EC, January 30, 2008, No 39.326

THE COMMISSION OF THE EUROPEAN COMMUNITIES,

Having regard to the Treaty establishing the European Community, Having regard to Council Regulation (EC) No 1-2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty (1), and in particular Articles 23 paragraph 1 lit e) and Article 20 paragraph 2 lit d) thereof, Having regard to the Commission decision of 29 September 2006 to initiate proceedings in this case, Having given the undertaking concerned the opportunity of being heard pursuant to Article 27(1) of Regulation (EC) No 1-2003 and Article 12 of Commission Regulation (EC) No 773-2004 of 7 April 2004 relating to the conduct of proceedings by the Commission pursuant to Articles 81 and 82 of the EC Treaty (2), After consulting the Advisory Committee on Restrictive Practices and Dominant Positions (3), Having regard to the final report of the Hearing Officer (4), Whereas:

I. Facts

A. Undertaking concerned

1. E.ON AG (E.ON) is an energy company (mainly electricity and gas) operating in Germany and other Member States of the EU. E.ON Energie AG (EE) is a wholly-owned subsidiary of E.ON, inter alia operating in the electricity generation and supply sector and in the trade in electricity. E.ON is based in Düsseldorf, while EE is based in Munich.

B. Actual course of events

2. Having come across evidence of anti-competitive conduct by the abovementioned undertakings, the Commission decided on 24 May 2006 to conduct an inspection pursuant to Article 20 of Regulation (EC) No 1-2003 on the premises of E.ON and of the undertakings controlled by it (5). This explicitly included EE. It was decided that the inspection would start on 29 May 2006.

3. The decision specifies that the Commission may, pursuant to Article 20(2)(d) of Regulation (EC) No 1-2003, seal any business premises and books or records during the inspection. The decision also contained a reference to Article 23(1)(e) of Regulation (EC) No 1-2003, which empowers the Commission to impose a fine where seals affixed by Commission officials are broken.

4. The inspection of EE began at 14:00 on 29 May 2006 at EE's business premises situated at Brienner Straße 40, 80333 Munich. It was conducted by four representatives of the Commission (6) and six representatives of the Bundeskartellamt (7) - all carrying their staff cards. EE stated, when the inspection decision was handed to it, that it would cooperate with the Commission and submit to the inspection. The certificate of service was signed by EE (8). The inspection team began examining the business documents at about 15:00.

5. The documents found and pre-sorted on that day were stored at about 19:00 - without having been copied or (completely) catalogued - in room G 505, on the fifth floor of building section G, made available to the Commission by EE for that purpose. The room was then locked by the leader of the inspection team, in the presence of company representatives, lawyers and the other members of the inspection team, and secured against unauthorised access by means of an official seal of the type used for this purpose by the Commission (9). The record of sealing (10) was also signed by a representative of EE, who was informed of the significance of the seal and the consequences of breaking it. When the inspection team left the business premises of EE at about 19:30, the seal was intact and affixed tightly to the door and door frame of room G 505 (11).

6. The seal affixed to the office door is a blue adhesive film with yellow stripes along the upper and lower edge and the yellow circle of stars of the European Union. In the lower yellow area, it is stated that the Commission may impose a fine if the seal is broken. The label stock material used to produce the seal was produced by 3M in December 2002 (12). The Commission had the abovementioned details printed on the label stock material by a printing firm in the first quarter of 2004, and the product has, since 1 May 2004, been used by the Commission to seal premises and items of furniture during inspections carried out under Article 20 of Regulation No 1-2003.

7. The seal affixed by the Commission is a security seal which has been designed especially for security-relevant application (e.g. sealing of flight recorders, transport boxes and rooms). Unlike paper seals, a breach of seal in case of the used plastic seals is not shown by a rupture of the seal. Once the seal has been affixed, its removal from the surface becomes detectable. It is impossible to reaffix it without this being apparent. When the product is removed, the white adhesive remains stuck to the surface displaying twelve-point (approx. 5 mm) high VOID markings spread over the entire surface of the film. The detached seal becomes transparent at these points, so that the VOID markings are clearly evident on the seal. In particular, the number and size of the VOID indicia make it de facto impossible to affix the seal in exactly the same position in which it was previously affixed. Even if that is successfully accomplished, the VOID message remains clearly visible.

8. When the inspection team returned at around 8:45 on 30 May 2006, it was found by the inspection team together with the company representatives and the lawyers representing EE that the condition of the seal affixed to the office door of room G 505 had changed significantly.

9. Although the seal was still on the office door in the morning of 30 May 2006 and was not torn or cut, the VOID message was clearly evident over its entire surface. Furthermore, the seal had been displaced by about 2 mm upwards and sideways, as was evident from the fact that traces of smeared adhesive were evident around the edge of the seal (13). Lastly, traces of adhesive were also evident on the rear of the detached seal, but beside the transparent sections to which they were originally attached (14). The relevant part of the report is worded as follows (15):

- The entire seal was displaced by about 2 mm upwards and sideways so that traces of adhesive were visible on the lower and right-hand edges of the seal.

- The "VOID" message was clearly visible over the entire surface of the seal, which was nevertheless still affixed diagonally from the door frame to the surface of the door and was not torn.

- After the door was opened by the Commission officials (Mr [...]*), the seal remained intact, i.e. it did not tear, and on the rear of the seal (the adhesive side) white traces of the "VOID" message were evident.

- When the seal is removed, the white "VOID" message normally remains on the substrate, and that was the case here, too, the message being in fact on the door surface.

- numerous white traces were visible on the adhesive surface of the seal, not on, but beside the corresponding transparent sections of the "VOID" message on the rear of the seal.

10. At 9:15 the door was opened by the leader of the inspection team without further damaging the seal. The part of the seal which had been affixed to the door was removed, while the part of the seal which had been affixed to the door frame remained affixed without change.

11. As mentioned, the documents stored in the room had not yet been catalogued, mainly because of their quantity. The inspection team was therefore not able to ascertain whether the pre-sorted documents stored there were still complete.

12. The leader of the inspection team drew up a breach-of-seal report. (16) The report was signed by a representative of the Commission and a representative of the Bundeskartellamt. The other members of the inspection team were also witness to the damaged state of the seal according to the breach-of-seal report (17).

13. The in-house and external representatives of EE who were present on that day during the inspection did not dispute the altered state of the seal (18), but they nevertheless refused to sign the report on the breach of the seal. Instead, on the evening of 30 May 2006, EE drew up a supplementary statement setting out the following three points (19):

"1. When the door was opened, no change in the documents stored in the room was ascertained.

2. On removal of the seal on the evening of 30.5 prior to resealing, the VOID message on the door frame was completely intact.

3. Mr [...]* was present when the seal was affixed on the previous evening and had the impression that the process was unusually protracted."

14. The inspection resumed on 30 May 2006 at 9:30, while parallel to this the circumstances surrounding the breach of seal were investigated. During the afternoon of that day, a number of digital photographs of the seal still attached to the door frame were taken. The pictures - especially of the insofar unchanged part of the seal still adhering to the door frame (20) - confirm the statements made in the breach-of-seal report, in particular the extensive appearance of the VOID messages, the displacement of the adhering part of the seal of approx. 2 mm and the traces of smeared adhesive around the edge of the seal. (21) Room G 505 was resealed and a memorandum was drawn up indicating the possible legal consequences of a breach of seal. The affixing of the seal on the same door on the second day was without incident. The seal showed no signs of tampering on the morning of 31 May 2006. The inspection ended on the evening of 31 May 2006.

C. Course of the proceedings

15. In order to further clarify the facts of the matter, the Commission sent EE on 9 August 2006 a request for information pursuant to Article 18 of Regulation No 1-2003, to which EE replied by letter dated 23 August 2006. Further requests for information were sent on 29 August 2006 to the manufacturer of the label stock material, 3M Europe S.A./N.V. ("3M"), on 31 August 2006 to the cleaning firm [...]* which cleans EE's premises and on 1 September 2006 to [...]*, the security firm which looks after the security of EE's premises. In addition, all ten members of the inspection team filled in questionnaires with their comments regarding the affixing of the seal and its condition on the morning of 30 May 2006 (22).

16. In its reply to the request for information from 9 August 2006, EE informed the Commission that a policy of strict control of access to the relevant building existed (23) and that, as far as the company was aware, "no unauthorised persons were in the buildings" during the relevant period (i.e. in the night from 29 May to 30 May 2006).

17. As a possible explanation for the change in the seal, EE suggested that a "partial loosening of the seal" might have occurred as a result of "poor adhesion of the seal to the surface (e.g. because of the building materials used)". EE also stated that a cleaner working for the cleaning firm [...]* had wiped the seal with a damp micofibre cloth and cleaning product and might possibly have displaced the seal slightly as a result (24).

18. In reply to the request for information sent to it, the cleaning firm [...]* stated that the cleaning product "Synto" had been used in cleaning the office walls, that the cleaner concerned had not been informed of the significance of the seal and that she had "not noticed any change to the seal in cleaning the door" (25). [...]* sent a bottle of the original cleaning agent to the Commission.

19. Moreover, in discussions with the Commission on 29 August 2006, EE stated that only the leader of the inspection team and the security service had keys to room G 505 and that the security guard had not given the key to the cleaning woman (or any other person). EE concluded from this that none of the pre-sorted documents had been removed from the room. However, in reply to the request for information sent to it, [...]*stated that, in addition to the key held by the inspection team, 20 other keys to the same room were in circulation, held by employees of [...]* and other firms working for EE (26).

20. On 2 October 2006, the Commission sent EE the statement of objections. In it, having assessed the available evidence, the Commission came to the preliminary conclusion that the seal had been broken and that such breach must be attributed to EE because of its organisational control of the relevant office building.

21. By letter dated 10 October 2006, EE asked the Commission to send it the original seals so that they could be examined by an expert. Though the Commission in principle was willing to transmit the original seal, it turned down the request in this concrete case. Indeed, the Commission has to safeguard, in view of the risk of a possible copy of the original seal, that the seals do not get into the hands of third parties without supervision. Even inside the Commission's premises, the seals are subject to strict control and only handed over to Commission officials in case of an inspection. It was also not possible for the Commission to accept the proposal of marking the seals (by punching them), since such a modification might have had an influence on the characteristics of the seal. EE was not prepared to accede to the Commission's suggestion that Commission representatives be allowed to be present during the examination. The Commission had to insist on this condition since otherwise tampering with the sample and the danger of reproduction of official seals could not be ruled out. (27).

22. EE was allowed access to the file by the Commission. Thus, by letter dated 11 October 2006, the Commission sent EE a CD-ROM on which all accessible documents on file were stored.

23. EE replied in detail to the statement of objections by letter dated 13 November 2006. In an annex it enclosed among other things a scientific and technical opinion. Essentially, EE argued that the statement of objections was based on incorrect or unsound, i.e. scientifically and technically unsubstantiated, facts and that the legal assessments, particularly as regards the question of intention or negligence in the breach of seal, were not correct (28).

24. On 6 December 2006, at EE's request, an oral hearing took place with the Hearing Officer. At it, EE put forward the concerns it had already argued in its reply to the statement of objections regarding a series of factual and legal points made by the Commission. Using video evidence which it had produced itself, EE argued in particular that external factors such as vibrations in the office wall and shaking of the locked door could have resulted in the VOID markings. Irrespective of the question whether and to what extent the conditions shown in EE's film did actually correspond to the real conditions on 29/30 May 2006, the film shows that vibrations only lead to the appearance of VOID-signs in the immediate area around the gap between door and frame, but not over the whole surface of the seal. EE proposed that further tests be carried out to determine the performance of the seal, the more so as the guarantee period of the seal had expired. Following the hearing, the Commission sent a further request for information to EE on 6 December 2006. By letter dated 18 December 2006, EE provided the information and documents requested.

25. The seal manufacturer 3M also attended the hearing on 6 December 2006 and made a number of comments of its own on the shelf life, performance and handling of the seals. On 8 December 2006, the Commission sent 3M a request for information asking it to confirm in writing some of the comments made at the oral hearing. These comments related to:

- the significance of the technical bulletin as regards the performance and use of the seals,

- its own experience with use of the type 7866 labels,

- the date of manufacture of the relevant label stock material,

- the possible effects of cleaning products on the seals, and in particular the likelihood of false positives (i.e. the appearance of VOID markings when the seal has not been broken) occurring under the influence of such external factors,

- the possible effects of vibrations on the seals, and

- the likelihood of false positives.

26. By letter dated 21 December 2006, 3M replied to these questions. It confirmed that, although the two-year period indicated in the 3M technical bulletin had been exceeded, this did not affect the performance of the seal used (29). The two-year period was not intended to be a definitive guide to the shelf life of the product, but was rather intended solely to limit the warranty guaranteed by 3M as regards further processing and printing. 3M also confirmed that properly stored seals remained fully functional even beyond the two-year period (30) and in particular dismissed the possibility that the seal could be affected by exposure to a cleaning agent (31). In more than 20 years of production, there had never been complaints in relation to false positives. Other explanations for false positives than a breach of seal were out of the question.

27. On 28 March 2007, EE sent the Commission a expert opinion drawn up by [the Institute*] on "investigations into the reaction of sealing films to shearing and peeling forces" ("[Institute*] Expert Opinion I"). Pulling-shearing stresses are defined as "forces acting tangentially to the seal surface", such as may be produced by movements to the locked door. Compression-shearing and peeling forces acting on the seal, by contrast, may be caused "by movements in the door leaf of the locked door in the direction of the door opening" (32). The expert opinion was based on tests on test pieces of 10 mm to 40 mm (33) of the (unprinted) label stock 7866 available commercially, but not on the original seal used by the Commission, which are printed with a special image (34) and have a size of 90 x 60 mm, hence 13 times bigger than the seal used by the [Institute]*.

28. The experts of EE came to the following conclusion: "repeated and rapid shearing stresses on the properly affixed sealing film result in creeping of the film. Alongside the film, traces of adhesive become visible in proportion to the path of the shearing stress. ... A cyclical compression-shearing and peeling load on the properly affixed sealing film results in visible defects in the sealing film affixed to the substrate (fragments of letters in the film) and in visible adhesive parts alongside the film in proportion to the path of the compression shearing load" (35).

29. This expert opinion was followed by a second [Institute]* expert opinion (submitted by EE on 6 June 2007) on "investigations into the reaction of sealing films to pulling-shearing stresses, compression-shearing stresses and peeling stresses following the use of Synto" ("[Institute]* expert opinion II"). This expert opinion was once again not based on the original seals used by the Commission, but on 10 mm to 40 mm test pieces of the commercially available label stock.

30. The experts came to the following conclusion: "After use of the cleaning product Synto and after a dwell time of about one hour, when the sealing film is peeled off, the formation of threads is apparent together with incomplete detachment of the seal letters. ... After use of Synto, following rapid pulling-shearing movements, fragments of the VOID lettering are apparent on the remains of the adhesive alongside the film. ... After use of Synto, following slow pulling-shearing movements, gaps in the security film are apparent between the aluminium extruded profile and the door leaf. ... With cyclical compression-shearing load on the sealing film that has suffered prior damage through Synto, alongside the sealing film on the extruded profile, remains of adhesive are visible as a result of a 2 mm displacement in the film and gaps are apparent in the sealing film (VOID markings) between the aluminium extruded profile and the door leaf, and VOID markings are apparent in the film affixed to the door leaf " (36).

31. On 11 April 2007, the Commission asked Dr Georg Krüger, a sworn expert in adhesives technology and the material behaviour of plastics, to draw up an expert opinion on certain aspects of the performance and handling of the seals ("Krüger expert opinion I"). The precise scope of the questions underlying this expert opinion was notified to EE by the Commission by letter dated 23 April 2007 (37). The questions were as follows:

1. What are the effects on the performance of the seals if seals are used 3.5 years after production of the label stock material and 2.5 years after production of the seals as final products and if, in the 3M technical bulletin for the label stock material, a two-year shelf life and a one-year guarantee period, each as from production of the label stock material, are specified? In particular it should be ascertained whether, as a result of the shelf life, a "false positive" can occur, i.e. VOID markings become visible on the entire upper surface of the seal without the seal having been removed?

2. It should also be ascertained whether a "false positive" (definition, see above) can occur as a result of the circumstances listed below under (a) to (e) (individually or in combination with one another, including the shelf life referred to under 1.).

(a) Use of the seal on a door made of painted sound-insulating door leaves and a door frame made of anodised aluminium, though these materials are not expressly mentioned in the 3M technical bulletin.

(b) No prior cleaning of the substrate on which the seal is to be affixed, although this is indicated in the 3M technical bulletin (please ensure, in so far as relevant, that the substrate is properly cleaned and is not contaminated).

(c) Rapid removal of the seal from its backing film or removal at a steep angle.

(d) Effect of the cleaning agent "Synto" made by the Swiss manufacturer Pramol Chemie AG on a cloth used by the cleaning staff.

(e) Play of some 2 mm between door and door frame or vibrations in the office wall in which the door is set.

Please bear in mind in the investigation that the seal is affixed in a normal office building, immediately after the affixing of the seal no VOID markings are evident on the upper surface of the seal and only some 14 hours later are VOID markings ascertained on the entire upper surface of the seal.

3. Is it to be expected that the seal can be displaced on the substrate as a result of the use of the cleaning product referred to under (d) or as a result of the influences described under (e), and that as a result traces of adhesive may become visible along the edge of the seal on the door and door frame or that traces of adhesive may form on the underside of the seal?

32. On 26 April 2007, the expert carried out appropriate tests in EE's office building in Munich in the presence of EE representatives and of two external experts acting on behalf of EE. The tests were carried out on the door to which the seal had been affixed on 29 May 2006. In addition, the expert carried out further tests in his laboratory. In his expert opinion of 8 May 2007, the expert came to a series of findings and conclusions which are summed up as follows:

1. Proper functioning of the seal:

"The shelf life of 3.5 years does not affect the proper functioning. (...) A false positive (...) due to the 3.5 year shelf life (...) can be ruled out."

2. Circumstances that could possibly have an impact on the proper functioning of the seal:

a) "The perceived state of pre-treatment (of the door and door frame) can be excluded as a reason for a false positive."

b) "Therefore it does not make a difference concerning the proper functioning of the seal if the surfaces (of the door and the door frame) had been cleaned with Synto or not. A false positive cannot be caused by the cleaning with Synto."

c) "It does not make a difference whether the EU-seals are removed rapidly or slowly from their backing films. A false positive cannot be caused by such a handling."

d) Furthermore, no VOID-signs were visible after the "cleaning" of the seals with Synto, i.e. a false positive cannot be caused by an external treatment with Synto."

e) Mechanical loads like those under the conditions in EE's premises can be ruled out for a false positive."

3. Rest of the adhesive at the edge of the seal and the under surface of the seal respectively:

"Neither in EE's premises nor in the laboratory could it be perceived that - considering 20mm amplitude of adhesion - the affixed seals were displaced. Therefore a displacement of the seal causing the visible rests of the adhesive can be ruled out."

These conclusions are described in detail in the legal assessment below. As to the details, it is referred to the expert opinion.

33. By letter dated 6 June 2007, the Commission informed EE of the additional facts ascertained since the statement of objections on the basis of 3M's comments and the Krüger expert opinion I and called on EE to respond to them in writing. The Commission also informed EE that the finding, already included in the statement of objections, that the seals were fully functional and that no false positives occur in their use was confirmed by the expert opinion and by 3M's comments and granted EE access to the respective files.

34. On 6 July 2007, EE sent the Commission a written reply and requested a further oral hearing "on the facts and (supposed) evidence newly adduced by DG Competition, pursuant to Article 12 of Regulation No 773-2004" on the grounds that this was the only way of ensuring that the right to be heard pursuant to Article 27 of Regulation No 1-2003 was fully ensured. EE did not claim that the Commission's letter of 6 June 2007 contained new objections. The Commission took the view that a further hearing was neither legally called for nor appropriate and refused the request by letter from the competent director dated 12 July 2007. In a further letter dated 18 July 2007, EE applied to the Hearing Officer, who also refused the request for a further hearing by letter dated 26 July 2007.

35. By letter dated 1 October 2007 - almost three month after the deadline fixed by the Commission for written responses to its letter dated 6 June 2007 - EE sent the Commission a further written study prepared by the Natural Science and Medical Institute of the University of Tübingen ("[Institute]* study III") on "investigations into the reaction of sealing films to peeling stresses following aging, the use of Synto and humidity". In this study, the expert examined the adhesive power and the sensitivity to the appearance of VOID-letters of artificially aged sealing films under influence of Synto and humidity. The experts came to the "possible conclusion": (38) "A sealing film which is stored beyond its shelf life and therefore suffered prior damage shows a significant reduction of its adhesive power and a significantly increased sensitivity to the appearance of VOID-markings. (...) Humidity also significantly influences the response characteristic of the sealing film. The sensitivity to the appearance of VOID-markings as a function of humidity further increases if the sealing film has suffered prior damage due to artificial aging and the additional influence of the cleaning product Synto." This expert opinion was once again not based on the original seals used by the Commission, but on 10 mm to 40 mm test pieces of the commercially available label stock.

36. The Commission scrutinized the presented arguments again. The Commission sent another request for information to EE which was answered by EE on 25 October 2007. The commission also contacted the leader of the inspection team from 29/30 October 2007 to clarify details of the transport of the seals from Brussels to Munich. On 6 November 2007, the Commission received the respective statement. Furthermore, the Commission authorized M. Krüger to comment on the arguments and remarks contained in the letter of EE from 6 July 2007 as well as in the [institute]* expert opinion I and II. Hereunto, the necessary documents were sent to M. Krüger. The questions M. Krüger was put in charge to answer were the following:

1. Please comment on methods, analysis and conclusions of the [institute]* expert opinions sent to the Commission with letter from 6 June 2007 and 1 October 2007. In your statement, explain in particular the reasons why the named [institute]* expert opinions do not challenge your own expert opinion from 8 May 2007 on functionality of Commission seals with regard to methods, analysis and conclusions. If further tests should be deemed necessary to confirm/back up your previous assertion, explain them briefly.

2. Please respond to the afore mentioned question also in view of arguments/factors which E.ON - beyond the [institute]* expert opinions - asserted in its writ of 6 July 2007 (e.g. missing statistical relevance of your tests).

3. Please confirm that the combination of the factors/arguments brought forward by E.ON (respectively [institute]*) (amongst others, missing preliminary purification of the background, the use of the cleaning agent "Synto" on the seal, shaking/vibrations of the door, humidity, alleged "ageing" of the seals) cannot have led to a false positive of the seals without having the seal removed from the surface. Please confirm as well that the combination of the factors/arguments brought forward is not able to explain the other circumstances witnessed by the Commission on the morning of 30 May 2005 (traces of adhesive around the edge of the seal and on the underside of the seal. If further tests should be deemed necessary to confirm/back up these statements, explain them briefly.

37. Insofar as EE claims in its letter of 10 December 2007 that the questions were formulated as "leading questions" (39), this cannot cast doubts as to the credibility of the study and the neutrality of the expert. The purpose of the second study of Mr Krüger was indeed in the first line to verify whether the conclusions of his first study were put into questions by the [institute]*-study III or not. The exact formulation of the questions related to previous conversations with Mr Krüger over the phone and to his immediate oral comments. It may also be noted that Mr Krüger is an experienced expert who knew exactly that it was the Commission's task to obtain a neutral opinion on the results of the [institute]*. The conclusion that Mr Krüger has understood the questions as neutral is further confirmed by Mr Krüger's own re-formulation of the questions in his written opinion ("Krüger expert opinion II"). The questions are without exception formulated neutrally (Question 1: "Stellen die [institute]*-Ergebnisberichte (...) die Schlussfolgerungen des letzten Gutachtens in Frage?"; Question 2: "Stellen die von Freshfields (...) vorgebrachten Argumente die Schlussfolgerungen des letzten Gutachtens in Frage?"; Question 3: "Kann ggf. eine Kombination der vorgebrachten Faktoren/Argumente zu einem positiven Fehlverhalten führen und die Umstände, unter denen das Siegel aufgefunden wurde, erklären?"). (40) Furthermore, EE did not express any doubts as to the credibility of the Commission's expert.

38. In his expert opinion of 20 November 2007 M. Krüger arrived, amongst others, at the following statements and conclusions:

1. As to the weaknesses of the [institute]* expert opinions I and II:

a) The test arrangements chosen in the reports do not comply with the conditions of 29 and 30 May and are therefore unable to challenge the statement of the main expert opinion. (41)

b) The samples chosen by [institute]* are in a ratio of 1 to 13 to the surface of the seal actually adhered. This high deviation leads to non-proportional different test results. (42)

c) The EU-seals could be stretched by 0,03mm in maximum in the gap itself. The resulting shearing and peeling forces are so small that they cannot cause a false positive. (43)

2. Own test results to compare samples treated to "Synto" to untreated samples:

a) In a peel test with untreated strip of the same size there is no difference regarding the peeling forces and regarding the transfer of the VOID-marking (functionality). The test underlines that even in case of small seal-strips, the influence of Synto does not influence the adhesive forces over the edge areas. (44)

b) An effect of Synto on creeping in case of a cleaning of the EUseals with Synto in the morning of 30 May 2006 is to be ruled out from a physical point of view. All own tests after influence of Synto show that the adhesion, also in edge areas, does not change. In case of shear stress there is no false positive. (45)

3. Own test results on mechanic stress and extreme stress: The creeping of EU-seals, as asserted in the [institute]* expert opinion for samples of 4 cm/1 cm, requires such a force in case of seals of 6 cm/9 cm size that the film would be stretched. An irreversible stretching of the sealing film in the area of the gap takes place. It is only at the transition to the sheet, and only there, where VOID-markings become visible because of the overexpansion until above the yield stress. (46)

4. Own test results on the influence of humidity: Even under this extreme humidity stress, there are clearly visible VOID markings (47) (...); high humidity in the rooms of E.ON AG does not change the functionality. (48)

5. As to potential vibration as reason of a shift/displacement of seal: Due to the huge adherent surface, the adhesive forces exceed the potential shearing- and peeling forces resulting from the small relative moves of door and door frame. Even in case of heavy vibrations the maximum movement between frame and door can only amount to 0,53 mm respectively the sealing film in the gap can be stretched by 0,03 mm. (49)

6. As to the combination of different factors: A combination of "pre-existing damage" because of the age, effect of Synto, humidity and mechanic forces as a result of vibrations and door shaking respectively does not cause VOID-markings to show up.

7. Conclusion:

The statements of the main expert opinion (50) are not challenged by the [institute]* expert opinions II and III. (51)

39. These statements are described in detail in the legal assessment below. As to the details, it is referred to the expert opinion.

40. With letter of 23 November, the Commission informed EE about the facts additionally identified since the letter of the Commission of 6 June 2007. These facts basically confirm that there was no doubt about the functionality of the seal. At the same time the Commission granted EE access to the respective files (Krüger expert opinion II and statement of the leader of the inspection team, M. [...]).

41. On 10 December 2007, EE sent another letter to the Commission in which EE commented on the documents sent to EE on 23 November. Added to this comment was a letter from the [institute]* ("[institute]*-letter"), which contained further comments on the Krüger expert opinion II and which criticised the opinion (52).

42. On 15 January 2008, the Commission received another submission from EE. To this submission, affidavits from those 20 persons were attached who declared that they were in possession of a key to room G.505 at the evening of 29/30 May 2006. Each of the 20 persons declares in these affidavits that they had either not been in EE's premises between 29 May 2006, 19:00h and 30 May 2006, 9:30h, or "had not opened the door during this period".

II. Legal assessment

43. Pursuant to Article 23(1)(e) of Regulation No 1-2003, the Commission may by decision impose on undertakings fines not exceeding 1% of their total turnover in the preceding business year where, intentionally or negligently, seals affixed in accordance with Article 20(2)(d) have been broken.

44. Insofar as EE is of the opinion that it is in principle for the Commission to deliver the facts necessary to establish a breach of the seal, the Commission agrees with EE that - independently of the question whether the in dubio pro reo principle is applicable in the framework of law of administrative sanctions - it is the Commission's duty to submit the facts to prove the breach of the seals (53). In this context, it is important to note that the very fact that the seal on 30 May 2006 showed all elements which are typical for a breach of a seal can in itself be considered as sufficient evidence for a breach of the seal. Nevertheless, the Commission has carried out further investigations (inter alia by two expert opinions) in order to exclude that in this specific case the seal did - in an atypical manner - not work, as claimed by EE. In cases of a breach of a seal the Commission has to rely on documentary evidence to prove the infringement (such as changed state of the seal, minutes of the breach, witness statements on the damaged seal, photos). It cannot be sufficient simply to claim that alternative explanations are possible, if the Commission has established the infringement with documentary evidence (54). EE would have to submit evidence that the Commission's evidence it not sufficient to prove the breach of the seal and to submit further convincing evidence to show that an - atypical - alternative explanation for the facts was possible (55). The mere claim that an alternative scenario was theoretically possible cannot be regarded as sufficient in this contex. (56).

1. Seal affixed by Commission officials pursuant to Article 20(2)(d) of Regulation (EC) No 1-2003

45. Article 20(2)(d) of Regulation No 1-2003 provides that business premises may be sealed for the period and to the extent necessary for the inspection.

a) Inspection carried out by decision

46. The inspection of EE which began on 29 May 2006 at EE's offices was conducted on the basis of the Commission's decision of 24 May 2006. No appeal was made against that decision.

b) Business premises

47. The sealed room G 505 was situated in EE's offices at Brienner Straße 40 in Munich and therefore constitutes business premises.

c) Sealing by Commission officials

48. The sealing was carried out by the leader of the inspection team, a Commission official (M. [...]*). Three other Commission representatives (Ms [...]*, M. [...]*and M[...]*) and six representatives of the Bundeskartellamt (Ms [...]*, M. [...]*, M. [...]*, M. [...]*, M. [...]* and M. [...]*), all bearing their staff cards, were also present at the sealing.

49. Prior to sealing, the representatives of the Commission and of the Bundeskartellamt checked that the surface was clean, so that particular cleaning of the relevant door and door frame was not necessary. Furthermore, there were no indications of prior damage to the seal when it was removed from its protective film. The Commission representatives did not notice any damage, nor can any of EE's staff recall difficulties in the seal removal and affixing process (57). From the mere fact that the inspection team stayed in front of the respective door some minutes right after sealing can in no case be deducted that there occurred problems in sealing. On the contrary, the additional waiting period only confirmed that the seal adhered firmly to the door which was not disputed by EE.

50. The sealing thus took place in a proper manner. The seal adhered perfectly to the substrate consisting of the door and door frame and, after it had been applied (including an appropriate waiting time), did not display any VOID markings on its blue and yellow upper surface. This was verified by several members of the inspection team on the evening of 29 May 2006 (58).

51. It was also confirmed by EE's signature on the record of sealing. The latter contains no reference to any supposedly improper affixing of the seal or to possible appearance of VOID-markings. EE disputes having confirmed that the seal was properly affixed in this way (59). However, this claim lacks credibility since it can be assumed that EE would have objected to the sealing, of whose meaning EE was perfectly aware, on the evening of 29 May 2006 had it been manifestly deficient in any way.

52. It was only on the following day that EE submitted, during the drawing-up of the breach-of-seal report, that the affixing of the seal had, in its estimation,required an unusually long time (60). However, this argument is not valid. Even if the affixing of the seal were to have taken longer than necessary - and this is clearly contradicted by the perception of the inspection team (61) - this would not call into question the proper affixing of the seal. Furthermore, neither on the day of the affixing of the seal, nor on the following day or in response to the request for information, did EE claim that the seal had not been properly affixed.

53. It was not until much later, in its reply to the statement of objections, that EE submitted that insufficient adhesion or prior damage to the seal could not be ruled out, so that the proper functioning of the seal was not guaranteed. A primary reason for this was the fact that the Commission representatives had not cleaned the substrate (door and door frame) prior to sealing or checked whether there was sufficient adhesion, although these preliminary actions were specifically mentioned for the affixing of seals in the technical bulletin (62). The same could also happen if the seal was removed too quickly or at too steep an angle from its liner (63).

54. However, these arguments put forward by EE are not valid, since they contradict the observations made on 29 May 2006 by the inspection team and by the representatives present on behalf of EE. Any sign of inadequate adhesion of the seal would inevitably have been noticed by those present. It was also in EE's particular interest to observe the affixed seal closely, since by signing the record of sealing it acknowledged that it had been informed of the significance of sealing and the possible legal consequences of a breach of seal. And yet, EE had no comment incorporated in the record of sealing that pointed at problems when the seal was affixed.

55. Furthermore, the tests carried out on the spot on 26 April 2007 by the expert appointed by the Commission showed that all the affixed seals, which were produced at the same time as the original seals used on 29 May 2006, adhered firmly to the substrate, regardless of whether the substrate had been previously cleaned or not (64). EE disputes the validity of the results of the expert opinion, arguing inter alia that "the investigations lacked the statistical reliability necessary for a scientific analysis and hence for a legally viable production of evidence" (65). However, this view is out of step with reality. As EE could see for itself, all the seals used in the 2006 inspection, originating all from the same production, adhered perfectly. The expert appointed by the Commission placed eight seals on doors in EE's presence during his on-the-spot tests on 26 April 2007. He carried out stress tests on more than 50 original seals from the Commission. In no single one of these tests having been carried out according to the relevant test procedures, a positive false occurred. (66) Nor does the argument as to lack of statistical reliability hold up inasmuch as the Commission has been using these seals , originating all from the same production, since May 2004 in numerous investigations, and no doubts whatsoever have so far arisen as to their adhesion. This also tallies with the observations of the manufacturer 3M, which has been producing the seals since 1985 and which in all that time has not had a single complaint about poor performance (67). It would be in 3M's interests to make improvements to its products if there were inadequate adhesion. What is more, the appearance of VOID markings over the entire surface of the seal, as was the case on the morning of 30 May 2006, can be explained only by a breaking of the seal (68). The seal manufacturer 3M has explicitly pointed out that no VOID marking would appear if the seal did not adhere to a certain surface once. (69) Hence the question of the need for a more comprehensive statistical reliability does not arise.

56. The manufacturer M3 also put it straight that the indications given by 3M in the technical bulletin are no instructions always necessary for technically correct performance. The advice on the handling of the seals (type of substrate, pre-testing and cleaning of the substrate, how to remove the seal from its liner, waiting time or "dwell time") are merely recommendations intended to prevent inadequate adhesion of the seal (leading to the consequence of detachment of the seal without appearance of the VOID marking). These recommendations are primarily of relevance to use of the labels under extreme conditions, for example where the relevant surface is contaminated by oil or grease. However, dust normally found in an office would not be expected to affect the performance of the labels (70). According to the seal manufacturer, pre-testing is necessary only where the product is used on certain low-adhesion surfaces such as silicon or Teflon. However, where the seals are used on ordinary office doors made of (painted) aluminium, they can be expected to function properly. This was unequivocally confirmed by the tests carried out by the Commission-appointed expert, both in situ and in the laboratory, on the sealing products and their adhesion behaviour on the substrate actually present. This follows in particular from the finding that, owing to the technical properties of the seals, VOID markings can appear only if the seal first adhered perfectly and subsequently was removed from the substrate (71).

57. The technical bulletin does not therefore, as EE alleges (72), contain any indications of difficulties of application, which were of any relevance in the present case, but is rather intended to draw the user's attention to particularities in the use of the seals on various, in particular "problematic", surfaces, and to prevent a detachment of the seal in those cases.

58. Thus, it must not be assumed that the fact that there was no prior cleaning or pre-testing of the surface can be regarded as unduly sealing that can result in a false positive in the seal affixed to the surface (i.e. the appearance of VOID markings without breach of the seal). The missing cleaning and testing cannot explain the condition in which the seal was found on 30 May 2006. As mentioned, it is at most possible that failure to carry out prior cleaning might have resulted in a "false negative" (i.e. detachment of the seal without appearance of VOID markings) (73). However, such circumstances are not relevant here. This has been confirmed by M. Krüger's tests to the full extent as well.

59. The manufacturer's indication on the waiting time between affixing the seal and its proper adhesion is similarly only a guideline, and the only risk posed by not observing it would be at most a possible false negative. (74)

60. Prior damage to the seal, postulated by EE as a possible explanation for the appearance of the VOID markings, would, according to the manufacturer, be possible at best if the seal were subjected to extreme load conditions or badly mishandled (75). However, the leader of the inspection team did not remove the seal in an absolutely atypical way, i.e. too quickly or at too steep an angle from its liner when affixing it, nor does EE claim that he did. Moreover, in the light of the further investigations and tests carried out by M. Krüger (76), the possibility can be ruled out that even removing the seal too quickly or at too steep an angle from its liner would lead to an appearance of the VOID markings. Furthermore, the VOID markings would have appeared immediately under such a mistreatment and not only on the next day. (77)

61. Moreover, any mishandling of the sealing film like touching it with a finger, when affixing it would at most produce partial damage, whereas the VOID message was visible over the entire surface of the seal on the morning of 30 May 2006 (78). Consequently, there can be no countenancing EE's claim that prior damage to the seal could have been caused by the leader of the inspection team touching the adhesive surface of the seal with his fingers when detaching the seal from its backing - something which could have occurred without his knowing (79). In such an event, the VOID markings appear immediately and only partially, i.e. at the points where the adhesive was touched. If the seal is handled properly, as it was in this case, such damage is possible at most within a 1-2 mm perimeter around the edge. EE could have had such partial prior damage noted in the record of sealing. However, it did not do so.

62. EE also argues that the seal used was not fully functional on the grounds that the two-year shelf life specified in the manufacturer's technical bulletin had been exceeded, since the seals were used 3.5 years after the stock material had been produced and 2.5 years after the final sealing products had been produced, despite the fact that a two-year shelf life and a one-year guarantee period, as from the production of the label stock material, were specified for the product in 3M's technical bulletin (80).

63. However, the investigations showed that time limits are no evidence for the technical performance of the seals. In fact it has been shown that all of the more than 50 original seals, which were tested by M. Krüger and originate from the same supply, i.e. with the same production date and characteristics as those used in the May 2006 inspection, functioned perfectly despite their shelf life, which is not disputed. Even in November 2007, i.e. nearly five years after production date, the performance of the seals was undisputed. (81) Consequently, the fact that the shelf life specified in 3M's technical bulletin was exceeded does not have any material significance as regards the performance of the seal. According to the manufacturer, the two-year limit on the shelf life of label stock 7866, used for the Commission's seals, specified in 3M's technical bulletin applies only to converters who process the labels, but not to final users of the printed seals. (82) The two-year limit is set primarily for commercial reasons and should not be understood to mean that the technical performance of the seals cannot be relied upon once their shelf life has expired. In 3M's experience, the performance of the seals can always be relied upon beyond the two-year period. In particular, a number of 3M's large business customers (e.g. converters) give a shelf life for their converted product of up to ten years to their own customers (the end users) (83). The fact that VOID markings appeared extensively on the seal surface is also to be regarded as evidence of the functional capability.

64. EE's claim that this conflicts with what is known on the "general" age-related deterioration in the performance of pressure sensitive adhesives is not valid (84). The quotation from the Handbook of Pressure Sensitive Adhesive Technology only describes the potential interaction between the backing and the adhesive during the ageing process but does not explain to what extent actual impairments of performance can occur and how these become obvious having regard to a certain time lapse. Furthermore, in making these comments EE is disputing the production and technical know-how of 3M, a well known manufacturer of those products, which has decades of experience in dealing with the relevant adhesives. This argument is not substantiated any further and is thus not credible.

65. Furthermore, EE cannot argue that it was unable to investigate the effects of age-related deterioration because the Commission did not provide a sample of the original seals (85). The Commission was in principle willing to provide original seals to EE on condition that Commission staff were allowed to be present at the tests (86). However, EE did not avail itself of this opportunity. The Commission had to insist on this condition since otherwise tampering with the sample and the danger of reproduction of official seals could not be ruled out. Moreover, the Commission appointed expert carried out his tests on the spot both on the original seals and in the presence of representatives of EE and of the experts and lawyers appointed by EE (87).

66. -The arguments brought forward by EE relating to a potential influence of the ageing process on the performance of the seals are not convincing as well. The conclusion drawn from the series of tests of the [institute]* expert opinion III is only that in peeling-tests following an artificially aging of the sealing film corresponding to a shelf period of three years, the adhesive power is reduced up to 50% of the original value and that the response sensitivity to the appearance of the VOID-markings raises. (88) However, it is not asserted or satisfactorily shown that the performance of the seals could be affected by ageing in such a way that it could have led to the concretely asserted displacement of the seal of 2 mm or to the appearance of VOID markings on the whole seal surface. It is undisputed that such damage symptoms cannot be explained by the mere reduction of adhesive force, especially because the expert opinion itself claims that "no correlation can be established between adhesive power and the sensitivity to the appearance of the VOID-signature. (89) Besides, the [institute]* expert opinion III is also unable to explain the other factors that indicated a breach of seal in the morning of 30 May 2006 (e.g. the traces of adhesive on the back side of the seal (90)). It has to be kept in mind that EE alleges the seal had not been removed from its background at any time and that its condition was to due to other reasons, like the "reduction of adhesive force".

67. Furthermore it has to be pointed out that the test conditions of the [institute]* expert opinion III did not - like the other [institute]* expert opinions - correspond to the technical circumstances as observed on the spot (condition of the seal, size of the seal, distance between door and door frame, application of the cleaning product etc.). This is especially true for the "artificially aged" seal material also used by the [institute]* which cannot serve as a substitute for the original material. The Krüger expert opinion II thus proves that the ageing method chosen by [institute]* does not bear up under scientific standards already because the ageing adhesive polymers of the Commission seals are covered by a protection film. The "ageing formula" chosen by [institute]* holds only true to non protected polymers. (91) In this context EE had previously acknowledged that it was unable to investigate the effects of age-related deterioration because they did not have the original seals. (92) Besides, the diagrams of the [institute]* expert opinions I and II lack periodically recurring "power peaks" like they can be determined in the expert opinions of M. Krüger. The absence of these peaks, which are caused by tearing out the VOID markings, also shows that the testing conditions chosen by [institute]* were not representative. (93) However, the Commission expert has tested the functionality of the original seals on the original door in EE's premises in the presence of EE and its experts and he did not establish doubts as to the functionality.

68. At all events, the possibility that the supposed "storage" of the seals beyond their shelf life would lead to false positives is excluded. At most, exceeding the shelf life might result in a false negative (detachment without VOID marking) or in VOID markings immediately after removing from the protection film. The fact that all seals used in the tests adhered correctly to the substrate is, however, evidence of the fact that they were functioning normally, so that the VOID markings can only appear if the seal is removed (94). EE's tentative explanation that insidious prior damage had occurred (95) is implausible.

69. It should be pointed out, moreover, that during the course of the inspection on 29 May 2006 three further seals from the same batch were affixed on EE's business premises and on the morning of 30 May 2006 no VOID markings were apparent (96).

The same holds true for the seals affixed that same day on the premises of other E.ON AG group companies. Also the seal that was affixed on the evening of 30 May 2007 stuck at the door the next morning, without showing any unusual signs (97).

70. On the basis of the above, there is no doubt as to the fact that the seal was duly affixed on 29 June 2006 and that it was a duly sealing according to Article 20 para. 2 lit. d) of Regulation (EC) 1-2003.

(d) Necessity of the type and extent of the sealing

71. The sealing of the office was carried out in order to ensure that the documents which had been pre-sorted, but not yet copied and catalogued, on the first day of the inspection were kept in a central location to which EE did not have access. It was intended to allow the inspection team to copy the pre-sorted documents the next day and to completely catalogue them in the lists which had been provided for. It was to be ensured that the documents were complete and intact the following morning. Even if - as EE asserts - parts of the pre-sorted documents had already been catalogued, this would not have made it possible to check that all the documents were still there.

72. It was in particular necessary to seal the room because simply locking it without a seal would not have provided sufficient assurance that the room would not be entered during the night by EE to examine the documents stored there. Although the leader of the inspection team had a key to the room during the night of 29 to 30 May 2006, EE and the security firm [...]* employed by EE had a further 20 keys in circulation with which room G 505 could have been opened (98).

73. The sealed room was a conference room made available to the inspection team by EE for the purposes of the inspection, and sealing it did not disrupt the normal business of the company. The seal was to remain in place for only one night, since the inspection was to be continued the next morning. The seal was therefore affixed for the period and to the extent necessary.

2. Breach of the seal

74. The condition of the seal on the morning of 30 May 2006 points clearly to the conclusion that the seal was removed from the office door during the night, thus allowing the office door to be opened during that period. This conclusion follows from the nature and physical properties of the seals, which as security-relevant indication film were produced by the manufacturer in a way that an extensive appearance of VOID markings, parts of glue alongside and on the back of the seal cannot be explained by any other reason than by the removal of the seal (99).

75. The VOID marking was not only visible in area of the edges (approximately few millimetres next to gap), but over the entire surface of the seal when the inspectors entered the building at 8:45 on the morning of 30 May 2006 (100). This occurs only if the seal is removed, since the message appears only if the adhesive layer on the back of the seal is damaged. Moreover, the traces of adhesive on the back of the seal alongside the transparent parts and the edges of the seal can be explained only by the fact that the seal was not reaffixed precisely in the original position. The argument put forward by EE that when the seal was removed the VOID markings on the door and door frame were wholly intact (101), is therefore irrelevant, because a "blurring" is unlikely in case of removal and re-affixing - in such a case, slightly displaced traces of glue on and alongside the seal are more likely, as in the present case. It is in this respect that the hint of EE to be found in the letter of 10 December 2007 is wrong that the commission had not shown how it could have come to traces of glue alongside the seal without a "creeping" of seal. (102) Already in its submission of statement of objections, the commission argued that the determined traces of glue can only result from a removal of the seal in combination with re-affixing in not exactly the same place as the original one. (103) The respective conclusion of the seal manufacturer 3M according to which a displacement of the seal clearly points towards removal and subsequent re-affixing of the seal, has already been made available to EE. (104)

76. The unambiguously documented damaged condition of the seal cannot be put into question by EE's assertion that the Commission failed to take evidence-protection measures in respect of the condition of the seal on the morning of 30 May 2006 (105) is incorrect. When the inspection team entered the building at 8:45 on the morning of 30 May 2006 it saw the VOID markings over the entire surface of the seal (106). The condition of the seal was ascertained by being inspected by eight people (107). Several team members double-checked that the seal on room G 505 was actually broken by comparing it with the other, undamaged seals affixed the previous day. This finding was conclusive as the seal in question, in contrast to the other seals, displayed VOID messages, namely over its entire surface, and traces of glue around the edges. The fact that some inspectors compared the damaged seal to an intact one first to get a better idea of the deviation of the seal in question from its normal condition does not - contrary to the view of EE - put in question the has been removed and replaced slightly unaligned. This would give rise to the showing of glue parts at the back of the seal.".. presence of the characteristics of damage laid down in the report. As this was the first case of breach of seal and as it was a modern seal which does not display a breach of seal by rupture, it seems understandable that the inspectors wanted to reassure by comparison with the other seals. The leader of the inspection team and inspectors from the Bundeskartellamt have, in the presence of EE's representatives, drawn up an protocol on the breach of the seals. The condition of the seal described therein, in particular the extensive appearance of VOID-markings, was unanimously confirmed by questioning the inspection team. (108) The photographs taken at the place of the breach of seal at a later time support the statements laid down in the report as well. They show that at least on the right part of the seal, which still adhered unchanged to the door frame, VOID markings were extensively visible and that there were traces of glue alongside the seal. (109) It is already the report as such that constitutes sufficient documentation on the state of the seal as it was found on the morning of 30 May 2006. Hence the evidence was immediately protected in the presence and with the involvement of EE. If the representatives of EE, who were able to get an idea of the condition of the seal right there, had wanted to challenge several statements of the report, it would have been to be expected that they would have done so in their supplementary declaration to the breach-of-seal report (e.g. by clarifying that die VOID markings did not appear in an extensive way). However, this did not happen. EE did not dispute the factual findings of the inspection team (including the extensive appearance of VOID markings) in the breach-of-seal report. EE's comments in the given declaration only related to other factual circumstances (inter alia, no blurring of the VOID markings, long duration of the sealing process). (110) Consequently, the issue of the immediate protection of the evidence does not turn on the fact that the Commission did not take digital photographs of the damaged seal until the afternoon of 30 May 2006 in addition to the report. For these photographs, taken as supplementary measure for conservation of evidence on own initiative of the Commission team, are just one of several instruments of proof available. Coming as they do on top of the statements of the inspection team and the breach-of-seal report, they simply constitute a further documenting of the taking of evidence.

(a) No significant influence of vibrations

77. EE has argued in different writs that vibrations of the closed door and the door frame respectively could be the cause for the damage found and has referred to different tests of the [institute]*. The Commission, having carefully scrutinized whether vibrations could have led to false positives, rules this out as a cause.

78. EE argues in particular that a positive false could have resulted from intensive shaking of the door or from vibrations, e.g. from frequent opening and closing of doors close-by. However, the tests commissioned by the Commission and die questioning of the seal manufacturer proved without any doubt that a positive false resulting from vibrations was out of question in the present case.

79. Even a partial "loosening" of the seal - as claimed by EE (111) - seems conceivable only if there had been significant play in the locked door and if it had been shaken violently. However, this does not correspond to the actual circumstances on the spot. The measurements at the closed door of the room G 505 showed that the door could only move back and forth about 0,53 mm. Under these conditions, i.e. in case of a crack between door and frame of 4 mm and a "play" of the door of only 0,53 mm, the Commission seal could not stretch for more than 0,03 mm (112), which literally is just a hair's breadth. Already a layperson knows that the forces impacting on the seal in such a system must be rather small. This was confirmed by the tests and measurements carried out by M. Krüger. As fat as it is alleged in the [institute]*-letter of 10 December 2007 that the door allowed a minimum displacement of approx. 2 mm "according to [institute]*-observations", this objection cannot put in question that even then the play of the closed door would still be small and that it could at best lead to a minimum stretching. [institute]* omits giving evidence for the assertion the play would have been 2 mm (rather than 0,53 mm, like according to M. Krüger's measurements documented on video). Furthermore, EE has never put into question the Commission's observation that the door could only move by 0,53 mm. This is despite EE was informed of this observation already on 6 June 2007 when the first study of Mr Krüger was sent to EE and despite the fact that EE was even present when Mr Krüger carried out its tests and measured the door, as taped on video. Even under the presumption that the play of the door was wider and actually of 2 mm, the stretching would remain very small in such a system (0,47 mm (113)). Even to a layperson it must be obvious that such a "micro-stretching" is unable to explain the extensive appearance of VOID-markings and the displacement of the seal of 2 mm.

80. The tests carried out by [institute]* which seem to suggest a "creeping" of seals (114) do not come up to the real circumstances on the spot. The [institute]* itself concedes that different seal sizes can have an impact in the results of the tests and the conclusions for a seal of a certain size can only be relevant for a seal of the same size (115). Apart from the fact that the Commission seals are 13 times as big as the ones tested by [institute]* and apart from other deficiencies and ambiguities of the [institute]* expert opinions (116) (deviations from real circumstances on the spot), the expert assigned by the Commission proved that the forces impacting on the seal in maximum were much smaller than the ones according to the [institute]* expert opinions. (117)

81. Furthermore, all practical tests on the spot and in the laboratory confirmed that the damage from 30 May 2007 cannot result from vibrations of the door. (118) A detachment or even displacement of the seal proved not to be possible when a number of attempts were made to do so on the spot in the presence of company representatives. Even in case of extreme stress applied to the seal, VOID signs appeared only in the area right at the crack in the door where the seal had detached itself from the surface because of the small game (approx. 1 mm around the crack in the door), but not on the entire surface of the seal. (119) Other than put forward by the EE, a positive false cannot occur even under "extreme conditions". (120) The seal manufacturer, 3M, also confirmed that in case of extreme shaking VOID markings can appear at best alongside a "vertical" stripe along the crack in the door (i.e. where the seal detaches itself from the door slightly because of displacement). In contrast, 3M rules out that a damage such as the present one could have resulted from shaking. (121)

82. Relating to the concrete circumstances of the incident, EE could name no person having shaken the door. (122) Even presumed there had been a shaking of the door, there would be no explanation as to the reason why the VOID markings on 30 May 2007 were visible on the entire surface of the seal and not only on the assumed loosened parts around the crack in the door - e.g. like it is clearly shown on the film presented by EE in the hearing on 6 December 2006. (123) In fact, the video submitted by EE at the occasion of the oral hearing, which was intended to show the effects of strong vibrations, shows that even after strong vibrations, the seal was detached from the door/the frame only within a very small area. The vibrations did not result in the appearance of extensive VOID-signs over the whole seal. Neither did the vibrations result in glue parts on the back of the seal, nor in a displacement of the seal.

83. Even intensive shaking of the locked door thus produces no visible effects on the seal. The same is also true of vibrations in the office wall caused by intensive movement in the neighbouring doors (124). In particular, such external influences do not produce any false positives in the seal (125). Also the other seals in the other rooms in EE's premises were fully intact (126).

(b) No significant influence of the cleaning product

84. Nor is EE's supposition that the seal might have been displaced when a cleaning woman wiped it with a damp cloth (127) convincing since the seal was affixed firmly to the door and door frame and cannot be displaced by mere wiping. Moreover, EE could not rely on this supposition inasmuch as responsibility for informing the cleaning company about the significance and handling of the seal and hence for avoiding such actions lay with EE itself (128). Furthermore, EE's testimony is inconsistent. In answer to the question as to how EE can explain the breach of seal (129), EE referred to the statement by the cleaning staff to the effect that it could not be ruled out that the seal might have been displaced when it was wiped. On the other hand, EE distanced itself from this statement when it argues later that the seal had not been displaced by being wiped (130).

85. A chemical reaction with the cleaning product used, which could have led to a positive false, can also be ruled out (131). The external effect of the cleaning product "Synto" on the surface of the seal does not have any impact on the functioning of the seal as proved by M. Krüger in comprehensive test series (132). Only the colour-printed surface of the seal is affected, with small amounts of colour coming off and remaining on the cloth. However, no false positive reaction in the seal is apparent after contact with "Synto" (133). The assertion that M. Krüger had not used the original cleaning product in his tests, since the [institute]* had tested the product and observed that the product was a water-free solvent with 2- (2ButoxyEthoxy)ethanol as its main component, is also unfounded. In this respect, the Commission stresses that it had obtained the cleaning agent "Synto" directly from the very same service company which had cleaned the rooms of EE in May 2006. Only this cleaning agent was used in the different test series. (134) It was not necessary to order more bottles of Synto, as offered by EE, since sufficient Synto was left for the tests. The Commission does not know which cleaning agent the [institute]* has used and tested, because the only cleaning agent sold under the name "Synto" ("Synto Forte") which is more or less free of water is based on glycol alcohol and is not sold in bottles of 1 litre. It is also not used as a general cleaning agent, but as a special solvent to treat particularly dirty spots (135).

86. Contrary to what EE contends (136), a false positive did not occur even after a lengthy waiting period following external impact of the cleaning product on the sealing film. (137)

87. By contrast, the claim by the [institute]* experts (138) that some "creep" in the sealing film may be produced by compression and shearing stresses combined with the effect of the cleaning product "Synto" is unconvincing. Apart from the fact that the respective "interpretation" to be found in [institute]* expert opinion II did not form the subject matter of the accredited test, (139) the expert report is not able to challenge the results of the own expert opinions as the laboratory tests were not conducted under the same conditions as obtained on 29 May 2006 when the seal was affixed on EE's premises. The differences relate inter alia to the size of the seal, the application of the cleaning product and the distance between the door and the door frame. Nor does the expert opinion explain how the cleaning product Synto could have diffused throughout the adhesive on the sealing film, especially since the seal used by the Commission is substantially larger than the label stock material used by [institute]* (140) and the expert opinion itself states that the cleaning product completely evaporated "within 15-20 minutes" (141). A softening of the adhesive, as alleged by the [institute]* expert opinions, could have happened only at the edges because the sealing film is made of PET and impermeable to moisture. (142) Diffusion "through the film" therefore is not possible (143) which is conceded by [institute]* as well. (144) The tests carried out by the Commission's expert M. Krüger on Synto confirmed that under practical conditions, the seal is in principle impermeable to moisture. (145) The consequences of an impact from the edges on a seal which is only 1 cm broad (as used in the test series carried out by [institute]*) of course differ considerably from the effects on a seal which is 6 cm broad (where the cleaning product would have to get 3 cm deep under the firmly affixed seal film to be able to loosen the adhesive). All tests carried out by Mr. Krüger confirmed that an external effect of Synto does not affect the performance of the seals. (146) This is even true for small samples of 1 or 2,5 cm in breadth (9) and for extreme exposure to Synto of 2 hours. Even if one assumes in EE's favour that the conditions underlying [institute]*'s laboratory tests corresponded to the actual conditions, this could still not explain the Commission's findings (residues of glue next to the seal, extensive VOID markings, white traces of glue on the adhering part of the seal) on the morning of 30 May 2006 (147). It has also to be stressed that the effects of humidity through the edges on a seal of only 1cm (as used by the [institute]*) differ, of course, significantly from the effects on a seal of 6 cm width (on which the cleaning agent would have to move 3cm under the seal to solve the glue of the whole seal).

88. In view of these observations it can also not be said that the Commission's seals showed a "tendency" to "false positives" (148). None of the tests by the [institute]* can explain the signs that were visible on the seal on 30 May 2006 (VOID-sign over the whole seal, parts of glue next to the seal and at the back of the seal), which are typical for a broken seal. The same is true for the numerous tests of Mr Krüger, which never resulted in a "false positive" as observed on 30 May 2006.

c) No significant influence of stress in affixation

89. The VOID markings which on the morning of 30 May 2006 were wholly evident over the entire seal can neither be explained by the fact that the seal had been affixed "under stress" (149) There is no indication for the seal having been affixed in an atypical way, i.e. under stress, and that it had been under stress immediately after it had been affixed. Neither when the seal was affixed nor when EE drafted its declaration for the protocol on the evening of 30 May 2007 (150), EE has mentioned that the seal was affixed, in an untypical manner, under tension. The Commission seals are not elastic, but made of material guaranteeing tensile strength and can therefore be stretched only for a minimum at all if very high forces are applied. (151) Moreover, the expert authorized by the Commission proved that the seal, even if stretched for 4 mm, is not displaced but only slightly stretched in the area of the crack in the door if at all. Under no circumstances VOID-signs can appear on the whole seal, neither can the other signs observed on 20 May 2006 appear. The fact that forces of 660N (or 66 kilograms) were necessary to effect such a manual stretching and that the seal in such a case would have to be touched extensively by some fingers (also on the back side of the seal, causing probably immediate VOID markings in the points of contact) shows that affixing under stress can be ruled out as a cause for the damage, since the seal was fully intact on the evening of 29 May 2006. Furthermore, the company 3M, producing the seal material since 1985, does not know of any example of a positive false due to stress so far. (152)

90. Nor can any significance be attached to EE's argument that, on the morning of 30 May 2006, the sealing film did not show any "crease" such as was partly evident when the seals were removed in the tests carried out on 26 April 2007 by the expert. (153) A "crease" is not a "necessary consequence" of detaching the seal film or a necessary indication for a breach of seal. (154) Even if it may be possible that a "crease" might appear as a result of a certain manner to remove the seal, the seal is technically designed in such a way that any movement is revealed only by the appearance of VOID markings - and not by the appearance of a "crease". The fact that in case of a certain way of removing a "crease" may appear additionally to the appearance of VOID markings does not mean in reverse that the lack of such a crease rules out a breach of seal. Other methods removing the seal without a "crease" are offhand imaginable (e.g. detachment from the middle onwards by using a straight edge or other auxiliary means, soft opening of the door, etc.).

d) No significant influence of humidity and of combination of different factors

91. The possible influence on the response characteristic of the sealing film due to the existing humidity (155) - possibly in combination with other factors (i.e. Synto and age-related deterioration) - as is claimed by the experts in the [institute]* expert opinion III, is also not convincing. The assertion that humidity might have caused a false positive was mentioned for the first time in a letter by EE dated 1 October 2007, which is more that two months after the end of term for comments on the additional letter of the Commission. The Commission scrutinized the arguments brought forward nonetheless and came to the conclusion that humidity is to be ruled out as a cause for a "false positive". As already outlined (156), the tests, on which this expert opinion III is based on, were not conducted under the same conditions as observed on 29 May 2006 when the seal was affixed on EE's premises. These relate inter alia to the too small size of the seal, or the play between the door and the door frame which in effect was very small. (157) To carry out the tests on effects of humidity according to [institute]* expert opinion III, "pieces of sealing film were put into a fridge at room temperature and relative humidity-value of 70%, 60% and 50%". [institute]* does neither give a convincing explanation for the fact that only small pieces of seal were used not does it become clear whether the seals were exposed to humidity with or without protection film (158) and for what time they were exposed to humidity. If the seals were exposed to humidity without a protection film, these tests would be of little relevance. It is uncontested that humidity could not take effect right on the adhesive surface. In reality the seal was protected by a moisture repellent protection film. If the seals were put in the fridge with a protection film on them, as it is suggested in the [institute]* letter dated 10 December 2007 (159), humidity could primarily only enter from the edges. Even if small quantities of moisture may penetrate the PET of the seal as well as the protection film in cases of long-term exposition to humidity, this rather theoretical observation cannot put into question the conclusions of the Commission. The assumption that the seal was damaged significantly in its functionality by humidity that entered through the PET-film as well as through the siliconised protection film on the back side during the transport (160) seems farfetched. The company 3M indicates in the data sheet to its seal material "7866" under the point water resistance that the seals can during at least 168 hours (i.e. during a whole week) be exposed to extreme humidity of 90%, without any effect on the functioning of the seal (161). If the security seals produced by 3M were such sensitive that even in non-tropical countries such as Germany already on a warm spring day like 29 May 2005 humidity could enter the seals offhand and could suspend their performance, the performance could hardly be assured in countries having a much higher humidity. However, according the manufacturer, the seal material of the company 3M has been used worldwide - i.e. also in countries with a much higher humidity than Germany - since 1985 as security seals but no false positive has ever occurred according to the manufacturer. (162) Besides, the results of the [institute]* expert opinion III do not serve as proof of the fact that the damage found on 30 May 2006 could be due to effect of humidity.

92. Indeed, the experts of [institute]* alleged a reduction of adhesive force and a "significantly high sensitivity" after the small pieces of seal had been exposed to an increased humidity-value of 60% and 70%. These theoretical assessments are however not an explanation for the condition of the seal on the morning of 30 May 2006, when the VOID-signs were visible on the entire surface of the seal. They also do not explain the further observations on the condition of the seal made by the inspection team on the morning of 30 May 2006 (displacement of the seal, traces of glue alongside the seal and on the adhering surface. In particular, they do not explain why the other seals used by the Commission at the same day were unaltered although they were exposed to the same conditions of stocking and transport and thus to the same humidity. The Commission expert has already laid down in his first expert opinion under realistic conditions and using original seals on the original door that the original seals in principle did function properly.

93. To check the alleged - however, not further substantiated - possibility of an effect of humidity on the seal, the Commission expert carried out a series of additional tests. (163) Although those tests were mostly carried out under extreme conditions (e.g. "watering" the non affixed seals for several hours (164)), a positive false could not be proven in any case, although the effect of forces highly exceeding the ones possible in reality was tested (165). The high moisture resistance is, like the Krüger expert opinion II ascertained (166), inter alia due to the fact that the adhesive used in the Commission seals belongs to the group of the Acrylates. Within the different types of adhesives, these are the ones who are the most resistant to humidity. The theoretical possibility mentioned by EE that acrylate adhesives can be sensitive to humidity as well is irrelevant in the present case as is proven by the smooth use of the other affixed seals, by the experience the manufacturer 3M has with the seals as well as by the tests carried out by M. Krüger. Even if theoretically, a certain degree of humidity might have entered into the otherwise water-repellent seal (167), there is no indication that this humidity could account for the found damage.

94. Apart from the fact that an exposure to humidity would not be relevant to the performance of the seals at all, EE did neither state nor prove why the seal should have been exposed to humidity at all. The expert opinion does not contest the correct storage and transport of the seal as assessed by the Commission, but only hints towards the possibility that the seal could have been exposed to humidity changes during transport (allegedly in the luggage compartment of the car (168)) which is only theoretical. The Commission questioned the leader of the inspection team ([...]*) for details of transport to exclude the possibility that the seals in question were in fact exposed to a higher humidity. In his statement, dated 6 November 2007, he explains that the seal in question was at no time in a luggage compartment, but either in a drawer in an office or in a brief case. In doing so, the seals were never exposed to considerable humidity or were treated or transported in any other "atypical" way. If it all, the humidity could have entered the seal during the night after affixing - and if so, only from the edges. Such an effect, however, has not been made plausible and even if, it would not have any effect on the performance of the seal. For the same reason, the argument of EE that the night of 19 May 2006 might have been a night of an extremely high humidity aboveaverage (169), cannot cast doubts on the functionality of the seals. Already the fact that there were no VOID markings or other atypical characteristics on any other of the used seals, so that the other seals functioned properly in spite of the increased humidity shows that increased humidity in the area outside EE's premises (170) on the day of the inspection was irrelevant to the performance of the seals. Furthermore, Germany, as already mentioned, belongs to the countries having a comparably low humidity. The seal manufacturer does not know of any false positives from countries having a considerably higher average humidity than Germany as well.

95. As far as EE in its last writ of 10 December 2007 holds that the significance of M. Krüger's tests would be diminished by having used powder coated sheet plates instead of anodized aluminium in his tests, this argument is also not plausible. (171) As can be seen from the pictures that were taken of the seal, the seals were affixed on painted ground and not, as in the tests by the [institute]*, on non-painted aluminium. The results of M. Krüger's first study are largely based on tests on the original surface, with original seals. Should there indeed be differences between the results of the studies of M. Krüger and the [institute]* because of the different surfaces, the only difference that one might have expected would be that the adhesive behaviour of the coated sheet plates would be worse than the one of aluminium because of the alleged interactions with the aggregates of the lacquer. Under no circumstances one would expect that the seals used by M Krüger would stick more strongly on the (painted) surface than in the tests on unpainted surfaces by the [institute]*. Since M Krüger concluded that debonding appearances or "creeping" cannot be observed on a painted surface, these results would apply a fortiori for the non-painted surfaces used by the [institute]* (which, however, did not correspond to the real situation).

96. As far as EE finally argues that at least a combination of different factors (e.g. effect of Synto and humidity and ageing of the seal) could have caused the damage of 30 June 2006, the Commission refers to the above mentioned statements of the manufacturer 3M, its own experiences and to M. Krüger's expert opinions. Thus, an "unfortunate chain" of potential other factors cannot have caused the found damage.

97. In this connection, it has to be taken into consideration that all tests carried out by M. Krüger were done on original seals. In testing the effect of Synto or of humidity in earlier tests, there were always two parameters (potential "ageing" and effect of Synto; potential "ageing" and humidity) being considered at the same time. In his last expert opinion, M. Krüger showed that a combination of all conceivable factors ("ageing", effect of humidity and Synto (172)) cannot be the cause of damage of 30 June 2006. He ascertains: "Also the combination of humidity, synto-effect and shaking as well as vibration does not lead to a false positive". (173) This is true even in case these factors should cumulate under "extreme conditions" (dipping in water, effect of Synto for hours, stretching for more than 0,1mm). (174) Even in this respect, there are no doubts as to the functionality of Commission seal. A combination of the factors/arguments put forward by EE (inter alia, effect of Synto, vibrations, humidity, alleged covering of the seals) thus could not lead to a false positive which would explain the damage of 30 May 2006.

(e) No relevance of the affidavits submitted by EE

98. Finally, also the affidavits of the 20 persons who were in possession of a key for room G.505, which were made and submitted at a very late point in time of the procedure, cannot put into question the conclusions with respect to the broken seal. In fact, an affidavit as such is nothing more than a unilateral declaration to make certain facts credible, with a very limited weight as evidence in the current administrative procedure, all the more since they were made only one and a half year after the inspections. Even if one were to assume that the employees of [...]* and the other firms working for EE, were perfectly neutral and not to be counted to the "sphere" of EE - which does by no means appear to be certain -, the content of the declarations would still not be corroborated by facts found by the Commission, which clearly point at a breach of the seal and a possible opening of the door. It should be noted that Article 23 of Regulation (EC) 1-2003 does not require the proof that the sealed door has been opened. Not the opening of a door, but the breaking of the seal is the object of the fines provision in Article 23. Hence, the clear evidence for a breach of the seal as found in the morning of 30 May 2006 cannot be put into question by a declaration that a key holder has not opened the door. Moreover, EE did not at all provide clear evidence that the door to room G.505 had not been opened. From the affidavits of the 20 key holders one might derive at most that these persons have not themselves the door. The affidavits do not exclude that other persons might have obtained the keys from the key holders to open the door or that the door might have been opened by other means.

(f) Conclusions

99. The purpose of sealing is to prevent (the possibility of) evidence being lost during the inspection. Sealing is thus a guarantee of the effectiveness of the Commission's investigatory activities during the course of inspections. It is intended to ensure that objects or rooms remain intact and to prevent the opening of the sealed object or room by unauthorised persons, or to make it evident that such opening has occurred. The sealing of rooms is also intended to ensure that the relevant rooms do not have to be placed under permanent guard (e.g. overnight).

100. Consequently, any change in the seal which indicates that the sealed object or room has been opened or could be opened must be regarded as a breach of seal. The criterion must not be that the seal was damaged (e.g. by cutting or buckling) or destroyed (e.g. by being completely removed). The criterion is rather that, despite sealing, the undertaking obtained the possibility of gaining access. It is therefore irrelevant that a key to the room was provided to the inspection team, particularly since there were a number of other copies. Ultimately, a breach of seal within the meaning of Article 23(1)(e) of Regulation (EC) No 1-2003 does not require proof that the door was actually opened or that documents stored in the sealed room were altered or destroyed.

3. Intention or negligence

101. Leaving aside cases where forces of nature are involved, it must be assumed in principle that a seal can be removed only by a deliberate action. This is also borne out by the fact that, after it had been removed, the seal was obviously reaffixed in such a way as to conceal the fact that it had been broken. Account must also be taken of the fact that only persons authorised by EE (including the staff of [...]*, a wholly-owned EE subsidiary) were in the building.

102. At the very least, it can be concluded that the breach of the seal happened through negligence. It must be borne in mind that, when the seal was being affixed, representatives of EE were informed by the inspection team leader of the significance of the seal and the consequences of any breach of it (175). This is also indicated on the seal itself. Furthermore, EE confirmed that the condition of the seal might have been changed by a [...]* employee using a cloth and cleaning product (176).

103. It is EE's responsibility to organise its own business sphere in such a way as to ensure that the instruction that the seal should not be breached is complied with. In view of the extensive security measures taken by EE, the possibility can be ruled out that during the night in question unauthorised persons gained access to the building and altered the condition of the seal in the manner described (177). Nor did EE itself either claim or provide any evidence of unauthorised entry into the building by third parties. EE cannot therefore argue, however, that the seal was broken inadvertently, particularly since not all the persons in the buildings (for example the cleaning service) were informed by EE of the existence of the seal and the need to keep it intact (178). It should, furthermore, be noted that EE took additional security measures for the second night so as to prevent any breach of the seal.

4. Legal consequences

104. Since the conditions for a breach of seal within the meaning of Article 23(1)(e) of Regulation (EC) No 1-2003 are met, the Commission can impose on EE a fine of up to 1% of its turnover. The level of the fine is determined in particular by the seriousness of the infringement and the particular circumstances of the case.

105. Quite apart from this specific case, breaches of seals must in principle be regarded as a serious infringement. The use of seals is intended to prevent (the possibility of) evidence being lost during the inspection, thus undermining the effectiveness of the inspection. It must also be borne in mind that the fine must have a deterrent effect. It cannot pay off for an undertaking involved in an inspection to break a seal.

106. When calculating the amount of the fine which is necessary to effectively deter companies, the Commission has to take into account among others the size of the company as well as its previous practice in cases of breaches of procedural provisions. Determining the level of the fine it also has to be considered that Art. 23 para. 1 of Regulation (EC) No 1-2003 has been reinforced with respect to its corresponding provision of the former Regulation 17-62. The latter provision which contained a limit of 5000 Euro maximum amount of the fine for violation and did not allow to take into account the size of the relevant undertaking (179), has been replaced by a significantly reinforced provision. The maximum limit of the fine can now be fixed of up to 1% of the undertaking's turnover.

107. It must also be borne in mind that inspections are generally ordered only where there is a suspicion of infringements of the competition rules and other means of establishing the facts appear less effective. This was the case in this specific instance. The inspection carried out on 29 and 30 May 2006 was particularly important for this investigation, since the Commission had to act promptly in following up indications of breaches of the competition rules.

108. Furthermore, generally when documents are stored in a sealed room (in the actual case all documents found on the first day of the inspection were stored in this room) or not yet inspected rooms are sealed, the breach of the seal may seriously jeopardise the purpose of the Commission's inspection. It was of particular importance that the seal remained intact, since the documents stored in the room were not (completely) catalogued and copied and it was no longer possible to ascertain when checking them on 30 May 2006 whether some documents were missing (180). EE was aware of this fact from the outset and at no time during the procedure was it contested.

109. On the other hand, with respect to the level of fine, the Commission has considered the fact that this new provision of Regulation (EC) No 1-2003 is being applied for the first time. However, this circumstance cannot result in a level of fine which would risk undermining the deterrent effect of Art. 23 para. 1 of Regulation (EC) No 1-2003.

110. Furthermore it has to be considered that EE - in the case at hand - is one of the largest energy producers in Europe which has at its disposal extensive legal expertise with regard to antitrust law ("inhouse" as well as external). The reinforcement of the fining possibilities for procedural infringements had already been adopted within Regulation No 1-2003 more than three years before the inspection took place and EE had been informed, both in the inspection decision and in the record of sealing, about the significant scope of the possible level of fine in the case of a seal breach. EE must therefore have been aware of the particular importance of the attention of the seal in case of inspections. This applies in particular as sealings had already taken place in the same group a few weeks previously (181). One could expect from EE, which has at its disposal a separate security service, to take the necessary measures to secure the seal so that a breach of seal is prevented from occurring. In any case EE should have ascertained that all the persons in the building who had access to the relevant section were explicitly informed of the existence of the seal and the significance of a seal breach and that these relevant persons were prevented from working on the sealed doors and from damaging the seal. (182)

111. The fact that the Commission did not provide any specific evidence that the door to the room had actually been opened or that documents had indeed been removed cannot be regarded as a mitigating circumstance (183). It is precisely the object and purpose of a seal to ensure that the Commission does not have to provide such evidence, which lies primarily within the sphere of the undertaking. In any case, it was impracticable for the Commission to provide such additional evidence as not all of the documents stored in room G.505 had been copied and catalogued before.

112. Nor can EE claim to have cooperated with the Commission beyond the extent required (184). EE merely replied to requests for information sent by the Commission and contributed little to further clarification of the case, and indeed still disputes the breach of seal.

113. The level of the imposed fine which corresponds to approximately 0.14% of EE's turnover is determined by the seriousness of the infringement of the EC procedural rules in competition cases and by the particular circumstances of the case. In the light of all the relevant factors described, the fine is set at EUR 38 000 000,

Has adopted this decision:

Article 1

E.ON Energie AG broke a seal affixed by representatives of the Commission pursuant to Article 20(2)(d) of Regulation No 1-2003 and at least negligently infringed Article 23(1)(e) of Regulation No 1-2003.

Article 2

A fine of EUR 38 000 000 is hereby imposed on E.ON Energie AG for the infringement referred to in Article 1.

The fine shall be paid in Euro within three months of the date of the notification of this decision to the following account:

Account no 642-0029000-95 of the European Commission Banco Bilbao Vizcaya Argentaria S.A. (BBVA) Avenue des Arts, 43 B-1040 Bruxelles

(SWIFT-Code: BBVABEBB - IBAN-Code: BE76 6420 0290 0095)

After expiry of that period, interest shall automatically be payable at the interest rate applied by the European Central Bank to its main refinancing operations on the first day of the month in which this Decision is adopted plus 3.5 percentage points.

Article 3

This Decision is addressed to:

E.ON Energie AG Brienner Straße 40 D- 80333 Munich Germany

This Decision shall be enforceable pursuant to the first paragraph of Article 256 of the EC Treaty.

Notes:

1 OJ L 1, 4.1.2003, p. 1, as last amended by Regulation No 411/2004.

2 OJ L 123, 27.4.2004, p. 18.

3 OJ C ...,...2008, p.

4 OJ C ...,...2008, p.

5 File, pp. 1 et seq. The decision was served on EE on 29.5.2006.

6 Mr [...]*[...]*, acting as team leader, Mrs [...]*, Mr [...]* and Mr [...]*.

7 Mrs [...]*, Mr [...]*, Mr [...]*, Mr [...]*, Mr [...]* and Mr [...]*.

8 File, p. 7.

9 For details of the sealing process see paragraphs 4253-43 below. Furthermore, other rooms of EE were sealed the same day as well. These seals remained intact.

10 File, p. 8.

11 Replies by six inspectors ([...]*) to question No 3 in the questionnaire for inspectors. File, pp. 13 et seq.

12 3M's reply to the request for information sent on 6.12.2006, p. 6.

13 File, p. 12, picture 006.jpg.

14 See the replies of ten inspectors ([...]*) to question No 6 in the questionnaire for inspectors, file, pp. 13 et seq.

15 Extract from the annex to the record of a sealing of 30 May 2006, file, p. 10.

16 File, pp. 9 et seq.

17 See footnote 11 above.

18 In the replies to questions 15 and 16 in the request for information sent on 9.8.2006 (file, pp. 61 et seq.), EE also confirmed the condition of the seal as described and confined its comments to possible explanations as to why this might have occurred.

19 File, p. 11.

20 This is also conceded by the [institute]*-document from 10 December 2007, p.7: "Captured in a photograph is only the condition of the film of the seal on the door frame."

21 File, p. 12: Images 000.jpg and 001.jpg show the position of the seal on the door frame (after opening the door, after the left part of the seal was removed from the door by Commission officials). Image 002.jpg shows the rear of the seal, images 003.jpg and 005.jpg show the - insofar unchanged - seal affixed to the door frame with the VOID message extensively evident, and image 006.jpg shows a detail of the right half of the seal not having been removed by the Commission on whose edge traces of adhesive in the form of the VOID message are evident.

22 File, p. 13 et seq.

23 (Summary of) EE's reply to the request for information sent on 9.8.2006, file, pp. 56-58: EE staff and security personnel have cards allowing them automatic access. Visitors and cleaning staff are allowed access only after specific prior authorisation. The building is monitored by video cameras and is equipped with movement sensors on doors and windows and a burglar alarm system. The building is also guarded at night by security staff.

24 See EE's reply to the request for information sent on 9.8.2006, file, p. 62.

25 [...]* reply to the request for information sent on 31.8.2006, file, pp. 217 et seq.

26 [...]* reply to the request for information sent on 1.9.2006, file, pp. 223 et seq.

27 See the Commission's letter of 11.10.2006.

28 The various arguments put forward by EE are dealt with in greater detail in this Decision in the legal assessment.

29 3M's reply to the request for information sent on 6.12.2006, pp. 3-4.

30 The seals are properly stored in a Commission filing cabinet. They are not exposed to humidity or wide fluctuations in temperature.

31 3M's reply to the request for information sent on 6.12.2006, p. 6.

32 [Institute]* expert opinion I, p. 4.

33 [Institute]* expert opinion I, p. 6.

34 See above, paragraph 6.

35 [Institute]* expert opinion I, p. 25.

36 [Institute]* expert opinion II, pp. 31 and 32.

37 In preparing the commissioning of the expert opinion, the Commission had already, by letter dated 2.2.2007, given EE the opportunity to comment on the scope of the investigations and their conduct. EE submitted its comments by letter dated 9.2.2007.

38 Here, as well as in the other expert opinions ([Institute]* expert opinion I and [Institute]* expert opinion II) the interpretation did not form part of the accredited test, see paragraph 69.

39 Letter of EE of 10 December 2007.

40 Vgl. Krüger expert opinion II, p. 2, point 2; p. 17, point 3; p. 18.

41 Cf. p. 16, paragraphs 3-7 Krüger expert opinion II.

42 Cf. p. 3 Krüger expert opinion II.

43 Cf. p. 4/5 Krüger expert opinion II.

44 Cf. p. 9 Krüger expert opinion II.

45 Cf. p. 12 Krüger expert opinion II.

46 Cf. p. 13 Krüger expert opinion II.

47 Only when the seal was removed, VOID-signs appeared. This was not the case in the tests with vibrations/elongations.

48 Cf. p. 15/16 Krüger expert opinion II.

49 Cf. p. 17 Krüger expert opinion II.

50 Which is Krüger expert opinion I.

51 Cf. p. 4 Krüger expert opinion II.

52 See paragraphs 79, 80 and 91

53 T-44-02 OPPO, judgement of 27 September 2006 - Dresdner Bank v Commission, nyp, § 59; Case C-185-95 P Baustahlgewebe v Commission [1998] ECR I-8417, § 58, and Case C-49-92 P Commission v Anic Partecipazioni [1999] ECR I-4125, § 86. ; Joined Cases T-67-00, T-68-00, T- 71-00 and T-78-00 JFE Engineering and Others v Commission, [2004] ECR II-2501, paragraph 179.

54 Joined Cases T-67-00, T-68-00, T-71-00 and T-78-00 JFE Engineering and Others v Commission, [2004] ECR II-2501, paragraph 187.

55 Idem.

56 Furthermore, even the possibility of an alternative explanation of the facts has not established by EE, since none of the tests carried out by the [institute]* or M Krüger resulted in a "false positive" reaction that is similar to the state of the seal on 30.5.2006.

57 EE's reply to the request for information sent on 6.12.2006, p. 3.

58 Replies of six inspectors ([...]*) to question No 3 in the questionnaire for inspectors, file, pp. 13 et seq.

59 EE's reply of 13.11.2006 to the statement of objections, p. 11.

60 EE's reply to the request for information sent on 9.8.2006, file, p. 60.

61 Replies given by seven inspectors ([...]*) to question No 2 in the questionnaire for inspectors, file, pp. 13 et seq.

62 EE's reply of 13.11.2006 to the statement of objections, p. 9.

63 EE's reply of 13.11.2006 to the statement of objections, p. 23.

64 Krüger expert opinion I, pp. 8 and 18. In subsequent tests in which the seal was affixed to non precleaned surfaces, a peel off of the seal never occurred, cf. for example Krüger expert opinion II, p.19.

65 Letter of 6.7.2007, p. 6.

66 Cf. Krüger expert opinion II, p. 18.

67 3M's reply to the request for information sent on 6.12.2006, p. 4.

68 See, in particular, paragraphs 65 and 68.

69 3M's reply to the request for information sent on 6.12.2006, p. 4.

70 3M's reply to the request for information sent on 6.12.2006, p. 4.

71 Krüger expert opinion I, pp. 4-8.

72 Letter of 6.7.2007, p. 7.

73 3M's reply to the request for information sent on 6.12.2006, pp. 4 and 5.

74 3M's reply to the request for information sent on 6.12.2006, p. 5.

75 3M's reply to the request for information sent on 6.12.2006, p. 5.

76 Krüger expert opinion I, pp. 9 and 18.

77 Krüger expert opinion I, p. 9.

78 Replies by seven inspectors ([...]*) to questions Nos 5 and 6 in the questionnaire for inspectors, file, pp. 13 et seq. See also image 003.jpg, 005.jpg and 006.jpg, which show VOID markings extensively spread over the part of the seal remaining on the door frame.

79 Letter of 6.7.2007, p. 12.

80 EE's reply of 13.11.2006 to the statement of objections, p. 17.

81 Cf. Krüger expert opinion II, p. 17.

82 3M's reply to the request for information sent on 6.12.2006, p. 4.

83 3M's reply to the request for information sent on 6.12.2006, p. 4

84 Letter dated 6.7.2007, pp. 9 and 10.

85 Letter dated 6.7.2007, p. 4.

86 See paragraph 21 above.

87 See paragraph 21 above.

88 [Institute]* expert opinion III, p. 19. The examination of shearing forces (i.e. forces that operate in a parallel way to the seal surface) did not form part of the series of tests. EE's claim in its submission of 10.12.2007, that Mr Krüger had misinterpreted the concept of "shearing forces", is obiously unfounded. Although the short explanation on page two of the Krüger Expert Opinion II may be ambiguous in this respect, the following explanations ("the whole sticking area resists against a displacement of the seal") clarify that Mr Krüger used the notion of shearing forces in a correct manner. Mr Krüger's conclusion, that pure shearing forces are practically inexistent, therefore remain valid and have not been contradicted by EE.

89 [Institute]* expert opinion III, p. 26.

90 See also the answer of 3M of 21 December 2006, p. 7:[A breach of the seal...] "would give rise to showing of glue parts at the back of the seal.".

91 Krüger expert opinion II, p. 6.

92 See paragraph 56.

93 Cf. p. 3 of the expert opinion of 20 November 2007.

94 3M's reply to the request for information sent on 6.12.2006, p. 4; Krüger expert opinion, pp. 7 and 18.

95 Letter dated 6 July 2007, p. 11.

96 See above, paragraph 12.

97 See above, paragraph 14.

98 [...]*'s reply to the request for information sent on 1.9.2006, file, p. 223.

99 Krüger expert opinion I, p. 18; the respective conclusion was confirmed in the Krüger expert opinion II once again, see p. 20; cf. also 3M's clear reply from 21 december 2006, point 5: "When there is a displacement of the adhesive and seal of 1 mm visible at the edges this is an indication that the seal

100 Replies by eight inspectors ([...]*) to questions No 5 and No 6 in the questionnaire for inspectors, file, pp. 13 et seq.

101 See paragraph 13 above.

102 Letter of 10 December 2007, p. 8.

103 Statement of objections, paragraph 31.

104 3M's reply of 21.12.2006, point 5: "When there is a displacement of the adhesive and seal of 1 mm visible at the edges this is an indication that the seal has been removed and replaced slightly unaligned."

105 Letter dated 6.7.2007, p. 15, letter dated 10.12.2007, pp. 7 and 15.

106 See paragraph 9.

107 See paragraph 9.

108 See in particular the explicit confirmation of the extensive appearance of VOID markings by the inspectors [...]*, file pp. 13.

109 File p. 12: Image 002.jpg shows the back of the seal, images 003.jpg and 005.jpg show in parts the - insofar unchanged - seal still adhering to the door frame on which the VOID marking is extensively visible, image 006.jpg shows a detailed view of the right part of the seal which has not been removed by the Commission at the edges of which traces of glue of approx. 2 mm in form of the VOID marking are recognizable.

110 Cf. paragraph 13.

111 EE's reply to the request for information sent on 9.8.2006, file, p. 62.

112 From the formula 42 + 0,532 = 4,03492 it follows that stretching (difference of distance of door and frame in passive state and in case of maximum "play" of 0,53 mm) constitute more than 0,0349 mm.

113 This follows from the formula 42 + 22 = 4,472.

114 Cf. e.g. [Institute]* expert opinion II, p. 29.

115 [Institute]* submission of 12.10.2007, p. 2..

116 Krüger expert opinion II, pp. 2-7.

117 Krüger expert opinion II, p. 14.

118 Krüger expert opinion I, pp. 11-14; confirmed by Krüger expert opinion II, pp. 14, 20.

119 Krüger expert opinion I, p. 15.

120 Krüger expert opinion I, pp. 11-14; confirmed by Krüger expert opinion II, p. 18 (Manifold stress and stretching for 0,53 mm of a seal foil treated and watered by Synto).

121 3M's reply of 21 December 2006, p.6.

122 EE's reply to request for information sent on 6 December 2006, p. 2.

123 This is supported by the video played by EE in the hearing that shows s seal after considerable door vibrations on which VOID markings became only visible on the loosened part of the seal.

124 As confirmed by EE in writing on 24.4.2007 and 3.5.2007, no changes have been made to the relevant door since 30.5.2006.

125 Krüger expert opinion I, pp. 12-15 and 18. The statement by a representative of EE (Mr [...]*) that a "D" was visible at the lower edge of seal 6 could not be confirmed by the other persons present at the investigation.

126 See above, paragraph 14.

127 EE's reply to the request for information sent on 9.8.2006, file, p. 62.

128 See paragraph 74 below.

129 EE's reply to the request for information sent on 9.8.2006, file, p. 62.

130 EE's reply of 13.11.2006 to the statement of objections, p. 19: "E.ON Energie AG has therefore not itself conceded that there may have been a displacement of the seal; it has simply repeated Mrs [...]*'s statement."

131 Krüger expert opinion I, p. 10; cf. 3M's reply to the request for information sent on 29.8.2006,

file, p. 206.

132 Krüger expert opinion I, pp. 10 et seq.; Krüger expert opinion II, pp. 8 et seq.; 19.

133 Krüger expert opinion I, pp. 8 and 10¸ Krüger expert opinion II, pp. 8 et seq; 19.

134 From the mere characterization of Synto as "watery dilution" it cannot be deduced that M. Krüger had a different cleaning product ("probably a different version"), see letter sent on 10 December 2007, p. 10.

135 See product description by Synto Pramol-Chemie AG: http://www.pramol.com/index.php?cPath=37&sort=2a&page=3&osCsid=d3b587006eeec23da3a967 00a33897b5.

136 Letter from EE dated 6.7.2007, p. 4.

137 Krüger expert opinion II, pp. 8 et seq.; 19.

138 [Institute]* expert opinion II, pp. 31 et seq.

139 [Institute]* expert opinion II, p. 33.

140 The material used by the [INSTITUTE]* has a size of 1 cm to 4 cm whereas the Commission seals have a size of 6 cm to 9 cm. The outcome of this is a different relation between the circumference and the surface. As a consequence the results found, e.g. concerning the diffusion of Synto or the much smaller adhesive force, are not applicable, see Krüger expert opinion II, pp. 3-4.

141 [Institute]* expert opinion II, p. 7.

142 Cf. the product information of the security films 7380, 7866, 7384 of 3M, point 7 "Stress with chemicals and dilutions".

143 Krüger expert opinion II, pp. 4-5; 14.

144 Cf. E.g. [Institute]* expert opinion III, p. 8, which stresses in describing the tests with Synto that "the whole edge of the sealing film came in touch with a defined amount of Synto" (translation done by the Commission).

145 Krüger expert opinion I, pp. 10 et seq.; Krüger expert opinion II, pp. 9-12; 14-17; 19-20. The mere assertion of EE in reply to the request for information of the Commission sent on 19 October 2007 saying that "it could be presumed that water molecules also diffuse through the seal film as there are no polymers being impermeable to water vapour" is only of theoretical nature. Besides, this is contrary to the assumption of [Institute]* that a possible impact over the edges is still relevant regarding the question of impact of moisture, see [Institute]* expert opinion III, p.8.

146 Krüger expert opinion I, p. 10; confirmed by the tests in Krüger expert opinion II, pp. 9-12; 19-20.

147 See paragraph 9 above.

148 See letter of EE of 10.12.2007, p. 6 and 13.

149 Cf. EE's letter sent on 6 July 2007, pp. 22/23; confirmed by Krüger expert opinion II, p. 12.

150 See above, paragraph 13.

151 Cf. the tests in the Krüger expert opinion I, pp. 11-17; confirmed by Krüger expert opinion II, p. 12.

152 3M's reply to the request for information sent on 6 December 2006, p. 6: "3M is not aware of having received any complaints about the functioning of the Product or converted products concerning the appearance of void signs after application in circumstances where a "false positive" indication id alleged."

153 E.g., see letter dated 6July 2007, pp. 4 and 5; letter dated 10 December 2007, p. 8.

154 Letter dated 6 July 2007, p. 5.

155 [Institute]* expert opinion III, p. 39.

156 See paragraph 66 above.

157 Cf. already paragraph 66 et seq. above.

158 The reason laid down in the letter of [Institute]* dated 10 December 2007 (p.2) according to which seal material should be saved, is not convincing in this respect.

159 Letter of [Institute]* dated 10 December 2007, p.4.

160 Letter of [Institute ]* dated 10 December 2007, p.6; letter of EE dated 10 December 2007, pp. 11/12.

161 Technical data sheet, p. 197 of the file.

162 3M's reply to the request for information sent on 6 December 2006, p.6: "3M is not aware of having received any complaints about the functioning of the Product or converted products concering the appearance of void signs after application in circumstances where a "false positive" indication is alleged."

163 In this respect the reproach of EE in the letter dated 10 December 2007 (p. 12) that the Commission had failed to check the seals on false positives goes astray.

164 See Krüger exper opinion II, p. 15.

165 See Krüger expert opinion II, pp. 15-17; 19/20.

166 Krüger expert opinion II, p. 15.

167 See paragraph 86 above. Even if the seal has not been affixed yet, the adhesive surface of the seal is protected from the entering of humidity by a siliconised protection film.

168 Cf. [Institute]* expert opinion III, p. 39.

169 See EE's reply of 25 October 2007 on the Commission's request for information dated 19.10.2007.

170 In the absence of data in the premises of EE, EE only claims that there was an increased humidity outside the premises. EE has not shown that the same level of humidity was to be found inside the building, see reply of EE to the Commission's request for information of 19.10.2007.

171 EE's letter dated 10.12.2007, p. 13; [Institute]*-letter dated 10.12.2007, p. 5.

172 Cf. [Institute]* expert opinion III, pp. 33-37.

173 Krüger expert opinion II, p. 21.

174 Krüger expert opinion II, p. 18/19.

175 What is more, a German undertaking can be assumed to know as a general rule that breach of seal is a punishable offence; see Article 136(2) of the German Criminal Code.

176 EE's reply to the request for information sent on 9.8.2006, file, p. 62; "After Mrs [...]* noticed that room G.505 was locked, she went to the security office and asked the security personnel to unlock the door. ... Upon further questioning, Mrs [...]* said that she then returned to building section G, 5th floor, cleaned any finger marks off the door and door handle, etc. of the sealed conference room (as it was her job to do) and in so doing also wiped the seal." The Commission concludes from this that, when she tried to open the locked and sealed door, Mrs [...]* was able to ascertain only by shaking the door that room G.505 was locked. Even if, according to information provided by 3M and the opinion given by the expert Dr Krüger, it is impossible that a mere wipe (without any other action) could have altered the seal, it was at the very least negligent of EE not to have prevented such an act.

177 EE's reply to the request for information sent on 9.8.2006, file, pp. 56 et seq.; reply given by [...]* to the request for information sent on 1.9.2006, file, pp. 224 et seq.

178 See paragraph 19 above.

179 See Art. 15 (1) of Regulation (EC) 17/62, JO 13 of 21.2.1962, 204-211.

180 See paragraph 70.

181 The Commission had already carried out inspections at E.ON's premises on 16 and 17 May 2006, see the Commission's press release MEMO/06/205.

182 See paragraphs 19 and 74 above.

183 EE's reply of 13.11.2006 to the statement of objections, p. 39.

184 EE's reply of 13.11.2006 to the statement of objections, p. 38.