison, but instead the opportunity cost principle should be applied to this analysis, the Commission reiterates its remarks already made above (263) regarding the constraints of applying the opportunity cost concept.

(431) [...]*. These contracts expire in [...]* at the latest but can be terminated by either party as from [...]* with four years' notice as of [...]*. This implies a threat that DONG could give notice for [...]* or a subsequent year in order to terminate the supply agreements early. In order not to risk losing their supply contracts, [...]* have no strong incentive to compete aggressively with DONG in Denmark by turning around gas at Ellund.

(432) [...]* (264). [...]* (265).

(433) Furthermore, the market investigation has not indicated that [competitor A]* is present to any relevant degree on the Danish wholesale market. (266) As regards [competitor B]*, there is no indication that it is active on the wholesale market other than via sales to its retail company [...]* - who in turn has been active on the Danish industrial and commercial market since 2003 (267) - [contains information on DONG's contractual relationship with a third party]*. However it must be considered that [competitor C]* presence in Danish downstream markets is a very limited one with market shares in the large customer and decentral CHP market of less than 5%. [Contains information on DONG's contractual relationship with a third party]* (268) to [contains information on DONG's contractual relationship with a third party]* signal that [competitor C]* is focusing on acting as a wholesale agent/facilitator for the activities of [competitor C]* in Sweden, and that the turn-around procedure is not a fully satisfactory solution.

(434) The incentives of [competitors]* are further reduced by DONG's potential for retaliation by expanding its activities in Germany, for example via its 25.1% shareholding in Energie und Wasser Lübeck (EWL) or its wholesale supply joint venture E-Nord. All three companies would risk losing customers in Germany to DONG, which could be very competitive with its own Danish gas, in particular as DONG has direct access to North-South capacities in DEUDAN through its subsidiary DANGAS which has a co-controlling share of 49% in DEUDAN (269). [...]*: [...]* (270)

(435) In its reply to the SO, DONG has pointed out that it has already established a presence in Germany and that consequently the potential for retaliation voiced by the Commission in its SO is no longer present. The Commission disagrees with this statement. [...]*.

(436) Taking into account all the discussed elements, namely transport and re-entry costs, higher price levels in Germany and the described disincentives, the Commission concludes that the possibility of re-entry of gas at Ellund into Denmark does not constitute a competitive constraint for DONG on the Danish wholesale market.

(437) The same is true for those swaps undertaken that could possibly be offered by third parties to [competitors]*. As explained in detail above, [competitors]* are very unlikely to turn-around gas into Denmark on their own account. It is therefore even more unlikely that they would turn around gas into Denmark in order to swap it with third parties' gas abroad. Correspondingly, such a swap has never happened and there are no indications that any of the parties receiving the gas at Ellund would be willing to trade it to potential competitors. The market investigation shows that even though interest has been shown for swaps with either of the companies receiving gas at Ellund, they were not interested in executing such a transaction (271). Furthermore, the likelihood of such swaps is further reduced as those counterparties must also make gas available at the specified locations, in the time period and in the particular quantities that are of interest to either [competitors]*.

(438) In its reply to the SO, DONG argues that the fact that no swaps have yet occurred between the companies having gas at Ellund and others, in particular HNG-MN, who have explicitly inquired about it, is not conclusive. In the case of HNG-MN, whose inquiry took place before liberalization, the market dynamics of today could be wholly different. In the Commission's view, the fact that the market investigation has not shown any indication on the existence or the intentions towards such arrangement render the likelihood of such swaps, and even more DONG's argument on their potential significance, highly speculative.

(439) DONG has also submitted information (272) on the rise in capacity reservations for the entry point Ellund for the year 2006, reaching an approximate value of 1 bcm, based on the figures of Energinet.dk (273). Firstly it must be noted that capacity reservations do not automatically translate into flows. The comparison of the flow and capacity reservation figures for the entry point Ellund for the years 2004 and 2005 (274) show that the actual capacity utilization was at 72% and 71% (275) for the two years respectively. One reason for this is the pricing scheme for capacity reservations applied by the TSO where one-year reservations are significantly cheaper on a unit basis than shorter reservations, whose price rises progressively, especially in the winter season. Furthermore it must also be noted that a substantial part of the reservations made for the entry point can be considered reservations for transit towards Sweden.

- Constraints that could arise of a liquid Danish spot (or secondary wholesale) market are hypothetical due to the illiquidity of the wholesale market in Denmark and swaps with DONG do not constrain DONG's market power

(440) The Commission's market investigation has clearly shown that Danish market participants almost unanimously view the Danish wholesale market as a highly illiquid market in which it is very difficult to conduct short term exchanges of natural gas. Since liberalisation, market participants have only seen first steps towards a higher liquidity of this market (of which the GTF was seen as an important step), and believe that in the absence of the merger liquidity would rise. Market participants also expressed their opinion that such liquidity is a necessary prerequisite for well functioning Danish natural gas markets. One of the wholesale market players commented that the drop in liquidity and the disappearance of E2 have effects on the Danish market for gas: "Liquidity inside the Denmark is important for the development of a competitive wholesale and trading market. Presently his liquidity is very limited. The situation can only be improved if a sufficient number of counter parts are remaining on the Danish market. Presently DONG, E2 and Elsam are the only competitors that could improve market liquidity by using their diversified portfolios and taking on spot opportunities. The disappearance in particular of E2 will have a measurable effect on the Danish gas markets." (276) Such liquidity would be beneficial to other wholesale market participants as it would reduce the costs of and their dependence on DONG (277). By contrast the current low level of such liquidity is a factor strengthening DONG's dominance.

(441) DONG itself has little economic incentive to increase this liquidity. A significant non-price factor in assessing the dominant position of DONG on the wholesale market is that it is only selling gas to a few companies on the market, namely [customer]* (278), [customers]* (279).

(442) Swaps into Denmark concluded with DONG can only be a weak substitute for such liquidity independent of DONG. Thus wholesale market participants can either enter into gas supply or sale agreements with DONG or perform physical gas swaps [...]* the ones with [competitors]* (280), both of which operations necessitate that DONG agrees to the deal. Swapping furthermore also requires the other party to have gas available at the locations, in the time period and in the particular quantities that are of interest to DONG. As swaps require the consent of both parties, DONG has clear control over the volumes, the parties and all other specifications of any swap deal and can execute them in line with its business strategy considering both its international and domestic positions and commitments. As swaps are performed to lower transportation costs or overcome other transportation obstacles such as congestion, DONG may have had an interest in entering into swaps for quantities which it could not (i) transport through its own part of the DEUDAN pipeline or (ii) sell in Denmark or Sweden. However the new pipeline, owned jointly by DONG and the DUC partners, connecting the Danish offshore fields to the NOGAT pipeline from where the gas can be transported to the Netherlands (where DONG has recently acquired a gas retail subsidiary Intergas) reduces any such incentive for DONG and gives it increased opportunities to sell its surplus gas other than through swaps.

(443) In spite of DONG's claim that this latter pipeline connection towards the Netherlands does not affect DONG's willingness to swap, it is reasonable to assume that DONG's incentive to swap will have decreased as a result of the pipeline, as via it, it can deliver its gas to the Dutch or German market even without swapping, and is not only reliant on the capacity it has at its disposal in the DEUDAN pipeline (281).

- Constraints from new pipeline capacity or other import facilities are not immediate

(444) The only advanced project for building new pipeline capacity to Denmark is the Baltic Gas Interconnector project, for which authorizations are almost complete from the three participating countries and an investment decision is soon to be taken. After such investment decision the technical construction can be expected to take less than three years. It would leave the German territory at Rostock (a mere 100 km away from the planned Russian/German Baltic Pipeline and likely easily connectable, particularly as one of the operators, E.ON-Ruhrgas, is in both pipeline consortia) and would go offshore to Southern Sweden with a branch off to E2's gas fired power plant in Avedøre. Once it is built, this pipeline could be a competitive constraint on DONG's dominant position. However, pending the decision whether this pipeline will be built, its realisation cannot currently be taken for certain.

(445) Another project, the LNG terminal planned by E2, is under consideration but no decision has been taken as to the realization yet.

(446) Therefore, the constraints imposed by those projects on DONG's dominance do not currently effectively constrain its dominant position.

- Conclusion on dominance in Denmark

(447) For these reasons, DONG is the dominant player on the market for wholesale supplies for Denmark.

DONG's position in regard to wholesale supplies for Sweden

(448) In a recent decision by the Swedish Competition Authority (282) the Swedish wholesale and retail market was thoroughly analysed. A summary of certain important elements of the decision can be given as follows. ??Swedish gas consumption has been relatively steady with a rise from 855 mcm in 1997 to about 1 bcm in 2003. The Swedish market is set to expand. There are plans to expand the grid to cover larger part of Southern Sweden. Consumption is expected to expand considerably and according to views expressed by market participants could reach 1500 mcm by 2008. Current capacity at the DK-SWE interconnector is estimated to around 1.4 bcm/year. Investment in compressors and other equipment can increase capacity to 2 bcm. [...]* of Swedish gas comes from the Danish part of the North Sea (DONG), the remaining [...]* - through contractual imports - from Germany. ??In 2003 there were three active wholesalers on the Swedish market. DONG sold [70-80%]* of all gas on a combined Danish and Swedish wholesale market. Sydkraft (now E.ON Sverige) accounted for [5-15%]*, while NOVA, to be acquired by DONG, had around [5-10%]*. Other players had the remaining market shares with no player having more than [0-5%]*.

(449) These findings have been largely confirmed in market investigation in this case. In addition, some new developments have occurred since the adoption of the SCA decision by the Swedish Competition Authority. [...]* (283). This contract started in October 2005 [...]* After DONG's acquisition of NOVA (284) the contract was taken over by DONG and executed until its termination. [...]* (285)

(450) [...]* (286).

(451) Competitive constraints on DONG in Sweden are higher than in Denmark only insofar as DONG has an international player, E.ON as a competitor in Sweden, whose comparatively strong position in the Swedish gas retail area is to be noted. (287) Although, there are no firm indications that [...]* import possibilities of E.ON from Germany (as discussed above) is sufficient to overcome DONG's dominance in wholesaling of gas for Sweden, E.ON's involvement in the BGI project could already today result in a stronger position vis-à-vis DONG in Sweden than would otherwise be expected.

(452) For the purpose of this case it can, however, remain open whether DONG is in a position of single dominance (288) with regard to a Swedish or Danish-Swedish market for wholesale supplies for Sweden. This is because, as already mentioned above (289), any negative effect of the operation on the wholesale market for Sweden would be an effect derived from a harmful effect of the operation on the wholesale market for Denmark. Therefore, if, as will be shown below, such harmful effect is remedied by commitments submitted by DONG, it is not necessary to examine the Swedish market separately as such remedies would automatically also resolve any problems on a market for Sweden.

Strengthening of DONG's dominance on the gas wholesale market by removal of actual and potential competition

- E2's and Elsam's commercial interest to become active players in Danish natural gas wholesale market

(453) E2 is a significant player on the Danish gas wholesale market stemming from the fact that it is the single largest consumer of natural gas in the country. It had a consumption of [...]* mcm (i.e. [...]* of total gas consumption in Denmark) in 2004 (290). Energi E2 along with Elsam have long-term contractual relations with DONG for the delivery of gas that make up the core of their gas purchasing volumes. These contracts - the "1991 aftalen" and the "2000 aftalen" in the case of Energi E2 and "1994 aftalen" in the case of Elsam (291) - [...]* (292). The two DONG contracts with Energi E2 have already been shortened (293) and are now running until [...]* in the case of the "1991 aftalen" (with and ACQ of 300 mcm) and [...]* in the case of the "2000 aftalen" or the Option Agreement (with a ACQ of 135 mcm). [...]* (294). Without these long-term contracts E2 would have had to source gas from the wholesale markets in much larger volumes thereby positively affecting even further the development of a gas wholesale market in Denmark. In the case of a further shortening of these contract periods and in any case from [...]* E2 and Elsam would necessarily be important actors on the Danish wholesale market.

(454) In its reply to the SO, DONG disputes the analysis of E2 and Elsam being potential competitors and states that they are not the most important competitive constraints on DONG. DONG states that neither E2 nor Elsam is comparable to [competitors]*, in terms of (i) access to gas in Denmark, (ii) guaranteed access to flexibility, (iii) organisational structure, (iv) experience as gas seller and (v) on the volume of gas sold annually to Danish customers.

(455) In the Commission's view, considering DONG's entrenched dominant position on the Danish wholesale market, E2 and Elsam do not necessarily have to be the single most important competitive constraint to DONG. For the existence of serious competition problems it suffices if they are an important constraint, the removal of which would result in a significant impediment to effective competition (and thus in this case a strengthening of DONG's dominant position). In the Commission's view, these competition problems follow from two independent but mutually reinforcing effects: the customer foreclosure effect (vertical effect: on this see further below) and the elimination of E2 and Elsam as potential competitors (horizontal effect, discussed in this section).

(456) Regarding the issue of access to gas in Denmark (cf. DONG's argument [i] in recital (454)), the Commission's further market investigation (295) has shown that a re-entry procedure (similar to the one at Ellund) could be realized - if requested by E2 and/or Elsam- in the case of the three central CHP plants, connected to the TSO's network, two of them owned by E2 and one owned by Elsam. (296) This could be done within a 12 month period which would include a change in the grid code (297). Such change (or, alternatively, a change in the delivery clause) would enable the two generators to access significant quantities of gas for resale on the Danish market. It is noted that E2 purchases significantly more gas than either [...]* receive in Ellund and that both E2 and Elsam purchase significantly more gas than any of [...]* have sold in Denmark, individually or combined, in the past. Already currently (and irrespective of the alleged "practical difficulties" of turning around long-term gas at the gate of its power plants) E2 sources a short-term gas portfolio that surpasses the retail sales of [...]*, combined, in 2004 in Denmark. (298)

(457) With regard to DONG's argument [ii] in recital (454), Elsam and E2 would thus also be able to access the flexibility in their contract with DONG in combination with the large flexibility of their gas-fired power plants. It is furthermore noted that E2, after DONG, is the second most important company as regards storage capacity booking in Denmark.

(458) As regards DONG's argument (iii) in recital (454[...]* (299).

(459) Concerning arguments (iv) and (iv) in recital (454), even though E2 has so far not sold large volumes on the market, this firstly does not discount its ability to trade gas on the wholesale market and, secondly, does not mean that it does not have significant potential -as [...]* comments from the market investigation support - to enter the wholesale market more strongly also on the selling side. (300)

(460) DONG furthermore sets out that the SO has not provided convincing evidence to suggest any intention of Elsam to enter as a competitor on the wholesale market.

(461) The Commission notes that the fact that Elsam's internal papers do not explicitly mention that it had strong entry intentions, does not, while obviously being relevant evidence, eliminate the commercial incentives which Elsam would have in particular after the expiry of its long-term contract with DONG in [...]* to enter gas wholesaling at least on a spot market basis (i.e. in terms of short-term trading of certain quantities to optimise its supply/demand balance and profit from situations in which higher margins can be achieved in gas than in electricity). The Commission furthermore notes that after the 2004 acquisition of NESA, Elsam also has a retail company that, based on its broad customer base, could potentially successfully enter into the retail gas market, such acquisition (absent the process leading to the currently proposed operation commencing shortly thereafter) having likely had a further strengthening impact on Elsam's commercial incentives to enter wholesale trading of natural gas in Denmark. It is furthermore plausible that Elsam would have had an incentive, encouraged both by its existing participation in E2 and by competition with/benchmarking to E2 that it would itself have followed E2's further advanced plans of entry. It is finally noted that Elsam's negotiating position regarding any further gas contract with DONG would have been improved by the mere threat of entry. Therefore, while the likelihood and weight of Elsam's potential entry into natural gas wholesaling in Denmark is inferior to that of E2's entry, this does not imply that it is to be neglected in the Commission's analysis.

(462) In its reply to the SO, DONG has furthermore noted that the [...]* of E2's and Elsam's gas supplies from DONG would most likely cease to exist upon the termination of the contracts in [...]*, inferring that the two companies would not have any reason to engage in gas trading activities from that point on.

(463) The Commission does not share this view. The two single largest players on the market are always likely to have the most favourable terms and also have the required critical size to warrant own import operations. Reference is made in this context to the above discussion of the BGI pipeline project, in the context of the NEGP pipeline project which is likely to facilitate access to such gas. Through this gas, together with the flexibility of their central CHP plants, Elsam and E2, independently or jointly, could very well participate as important players on the Danish wholesale market.

(464) Moreover, the Parties claim that the heating requirements and the regulated prices of heating restrict the ability of the power plants to switch fuels from gas to oil.

(465) The Commission's analysis of the fuel use at AVV2 (301) challenges this assertion as it clearly shows that E2's largest power plant burns gas mostly in the summer, in condensing mode, without heat production and oil mostly in the winter, in backpressure mode which goes along with heat production. Furthermore, the analysis of heat production at E2's central CHP plants shows that heat production does not necessarily provide a constraint for the power plants even in the winter months as heat production is not at the capacity limit (302). In the winter months of the calendar year 2004 (303) the highest average capacity utilization was at Avedøre, with [...]*, Amager was at [...]* and, H.C. Ørsted and Svanemølle were below [...]*. This, combined with the fuel flexibility both intra-plant (Avedøre, H.C. Ørsted) and between plants (Avedøre, H.C. Ørsted, Amager and Svanemølle all being connected to the interconnected heating grids in the Greater Copenhagen area) confirms that the restrictions placed upon Elsam's and in particular E2's gas reselling flexibility are not particularly high.

(466) [...]* (304). [...]* (305).

(467) [...]* (306) the overall concept of the flexible production potential of E2 (and also similarly for Elsam) has not changed conceptually and therefore the Commission maintains its view of the importance and relevance of the significant flexibility potential of those power plants.

(468) Furthermore E2's aim to develop a gas trade portfolio through the Baltic Gas Interconnector project or the [...]* project is stated in the main commercial tasks for 2004 (307). Additionally E2 is chairman of the Operating Committee of BGI" (308) [...]* (309) [...]*. Nevertheless, E2's initiative towards these projects clearly demonstrates its strong incentive to become an independent player on the Danish natural gas market.

(469) The Parties have cited the DCA's HNG-MN decision claiming that in its market investigation it did not include CHPs in the wholesale market. While it is correct that E2 is not mentioned in the decision as a player on the Danish wholesale market for gas, the reason for this is that none of the 33 largest buyers (70 percent of the market) that were asked in the market investigation mentioned E2 as their supplier. Obviously this is not the same as saying E2 is not a player on the market. In addition, the DCA's market investigation showed that three key players on the Danish gas market consider E2 as a player on the market, despite the fact that none of these players buy gas from E2. [...]* (310).

(470) As a player on the wholesale market, Energi E2 has acquired a shipper's license in the Danish transmission system, reserving capacity in the period 2002-2005 as well as in the E.ON Ruhrgas network in the period 2003-2005 (311). Shipper status in a transmission system means managing own gas transports and as such involves significant system know-how. [...]* (312).

(471) The market investigation has shown that players on the Danish gas market believe that an E2 exit from the wholesale market could cause a sizable to substantial drop in liquidity. Furthermore E2 has been highly ranked by the market players as a strong potential entrant in the gas wholesale market. Additionally many players see that E2 would have potentially entered the gas wholesale market in the absence of the merger (313). If one considers the emergence of a "nascent" segment of the wholesale market in Denmark, encompassing non-captive deliveries at the current entry and exit point as well as the GTF, it can be concluded that E2 has already so far been a very important player on this market.

(472) The Commission has found strong evidence of E2's imminent entry intentions314: [Contains information on tender offers submitted by E2]*.

(473) [Contains information on tender offers submitted by E2]*.

(474) [Contains information on tender offers submitted by E2]*.

(475) In its reply to the SO, DONG dismisses E2's activity in the wholesale market as sporadic and points to the fact that E2 itself rejected two requests for offers in November 2005. The Commission notes that one of the possible reasons for the rejection of the offers was the ongoing merger proceedings with DONG which at the time had been ongoing for almost a year and could well have influenced E2's strategic decision-making.

- Impact on constraints to DONG's dominance

(476) As argued above there are five potential sources of competitive constraint to DONG's dominance on the Danish wholesale gas market :

(a) from operators in the Danish offshore area,

(b) from operators importing gas from Germany south of the DEUDAN pipeline,

(c) from operators willing to turn around gas at Ellund,

(d) from a liquid Danish wholesale market,

(e) from new pipeline capacity or other import facilities.

(477) As a result of the proposed operation DONG's position on the Danish wholesale market will be strengthened by removing E2 as an important source of potential and actual constraint concerning constraints (b), (d) and (e) but to some extent also constraints (a) and (c) and Elsam as an important source of potential constraint regarding constraints (a)-(d). This is for the reasons.

- regarding constraints from operators in the Danish offshore area

(478) It has been argued above that these operators only exert a weak competitive constraint on DONG. However, it has also been indicated to the Commission by offshore gas players that the most likely motivation for offshore competitors of DONG to supply gas to the Danish on-shore region independently from DONG would be customers like E2 and Elsam. (315) This is because both constitute very large customers that can be supplied with limited previous downstream presence. All that would be required from such offshore players would be delivery of regular or irregular quantities of gas to the Danish onshore entry point of Nybro. Supplies to these then de facto wholesale customers would likely cause wholesale entry of these players also to other customers. The elimination of this source of potential threat to DONG's position via the merger can also be characterized as an aspect of customer foreclosure, as the power plants are removed from the market, an aspect that will be discussed in the customer foreclosure section below. In this section on elimination of potential competition by Elsam and E2 it is only relevant to the extent that wholesale gas sourced from DUC could have been resold by E2 profiting from its means of flexibility.

- regarding constraints from operators importing gas through the DEUDAN pipeline

(479) [...]* (316).

(480) Indeed E2 has, in recent years, built up a considerable portfolio of short term and medium term gas purchases in competition to supplies by DONG. It has done so by contracting gas with German players [...]* in Germany who have then delivered the gas to E2 at Ellund or at the Danish GTF and by purchasing such gas in Germany [...]* (317) and transporting it on its own account to Denmark. The quantities which it has thus contracted independently from DONG are impressive: on average above [...]* annually from 2003 to 06/2005. Table 13, giving a complete overview of E2's purchases since 2002, shows that the independent wholesale element ("own delivery") of its purchases has increased from [...]* in 2002 to [...]* in the first semester of 2005. There have been several companies that have benefited from and have thus themselves become wholesale entrants for supplies to and for Denmark.

Table 13: Supplies to E2

[...]*

(481) In its reply to the SO, DONG states that there are indications that E2 is only a player on the gas procurement market (318) instead of it being a player on the wholesale market. Furthermore, they state that in the advanced state of liberalization that characterizes the Danish gas market, "booking and management of transport, balancing and even storage are not services that are naturally linked only to wholesale."

(482) The Commission notes that there is no evidence to substantiate this claim and in fact the market investigation has shown exactly the opposite, namely that only the very few players active on the wholesale market manage the above-mentioned complex services for themselves and/or for others.

(483) DONG also states that gas purchased by E2 outside the long-term agreements with DONG is used for the supply of its decentral plants, explaining this with the substantial practical difficulties of using more than one supplier at a single exit point. However, the TSO has provided the Commission with contrary information on these alleged practical difficulties in its market investigation, stating that it is possible, without great difficulty, to manage the deliveries of more than one supplier to the gates of one of the central CHPs connected to the high-pressure grid. (319) .

(484) E2 has even begun to be involved in selling gas on the wholesale market outside Denmark. [...]* (320). [...]* (321) [...]* (322). This shows E2's growing ease of purchases and sales even outside Denmark, testifying to the importance of its activities outside Denmark which have enabled it in the past to purchase cheap gas on its own account and transport it to Denmark, thereby reducing its own dependence on DONG. As shown in the purchases and sales of E2 these transactions have made it possible to invite to the market new participants not tied to other partners via long-term contracts.(On this aspect see the section dealing with customer foreclosure part).

(485) This E2-proper and E2-induced import activity has constituted both a step by E2 towards becoming an actual competitor to DONG in Denmark (horizontal effect) and an incentive for other wholesale players to enter the Danish market (vertical effect), a significant threat to DONG's dominance in Denmark. It can be assumed that in the process of the further opening of markets Elsam's interest would be similar to the proven interest of E2 in further increasing the competitive constraint on DONG. The merger will eliminate such competition.

- regarding constraints from operators willing to turn around gas at Ellund

(486) In spite of its aggressive procurement strategy, E2 has not succeeded in obtaining turn around gas at Ellund since 2003 from either [Competitors]* (which may in itself be taken as an indication that this gas is not readily available and is therefore a weak constraint for DONG's position on the Danish wholesale market). However, in 2002, E2 obtained more than [...]* of natural gas from [Competitors]* at Ellund which appears to have been turn-around gas. Therefore, it can be assumed that, E2 and by implication Elsam, would constitute possible incentives for turn-around gas to be delivered to Denmark in competition with DONG.

- constraints from a liquid Danish wholesale market

(487) The Commission's Issues Paper on the Energy Sector Inquiry (323) underlines that liquid wholesale markets are necessary for new entrants to enter the market and this holds true in the case of the Danish wholesale market as well. Short term trades for balancing positions via a market point or hub such as the GTF are also very important in relation to transparency on prices in the market. Some of the more developed and more liquid gas hubs in Europe such as the TTF (324), the NBP (325), Zeebrugge and Emden have price quotations that are already used as the basis for contracts. Consequently trades by a company like E2 (or potentially Elsam), whose activities would tend to fuel such short term wholesales (326) are very important contributions not just to the liquidity but also to the transparency of the Danish wholesale market.

(488) The wholesale market in Denmark is still at an emerging stage. As mentioned, the Danish TSO, Energinet.dk, has set up a virtual hub (327), the Gas Transfer Facility (GTF). The service is especially suitable for "shippers" (328) who have either a surplus or a shortfall of gas available in the transmission system. According to Energinet.dk (statistics submitted in the notification) approximately 9% of the total Danish consumption is traded on GTF. (329) However, (bilateral) ad-hoc transactions also take place outside the GTF, for example, in order to sell excess supplies or to cover shortages of gas. Both Elsam and E2 have used the GTF to transfer gas (330) and both have the necessary shipping licences. Elsam is currently not active on the GTF but has been in the past. Until 2005, E2 traded gas on the GTF, which is also one of E2's delivery points. (331)

(489) Market participants in the Commission's market investigation have stressed the crucial role that the development of a liquid wholesale market in Denmark would play for the proper-functioning of competitive Danish gas markets. The role Elsam and E2 would play in such a development towards more liquidity was considered to be essential. As power plant operators Energi E2 and Elsam are able to switch fuel on short notice in several of their plants. Thus, they have been, and would in the absence of the merger continue to be, particularly well placed to trade their gas supplies. (332) As a consequence of their likely disappearance as traders, the emergence of a liquid Danish wholesale/trading market would be considerably hampered and DONG's dominance on this market could be entrenched. (333) Market participants did not assume that this liquidity could be replaced readily by other market players.

(490) DONG's' comment in its reply to the SO that the lack of liquidity of the GTF is not a specific result of the merger is valid, but it is likely that a liquid hub can have the most significant effect on the development of a liquid wholesale market and the market investigation has confirmed E2's important past and expected future role in bringing liquidity to the hub, be that on the purchaser or the supplier side. Additionally, as already discussed earlier (334) the TSO has confirmed the possibility of changing the grid code - if requested - in order to facilitate the re-entry of gas into the system at the gates of some central CHPs, rendering DONG's argument void that the fuel switch for several CHPs is irrelevant given the practical impossibility of selling unused gas at the GTF.

- competitive constraint from new pipeline capacity or other import facilities

(491) [...]* (335) [...]* (336). A clear goal of the project is to enhance competition and secure diversity of supply (337) in the Danish and Swedish gas markets. Members of the consortium are E2, E.ON Sverige (Sydkraft) [...] and [...]. Observing parties or nonvoting "Associates" are [...] (338).

(492) The Baltic Interconnector Project was started in 1998 and the feasibility study, with 50% co-financing by the Commission, was concluded in 2000. Since then seabed surveys, ecological studies and public planning processes have been finalized and the risk analysis has been conducted (339). It is currently in the phase of receiving authorization from the countries Member States concerned. Germany and Sweden have already issued their authorization for the project and the "permission of the Danish

Energy Agency is expected before the end of this year". (340)

(493) The BGI is planned to be an offshore pipeline running from the VNG transmission system area in Germany (Dierhagen or Börgerende) to Denmark with a connection to Avedøreværket and to Sweden with a connection near Trelleborg on the southern coast. Its planned length is 210 km, with a diameter of 28" and an initial capacity of 5 bcm/year that can later on be expanded to 10 bcm/year (341). The main flow is planned to be in a south-north direction. As the current consumption of E2 is around [...]*, the planned initial capacity can serve as the basis of importing those volumes required for own consumption, and leaving sufficient capacity for importing gas with a view to reselling it on the Danish market.

(494) Members of the consortium consider that the merger poses significant threats for the projects. "E2 is driving party of the project and merger will affect interest to continue - worst case is close of project" (342).

(495) In the Commission's view, the realisation of the BGI project would lead to significant competition to DONG's entrenched wholesale dominance in Denmark and Sweden. E2 can be expected to use this import facility to vertically integrate upstream by securing an upstream contract (for example, with Russian gas) (343). E2 would deliver gas to its power plants itself and have every incentive to fiercely compete for further wholesale (and retail) customers in Denmark (and even Sweden) in competition with DONG. Given this clear threat to its dominant position in Denmark (and the threat to its possibly dominant position in Sweden), DONG has a clear interest in attempting to block this project. Post-merger, as it will be one of only two leading partners, it will very likely be in a position to achieve such a goal.

(496) Even if DONG, post-merger, would not be in a position to achieve this strategic goal of blocking this new capacity (or were to decide that such new capacity, if controlled would be in its own interest), it would nevertheless be the leading partner together with E.ON and would very likely manage to retain full control over at least the Danish branch of the pipeline.

(497) In its reply to the SO, DONG has argued that there is an inconsistency in the Commission's arguments stemming from the fact that, on the one hand, it has stated that there are serious disincentives for E.ON to aggressively compete with DONG (344), while, at the same time, stating that E.ON is a leading partner in the BGI project together with E2. However, the Commission's argument referred to E.ON Ruhrgas competing with DONG on DONG's home market in Denmark (while always conceding that DONG and E.ON are together the two strongest players on the Swedish market). E.ON's participation in the pipeline project could well lead to its focussing on the Swedish market rather than Denmark. Competition with DONG for Denmark would be much more likely to arise from E2's participation (as shown also by the fact that the Danish branch of the pipeline is to connect at E2's central CHP Avedøreværket). Thus E.ON's participation in the BGI project is not in contradiction to the statement that it has a disincentive to enter aggressively into the Danish market because of the retaliation potential that could be exercised by DONG. The fact that E2 has not pushed the project aggressively recently, may well be explained by the ongoing merger process.

(498) In addition to Energi E2's clear involvement in the BGI project, it has also been sounding out the possibilities of building a liquefied natural gas (LNG) terminal in Denmark at one of its facilities (345)*. [...]* (346). [...]* (347). [...]* further confirms that E2 seriously intends to enter gas wholesaling in Denmark and, given the limited size of the Danish market and the availability of transport capacity between Sweden and Denmark, also in Sweden.

(499) In response to the decision pursuant to Article 6.1.(c) of the Merger Regulation (348) DONG has criticized what it regards as the Commission's asymmetric standard for assessing the Baltic Gas Interconnector and other pipelines, in particular the DEUDAN. DONG's criticism is unfounded. It should be noted that the two pipelines are conceptually completely different. The DEUDAN pipeline, with a large participation by DONG, has southwards flow (349) with possibilities to gain access to contractual interruptible (350) northwards capacity. The current business rationale and model of this pipeline is directed at exporting gas from Denmark, whereby the partners have only limited commercial and strategic interest in achieving significant contractual northbound flow. By contrast the BGI is conceived to become a gas import pipeline with a prevailing northwards flow, providing uninterruptible capacity, owned, in the absence of the merger, by competitors of DONG with a strong commercial interest in boosting the capacity use in this pipeline quickly. The two pipelines are, as such, not comparable in their objectives.

- Vattenfall cannot compensate for the loss of actual and potential competition

(500) The likelihood of a significant impediment to effective competition as a consequence of the removal of Elsam and E2 as actual and potential competitors on the gas wholesale market will not be reduced by the announced divestiture of some Danish power plants to Vattenfall. According to the information contained in the notification, Vattenfall will only acquire a small share (351) of Elsam and E2's gas-fired power plants and is therefore unlikely to be able to outweigh the loss of competitive pressure caused by the integration of Elsam and E2 into the DONG group. Altogether the plants to be divested to Vattenfall include two decentral CHP plants (Helsingør and Hillerød) as well as the Fynsværket, of which only unit 3 (FYV B3) is a gas-fired unit. The annual consumption of these three power plants was [...]* mcm in 2004 (352) which is only [...]* of the gas currently purchased by E2 and Elsam. It should be noted that gas consumption has dropped greatly for FYV B3. [...]* (353). Furthermore it should be noted that FYV B3 is the most expensive block of the Elsam production facilities and is generally the last block to enter into production (354). With regard to the two decentral power plants, it should be noted that while Vattenfall will have to acquire the necessary volumes for the fuelling of the Hillerød and Helsingør plants, these will not provide any of the flexibility that an operator gains by acquiring central power plant. Decentral power plants are quite inflexible regarding their consumption of gas; they are de facto shut down in the summer and are only operational when producing heat and electricity together in the winter. Thus Vattenfall will not be able to replace the actual position of E2 nor even the potential position of Elsam as a competitor on the Danish wholesale market for natural gas (355).

(501) While it is correct - as pointed out by DONG - that Vattenfall has recently created a gas trading desk in Germany, the inference the Parties make about the role of this trading desk on the Danish market is unsubstantiated as there are no indications that this trading desk will deal with any market than the German market. Finally DONG has also alluded in its reply to the SO to Vattenfall's general advantage over Elsam or E2 as regards its independence, who in contrast DONG claim held substantial shareholdings in the latter two - on this issue reference is made to the earlier analysis in this decision on the issue of cross-shareholdings (356) and remarks on this issue further below.

- Conclusion on potential competition on the Danish wholesale market

(502) For the above reasons, the proposed concentration removes very significant competitive constraints on DONG, and in particular removes E2 as an actual and Elsam as a credible potential competitor, on the Danish wholesale market for natural gas. Given DONG's entrenched dominant position on such a market, this is likely to lead to a significant impediment to effective competition, in particular through the strengthening of DONG's dominant position.

Strengthening of DONG's dominance on the wholesale market through the customer foreclosure effect of the merger

The merger forecloses the demand of the two most important Danish customers

(503) In addition to its horizontal effects, the merger leads to customer foreclosure due to vertical integration of DONG with Elsam and E2, as other suppliers are unlikely to be able to secure supplies to DONG's future electricity generation subsidiaries. (357) Elsam and E2 are, by far, the most important Danish customers accounting for about [...]* of total Danish consumption.

<emplacement tableau>

(504) The Commission, in the Statement of Objections, has stated that the foreclosure of this demand after the merger will make it more difficult for competitors to DONG to enter the Danish natural gas sector, whether as wholesale suppliers or as suppliers of final customers, thereby raising barriers to entry to all of these markets.

(505) In its Reply to the SO DONG submits that an analysis of the customer foreclosure theory should be carried out with reference to total consumption in Denmark which it acknowledges is a concept implicit in the Commission's analysis as set out in the SO. Reference to a market that is narrower than total Danish consumption would not be meaningful. The alleged customer foreclosure effect should be analysed in the context of Elsam's and E2's overall gas purchases.

(506) The Commission notes that DONG's view on the scope of analysis of the customer foreclosure effect is not in contradiction to the view set out in the Statement of Objections as regards the impact of the customer foreclosure effect on the Danish wholesale market (which likewise encompasses all wholesales in and into Denmark which are destined for further supplies to Danish customers). As regards individual retail markets, the Commission, however, maintains that in addition to having an effect on an overall market, the customer foreclosure of Elsam and E2 can also have an effect on entry incentives into individual markets, if these incentives are negatively impacted by lowered entry incentives into either the vertically related wholesale market or into neighbouring retail markets. The Commission agrees that the customer foreclosure effect should be analysed in the context of Elsam's and E2's overall gas purchases.

(507) The notifying party furthermore argues that such an effect will not arise, or will only arise to a small extent, due to (a) due to the existing supply agreements by DONG with (b) practical obstacles of dual supplier supplies to Elsam's and E2's power plants, (c) other reasons such as the lack of likelihood that Elsam would start sourcing additional quantities from third parties and (d) the divestment of Elsam and E2 assets to Vattenfall. Each of these arguments will be discussed below. Irrespective of the likelihood of the shortening or illegality of the existing longterm contracts with Elsam and E2, the certainty of their foreclosure after [...]* has a prior effect on competition

(508) The Commission notes that the legality of these long-term contracts is not beyond doubt. In particular, the long-term supply contract between DONG and Elsam raises, in the light of its possible quasi-exclusive character, some doubts as to its legality. Yet, with regard to the necessity for speed, which characterises the general scheme of the Merger Regulation, the question whether long term supply contracts for gas are compatible with the Community competition rules cannot be decided in this merger control decision. Such findings in this case could not be made without undertaking a detailed investigation into the matter, in particular on foreclosure effects and possible justifications for the contracts concerned. Moreover, DONG's long-term contracts with Elsam and E2 have previously been shortened. However, given the pricing arrangements currently favourable to them, Elsam and E2 would, also in the absence of the merger, have been unlikely to challenge these contracts, and it is uncertain whether they would have been successfully challenged in a timely manner by any third party before [...]* (359).

(509) However, the Commission does not agree with DONG's contention that these quantities therefore have to be set aside for the purpose of the competitive effects of the merger. The expiry of the largest part of these contracts in [...]* (representing [...]* ACQ out of a total of [...]* ACQ) is a certainty, which is not contested by DONG. This certainty affects competitors' assessments of the profitability of an entry into the Danish wholesale and retail markets even prior to that date. This follows, firstly, from the fact that any significant entry into Danish markets obviously has to be built not only on a short-term but also on a mid-term perspective and, secondly, from the fact that negotiations on such large quantities are highly likely to commence a considerable time before [...]* and at the latest in [...]*. It would therefore be incorrect not to attribute an immediate effect on competition in Danish wholesale and retail gas markets to these (foreclosed) volumes, representing above [...]* of total Danish demand (above [...]* if combined with the E2 "top-up" quantities). Elsam's and E2's power plants could have dual-supplier contracts before [...]*

(510) DONG has argued that for practical reasons linked to Energinet.dk's balancing regime there is a practical impossibility for Elsam and E2 to source additional quantities from third party suppliers for their central CHPs, which are in any event partly supplied by DONG until [...]* under the long-term contracts. In DONG's view, this practical impossibility further reduces the foreclosed quantity. Evidence provided by Energinet.dk contradicts this information. The Commission therefore concludes that Elsam and E2 could, even before [...]*, have elected to source (further) quantities from third parties for their central CHPs.

Irrespective of whether Elsam would have started to source quantities from third parties in the short term, E2's short term third party demand ( "top-up" quantities) would likely have been considerable

(511) DONG argues that Elsam is not likely to have started sourcing additional gas from third parties as it has so far not done so. The Commission agrees that Elsam has so far sourced natural gas from third parties to a much more limited extent than E2, since, indeed, it has only engaged in some swapping and short term purchasing arrangements with E2 on the basis of both parties' long-term DONG contracts. However, as argued elsewhere in this decision, Elsam had significant incentives to enter Danish gas markets. Furthermore, these incentives significantly increased shortly before the process leading to the proposed merger for example, due to the acquisition of NESA in 2004 and the complete opening up of Danish retail gas markets in 2004) and in general due to the progress of the opening of markets in natural gas and electricity. Dry-year scenarios (or very cold winters), which cannot be foreseen with certainty but which nevertheless have a certain likelihood, would have increased the likelihood of Elsam's interest in, or even need for, additional supplies, going beyond the quantities contracted with DONG on the basis of the long term contract. (360)

(512) Supplies to Elsam by DONG's rivals would thus at least have constituted a potential constraint on DONG's behaviour and would have provided an entry incentive and possibility to DONG's competitors. The Commission therefore maintains that the foreclosed volumes potentially sourced by Elsam even before [...]* need to be considered foreclosed demand raising barriers to entry.

(513) Irrespective of this foreclosed demand by Elsam, the foreclosure of the demand by E2 is considerable even irrespective of the above considerations concerning the long-term demand. In addition, it follows from the Commission's arguments concerning the likelihood of E2's entry into retail supply markets, in particular into the market(s) for the supply of natural gas to large industrial customers and decentral CHPs, that E2's future foreclosed demand can be expected to have been greater than a mere linear projection of past demand into the future.

(514) In its Reply to the SO, DONG does not effectively contest that the foreclosure of E2's short term demand is higher than what is offset by the divestiture of the two decentral CHPs to Vattenfall. DONG states E2's short-term purchases as [...]*, [...]* and [...]* mcm for 2002, 2003 and 2004, respectively, constituting [...]* to [...]* of Danish demand.

(515) As to DONG's argument that E2's short term demand has partly been satisfied at Emden in Germany, it can be noted that by far the largest part (361) of E2's short term demand from 2002 to 2004 was satisfied not at Emden but either at the Danish border in Ellund or within Denmark, usually at the GTF, and that, even for the quantities purchased at Emden, the use of these quantities for Denmark (and therefore the constraining effect on DONG in Denmark) was obvious.

The divestment of assets to Vattenfall will only have a limited compensating effect

(516) Vattenfall will acquire three gas-fired power plants: namely the decentral power plants Helsingør and Hillerød, currently owned by E2, and one central CHP, Fynsværket, owned by Elsam.

(517) [...]* However, the expected gas consumption of Fynsværket is low. The plant is expected by Elsam to be largely on stand-by mode in 2006 and 2007, operating only during parts of 4 months each year and burning annually no more than [...]* and [...]* in 2006 and 2007, respectively, (362) thereby amounting to clearly less than [...]* of Danish consumption. The [...]* (363).

(518) As regards the two decentral power plants acquired by Vattenfall, which constitute the entire former Elsam and E2 demand that is (prior to [...]*) not likely to be foreclosed (364), their demand is in any event significantly below the immediately foreclosed quantities of E2's short-term needs, and can therefore, even in a best-case scenario, not offset the customer foreclosure effect following from these plants.

(519) In addition, the incentive for a third party to enter Denmark to supply E2 and/or Elsam, is much greater than to supply Vattenfall, given the greater potential for future sales to E2 and/or Elsam as such volumes providing a better foothold for further expansion in DK than any sales to Vattenfall.

Conclusion on customer foreclosure for the wholesale market

(520) The Commission considers that even if the foreclosure effect was limited to the foreclosed E2 short term volumes, such foreclosure would constitute a significant strengthening of DONG's dominant position (and as such also a significant impediment to effective competition) on the Danish wholesale market by creating further disincentives to third parties' entering this market on a significant scale. This is because nearly all independent (non-captive) third party demand in Denmark has so far come from E2, including satisfaction of part of this demand by companies who also supplied their own affiliates active in Denmark at wholesale level (like BEB (365), Statoil or E.ON). Therefore the foreclosure of this demand will not only have the effect of raising barriers to entry to companies not yet present on the Danish wholesale market but will actually lead to a complete drying up of such non-captive market, harming both existing competitors to DONG on the market (such as RWE, Statoil, E.ON and EnBW (366)) but also retail suppliers and wholesale customers such as Statoil Gazelle and HNG-MN (367).

(521) Moreover, the Commission considers that the customer foreclosure effect of the merger cannot be limited to E2's foreclosed short-term demand due to (i) the possibility for Elsam to source short-term volumes with third parties, (ii) the likelihood that the definite expiry of the long-term contracts in [...]* will have a stimulating effect on the Danish wholesale market even before that date and (iii) the certainty that these contracts will be terminated in any event as of [...]*.

(522) As set out in the Statement of Objections the shrinking of the accessible market and the increased difficulty for achieving scale effects e.g. as regards customer portfolio, flexibility requirements, storage costs, gas procurement and gas transport/supply chain management will lead to an increase of entry barriers into the Danish wholesale market and will therefore lead to a significant impediment to effective competition, in particular as regards the strengthening of DONG's dominant position. The proposed merger will raise barriers to the Danish wholesale market because it eliminates a company that has purchased important quantities of natural gas on the wholesale market with a view both to using it internally and to supplying natural gas to other users. The elimination of this company generally reduces entry possibilities into the Danish wholesale market and thereby raises barriers to entry to the market.

(523) The Commission also considers that such foreclosure of customers can have an effect on the wholesale market for Sweden, as it reduces incentives for a combined entry into Danish and Swedish wholesale markets. However, it is not necessary to conclude on this point as, due to the close links between these markets, any negative effect on the market in Sweden would also be remedied if the negative effect on the market in Denmark were remedied.

Customer foreclosure and retail markets

(524) As regards the impact of the customer foreclosure on the market for supplies to central CHPs, this will be discussed in the context of the assessment of this market. For the other retail markets, however, it appears appropriate to discuss the effect of customer foreclosure on these markets already at this stage as this effect is closely linked to the one just discussed for the wholesale market.

(525) Regarding the remaining retail supply markets for large and small customers the Commission has set out in its Statement of Objections that the proposed merger also makes entry into these Danish markets, in particular the market(s) for the supply of natural gas to large industrial customers and/or decentral CHPs, and the market for supply to small customers (small business customers and household customers) more difficult. This is because of shared requirements for operating on supply markets, such as wholesale procurement of gas, storage/flexibility management of gas and supply chain management. The merger thus, due to the size of the foreclosed demand and the reduced likelihood that other operators could benefit from the achievable cross-market cost savings between other supply markets and the supply market to central CHPs therefore, makes entry into these other supply markets more difficult and thereby raises barriers to entry. This is aggravated by the fact that - as the market investigation has shown - these other markets in Denmark will show only very small growth (368) for all customer groups except decentral CHPs which is even a shrinking market.

(526) With reference to the impact of the concentration on other retail markets, DONG states, first, that, to date, other suppliers have not been foreclosed from other gas markets, which will not change after the merger. Second, it states that arguably central CHPs are not customers that a new entrant (or a potential entrant) would target as a means of entering the Danish gas market. Third, it claims that critical volume for entry into the Danish market is "relatively low" due to the fact that the Danish market is fully liberalised and complies with all Community gas Directives.

(527) As to the first argument, reference can be made to the fact that affiliates/group companies of the companies active in Danish supply markets, have hitherto supplied E2 [...]* and that therefore the elimination of this supply possibility will also be expected to negatively impact on the possibility of these companies/groups to remain and/or develop their presence on the Danish market. Secondly, while acknowledging that large power producers are a specific customer group with specific supply needs, it needs to be noted that (as set out in the reply to the SO) at least for a certain type of supplies to these power producers entry barriers from the wholesale level are comparatively low (i.e. if these supplies are not considered wholesale sales themselves) so that this kind of supplies, [...]*, are quite attractive to new entrants. Third, the Commission agrees that full implementation of Community liberalisation Directives lowers barriers to entry and that the structural unbundling of the TSO (Energinet.dk) and the supply company (DONG) also facilitates entry. However, this unbundling which has already occurred, is, as regards transmission, not merger specific and therefore does not have an immediate impact on the analysis as to whether the proposed concentration will raise entry barriers.

(528) For these reasons, the Commission maintains that barriers to entry on retail markets are affected by the customer foreclosure effect of the merger. However, this raising of entry barriers on markets for the retail supply to large customers and for small customers follows from the foreclosure of customers who are to a large extent either wholesale customers or customers on a specific market for central CHPs (and thus customers on markets other than the retail supply to large customers and to small customers). By consequence, this customer foreclosure effect can also be overcome by measures primarily affecting other markets but facilitating entry on retail markets (such as the provision of wholesale liquidity). Therefore, since, as will be shown below, such concerns would in any event be remedied by the remedies submitted by DONG for the wholesale market, it can in the end be left open whether the merger raises concerns on the markets for sales of natural gas to large business customers and to small customers due to foreclosure effects.

Impact on the wholesale market for Sweden

(529) The competitive analysis of the gas wholesale market does not change significantly for Sweden. According to the findings of the Swedish Competition Authority (SCA) in its DONG/Nova merger decision of October 2004 (369), the annual gas consumption in Sweden is approximately 1 bcm (370), i.e. 20-25% of the Danish gas consumption. In the same decision the SCA also found that DONG supplied [70-80%]* of the overall Danish-Swedish gas consumption (increased by further [...]* following the acquisition of Nova). This market position is similar to DONG's position in Denmark where it supplies approximately [70-80%]* of the annual consumption. In regard to gas wholesale and imports Sweden currently entirely depends on Denmark.

(530) Elsam and E2 are credible competitors to DONG also on a Danish-Swedish gas wholesale market. [...]*. This proves that in the absence of the merger E2 would be a credible competitor on the Swedish wholesale market. (371)

(531) However, for a full assessment as regards Sweden, also the competition constraint that DONG is likely to face on the wholesale market for Sweden from other players having access to natural gas through central CHPs (Göteborg Energi and E.ON) needs to be considered. These factors and E.ON strategic focus on Sweden indicate that any harmful effect of the concentration on the Swedish market would be less important than the harmful effect on the Danish market.

(532) Therefore, as argued above, it is, in view of the impact of the proposed remedies as examined below, not necessary to come to a conclusion regarding the impact of the concentration on the market for wholesales of natural gas for Sweden.

3. The market for the supply of natural gas to central CHPs

(533) In order to assess the effect of the proposed concentration on the market for supply of natural gas to central CHPs it is necessary to consider whether the concentration leads to competitive concerns on this market through either a) customer foreclosure or input foreclosure (vertical effects) or b) elimination of potential competition (horizontal effects).

a) Customer foreclosure/input foreclosure

(534) There is a potential for customer foreclosure with respect to the market for the supply of natural gas to central CHPs where there is a strong likelihood that Elsam and E2 as customers will be removed from the market, thereby strengthening DONG's position on the market and raising barriers to entry (through removal of the largest part of demand) which could, in theory, be to the detriment of the gas-fired plant divested to Vattenfall (Fynsværket) as well as the three gas-fired central CHPs in Sweden (Heleneholmsverket and a new plant both owned by E.ON and the Rya plant owned by Göteborg Energi). For these rival plants also the issue of input foreclosure needs to be considered.

(535) However - at least until [...]* - there would not be any central CHP customers that would suffer harm on either the Danish or the Swedish markets for supply to central CHPs.

(536) Fynsværket, acquired by Vattenfall, is the only non DONG-owned gas fired CHP customer left on the Danish market. Fynsværket's new owner Vattenfall, has the right, on the basis of the agreements between DONG and Vattenfall, to take over part of its former owner Elsam's supply agreement with DONG, which lasts until [...]*. This agreement is at current energy prices very favourable, and would give Vattenfall protection from any possible negative effects. Post [...]* it is uncertain whether Fynsværket will use any gas at all, inter alia due to large investments in a new, straw fired plant at Fynsværket (372). In any event, Vattenfall would post [...]* be protected if the functioning of the wholesale market is reinforced through the Gas Release Programme proposed by the parties to remedy the Commissions concerns on the gas wholesale market.

(537) According to the Commission's information, it is highly unlikely that new gas fired central CHPs will be built in Denmark in the foreseeable future, thereby eliminating any potential adverse effects of the concentration on such potential new power plants.

(538) The Swedish central CHPs (Rya and the two plants owned by E.ON) would also not be affected by the concentration. The Rya plant has already signed a contract with DONG which lasts until [...]* and is therefore protected from any adverse impact of the concentration until then. (373) The two plants owned by E.ON, (374) would be protected if the functioning of the wholesale market is reinforced through the Gas Release Programme proposed by the parties to remedy the Commission's concerns on the gas wholesale market. According to the Commission's information no further gas-fired central plants other than the one owned by E.ON are expected to be built in Sweden in the foreseeable future.

(539) For these reasons, no significant impediment to effective competition can arise as a consequence of vertical effects having an impact on the markets for supply of natural gas to central CHPs in Denmark and Sweden.

b) Potential competition

(540) It should further be considered whether the concentration leads to the elimination of potential competition by E2 or Elsam.

(541) As mentioned above, the Commission considers it highly likely that E2 and/or Elsam would have been well placed to expand their activities on, or enter, the market for wholesale of gas for Denmark and/or Sweden.

(542) However, as also pointed out by the parties in their reply to the SO, it is less likely that E2 and/or Elsam could be considered potential entrants on the market for supply to central CHPs since it would seem unlikely that other central CHPs would be willing to satisfy their gas needs by supplies from direct competitor(s) E2 and/or Elsam. Such an arrangement would give E2 and/or Elsam insight into the input costs of its competitors on the electricity market. In addition, it would make Elsam and/or E2 stronger vis-à-vis their competitors.

c) Conclusion on the market for supply to central CHPs

(543) For these reasons, the Commission considers that the concentration does not lead to a significant impediment of effective competition on the markets for supplies to central CHPs in either Denmark nor Sweden.

4. Market(s) for supplies to industrial customers and decentral CHPs

DONG is dominant and will remain dominant in the foreseeable future

DONG has a very high market share and benefits from its wholesale and storage dominance

(544) DONG's market share on the market(s) for supplies of gas to industrial customers and decentral CHPs can be seen from table 16.

<emplacement tableau>

(545) As can be seen from Table 16 above, DONG's market share was around [60-70%]* in 2004.

(546) The parties point to the fact that DONG has lost considerable market share since the opening of the market. It is true that DONG's market share has gone down from [75- 85%]* in 2001 to around [60-70%]* in 2004, which is the last full year for which market data is available. However, this drop in market share cannot change the fact that DONG is dominant on the relevant market.

(547) In the reply to the SO, the parties estimate that DONG's market share has fallen to less than [55-65%]* in 2005 and is likely to fall further in 2006. The parties submit further, that a market share above 50% does not imply dominance, especially in a market characterised by tendering (376). Finally DONG submits that the Commission "has previously accepted that relatively high market shares are not problematic where there has been a gradual reduction in those shares".

(548) The market share figures for 2005 provided by the parties are mere estimates, have not been substantiated and could not be verified by the Commission. These are therefore, uncertain and do not distinguish between the possible separate markets for supplies to decentral CHPs and large business customers. However, even on the assumption that DONG's estimates were correct, this would for the following reasons not change the finding of dominance on a combined market for supply of gas to decentral CHPs and industrial customers.

(549) First, as the Court of First Instance of the European Communities (" the CFI") affirmed in its recent judgment in General Electric377, market shares also play an important role in evaluating market power in markets characterised by tendering if they remain at a high level for a sufficiently long period of time. DONG's market share has been consistently high (above or around [55-65%]*), which is a clear indication of dominance. Erosion of an incumbent's market share during the first years after liberalisation is not in any way unusual, but has also been the experience in other countries, e.g. the United Kingdom, Italy and Spain and in other sectors post liberalisation, e.g. telecoms. This does not in itself mean that these companies are no longer dominant.

(550) The Commission acknowledges that the alleged drop in market share in 2005 seems to be larger than the years before. However this could, if indeed correct, be explained by the fact that 2005 was the first "full year" of liberalisation since some "old" preliberalisation contracts were still running in 2004. Furthermore in 2005, E.ON bought (378) 63 decentral CHPs, which might represent a gas consumption in the range of around 50-60 mcm (379). These effects are "one-offs", and it must be presumed that the remaining customers are the less price elastic customers, whereas DONG will have the chance to win back some of the elastic customers, when these are tendered in autumn 2006.

(551) As regards 2007, the possibility for HNG/MN to source sufficient quantities of gas independent from DONG at fully competitive prices in order to sustain its important large business customer business (with [10-20%]* market share in 2004) is uncertain. It is likely that DONG is able to take at least a certain portion of those customers which it has so far supplied only indirectly (via HNG/MN) and with regards to whose consumption patterns and other competition-relevant information it has perfect knowledge (surpassing the business knowledge of any competitor). (380) Such a business decision which it can take any time (381) would obviously considerably increase its market share.

(552) There is thus no reason to believe that DONG's market share would continue to drop at the rate alleged for 2005 and would fall below [45-55%]* in the foreseeable future.

(553) As mentioned above, DONG is also dominant in gas wholesale. DONG furthermore controls storages and access to offshore gas as well as part of the import pipes. This dominance upstream further supports DONG's position in this market/these markets of supply to large business customers and decentralised CHPs, since storage facilities and access to gas are important competitive parameters in the context of supplies to these customer groups. It should further be underlined, that DONG supplies all its competitors, except, partly Statoil Gazelle, with wholesale gas and provides storage to all competitors.

(554) In their reply to the SO, the parties claim that the Commission has failed to establish any meaningful link between DONG's position on the wholesale market and dominance on the market for supplies to decentral CHPs and industrial customers. DONG's position on the wholesale market is of limited relevance to supply on this market since DONG either does not supply its competitors on this market (for example Statoil Gazelle) or the supply agreements have been approved by the Danish Competition Authority. [Competitors]* are able to use their volumes at Ellund and Statoil Gazelle purchases its gas from Statoil. HNG/MN is therefore the only competitor sourcing its gas from DONG and that supply contract has recently been reviewed by the Danish Competition Authority, as a result of which HNG-MN's gas volumes also must be considered independent from DONG as per 1 January 2007.

(555) Finally, the parties submit that DONG is not dominant on gas wholesale or storage markets.

(556) As to the parties' arguments on DONG's dominance on wholesale and storage markets, reference is made to the relevant sections on wholesale and storage in this decision.

(557) The CFI has in its judgment in General Electric confirmed the relevance of vertical ties to an assessment of dominance (382). In this case it is obvious that a strong position on the gas wholesale market confers more flexibility on the retail market, e.g. gas that is not sold on retail market can be sold on the wholesale market and vice versa.

(558) Furthermore, on a non-liquid market such as the Danish wholesale gas market, access to gas is important. DONG's competitors depend on DONG wholesale supplies to a much larger extent than DONG depends on its competitors. [...]*. Deliveries to [...]* and [...]* at Ellund are made by DONG, at a certain margin to DONG. There are therefore clear advantages for a retail supplier to be vertically integrated and have a strong position upstream at wholesale level.

Its competitors cannot challenge DONG's dominance

(559) HNG/MN and Statoil Gazelle are small companies. HNG/MN is a financially weak competitor. There exist strong ties between DONG and HNG/MN, which sources all its current gas from DONG under an agreement (for further details see below) which is not favourable to HNG-MN's competitiveness on this market and has led to losses in market share by HNG/MN.

(560) E.ON is the competitor with the highest amount of wholesale gas [...]*. However, its very modest market share in the Danish market(s) for sales to large business customers and decentralised CHPs shows that E.ON, has so far been focusing and is, in the future, likely to focus its Nordic efforts on the Swedish market, where it already has a strong local subsidiary, E.ON Sverige (formerly Sydkraft). The Danish market is relatively small compared to E.ON's home market in Germany, where entry by DONG into retail markets could have potentially very damaging effects for E.ON. [...]*:

"[...]*" (383)

(561) Shell's market share has grown since they entered the market, but the company is still a marginal player with some [0-5%]* market share on the combined market for industrial customers and decentral CHPs. It is not likely that Shell in the short term will be in a position to seriously contest DONG's position, [...]*.

(562) In their reply to the SO, the parties submit that the constraints on DONG from these competitors are indeed strong.

(563) As regards Statoil Gazelle, the parties claim that the company has [10-20%]* of the market, that it obtains its gas from suppliers other than DONG and that the size of the company is irrelevant as its parent companies are Statoil, a major integrated oil and gas company and Naturgas Fyn, the incumbent gas distributor on the island of Funen.

(564) A company holding [10-20%]* of the market can only in very exceptional circumstances can constrain a company holding [60-70%]*. Such circumstances have not been argued by DONG and are indeed absent. Furthermore, Statoil Gazelle is more vulnerable than DONG as witnessed by its higher own-customer switching rates (see below under market for supply to small business and household customers). This is substantiated by a quote from a report (384) on the privatisation alternatives for DONG made by Rothschild and commissioned by the Danish Ministry of Finance:

"Although the entire market was opened on 1 January 2004, DONG has so far lost none of its household customers and has gained only a few of such customers in the other distribution companies' areas. It has, however, taken industrial and commercial customers accounting for around 30% (by volume) of Naturgas Fyn's market" .(emphasis added).

(565) It is true that one of the parent companies of Statoil Gazelle, Statoil, has large quantities of gas, but the company encounters high costs and other obstacles when shipping that gas to Denmark, [...]*. Finally, it is true that Naturgas Fyn is the incumbent distributor on Funen, but this distribution area is by far the smallest of the five Danish distribution areas.

(566) As regards HNG/MN, the parties submit that the evidence of size is irrelevant and that the supply agreement between DONG and HNG/MN has recently been reviewed by the Danish Competition Authority and has been amended to remove any possible foreclosure effects.

(567) The Commission notes that the situation on the wholesale market (see relevant parts of this decision on the wholesale market) makes it very unlikely that HNG/MN will be in a position to negotiate a wholesale contract independent of DONG that will allow it to be competitive on this market even after 2007. It is thus very likely that HNG/MN will be wholly or partly dependent upon supplies from DONG even after 2007. The financial weaknesses of HNG/MN has not been remedied by the decision from the Danish Competition Authority and will most likely result in the continuation of the risk avoidance and defensive strategy which the company has followed hitherto.

(568) As regards, E.ON/Ruhrgas, DONG submits that

- the fact that E.ON currently has a relatively small share of the market for supplies to decentral CHP's and industrial customers, is not probative as to whether this share is likely to increase in the future;

- the Commission's conclusions from the fact that the Danish market is relatively small compared to E.ON's home market are counterintuitive and that this theory of importance of size is at odds with its earlier finding that E.ON will focus on the much smaller Swedish market;

- the suggestion that DONG could materially have an impact on the German market is far fetched and has not hindered E.ON from entering the Swedish market. In any event the scenario from the internal paper which the Commission quotes has not been borne out, since DONG has acquired 25.1% of the shares in Stadtwerke Lübeck;

- E.ON has intensified its operation in Denmark in 2005, established a sales department, advertised for four more key account managers and acquired 60 decentralised CHP's in 2005.

(569) The Commission would like to emphasize the following points:

- The liberalisation of the Danish gas market is not in its infancy since the first market opening took place in 2000. In the period since then, E.ON has not entered the Danish market strongly which is in marked contrast to its position in Germany and Sweden.

- The fact that the Swedish market is smaller than the Danish market is irrelevant for the purposes of evaluating the entry intentions of E.ON, since entry in Sweden took place through a joint electricity/gas acquisition (Sydkraft), which is not possible to replicate in Denmark post merger. The Swedish market may also be more interesting to enter for E.ON in view of the market's growth potential.

- [...]*.

- The fact that E.ON has set up a Danish subsidiary merely confirms the Commission's geographic market definition and should rather be seen in the context of E.ON acquiring 60 decentralised CHPs, thereby wanting to selfsupply its decentral power plants and managing their electricity and heat output. Even if the establishment of the Danish subsidiary should be seen as the result of a deliberate entry strategy on Danish gas markets, it would not effectively constrain DONG until, possibly, such time as E.ON becomes independent of DONG in its Swedish and Danish gas wholesale supplies, e.g. through the BGI.

(570) [Contains information on competitor and availability of gas]*.

(571) [Contains information on competitor and availability of gas]*.

(572) As to the advantages of Shell offering gas and diesel/lubricants, it could be expected that such advantages would have materialised by now on the Danish market. However, as can be seen from Table 16 above, Shell still has very low market shares. Other indications of DONG's entrenched dominance pre-merger

(573) DONG's customer switching rate is clearly the lowest among the large players on the market. According to figures provided by the Danish Competition Authority (385), around [...]* of DONG's customers in this market/these markets switched supplier in 2004. However, the average switching rate for the combined market for large industrial customers and decentral CHPs was 30%, and no company other than DONG had a lower switching rate than 25%.

(574) In their reply to the SO, the parties submit that the rate of tendering is more indicative than the rate of switching. In any event, according to DONG's own data, [...]* of DONG's large industrial customers (including decentral CHPs) have switched supplier in 2004 or the years before and by the end of the third quarter 2005 the rate had gone up to [...]*. The parties further claim that the Commission, in its 2005 benchmarking report (386), registers cumulative switching rates in Denmark since market opening in the range of 20-50%.

(575) The Commission does not deny the importance of tenders but the success of DONG in a high number of these tenders is, in particular when seen over time, a highly relevant indicator of market power. The fact that an additional [...]* of DONG's customers switched supplier during the first three quarters of 2005, (as opposed to [...]* in the whole of 2004) does not change the conclusion, that DONG's switching rates are lower than its competitors. As mentioned above, the one-off acquisition by E.ON of 60 decentral CHPs could explain part of the switching in 2005. Finally the fact that the Commission benchmark report estimates cumulative switching in Denmark since market opening to be in the range 20-50% is not contrary to the argument that DONG's switching rates in 2004 were [...]* and clearly the lowest.

(576) The market investigation showed a clear interest on the part of large gas customers to be able to buy electricity and gas from the same supplier, so called dual fuel offers (387). DONG is the best placed company to offer such dual fuel products (and indeed does so), since it is already active - albeit to a much lesser extent than it will be post merger - in the electricity area.

(577) Finally DONG has been able to develop relatively close relations with a number of decentralised CHPs. [...]*. These commercial relations, as they cannot be easily replicated by DONG's competitors in natural gas, further limit possibilities for competitors to effectively challenge its dominant position. (388)

(578) In their reply to the SO the parties, submit that these agreements with the decentral CHPs:

- only concern [...]* out of 120 gas-fired plants and can therefore not constitute the basis for finding of dominance;

- overstate the involvement of DONG in the decentral CHP's since in 12 of the [...]* agreements DONG only forwards gas and power nominations to the respective TSOs and receives/pays imbalance penalties;

- should be compared to the 86 decentral CHP's that have CHP agreements with DONG's competitors which are national power distributors and which could form a good platform for gas sales and further competition in the gas market.

(579) However,

- The Commission does not argue that these agreements create DONG's dominance, but that they strengthen its already dominant position. This is so even if these agreements relate to only [...]* of DONG supplied decentral CHPs. No other gas supplier has similar agreements.

- The argument that DONG's involvement is overstated is largely irrelevant since what matters is the customer benefit and the associated loyalty effect.

- The exclusive presence of power companies in providing CHP agreements confirms an advantage that other gas players do not have. As to the alleged potential entry by these power companies in gas retail, such entry requires access to gas on a liquid gas wholesale market. For reasons discussed above in the part on wholesale gas this is not the case in Denmark.

The proposed operation strengthens DONG's dominance due to the raising of entry barriers

(580) The concentration will lead to a strengthening of DONG's dominant position on the market for supplies to decentralised CHP's and large industrial companies. This is also confirmed by the market investigation, where a clear majority of respondents expected that DONG would strengthen its dominance, prices would go up and it would be more difficult to enter the market as a result of the concentration. (389)

(581) For reasons already indicated above in the discussion of the customer foreclosure effects of the proposed merger on the market for supplies to central CHPs it will be more difficult for other players to attain critical size in gas supply if some [...]* of total Danish gas consumption is de facto removed from the market.

(582) It must be considered that DONG's already privileged access to decentral CHPs will be further strengthened as a consequence of the merger. A large number of decentral CHPs responding to the Commission's market investigation have confirmed that they would be interested in entering into a privileged commercial relationship with a company that would guarantee them margins on the heat production side and would take over the marketing of their electricity output. No company other than DONG - which, as mentioned above, has already developed in this direction - would postmerger be better placed for making such offers to Danish CHPs, which thereby would become even more loyal to DONG and consequently less accessible to DONG's competitors. The proposed merger (due to its integration of by far the strongest Danish gas and electricity companies) would strengthen both DONG's ability and incentives to pursue this strategy successfully and obtain partial control over the economic operation of decentralised CHPs both on the gas and on the electricity side. The increase in DONG's incentives follows from its stronger involvement in the electricity market. Post merger, DONG will have an increased strategic interest in having influence over its own electricity wholesale competitors. This will in turn also increase its incentive to give such offers to central CHPs.

(583) In its reply to the SO, DONG submits that the effect on decentral CHPs is not merger specific since DONG already has the ability to offer services to this category of consumers before the merger. The parties further argue that DONG is not in a privileged position compared to other gas supply companies since it has agreements with only [...]* out of 330 decentral CHPs and that E.ON is in an equally good position since it acquired 60 decentral CHPs in 2005 and now supplies 70 decentral CHPs.

(584) The argument that the effect is not merger specific is not in any way reasoned or substantiated by the parties and the Commission therefore refers to its reasoning above (see recital (582)).

(585) The Commission does not deny that E.ON has the ability and incentives to provide similar services to decentral CHPs, however such ability and incentives are much smaller than DONG's. First, DONG has a first mover advantage and long-term customer relations. Secondly, E.ON is only a small player on the Danish electricity market and a much smaller player on the Danish gas market than DONG. Finally, the acquisition of the 60 decentral CHPs was a one-off possibility following the commitments made by Elsam to the Danish Competition Authority.

(586) As mentioned above, the merger also raises the risk of gas storage input foreclosure or an increase in rivals' costs in connection with storage. This will further increase entry barriers to the market for gas supplies to decentralised CHPs and large industrial companies.

(587) The concentration is furthermore also likely to raise entry barriers on wholesale markets (cf. discussion on wholesale market above) thereby making it more difficult for DONG's competitors in the market for supplies to decentralised CHPs and large industrial companies to source supplies competitively.

(588) DONG will post merger also be in a unique position to offer dual fuel products. It is a position that cannot be matched by other competitors alone, inter alia, because of DONG's unique combined strength in Danish electricity and gas wholesale and retail markets. DONG's privileged access to this customer group due to its position in electricity retailing (390), combined with the fact, as shown in the Commission's market investigation, that competitive dual fuel (gas/electricity) offers would be met with a high level of interest among large electricity and gas customers (391), will lead to a strengthening of DONG's dominant position. This is due to the fact that the operation eliminates the possibility of competitor matching DONG's pre-merger ability and the incentive to offer dual fuel products, e.g. via competing offers made or backed by Elsam-NESA or by E2, perhaps in combination with its major shareholder KE.

(589) [Contains information on dual fuel products and commercial importance hereof]* (392). [...]*.

(590) There is only one further electricity retailer to large customers that is equally strong on the Danish market - EnergiDanmark, who would be capable of teaming up with a gas supplier for dual fuel offers in gas and electricity. EnergiDanmark already supplies a financial gas product for customers with an annual consumption above 300,000 m3, where large gas customers can fix their gas price with EnergiDanmark for a certain period. However, other than KE (supplying city gas) it does not have actual physical sales to gas customers. There has been no indication that it intends to follow KE's explicit multi-utility strategy, cf. below.

(591) Vattenfall is not active in electricity retailing in Denmark and therefore has no Danish customer base.

(592) In its reply to the SO, DONG denies that post merger it will be in a unique position to offer dual fuel products, for the following reasons:

(a) It does not follow from the market investigation that those customers would be prepared to accept in any way inferior terms from their supplier in order to obtain dual-fuel supplies. Indeed, certain respondents who did reply in the affirmative did caveat their response to this effect. It is, therefore, questionable whether an ability to offer dual-fuel would provide a genuine competitive advantage that would be material in comparison to the terms of supply offered by each supplier.

(b) In DONG's experience, large customers are generally more interested in combined supply of gas/oil than combined supply of gas/electricity.

(c) Even if there is a dual fuel effect it amounts to an efficiency that could be matched by other energy suppliers. Numerous competitors active in electricity supply could partner with gas suppliers (given the significant competition that exists on the gas market). Elektra, which merged with EnergiDanmark in January 2005, has teamed up with Statoil Gazelle in order to supply business and industrial customers with both electricity and gas. Other electricity retailers to large customers with a strong position on the market include Sydvest Energi, ScanEnergi and NVE/SEAS. NVE/SEAS are co-operating with Statoil and ScanEnergi is co-operating with Shell.

(d) Elsam-NESA and E2-KE would have been influenced in any decision whether to offer competing offers to DONG by the pre-existing ownership links with DONG.

(e) The argument relating to dual fuel products is tantamount to advancing an argument that a SIEC will result through conglomerate concerns. The Commission has not satisfied the burden in respect of such an argument (see judgment in Tetra (393)), especially given the fact that the ability to offer dualfuel packages appears to constitute an efficiency.

(593) The Commission maintains that DONG is in a unique position to offer dual fuel products and that this is a very significant effect of the merger.

(a) The customer caveats mentioned by DONG are not contrary to the Commission's arguments. The Commission does not allege inferior terms, but merely concludes that DONG would be in a unique position to offer dual fuel products.

(b) DONG does not substantiate its claim that large customers are generally more interested in combined supply of oil/gas. If it were true, Shell's market share would have been much larger than the [0-5%]* in 2004. Even if customers were equally interested in gas/oil supplies and in gas/electricity supplies this would not eliminate the strengthening that comes from being in a position to offer either of them. Furthermore DONG sells oil on the spot market from the Danish production fields in the North Sea, owns the upstream oil pipes from the fields to DONG's oil terminal in Fredericia, and would therefore, it seems, have been able to also sell gas/oil, had it been demanded by its customers or been commercially attractive.

(c) The Commission acknowledges Statoil Gazelle's dual fuel alliance with Elektra (now EnergiDanmark). The Commission itself has considered the possibility of Elsam-NESA, E2 and KE forming alliances and the post merger absence of this teaming-up possibility for other gas players as a merger specific effect. This alliance between Statoil Gazelle and Elektra is evidence of the likelihood that such alliances would have happened in the absence of the merger put competitive pressure on DONG. The elimination of such possibility for gas players as regards Elsam-NESA, KE and E2 is a merger specific strengthening of DONG's position. The press release quoted by DONG is further evidence of the interest of large customers in dual fuel gas/electricity offers (394) . It should however, be noted, that Energi Danmark, in contrast to DONG, is currently not offering physical dual fuel gas/electricity deliveries to its customers (395). NVE/SEAS' alleged cooperation with Statoil and ScanEnergi's co-operation with Shell, has not been substantiated by DONG. The Commission has no evidence on the file that would suggest any such cooperation between Shell and ScanEnergi (396) or between NVE/SEAS and Statoil. Even if such cooperation takes place or is possible in the future, this does not invalidate the argument that DONG's rivals cannot achieve synergies similar to DONG's, even with such cooperation, since DONG has its own flexible generation capacity and a privileged access to wholesale gas at very competitive prices. ScanEnergi has a very small electricity customer base. SEAS-NVE's electricity customer base is bigger but clearly smaller than the merged entity's, which is even more so for large customers than for small customers. In the absence of the merger, as a major shareholder in E2, NVE-SEAS would have been a likely participant in a dual fuel marketing initiative by E2 and its other major shareholder KE. Without these ties, their potential in this direction will be much smaller.

(d) As to Elsam-NESA's and E2's disincentives due to minority crossownerships, reference is made to the general discussion at recitals (283)-(292).

(e) The Tetra case (concerning the creation of a dominant position through conglomerate effects), as well as the GE case, (concerning the strengthening of a dominant position through conglomerate effects) cannot be compared, in this respect, to this case. The effect that is discussed in this case is an immediate and structural effect that does not depend on the behaviour of DONG post merger (or only to the extent that it will continue to offer dual fuel supplies and that it will strengthen its own dominant position, by exploiting, in a normal commercial way, its improved position, thereby structurally preventing, or making more difficult, a comparable improvement of other players.) It must also be emphasized that any efficiencies due to dual fuel supplies are pro-competitive provided that they are accessible also to DONG's competitors. However, in this case, the availability of these efficiencies for competitors is reduced. DONG has not set out an efficiency defence regarding this aspect that would have made it possible for the Commission to verify whether the criteria laid down in the relevant provisions are met.

The proposed operation strengthens DONG's dominance due to elimination of potential competitors

Entry by electricity generators

(594) Both Elsam and E2, as electricity generators, have the ability and the incentive to enter the market for supplies to large business customers and to decentral CHPs. They have access to large quantities of natural gas at competitive prices at which they could achieve good margins with sales to these customer groups. As shown above, at least E2 has the ability and incentive to embark on active gas procurement strategies for the sourcing of additional quantities. Procurement of extra quantities for selling on other markets such as the market(s) for supply of natural gas to large business customers and decentral CHPs would not expose them to particular commercial risks and would allow them to achieve further scale effects on gas procurement and supply chain management. They also, both as wholesale players booking storage capacity and as gas-fired power plants operable as "virtual gas storages", have access to storage and means of flexibility. Moreover, they have the necessary "energy" brand. Consequently, a clear majority of respondents to the Commission's market investigation have stated that Elsam and E2 could sell on extra quantities of gas to them. (397) The entry of electricity generators into sales of natural gas to large customers is also known from examples in other Member States such as the United Kingdom and Spain. Reference can also be made [contains information from internal documents]* (398). The Commission has found strong evidence for such entry opportunities and intentions by E2. [...]* (399).

(595) In their reply to the SO, the parties disagree - for the following reasons - that E2 and/or Elsam would have been potential competitors on the market for supply to decentral CHP's and large industrial customers.

(a) Large industrial customers have little or no interest in the combined supply of gas and electricity. These customers are generally more interested in combined supply of gas/oil than combined supply of gas/electricity. In this regard, companies such as Statoil and Shell have an important competitive advantage compared to DONG as they often use their existing relationship as an oil supplier to offer the supply of gas to their industrial customers

(b) Elsam and E2 are not well placed to act as resellers of gas purchased under the long-term supply contract with DONG or from third parties on the open market due to their practical impossibility to resell gas delivered at power plant gate.

(c) Any flexibility from Elsam's and E2's central CHPs is very limited and can in any case, for practical reasons, not be used for supplying third parties.

(d) Reference to Elsam's and E2's energy brand is overstated, given that neither of them is active in retail. In addition Elsam and E2 do not have the customer base.

(e) Elsam and E2 could become over-exposed in relation to gas prices if they were to act as resellers, which would jeopardise their core activity as electricity generators.

(f) The fact that "a clear majority" of respondents to the Commission's market investigation stated that Elsam and E2 could sell on extra quantities of gas to them is irrelevant to the question of whether Elsam and E2 in fact have the ability and economic incentive to sell gas on to decentralised CHPs and large industrial customers.

(g) The Commission provides no evidence as to why it should be concluded that the experience in the United Kingdom and Spain might be, or would be likely to be, repeated in Denmark. No comparison of market conditions in these Member States has been set out in the SO to support such a contention.

(h) [...]*

(i) [...]*

(j) The Commission ignores the existing relationship between DONG and Elsam/E2. Elsam and E2 are unlikely to act against the interests of a substantial shareholder.

(596) The Commission rebuts each of the parties' arguments and therefore concludes that Elsam and/or E2, as a direct effect of the concentration, in fact is/are removed as potential competitors on the market for supply to decentral CHPs and large industrial customers.

(a) The market investigation has shown that industrial customers have a significant interest in offers of combined supply of gas and electricity. This is irrespective of any interest in combined gas/oil supply. Any such effect can be expected to have materialised already as Statoil Gazelle and Shell have been market participants for some time already. DONG, in this context, does not refer to any other oil companies that would be well placed as potential entrants. Reference is also made to recital (593) above.

(b) As regards the alleged practical impossibility to resell gas delivered at the gates of power plant, reference is made to the discussion of this issue in the part on wholesale at recital (456).

(c) The flexibility opportunities offered by the central CHPs are considerable, as shown above in the part on flexibility. It can furthermore be noted that fuel flexibility is one of the announced research priorities of Elsam (400) and an important part of E2's Strategy (401). The flexibility provided by the central CHPs can indirectly be supplied to third parties via supply agreements that are flexible for the third party. [...]*.

(d) As regards the strength of the energy brand, the Commission notes, that E2 clearly has promoted its brand and has chosen a brand logo which it advertises on a website, in brochures and through visual public displays. Large and expensive advertising campaigns with famous Danish actors on nationwide television can also only be seen as a desire on the part of E2 to promote its brand. Elsam has likewise promoted its brand on the website, brochures, public displays and advertising campaigns. On the website, Elsam is marketed as "Denmark's leading Energy Company" its slogan is "Essential Energy" and the logo, representing a red galaxy402, is visible in sponsorships of the Danish Olympic Sailing Team both for Athens 2004 and Beijing 2008. Elsam has furthermore, through NESA, already entered electricity retail to large customers and therefore already has a customer base. As for E2, it could partner up with its large shareholder KE, who already is active in electricity retail. [...]*.

(e) As to the alleged over-exposure in relation to gas prices, the Commission notes that both and, in particular Elsam, pursue a stated policy of product innovation outside the core area but related to it (403). This is certainly true for gas trading. It should be noted that Elsam currently trades coal. E2 has already shown that it is willing to act as reseller, cf. recital (472)ff and (484) above. Risks associated with such activities are a normal business situation. Companies can insure/hedge against risks. Complementary fuel trading is in itself an insurance instrument.

(f) The Commission considers that views of customers as to whether Elsam and E2 could sell gas on to them also reflect some readiness to purchase such gas and are therefore highly relevant.

(g) As regards the experience from other Member States the Commission is of the view, that - in spite of obvious national differences - there is an underlying industrial gas/electricity cross entry logic that is applicable in all Member States and has had a very positive effect on competition in the common market.

(h) [...]*

(i) The offers made and solicited reflect both a confidence and willingness to sell and a customer's view regarding potential viable suppliers. This is irrespective of whether actual sales followed. It is, for instance, hardly surprising that E2 did not follow up on a requested bid in 2005, in view of the merger project.

(j) As to Elsam-NESA's and E2's disincentives to act against the interests of a minority shareholder, due to minority cross-ownerships, reference is made to the general discussion of this issue at recitals (283)-(292).

Entry by electricity supply companies

(597) Not only electricity generators such as E2 and Elsam but also strong electricity supply companies such as NESA and KE would have the ability and incentive to enter into the supply of natural gas to these customer groups. [...]* (404). These electricity supply companies have strong energy brands and can achieve cost synergies for their sales and service organisations and their overhead costs and could achieve increased customer loyalty by sales of duel fuel. They have access to a large customer base of industrial customers which, as stated above is interested in competitive offers of dual fuel sales of electricity and natural gas. Also Statoil Gazelle considered KE, Elsam and Energi E2 potential competitors on Danish gas markets. In Statoil Gazelle's business plan from November 2002 (405), KE and Elsam are mentioned as potential competitors on Danish gas markets alongside Shell, Fortum, Ruhrgas and EnergiDanmark. In the same document, Statoil Gazelle analyses the marketing strategies of their (potential) competitors: DONG, EnergiDanmark, Elsam, KE and Energi E2.

(598) In the absence of the merger, Elsam-NESA and KE, E2 with each other or with other partners, could have combined these incentives and would thus have been in a very strong position with regard to potential entry into supplies of natural gas to these customer groups. They would also have been available as partners of natural gas retailers competing with DONG and wishing to strengthen their competitive position vis-à-vis DONG by offering their customers a better choice of products. The merger eliminates these extremely well-, if not best-placed, competitors to DONG.

(599) In their reply to the SO, the parties deny - for the following reasons - that KE and/or NESA are potential entrants on the market for supply to decentral CHPs and industrial customers

(a) The Commission ignores the existing relationship between DONG and NESA. NESA is unlikely to act against the interests of a substantial shareholder.

(b) [...]*

(c) The reference to Statoil Gazelle's November 2002 business plan considering KE and Elsam as potential competitors on the market for the supply of gas to decentralised CHPs and large industrial companies is over three years old, and should be regarded as an essentially historic document. Statoil Gazelle's Business Plan was made before the full liberalisation of the Danish gas market and before new competitors had entered the gas market. None of the electricity companies mentioned in the business plan have so far actually entered the gas market. Statoil Gazelle has successfully gained approximately [10-20%]* of the market.

(d) There remain numerous other electricity retailers operating on the market, such as EnergiDanmark, Sydvest Energi, ScanEnergi and NVE/SEAS, that could partner with a competitor gas supplier to DONG in order to provide a dual-fuel package.

(600) The Commission is of the view, that the evidence for the potential entry of KE and/or NESA on the market for supply to decentral CHP's and industrial customers is clear.

The arguments made by the parties are addressed below.

(a) As regards DONG's pre-merger ownership share in NESA, see above in paragraphs (283)-(292).As regards DONG's internal strategy documents see recital (596) (h) above.

(b) Even though Statoil Gazelle's strategy and market assessment document is 3 years old, it still constitutes important evidence. The fact that Statoil Gazelle has a certain market share does not disqualify its pre-merger evidence, quite on the contrary, its market experience makes it more credible. The fact that the forward looking statements of Statoil Gazelle regarding electricity companies have so far not materialised does not mean that they would not have done so in the absence of the merger. DONG's own example concerning Elektra (see recital (592) above) also shows that electricity companies have tried to enter.

(c) As regards DONG's internal strategy documents see paragraph (596) (h) above.

(d) Even though Statoil Gazelle's strategy and market assessment document is 3 years old, it still constitutes important evidence. The fact that Statoil Gazelle has a certain market share does not disqualify its pre-merger evidence, quite on the contrary, its market experience makes it more credible. The fact that the forward looking statements of Statoil Gazelle regarding electricity companies have so far not materialised does not mean that they would not have done so absent the merger. DONG's own example concerning Elektra (see paragraph (592) above) also shows that electricity companies have tried to enter.

(e) The Commission does not deny the possibility of EnergiDanmark, Sydvest Energi, ScanEnergi and NVE/SEAS partnering with a gas supplier; however none of them, with the sole exception of EnergiDanmark can offer a comparable electricity customer base with large business customers to the combined activities of NESA, KE and FE.

(601) In their reply to the SO, the parties furthermore submit that the Commission ignores the emergence (as a result of the merger) of Vattenfall as a wholly independent operator in the market. To the extent that either Elsam or E2 could be said to have been in an advantageous position to enter the supply market as a result of their purchasing expertise in relation to gas, this advantage will now pass to Vattenfall. The disposal of gas-fired generation capacity to Vattenfall will therefore mean that it will have the option of terminating its long-term gas contract with DONG, thereby enabling it to negotiate contracts with third parties, if it so chooses, which may allow it to resell gas to large industrial companies and decentralised CHPs. Finally the parties are of the opinion that unlike Elsam and E2 pre-merger, Vattenfall will be an independent operator on the market, unconstrained by minority participations.

(602) The Commission does not ignore this "Vattenfall effect" and has analysed its impact at many points (in particular by not raising objections on electricity wholesale in Denmark). However, it needs to be considered that Elsam's and E2's advantageous position only passes on to a very minor extent to Vattenfall as Vattenfall only acquires one gas fired central power plant with a very small future consumption, if any consumption at all. Vattenfall will likely not exercise its termination option for the Fynsværket plant as it can benefit from the very low gas price in the contract. On the other hand it is not to be expected that the splitting of this contract and the partial transfer to Vattenfall will provide Vattenfall with significant excess gas, thereby making it much worse-placed to perform flexibility services to customers.

(603) Regarding Vattenfall's two gas fired decentral power plants, it is more likely than not that these are too small to justify own supplies, thereby making own procurement of gas by Vattenfall in Denmark, a more unlikely option. However, even if it were assumed that Vattenfall could profitably decide to supply its own decentral gas plants itself it would lack much of Elsam's and E2's entry advantages, e.g. regarding storage/flexibility. Vattenfall will also need some time to integrate its recent acquisitions. After its entry into German electricity and gas markets it took almost 6 years until it started gas trading operations (406). Also Vattenfall's historic lack of interest in natural gas in Sweden needs to be taken into account. In Germany it sold Hanse Gas, its first acquisition in the German natural gas sector, to E.ON. Finally, and importantly, unlike Elsam-NESA, KE and FE Vattenfall lacks any retail customer base in Denmark.

(604) As to the greater independence of Vattenfall, it should be mentioned that Vattenfall, too, will not be completely independent from DONG post-merger, as they have joint ownership of the wind power facilities of Horns Rev as well as other ancillary facilities. This would, if one were to accept DONG's logic with respect to ownership participations, also make it less likely that Vattenfall will compete with DONG in a new market in a new business area. In any event, even assuming that DONG's point were valid, the weight of the argument is comparatively insignificant in view of all other Vattenfall shortcomings as an effective potential gas competitor in Denmark listed above.

Conclusion on supply to large business customers and decentral CHPs

(605) For these reasons, it is therefore concluded that the proposed operation will lead to a significant impediment to effective competition, in particular as a result of the creation or strengthening of dominant positions on the market or markets for the supply of natural gas to large business customers and decentralised CHPs.

5. Market for supplies of gas to households and small businesses DONG's dominance in two regional areas

(606) A joint market for supply to non-metered customers comprises about 320,000 households as well as about 16,700 small businesses with an annual consumption of around 900 mcm, of which a little more than 50 per cent relate to households.

(607) Sales on this market require a national sales office as well as large investments in sales force, technical service centres, IT equipment, billing systems etc. Customer fidelity and brand loyalty is high and financial switching incentives are as yet low. Barriers to entry are therefore high. This may also explain the fact that no sustained overall entry from competitors has occurred on these markets since market opening on 1 January 2004, despite entry on other gas supply markets. (407)

(608) Denmark is divided into five distributions areas, each area having a regional distribution company (which also is responsible for the regional grid), cf. table 17. The regional distribution companies have both a USO408 company and a trading company. There has been a concentration process among the formerly five distribution companies with the result that there are currently three incumbent competitors on these markets as outlined below, each having above 98% market share in their area.

<emplacement tableau>

(609) DONG has been active in regional distribution since the acquisition of Naturgas Syd in 1999. In 2001 DONG acquired the regional distribution company Naturgas Sjælland.

(610) HNG and MN were originally separate distribution companies serving the area of greater Copenhagen and Mid/North of Jutland respectively. Since 2003 the companies have entered into a close commercial relationship with e.g. a common trading company and a common USO company. They also have a common supply agreement with DONG. During the market investigation, the companies have delivered joint answers to questionnaires, and the other respondents in the market investigation consider HNG and MN as one entity. For the purposes of this decision, the companies are therefore considered as one entity.

(611) Statoil Gazelle is a joint venture between Naturgas Fyn, the incumbent distribution company on the island of Funen and Statoil Denmark. The company was established in 2003 and Naturgas Fyn owns 70 % of the company, Statoil Denmark the rest. It is the only distribution company that does not source its gas from DONG.

(612) As mentioned above, HNG/MN entered into an exclusive supply agreement with DONG in 2003. The underlying reason for the agreement was that HNG/MN was heavily indebted at that point in time. Under the agreement HNG/MN undertakes to source all of its gas requirements with DONG until 31 December 2008. Under the supply agreement, HNG/MN pays different prices according to which customer group HNG/MN ultimately supplies (destination clause). For supplies to non-metered customers, the prices HNG/MN has to pay to DONG are higher than the prices for supplies to metered customers.

(613) On 21 December 2005, the Danish Competition Council took a binding commitments decision, whereby DONG commits itself to terminate the contract as per 1 January 2007. The parties can renegotiate a new supply agreement, which however must not contain exclusivity provisions or provisions as to the final destination of the gas (different prices).

(614) Even though the supply agreement between DONG and HNG/MN is to be terminated as per 1 January 2007, DONG has the right to renegotiate the terms, and in case a new agreement has not been reached on 1 June 2006, HNG/MN has an option to be supplied between 300 mcm and 600 mcm, by DONG on terms specified in the commitment decision. It is therefore highly likely that DONG will continue to be HNG/MN's main supplier after 2007.

(615) In view of their high market shares (409) and the high entry barriers (410) the regional distribution companies are currently dominant in their respective supply areas. DONG is therefore dominant in its two distribution areas in South Jutland and South Zealand.

(616) DONG's dominant position is furthermore reinforced by the fact that DONG controls storages, and access to gas from the Danish part of the North Sea and is dominant on the wholesale market as outlined above. The fact that HNG/MN is supplied by DONG is clear evidence of this upstream strength which limits the ability and incentive of HNG/MN to compete downstream.

(617) The fact that its main competitor HNG/MN is relatively vulnerable financially, while DONG is state owned and benefits financially from its oil revenues, further exacerbates the dominance of DONG on the regional market for supplies to nonmetered customers in South Jutland and South Zealand.

(618) In their reply to the SO, the parties seek to refute that DONG is dominant on regional markets, with the following arguments:

(a) In a historic context, it is only natural that the regional distribution companies in the newly opened gas market are still the largest suppliers.

(b) DONG is not dominant in storage and access to gas and cannot use them to raise rivals' costs.

(c) The market has been liberalised relatively recently. USO tariffs give customers little or no incentive to switch supplier and a considerable part of the gas price comprises tax. Since 2004, customer loyalty has declined, and correspondingly the switching rate of small business customers has increased. Suppliers will have, from 1 January, 2006, the ability to send one bill instead of two bills to the customer.

(d) Denmark is identified as one of the few Member States having some degree of competition for households.

(e) Statoil Gazelle's has initiated advertising campaigns directed towards household customers in the Zealand area. The present competitors to DONG are Statoil Gazelle, Shell and HNG/MN who have engaged in competition for customers outside what have been broadly their traditional regional focus areas.

(619) The Commission maintains that DONG is dominant on two of the five regional areas. The parties' arguments are all rebutted below:

(a) DONG is not only the "largest supplier", but is clearly dominant on the two home markets with extremely high markets shares (above [90- 100%]*). This also shows that the first two years of liberalisation have had virtually no impact on DONG's market share.

(b) As to DONG's dominance on storage and wholesale markets, reference is made to the relevant parts of the decision.

(c) There is no evidence of a marked increase in switching rates on the household and small business customers market. The figures supplied by the parties show that approximately 600 customers switched supplier in 2004, whereas approximately 850 customers did so in 2005. The Commission acknowledges that the one bill option (previously customers switching had to pay two bills, one to the distribution system operator, one to the supplier) may have some positive effect on future switching but this will in itself not lead to a strong erosion of market power. Even if it would lead to increasing switching rates this would perhaps somewhat increase the competitive pressure on the dominant company but not threaten its dominant position. This is also true if, as is likely, the small business' switching rate is somewhat higher than that of households. [...]*.

(d) Nor does the fact that Denmark may be among the minority of Member States in which competition for household customers is already possible as such dissolve the incumbents' continued dominance. Reference can be made to the liberalisation process for electricity, which has much further than for natural gas, and which has not led to the disappearance of dominant positions for small customers.

(e) The fact that SGA tries to win customers in other areas, that HNG/MN allegedly have engaged in competition for customers outside their own areas (which, however, is not in any way substantiated by the parties and is doubtful in the light of other evidence), and that Shell has, recently, started supplying in one of the two possible submarkets (that for small business customers) is not proof of absence of dominance in those areas. If it were so, no dominance could be established in markets with more than one supplier.

(f) Reference is furthermore made to the Judgment of the Court of First Instance in General Electric confirming the importance of high market shares if sustained over a sufficiently long period of time for an assessment of dominance. (411)

(620) [...]* (412).

(621) For these reasons, the Commission considers that DONG is dominant in its two regional areas.

Alternative market definition: DONG's joint dominance on a national market

(622) Only four suppliers are active on the market for supply to small businesses and households. Of these, DONG and HNG/MN are clearly the largest with market shares of [25-35%]* and [55-65%]* respectively, i.e. the C2 ratio which indicates the market shares of the two leading competitors is [85-95%]*. The only other significant competitor, NGF/SGA, has [5-10%]* (413), whereas Shell has only a very minor market share [0-5%]* and is only active on the segment for supply to small businesses. Consequently also the overall degree of concentration in the markets is very high (HHI: 5900), which facilitates the emergence and persistence of a dominant oligopoly. Values on the assumption of separate markets for household customers and small business customers are similar.

(623) As mentioned above, there exist strong ties between DONG and HNG/MN. DONG, as a supplier to HNG/MN knows the gas costs of HNG/MN. DONG also knows the split (metered/non-metered) of HGN/MN's customers. HNG/MN has furthermore an explicitly stated strategy of focusing on supplying its USO customers at reasonable prices and has not been active in trying to gain customers in the other distribution companies' areas. The level of competition, as evidenced by the low switching rates, is very low. All of this leads to high price levels.

(624) The market is very transparent also because there are published prices and discounts from published prices are not a regular market feature. Incumbents know about each switching customer. Any deviation from oligopolistic behaviour could therefore be detected easily. The fact that the product is very homogenous also facilitates tacit coordination about staying out of each other's supply areas.

(625) The market is furthermore characterised by very high customer loyalty rates and, correspondingly, very low switching rates. This leads to high customer acquisition costs, which result in high entry barriers, thus minimising the risk of the emergence of strong mavericks. Furthermore, the strong retaliation potential of DONG and HNG/MN in this high margin market reduces the possibilities of SGA or Shell attacking DONG and HNG/MN in their supply areas. Moreover, SGA is not a strong fringe player (see recitals (563) - (565)) whereas Shell is not present in the household segment and has only shown minimal activity in the market for small businesses.

(626) Moreover, there are effective retaliation mechanisms against a deviating oligopolist (namely by targeting customers in the other oligopolists' area, thus exerting pressure on the incumbent's margins) and a high joint interest in maintaining the status quo. This is because margins in the market(s) are high compared with margins in the other gas supply markets, which makes retaliation by competitors especially painful in terms of financial losses. This combined with the fact that the strongest player in terms of market share, HNG/MN, is in a weak financial position and will not, also for this reason act aggressively (except in retaliation scenarios) and the fact that retaliation for all market players is easy to implement leads to the conclusion that the market situation on an assumed nation-wide market can be characterised as one of joint dominance.

(627) In its reply to the SO DONG submits that the Commission, for the following reasons, has not established that the criteria for joint dominance as set out in the judgment in Airtours (414) have been met.

(a) The close links between DONG and HNG/MN have effectively been broken following the recent decision of the DCA. Furthermore, HNG/MN is actively competing for customers in DONG's traditional distribution area.

(b) Contrary to the Commission's assertion, the market is not transparent. Especially on the business segment, prices are not published, but negotiated.

(c) The market has been liberalised relatively recently. USO tariffs give customers little or no incentive to switch supplier and a considerable part of the gas price comprises tax. Since 2004, customer loyalty has declined, and correspondingly the switching rate of small business customers has increased. Suppliers have, from 1 January 2006, the possibility of sending one bill to the customer.

(d) There are no effective retaliation mechanisms since customers are reluctant to switch.

(628) The Commission maintains that, as regards the notified concentration, joint dominance is established to the requisite legal standard.

(a) The links between HNG/MN and DONG have not yet been broken and it remains to be seen whether - which in the Commissions view is unlikely - they will be fully broken, see also recitals (567) and (614) above.

(b) The assertion that HNG/MN are actively competing outside their own areas is not in any way substantiated and contrary to other information known to the Commission. For instance the mission of HNG/MN is stated in its 2004 annual report as the following:

"HNG and Midt-Nord are municipally owned companies undertaking the distribution and supply of natural gas in the owner municipalities in an economic and efficient way....In their capacity as municipally owned companies, HNG and Midt-Nord focus their activities on local social and environmental aspects."

(c) At least the household part of market is indeed very transparent. On www.gasprisguiden.dk all prices for customers with an annual consumption below 10,000 m3 can be compared. Given the very low switching rates, the companies are aware of who switches and whereto. It is correct that for small business customers, prices are not published. Shell is only active in this business segment/market. However this does not, in itself, make the market in transparent. The incumbents have a very good knowledge about past customers and are able to detect very quickly any targeted action to acquire customers. In addition the DSOs, which are accounting-unbundled affiliates of the incumbents, know exactly to which company any switching customers have switched. Such switching customers would not have a strong incentive to keep prices confidential. In the hope of even lower prices by the incumbent they (or a sufficient percentage of these customers) will be prepared to show the incumbent any price offer they have received.

(d) The high margins in this market, which are also acknowledged by DONG, could be used to make it more attractive for customers to switch. Therefore, retaliation possibilities are high if companies are willing to give up part of their high margins.

(629) [...]* (415).

DONG's dominance will be strengthened

By raising entry barriers

(630) As explained above, the concentration will remove Energi E2 and Elsam as the customers on the natural gas markets. This raises barriers to entry not only on the markets for supply to central CHPs and to large business customers and decentralised CHPs but also on the market(s) for supply to small business customers and households due to smaller overall incentives to enter the Danish (or Danish-Swedish) wholesale market and due to smaller incentives to enter retail supply markets due to higher storage costs and to greater difficulties in achieving critical size and achieving the economies of scale and scope necessary for entry to be financially viable. Estimates on the minimum viable size by market participants have been given which assume a value of several hundred mcm for the non-metering customers market(s). Given that the total size of the non-metering market(s) in Denmark is only about 900 mcm it is clear that a company wishing to enter would have to sell gas in other Danish gas markets as well.

(631) Furthermore, the operation removes very large customers on a different market whose flexibility could well be balanced against the flexibility requirements of household and small business customers. Thus, it would be possible to sell more gas to central CHPs in the summer when small customer demand in Denmark is considerably lower which makes access to decentral CHP customers more difficult. (416) Secondly, as also shown above the operation also strengthens DONG's dominance on the market(s) for supply of natural gas to large business customers and to decentral CHPs (the latter also partly due to a type of customer foreclosure effect), which makes joint entry into that market and the market(s) for non-metering customers less profitable and less likely. The raising of entry barriers in other Danish natural gas markets therefore has, as a knockon effect, the consequence of also raising barriers to entry on the non-metering market(s).

(632) Furthermore the likely reduction of liquidity on the Danish wholesale gas market described above increases the ability of DONG to foreclose access to wholesale gas to its competitors thereby further raising the costs and entry barriers for potential entrants to the market for the supply of gas to small business and household customers.

(633) In its reply to the SO DONG denies that entry barriers will be raised, for the following reasons:

(a) There is no connection between changes in upstream markets and markets for larger customers and the ease of entry to the small business/household market(s).

(b) If the same suppliers are likely to enter the wholesale market, the market for supply to CHPs and the market for supply to small business customers and households, then it is clear that any foreclosure effect should not be measured purely against the demand for gas from central CHPs.

(c) If the customer foreclosure effect is considered with reference to the market for supply to central CHPs only, then it has not adequately explained how an effect on the market for the supply to small business customers and households will arise. Minimum viable size for the nonmetering customer market(s) is not substantiated.

(d) Supply to small businesses and retail customers can itself justify entry into gas retail independently of any potential opportunity to supply to central CHPs. Despite the fact that central CHPs are not contestable customers, entry has nevertheless occurred on other gas markets, including those for small business customers and household retail customers rather than central CHPs would be considered as more attractive to a new gas supply entrant, because of higher margins and smaller quantities needed.

(634) For the following reasons the Commission does not agree with DONG's view:

(a) This decision establishes an explicit link between the wholesale market and retail markets for instance in the section on customer foreclosure. Easy access to gas is a requirement for activity on the retail market. The dominance of DONG on the wholesale market therefore puts it in a position to act more independently on the small customers' retail market(s). To satisfy its customers' storage/flexibility requirements is an obvious need for any retail supplier. The mere fact that DONG is both the storage owner and the biggest storage customer makes it easier for DONG as a retail supplier to satisfy these storage needs, even in the absence of any discriminatory behaviour as there will for instance always be an existing standard storage contract that allows for a sufficient flexibility margin for even a large number of changes in DONG's large customer portfolio. This is not necessarily the case for other smaller suppliers for whom individual changes have a much higher relative importance and which therefore may have to negotiate extra contracts, which on the basis of information provided by DONG are somewhat more expensive (even if these higher prices were to be applied in a non-discriminatory fashion). If, as a merger-specific effect, the prices for these contracts were raised (as discussed above in the section on storage/flexibility), new entrants would be likely to suffer most. If as a merger specific effect access to wholesale gas becomes more difficult or if the Danish wholesale spot market becomes more illiquid (as discussed above in the section on wholesale) entry will be more difficult for newcomers, thereby raising entry barriers. Both of these effects have been shown in the relevant sections above. Reference is also made to the sections at paragraphs (338)-(373) and (453)-(532) above.

(b) Reference is made to general questions and foreclosed volume in the customer foreclosure part. The Commission does not disagree with DONG's argument but fails to understand how this contradicts the Commission's assumption that even if customer foreclosure is measured against the whole Danish gas demand this obviously also has an impact on all parts thereof.

(c) Reference is furthermore made to the findings on minimum viable scale in the customer foreclosure part. It is true that minimum viable scale for small customers is not substantiated. However, as the foreclosure effect generally makes it more difficult to overcome minimum viable scale obstacles, the exact size of any minimum viable scale is of less importance. This is irrespective of the question whether minimum viable scale for the small customer market(s) itself is smaller than for other gas supply markets, which in any event is unlikely (see point (d) below).

(d) As explained in recital (607), there are high entry barriers (e.g. investments) to the market(s) for supplies to small business and household customers. Therefore, this market/these markets cannot be seen as more attractive to a new gas supplier than supplies to central CHPs and/or large industrial customers/decentralised CHPs, which is also confirmed by the entry of e.g. E.ON on the other markets, and the absence of entry by Shell on the household customer market. It is true, that Shell has entered the markets for supplies to large and small industrial customers, but Shell is still only a marginal player, for whom according to the Commission's information there is no indication that it has any intention of entering the household segment or market. Apart from Shell, no one has entered the small business customer market. (417) No company which has entered the market for supply to small customers only. Therefore, and in the absence of contrary evidence, it can be assumed that independent entry into the small customer market(s) alone is not a viable option or at least one for which entry barriers are higher than for combined entry with other markets, most importantly the one for supply to central CHPs.

By eliminating potential competition

(635) A further strengthening of DONG's dominant position could emerge from the fact that two potential competitors KE and, to some extent, NESA (part of Elsam) are eliminated.

(636) NESA and KE are electricity retailers, both with a base in the greater Copenhagen area. They have already sales offices, large sales forces, IT equipment and billing systems in place. They have large customer portfolios (418), which could be built upon with the aim to enter gas retail markets. Both companies also have considerable brand strength both at a national and especially at a regional level. Thus, NESA and KE do not face the high barriers to entry that most other entrant would meet.

(637) In addition, KE already has experience in selling citygas which from 2007 will consist of 50% air and 50% natural gas, in its area. It is thus already a multi-utility energy company and one of its strategic goals is to become one of if not the leading Danish multi-utility company, thereby entering into direct competition with DONG which has pursued a similar objective and has invested strongly in brand building.

(638) Since its acquisition by Elsam in 2004 NESA has had access to a considerable amount of natural gas via Elsam's power plants.

(639) On the other hand, the fact that KE already sells city gas, heating and electricity limits to some degree the potential of KE to use its existing customer base to also supply natural gas, since many (175,000) of its customers will already have their energy requirements for cooking covered by city gas and their heating requirements covered by district heating (97% of Copenhagen's heating requirements are covered by district heating supplied by KE (419)).

(640) It also has to be taken into account that NESA's and KE's distribution areas do not overlap with those of DONG. NESA and KE have to some extent acquired electricity customers outside their areas which could have been used to compete with DONG for gas supplies to small businesses and households, but the bulk of NESA's and KE's electricity customer base could only have been used to compete with HNG/MN whose gas distribution areas geographically overlap with those of NESA and KE, and not with those DONG.

(641) On an assumption of regional markets, NESA and KE could therefore have used only part of their electricity customer base to compete with DONG for gas supplies to small businesses and households directly. Accordingly, the size of their electricity smallcustomer base to some extent exaggerates the real potential competitive threat exerted on DONG by KE and NESA.

(642) However on an assumed national market for supplies to small businesses and households, KE and NESA could have used the potential of their electricity customer base to compete with the jointly dominant companies for gas supplies to small businesses and households, and could thus have had a destabilising impact on the oligopoly, only subject to the caveats regarding KE's supplies of city gas and district heating mentioned above at recital (639).

(643) [...]* NESA or especially KE could have entered this market in the absence of the merger.

(644) [...]* (420),[...]* (421).

(645) Also DONG considered prior to the merger that KE or NESA could be a future competitor on retail gas markets to small customers. [...]* (422).

(646) [...]* (423) As mentioned above, not only DONG and KE but also Statoil Gazelle considered KE to be a potential competitor on Danish Gas markets.

(647) Indeed DONG's own multi-utility strategy shows that there are considerable advantages in the form of synergies to be gained by the supply of to retail customers, both to large customers and to small customers.

(648) The Commission's market investigation has confirmed that interest by customers in dual fuel (gas/electricity) offers is considerable. In light of this, Elsam-NESA and especially KE could, in the absence of the merger, and perhaps in a different form of consolidation scenario from the one proposed, have been entrants into gas supply via dual fuel offers both to some of their existing customers and to new ones. In their reply to the SO, the parties doubt whether the ability to offer dual fuel products provides a genuine competitive advantage. In the view of the parties, customers in Denmark are often (and increasingly) provided with district heating. Furthermore, the parties argue that small businesses are also sensitive to price and would not be prepared to accept in any way inferior terms from their supplier in order to obtain dual-fuel supplies. Finally the parties consider that there are only marginal advantages (cost savings) in being able to provide dual-fuel.

(649) The Commission takes note of the fact that DONG in its notification (Form CO) itself has suggested that small customers are more likely to be interested in dual fuel offers than large customers. The Commission acknowledges that many small customers may be served by district heating, as also mentioned above in recital (639). But there are still more than 300,000 households in Denmark that are heated by gas. Furthermore, cooking requirements cannot, outside city gas areas, be satisfied by district heating.

(650) The fact that small businesses are also, to some extent, price sensitive cannot be used as an argument against advantages of dual fuel supplies, especially in the case of DONG, who with gas and electricity in-house would be one of only a few companies, and would certainly be the strongest company in terms of retail sales, able to offer dual fuel at competitive prices. Reference is also made to the discussion of this point above in recital (593).

(651) [...]* Reference is made to the discussion of the issue of advantages of dual fuel supplies at recital (589).

(652) In their reply to the SO, the parties submit that the reference to considerable brand strength is misplaced, since the KE brand does not form part of the notified operation.

(653) The Commission acknowledges that the KE brand will remain with the local authority of Copenhagen. However, more importantly, the commercial companies of KE, namely KE Marked, KE Kunde and KE Tele are all transferred to DONG, whereas the local authority of Copenhagen only keeps the USO companies, including city gas, water, heating and drainage but not electricity. Therefore the ability of the part of KE remaining with the city of Copenhagen to become a competitive multi utility supplier will be diminished as a result of the merger. NESA also has a strong brand as part of the Elsam group, cf. discussion above at recital (596).

(654) In their reply to the SO, the parties deny that there exists credible evidence that either NESA or KE are potential competitors since neither NESA nor KE have entered retail gas markets. The parties further argue that the Commission's reliance on a series of outdated internal strategy papers is irrelevant, given the subsequent market developments.

(655) The Commission acknowledges that neither NESA nor KE have so far entered the retail gas market. However, the Commission is of the view that the internal papers, which indicate that such entry is quite likely, are not outdated. They are from [...]* and when assessing the actuality of the documents it needs to be taken into account that the this merger has been planned at least since early [...]* and that KE has been put up for sale even longer, with both DONG and Vattenfall having shown interest also before this merger. It is likely that such sales processes and plans have had a stifling effect on important business decision such as entry into gas markets, especially if one of the likely buyers already has gas retail as a core business.

(656) The Commission also acknowledges that the file only contains indications of entry into gas retail by KE and not by NESA. However, in the absence of the merger NESA would have had similar incentives to KE and even greater ability to do so (easy access to gas through its controlling parent company Elsam).

(657) DONG further submits in its reply to the SO that there are other potential entrants such as EnergiDanmark, Scanenergi or Nordjysk Elhandel.

(658) The Commission acknowledges that other electricity supply companies could enter the market for supply of gas to small business and household customers. The Commission also acknowledges that there might be other potential competitors (or alliances of competitors) with broadly comparable small-customer customer portfolios to those of KE and NESA, especially when taking the issues of city gas and district heating mentioned above under recital (639) into account.

(659) However, the Commission is of the view that KE and NESA have a special position when it comes to access to gas, that would be difficult for other electricity companies to duplicate. KE has privileged access to gas through its major interest in E2 and NESA through its parent company Elsam. NVE-SEAS would also pre-merger have had privileged access to gas through its shareholding in E2, and this is also true for some of the other electricity companies holding shares in Elsam. All of these possibilities would, however, be removed by the merger.

(660) The Commission therefore concludes that the merger, to a certain extent, eliminates potential competition in particular by eliminating one (KE) and possibly two (NESA and KE) potential competitor(s) while at the same time drying up the sources of competitive wholesale gas accessible to alternative potential entrants.

(661) The effects outlined above will take place not only on the regional markets where DONG is dominant, but on all regional markets in Denmark thereby also jeopardising increases in competition in an increasingly liberalised Danish market. While prior to the merger KE and NESA would have had incentives to enter these areas with a dual fuel (natural gas/ electricity) offer, this incentive will post merger likely be reduced as DONG can be expected to be unwilling to risk retaliation or significant weakening of in particular HNG/MN (to whom it may expect to continue to sell on wholesale level, achieving similar margins with less effort)424. Entry by DONG (through NESA and or KE) into HNG/MN's territories would thus be likely to have a negative effect on its own overall (upstream and/or downstream) margins and ultimately profits. This analysis therefore applies irrespective of whether the small customer market(s) is/are seen as regional or as national.

Conclusion on the supply of natural gas to households and small business customers

(662) It is therefore concluded that the proposed merger will significantly impede competition on the market or markets for the supply of gas to households and small business customers, irrespective of whether these markets are/this market is defined as regional or as national, in particular through the strengthening of a dominant position.

6. Conclusion on the Assessment of the Natural Gas Markets Prior to Consideration of Modifications to the Notified Concentration

(663) The Commission concludes that the transaction leads to a significant impediment to effective competition, in particular through the strengthening of dominant positions on the following product markets:

- Supply of wholesale gas for Denmark (and potentially also for Sweden)

- Gas storage or gas flexibility (irrespective of whether only for Denmark or also for Sweden)

- Supply of gas to large business customers and decentral CHPs, as 1 or 2 markets, in Denmark,

- Supply of gas to small business customers and/or households, as 1 or 2 markets, in Denmark

(664) For these reasons, the Commission has come to the conclusion that the concentration as notified would significantly impede effective competition in a substantial part of the common market within the meaning of Article 2 (3) of the Merger Regulation, and that it would thus be incompatible with the common market and the functioning of the EEA Agreement unless the Commission finds, pursuant to Article 8(2) of the Merger Regulation that following modification by the undertakings concerned the notified concentration fulfils the criterion laid down in Article 2(2) of this Regulation, in which case it is to be declared compatible with the common market. As DONG has submitted such modifications on 1 March 2006, a date within the statutory deadline provided for in Article 19 (2) of Commission Regulation (EC) no 802/2004 (425), this question will be examined in the remedies section of this decision.

VII/3 PART B ELECTRICITY

1. Electricity wholesale

(665) Generation of electricity in Denmark is primarily based on coal or gas-fired CHPs. Some CHPs are also biomass-fired. A relevant distinction can be drawn between large central CHPs, operated by Elsam and E2 and small decentralised CHPs operated only to a lesser extent (426) by Elsam and E2 and having a dispersed ownership. A certain amount of Danish electricity production is generated by windmills. The shares of different fuels /sources of input in electricity generation in Denmark are as follows:

Coal 45%

Gas 26%

Windmills 17%427

Bio fuel 5%

Waste 3%

Oil 3%

(666) The role of natural gas in power production increased significantly in 1990s where its share rose from less than 3% in 1990 to almost 23% in 1999 (428). Roughly half of the natural gas used in Denmark is used by heat and power producers and, of that, roughly half is used by decentralised CHPs, one third is used by central facilities and the rest is mainly used by private heat and power producers. In addition to gas (47.0 PJ in 2004), decentral CHPs also use renewable sources and waste as input (18.8 PJ in 2004).

- Assessment on the assumption of separate East Dk and West Dk wholesale markets - horizontal effects

(667) E2 is the generation incumbent in East Denmark and Elsam is the generation incumbent in West Denmark. Their respective shares of all electricity generated in their areas in 2004 were as follows: [60-70%]* (Elsam, West Denmark), [70-80%]* (E2, East Denmark). The remainder is produced by a large number of mainly municipality-owned or cooperatively-owned decentral CHPs and by third-party wind power. Those decentralised CHP plants who are obliged to sell their electricity on the open market (the current threshold of >10 MW will be lowered to >5MW) tend to place price-independent bids at Nord Pool with the help of Nord Pool agents (such as Markedskraft, Nordjysk or DONG) aiding them, against a fee, in this process. Due to their low flexibility of electricity production, they tend to sell at any price as they produce electricity as a 'by-product' whenever they need to produce heat. (429) This inflexible behaviour places only limited constraint on Elsam and E2 whenever their respective areas are isolated from other Nord Pool areas. In addition the capacity of these operators is insufficient to meet Danish demand. As regards windmills, the production is either purchased by the TSO at prices provided for by national law (430), and thereafter sold to Nord Pool or sold directly to Nord Pool on price-independent bids, i.e. bids that will always be paired with demand. Due to these circumstances, the effective constraint which the windmills and the decentralized CHP plants can exert on Elsam and NESA will necessarily be low. Prima facie therefore E2 and Elsam pre-merger hold a dominant position in their respective Nord Pool areas in electricity wholesale whenever these areas are isolated from other Nord Pool areas. (431)

(668) The notifying party argues that post merger, this market power will be reduced as Elsam and E2 will not be combined in their entirety but a substantial amount (about 25%) of their generation capacity both in East Denmark and in West Denmark will be divested as part of the proposed transaction to Vattenfall (432). The Vattenfall transaction, although a separate concentration also subject to Community merger control (433), is conditionally linked to the acquisition of the remaining parts of Elsam and E2. The notifying party argues that this divestiture of parts of Elsam and E2 to Vattenfall is actually a "pre-merger remedy" to competition problems which may otherwise have arisen. Tables 18 and 19 indeed show that the merger leads to very small horizontal overlaps in both East and West Denmark, which will most likely be substantially surpassed by the loss of market share due to the Vattenfall divestitures in both East and West Denmark.

<emplacement tableau>

(669) The proposed transaction, through the conditionally linked Vattenfall transaction, thereby creates a viable competitor to the area incumbents, reducing their market shares substantially. The competitive loss in potential competition and in (the very small actual competition) is indeed likely to be substantially lower than the competitive gain by the additional competitor Vattenfall.

(670) To establish this, the Commission has considered the likelihood of large power plants being built in Denmark. The views expressed in the course of the Commission's market investigation show that as Denmark is already a net-exporter of electricity and as no strong demand increase is expected in Denmark no major newly built capacity is expected in Denmark in the foreseeable future. If anything, a conversion of existing coal and especially oil-fired capacity (owned by Elsam and E2) to other fuels (gas or biomass) or more modern generation technologies was considered possible. A further increase of the percentage of wind power and biomass generation in the Danish generation mix was considered likely, whereas no clear picture emerged as to the likelihood or not of higher gas consumption for the purpose of power production in Denmark in the future. Therefore the general probability for any green-field entry in Denmark with new large coal or gas fired power plants in competition with Elsam's and E2's current plants is very low.

(671) The Commission has considered the extent to which, in spite of the general unlikelihood of any entry, (i) Elsam and E2 would have been potential competitors in each other's markets and (ii) DONG would have been likely to enter the East and West Danish markets to a greater extent. In this context, the Commission has examined internal documents. No intentions by Elsam and E2 to enter into each other's territories could be found. A study by DONG has examined but rejected the possibility of entering electricity generation via a gas-fired power plant. As the study dated from 2003, a sensitivity analysis to today's changed circumstances has been conducted which has led to the result that the economic incentive for DONG to enter gas-fired electricity generation in Denmark has not improved and that it is unlikely to improve in the foreseeable future. Therefore the elimination of the potential competition by Elsam, E2 and DONG cannot counterbalance the much more certain and immediate positive effect of Vattenfall's entry.

- Assessment of vertical effects on the assumption of separate East Dk and West Dk wholesale markets

(672) The Commission has also investigated whether the vertical integration of Elsam and E2 with DONG, the dominant supplier of an important fuel in electricity production, could lead to negative effects on competition. DONG is the main supplier of gas to decentral CHPs in Denmark, which prior to the merger constitute a real, albeit weak, competitive constraint on Elsam and E2.

(673) Potential negative vertical effects could either take the form of (a) DONG post-merger having the ability and incentive to raise the cost of gas to decentral CHPs thereby reducing the competitive constraint on the merged entity's internal electricity production, or by (b) DONG directly or indirectly controlling the behaviour of decentral CHPs through their contractual relationship.

(674) The Commission has found that neither of the two potential types of vertical concerns are likely to lead to a significant impediment to effective competition.

(a) raising rivals' cost

(675) The Commission initially notes that the decision by decentral CHPs whether to operate or not is essentially determined by the local demand for heat. The heat production is sold at a cost covered basis and the electricity has historically been offered to Nord Pool as price-independent production. In a fully liberalised setting, decentral CHPs will participate on Nord Pool on a market basis. While electricity is essentially a by-product of the heat production, there is some flexibility in some decentral CHPs as to when exactly to produce heat and electricity (the heat necessary can be stored for short periods in accumulators) or as to whether to be able to produce electricity when no heat is needed (through cooling facilities) or possibly even whether, when in operation, to produce electricity at all.

(676) The relevant question is thus to what extent the merged entity would be able to induce the decentral CHPs to reduce electricity output by raising their input prices. (436) Energinet.dk has assisted the Commission in assessing the likely effect on the electricity production from a potential increase in gas prices for the decentralised CHPs. This is done within the framework of the SIVAEL (437) model, which contains very detailed information of each production facility in Western Denmark (438) including the fuel efficiency, the heat storage facilities and the technical abilities on each production site to produce heat only or heat and electricity combined.

(677) The simulation in the SIVAEL model confirmed that in order for the reduction in output to be of any importance very significant increases in gas prices would be required. It is the Commission's view that such increases based on merger-related vertical effects would not be realistic in the light of the competitive constrains which, despite its dominant position, would nevertheless exist on DONG after the merger. This conclusion is reinforced by the remedies offered by DONG (see below in the Remedies section) which will increase the availability of alternative sources of gas to decentral CHPs after the merger.

(678) Furthermore, even if the merged entity post merger had an incentive to seek to increase prices of gas to decentral CHPs, it is unlikely that this would translate into higher electricity prices in the light of the increased competition in East and West brought about by the divestiture of the power plants to Vattenfall (439).

(b) direct or indirect control of decentral CHPs

(679) For some decentral CHPs the current contractual relationship with DONG effectively corresponds to DONG obtaining control of the electricity production including balancing of the plant (subject to the heat production requirements) including the economic risks relating to the production. For other CHPs, DONG is controlling the production of electricity including balancing (subject to the heat production requirements), but these other CHPs do not hold the economic risk relating to the electricity production (440).

(680) In 2002 DONG and Elsam discussed forming a joint venture, the purpose of which was to offer risk-covering services to the decentral CHPs once these were going to operate under market based conditions. These services would include guaranteed prices for the electricity, for gas purchases, selling of balancing services etc. [Contains information from internal papers]* (441). The joint venture never materialized, but the project and its assessment by Elsam illustrate that contracts can allocate the strategic risk from the CHP to the supplier of gas in a way that amounts to a situation similar to that of complete control of electricity output.

(681) However, when the Danish Competition Authority in March 2004 approved the acquisition by Elsam of NESA, Elsam and NESA undertook to divest all their shares in decentral gas-fired CHPs corresponding to a production capacity of 230MW. In addition Elsam is required to abstain from running or controlling decentral CHPs for 121/2 years. Due to this commitment, DONG is unlikely in the future to be able to exercise any direct influence on decentral CHPs.

(682) The consideration of potential vertical effects of the proposed concentration therefore does not change the finding of the absence of any significant impediment to effective competition on separate East Danish and West Danish electricity markets.

- Possibility of a separate market for bilateral wholesale supplies

(683) The above considerations and findings concerning horizontal and vertical effects of the proposed merger also apply to bilateral wholesales of electricity to customers within both East and West Denmark assuming a separate market for wholesale customers who may not have direct access to the NordPool wholesale market. Irrespective of what the parties' pre-merger market shares on such a potential market are, their market power will be more effectively constrained after the merger due to the creation of a second important source of physical supply within both East and West Denmark, Vattenfall.

- Assessment on wider than Danish markets

(684) In the Nord Pool zone, and even in a Nord Pool zone encompassing only Sweden and both parts of Denmark, Elsam and E2 are comparatively small companies, with a combined market share of below 10%, while the most important players on a pan- Nordic scale are Vattenfall, Statkraft (Norway) and Fortum (Finland). The notified operation thus cannot have any significant negative impact on such a wider market.

(685) The impact of the conditionally linked Vattenfall divestment on such a market has been examined in a separate decision (442) and was also found not to lead to any competition problems. Vattenfall's share on such "worst case" Danish-Swedish markets was only slightly increased. In addition, such constellation was rare compared to constellations including also either Norway or Finland or both and thereby further diluting Vattenfall's market share (and market power) on such possible wider markets.

- Assessment post 2010 (Great Belt Interconnector)

(686) While the overall effect of the merger until 2010, the foreseen date for the entry into operation of the Great Belt Interconnector cable, is thus clearly positive, the effect of the merger on the otherwise positive impact of this Great Belt Interconnector needs to be examined. In the absence of the merger, such an interconnector would likely have led to increased competition in both East and West Denmark by providing the opportunity to Elsam and E2, in particular, but also to other operators, to directly 'export' (via NordPool) electricity into the other Danish area. This would have had a constraining influence on the incumbent in this other area. After the merger such incentives may be reduced as it cannot be excluded that DONG and Vattenfall, the main players in both areas, will not use the Great Belt Interconnector to the same extent as Elsam and E2 would have done in the absence of the merger.

(687) However, even assuming complete absence of such incentives for Vattenfall and Elsam (which is likely only in a collusive situation) other players would be left who, e.g. with price independent bids, would be able to exert at least some degree of competitive pressure on the neighbouring Danish area even in situations in which East Denmark and West Denmark were isolated from the rest of NordPool.

(688) What is even more important is that the Great Belt Interconnector is not expected to have a capacity of significantly more than 600 MW. Therefore the constraint that could be provided, in the absence of the merger through these 600 MW from the neighbouring Danish area will, through the merger and the conditionally linked Vattenfall operation, exist not only much earlier but - as regards West Denmark - to a much higher degree and - as regards East Denmark - to the same degree within the same area. (In West Denmark, Vattenfall will acquire a capacity of 1351 MW of central CHPs, Fynsværket and Nordjyllandsværket, and a total of 309 MW land-based and sea-based wind power. In East Denmark, Vattenfall will acquire 611 MW, of which 477 MW central CHPs). The net effect of these constraints (i.e. the hypothetical, though unlikely, complete neutralisation of the competition-enhancing effect of the Great Belt Interconnector balanced against the earlier competition enhancing effect stemming from competition through Vattenfall's newly acquired capacity) can therefore be expected to be a positive one even after 2010.

(689) This has been confirmed by a model simulation carried out by the Danish TSO Energinet.dk comparing an absent-the-merger scenario in a post 2010 situation (i.e. assuming the operation of the Great Belt Interconnector and area incumbencies by E2 and Elsam) with a post-merger scenario in 2010 (i.e. assuming the operation of the Great Belt Interconnector and DONG and Vattenfall operating both in East and West Denmark). The results of this model have shown net positive effects of the postmerger situation on Denmark.

2. Ancillary services (690) Regarding ancillary services, the Commission's market investigation has shown that

Elsam and E2 have had to face little competition in their respective areas. Postmerger DONG will be faced, in each area, with a serious competitor - Vattenfall. Even if an allocation of a specific market share to the assets acquired by Vattenfall is not possible (as Elsam and E2 have not provided these basis in a form attributable to individual plants) the Commission has verified that the assets acquired by Vattenfall are suitable for providing a very wide range of competitive ancillary services. The merger, therefore, is very unlikely to lead to competition problem on the markets for ancillary services in East Denmark and West Denmark as the competitive pressure exerted by Vattenfall on the merged entity will in any event be larger than the competitive pressure with which Elsam and E2 were faced prior to the proposed concentration, in their respective areas.

3. Financial derivatives of electricity

(691) On a Nordic market for trading in financial derivatives of electricity the merging parties have a combined market share of < 10%. This small market share dispels all concerns on such a market.

(692) Regarding potential area markets for CfDs for East and West Denmark, E2 and Elsam are clearly the major sellers of CfDs in their respective areas. (443) This indicates that it is in the interest of producers (and particularly if, as in the case of E2, these producers do not have final customer sales in the area) to sell CfDs, e.g. in order to hedge against the risk that area prices could fall below the level of the average NordPool price level. Both in East Denmark and in West Denmark, the operation will, in the first place, lead to some minor increase in this share of all sales as other parties to the concentration have also sold CfDs. However, it is to be expected that the divestment of generation assets to Vattenfall will have the effect of introducing Vattenfall as a major seller of CfDs in both East and West Denmark. (444)

(693) The operation will therefore not lead to any significant impediment to effective competition on any possible market or submarket of financial derivatives of electricity.

4. Retail supply of electricity to business customers

(694) The market investigation has confirmed that the current competition level in the Danish market for electricity supplies to metered customers is high. The merged entity will combine the second strongest (Elsam-NESA) and the third strongest (KE) competitors on the market with DONG's and FE's much smaller activities. DONG has entered the markets (through the acquisition of a small local electricity supplier and distribution company and other nationwide market entry) and has gained some small market share. FE only has a very small customer base. The parties' post-merger combined market shares will be [20-30%]*.

<emplacement tableau>

(695) After the merger, the merged entity will largely close the gap that currently exists between the number 1 player, EnergiDanmark, and NESA. Energi Danmark is a trading company formed jointly by some West Danish local incumbents which has since acquired two East Danish competitors (Disam and Elektra).

(696) A number of further competitors are considered viable alternatives by customers. These include SEAS-NVE and Nordjysk Elhandel (each with around 5%) but also new entrants which do not have links to distribution companies such as OK, whose growth rate has been high, or Scanenergi. A large number of small companies, often affiliated to local USO companies will remain in the market (e.g. Sydvest Energi, Energi Midt, Energi Nord, Elgas, Elro or NOE). Entry by Vattenfall is also possible and has become more likely as a result of the proposed transaction.

(697) While the parties, in particular DONG, Elsam/NESA and KE have, to some extent, competed with each other, the market investigation has not shown strong unilateral effects between the parties to the concentration. Customers have not complained of a scarcity of offers.

(698) For these reasons the concentration does not give rise to competition concerns on the market for supplies of electricity to Danish business customers.

5. Retail supply of electricity to small standard load profile customers

(699) The operation will also lead to some market share overlaps in the market for retail supply of electricity to small customers, mainly due to the combination of Elsam- NESA's activities with those of KE, whereas the operations of FE and of DONG (who has recently entered the market) are rather small.

<emplacement tableau>

(700) The competitive landscape in this market in Denmark is quite fragmented with only a small number of household customers having so far switched from their incumbent local supplier (2% in 2003, 1% in 2004 and 0.5% in the first half of 2005).

(701) On a national, Danish, market DONG will become the largest competitor, albeit with a limited market share of around [25-30%]*, followed by EnergiDanmark (and its 6 parent companies) with around [15-20%]*. SEAS-NVE is the third largest competitor with around [10-15%]* market share, whilst 5 further competitors have market shares between 5% and 10%. In addition to the old regional incumbents a number of new suppliers, e.g. OK, are also offering their services to small Danish customers.

(702) The Commission has analysed the companies gaining customers through switching and has found that although NESA and KE were among the more important companies to which customers have switched, other competitors such as in particular EnergiDanmark (447) and OK (448), which entered in 2004, have comparable customer acquisition rates.

(703) On the assumption of regional markets, the Commission has verified whether the operation would lead to a significant strengthening of NESA's, KE's and FE's positions in their respective incumbency areas. The information received has shown that third party competitors such as OK and EnergiDanmark are as well placed competitors to the area incumbents as are the other parties to the merger. The information also shows that a previous market exit by an important new supplier (namely Vattenfall's exit in 2003) did not lead to an increase in the incumbents' market share but rather could be compensated by increased market shares of other entrants.

(704) In the light of this, the operation does not lead to a significant impediment to effective competition, in particular as a result of the strengthening of a dominant position on either a national market or regional markets for the supply of electricity to small customers in Denmark.

5. Conclusion on competitive assessment for electricity markets

(705) For the reasons stated above, the concentration does not lead to a significant impediment to effective competition, in particular as a result of the creation or strengthening of a dominant position on any possible affected market for electricity wholesale, ancillary services, financial derivatives of electricity, retail supply to business customers or electricity supply to small customers.

VIII. COMMITMENTS

(706) With a view to addressing the competition concerns set out in this decision, DONG has submitted a set of Commitments ("the Commitments") to the Commission on 30 January 2006. Those Commitments were amended on 1 March 2006. They are composed of two parts: a storage divestiture and a Gas Release Programme.

A. DESCRIPTION OF THE COMMITMENTS

1. Storage divestiture

(707) With a view to addressing the competition concerns on the storage/flexibility market, DONG has submitted a commitment to divest its gas storage facility in Lille Torup in Jutland. This is the larger of DONG's two storage facilities, and has a working volume of at least 400 million m3 (mcm) and a total storage gas volume of 710 mcm. Its daily injection and withdrawal capacity amounts to 3.6 and 7 mcm, respectively.

(708) The Divestiture Business encompasses the personnel and all assets related to the Lille Torup storage facility. These assets include all tangible and intangible assets (including intellectual property rights), all licences, permits and authorisations, and all contracts, leases, commitments and customer orders as well as all customer, credit and other records.

(709) In the amended Commitments DONG commits to enter into an Interconnection and Operating Balancing Agreement with the Danish TSO Energinet.dk with a view to ensuring that conditions for use by Energinet.dk of DONG's storage facilities at both Danish storage facilities are not made less favourable compared to the pre-merger situation. As part of the divestiture, the purchaser will, as far as the Divestiture Business is concerned, enter into DONG's position in that agreement with Energinet.dk.

(710) DONG commits to enter into a final sale and purchase agreement with an appropriate purchaser for the Divestiture Business, within six months from the date of the adoption of this decision. If, by the end of that period, DONG has not entered into such an agreement, the Divestiture Trustee will have an exclusive mandate to sell the Divestiture Business within the following three months. The purchaser has to be approved by the Commission.

(711) The closing of the sale of the Divestiture Business must take place no later than 1 May 2007. Once a final binding sale and purchase agreement has been concluded, only the purchaser will be entitled to sell storage capacity in the Lille Torup storage facility for the gas storage year 2007/2008 which runs from 1 May 2007 to 30 April 2008. DONG commits not to acquire, for a period of ten years from closing of the sale to the purchaser, direct or indirect influence over whole or part of the Divestiture Business unless the Commission has previously found that the market structure has changed to such an extent that a re-acquisition should no longer be excluded.

2. Gas Release Programme

(712) With a view to addressing the competition concerns raised in this decision, DONG has submitted, as a commitment, a Gas Release Programme in order to make natural gas available to third parties in Denmark. The amount of gas to be released will be 400 mcm per year, (a total of 2,400 mcm) to be auctioned in the years 2006 through 2011. According to the amended Commitments, the volumes released will, each year, be divided into ten lots of 40 mcm to be delivered equally over two delivery periods. The delivery period for both primary and secondary auctions to be held in 2006 will be, for practical reasons, the calendar years 2007 and 2008. As for the auctions to be held in 2007 and later, the delivery period will comprise the two following gas years (the first one starting in October of the year of the auction). The volumes correspond to a significant proportion of DONG's sales in Denmark in 2005 and approximately 10% of total Danish consumption. In the event that market conditions change significantly, e.g. DONG's upstream supplies will fall below a certain level during the course of the Gas Release Programme, DONG may, under certain conditions, apply to the Commission to have the Gas Release Programme terminated with respect to the 5th and 6th auctions.

(713) The Gas Release Programme foresees a two-step auction process: "Primary Auctions" to be held no later than April (for 2006: August) and "Secondary Auctions" to be held in June (for 2006: October) of the same year. In the "Primary Auction", DONG will make available gas at the virtual trading point/hub in Denmark (GTF) and, as in a swap, successful bidders will make available the same volume of gas to DONG at one of the specified gas hubs in Germany (Emden), the Netherlands (TTF), Belgium (Zeebrugge) or the United Kingdom (NBP). The "Primary Auction" determines the swap fee at which demand meets the number of auctioned lots. With a view to attracting bidders, DONG pays a separate "compensation" fee of at least 0.33 EUR/MWh to every successful bidder, which may be offset in the swap fee achieved by the auction. The "compensation" fee is intended to "compensate" for any price differences between the GTF and the other hubs. The gas will be divided into ten identical lots of 40 mcm each and auctioned with reference to the swap fee. No single bidder can bid for more than half of the lots auctioned in one year (Primary and Secondary Auction combined). The Gas Release Programme provides (both for the Primary and Secondary Auction) flexibility to the market with a take-or-pay obligation of 90% of the annual contract quantity and with daily minimum and maximum rates of 50% and 110%, respectively, of the daily contract quantity. It will provide further flexibility to the market by allowing the successful bidders to choose to swap flexibility by providing DONG with the same flexibility terms at the respective re-delivery point as they wish to obtain from DONG at the GTF.

(714) Any quantities not sold in the Primary Auction, will be sold in a Secondary Auction to be held later in the same year. In the Secondary Auction, DONG makes gas available to third parties in Denmark against payment in cash instead of re-delivery of gas. Lots that have not been sold in the Secondary Auction will be carried forward to the Primary Auction of the following year. In the Secondary Auction, a minimum price is to be set as a certain percentage of an indexation reflecting the structure of the indexation of DONG's contracts for purchase of gas from the Danish sector of the North Sea.

(715) In addition, the Commitments foresee a customer release clause according to which existing direct customers of DONG which participate in the auction process or which purchase gas from a trader/wholesaler who was awarded lots in the auction, are entitled to reduce their contractual purchase obligation vis-à-vis DONG by the amount of gas they will have purchased as a result of the Gas Release Programme (i.e. themselves or from a successfully participating wholesaler).

(716) In the event that a third party has reason to believe that DONG has not complied with the commitments a mediation procedure will be put in place. The mediation procedure will be overseen by the Monitoring Trustee, who will be entitled, under certain conditions, to appoint additional professionals to assist in the mediation process.

B. ASSESSMENT OF THE COMMITMENTS SUBMITTED

1. Effect of the proposed commitments on the storage/flexibility market

(717) The main effect of the proposed Commitments on the storage/flexibility market emanates from the divestiture of the storage facility in Lille Torup. In addition, the Gas Release Programme will increase the liquidity of the Danish wholesale market and thereby also provide new sources of flexibility.

(718) The divestiture of the larger of the two Danish storage facilities which has a capacity of approximately 400 mcm with no capacity reservations exceeding 1 year, will result in a new entry on the Danish storage/flexibility market. This new constellation will increase competition on this market, whether regarded as a storage market or a wider market for flexibility.

(719) The Danish Energy Regulatory Authority, DERA, confirmed that the implementation of the storage divestiture would be possible under the Danish Natural Gas Supply Act. According to this Act, the purchaser has to apply for a licence which requires that the applicant must have the necessary expertise and economic background.

(720) Most respondents to the market test consider, in the light of their experiences in other countries, that the two storages can be operated separately by two different operators without any technical or other difficulties. Some respondents indicate that, currently, it is common practice in Denmark to book storage capacity in a way that does not distinguish between the two storage facilities as these are operated by DONG Storage as one virtual facility. However, according to DONG Storage, commercial users can serve all their operations and customers from one storage facility in Denmark, which appears convincing in the light of the Danish gas network structure and regulatory framework.

(721) The Danish Transmission System Operator Energinet.dk, however, emphasised that it needs to rely on both Danish storages to physically operate the Danish transmission system. This specific need results from Energinet.dk's statutory obligations regarding pressure, balancing and emergency supply. The necessary access is currently provided under an Interconnection and Operating Balancing Agreement ("IOBA") between Energinet.dk and DONG Storage. As DONG has proposed to conclude an additional IOBA and that the purchaser of the Divestiture Business will also enter into such an agreement, the necessary interconnection and operating balancing between the two Danish storages will be ensured also in the future. Therefore, the storage divestiture will not put at risk the secure operation of the Danish gas transmission system.

(722) A small number of respondents to the market test have indicated that the two Danish storage facilities, Lille Torup and Stenlille, have somewhat different technical characteristics. (449) However, it appears from the market test that Lille Torup is generally perceived as the more flexible and better performing one.

(723) The divestiture of the Lille Torup storage facility is generally viewed by market participants as having a positive impact on competition with respect to storage/flexibility in Denmark. In particular the ownership unbundling of the storage infrastructure from DONG's gas sales operations is considered as pro-competitive, including by energy consumer associations. In this context, several respondents suggest that the non-commercial TSO Energinet.dk would be an appropriate purchaser of the Lille Torup storage facility. Some respondents expect that such a divestiture could lead to a regulated third party access regime for storage capacity. It has also been suggested that the ownership by an operator not involved in downstream supply activities would lead to an improvement of the products offered as the new owner would have an increased incentive to be fully responsive to the market's needs. The Commission considers this a strong likelihood.

(724) Although some respondents to the market test advocate a divestiture of both storage facilities, the Commission considers that the ownership unbundling of the larger storage facility is necessary but also sufficient to address the merger-specific competition problems identified in this decision with regard to storage/flexibility.

(725) In the course of the market test, Naturgas Fyn - which has been admitted as an interested third party to the case (Cf. recital (16)) - has submitted to the Commission that they do not consider either the storage divestiture or the Gas Release Programme, or to that effect, any other possible remedy to be submitted by DONG, to sufficiently address the competition concerns raised by the Statement of Objections, and therefore essentially call for a prohibition decision. (450)

(726) The storage/flexibility volumes which will be made available to the market by a third party operator following the proposed divestiture, amount to more than 57% of the Danish storage capacity and therefore for a very large part of the overall flexibility available in Denmark. Accordingly, a large number of respondents to the market test considered the divestiture sufficient to compensate for the removal of Elsam's and E2's flexibility from the market. The Commission shares this view. The elimination, as an effect of the proposed concentration, of one flexibility tool independent from DONG (flexible demand) is compensated by the creation, as a result of the storage divestment remedy, of a new, and in general superior, source of flexibility independent from DONG (the Lille Torup storage facility).

(727) The Commission also considers that the arguments put forward by some market participants that a storage facility does not exactly mirror the flexibility that can be provided by decentral CHPs are not well-founded. What is important is that the very flexible Lille Torup storage facility has the ability to provide all flexibility services which flexible consumption of central CHPs can provide both in terms of seasonal and of short-term flexibility. With regard to short-term flexibility it should be noted that the Lille Torup storage facility historically has been a significant source for such services whereas Elsam's and E2's central power plants hold a significant potential but have so far only to a limited extent released this potential. In the Commission's assessment, a suitable buyer of the Lille Torup storage facility will not only have the ability but also incentives to provide all services which in the absence of the merger, the owners of Elsam and E2 would have been likely to provide.

(728) A positive impact on flexibility available in Denmark will also emanate from the Gas Release Programme, which thus complements the flexibility offered to the market by the Lille Torup storage facility. It is noted that the degree of flexibility in the auctioned volumes is comparable to the flexibility which current competitors to DONG have access to in Denmark. This available supply contract flexibility has so far clearly reduced these competitors' storage needs. The standard contract terms comprise a considerable flexibility with a take-or-pay obligation of 90% of the annual contracted quantity and a daily minimum quantity of 50% and a daily maximum quantity of 110%. Competitors wishing to sell the volumes obtained through the Gas Release Programme to end customers in Denmark directly will therefore have much reduced storage needs. The flexibility terms of the Gas Release Programme will allow successful bidders to compete with DONG for all groups of customers with different flexibility requirements. Thereby the customers of these purchasers will also benefit from these additional sources of flexibility. In addition the Gas Release Programme will increase the general liquidity of the Danish wholesale market. The purchasers of the 400 mcm gas released annually will be able to pass on the flexibility terms they receive from DONG not only directly to their end customers but also to intermediate (wholesale) customers of flexibility. As this additional flexibility is provided for 10% of the Danish total market volume, it will have a significant positive effect on the Danish market for storage (or for flexibility).

(729) This additional flexibility through the Gas Release Programme thus further compensates for a significant part of the flexibility which could have, in the absence of the merger been provided to the market by Elsam and E2. Therefore this additional flexibility further addresses the concerns expressed by those respondents to the market test which considered that the divestiture of one storage facility would not be sufficient to remedy the competition concerns regarding flexibility/storage.

(730) The divestiture will also have a positive impact on adjacent or downstream markets, in particular the gas wholesale market. The divestiture of one storage facility will enable DONG Trade's competitors to satisfy their flexibility needs without relying on their main competitor. This new ability to acquire storage services which constitute an important and vital input for any gas wholesale (and frequently also retail) operation, independently from DONG will significantly increase the competitiveness of these competitors. The improvement of their competitive situation will be twofold: first, they will no longer need to provide the DONG group with information about their flexibility needs. Companies are usually very uncomfortable about providing such sensitive information to companies linked to their competitors as they are afraid that this information would allow conclusions on their customer profile. Second, and more importantly, DONG's wholesale (and retail) competitors will no longer need to pay for their storage to the DONG group. They will thus no longer have to contribute to the revenues and profits of the DONG group. In this respect, the provision of storage by an operator independent of DONG will contribute to a level playing field between DONG and its competitors.

(731) In conclusion, the Commission considers that all flexibility elements of the Commitments are sufficient to compensate for the merger specific loss of competitive potential stemming from the CHPs and to ensure that DONG's ability to increase rivals' storage costs will not be significantly improved.

(732) Both elements of the Commitments, i.e. the storage divestiture and the Gas Release Programme, will also have a positive impact on the Swedish markets. Competition between DONG and the future storage operator of Lille Torup will also benefit Swedish storage customers. The same applies to the effects of the Gas Release Programme.

2. Effect of the proposed commitments on the wholesale market for natural gas

(733) The Commission has concluded that the Gas Release Programme offered by DONG, in combination with the storage divestiture, is sufficient to remove all the competition concerns identified by the Commission in relation to the wholesale market.

(734) As to the Gas Release Programme, the quantities offered of 400 mcm per year to be swapped or sold via either the primary or the secondary auction make up approximately 10% of Danish demand. These quantities will result in substantial volumes of gas being sold to independent third parties in Denmark. The Gas Release Programme will be complemented by a customer release clause, which entitles DONG's customers who acquire gas quantities directly in the Gas Release Programme or from suppliers who acquired the corresponding gas quantities in the Gas Release Programme to request a release from their purchase obligations vis-à-vis DONG. The combination of these measures will thereby resolve both competition concerns raised on the gas wholesale market, namely customer foreclosure and removal of potential competition.

(735) First, with regard to customer foreclosure, these quantities will, in combination with the customer release provision, compensate for the volumes which E2 has sourced in the past years on a short term basis both inside and outside Denmark, and which accounted for approximately [5-10%]* of the Danish demand. The released quantities also compensate for any additional volumes which would potentially have been purchased by both Elsam and E2 prior to the expiry of their long-term supply contracts with DONG in [...]*, taking into account the effect of the divestiture of gasfired decentral plants to Vattenfall. Moreover, the released gas quantities also compensate for the further quantities which Elsam and E2 could realistically be expected to have purchased from other suppliers than DONG from [...]* onwards. Although the volumes currently purchased from DONG by Elsam and E2 under these long-term supply contracts correspond to [15-20%]* of total Danish demand, it cannot be assumed that Elsam and E2 would in the absence of the merger have purchased more than half of their demand (i.e., on the current basis, ca [...]*) from third parties (and not from DONG). Indeed, the Gas Release Programme provides for an annual volume of 400 mcm of gas which will be certainly released every year, and for which customers will be released from their contracts with DONG, whereas the future purchasing behaviour of Elsam and E2 is uncertain. Therefore, an annual gas release volume of 400 mcm appears an appropriate and proportionate quantity in view of the current market structure and the likely future developments. It can therefore be concluded that, overall, these quantities are adequate to offset any negative effect of the merger for the foreseeable future stemming from customer foreclosure.

(736) The quantities will, secondly, compensate for the elimination of potential competition by Elsam and especially E2. It is, however, difficult to establish with sufficient certainty what quantities would have been sold by these companies in the foreseeable future. What is clear from table 9 in recital (387) above is that all imports by and for E2 plus all competitors' market shares (except for swaps with DONG) together (i.e. all independent competition to DONG taken together), amounted to about 10% of Danish demand in 2004. The Commission's analysis has not claimed that E2 and Elsam would be the only, nor even that they would necessarily be the closest, competitors to DONG on the Danish wholesale market. The Commission's analysis has also underlined that DONG's dominance would not be threatened by the entry of Elsam and/or E2. In view of these circumstances, an additional 10% independent source of natural gas in Denmark can be expected to offset the harmful effect of the merger by providing sufficient incentives for DONG's existing actual competitors and DONG's potential competitors to enter and/or expand their activities. This extra incentive is for a sufficiently long period of time (6 years, deliveries ending in 2012).

(737) It should also be considered that these quantities will in themselves have a daily flexibility of between 50-110% of DCQ (451) and an annual flexibility of 90-100% of ACQ (452), with the additional possibility for a swapping partner to DONG to bilaterally negotiate other, even more convenient terms. These provisions will make this gas attractive for all purposes of the Danish wholesale market. The purchaser can decide to sell on the entire capacity to a wholesale trader, a wholesale customer or a final customer, or to use it itself, or it can decide to sell on or use only part of the volume and use the extra volumes for spot trades in Denmark, thereby adding substantial liquidity to the Danish market, presumably in particular the GTF, which according to the Commission's analysis would otherwise have been negatively affected by the proposed operation.

(738) An element of specific importance is the customer release provision which provides an incentive for the purchasers of this gas to use it in Denmark, facilitates entry, and addresses the customer foreclosure concern.

(739) Regarding the preference (due to the primary auction) for swaps with players established at other North-Western European hubs, this firstly does not limit the possibility of Danish wholesalers and wholesale customers to gain access to this gas as such international players might well want to ensure Danish "counterparties" to their activities. Secondly it contributes to integrating north-western European hubs with benefits for the development of competition around all of these hubs. A particular advantage for the GTF from a Danish wholesalers' and wholesale customer's perspective is that the results of the auction will make Danish wholesale transactions (also in comparison to other hubs) more transparent.

(740) The provisions concerning the secondary auction ensure that even in the absence of a sufficient number of bidders wishing or able to swap, the effectiveness of the remedy is nevertheless ensured. In particular, any lots not sold in the primary auction will be automatically transferred to the secondary auction of the same year which will take place sufficiently early to make sure that the purchaser has enough time to secure the sale of the acquired volumes to its customers.

(741) Regarding the duration of the Gas Release Programme (expiring in 2012) a large number of market participants have qualified this as sufficient for offsetting the harmful effects of the merger on the Danish wholesale market. This can be explained by the fact that there is a certain element of advancing positive "liquidity effects" that in the absence of the merger would likely have taken as long to mature and which, once established might be expected to have some self-regenerating and self propagating effect. (453)

(742) As regards the implementation of the Gas Release Programme, it is important to ensure that all participants are admitted at transparent and non-discriminatory terms and the swap/sale is made under competitive conditions. The provisions governing the auction appear to be sufficiently clear and impartial (and safeguarded by the function of the Monitoring Trustee and the independent third party conducting the auction) as to ensure a successful non-discriminatory conduct of the auction.

(743) The market test has confirmed that the annual gas release volume of 400 mcm is considered as sufficient by a large number of respondents. Following the feed-back in the market test, the Commitments have been adapted with respect to the delivery period corresponding to the Gas Year (October-September), except for the lots auctioned in 2006 which, due to technical problems of a timely organisation of the auctions in 2006, are to be delivered in the following two Calendar Years. The Commitments as amended further take account of a series of well-founded comments made by respondents to the market test. (454)

(744) Regarding the reservation price for the secondary auction, it can be observed that this is clearly below the price of the contracts DONG has with [Competitors]* .

(745) Also the storage divestiture remedy will have a beneficial effect on the Danish wholesale market. This is because access to non-discriminatory storage capacity independent from DONG is an essential facilitator of third parties' wholesale operations in Denmark. As explained above, it is crucial for DONG's competitors that they do not cross-finance DONG through their storage fees.

(746) For these reasons it is concluded that the commitments are sufficient to remedy the competition problems identified on the Danish wholesale market.

(747) The Commission moreover notes that, as regards the BGI pipeline project, DONG has stated that "[its] acquisition of Energi E2 A/S will not in itself have a negative impact on the continuation of the BGI project. The project will continue to be evaluated solely on the basis of its economic merits, compared to alternative means of transporting gas to Denmark including investments in new infrastructure linking Denmark to other gas markets" (455).

3. Effect of the proposed commitments on natural gas retail markets

a) Market(s) for supplies to industrial customers and decentral CHPs

(i) The raising of entry barriers (vertical problems)

(748) The Commission finds that the Gas Release Programme remedies any potential concerns relating to the raising of entry barriers on the market(s) for supplies to industrial customers and decentral CHPs.

(749) More specifically, the Gas Release Programme addresses the concern relating to customer foreclosure effect and the attainment of critical size, through the customers release part of the programme, whereby customers representing an annual consumption of up to 400 mcm can be made available for other suppliers. The 400 mcm volume auctioned amounts to 17% of the whole market for supplies to industrial customers and decentral CHPs and is therefore likely to offset any potential negative vertical effects as set out in the competitive assessment.

(750) The annual amount of gas to be released will have an important impact on liquidity on the Danish wholesale gas market and can be expected to have positive effects on trading at the GTF, thereby facilitating access to wholesale gas on competitive terms for DONG's competitors. Since access to wholesale electricity is already possible for DONG's competitors via Nord Pool, the effects of the Gas Release Programme will also facilitate the possibility of competitors which are active, or wishing to become active, on the market for supplies to industrial customers and decentral CHPs to offer dual fuel natural gas/electricity supplies in competition with DONG.

(751) Any concerns relating to the removal of flexibility of central CHPs as a balancing possibility to the flexibility requirements of suppliers and input foreclosure, are remedied by the flexibility provisions of the Gas Release Programme as well as the divestiture of the Lille Thorup storage facility, which introduces competition between the two Danish storages, eliminates the risk of input foreclosure and thereby facilitates entry.

(752) The Commission therefore concludes that as a result of the Commitments submitted entry barriers will not be higher as compared to the pre-merger situation.

(ii) Elimination of potential competition

(753) The Commission considers that the overall lowering of entry barriers on this market/these markets will have the effect on offsetting the loss of potential competition on this market/these markets.

(754) Regarding the Gas Release Programme this is in particular achieved in two ways. First, the improved access to gas under the Gas Release Programme with the associated flexibility facilitates entry by other potential competitors. Secondly, the customer release mechanism ensures that potential competitors wanting to enter the Danish gas market who have acquired gas through the release programme, will have comparatively easy access to customers.

(755) Second, the storage divestiture remedy will facilitate entry by introducing competition between the two Danish storages, eliminating the risk of input foreclosure and strengthening overall confidence in non-discriminatory access to storage facilities.

b) Market(s) for supplies to small businesses and households

(i) Raising of entry barriers

(756) For the same reason as indicated above regarding the market for supplies to industrial customers and decentral CHPs the Commission finds that the release programme remedies the concerns relating to critical size and liquidity of wholesale gas. The annual volumes of the release programme correspond to 45% of the market(s) for supplies to small businesses and households.

(757) Furthermore, the concerns identified relating to the removal of flexibility of central CHPs as a balancing possibility to the flexibility requirement of small businesses and households, are remedied by the flexibility provisions of the Gas Release Programme as well as the divestiture of the Lille Thorup storage facility, which introduces competition between the two Danish storages.

(758) For the same reasons as outlined above for the market(s) for supplies to large business customers and decentral CHPs, (456) the commitments submitted will prevent any raising of entry barriers through the risk of foreclosure of access to storage/flexibility and, potentially, customer foreclosure.

(759) The Commission therefore concludes that as a result of the Commitments submitted entry barriers will even be lowered as compared to the pre-merger situation.

(ii) Elimination of potential competition

(760) The Commission considers that the overall lowering of entry barriers on this market/these markets will have the effect of offsetting the loss of potential competition on this market/these markets.

(761) Regarding the Gas Release Programme this is in particular achieved in two ways. First, the improved access to gas under the Gas Release Programme with the associated flexibility facilitates entry by other potential competitors. (It is recalled that the competitive assessment identified several other potential entrants whose main problem was access to competitive wholesale gas.) Secondly, the customer release mechanism ensures that potential competitors wanting to enter the Danish gas market who have acquired gas through the release programme will have comparatively easy access to customers.

(762) Secondly, the remedy submitted on the storage/flexibility market will facilitate entry by introducing competition between the two Danish storages, eliminating the risk of input foreclosure and strengthening overall confidence in non-discriminatory access to storage facilities.

4. Overall conclusion on Commitments

(763) For these reasons, the Commission concludes that commitments submitted by DONG on 30 January 2006, and amended on 1 March 2006, are sufficient to remedy the competition problems identified in the assessment of the impact of the notified concentration.

IX. CONCLUSION

(764) It is concluded that the commitments submitted by the notifying party are sufficient - within the meaning of Art. 8(2) of the Merger Regulation - to address the competition concerns raised by the notified operation. Accordingly, subject to compliance with the Commitments submitted by the notifying party, the notified operation should be declared compatible with the common market and the functioning of the EEA Agreement.

HAS ADOPTED THIS DECISION:

Article 1

The notified operation whereby the undertaking DONG A/S acquires within the meaning of Article 3(1)(b) of Regulation (EC) No 139-2004, sole control of the undertakings Elsam A/S, Energi E2, Københavns Energi Holding A/S and Frederiksberg Elnet A/S is hereby declared compatible with the common market and the functioning of the EEA Agreement.

Article 2

Article 1 is subject to full compliance with the conditions set out in Section B.II. (Storage Divestiture) of the Annex.

Article 3

Article 1 is subject to full compliance with the obligations set out in Section B.I. (Gas Release Programme) and the remainder of the Annex.

Article 4

This decision is addressed to:

DONG A/S

Agern Allé 24-26

2970 Hørsholm

Denmark

Notes

1 OJ L 24, 29.1.2004, p. 1.

2 OJ C ...,...200. , p....

3 OJ C ...,...200. , p....

4 Cities of Helsingør, Roskilde, Hillerød and Slagelse (SK-EL) and the cooperative NVE-SEAS.

5 Cf. Section 10.1.3 of the "Conditional Shares Transfer Agreement" between DONG and the City of Frederiksberg in connection with Section 12.1.3 of Annex 1.1 to that Agreement.

6 Cf. Section 3.1 (ii) of the respective "Shares Transfer Agreements" with the four municipalities and Section 3.1 (f) (ii) of the "Shares Transfer Agreement" with the cooperative.

7 Cf. Commission Decision of 22 December 2005, Case COMP/M.3867-Vattenfall/Elsam and Energi E2 assets.

8 The assets acquired by Vattenfall in Denmark are the following: West Denmark: Nordjyllands central CHP (665 MW), Fyns central CHP (686 MW), 60% of Horns Rev wind farm (96 MW) plus Elsam's onshore wind activities consisting of wind mills spread across Western Denmark (213 MW). East Denmark: Amager central CHP (477 MW) and Hillerød and Helsingør decentral CHPs (together 135 MW). Fyns central CHP is the only gas-fired central CHP acquired by Vattenfall but only uses a very small quantity of gas.

9 Only a very small amount of this, all wind power, is situated outside Denmark: Sweden (28MW), Germany (32MW), United Kingdom (90 MW), Poland (30 MW).

10 Turnover calculated in accordance with Article 5(1) of the Merger Regulation and the Commission Notice on the calculation of turnover (OJ C66, 2.3.1998, p. 25).

11 Directive 98-30-EC of the European Parliament and of the Council.

12 Consolidated Act no. 526 of 11 June 2002 on the use of the Danish subsoil.

13 Act No. 1384 on Energinet Danmark, passed on 14 December 2004.

14 RfG 6.0.

15 bcm: billion m³.

16 Source: Energinet.dk

17 Cf. Energinet.dk Annual Report 2004; [Information concerning replies by DONG to information request by the Commission]*.

* Parts of this text have been edited to ensure that confidential information is not disclosed; those parts are enclosed in square brackets and marked with an asterisk.

18 The Dansk Undergrunds Consortium (DUC) joint venture was formed in 1962 by A.P. Møller - Mærsk A/S (46%), Shell Olie- og Gasudvinding Danmark B.V. (39%) and ChevronTexaco Denmark Inc. (15%) and holds the Sole Concession of July 1962. Maersk is the operator of the field. (Cf. Form CO)

19 Amerada Hess ApS (57.479%), Danoil Exploration A/S (1.584%), Denerco Oil A/S (6.562%), DONG E&P (34.375%) form the Syd Arne group. Amerada Hess is operator.

20 DONG E&P (45.594%), Denerco Oil A/S (38.904%), Denerco Petroleum A/S (17.502%).

* Parts of this text have been edited to ensure that confidential information is not disclosed; those parts are enclosed in square brackets and marked with an asterisk.

21 Cf. Form CO Annex A and the response of DONG of the Commission's questions of December 8, 2005. [...]*.

22 Northern Offshore Gas Transport

23 Cf. DONG's response to the Commission's questionnaire of 30 November 2005.

24 Regulators' Annual Report to the European Commission - Contribution for Denmark compiled by the Danish Energy Regulatory Authority (DERA) "DERA Annual Report".

25 DERA Annual Report.

26 STAT-USA - The Danish Oil and Gas Market http://strategis.ic.gc.ca/epic/internet/inimrri. nsf/en/gr118754e.html.

27 Only three end customers are connected directly to the transmission grid. These are the large (central) power plants: H.C. Ørstedsværket, Avedøre II and Skærbækværket.

28 Cf. Response of DONG to the Commissions' Questionnaire of 30 November 2005.

29 Dangas has a capacity share of 43263 m3/h (12.8%), BEB has a capacity share of 92095 m3/h (27.2%) and E.ON Ruhrgas Transport has a capacity share of 203000 m3/h (60%). Source: www.dangas.de, www.eon-ruhrgas-transport.com, www.beb.de.

30 Cf. Form CO.

31 The term "exit zone" is used collectively for the interfaces of the transmission system with the individual lower pressure distribution networks and the three large power plants which are connected directly to the high pressure grid.

32 The Capacity charge, i.e. the entry and the exit fees are both 15.55 DKK/kWh/h/year for a yearly reservation. The commodity component is 0.00193 DKK/kWh to be paid once. Cf. http://www.gas.energinet.dk/uk/Energiservice/tpa/prisblad_uk.pdf.

33 The interruptible emergency tariff is 7.45 DKK/kWh/h on an annual basis while the firm emergency tariff is 21.35 DKK/kWh/h. Cf. http://www.gas.energinet.dk/uk/Energiservice/tpa/prisblad_uk.pdf.

34 "Line pack" allows variations in the volume available in the gas pipeline network by means of an increase or decrease of pressure in the network.

35 For "short" imbalances exceeding the tolerance level, shippers are charged the gas price plus a premium of 50%. For "long" imbalances exceeding the tolerance level, shippers are paid 50% of the gas price. In summer the tolerance level is equal to 15% of the daily maximum capacity (corresponding to 360% of the maximum hourly capacity). In winter the tolerance level is equal to 5% of the daily maximum capacity (corresponding to 120% of the maximum hourly capacity).

36 Following the acquisition of Naturgas Sjælland in 2001, and Naturgas Syd in 1999.

37 HNG/MN can be considered for the purposes of this communication to be one entity. Cf. recital (610).

38 Cf. Form CO.

39 Cf. Form CO.

40 Cf. Form CO, Annex B.

41 In the context of the Danish Competition Authority decision on the acquisition of Naturgas Sjælland by DONG Naturgas of February 2001.

42 Cf. Form CO, Annex B.

43 Some minor exceptions have remained, particularly regarding small combined heat and power plants (decentral CHPs) and wind power. However, these are being successively lifted. From January 1, 2005 all units with a capacity over 10 MW have to sell their power to the market, while as of January 1 2007 this limit will be lowered to 5MW. All plants, even smaller ones may already sell the power to the market if they wish.

44 This is possible because Nord Pool has a de facto monopoly over the actual use of the available interconnector capacity. Only when the actual physical capacity in one direction is used to 100%, taking all counterflows into consideration, a price splitting mechanism sets in, which separates the markets on both sides of the fully utilised (i.e. 'congested') interconnection, thereby preventing any further load on the interconnector.

45 The grid company's tariffs are determined by income caps set by the regulator. The tariffs are separated into an entry charge (generation) and an exit charge (consumption). The main part of the charge is put on the exit charge. Tariffs vary across the day but not by location. All charges are put on the flow of electricity (contrary to fixed tariffs or capacity reservation charges).

46 These are: SEAS-NVE Net A/S, Sydvest Energi Net A/S, TRE-FOR Elnet A/S, NRGi Net A/S, NESANET A/S, Københavns Energi, HEF Net A/S, Energi Fyn Net A/S, and Energi Danmark ESS Net A/S.

47 COMP/M.1383-Exxon/Mobil, 29.09.1999; COMP/M.3410-Total/GdF, 08.10.2004, recital 18.

48 Notification, p. 146. See also the decision by the United Kingdom Competition Commission referred to by DONG.

49 Cf. Responses to the Commission Questionnaire to Danish shippers and suppliers of natural gas.

50 Cf. also the reply given by Energinet.dk itself in regard to its own flexibility needs which qualifies the capacity of line pack storage and flexibility as insufficient for needs going beyond very short-term flexibility needs.

51 DONG, in reply to the Commission's information request in phase II, regarding its own storage needs compared to its upstream supply flexibility.

52 For example, one market player qualified this type of flexibility provision, in particular by central CHPs as "very useful for weekly balancing" and in regard to seasonal balancing as "very useful because the costs for storage will be reduced dramatically if power plants use gas in the summer and are interrupted in winter" (SGA). Another market player qualified it as "suitable if economically attractive" for both short term and seasonal storage (Dansk Shell).

53 A price for such service would most likely take the form of a rebate on the gas supply price but could also be structured as a cash re-payment by the supplier when it uses this service a customer offers.

54 A notable exception is Elsam's decentralised CHP Herning, which uses both gas and biomass as fuels. For the purpose of this decision, the flexibility provided by this very large decentralised CHP is therefore aggregated with the flexibility of Elsam's central CHPs as discussed in recital (64) f. below.

55 A phrase used by Gas Natural in case M.3440 - EDP/ENI/GDP.

56 E2 has itself booked storage capacity with DONG (testifying to its serious intention to become a player in the gas markets). In this context it should not be neglected that liberalisation of Danish gas markets is still in its early stages, as evidenced, inter alia, by the low degree of liquidity in the wholesale markets and the very initial development of the GTF hub. The fact that E2, like DONG, has purchased short-term flexibility therefore, cannot be interpreted to mean that it cannot also offer short-term flexibility. In the course of further progress in the opening up of the markets, central CHPs would with a high degree of certainty: (a) have become more aware of this commercial possibility; and (b) would have found more opportunities to profitably market their flexibility potential.

57 [...]*

58 [...]*

59 Energinet.dk, reply of Oct 9,2005, question 24.

60 For this gas, the Danish onshore grid has an entry point at Nybro in Jutland. Although this gas is "Danish gas" also at the offshore fields, it is thus nevertheless "imported" with regard to the onshore grid (with a different ownership and different utilisation rules).

61 It should be noted that with the DCA's decision in December, 2005 DONG has agreed to free the customer from its contractual ties as of 2007. This of course does not mean that it will not be supplying the company after that date.

62 [...]*

63 [...]*

64 The addition of such TSO fees is commonly called "pan-caking".

65 It is to be noted that this feature of a liquid wholesale market is largely absent in Denmark.

66 CHP: Combined Heat and Power Plant.

67 The term 'physical imports' is understood as deliveries into the Danish onshore system/transmission grid. A Danish producer in the Danish part of the continental North Sea shelf thus only becomes a direct participant in the Danish wholesale market if it delivers the gas into the Danish onshore grid. Currently all gas supplied from the Danish exploration platforms in the North Sea to Denmark (which is not produced by DONG anyway) is bought by DONG at the offshore platforms. DONG then "imports", i.e. ships the gas through its pipeline to the Danish shores at Nybro and sells it in Denmark or exports it further to Sweden and Germany. Due to these particularities of the Danish gas exploration structure, producers are currently not active on the wholesale market in Denmark.

68 DONG has contested the lack of necessity of examining the German wholesale demand. This issue is discussed in detail in the relevant geographic market section below.

69 It is noted that such market may well be considered to be similar to or largely identical to the market for wholesale (supply) of natural gas for Denmark.

70 Cf. Issues Paper of the Commission's Energy Sector Inquiry, November 2005.

71 [Contains information on price data submitted by E2]*

72 See below in the respective sections of the assessment.

73 Cf. both form CO and the Reply to the Statement of Objections. No new arguments are set out or new evidence is adduced in the Reply to the Statement of Objections.

74 [...]*

75 See DONG's reply of 7 December 2005, question 7. This also becomes evident by comparing the prices in Table 5 and Table 6 below.

76 [...]*

77 See DONG's reply of 7 December 2005, question 28 and question 7 (on other cost).

78 See DONG's reply of 7 December 2005, question 2.

79 [...]*

80 [...]. See also results of the market investigation by the Danish competition authority in the antitrust case on DONG's supply contract with HNG/MN.

81 [...]*

82 It should be noted that there is one (small) group of decentralised CHPs which is linked with industrial consumption patterns. These are CHPs in which heat/steam is produced for industrial processes. [...]*

83 [...]*

84 [...]*

85 [...]*

86 The relevance of this threshold was explained above (cf recital. (107): Metering requirements only exist for customers with an annual consumption above 0.3 mcm. The customer groups below, households and small business customers do not have such requirements. In the Commission's market investigation this value has been seen as constituting the most appropriate dividing line between large and small customers.

87 There is a possibility for customers with an annual consumption below 300,000 m3 to have gas supplied on "metered terms". This, however, requires the installation of metering equipment and is only used by a few businesses that have a large enough consumption that can compensate for the costs of metering equipment. The impact of this possibility on market definition is therefore negligible.

88 Decision in the DONG-HNG/MN antitrust case.

89 [...]*

90 Other costs include transmission, storage and overhead.

91 It is 10 mcm in Sweden as compared to 700 mcm in Denmark.

92 E.ON's gas-fired power plant in Malmö is comparatively small and old and, importantly cannot operate in condensing mode, which means that it inevitably needs to produce heat, thus severely limiting its usability for flexibility services.

93 DONG Annual Report 2004, p. 15. Dutch tariffs for large injection and withdrawal capacity are even twice as expensive as Danish tariffs.

94 See for instance the level of congestion in the BEB area.

95 M.1383, recital 262; M.3086, paragraph16.

96 [...]

97 Danish Energy Agency, Monthly Natural Gas Statistics.

98 These figures are net of all imports or re-imports (turn-arounds).

99 As to the reasons for this, reference is made to the discussion of turn around gas in the "wholesale" section.

100 The Commission acknowledges that this booking of capacity by potential turn-around flexibility companies is only an indication of its lesser degree of substitutability in comparison to intra-Danish sources. As noted above, E2 itself, with access to large flexibility resources in Denmark through its power plants, has booked storage capacity with DONG. While this might, theoretically, on the one hand indicate that both flexible supply contracts and central CHP flexibility are less than perfect substitutes for physical storage, this booking of physical storage capacity might on the other hand also be due to, as stated above, the rather early stage of market opening in Denmark.

101[...]*.

102 450.000 and 500.000 for Stenlille and Lille Thorup respectively.

103 Reply to the Statement of Objections recital 7.19.d (iii).

104 Energinet.dk, reply of Oct 9, 2005, question 23.

105 Cf. DONG's Annual Report for 2004.

106 The Article 6(1)(c) decision left the question open whether Denmark and Sweden constitute a common or separate relevant geographic market(s).

107 In the notification, DONG had not defined any wholesale gas market. The notification submitted, however, that even the "retail market for sale of natural gas to central CHP stations" includes not only Denmark but also Sweden, Germany, the Netherlands, Belgium and the United Kingdom.

108 See e.g. the DONG/NOVA gas decision by the Swedish competition authority.

109 [...]*

110 This does not deny the possibility, underlined by DONG, that after depletion of the Danish fields the flow direction between Denmark and Germany, with some extra investment, can be reversed. However, [...]*. Secondly, as to the Danish-Swedish flow direction, such direction cannot be changed before Sweden is connected to any other pipeline carrying more gas than is needed in Sweden to Sweden. The only possible such connection to materialise in the foreseeable future with some degree of likelihood (but not certainty is the BGI (see below). However, even if, as is possible, the BGI will be built in the foreseeable future, it nevertheless cannot be foreseen whether this will change the flow direction between Denmark and Sweden as it is planned that the BGI will also be connected to Denmark (at Avedøreværket), making the percentage of physical gas that will flow through it to Sweden and Denmark, respectively, uncertain.

111 Cf e.g. the issue of whether counterflow capacity (northbound) through the DEUDAN pipeline is available on an interruptible or non-interruptible basis.

112 E.g. there are clear indications on the Commission's file that DONG already uses its contractually secure physical export capacity to Germany and the Netherlands to a high degree, which obviously limits DONG's ability to export more.

113 This includes all gas that was entered into the Danish TSO system at the Ellund, either contractually transported to Ellund from south of the DEUDAN pipeline, or turned-around. [...]*.

114 Small but Significant Non-transitory Increase in Prices (see Commission Notice on Market Definition).

115 Average 2% of their total supplied volume, with only Zeebrugge and NBP experiencing higher volumes.

116 [...]*

117 DONG has pointed to the temporal difference of the two contracts, and the fluctuation of prices at hubs, which in their view results in an incorrect comparison.

118 Cf. Issues Paper of the Commission's Sector Enquiry and, importantly, the Commission's Fifth Benchmarking Report.

119 DONG's remarks can be understood as setting forth that Energinet.dk, though also owned by the Danish state, is more clearly separate and dissociated from DONG's supply activities than BEB is from the supply activities of its parent companies Shell and ExxonMobil. However, reference must also be made to the fact that DONG states that the adjacent North German BEB system is an entry-exit system just as that of Energinet.dk. As to the difficulties of transmission within the German "patchwork" of TSOs this can be considered to be an obstacle in both directions.

120 Cf. paragraph 4.8.d (iv) of the reply to the SO.

121 Cf. replies to questionnaire to Danish shippers; replies to flexibility question involving an indirect comparison between wholesale import of gas and Danish wholesale gas.

122 "Of 38 customer answers 8 had considered and 1 had voiced general interest, so 24% considered it"

123 Cf. the Commission's Notice on Market Definition.

124 On the basis of the publicly available tariffs, the transport costs from Emden to the Danish border at Ellund [...]* amount to at least 8-9% of the purchase price. On the basis of the figures provided by [...]*, they amount to at least 10% of the purchase price in 2005, and even more for 2004. In addition, any crossborder transport triggers entry fees into Energinet.dk's network, and possibly further exit fees and commodity charges in that network.

125 "Pancaking" refers to the fact that gas transport in Germany frequently needs to cross the networks of various network operators who each charge transport fees, thereby leading to an accumulation of transport costs.

126 [...]*

127 The comparison e.g. does not take account of factors such as connection/lack of connection to the high pressure grid or not, transit through other grid areas (e.g. the E.ON Hanse net) etc.

128 [...]*

129 Cf. HNG's reply to the Commission's questionnaire.

130 Cf the Eurostat prices quoted in the assessment sections on sales to large business customers and to small commercial customers.

131 [...]*

132 Cf. recital (369) on Energinet's statement.

133 http://www.hydro.com/en/press_room/news/archive/2003_09/maersk_gas_en.html

134 One of the technical constraints on delivering non-DONG DUC origin gas to Denmark is the 1 mcm/day of free capacity on the upstream pipeline, another is the difficulty of counterflows through the NOGAT pipeline (and through the F3 pipeline connecting the NOGAT system to the Danish offshore system). Other constraints of an economic nature (price, prior investment, costs of capacity reservations, etc.) have prompted the sales to the Netherlands. A further commercial constraint is brought about by the transaction itself by removing from the market the most likely large customers for gas producers.

135 Reply of 07.12.2005, question 10.

136 Dragør is the exit point of the Danish energinet.dk gas network at the Swedish border.

137 The Danish contract in 2003 had high flexibility at higher price, while the Swedish contracts from 1987 and 1989 were between DONG and Swedegas (later called NOVA), the sole importer to Sweden until 2004.

138 In a decision of 2004 (on DONG/NOVA) the Swedish Competition Authority ("SCA") found that the Swedish wholesale market did not include Germany: Transport cost were found to be higher in Germany, this combined with pan-caking made it much more expensive to buy gas from others than DONG for the Swedish market. The Danish-Swedish gas market was isolated due to high transport cost and the price influence from European hubs was modest. The geographic wholesale market was therefore defined as Sweden and Denmark. Cf. in that respect also recital 16 of the Commission's Notice on the Definition of the Relevant Product Market.

139 No gas has ever been exported from Sweden to Denmark via the only entry/exit point at Dragør. Since Sweden has no production of gas, natural gas has only been transported from Denmark to Sweden.

140 Cf. DONG's recent contractual arrangements regarding delivery of gas to Dragør.

141 There is possibly one (!) contract that is concluded for a supplier established in Sweden/Germany. However this supplier is also established in Denmark.

142 In a different context, DONG's Reply to the SO has, with regard to E.ON adduced further evidence for this need.

143 Eurostat, Statistics in Focus (Environment and Energy), 4/2005: Gas Prices for EU industry on 1 January 2005. Regarding Sweden and Denmark see also DONG's Reply of 7 December, question 11.

144 Industrial customer nomenclature: according to yearly consumption and load factor: I3-1 to 41860 GJ/a (200 days, 1000 hours); I4-1 418600 GJ/a (250 days, 4000 hours). In its Reply to the SO, DONG has adduced some further evidence concerning this price difference.

145 Industrial customer nomenclature: according to yearly consumption and load factor:. I1 up to 418.6 GJ/a; I2 to 4186 GJ/a.

146 Including storage and Transmission (equalling approximately 0.35 DKK/M3)

147 See Commission decisions in cases IV/M.2890 EDF/Seeboard, IV/M.3007 E.ON/TXU, IV/M.3268 Sydkraft/Graninge, COMP/M. 3440 EDP/ENI/GDP, IV/M.2684 EnBW/EDP/Cajastur/ Hidrocantabrico.

148 In autumn 2005 also Vattenfall's grid area in Eastern Germany will - to some extent - be joined as a separate Nord Pool bidding area, called Kontek.

149 Nord Pool accounts for more than 40% of the physical electricity consumed and produced in the Nordic region. For Denmark this percentage is even higher (51 % in East Denmark, 65 % in West Denmark, in 2005).

150 Gigawatt hour.

151 This is irrespective of Nord Pool's attempts to make its trading offer more flexible which is unlikely to reach the required flexibility due to the inherent constraints of an electricity exchange to offer sufficiently standardised products in order to achieve the necessary level of liquidity.

152 Cf. COMP/M. 2947 - Verbund/EnergieAllianz where the possibility of such a separate market for (wholesale) supplies to small electricity retailers was discussed but left open.

153 See e.g. case COMP/M. 3268 Sydkraft/Graninge or case COMP/M. 3440 EDP/ENI/GDP.

154 See cases COMP/M. 3440 EDP/ENI/GDP and COMP/M. 2947 - Verbund/EnergieAllianz. It can be noted that also the Danish Competition Authority, defined separate markets in this area. In its decision on the

Elsam/Nesa merger in 2004, it defined separate markets for wholesale supplies, regulating power and reserve capacity respectively (The Danish Competition Authority: The merger between Elsam and Nesa, March 2004).

155 MW: megawatt.

156 This service can also be termed "reserve capacity" and there are indeed some functional overlaps with the "emergency reserve capacity" service.

157 MWh: megawatt hours

158 While downward regulation can be performed by taking off production capacity or by switching on extra consumption, upward regulation can be performed by switching on extra generation capacity or by taking off consumption.

159 The market share of Eltermin in all financial contracts in the Nordic area has increased from 25% in 2000 to 38% in 2005. As regards the CfD's the market share of Eltermin is significantly larger as about 83% of all trades conduced since 2004 for the price areas Denmark East and West, have been conducted there. The volume of sales on Eltermin is 3.5 - 5 times the volume on Elspot.

160 E.g. the financial market provides market players with tools to reduce the risk of varying physical electricity wholesale prices, whereas expectations for the physical electricity wholesale market strongly influence prices on the financial market.

161 Cf recital (243) above.

162 Norway, Sweden, Finland, Denmark West, Denmark East and Syger (comparing the Phelix price Germany to the system price Nord Pool).

163 The dividing line between the two customer groups has been lowered from 0.2 GWh annual consumption to 0.1 GWh annual consumption.

164 See e.g. EDP/ENI/GDP.

165 E.g. in Sydkraft/Graninge for Denmark East/Sweden, based on interconnection congestion figures which were substantially lower than those for 2004 and 2005 (see below fn.167).

166 Yearly averages were 2.0% of the time in 2003, 6.0 % in 2004 and 14.0 % in 2005 (till Nov). Based on Nord Pool data.

167 48.9% of the time in 2003 the interconnectors to both Norway and Sweden were congested, 30.5 % of the time in 2004 and 51.3 % in the first nine months of 2005. These figures are also higher if calculated using a shorter timeframe. In 2005, until November, the interconnectors between West Denmark and either South Norway or Sweden were even congested in 60% of hours. The interconnection congestion between West Denmark and Sweden in this period was 56%, interconnection congestion between West Denmark and Norway South was also 56% and simultaneous interconnection congestion between West Denmark and both Norway South and Sweden was 52 %.

168 It is furthermore noted that financial derivatives can only provide imperfect hedges against deviating Nord Pool area electricity wholesale prices. In order to attain the maximum achievable hedge a customer needs to purchase two products, a hedge against the Nord Pool system price and a hedge against a deviating area price. For the latter purpose such customer needs to purchase CfDs which are only available on 1 year, 3 quarter and 2 month contract basis for all CfD areas, whereas the customer may have the need to hedge against the price risk in particular hours and at much shorter temporal distance to the spot market auctions.

169 This is due to a number of factors: (i) reserved volumes on the East Denmark-German connection; (ii) maintenance works; (iii) indications that capacity on the West Denmark-German interconnector (connecting a Nord Pool area with a non- Nord Pool area) is not allocated in an optimal way (which results in electricity being transmitted in the "wrong" direction, i.e. from the high-price area to the lowprice area), thus reducing the competitive pressure from this interconnector; (iv) decisions by the Swedish TSO Svenska Kraftnät to significantly lower the capacity on the interconnectors between Denmark West/East and Sweden in times of high north/south flows within Sweden leading to intra-Swedish grid imbalances; (v) the Danish transit function between the zones dominated by hydropower (Norway but also Sweden) and the zones dominated by thermal power (Germany) which leads to a tendency that the available interconnection capacity is used in a "one-sided" way at any given moment (this which means that constraints on Danish prices will only come from either north ,in times of hydropower surplus/lower hydropower production costs, or from south ,in times of hydro power shortage/lower thermal power production costs, provided there is no import congestion).

170 In 2004 this was the case in 26% of hours, in 2005 in 32% of hours.

171 Physically, the only Nord Pool area except Sweden connected to Denmark is South Norway. However, South Norway is only linked to West Denmark. Therefore any combination of zones including Denmark and South Norway would either have to include Sweden as well or would not include East Denmark. In both cases the parties' shares would logically, have to be lower than in the scenarios to be examined, West Denmark (stand-alone) or East Denmark + West Denmark +Sweden combined. A common price East Denmark + West Denmark +Sweden has existed in 2004 in 57 % of hours and in 2005 in 41 % of hours. However, the number of hours in which such an area was isolated from all the remaining Nord Pool areas was very low. A consideration of this area "East Denmark, West Denmark and Sweden only" is therefore a hypothetical worst case possibility rather than reflecting a commercial reality. E.g. in 39% (2004) and 41 % (2005) of the times the common price area East Denmark, West Sweden included at least also Norway.

172 Only if such a scenario leads to competitive concerns then the question of the significance of this specific combination in comparison with all other, wider, combinations (in which competition problems would necessarily be lower) needs to be analysed, and if necessary balanced with the effect of the concentration on the narrow East Danish and West Danish markets, in view of the likely effect of the merger on consumers.

173 The Danish Competition Authority has recently ruled in favour of all-Danish retail markets for large (> 0.2 GWh) customers while smaller customers (< 0.2 GWh) were considered on a narrower, regional or even local market due inter alia to their inertia in switching supplier for these smaller customers.

174 E.g. Vattenfall, the largest Swedish supplier, is not currently active in Denmark as a retail supplier to industrial customers.

175 Cf. Directive 2003-87-EC of the European Parliament and of the Council of 13 October 2003 establishing a scheme for greenhouse gas emission allowance trading within the Community and amending Council Directive 96-61-EC. This scheme has various international aspects transcending EU boundaries, in particular the JI (joint implementation) and the CDM (clean development mechanism) but it is uncertain whether these links are already strong enough to warrant a widening of the geographic market.

176 Emineral A/S. Post-merger, Vattenfall will become a 23.7% shareholder in Emineral A/S.

177 There is post-merger even likely to be a reduction of the parties' market share in the Copenhagen area, due to the split up of E2's activities between DONG and Vattenfall.

178 Points 1.15 and 2.4 et seq. Of DONG's reply to the SO.

179 Cf. Figure 1 in Point 2.5 of the Reply.

180 Cf. in that respect also judgment of the CFI of 14.12.2005, T-210-01 (General Electric/Commission), at recital 118 et seq.

181 The Commission's assessment of the effects of the concentration is thereby based on the explicit assumption that, after completion of the concentration, DONG and all its subsidiaries will be bound by all remedies submitted to the Danish Competition Authority by any company forming part of the transaction.

182 Cf. Judgment of the Court of First Instance in the case GE, T-210-01 recital 115 stating that very large market shares are in themselves evidential of dominance.

183 See DONG's agreement with Energinet.dk.

184 See below in the discussion of the competitive assessment on the decentral-CHP supply market/segment: DONG's electricity agent contracts.

185 Even on the assumption that this contract is illegal, this illegality may not necessarily concern the storage/flexibility aspect of this contract as far as industrial customers are concerned.

186 [...]*

187 Cf also the different flexibility provisions: DONG's upstream contract has a higher flexibility (Daily Minimum Quantity: [...]*; daily Maximum Quantity: [...]* than [...]* contracts with DONG at Ellund (Daily Minimum Quantity: [...]* Daily Maximum Quantity: [...]*.

188 [Contains considerations on E2's seasonal flexibility]*

189 Standard package 1 allows the storage customer to inject and withdraw the reserved storage volume for a minimum of 200 and 100 days, respectively. Standard package 2 allows the storage customer to inject and withdraw the reserved storage volume for a minimum of 40 and 20 days, respectively. Both packages can be purchased with and without filling restrictions (i.e., restrictions on the amount of gas in storage during the winter period).

190 See DONG's reply of 9 December 2005 to question 5 of the Commission's Article 11 letter (DONG I) of 8 December 2005.

191 For detailed figures on the use of storage contracts see recital (316) below.

192 Form CO p. 155.

193 See "Forhandlede vilkår", www.dong.dk.

194 In its reply of 9 December 2005, to question 5 of the Commission's Article 11 letter (DONG I) of 8 December 2005, DONG states that the storage tariffs have so far never been changed during the storage year. However, for the reasons stated in this section, DONG's ability and incentive to increase storage tariffs will be greater following the proposed transaction.

195 A significant number of contracts do not start in May but rather in autumn (beginning of the "gas year" is October).

196 Cf. DONG's response of 12.12.2005.

197 Plus 1% rounding error.

198 The demand by Energinet.dk is however, functionally different from the demand of suppliers as Energinet.dk's demand is used for balancing the pressure in the grid and not for supply purposes.

199 Par. 24 of the Decision

200 "§ 38. Priser og betingelser for ydelser fra transmissions-, lager- og LNG-selskaber skal fastsættes således, at der ikke diskrimineres mellem systembrugerne. Ved prisfastsættelsen tages hensyn til selskabernes omkostninger samt til, at der skal kunne opnås et rimeligt afkast af den i selskaberne investerede kapital, jf. dog § 37 d. "

201 Recital 236 of the Decision

202 Translation of recital 93.

203 Email received by the Commission.

204 Decision of 25 April 2005 Gastra - priser for adgang til transmissionsnettet 2001-2004

205 Recital 115

206 See recital 153 of Decision of 15 March 2002 "Priser for benyttelse af DONG Naturgas A/S' transmissionssystem"

207 See recital 103 of Decision of 29 August 2005 "Indtægtsrammeregulering af naturgasdistributionsselskaberne - fastsættelsen af forrentningssatser for 2005 samt udmelding af indtægtsrammer for 2005"

208 The reasonable rate of return was 7.4% on 3.1bnDKK equal to 225mDKK. A fall in the rate of return by 1.74%-point (5.1%-3.36%) would result in a fall to 172 mDKK.

209 See recital 18 of Decision of 31 October 2005: "Fastsættelse af effektiviseringskrav i forbindelse med fastsættelse af indtægtsrammer for naturgasdistributionsselskaber".

210 For the long-term agreements: cf. Elsam's response of 02.11.2005 (21:04 h) to the Commission's request for information; questionnaire to gas customers; Response to question 16 which also covers E2's longterm agreement with DONG.

211 [...]*

212 [...]*

213 Par. 4.41

214 par 6.2 (b)

215 page 14

216 Germany and the Netherlands respectively

217 Agreed minutes of telephone conference with Energinet.dk on 22.1.2006.

218 Table 2 in the Rothschild report to the Ministry of Finance "Privatisation alternatives for DONG A/S".

219 The NEGP connects Russia through the Baltic Sea with Northern Germany.

220 This approach is, for instance, reflected in recital 12 of the Commission's Notice on the concept of full function joint ventures.

221 Cf. Sector Inquiry into the Natural Gas Market; Benchmarking Report.

222 [...]*

223 Cf. Rothschild report submitted by DONG on December 2, 2005

224 This timeframe is confirmed (and even surpassed) by another, third-party forecast on the Commission's file which however, needs to remain confidential vis-à-vis DONG.

225 At December 2005

226 It must be considered that the Danish branch is foreseen to connect at E2's Danish power plant Avedøreværket, which is after the proposed merger controlled by DONG.

227 The figure is calculated for 2004 by taking the 4,2-4,5 bcm gas consumption in Denmark and subtracting from it the (max.) [...]* bcm in imports from Germany either through re-entry or contractual counter-flow on the DEUDAN pipeline as well as the volumes of the two swap agreements DONG had with [Competitors/customers]* and [Competitors/customers]*. (Cf. Energinet.dk response of October 9 to the Commission's questionnaire and DONG's response to the Commission's questionnaire of December 14, 2005). In 2004, about [...]* would need to be added if the sales DONG made to third parties in Denmark through swaps were attributed to DONG. The percentage that would need to be added for the first half of 2005 can be assumed to be higher than [...]* (as DONG's swaps in 2005 were [...]* higher than in 2004.)

228 Considering the first half of 2005, the last period for which a full assessment of volumes could be done considering the timing of the market investigation, this percentage is ca [0-5%]* higher

229 An exception being the volumes in conjunction with the Commission's DUC-DONG case (Cf. recital (398)).

230 One immediate such constraint is the DEUDAN interconnector capacity which DONG itself has stated that it is fully utilizing in its own "TSO regime". Furthermore capacity constraints in Germany are also critical just as the acquisition of customers to whom the excess gas can be sold to.

231 Cf. Judgment of the Court of First Instance in the case GE, T-210/01 recital 115 stating that very large market shares are in themselves evidential of dominance.

232 Cf Case COMP/E-4 38.187 DUC-DONG.

233 Cf. DONG's response to the questions of the Commission of November 30, 2005

234 Shell Olie og Gasudvinding Danmark BV.

235 Cf Shell's response to the Questionnaire to gas producers.

236 [...]*

237 Cf. Form CO, Annex A

238 This involves shipping on the NOGAT from Den Helder to NAM F3B, an onwards on the Tyra-F3 pipeline to Tyra and finally on the Tyra-Nybro pipeline to Nybro.

239 Cf. Response from Shell to the Commission's Gas Producer Questionnaire, phase II, as well as the response from NOGAT B.V.

240 Cf. DONG's response to the questionnaire of the Commission of November 30, 2005.

241 Cf. Submission to the Commission of 23/11/2005 of "Gasgruppen".

242 As the pipeline connecting the Tyra platform and the NOGAT system was only operational for half a year, figures for 2004 and the first half of 2005 are comparable. From those it seems clear that this volume has increased from [...]* mcm to [...]* mcm. Cf. DONG's reply to the Commission's questionnaire of 13 December, 2005.

243 Cf also views expressed in the market investigation considering hubs in Northern Germany or Benelux only as weak substitutes for wholesale gas in Denmark (Danish shippers questionnaire, in the context of flexibility).

244 GTE Report "Calculation of Available Capacities with Reference to Five European Transport Routes", 15 December, 2005.

245 "Capacity in south-north direction has been developed by building two compressors at the southern end of the DEUDAN transmission system in Quarnstedt. However, during the last few years, flows were solely from north to south. Therefore, the compressors in Quarnstedt were put into conservation mode. Currently, BEB and DANGAS offer interruptible capacity from south to north only (a). Should there be serious requests by shippers for firm capacity in this direction, the compressors in Quarnstedt would be put back in operational mode and these capacities could be contracted on a firm basis."

246 Cf. Figure 3 comparing those prices in the market definition section.

247 These additional costs amount, e.g. for a transport from Emden to Ellund, to at least [...]* but probably more than [...]* of the purchase price. On the basis of the publicly available tariffs, the transport costs from Emden to the Danish border at Ellund [...]* amount to at least 8-9% of the purchase price. On the basis of the figures provided by [...]* on 14.12.2005, they amount to at least 10% of the purchase price in 2005, and even more for 2004. The use of the DANGAS part of the DEUDAN pipe does not result in a reduction of those costs as such a scenario would trigger additional exit and entry costs for the switch of the network at the southern end of the DEUDAN pipeline near Quarnstedt. In addition, any cross-border transport triggers entry fees into Energinet.dk's network, and possibly further exit fees and commodity charges in that network.

248 As mentioned earlier the direction of the DEUDAN flow has been exclusively in a southerly direction, therefore companies can only contractually import gas into Denmark through the DEUDAN pipeline.

249 Cf. Response of DONG/Dangas to the questionnaire of the Commission on November 30, 2005. Entry charge at Quarnstedt and exit charge at Ellund for the Dangas part of the DEUDAN is the same at 13.72 EUR/m3/h for an annual tariff.

250 Cf responses to the market investigation by BEB (Gas Competitor questionnaire) and the response by DONG/Dangas to the Commission's questions of 1 December 2005

251 It can also be noted that the current entry capacity into the Energinet.dk system is merely 1.75 bcm/year. (Cf. http://www.gas.energinet.dk/uk/Energiservice/Technical_plants/cap2netUK.pps#266,7,Capacity and Reservations - Entry Ellund)

252 [...]*

253 Cf. Recital (172)

254 Cf. Recital (419)

255 It is also noted that the fact that the maximum import capacity at Ellund is 1.7 bcm, combined with the fact that this gas is at least as likely to be destined for the growing Swedish market than for Denmark (and that therefore the relevant technical maximum percentage of import capacity as of demand needs to be related to Denmark and Sweden, means that almost 70% (and with the expected growth of the Swedish market more than 70% of Swedish demand cannot be imported from Germany, whether through turnaround or otherwise (except through swaps with DONG, obviously concluded by DONG in a way compatible with its commercial interests and therefore not likely to exert extra pressure on DONG.)

256 According to Energinet.dk's website the re-entry fee amounts to 15.55 DKK/KWhg/hour/year, to which an exit fee of the same amount, a commodity charge and other fees have to be added if the gas is not sold at the GTF.

257 All prices excluding VAT and other taxes.

258 Therefore, the Commission considers the Eurostat statistics to reflect best the actual price levels and their developments.

259 Prices to small (annual consumption of 4,186 GJ = 110,000 m3) and very small industrial customers (annual consumption of 418.6 GJ = 11,000 m3) are higher in Denmark than in Hamburg.259 However, this customer group is much smaller and sales to large and very large customers therefore make up for the bulk of the gas supplies to industrial customers.

260 DONG's reply to the Commission's Questionnaire of December 8, 2005

261 [...]*

262 In addition to the margin advantage DONG has, capacity reservations must be made and an entry capacity charge must be paid for all planned gas deliveries via a simple entry procedure or a re-entry in the Danish transmission system, just as the same rules apply with regards to reservations and capacity charge payment for gas leaving the transmission system in the national exit zone. Furthermore all actual gas deliveries must pay an additional commodity component upon entry of the gas. Thus any gas re-entered into the system will by the logic of the system be more expensive than gas not taken out of the system first by the magnitude of an entry fee if sold at the GTF and en entry and exit capacity reservation fee and a volume fee if sold to any exit point in the system. As the standard yearly capacity reservation becomes progressively more expensive for periods shorter than 12 months, a further price disadvantage occurs in comparison to DONG who has long-term capacity reservations depending on the length of the capacity reservation. Essentially this means that the extra transmission capacity reservation - assuming a best-case scenario of a full year's reservation - results in on average a ca. 5% increase in costs for the sale of turned-around gas..

263 Cf. recital (393)above.

264 Cf recital (382) above.

265 [...]*

266 Cf. also [competitor]* response to the Commission's questionnaire on Gas Competitors and follow-up questions

267 Cf. [competitor]* reply to the questionnaire for Gas producers

268 Cf. Response of DONG to the questionnaire of the Commission of October 7, 2005.

269 It must be noted however that the capacity share of Dangas in the DEUDAN is only 12.8% as compared to its 49% ownership share.

270 [...]*

271 Cf Response of HNG/MN to the Danish shippers, supplier questionnaire

272 Cf. Document submitted by DONG at the meeting with the Commission of January 18, 2006, after its Reply to the SO.

273 http://www.gas.energinet.dk/uk/index.asp

274 First nine months of 2005.

275 Calculations based on capacity reservation and flow figures provided by Energinet.dk

276 Cf. Response to the commission's questionnaire to Danish suppliers and shippers

277 In the last two months liquidity has notably dropped on the GTF which can have a further effect on the strengthening of DONG's dominance. (Cf. Energinet.dk's submission to the Commission of December 14, 2005

278 [...]*

279 [...]*

280 [...]*

281 [...]*

282 Case 556-2004.

283 Cf DONG answer to questions of the Commission of October 7

284 Cf. Swedish Competition authority decision in 2004

285 [...]*

286 [...]*

287 In 2003, Sydkraft had a retail market share of [40-50%]*

288 Even if it were not, then the existence of a jointly dominant position with E.ON would still need to be investigated.

289 Geographic market definition.

290 cf Form CO; A later submission by E2 to questions of the Commission of November 30 has specified the individual quantities whereby the total was [...]*

291 With an annual contracted quantity of [...]*. Cf From CO and Answers to the Commissions questions of November 30 to E2

292 Cf Form CO - price development

293 Decision of the Danish Competition Authority on the acquisition by DONG of Naturgas Sjælland

294 Cf. Form CO

295 Cf. Agreed minutes from telephone conversation with Energinet.dk

296 Avedøre and H.C. Ørsted - E2, Skærbæk - Elsam

297 This code has already been changed several times to adjust it to regulatory changes. According to the TSO introducing further changes should not be problematic. In such a procedure entry/exit fees would have to be paid twice just like in the case of the Ellund turnarounds.

298 [...]* has no such sales.

299 [...]*

300 Reference can also be made to the evidence submitted by DONG concerning Vattenfall's recent setting up of a trading desk in Germany. Such a step would even be easier for E2 in Denmark.

301 Cf. Section on the storage and flexibility market

302 The calculation was based on E2's reply of September 1, 2005 to the Commission's questionnaire in terms of the heat production data and information from E2's web site on the heat production capacity in the particular power plants.

303 January, February, November and December

304 [...]*

305 [...]*

306 Cf. Section on the storage and flexibility market

307 [...]*

308 Cf. the reply to question 4 of the Commission's questionnaire on the Baltic Gas Interconnector submitted by E2 on 15 November 2005.

309 Cf. E2 reply to the Commission's questionnaire of December 7, 2005.

310 [...]*

311 E2 answers to the Commissions questions of October 7

312 E2 answers to the Commissions questions of November 30

313 Cf responses to the Danish Suppliers and shippers of natural gas questionnaire

314 Cf E2's reply of 14.12.2005 to the Commission's questionnaire of 13.12.2005.

315 Cf. the response from Amerada Hess to the questionnaire of the Commission for Gas Producers

316 [...]*

317 The Parties state in their reply that the Commission has used selective account as regards [...]*, firstly noting that it has already committed its gas to Germany and later not identifying any problems in relation to sales to E2. The Commission however points out, the Parties argument is based on a misreading of the information as E2 has purchased from [...]* only in the framework of its gas release programme at Emden [...]*

318 Cf. E2's Annual Report 2004.

319 "As a starting point RFG and Energinet.dk's IT system support the so called "pro-rata" principle, but if transport customers want another principle, Energinet.dk will implement these [...]" RFG §10.5.3: "If a Gas Supplier has Natural Gas transported by several Shippers in the Transmission System, the Gas Supplier's Delivered Quantity will be allocated to the Shippers on a pro rata basis in accordance with the Accepted Nominations in the Allocation Area for the Hour in question [...]* The three power plants directly connected to the transmission grid are per definition their own gas suppliers and can therefore use the abovementioned rules if they buy gas from more than one transport customer."

320 Cf .E2's Response to the commission's questionnaire of November 30, 2005

321 Cf. the competitive assessment section on the storage and flexibility market

322 Cf analysis of the storage/flexibility market.

323 http://europa.eu.int/comm/competition/antitrust/others/sector_inquiries/energy/issues_paper15112005.pdf

324 TTF: the Dutch gas hub.

325 NBP: the British gas hub.

326 [...]*

327 A virtual hub is a virtual marketplace on which gas can be traded on standard terms if that gas is physically delivered to any entry point of a transmission system, or any point within a sizeable part of a transmission system. Important virtual hubs in Europe are i.e. the National Balancing Point (NBP) in the United Kingdom or the Title Transfer Facility (TTF) in the Netherlands.

328 "Shippers" are companies having a license to nominate gas on Energinet.dk's transmission grid. All companies active on the wholesale level in Denmark therefore need a shipper's license.

329 The Commission acknowledges that DONG finds this figure rather high. However, irrespective of the precise percentage there is no doubt that such a trading/wholesale market was emerging in Denmark prior to the merger.

330 Cf E2's Response to the commission's questionnaire of November 30, 2005- E2 has been involved in swapping gas with Elsam

331 According to information provided by Energinet.dk the following companies are currently "active shippers on the Danish market": DONG, E2, Statoil Gazelle, Dansk Shell, Sydkraft and E.ON Ruhrgas, RWE Trading.

332 Cf e.g. the fact that all three customers currently connected to Energinet.dk's transmission system are Elsam and E2 power plants (H.C. Ørstedsværket, Avedøre II and Skærbækværket).

333 Cf e.g. the view expressed by HNG/MN in this regard fearing problems of competitiveness of all companies except DONG because of this reduction of liquidity as a result of the merger..

334 Cf. recital (456)

335 [...]*

336 [...]*

337 Cf. responses to the Baltic Interconnector questionnaire

338 Response of Energi E2 to the Baltic Interconnector questionnaire

339 Response of Energi E2 to the Baltic Interconnector questionnaire

340 Cf. E2's reply to the Commission's questionnaire on the Baltic Interconnector

341 Response of Energi E2 to the Baltic Interconnector questionnaire

342 Response of one of the consortium members to the Baltic Interconnector questionnaire

343 Possibly but not necessarily with gas flowing through the new Russian-German Baltic pipeline, which as mentioned is planned to enter Germany close to the German departure point of the BGI.

344 [...]*

345 Cf. E2's Annual Report for 2004.

346 [...]*

347 [...]*

348 Cf section 4.1.2 (iii) of the Response.

349 Cf. Response (Annex VIII) of DONG/Dangas to the questionnaire of the Commission from November 30, 2005.

350 Physical flow would require a reverse of flow which in view of the long-term capacity reservations in southward direction seem highly unlikely Cf. Response by DONG/Dangas to the Commission's questionnaire of November 30. A GTE Report cited by DONG in the reply to the SO)Cf. Recital (419)) states that BEB would essentially theoretically also be able to provide firm contractual south-north capacity if a serious request were made.

351 The only wholly or partly gas-fired assets that Vattenfall will acquire are the decentral power plants Hillerød and Helsingør (exclusively gas-fired) as well as block 3 of Fynsværket central CHP.

352 [...]*

353 Cf. Parties' response to the Commission's questions of December 2, 2005. It is understood that the forecasts made and cited have been made by Elsam and not Vattenfall, but it seems unlikely that with the change of ownership there can be expected significant changes to such figures.

354 Response of Elsam to the commission's questionnaire of November 30

355 The Parties again refer to the practical impossibility of re-entering gas sold at the gate, which has proven to be a false assertion on the basis of the information for the TSO. Additionally they again raise the difficulty of sourcing from two suppliers to one power plant gate, which assertion too has proven to be erroneous on the basis of the information form Energinet.dk. Finally they point out the sizable German operations of Vattenfall, which in the case of a broader relevant market including also Germany would result in Vattenfall in their view becoming an even stronger competitive constraint to DONG - the Commission has however demonstrated that a relevant market only encompassing Denmark or Denmark and Sweden can be upheld.

356 Cf paragraphs (283)-(292)

357 In its Reply to the SO, DONG does not dispute that other suppliers are unlikely to secure supplies to Elsam and E2 after the merger.

358 It is noted that other estimates arrive at somewhat higher consumption, estimates ranging from 4.0 bcm to 4.5 bcm.

359 It needs to be considered that the contracts are, indirectly, also favourable to Elsam's and E2's district heating customers.

360 E.g. it needs to be considered that an additional long-term contract which Elsam had with DONG (Levetids forlængelsaftalen) was only terminated quite recently (on 1 May 2002). The demand satisfied by this contract has not entirely disappeared.

361 Judging from fn 362 of DONG's Reply to the SO, about [...]* of this demand 2002-2004.

362 The main reason for this being long-term biomass burning obligations and new investments in a biofuel production unit. See Elsam's reply of 7 December.

363 This is also because under the current gas price conditions, the coal indexed supply contracts between DONG and E2 and Elsam are favourable to the customers.

364 Unless, Vattenfall and DONG still negotiate some transfer of the long-term "quota" that Vattenfall is to receive of DONG's contract with Elsam, which is according to the Commission's information not excluded

365 Taking into consideration that the supply business of BEB was split up between Shell and Exxon.

366 For this reason, the argument that minimum viable scale for companies already active in Denmark will not be affected by the proposed operation cannot be accepted.

367 HNG/MN's capability, due to the DCA's recent decision quoted by DONG in its Reply to the SO, to source at least part of its needs with third parties, will thus be reduced, thereby partly depriving the DCA's decision of its effect.

368 DONG's annual estimate is 0-1% for 2005-2009 in the case of other markets and an annual shrinkage of - 5%, from 2005-2009 for the decentral CHP market. This shrinkage is somewhat above the expectations of other market participants. Cf. DONG Reply on 7 December, question 9. Phase I market investigation, natural gas customers.

369 Case 556/2004.

370 Cf. Form CO: 0.91 bcm

371 Cf E2's reply of 14.12.2005 to the Commission's questionnaire of 13.12.2005. [...]*.

372 Fynsværket also has long-term biomass burning obligations following a political agreement. See Elsam's reply of 7 December.

373 As regards potential extra supplies, the Rya plant would be protected if the functioning of the wholesale market is reinforced as discussed elsewhere in this decision. Göteborg Energi can e.g. elect to purchase gas with any successful bidder in the primary auction or could purchase gas directly in the secondary auction. In addition, it is a partner in the BGI consortium and could, once this pipeline is built, source its own supplies from Germany

374 One of these plants, previously supplied by DONG has already found a new, presumably intra-company supplier. The other plant will not operate before 2009.

375 These figures are very similar to the ones provided by DONG itself (for 2004), which however also cover both the market(s) for large industrial customers and decentralised CHPs and the small business customer market. As the small business customer market is also small in overall volume, the distortion introduced by this wider scope explains some differences but does not decisively change the relations between the individual companies. [...]*. DONG was unable to provide market shares separately for large industrial customers and central power plants. However, the market investigation has clearly shown that DONG would have market shares well above [45-55%]*, and be dominant, also on the assumption of separate markets for decentral CHPs and large industrial customers.

376 In this respect the parties refer to an investigation carried out by the Danish Competition Authority according to which 26 out of 32 contracts were put out to tender in 2005.

377 See Case T-210/01 General Electric vs. the Commission, recital 149-150

378 As part of the Elsam-Nesa merger, the parties committed to sell all gas fired decentral CHP's. 63 of these has been bought by E.ON.

379 Assuming an average gas consumption of 1 mcm per MW installed capacity, which is roughly the ratio that comes out the Commissions market investigation.

380 On this point see also the submission of Statoil Gazelle and the information provided by DONG in form CO on past acquisition of customers.

381 It is noted that the decision by the DCA on the illegality of certain aspects of the long-term supply contract from DONG to HNG/MN has increased DONG's commercial freedom (and business rationale) to do so.

382 See Case T-210/01 General Electric vs. the Commission, recital 241-242

383 [...]*

384 Rothschild Study for Danish Finance Ministry: Privatisation Alternatives for DONG A/S (privatisation), May 2004

385 Cf. Draft decision by the Danish Competition Authority in the DONG-HNG/MN case.

386 Report form the Commission to the Council and European Parliament on progress in Creating the Internal Gas and Electricity market, SEC (2005)

387 Around 35% of the large gas costumers confirmed such an interest.

388 [...]*

389 See phase I gas customer questionnaire. Also in phase II decentral CHP questionnaire a majority expected negative effects, namely a reduction of competition.

390 DONG's market share on supply of electricity to large customers will be around [25-35%]* post merger

391 14 out of 34 gas (41%) customers answering the question confirmed such an interest cf. market investigation

392 Rothschild Report for Finance Ministry

393 See case T 5/02 Tetra Laval vs. Commission

394 "Elektra has many large industrial customers, who also demand gas at competitive prices. It has therefore been important for Elektra to find a partner with the necessary competencies to act professionally on the Danish gas market. This aim has been reached with the agreement with Statoil Gazelle"("Elektra har mange store erhvervskunder, der også efterspørger naturgas til konkurrencedygtige priser. Derfor har det været vigtigt for Elektra at finde en samarbejdspartner med den nødvendige kompetence til at agere professionelt på det danske naturgasmarked. Dette mål er nået med aftalen med Statoil Gazelle") Statement in Press release of 28 April 2004, made by the CEO of Elektra Tom Hefting. From StatoilGazelle's homepage.

395 What Energi Danmark offer in gas is "financial gas price fixing" for customers >0.3 mcm. There is nothing on physical deliveries, on the contrary they explicitly state that "you do not need to change your physical supplier". Therefore gas competitors to DONG will not profit from this offer.

396 Unless this refers to an abandoned past attempt by Shell to make dual fuel gas/electricity offers. If it does, then the fact that Shell has stopped making such offers suggests these companies were not the ideal partners (cf e.g. ScanEnergi's rather small electricity customer base - see electricity part of this decision). It is noted that, in the absence of the merger, Shell, like other gas companies, may have elected to team up with KE or Elsam/NESA.

397 Decentral CHP questionnaire, phase II.

398 [...]*

399 Cf E2's reply of 14.12.2005 to the Commission's questionnaire of 13.12.2005.

400 E.g Annual Report 2004: under the Heading "Flexibility of Fuel and Optimisation of Operation": At short sight the development efforts of Elsam comprise a number of projects ranging from the development of technology which can handle very different types of fuels without doing harm to the plant and its residual products to large optimisation tasks in the control rooms and processes. The driving force is the wish that the production plants are sufficiently robust and flexible to hold a mixture of fuels..." (product developed is the control room tool OCTOPUS) p.19. Similar in Elsam's 2003 Report: "it can be calculated how much power can be produced at the expense of heat"

401 Cf. E2 Strategy statement announced on E2's webpage: "E2 will improve fuel availability, infrastructure and flexibility".

402 http://www.elsam.com/multimedia/Elsam_2004_UK4.pdf: , Annual Report 2004: "In order to assume a highly recognisable and uniform image we have chosen the red galaxy as the overall symbol of all companies in the Group"

403 Annual Report 2004: "Elsam aims developing new, profitable business areas based on and supported by its core business." p.7

404 [...]*

405 Gazelle: Business plan, November 2002

406 1.1.2006, as submitted by a press release being on the Commission's file. Cf the acquisition of joint control in HEW, including - at the time - its gas activities, authorised on 20.03.2000 (cf. Case COMP/M.1842 - Vattenfall/HEW).

407 As to Shell's recent entry into the small business customer market or segment, it can be noted that this was carried out only after Shell had already established operations on the large customer market. On Statoil Gazelle's entry into neighbouring geographic markets, similar remarks can be made. In any event (i) both of these entries have only had a very limited impact and it is too early to state whether they will be sustainable, (ii) entry barriers for other companies will be higher, as no other companies are equally established in neighbouring product markets (large business customers) or neighbouring regional geographic markets, the only possible exception being HNG/MN which is in a weak financial position and, as argued elsewhere, dependent on DONG.

408 Universal Supply Obligation

409 Cf Judgment of the Court of First Instance in General Electric, T-210/01, recital 115.

410 Confirmed by the market investigation.

411 T-210/01, recital 115.

412 [...]*

413 Evidence provided both by DONG and a third party suggests that Statoil Gazelle may be rather in the role of a maverick or fringe player than part of such oligopoly.

414 See Case T-342/99 Airtours vs. The Commission.

415 [...]*

416 For example,. a short term contract by [...]* with [...]* has resulted in supply only during one specific non-winter month (September). Another contract with [...]* by [...]* foresaw supply from May to the end of September, exactly the months of lowest household gas demand.

417 Except for Statoil Gazelle who was already present on this market and possibly has entered neighbouring geographic markets (on the assumption of local markets).

418 Copenhagen Energy has about 350,000 customers; NESA about 530,000 customers.

419 Cf. www.KE.dk

420 [...]*

421 [...]*

422 [...]*

423 [...]*

424 It is noted that this mere increased threat of entry may encourage HNG/MN not to seek an alternative wholesale supplier, thus also strengthening DONG's position also on the wholesale market.

425 Commission Regulation (EC) No 802/2004 of 7 April 2004 implementing Council Regulation (EC) No 139-2004 on the control of concentrations between undertakings; Official Journal L 133 , 30/04/2004 P. 0001 - 0039

426 Most decentral CHPs are gas-fired. Of these, pre-merger, third parties control about 348 MW of generation capacity (of which 135 MW will be divested to Vattenfall), whereas third parties control 1335 MW. Source: DONG

427 Windmills are currently more important in West Denmark where their relative share is 20% whereas it is only 12% in East Denmark.

428 Energistatistik 2004

429 They can, however, increase their flexibility via the use of heat accumulators.

430 The prices paid to windmill producers for their production are specified in the Danish Electricity Supply Act

431 It has been argued that such a situation of dominance may also be present in situations in which the respective areas, in particular DK West, is not separated from other Nord Pool areas, namely in situations in which prices would fall in Denmark if interconnection congestion arises. It has been alleged that in such situations Elsam and possibly also E2 show their dominance by setting measures "avoiding" such interconnection congestion and thus keeping prices high. It is not necessary to discuss this issue further in this decision. If such alleged behaviour actually occurred it may be sanctionable under the provisions of Art.82 EC and of Danish Competition Law. While this would - due to a kind of cellophane fallacy - have an influence on the periods in which West DK or East DK can be considered separate wholesale markets (such behaviour, due to reducing the congestion rate, would obviously make such periods appear to be less frequent than they actually are) it does not have an impact on the assessment to be carried out in this decision as this decision will show that on any assumption (narrower or wider markets) there will not be competitive harm.

432 Cf. recital (10)

433 COMP/M.3867 - Vattenfall/ Elsam & E2 Assets, cleared on 23 December 2005.

434 These include a large number of small decentral CHPs and independent wind power producers.

435 These include a large number of small decentral CHPs and independent wind power producers.

436 This can be achieved by flexible CHPs with cooling facilities.

437 SImulering af VArme og EL (Simulation of heat and electricity). The model was originally developed by Elsam/Eltra as a planning tool.

438 Though the model only applies to West Denmark, the Commission has n° reason to assume that the results would be significantly different in East Denmark. In particular in the light of the fact that Denmark East appears to be better interconnected with other price regions than Denmark West.

439 It should be noted that Vattenfall would also not be liable to be hurt by such input foreclosure inter alia due to the gas supply contracts taken over from Elsam. Reference is made to the discussion in the part of the decision on gas supplies to central CHP's.

440 When DONG has the risk, it means that if prices on Nord Pool are high, DONG will benefit from it. When it does not have the risk, DONG pays the CHP a price for the electricity which is linked to the Nord Pool spot area price.

441 [...]*

442 COMP/M.3867 - Vattenfall/ Elsam & E2 Assets, Commission decision of 23 December 2005.

443 The shares are, however, confidential vis-à-vis the parties to the concentration.

444 This divestment also makes it possible that Vattenfall, in agreement with Nord Pool, will act as a further market maker in both East and West Denmark.

445 Metered customers are customers over 0.1 GWh annual consumption. Although not necessarily identical to large customers

446 OK's share was not estimated by DONG. The market investigation, however, showed that it has since its entry in 2004 been successfully acquiring new customers.

447 Including Elektra and Disam, which have been purchased by EnergiDanmark.

448 OK entered in 2004.

449 Reference was made to the fact that Lille Torup is a cavern storage whereas Stenlille is an aquifer storage.

450 In its submission to the Commission of 6 February, 2006 the Naturgas Fyn Group stated: "Given the above and the fact that any modified commitments that have any prospective of eliminating the competition concerns identified in the SO are likely to be so extensive and complex that it would not be possible for the Commission to determine with the requisite degree of certainty that effective competition will be restored in the market, NGF submits that the Commission has n° alternative but to prohibit the notified Concentration." Furthermore, on 9 February, 2006 Naturgas Fyn has submitted to the Commission a formal complaint alleging infringements under Art. 81 and 82 of the EC Treaty with regards to the long-term supply contracts between DONG and Elsam and E2 (COMP/B-1/39299). This complaint is being dealt with under the procedure set out in Regulation 773-2004.

451 Daily Contract Quantity.

452 Annual Contract Quantity.

453 Further relevant considerations on the appropriateness of the timeframe are first, the impact from 2011 of the Russian-German Baltic gas pipeline which could lead to an influx of gas into Northern Germany, which due to its geographical proximity could lead to increased competitive pressure on DONG and, secondly, the fact that by 2012 Danish gas reserves are not unlikely to enter into a phase of decline.

454 As already mentioned in recital (725), Naturgas Fyn have also considered the proposed Gas Release Programme to be ineffective to remedy the competition concerns raised in the case.

455 Statement received by the Commission from DONG on 10 March 2006.

456 Cf. Recitals (747)-(751) above.

CASE M.3868 - DONG/ELSAM/E2/KE/FE

COMMITMENTS TO THE EUROPEAN COMMISSION

submitted by DONG A/S

Pursuant to Articles 8(2) and 10(2) of Council Regulation (EC) No. 139-2004 (the "Merger Regulation"), DONG A/S ("DONG") hereby provides the following commitments (the "Commitments") in order to enable the European Commission (the "Commission") to declare the acquisition of sole control by DONG of Elsam A/S, Energi E2 A/S, Københavns Energi Holding A/S and Frederiksberg Elnet A/S (together with DONG the "Parties") compatible with the common market and the EEA Agreement by its decision pursuant to Article 8(2) of the Merger Regulation (the "Decision").

The Commitments shall take effect upon the date of adoption of the Decision.

This text shall be interpreted in the light of the Decision to the extent that the Commitments are attached as conditions and obligations, in the general framework of Community law, in particular in the light of the Merger Regulation, and by reference to the Commission Notice on remedies and under Commission Regulation (EC) No 802/2004.

Section A. Definitions

For the purpose of the Commitments, the following terms shall have the following meaning:

[BUSINESS SECRETS]*

Affiliated Undertakings: undertakings controlled by DONG, whereby the notion of control shall be interpreted pursuant to Article 3 of the Merger Regulation and in the light of the Commission Notice on the concept of concentration.

Authorized Re-delivery Points:

(i) NBP ("National Balancing Point", i.e. the virtual location in the English H-Gas transmission grid as operated presently by National Grid plc.),

(ii) ZBT ("Zeebrugge Hub", i.e. the virtual location in the Belgian H-Gas transmission grid as operated presently by Huberator SA),

(iii) TTF ("Title Transfer Facility", i.e. the virtual location in the Dutch H-Gas transmission grid as operated presently by Gas Transport Services B.V.) and

(iv) BEB-VP ("BEB Virtual Point", i.e. the virtual location in the H-Gas transmission grid as operated presently by BEB Transport und Speicher Service GmbH).

Calendar Year: the period from 1 January at 6.00 hours of any given year to 1 January at 6.00 hours the following year.

Closing: the transfer of the legal title of the Divestment Business to the purchaser.

Divestment Business: the Gas storage facility in Lille Torup as defined in Section B and the Schedule that the Parties commit to divest.

Divestiture Trustee: one or more natural or legal person(s), independent from the Parties, who is approved by the Commission and appointed by DONG and who has received from DONG the exclusive Trustee Mandate to sell the Divestment Business to a purchaser at no minimum price.

[BUSINESS SECRETS]*

Effective Date: the date of adoption of the Decision.

First Divestiture Period: the period of six months from the Effective Date.

Gas: Natural gas with quality and delivery specifications to be in accordance with the quality and delivery specifications applicable at any time at the Authorized Re-delivery Points and applied by the transmission system operator(s) at the Authorized Re-delivery Points. One cubic meter (m3) of Gas equals 12.21 kWh, expressed at gross calorific value.

Gas Year: the period from 1 October at 6.00 hours of any given year to 1 October at 6.00 hours the following year.

Imbalance Price: an average of daily market quotations for contracts at the respective Authorized Re-delivery Points with deliveries during the two full Gas Years (or, in the case of the Primary Auction in 2006, two Calendar Years) subsequent to the relevant Primary Auction, during the period of three calendar months immediately prior to the calendar month during which the Primary Auction shall take place, which will be determined in consultation with the Monitoring Trustee prior to each auction. The market quotations shall be those published in daily industry publications, such as Argus European Natural Gas Report, published by Argus Media, European Spot Gas Markets, published by Heren Energy, or Platts European Natural Gas Report, published by The McGraw-Hill Companies. Should industry publications not publish or cease to publish daily market quotations for such contracts at an Authorized Redelivery Point, then a suitable substitute for the above Imbalance Price for that Authorized Redelivery Point shall be found in consultation with the Monitoring Trustee. For contracts with deliveries at an Authorized Re-delivery Point where daily market quotations are available for only the first Gas Year after the relevant Primary Auction, an Imbalance Price based on those quotations shall constitute a suitable substitute.

Monitoring Trustee: one or more natural or legal person(s), independent from the Parties, who is approved by the Commission and appointed by DONG, and who has the duty to monitor DONG's compliance with the conditions and obligations attached to the Decision.

Personnel: all personnel currently employed by DONG Lager A/S and working at Lille Torup.

Primary Auction: an auction whereby the successful bidders agree to give to DONG the right to an equivalent annual contract quantity of Gas at an Authorized Re-delivery Point as further defined in Section B, Part I, paragraphs 1-18.

Secondary Auction: an auction whereby the successful bidders agree to pay for a certain annual contract quantity of Gas a certain auction price as further defined in Section B, Part I, paragraphs 19-21.

Trustee: the Monitoring Trustee and/or the Divestiture Trustee.

Trustee Divestiture Period: the period of three months from the end of the First Divestiture Period.

Section B. The proposed remedies The Commitments shall be implemented by DONG as set out in this Section B and in cooperation with the Trustee as further detailed in Section C and in accordance with the outline of the Commission's procedural duties as summarised in Section D.

I. Gas Release Programme

1. DONG undertakes to the Commission to implement a Gas release programme (the "Gas Release Programme") in order to facilitate access to Gas for third parties in Denmark. DONG will make available to third parties Gas in Denmark to be re-delivered by the said third parties to DONG at market locations in North West Europe. In practice, DONG intends to implement the Gas Release Programme through DONG Naturgas A/S.

The auction process

2. The amount of Gas to be released will be 400 million m3 per year to be auctioned in the years 2006, 2007, 2008, 2009, 2010 and 2011. These volumes correspond to 10% of the total Danish consumption in 2005 (which amounted to 3,994 million m3). The total quantity to be released is up to 2,400 million m3. Each year, the volumes released will be divided into lots of 40 million m3 to be supplied over two year delivery periods.

3. After the auction to be held in 2009 (i.e. the Primary and Secondary Auctions, cf. paragraph 19), DONG may apply to the Commission for termination of the Gas Release Programme if it can demonstrate, on the basis of data available from the Danish Energy Agency or other reliable sources that the market conditions have changed significantly.

[BUSINESS SECRETS]*

4. The first Primary Auction is planned to take place no later than August 2006 with supplies to commence on 1 January 2007 for the two subsequent Calendar Years 2007 and 2008. With effect from 2007, the Primary Auctions will take place no later than April with supplies to commence on 1 October of the same year for the two subsequent Gas Years. The first auction will be announced to participants in the North West European gas market and published on the website of DONG no later than two months after the Effective Date.

Sales method and auction procedure for Primary Auctions

5. The sales method will be a business-to-business internet auction. Each auction shall comprise 10 lots of 40 million m3 (equivalent to 488,444 MWh per lot at gross calorific value) to be delivered over two years with an annual contract quantity of 20 million m3 (together with any lots carried forward pursuant to paragraph 19 and/or paragraph 21 the "Auction Lots"). Lots are to be delivered by DONG in GTF (Denmark) and re-delivered in one of the Authorized Re-delivery Points.

6. The execution of the auctions will be carried out by an international IT-service provider. The auction procedure shall be handled so as to ensure that, subject to paragraph 24, DONG will not gain knowledge of the intermediary bids placed by participants to the auction. Subject to paragraph 24, only the number of lots awarded to each successful bidder and the final swap fees, as determined by the auction, shall be communicated to DONG.

7. The auction procedure will be designed so as to ensure a transparent and non discriminatory allocation of the offered quantities. At each annual auction, the final allocation of lots shall be determined only on the basis of bids submitted during the course of the auction. Bidding will be anonymous. Subject only to being pre-qualified (pursuant to paragraphs 11 and 12) for re-delivery at relevant Authorized Re-delivery Points, any bidder will be free to bid for re-delivery of lots at any combination of those Authorized Redelivery Points. However, no single bidder can bid for more than half of the Auction Lots in one auction.

8. As in recent European gas release programmes, bidding will progress by rounds during which bidders will be free to place binding bids for lots at one or more of the Authorized Re-delivery Points. Prior to each round, a set of swap fees will be announced, for which bidders then nominate the number of lots that they demand at each Authorized Re-delivery Point. The auction will start with an identical swap fee of zero across all Authorized Redelivery Points. In the event that aggregate demand across all Authorized Re-delivery Points exceeds the number of Auction Lots, a new round will be announced. Prior to each new round, bidders will be informed about total demand, aggregated over all Authorized Re-delivery Points, at the previous round. Then a new set of swap fees (which may differ across Authorized Re-delivery Points, so as to ensure that different demand conditions are adequately reflected) will be announced. The new swap fees will be at least as high as the swap fees in the previous rounds. Bidders will then nominate the number of lots that they demand at the new swap fees, for each of the Authorized Re-delivery Points. In addition, to give the auction participants increased flexibility in their bidding behaviour, they may place bids for swap fees between the swap fee of the current round and that of the previous round.

9. Each round will continue for a pre-defined time-period, and the auction will continue until aggregate demand across Authorized Re-delivery Points does not exceed the number of Auction Lots. If at the end of a round total demand should be less than the number of Auction Lots, a non-discriminatory clearing mechanism will be invoked, in order to ensure that the highest possible number of lots will be sold.

10. The final swap fees, as determined by the auction, will be fixed amounts to be paid per MWh. This payment will apply to the actual off take in GTF (Denmark). The swap fees at which the lots will be allocated can be equal to or greater than zero, depending on the result of the auction (however, cf. paragraph 16).

11. Potential bidders must register for the auction process and will be admitted to the auction process upon providing a first demand performance guarantee of EUR 200,000 per lot to be bid for. Successful bidders who acquire lots at the auction, will have to substitute this guarantee for a payment guarantee in the form of a bank guarantee which will be fixed in line with standard practice and with reference to the amount that the successful bidder will have to pay for Gas volumes acquired in the auction (including the Imbalance Price set prior to the auction). DONG will provide a similar payment guarantee to the successful bidder.

12. A shipper's license in GTF (Denmark) and at the relevant Authorized Re-delivery Points for which potential bidders intend to submit bids are needed to guarantee operational performance. DONG and all of its Affiliated Undertakings are excluded from placing bids in the auction processes, be it directly or indirectly through an agent. Furthermore, so as not to undermine the effectiveness of the Gas Release Programme, DONG, its Affiliated Undertakings and undertakings in which DONG holds more than 10% of the shares or votes shall be excluded from swapping and/or purchasing the Gas from the successful bidders, be it directly or indirectly through an agent, for delivery to DONG in Denmark.

13. The auction process and timetable, the design of the auction (including specific and technical bidding rules) and the details of the delivery and re-delivery contracts with the successful bidders (together the "Rules for Primary Auctions") shall be drafted in close cooperation with the Monitoring Trustee as early as possible after the Effective Date and forwarded to the Commission for approval before publication. The Commission shall respond without undue delay after receiving the draft of the Rules for Primary Auctions, cf. Section D. As a general principle, the auction shall be designed and the bidding rules drafted so as to attract the highest possible number of bidders and to avoid favouring redelivery at any particular one of the Authorized Re-delivery Points. DONG undertakes to fully abide by the Rules for Primary Auctions as approved by the Commission and as possibly modified according to paragraph 24 below.

Standard contract terms

14. Terms of delivery and re-delivery shall be subject to standard contract terms and include flexibility provisions to the effect that either party (the successful bidder and DONG) will only have to purchase and pay 90% of the annual contract quantity. The daily contract quantity of each lot is 669,041 kWh. At the GTF, where a "daily regime" applies, the maximum daily rate will be 110% of the daily contract quantity and the minimum daily rate will be 50% of the daily contract quantity. At the Authorized Re-delivery Points, provided that an "hourly regime" applies, the maximum daily rate and the minimum daily rate shall, in absolute terms, be divided by 24.

Imbalances between delivery and re-delivery are to be paid for on a monthly basis. The cost of imbalances to a successful bidder will be the Imbalance Price less the compensation plus the swap fee (cf. paragraphs 10 and 16), while the cost to DONG will be the Imbalance Price.

15. In addition to the flexibility provisions forming part of the standard contract terms (cf. paragraph 14) and as an alternative to standard storage packages, successful bidders are afforded the possibility of "swapping" flexibility: for each lot, the successful bidder is free to choose a higher degree of flexibility (i.e. lower "minimum daily rate" and "purchase or pay", cf. paragraph 14) which would then be applied both to the volumes to be delivered by DONG and the volumes to be re-delivered by the successful bidder at the Authorized Redelivery Point.

16. With a view to attracting and incentivising potential bidders to participate in the Gas Release Programme as detailed above, DONG shall pay a compensation to successful bidders to reflect the estimated price differences between GTF and each of the Authorized Re-delivery Points. Such compensation shall be announced prior to each auction, and shall amount to at least EUR 0.33/MWh. The compensation, which may be offset in the swap fee payment (cf. paragraph 10 above), will be paid to the successful bidder for volumes taken in GTF (Denmark).

17. The terms and conditions of the supply contracts for Gas delivered to and re-delivered by successful bidders in the auctions (e.g. nomination procedure, payment periods etc.) and in particular as regards the security of supply and penalties (e.g. situation of maintenance, force majeure, off-spec gas, interruptibility, etc.) shall be in line with standard practices at the delivery points.

18. Existing direct customers of DONG which participate in the auction process or which purchase Gas from a trader/wholesaler, who participates in the auction processes, are entitled to ask DONG, DONG being obliged to grant such request, for an extraordinary contractual modification in order to reduce their obligation to purchase gas from DONG by the amount of Gas they will have purchased as a result of the Gas Release Programme. The other terms and conditions of the Gas supply contracts between these direct customers and DONG shall not be modified, although, in absolute terms, the flexibility provisions in the existing contract would be reduced pro rata. The total amount of reductions must not be higher than the quantities sold in the auction processes.

Sales method and auction procedure for Secondary Auctions

19. Any quantities not sold in a Primary Auction as provided for in paragraph 4 above shall be carried forward to be sold in a Secondary Auction. In each Secondary Auction, DONG shall make Gas available to third parties in Denmark against payment in cash instead of redelivery of Gas. If a Secondary Auction is needed following the Primary Auction in August 2006, this Secondary Auction shall take place in October 2006 with supplies to commence on 1 January 2007 for the two subsequent Calendar Years 2007 and 2008. With effect from 2007, if a Secondary Auction is needed, it shall take place in June with supplies to commence on 1 October of the same year for the two subsequent Gas Years.

20. The Secondary Auction process and timetable, the design of the auction (including specific and technical bidding rules) and the details of the delivery contracts with the successful bidders (together the "Rules for Secondary Auctions") shall be in line with the Commission's practice and experience in recent gas release programmes under the Merger Regulation and shall be drafted in close cooperation with the Monitoring Trustee as early as possible after the Effective Date and forwarded to the Commission for approval before publication. The Commission shall respond without undue delay, cf. Section D. Each auction process will start at a minimum price to be stated in the auction rules before each Secondary Auction, provided, however, that this price shall amount to at least [Business secrets]* per cent of a proxy of "the Danish wholesale Gas price", defined as an indexation reflecting the structure of the indexation of DONG's contracts for purchase of Gas from the Danish sector of the North Sea. The said proxy shall be developed by DONG and reviewed by the Monitoring Trustee as set out in the Confidential Annex. The level of the indexation shall be calibrated to fit the historical development of the "average import price for natural gas in Germany" ("durchschnittlicher Grenzübergangspreis für Erdgas in die Bundesrepublik Deutschland gemäss Veröffentlichung des Bundesamtes für Wirtschaft und Ausfuhrkontrolle (zuzüglich Steuern) für jeweiligen Liefermonat") as calculated by BAFA. As a general principle, the auction shall be designed and the bidding rules drafted so as to attract a sufficient number of bidders.

DONG undertakes to fully abide by the Rules for Secondary Auctions as approved by the Commission and as possibly modified according to paragraph 24 below. Paragraphs 14, 17 and 18 shall apply mutatis mutandis to Secondary Auctions.

21. Notwithstanding paragraph 19 above, a Secondary Auction of 80 million m3 or less carried forward shall not be held. Instead, such lots shall be carried forward to the Primary Auction the following year. Any quantities not sold in a Secondary Auction shall be carried forward to the Primary Auction the following year. Moreover, the total quantity to be sold in the auctions (Primary and Secondary Auction) of one Calendar Year cannot exceed 480 million m3. Instead, any quantities in excess of 480 million m3 shall be carried forward to the next auction(s). Any quantities of 80 million m3 or less not sold in the last auction of the Gas Release Programme are taken out of the programme. If more than 80 million m3 remain unsold after the last auction, one additional Primary Auction and, if necessary, one additional Secondary Auction will be held in the following year.

Monitoring Trustee

22. A Monitoring Trustee and the Danish Competition Authority shall supervise the auction processes under the Gas Release Programme.

23. Within a period of one month of the conclusion of each annual auction process, DONG will inform and procure that the IT-service provider (cf. paragraph 6) informs the Monitoring Trustee and the Danish Competition Authority about the implementation of the Gas Release Programme and will make the respective tender and bidding documents, correspondence and contracts available to the Monitoring Trustee and the Danish Competition Authority for onward evaluation and reporting to the Commission.

24. Companies having participated or intending to participate in the Gas Release Programme may make comments and representations on the functioning of the programme to the Monitoring Trustee and the Danish Competition Authority. The Monitoring Trustee will evaluate such comments or representations and afford DONG the opportunity, on the basis of adequate information, to reply to the comments or representations and to adduce factual evidence to the extent necessary. The Monitoring Trustee shall prepare a report on the functioning of the auction to the Commission. From the second year of auctions to be held in 2007, for the purposes of the Commission's future general assessment of remedies in merger cases, the Monitoring Trustee's report shall also include information available to the Trustee on the effect of the Gas Release Programme on the wholesale and retail natural gas markets in Denmark. Should the findings of the Monitoring Trustee in his opinion give reason for concern as to the implementation of the remedy which cannot be resolved in an amicable manner, the Monitoring Trustee will state this in its written report to the Commission which will also be provided to DONG and to the company which had initially made the comment or representations, provided however, that business secrets of DONG must not be revealed to this company and that business secrets of this company must not be revealed to DONG. Provided that the Commission deems it necessary in order to resolve concerns identified by the Monitoring Trustee, the Commission may order DONG to propose modifications and/or improvements in the Primary Auction Rules and/or the Secondary Auction Rules on the basis of the experience gained from previous auctions. A Commission order to this effect must be issued at least six months before the auction at which the modifications and/or improvements are to take effect. Within one month of receiving such order, DONG shall submit its proposal to the Commission, the Monitoring Trustee and the Danish Competition Authority at the same time. Within one month of receiving DONG's proposal, the Danish Competition Authority and the Monitoring Trustee shall review the proposal and provide comments to the Commission, DONG and each other. The Commission shall respond without undue delay, cf. Section D.

II. Storage divestiture

1. In order to address the Commission's concerns in relation to gas storage/flexibility, DONG commits to divest, or procure the divestiture of, the Divestment Business, by the end of the Trustee Divestiture Period, as a going concern to a purchaser and on terms of sale approved by the Commission in accordance with the procedure described in paragraph 9. To carry out the divestiture, DONG commits to find a purchaser and to enter into a final binding sale and purchase agreement for the sale of the Divestment Business within the First Divestiture Period. If DONG has not entered into such an agreement by the end of the First Divestiture Period, DONG shall grant the Divestiture Trustee an exclusive mandate to sell the Divestment Business in the Trustee Divestiture Period in accordance with the procedure described in Section C, Part II, paragraph 3.

2. DONG shall be deemed to have complied with this commitment if, by the end of the Trustee Divestiture Period, DONG has entered into a final binding sale and purchase agreement, if the Commission approves the purchaser and the terms in accordance with the procedure described in paragraph 9 and if the Closing takes place no later than 1 May 2007. As soon as a final binding sale and purchase agreement has been concluded, only the purchaser shall be entitled to market and sell storage capacity for the storage year 2007/8 in the Divestment Business' Gas storage facility.

3. In order to maintain the structural effect of the Commitments, DONG shall, for a period of 10 years after Closing, not acquire direct or indirect influence over the whole or part of the Divestment Business, unless the Commission has previously found that the structure of the market has changed to such an extent that the absence of influence over the Divestment Business is no longer necessary to render the proposed concentration compatible with the common market.

Structure and definition of the Divestment Business

4. The Divestment Business consists of the storage facility in Lille Torup in Jutland, Denmark, covering underground facilities, facilities for injection and withdrawal of Gas, control systems and buildings etc. The storage facility consists of seven salt caverns with a working volume of at least 400 million m3 (total storage Gas volume amounts to 710 million m3) and a daily injection and withdrawal capacity under normal working conditions of 3.6 and 7 million m3, respectively.

The present legal and functional structure of the Divestment Business as operated to date is described in the Schedule. The Divestment Business includes:

(a) all tangible and intangible assets (including intellectual property rights), which contribute to the current operation or are necessary to ensure the viability and competitiveness of the Divestment Business;

(b) all licences, permits and authorisations issued by any governmental organisation for the benefit of the Divestment Business;

(c) all contracts, leases, commitments and customer orders of the Divestment Business; all customer, credit and other records of the Divestment Business (items referred to under (a)-(c) hereinafter collectively referred to as "Assets"); and

(d) the Personnel.

Due diligence

5. In order to enable potential purchasers to carry out a reasonable due diligence of the Divestment Business, DONG shall, subject to customary confidentiality assurances and dependent on the stage of the divestiture process:

(a) provide to potential purchasers sufficient information as regards the Divestment Business;

(b) provide to potential purchasers sufficient information relating to the Personnel and allow them reasonable access to the Personnel.

Reporting and other obligations prior to the divestment

6. DONG shall submit written reports in English on potential purchasers of the Divestment Business and developments in the negotiations with such potential purchasers to the Commission and the Monitoring Trustee no later than 10 days after the end of every month following the Effective Date (or otherwise at the Commission's request).

7. DONG shall inform the Commission and the Monitoring Trustee on the preparation of the data room documentation and the due diligence procedure and shall submit a copy of an information memorandum to the Commission and the Monitoring Trustee before sending the memorandum to potential purchasers.

8. Pending the divestiture, DONG undertakes not to decrease the service level in its retained storage facility in Stenlille or in the Divestment Business.

9. Before Closing, DONG shall enter into an agreement with Energinet.dk with a view to ensuring that the conditions for use by Energinet.dk (in connection with its TSO responsibilities) of DONG's storage facilities at Stenlille and Lille Torup are not made less favourable compared to the pre-merger situation. As part of DONG's sale of the Divestment Business, the purchaser shall enter into said agreement with Energinet.dk (unless Energinet.dk is itself is the purchaser).

The purchaser

10. In order to ensure the immediate effectiveness of the Commitments, the purchaser, in order to be approved by the Commission, must:

(a) be independent of and unconnected to the Parties;

(b) have the financial resources, proven expertise and incentive to maintain and develop the Divestment Business as a viable and active competitive force in competition with DONG and other competitors;

(c) neither be likely to create, in the light of the information available to the Commission, prima facie competition concerns nor give rise to a risk that the implementation of the Commitments will be delayed, and must, in particular, reasonably be expected to obtain all necessary approvals from the relevant regulatory authorities for the acquisition of the Divestment Business;

(d) provide a copy of any agreement, which is to be concluded between the Energinet.dk and DONG and which is to be entered into by the purchaser pursuant to paragraph 9 above (these criteria for the purchaser hereafter the "Purchaser Requirements").

Direct or indirect control or influence over the purchaser on the part of any organ of the Danish State shall not, of itself, be inconsistent with the foregoing requirements.

Commission approval

11. The final binding sale and purchase agreement shall be conditional on the Commission's approval. When DONG has reached an agreement with a purchaser, it shall submit a fully documented and reasoned proposal, including a copy of the final agreement(s), to the Commission and the Monitoring Trustee. DONG must be able to demonstrate to the Commission that the purchaser meets the Purchaser Requirements and that the Divestment Business is being sold in a manner consistent with the Commitments. For the approval, the Commission shall verify that the purchaser fulfils the Purchaser Requirements and that the Divestment Business is being sold in a manner consistent with the Commitments. The Commission may approve the sale of the Divestment Business without one or more Assets or parts of the Personnel, if this does not affect the viability and competitiveness of the Divestment Business after the sale, taking account of the proposed purchaser.

Section C. Trustee

I. Appointment Procedure

1. DONG shall appoint one (or two) Monitoring Trustee(s) to carry out the functions specified in the Commitments in Section B, Parts I and II (respectively). If DONG has not entered into a binding sale and purchase agreement one month before the end of the First Divestiture Period or if the Commission has rejected a purchaser proposed by DONG at that time or thereafter, DONG shall appoint a Divestiture Trustee to carry out the functions specified in the Commitments for a Divestiture Trustee. The appointment of the Divestiture Trustee shall take effect upon the commencement of the Trustee Divestiture Period.

2. The Trustee shall be independent of the Parties, possess the necessary qualifications to carry out its mandate, for example as an investment bank or consultant or auditor, and shall neither have nor become exposed to a conflict of interest. The Trustee shall be remunerated by DONG in a way that does not impede the independent and effective fulfilment of its mandate.

Proposal by DONG

3. No later than one week after the Effective Date, DONG shall submit a list of one or more persons whom DONG proposes to appoint as the Monitoring Trustee to the Commission for approval. No later than one month before the end of the First Divestiture Period, DONG shall, in the event of not having entered into a binding sale and purchase agreement for the Divestment Business, submit a list of one or more persons whom DONG proposes to appoint as the Divestiture Trustee to the Commission for approval. The proposal shall contain sufficient information for the Commission to verify that the proposed Trustee fulfils the requirements set out in paragraph 2 and shall include:

(a) the full terms of the proposed mandate, which shall include all provisions necessary to enable the Trustee to fulfil its duties under these Commitments;

(b) the outline of a work plan which describes how the Trustee intends to carry out its assigned tasks;

(c) an indication whether the proposed Trustee is to act as both Monitoring Trustee and Divestiture Trustee or whether different trustees are proposed for the two functions.

Approval or rejection by the Commission

4. The Commission shall have the discretion to approve or reject the proposed Trustee(s) and to approve the proposed mandate subject to any modifications it deems necessary for the Trustee to fulfil its obligations. If only one name per function is approved, DONG shall appoint, or cause to be appointed, the individual or institution concerned as Trustee, in accordance with the mandate approved by the Commission. If more than one name is approved, DONG shall be free to choose the Trustee to be appointed from among the names approved. The Trustee shall be appointed within five working days of the Commission's approval, in accordance with the mandate approved by the Commission.

New proposal by DONG

5. If all the proposed Trustees are rejected, DONG shall submit the names of at least two more individuals or institutions within five working days of being informed of the rejection, in accordance with the requirements and the procedure set out in paragraphs 1 and 4.

Trustee nominated by the Commission

6. If all further proposed Trustees are rejected by the Commission, the Commission shall nominate a Trustee, whom DONG shall appoint, or cause to be appointed, in accordance with a trustee mandate approved by the Commission.

II. Functions of the Trustee

1. The Trustee shall assume its specified duties in order to ensure compliance with the Commitments. The Commission may, on its own initiative or at the request of the Trustee or DONG give any orders or instructions to the Trustee in order to ensure compliance with the conditions and obligations attached to the Decision.

Duties and obligations of the Monitoring Trustee

2. The Monitoring Trustee shall:

(a) propose in its first report to the Commission a detailed work plan describing how it intends to monitor compliance with the obligations and conditions attached to the Decision.

(b) assume the other functions assigned to the Monitoring Trustee under the conditions and obligations attached to the Decision and the Commitments including the Confidential Annex;

(c) propose to DONG such measures as the Monitoring Trustee considers necessary to ensure DONG's compliance with the conditions and obligations attached to the Decision;

[BUSINESS SECRETS]*

(e) review and assess potential purchasers as well as the progress of the divestiture process and verify that, dependent on the stage of the divestiture process, (i) potential purchasers receive sufficient information relating to the Divestment Business and the Personnel in particular by reviewing, if available, the data room documentation, the information memorandum and the due diligence process, and (ii) potential purchasers are granted reasonable access to the Personnel;

(f) provide to the Commission and sending DONG a non-confidential copy at the same time, a written report (i) on the Gas Release Programme for each auction within one month of the completion of the auction process (also with a copy to the Danish Competition Authority) and (ii) on the divestment process within 15 days after the end of every month of the First Divestiture Period. In addition to these reports, the Monitoring Trustee shall promptly report in writing to the Commission, with a copy to the Danish Competition Authority (only as far as the Gas Release Programme is concerned) and sending DONG a non-confidential copy at the same time, if it concludes on reasonable grounds that DONG is failing to comply with these Commitments;

(g) within one week after receipt of the documented proposal referred to in Section B, Part II, paragraph 9, submit to the Commission a reasoned opinion as to the suitability and independence of the proposed purchaser, the viability of the Divestment Business after the sale and whether the Divestment Business is sold in a manner consistent with the conditions and obligations attached to the Decision, in particular, if relevant, whether the sale of the Divestment Business without one or more Assets or not all of the Personnel affects the viability of the Divestment Business after the sale, taking account of the proposed purchaser.

Duties and obligations of the Divestiture Trustee

3. Within the Trustee Divestiture Period, the Divestiture Trustee shall sell at no minimum price the Divestment Business to a purchaser, provided that the Commission has approved both the purchaser and the final binding sale and purchase agreement in accordance with the procedure laid down in Section B, Part II, paragraph 9. The Divestiture Trustee shall include in the sale and purchase agreement such terms and conditions as it considers appropriate for an expedient sale in the Trustee Divestiture Period. In particular, the Divestiture Trustee may include in the sale and purchase agreement such customary representations and warranties and indemnities as are reasonably required to effect the sale. The Divestiture Trustee shall protect the legitimate financial interests of DONG, subject to DONG's unconditional obligation to divest the Divestment Business at no minimum price within the Trustee Divestiture Period.

4. In the Trustee Divestiture Period (or otherwise at the Commission's request), the Divestiture Trustee shall provide the Commission with a comprehensive monthly report, written in English, on the progress of the divestiture process. Such reports shall be submitted within 15 days after the end of every month with a simultaneous copy to the Monitoring Trustee and a non-confidential copy to DONG.

III. Duties and obligations of DONG

1. DONG shall provide and shall cause its advisors to provide the Trustee with all such cooperation, assistance and information as the Trustee may reasonably require to perform its tasks. The Trustee shall have full and complete access to any of DONG's or the Divestment Business' books, records, documents, management or other personnel, facilities, sites and technical information necessary for fulfilling its duties under the Commitments and DONG and the Divestment Business shall provide the Trustee upon request with copies of any document. DONG and the Divestment Business shall make available to the Trustee one or more offices on their premises and shall be available for meetings in order to provide the Trustee with all information necessary for the performance of its tasks.

2. DONG shall provide the Monitoring Trustee with all managerial and administrative support that it may reasonably request on behalf of the management of the Divestment Business. This shall include all administrative support functions relating to the Divestment Business which are currently carried out at headquarters' level. DONG shall provide and shall cause its advisors to provide the Monitoring Trustee, on request, with the information submitted to potential bidders in the Gas Release Programme. On request DONG shall inform the Monitoring Trustee on possible bidders, submit a list of potential bidders, and keep the Monitoring Trustee informed of all material developments in the auction processes.

3. DONG shall grant or procure Affiliated Undertakings to grant comprehensive powers of attorney, duly executed, to the Divestiture Trustee to effect the sale, the Closing and all actions and declarations which the Divestiture Trustee considers necessary or appropriate to achieve the sale and the Closing, including the appointment of advisors to assist with the sale process. Upon request of the Divestiture Trustee, DONG shall cause the documents required for effecting the sale and the Closing to be duly executed.

4. DONG shall indemnify the Trustee and its employees and agents (each an "Indemnified Party") and hold each Indemnified Party harmless against, and hereby agrees that an Indemnified Party shall have no liability to DONG for any liabilities arising out of the performance of the Trustee's duties under the Commitments, except to the extent that such liabilities result from the wilful default, recklessness, gross negligence or bad faith of the Trustee, its employees, agents or advisors.

5. At the expense of DONG, the Trustee may appoint advisors (in particular for corporate finance, auction or legal advice), subject to DONG's approval (this approval not to be unreasonably withheld or delayed) if the Trustee considers the appointment of such advisors necessary or appropriate for the performance of its duties and obligations under the mandate, provided that any fees and other expenses incurred by the Trustee are reasonable. Should DONG refuse to approve the advisors proposed by the Trustee the Commission may approve the appointment of such advisors instead, after having heard DONG. Only the Trustee shall be entitled to issue instructions to the advisors. Paragraph 4 shall apply mutatis mutandis. In the Trustee Divestiture Period, the Divestiture Trustee may use advisors who served DONG during the Divestiture Period if the Divestiture Trustee considers this in the best interest of an expedient sale.

IV. Replacement, discharge and reappointment of the Trustee

1. If the Trustee ceases to perform its functions under the Commitments or for any other good cause, including the exposure of the Trustee to a conflict of interest:

(a) the Commission may, after hearing the Trustee, require DONG to replace the Trustee; or

(b) DONG, with the prior approval of the Commission, may replace the Trustee.

2. If the Trustee is removed according to paragraph 1, the Trustee may be required to continue in its function until a new Trustee is in place to whom the Trustee has effected a full hand over of all relevant information. The new Trustee shall be appointed in accordance with the procedure referred to in section I.

3. Beside the removal, the Trustee shall cease to act as Trustee only after the Commission has discharged it from its duties after all the Commitments with which the Trustee has been entrusted have been implemented. However, the Commission may at any time require the reappointment of the Trustee if it subsequently appears that the relevant remedies might not have been fully and properly implemented.

Section D. Outline of the Commission's procedural duties

Implementation of the Gas Release Programme

1. The Commission's procedural duties for the implementation of the Commitments in Section B, Part I above (the Gas Release Programme) can be summarised as follows:

2. Section B, Part I, paragraphs 13 and 20: Once DONG, in cooperation with the Monitoring Trustee, has drafted the Rules for Primary Auctions and the Rules for Secondary Auctions, respectively, the said rules shall be forwarded to the Commission for approval. The Commission shall respond without undue delay.

3. Section B, Part I, paragraph 24: In the event that the Monitoring Trustee should identify concerns regarding the functioning of the Gas Release Programme, and the Commission, on the basis thereof, orders DONG to propose modifications and/or improvements to the Primary Auction Rules and/or the Secondary Auction Rules at least six months before the auction, DONG shall provide its proposal to the Commission, the Monitoring Trustee and the Danish Competition Authority within one month of receiving such order. The Monitoring Trustee and the Danish Competition Authority shall review DONG's proposal and provide comments to the Commission, DONG and each other within one month. The Commission shall respond without undue delay.

4. Section B, Part I, paragraph 3: In the event that DONG applies to the Commission pursuant to Section B, Part I, paragraph 3 for termination of the Gas Release Programme, the Commission shall take a decision within two months of receiving a fully reasoned submission from DONG accompanied by the Monitoring Trustee's report.

Implementation of the storage divestiture

5. The Commission's procedural duties for the implementation of the Commitments in Section B, Part II above (the storage divestiture) can be summarised as follows:

6. Section B, Part I, paragraph 11: The Commission shall take a decision on the approval of the purchaser once DONG has reached an agreement with a purchaser and submitted a fully documented and reasoned proposal, including a copy of the final agreement(s), to the Commission and the Monitoring Trustee.

Section E. Mediation

1. In the event that a third party, notably but not exclusively a bidder in an auction under the Gas Release Programme, has reasons to believe that DONG is failing to comply with the requirements of the Commitments, the Monitoring Trustee may be instructed by the Commission to act as mediator to attempt to settle the dispute amicably. The Monitoring Trustee shall be allowed to appoint further professionals to assist him in the mediation process. The appointment of such professionals must be acceptable to DONG and the third party.

2. The procedure of the mediation process under the Monitoring Trustee stewardship shall be proposed by the Monitoring Trustee and agreed to by DONG and the third party. The Monitoring Trustee shall also set forward a proposal as to who bears the costs of the mediation procedure which shall take into account general mediation standards. The procedure shall correspond to the European Code of Conduct for Mediators as developed by the Commission an launched on 2 July 2004 in as far as appropriately applicable under the stewardship of the Monitoring Trustee. The procedure shall comprise a first phase of exchange of written observations between the parties. It is envisaged that the deadlines to reply to the observations shall be set by the Monitoring Trustee in a timely manner in order to accelerate the mediation process.

3. Following the exchange of written observations the Monitoring Trustee shall arrange for negotiations between DONG and the third party under his stewardship with a view to reaching an amicable solution of the dispute. Should such negotiations not produce an amicable solution, the Monitoring Trustee is then empowered to recommend a solution which shall become binding upon DONG and the third party unless DONG and/or the third party have lodged its/their opposition to this recommendation within one month from receiving a fully reasoned version of the recommended solution in the English language. In the latter case the Monitoring Trustee shall prepare a report for the Commission on the outcome of the mediation process. A copy of this report shall also be provided to the Danish Competition Authority and to DONG and the third party. Nothing in the mediation procedure shall affect the powers of the Commission to take decisions in relation to the Commitments in accordance with its powers under the Merger Regulation and the EC Treaty.

Section F. The Review Clause

The Commission may, where appropriate, in response to a request from DONG showing good cause and accompanied by a report from the Monitoring Trustee:

(i) Grant an extension of the time periods foreseen in the Commitments, or

(ii) Waive, modify or substitute, in exceptional circumstances, one or more of the undertakings in these Commitments.

Where DONG seeks an extension of a time period, it shall submit a request to the Commission no later than one month before the expiry of that period, showing good cause. Only in exceptional circumstances shall DONG be entitled to request an extension within the last month of any period.

Copenhagen, 28 February 2006

For DONG A/S:

Opinion

of the ADVISORY COMMITTEE on CONCENTRATIONS given at its 137th meeting on 1 March 2006 concerning a draft decision relating to Case COMP/M.3868 - DONG/Elsam/Energy E2

Rapporteur : HUNGARY

1. The Advisory Committee agrees with the Commission that the notified operation constitutes a concentration within the meaning of Article 3 (1)(b) of Council Regulation (EC) No 139-2004 and that it has a Community dimension.

2. The Advisory Committee agrees with the Commission that for the purposes of assessing the present operation, the relevant product markets are:

in the natural gas sector:

a) the market for storage or alternatively the market for flexibility of natural gas;

b) the wholesale market of natural gas;

c) the market for the supply of natural gas to central combined heat and power plants (CHPs);

d) the market or markets for the supply of natural gas to large industrial customers and to decentralized CHPs;

e) the market or markets for the supply of natural gas to non-metered small business customers and to households;

in the electricity sector:

f) the wholesale market of electricity (irrespective of whether it includes or not bilateral wholesales and ancillary services, constituting possibly separate markets),

g) the market for financial derivatives of electricity (irrespective of whether it includes or not CfDs constituting a possibly separate market) ,

h) the retail market for electricity sales to metered (business) customers,

i) the retail market for electricity sales to non-metered (predominantly household) customers;

for other markets:

j) the market for district heating,

k) the market for fly ash production,

l) the market for trading of CO2 emission rights.

3. The Advisory Committee agrees with the Commission that for the purposes of assessing the present operation, the relevant geographic markets are:

in the natural gas sector:

a) the market for storage, or alternatively the Danish market for flexibility of natural gas, for Denmark is Danish;

b) the market for storage or alternatively the Swedish market for flexibility of natural gas for Sweden is Swedish or alternatively Swedish-Danish in scope;

c) the wholesale market for natural gas for Denmark is Danish in scope;

d) the wholesale market for natural gas for Sweden is Swedish or alternatively Swedish-Danish in scope;

e) the market for the supply of natural gas to Danish central CHPs is Danish;

f) the market for the supply of natural gas to Swedish central CHPs is Swedish or alternatively Swedish-Danish in scope;

g) the market or markets for the supply of natural gas to large industrial customers and to decentralized CHPs is Danish in scope);

h) the market or markets for the supply of natural gas to small business customers and to households is/are Danish-national or Danish-regional in scope; in the electricity sector:

i) the wholesale market of electricity is East Danish and West Danish or wider; if ancillary services constitute one or more relevant markets then the geographic scope of such market(s) is East Danish and West Danish;

j) the market for financial derivatives of electricity encompasses the Nord Pool area if excluding CfDs; if CfDs constitute a separate market then the relevant markets are the individual NordPool areas, i.e. East Danish and West Danish;

k) the retail market for electricity sales to metered (business) customers is Danish in scope;

l) the retail market for electricity sales to non-metered (predominantly household) customers is Danish-national or Danish-regional in scope;

for other markets:

m) the market for district heating is local;

n) the scope of the geographic market for fly ash production can be left open;

o) the market for CO2 trading is EU-wide.

4. The Advisory Committee agrees with the Commission that DONG has a dominant position:

in the natural gas sector:

a) on the market for gas storage or gas flexibility for Denmark;

b) on the wholesale market of natural gas for Denmark;

c) on the market or markets for the supply of natural gas to large industrial customers and decentralised CHPs in Denmark;

d) on the market or markets for the supply of natural gas to small business customers and/or households on two of the five Danish regional areas (single dominance) or alternatively on a national market with HNG/MN? (joint dominance),

e) and that the question for the markets for gas storage or gas flexibility and gas wholesale for Sweden can remain open as any commitments which are sufficient to remedy the competition concerns on the respective markets for Denmark also solve such problems on the markets for Sweden.

5. The Advisory Committee agrees with the Commission that the transaction will give DONG the ability and incentive on the Danish gas storage or gas flexibility market to raise rivals' costs of storage post merger, increase storage tariffs in Denmark and to use the central CHP plants owned by Elsam and Energi E2 for flexibility purposes and that the transaction would thus strengthen DONG's dominant position on the possible Danish market for storage or on the possible Danish market for storage/flexibility and would lead to a significant impediment to effective competition.

6. The Advisory Committee agrees with the Commission that the transaction will strengthen DONG's dominant position on the Danish wholesale market especially through removing significant actual and potential competitive constraints to DONG and through foreclosure of demand and that this is likely to lead to a significant impediment to effective competition.

7. The Advisory Committee agrees with the Commission that the concentration does not lead to a significant impediment of effective competition on the markets for natural gas supplies to central CHPs either in Denmark or in Sweden.

8. The Advisory Committee agrees with the Commission that the concentration will lead to a significant impediment of effective competition on the market or markets of the supply of natural gas to large industrial customers and decentralised CHPs especially through raising entry barriers and eliminating potential competition.

9. The Advisory Committee agrees with the Commission that the concentration will significantly impede competition on the market or markets for the supply of natural gas to households and small business customers irrespective of whether these markets are defined as regional or as national in particular through the strengthening of a dominant position.

10. The Advisory Committee agrees with the Commission that the notified concentration does not lead to a significant impediment of effective competition, in particular as a result of the creation or strengthening of a dominant position on either any possible affected market for physical electricity or for financial derivatives of electricity.

11. The Advisory Committee agrees with the Commission that in the absence of competition concerns it is not necessary to examine further the effect of the concentration on the markets for district heating, fly ash production and CO2 trading.

12. The majority of the Advisory Committee agrees with the Commission that the commitments submitted by the parties concerning storage divestiture and the gas release programme are suitable and in combination with each other sufficient to remove the identified competition problems on the following markets:

a) market for gas storage or gas flexibility irrespective of whether only for Denmark or also for Sweden;

b) the wholesale markets of natural gas for Denmark and potentially also for Sweden;

c) the market or markets for the supply of natural gas to large industrial customers and/or decentralised CHPs in Denmark;

d) the market or markets for supply of natural gas to small business customers and/or households in Denmark.

A minority disagrees with point b).

13. The Advisory Committee agrees with the Commission that, subject to full compliance with the commitments offered by the parties the proposed concentration does not significantly impede effective competition and as a result the proposed concentration can be declared compatible with the common market and the functioning of the EEA Agreement.

14. The Advisory Committee asks the Commission to take into account all the other points raised during the discussion.

<emplacement tableau>

FINAL REPORT OF THE HEARING OFFICER IN CASE COMP/M.3868 - Dong/Elsam/Energi E2

(pursuant to Article 15 of Commission Decision (2001-462-EC, ECSC) of 23 May 2001 on the terms of reference of Hearing Officers in certain competition proceedings - OJ L162, 19.06.2001, p.21)

The proposed concentration

On 13 September 2005, the Commission received a notification of a proposed concentration pursuant to Article 4 of Council Regulation (EC) No 139-2004 of 20 January 2004 ("the Merger Regulation") whereby the Danish gas incumbent DONG acquires within the meaning of Article 3(1)(b) of the Merger Regulation control of two important Danish power generators (Elsam in the West; Energi E2 in the East) and of two electricity distribution companies, namely København Energi and Frederiksberg Elnet (the latter four companies are hereinafter referred to as "other involved parties").

The initiation of proceedings and provision of key documents

At the end of the first phase of the investigation, the Commission concluded that the concentration raised serious doubts as to its compatibility with the common market and with the EEA Agreement. On 18 October 2005, the Commission therefore initiated proceedings in accordance with Article 6(1)(c) of the Merger Regulation.

On 7 November 2005 DONG was provided with access to the "key documents" in the Commission file in accordance with chapter 7.2. of the "Best Practices on the conduct of EC merger control proceedings" ("Best Practices"), as determined by the Directorate General for Competition.

The Statement of Objections and the parties' reply

A Statement of Objections was sent to DONG on 19 December 2005. The other involved parties were provided with a non-confidential version of these objections. In the following days, access to the Commission's file was granted. DONG was asked to reply by 9 January 2006. DONG and the other involved parties submitted a joint reply on 5 January 2006.

In their reply of 5 January 2006, DONG considered that "their ability to respond to the Commission's concerns in the SO had been undermined by the Commission's approach to the provision on access to file". More specifically, they stated that "the level of redaction of third party submissions was not satisfactory and made it difficult for the parties to properly scrutinise or assess the documents". By letter of 16 January 2006, I informed DONG that subsequent to Article 8 of Commission Decision 2001-462-EC, ECSC of 23 May 2001 on the terms of reference of hearing officers in certain competition proceedings (O.J. No. L 162, 19.6.2001, p. 21), they were entitled to lodge a reasoned request for access to file with me should they consider that they had not received the information they were entitled to in order to prepare their defence. DONG has not reacted to my letter.

Neither DONG, nor the other involved parties requested to develop their arguments in a formal oral hearing.

The participation of a third party

Upon request, I admitted Naturgas Fyn Group as a third party according to Article 18(4) of Council Regulation 139-2004 on 22 December 2005. They were sent a non-confidential version of the Statement of Objections.

The commitments and the results of the market test

On 30 January 2006, DONG offered commitments. The results of the market test of these commitments conducted as from 1 February 2006 were mixed. Concerns were raised in particular with regard to the effectiveness of the two-step auction process provided for in the envisaged Gas Release Programme. As a result, the commitments were amended. I have not been asked to verify the objectivity of the enquiry.

The draft decision and the respect of the right to be heard

In the light of the commitments proposed and having analysed the results of the market test, the draft decision concludes that the proposed concentration is compatible with the common market and with the EEA Agreement.

In the light of the above, I consider that the rights to be heard of all participants to the present proceeding have been respected.

Brussels, 21 February 2006

Serge Durande