Language of document : ECLI:EU:T:2021:404

JUDGMENT OF THE GENERAL COURT (Eighth Chamber, Extended Composition)

30 June 2021 (*)

(REACH – Establishment of a list of substances identified for eventual inclusion in Annex XIV to Regulation (EC) No 1907/2006 – Entry of octamethylcyclotetrasiloxane (D4), decamethylcyclopentasiloxane (D5) and dodecamethylcyclohexsiloxane (D6) in that list – Articles 57 and 59 of Regulation No 1907/2006 – Annex XIII to Regulation No 1907/2006 – Weight-of-evidence determination – Manifest error of assessment – Proportionality)

In Case T‑519/18,

Global Silicones Council, established in Washington, DC (United States), and the other applicants whose names are listed in the annex, (1) represented by R. Cana and D. Abrahams, lawyers,

applicants,

supported by

American Chemistry Council, Inc. (ACC), established in Washington, represented by K. Nordlander and C. Grobecker, lawyers,

intervener,

v

European Chemicals Agency (ECHA), represented by M. Heikkilä, W. Broere and A. Hautamäki, acting as Agents,

defendant,

supported by

Federal Republic of Germany, represented by J. Möller, D. Klebs, S. Heimerl and S. Costanzo, acting as Agents,

and by

European Commission, represented by L. Haasbeek and R. Lindenthal, acting as Agents,

interveners,

APPLICATION under Article 263 TFEU for the annulment in whole or in part of the decision of ECHA of 27 June 2018 including octamethylcyclotetrasiloxane (D4), decamethylcyclopentasiloxane (D5) and dodecamethylcyclohexsiloxane (D6) in the Candidate List for eventual inclusion in Annex XIV to Regulation (EC) No 1907/2006 of the European Parliament and of the Council of 18 December 2006 concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH), establishing a European Chemicals Agency, amending Directive 1999/45/EC and repealing Council Regulation (EEC) No 793/93 and Commission Regulation (EC) No 1488/94 as well as Council Directive 76/769/EEC and Commission Directives 91/155/EEC, 93/67/EEC, 93/105/EC and 2000/21/EC (OJ 2006 L 396, p. 1, corrigendum OJ 2007 L 136, p. 3),

THE GENERAL COURT (Eighth Chamber, Extended Composition),

composed of S. Papasavvas, President, J. Svenningsen, R. Barents, T. Pynnä and J. Laitenberger (Rapporteur), Judges,

Registrar: B. Lefebvre, Administrator,

having regard to the written part of the procedure and further to the hearing on 30 September 2020,

gives the following

Judgment

 Legal framework

1        On 1 June 2007, Regulation (EC) No 1907/2006 of the European Parliament and of the Council of 18 December 2006 concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH), establishing a European Chemicals Agency, amending Directive 1999/45/EC and repealing Council Regulation (EEC) No 793/93 and Commission Regulation (EC) No 1488/94 as well as Council Directive 76/769/EEC and Commission Directives 91/155/EEC, 93/67/EEC, 93/105/EC and 2000/21/EC (OJ 2006 L 396, p. 1, corrigendum OJ 2007 L 136, p. 3) entered into force.

2        Annex XIII to Regulation No 1907/2006 (‘Annex XIII’) lays down the criteria for the identification of persistent (‘P’), bioaccumulative (‘B’) and toxic (‘T’) substances (together, ‘PBT’), and very persistent (‘vP’) and very bioaccumulative (‘vB’) substances (together, ‘vPvB’).

3        On 15 March 2011, the European Commission adopted Regulation (EU) No 253/2011 amending Regulation No 1907/2006 as regards Annex XIII (OJ 2011 L 69, p. 7). That regulation amended the former Annex XIII taking into account experience acquired at international level.

4        The preamble to Annex XIII now states:

‘This Annex lays down the criteria for the identification of persistent, bioaccumulative and toxic substances (PBT substances), and very persistent and very bioaccumulative substances (vPvB substances) as well as the information that must be considered for the purpose of assessing the P, B, and T properties of a substance.

For the identification of PBT substances and vPvB substances a weight-of-evidence determination using expert judgement shall be applied, by comparing all relevant and available information listed in Section 3.2 with the criteria set out in Section 1. This shall be applied in particular where the criteria set out in Section 1 cannot be applied directly to the available information.

A weight-of-evidence determination means that all available information bearing on the identification of a PBT or a vPvB substance is considered together, such as the results of monitoring and modelling, suitable in vitro tests, relevant animal data, information from the application of the category approach (grouping, read-across), (Q)SAR results, human experience such as occupational data and data from accident databases, epidemiological and clinical studies and well documented case reports and observations. The quality and consistency of the data shall be given appropriate weight. The available results regardless of their individual conclusions shall be assembled together in a single weight-of-evidence determination.

The information used for the purposes of assessment of the PBT/vPvB properties shall be based on data obtained under relevant conditions.

The identification shall also take account of the PBT/vPvB-properties of relevant constituents of a substance and relevant transformation and/or degradation products.

This Annex shall apply to all organic substances, including organo-metals.’

5        Annex XIII provides, inter alia:

‘1.1.2. Bioaccumulation

A substance fulfils the bioaccumulation criterion (B) when the bioconcentration factor in aquatic species is higher than 2 000.

1.2.2. Bioaccumulation

A substance fulfils the “very bioaccumulative” criterion (vB) when the bioconcentration factor in aquatic species is higher than 5 000.’

6        Annex XIII also states as follows:

‘3.2. Assessment Information

The following information shall be considered for the assessment of P, vP, B, vB and T properties, using a weight-of-evidence approach.

3.2.2. Assessment of B or vB properties

(a)      Results from a bioconcentration or bioaccumulation study in aquatic species;

(b)      Other information on the bioaccumulation potential provided that its suitability and reliability can be reasonably demonstrated, such as:

–        Results from a bioaccumulation study in terrestrial species;

(c)      Information on the ability of the substance to biomagnify in the food chain, where possible expressed by biomagnification factors or trophic magnification factors.’

 Background to the dispute

7        The first applicant, Global Silicones Council, is a non-stock corporation, established in the United States, representing companies which manufacture and sell silicone products throughout the world. The other applicants, whose names are listed in the annex, are companies which manufacture, sell and supply the chemical substances octamethylcyclotetrasiloxane (‘D4’), decamethylcyclopentasiloxane (‘D5’) and dodecamethylcyclohexsiloxane (‘D6’) on their own or in a mixture.

8        On 14 October 2014, the Executive Director of the European Chemicals Agency (ECHA) requested the ECHA Member State Committee (‘the MSC’) to deliver an opinion on the persistence and bioaccumulation of D4 and D5 in the light of the criteria laid down in Annex XIII.

9        On 17 April 2015, the United Kingdom of Great Britain and Northern Ireland submitted to ECHA a dossier based on Annex XV to Regulation No 1907/2006 (‘Annex XV’) proposing a restriction of the use of D4 and D5 in cosmetic products that were washed off in normal conditions of use. The restriction proposal was based on the purported PBT and vPvB properties of D4 and on the purported vPvB properties of D5.

10      On 22 April 2015, the MSC adopted an opinion according to which both D4 and D5 met the criteria set out in Annex XIII regarding the identification of vP and vB substances.

11      On 10 March 2016, the ECHA Committee for Risk Assessment (‘the RAC’) adopted an opinion on the basis of an examination of the United Kingdom’s restriction proposal and of the MSC’s opinion. The RAC, referring to the MSC’s opinion of 22 April 2015, stated that it had not reassessed the vP and vB properties of D4 and D5, but that it had examined the T properties of those substances. The RAC concluded that D4 fulfilled the identification criteria for PBT and vPvB substances set out in Annex XIII and that D5 fulfilled the identification criteria for vPvB substances, but that the latter, on the basis of the available evidence, did not meet the criteria set out in Annex XIII for toxicity.

12      A restriction on D4 and D5 was adopted by the Commission on 10 January 2018 (Regulation (EU) 2018/35 amending Annex XVII to Regulation No 1907/2006 as regards D4 and D5 (OJ 2018 L 6, p. 45)). By application received at the Court Registry on 2 April 2018, the majority of the applicants in the present case brought an action for annulment of that restriction. ECHA has intervened in support of the Commission in the case.

13      On 1 March 2017, the competent authority of the Federal Republic of Germany submitted dossiers under Article 59(3) of Regulation No 1907/2006 in accordance with Annex XV, proposing that D4 and D5 be identified as substances of very high concern on account of their PBT and vPvB properties. On 21 December 2017, the Commission, under Article 59(2) of that regulation, requested ECHA to prepare a dossier in accordance with Annex XV for the identification of D6 as a substance of very high concern. ECHA concluded in that dossier that D6 met the PBT and vPvB criteria.

14      On 8 March 2018, ECHA published the three dossiers prepared in accordance with Annex XV relating to D4, D5 and D6, respectively (‘the Annex XV dossiers’). On the same day, ECHA invited all interested parties to submit their comments on those dossiers, in accordance with Article 59(4) of Regulation No 1907/2006. ECHA received comments including, on 23 April 2018, those of the applicants.

15      Subsequently, the German competent authority and ECHA produced three documents (the ‘RCOM documents’), dated 12 June 2018, containing those authorities’ responses to all the comments received by ECHA in the course of the public consultation.

16      Having received comments concerning the identification of the three substances at issue, ECHA, in accordance with Article 59(7) of Regulation No 1907/2006, referred the dossiers to the MSC. In accordance with the MSC’s working procedures for the identification of substances of very high concern, the MSC was provided with an Annex XV dossier, a draft MSC agreement and a draft support document for each of the three substances.

17      At its 60th meeting, held from 12 to 14 June 2018, the MSC reached unanimous agreements on the identification of D4, D5 and D6 as substances of very high concern, on the ground that those substances met the criteria for identification as PBT and vPvB substances under Article 57(d) and (e) of Regulation No 1907/2006. The grounds for those agreements are set out in the respective support documents.

18      The support documents for D4 and D5 refer several times, including in the summaries concerning the persistence and bioaccumulation of the two substances, to the MSC’s opinion of 22 April 2015 on D4 and D5 and, concerning the toxicity of D4, to the RAC’s opinion of 10 March 2016. The data are supplemented in the support documents for D4 and D5 by references to new studies published after those opinions. The support document for D4 concludes that D4 has PBT and vPvB properties in accordance with Annex XIII. The support document for D5 does not evaluate data on the toxicity of that substance for human health or the environment. That document concludes that D5 is a vPvB substance and that, when it contains D4 in a concentration greater than or equal to 0.1% (weight by weight), it is also PBT.

19      The support document for D6 presents, inter alia, assessments on the degradation and distribution of that substance and refers to data provided in the support documents for D4 and D5. The support document for D6 does not assess the environmental and human-health toxicity of D6. It concludes that D6 is a vPvB substance and states that it also meets the criteria for PBT substances when it contains D4 in a concentration greater than or equal to 0.1% (weight by weight). In addition, given that D4 and D5 were identified as impurities in D6, it concluded, with reference to the MSC’s opinion of 22 April 2015 and the RAC’s opinion of 10 March 2016, that D6 also fulfils the criteria for vPvB substances when it contains D4 or D5 in a concentration greater than or equal to 0.1% (weight by weight).

20      On 27 June 2018, pursuant to Article 59(8) of Regulation No 1907/2006, ECHA adopted Decision ED/61/2018 regarding the inclusion of D4, D5 and D6 in the Candidate List for eventual inclusion in Annex XIV to Regulation No 1907/2006, as referred to in Article 59(1) of that regulation (‘the Candidate List’), on the ground that they had been identified as PBT and vPvB substances within the meaning of Article 57(d) and (e) of Regulation No 1907/2006 (‘the contested decision’), published the same day on ECHA’s website.

21      On the same day, the Candidate List published on ECHA’s website in accordance with Article 59(10) of that regulation was updated to include entries for D4, D5 and D6. The remarks on the respective Candidate List entries provide that D5 fulfils the criteria laid down in Article 57(d) of Regulation No 1907/2006 for a PBT substance when it contains a quantity of D4 greater than or equal to 0.1% (weight by weight). For D6, the remark states that that substance fulfils the criteria laid down in Article 57(d) of Regulation No 1907/2006 for a PBT substance when it contains a quantity of D4 greater than or equal to 0.1% (weight by weight). As regards D6, the remark also states that, in addition to its intrinsic properties, that substance also meets the criteria set out in Article 57(e) of Regulation No 1907/2006 for a vPvB substance when it contains a quantity of D4 or D5 greater than or equal to 0.1% (weight by weight).

 Procedure and forms of order sought

22      By application lodged at the Court Registry on 3 September 2018, the applicants brought the present action.

23      On 5 December 2018, ECHA lodged its defence at the Court Registry.

24      By documents lodged at the Court Registry on 10 and 19 December 2018 respectively, the Federal Republic of Germany and the Commission applied to intervene in support of the form of order sought by ECHA.

25      By document lodged at the Court Registry on 27 December 2018, the American Chemistry Council, Inc. (‘ACC’) applied to intervene in support of the form of order sought by the applicants.

26      By decision of 10 January 2019, the President of the Fifth Chamber of the Court granted the Federal Republic of Germany leave to intervene.

27      On 14 January 2019, the applicants, by separate document lodged at the Court Registry, requested that certain information in the file be treated as confidential vis-à-vis the Commission.

28      By order of 1 February 2019, the President of the Fifth Chamber of the General Court granted the Commission leave to intervene. He also ordered the Court Registry to send the Commission a non-confidential version of the pleadings and documents served on the main parties. Finally, he decided it was necessary to set a deadline for the Commission, first, to submit its observations on the applicants’ request for confidential treatment in its regard and, second, to submit its statement in intervention. By decisions of 10 April 2019, those two deadlines were set at 26 April and 22 May 2019 respectively.

29      On 28 February 2019, the applicants lodged their reply at the Court Registry.

30      By order of 8 April 2019, the President of the Fifth Chamber of the General Court granted ACC leave to intervene.

31      On 25 April 2019, ECHA lodged its rejoinder at the Court Registry.

32      On 22 May 2019, the Federal Republic of Germany, the Commission and ACC lodged their statements in intervention at the Court Registry.

33      By documents lodged at the Court Registry on 30 and 31 July 2019, respectively, the applicants and ECHA submitted their observations on the statements in intervention of the Federal Republic of Germany, the Commission and ACC.

34      By letter of 18 September 2019, corrected by letter of 24 September 2019, the applicants requested that a hearing be held.

35      By decision of 17 October 2019 and following a change in the composition of the Chambers of the Court, the case was reassigned to a new Judge-Rapporteur sitting in the Eighth Chamber.

36      By measure of organisation of procedure of 7 May 2020, the main parties were asked whether or not, in the light of the health situation connected to COVID-19, they were maintaining their request to be heard at a hearing. By letter of 27 May 2020, the applicants stated that they were maintaining that request.

37      On 17 June 2020, the Court (Eighth Chamber), acting on a proposal from the Judge-Rapporteur and by way of the measures of organisation of procedure provided for in Article 89 of the Rules of Procedure of the General Court, put written questions to the main parties and to the Federal Republic of Germany, inviting them to reply to those questions in writing. The parties complied with that request within the period prescribed.

38      On 24 June 2020, on a proposal from the Eighth Chamber, the Court decided, pursuant to Article 28 of its Rules of Procedure, to refer the case to a chamber sitting in extended composition.

39      On 7 August 2020, the Court (Eighth Chamber, Extended Composition), acting on a proposal from the Judge-Rapporteur and by way of the measures of organisation of procedure provided for in Article 89 of the Rules of Procedure, invited each of the parties to submit its observations on the other parties’ replies to the written questions of the Court of 17 June 2020. The main parties complied with that request within the period prescribed.

40      On a proposal from the Judge-Rapporteur, the Court (Eighth Chamber, Extended Composition) decided to open the oral part of the procedure.

41      The parties presented oral argument and answered the questions put to them by the Court at the hearing on 30 September 2020.

42      The applicants claim that the Court should:

–        declare the action admissible and well founded;

–        annul the contested decision in so far as it includes D4, D5 and D6 in the Candidate List as substances of very high concern;

–        alternatively, should the Court not annul the inclusion of all three substances, annul the contested decision with regard to one or more of those inclusions in the Candidate List;

–        order ECHA to pay the costs;

–        take such other or further measure as justice may require.

43      ECHA contends that the Court should:

–        dismiss the action for annulment;

–        order the applicants to pay the costs of both parties.

44      The Federal Republic of Germany supports the form of order sought by ECHA seeking dismissal of the action and an order that the applicants pay the costs.

45      The Commission, for its part, supports the form of order sought by ECHA seeking dismissal of the action and an order that the applicants pay the costs.

46      ACC submits that the Court should declare the action for annulment well founded.

 Law

47      In support of their action, the applicants rely on two pleas in law. The first plea alleges that ECHA manifestly erred in its assessment of the B properties of D4, D5 and D6 and in its assessment of the T properties of D5 and D6, that it exceeded its powers and that it infringed Article 59 of Regulation No 1907/2006. By the second plea, the applicants allege breach of the principle of proportionality.

 First plea in law:commission by ECHA of a manifest error of assessment in the assessment of the B properties of D4, D5 and D6 and in the assessment of the T properties of D5 and D6, exceedance of its powers and infringement ofArticle 59 of Regulation No 1907/2006

48      By their first plea, the applicants, supported by ACC, claim that ECHA committed a manifest error of assessment in the assessment of B and T properties, exceeded its powers and infringed Article 59 of Regulation No 1907/2006 in a number of ways.

49      ECHA, supported by the Federal Republic of Germany and the Commission, contends that the first plea must be rejected as unfounded.

50      As a preliminary point, it should be noted that the contested decision was adopted on the basis of Article 59 of Regulation No 1907/2006. As is apparent from Article 1(1) thereof, the purpose of that regulation is to ensure a high level of protection of human health and the environment, including the promotion of alternative methods for assessment of hazards of substances, as well as the free circulation of substances on the internal market while enhancing competitiveness and innovation.

51      According to the case-law, in an area of evolving and complex technology such as that in the present case, the competent EU authorities have a broad discretion, in particular as to the assessment of highly complex scientific and technical facts, in order to determine the nature and scope of the measures which they adopt, whereas review by the European Union judicature has to be limited to verifying whether there has been a manifest error of assessment or a misuse of powers, or whether the authorities have manifestly exceeded the limits of their discretion. In such a context, the European Union judicature cannot substitute its assessment of scientific and technical facts for that of the authorities of the European Union on which alone the FEU Treaty has placed that task (see judgments of 21 July 2011, Etimine, C‑15/10, EU:C:2011:504, paragraph 60, and of 30 April 2015, Polynt and Sitre v ECHA, T‑134/13, not published, EU:T:2015:254, paragraph 52 and the case-law cited).

52      In addition, it must be stated that the broad discretion of the EU authorities, which implies limited judicial review of its exercise, applies not only to the nature and scope of the measures to be taken but also, to some extent, to the finding of the basic facts. However, even though such judicial review is of limited scope, it requires that the EU institutions which have adopted the act in question must be able to show before the EU judicature that in adopting the act they actually exercised their discretion, which presupposes that they took into consideration all the relevant factors and circumstances of the situation the act was intended to regulate (see judgments of 8 July 2010, Afton Chemical, C‑343/09, EU:C:2010:419, paragraphs 33 and 34, and of 30 April 2015, Polynt and Sitre v ECHA, T‑134/13, not published, EU:T:2015:254, paragraph 53 and the case-law cited).

53      It is thus for the EU judicature to establish, in the light of the factors relied on by the applicant, whether the evidence relied on is factually accurate, reliable and consistent, whether that evidence contains all the information which must be taken into account in order to assess a complex situation, and whether it is capable of substantiating the conclusions drawn from it (see judgments of 22 November 2007, Spain v Lenzing, C‑525/04 P, EU:C:2007:698, paragraph 57; of 6 November 2008, Netherlands v Commission, C‑405/07 P, EU:C:2008:613, paragraph 55; and of 9 September 2011, France v Commission, T‑257/07, EU:T:2011:444, paragraph 87 and the case-law cited).

54      In the present case, the substances at issue were identified as being of very high concern under Article 57(d) and (e) of Regulation No 1907/2006 because it was concluded that they were PBT and vPvB substances. According to Annex XIII, a ‘weight-of-evidence determination using expert judgement shall be applied, by comparing all relevant and available information listed in Section 3.2 [of that annex] with the criteria set out in Section 1 [of that annex]’ and ‘[the] weight-of-evidence determination means that all available information bearing on the identification of a PBT or a vPvB substance is considered together’. According to that annex, the competent authorities must give appropriate weight to the quality and consistency of the data and assemble the results together in a single weight-of-evidence determination. It is thus apparent from Regulation No 1907/2006 that the identification of substances by means of a weight-of-evidence determination must be carried out on the basis of complete data, and that it requires the judgement of experts who have a broad discretion in the assessment of highly complex scientific and technical facts.

55      The first plea as put forward by the applicants has eight parts. The first, second and third parts concern the three substances at issue. The fourth and fifth parts concern only D4 and D5. The sixth and eighth parts concern only D6 and the seventh part concerns only D5.

 First part of the first plea: commission by ECHA of a manifest error of assessment in relying on the opinions of the MSC and the RAC without conducting its own assessment of the information available to it, thereby ‘importing’ the errors vitiating those opinions

56      The applicants claim that ECHA had a direct and independent duty to assess the information before it on the bioaccumulation of D4, D5 and D6. However, in the support documents for D4 and D5, ECHA referred extensively to the MSC’s opinion of 22 April 2015, particularly as regards the bioaccumulation of those substances in aquatic and terrestrial organisms, the field data and the conclusions on bioaccumulation in general. The support document for D6 contains several references throughout the assessment to the data on D4 and D5. ECHA did not assess whether those conclusions, which were reached in the context of a different procedure, were also valid for the procedure leading to the adoption of the contested decision. The applicants claim that the ‘EU legislator did not envisage a system of “cut and paste” or recycled reasoning [and that] procedures [should] not be blurred by the wholesale importing of the work product from one decision straight into another decision’.

57      ECHA, supported by the Federal Republic of Germany, disputes those arguments.

58      As a preliminary point, it should be noted that it is apparent from the file, and in particular from the support documents and the minutes of the MSC meeting held from 12 to 14 June 2018, that the MSC carried out an assessment of the vB properties of the three substances. It is apparent from the support documents for D4 and D5 that, in the context of an overall weight-of-evidence determination, the MSC took into account the data set out in the MSC’s opinion of 22 April 2015, including in its annexes, the assessment of that information set out by the MSC in that opinion and the data from new studies published after that opinion. The support document for D6 analyses extensive data relating specifically to the bioaccumulation of that substance and concludes that it fulfils the criteria for vB substances set out in Annex XIII. The mere fact that that support document contains references to data relating to D4 and D5 does not in itself demonstrate that ECHA committed a manifest error of assessment. The minutes of the MSC meeting which took place from 12 to 14 June 2018 indicate inter alia that, following a presentation of the results of the public consultation, the MSC discussed, in plenary session, a number of outstanding questions concerning the assessment of the P properties and of the B and T properties of the three substances at issue, before deciding to identify those substances as being of very high concern.

59      Moreover, the applicants’ line of argument according to which the MSC should not have referred to its opinion of 22 April 2015 because that opinion had been obtained in the context of a different procedure laid down by Regulation No 1907/2006 must be rejected. It is true that that opinion was obtained pursuant to Article 77(3)(c) of Regulation No 1907/2006 and that it was used in a procedure leading to the adoption of a restriction on the placing on the market of D4 and D5 provided for in Title VIII of Regulation No 1907/2006. However, that MSC opinion of 22 April 2015 concerned the vP and vB properties of D4 and D5 assessed in accordance with Annex XIII. The same rules and assessment principles as those required in the present case under Article 57(d) and (e) of Regulation No 1907/2006 were therefore applied. Consequently, neither the fact that the MSC confirmed its earlier opinion nor the fact that it referred to information and conclusions contained therein, because it considered that information and those conclusions still to be valid and consistent with its opinion, constitutes a manifest error of assessment.

60      Accordingly, the first part of the first plea must be rejected as unfounded.

 Fourth part of the first plea: commission by ECHA of a manifest error of assessment in reaching conclusions on the bioaccumulation of D4 and D5 which the evidence relied upon is incapable of supporting

61      In the fourth part of the first plea, the applicants, supported by ACC, claim that ECHA’s conclusions on the vB properties of D4 and D5 in the MSC’s opinion of 22 April 2015 and in the support document for D4 are the result of a manifestly incorrect assessment of the available data.

62      In the first place, the applicants argue that it is apparent from the available information that D4 does not biomagnify to increasing unpredictable concentrations in the food web and cannot therefore pose a threat to the main predators or to humans through biomagnification. According to the applicants, ECHA should have taken that into consideration in a weight-of-evidence assessment.

63      The applicants, supported by ACC, maintain that ECHA instead based its conclusions as regards the B properties on the bioconcentration factor (‘BCF’), disregarding other information and thus distorting the relationship between Sections 1 and 3.2 of Annex XIII. They claim that the BCF referred to in Section 1 of Annex XIII is merely a tool to protect against the risk resulting from biomagnification through the food chain.

64      In support of their reasoning, the applicants refer, first, to the conclusions reached in the MSC’s opinion of 22 April 2015 on the biomagnification of D4 and D5 based on a dietary test, in accordance with Guideline No 305 of the Organisation for Economic Cooperation and Development (OECD) Guidelines for the Testing of Chemicals (‘OECD Guideline No 305’). The MSC stated in that opinion that the results of such a test could not be directly compared with biomagnification under field conditions and that even if the biomagnification factor (‘BMF’) deriving from such a test was below 1, the BCF could still meet the B or vB criteria. The MSC claimed to have substantiated that conclusion by providing examples of other substances, without explaining why those examples were relevant for D4 and D5.

65      Second, the applicants assert that, in the RCOM document for D4, ECHA acknowledged that most field studies showed no biomagnification. However, that document stated that the lack of biomagnification could not counter the concerns caused by bioconcentration. Similarly, the MSC’s opinion of 22 April 2015 states that, according to Chapter R.11 on the assessment of PBT properties of November 2014 from ECHA’s ‘Guidance on Information Requirements and Chemical Safety Assessment’ (‘the ECHA Guidance’), the absence of such biomagnification should not be used to conclude that substances do not meet the B or vB criteria. The applicants claim in that regard that a non-binding ECHA guidance document cannot be of equal value to the mandatory criteria set out in Annex XIII, under which all information should be taken into account for the weight-of-evidence determination.

66      Third, the applicants refer to the study by Krogseth et al. (2017). That study, among others, shows that the substances at issue do not display trophic magnification. Metabolism and elimination of D4 by means of exhalation prevent the accumulation of D4 in terrestrial and other air-breathing organisms.

67      ECHA, supported by the Federal Republic of Germany, disputes those arguments.

68      As a preliminary point, it should be pointed out that biomagnification and bioconcentration are different phenomena, as ECHA, supported by the Federal Republic of Germany, rightly states. The legislature chose to lay down in Sections 1.1.2 and 1.2.2 of Annex XIII criteria for identifying B or vB substances in relation to their BCF in aquatic species. According to the second paragraph of the preamble to Annex XIII, the information set out in Section 3.2, including the results of a study on the bioconcentration or bioaccumulation of a substance in aquatic species, and the information on the ability of that substance to biomagnify in the food chain, where possible expressed by biomagnification factors or trophic magnification factors, are to be compared with the criteria set out in Section 1 of that annex on the basis of a weight-of-evidence determination using expert judgement.

69      According to the second paragraph of the preamble to Annex XIII, the weight-of-evidence determination is to be applied in particular where the criteria set out in Section 1 cannot be applied directly to the available information. A BCF resulting from a reliable study on the bioconcentration of a substance in aquatic species, such as that in the present case, may be directly – that is to say, numerically – compared with the criteria laid down in Sections 1.1.2 and 1.2.2 of Annex XIII. While recognising that this does not preclude the application of the weight-of-evidence determination in cases such as the present one, in which BCF data are available and relevant, it is clear that the information and data referred to in Section 3.2.2 of Annex XIII gain in importance in particular when the criterion defined by reference to the BCF cannot be applied directly to the information available.

70      Contrary to what is suggested by ACC, it is not apparent from recitals 5 and 6 of Regulation No 253/2011 that the weight-of-evidence determination is particularly relevant when the data show that a substance does not biomagnify to unpredictable concentrations in the food web. Recital 6 of that regulation, like the second paragraph of the preamble to Annex XIII, refers to the impossibility of directly applying the criteria in Section 1 to the information available. Where reliable BCF information is available, the criteria established by reference to the BCF may be applied to that information directly.

71      It follows from those provisions that the legislature chose to give priority to the results of reliable studies on the BCF of a substance in aquatic species or, at the very least, that the MSC, without committing a manifest error of assessment, considered that the BCF values had greater weight than the other data cited by the applicants.

72      The applicants have not demonstrated that ECHA’s assessment and conclusion on the vB properties of D4 and D5 in the present case are not in conformity with those provisions or are vitiated by manifest errors of assessment. Their arguments concerning the assessment of the information relating to biomagnification and trophic magnification do not take into account the importance expressly given by the legislature to the results of reliable studies on the BCF of a substance in aquatic species. Those arguments are thus based on a misinterpretation of the relevant provisions of Annex XIII.

73      Contrary to what the applicants claim, ECHA’s finding, as reflected in the RCOM document for D4, in the MSC’s opinion of 22 April 2015 and in the ECHA Guidance, that the absence of biomagnification did not refute the concern arising from bioconcentration, is consistent with the legal framework established by Annex XIII.

74      Moreover, it is apparent from the file, and more specifically from the support documents for D4 and D5 and the MSC’s opinion of 22 April 2015 to which those documents refer, that inter alia the data resulting from the field studies indicating that there was no trophic magnification in certain food webs were taken into consideration. The MSC however found that those data did not invalidate the other evidence supporting the conclusion on bioaccumulation. That position takes account of the fact that biomagnification and bioconcentration are two different pathways to bioaccumulation. Even if a substance does not bioaccumulate by means of biomagnification, it may nevertheless bioaccumulate by means of bioconcentration. Furthermore, the absence of biomagnification of a substance in a food chain does not prove that that substance does not biomagnify in other food chains.

75      For the same reasons, it must be held that the applicants have also failed to demonstrate that ECHA’s assessment of the Krogseth et al. (2017) study and other information indicating that there was no trophic magnification was manifestly incorrect.

76      As regards, more specifically, the results of the dietary test referred to by the applicants, the MSC explained in detail why it did not consider that BMFs below 1 resulting from certain studies demonstrated the absence of bioaccumulation. According to the MSC, the results of that test could not be compared directly with biomagnification under field conditions. In dietary tests carried out in accordance with the OECD Guideline No 305, fish are supposed to be placed in clean water, such that direct uptake from water is absent or negligible, whereas a BMF obtained in the field also implicitly includes the uptake of a substance directly from the water. Furthermore, results relating to other substances show that, even where BMFs obtained through dietary tests are less than 1, BCFs can still fulfil the criteria for B or vB substances.

77      That explanation appears plausible, particularly if one considers that biomagnification and bioconcentration are two different pathways to bioaccumulation, as ECHA rightly points out. Contrary to what the applicants claim, ECHA was not required to explain the reasons why the conclusions relating to other substances were relevant in the present case. The MSC referred to other substances only to show that it was not unusual for a substance with a BMF of less than 1 resulting from a dietary test to have a BCF fulfilling the criteria of B or vB substances.

78      Consequently, the applicants’ arguments relating to the nature of the BCF and its assessment in relation to the other data, such as the data concerning biomagnification and trophic magnification, must be rejected.

79      Next, the applicants claim that ECHA made a manifest error of assessment in rejecting the use of a fugacity approach, as is set out in the Mackay et al. (2015) study. That approach consists of ‘analysing monitoring data for representative food webs and [converting] the concentrations to fugacity or chemical activities to reveal their relative equilibrium status’. The applicants submit that ECHA should not have disregarded that approach on the ground that it ‘[had] not been accepted for regulatory decision-making worldwide and [had] not [been] validated for D4 and D5’. Indeed, Section 3.2.2(c) of Annex XIII recognises the relevance of bioaccumulation studies.

80      ECHA disputes those arguments.

81      It is apparent from the file and, in particular, from the RCOM documents that the fugacity approach, as is set out in the Mackay et al. (2015) study and in the Gobas et al. (2015b) study, which applied it to D5, was considered irrelevant to the assessment of bioaccumulation in essence for two reasons.

82      First, according to ECHA, that approach had not been sufficiently validated and there had been no agreement on the interpretation of fugacity ratios. That finding is not disputed by the applicants, who nevertheless claim that that should not have prevented ECHA from taking that information into consideration in the context of a weight-of-evidence determination. However, ECHA did not commit any manifest error of assessment in rejecting an approach which, in its view, had not been sufficiently validated and the interpretation of which was uncertain.

83      Second, in ECHA’s opinion, the proposed fugacity trigger of 1 was not equivalent to the bioaccumulation criterion provided for in Annex XIII and the results did not contradict the BCF values. Given that the fugacity approach is another approach which assesses biomagnification potential and that the Gobas et al. (2015b) study indicates that D5 undergoes metabolism which could prevent D5 from accumulating in a food chain, it must be held that, in the light of the broad discretion which it enjoys, ECHA’s assessment in that regard may likewise be considered not to be vitiated by a manifest error of assessment.

84      In addition, the applicants claim that ECHA relied on a misinterpretation of Annex I to Regulation No 1907/2006 (‘Annex I’), in so far as it stated the following in the RCOM documents for D4, D5 and D6:

‘uncertainties and policy choices are inherent to any quantitative risk assessment, but in the case of PBT/vPvB substances, uncertainty of exposure assessment and toxicity assessment is too high. Therefore, Section 4.0.1 of Annex I to [Regulation No 1907/2006] defines that for PBT/vPvB substances quantitative risk assessment cannot be carried out with sufficient reliability.’

85      In that regard, the applicants state that, according to Section 4.2 of Annex I, relating to the characterisation of exposure for the ‘PBT and vPvB assessment’, ‘an emission characterisation shall be conducted comprising the relevant parts of the exposure assessment as described in Section 5’ and that ‘in particular it shall contain an estimation of the amounts of the substance released’.

86      Thus, contrary to what ECHA considered, an exposure assessment may be carried out for PBT and vPvB substances. Moreover, ECHA cannot, according to the applicants, rely on rules applicable to substances whose PBT and vPvB properties have been established, in order to assess evidence with a view to deciding whether a substance is a PBT or vPvB substance.

87      ECHA disputes those arguments.

88      It is appropriate to reject those arguments of the applicants on the ground that Sections 4 and 6.5 of Annex I were not used as assessment criteria in the support documents, as ECHA correctly submits. The extracts from the RCOM documents cited by the applicants are responses to a comment about risk. However, the criteria set out in Annex XIII are based on the assessment of the hazards arising from the intrinsic properties of the substances and Article 57 of Regulation No 1907/2006 does not require, in respect of the substances concerned, an assessment of risk arising from the use of those substances to be carried out analogous to that required, in the context of the evaluation procedure, under Section 6 of Annex I to that regulation (see, to that effect, judgment of 15 March 2017, Polynt v ECHA, C‑323/15 P, EU:C:2017:207, paragraphs 24 and 25). Consequently, the exposure assessment under Annex I is not relevant in the present case.

89      When they claim that a risk assessment should be carried out by examining what actually happens in the environment when the BCF is at a level of concern, the applicants disregard the fact that the identification of substances as substances of very high concern forms part of a hazard assessment. According to the applicants, that risk assessment should examine whether the substances are present in water. If a substance were not present in water, it would not be available to bioconcentrate and could not, therefore, from their point of view, be considered bioaccumulative. Thus, according to the applicants, exposure helps to determine which conditions should be regarded as ‘relevant’ under Annex XIII.

90      That argument is, moreover, based on the applicants’ misinterpretation of Annex XIII. As is stated in paragraph 68 above, Annex XIII defines the criteria for identifying a substance’s bioaccumulation in relation to its BCF in aquatic species. There is no indication in that annex that the BCF is merely a ‘tool’ and that exposure and risks should be assessed in the context of an assessment of PBT and vPvB substances.

91      In the second place, the applicants, supported by ACC, add that a weight-of-evidence determination requires specific weight to be given to each piece of information, as is confirmed by Sections R.4.4 and R.11.4.1 of the ECHA Guidance. Yet ECHA failed to give specific quantitative or qualitative weight to each item of evidence, which constitutes an error of law and makes it impossible to examine the manner in which ECHA exercised its discretion to identify the substances at issue as PBT or vPvB substances.

92      ECHA disputes those arguments.

93      It must be stated that an obligation to assign numerical values to the various sources of data in the context of the weight-of-evidence determination is not provided for or indicated in either the second or third paragraphs of the preamble to Annex XIII, in any other provision of that annex, in the recitals of Regulation No 253/2011, or in the ECHA Guidance.

94      It is apparent in particular from the preamble to Annex XIII that ‘appropriate weight’ is to be given to the quality and consistency of the data, that the results are to be assembled together into a single weight-of-evidence determination and that it is necessary to compare the information listed in Section 3.2.2 with the criteria mentioned in Sections 1.1.2 and 1.2.2 of that annex. In order for ‘appropriate weight’ to be given to the quality and consistency of the data, it is not essential that an exact numerical value necessarily be given to each piece of evidence. It follows from the use of the words ‘by comparing’ that the information listed in Section 3.2.2 must be examined in the light of its similarities to and differences from the criteria. That does not necessarily imply a purely quantitative approach. Thus, the rules set out in Annex XIII allow the competent EU authorities to apply a quantitative or qualitative weight-of-evidence determination. The choice of the competent authorities between those options in a given case depends on the circumstances of the case, in particular the nature of the available information to be compared. For example, as Section R.11.4.1 of the ECHA Guidance states, the derivation of a conclusion property by property needs expert judgement, especially when very different types of information are available and when the information cannot be directly (numerically) compared with the criteria.

95      That finding, which is apparent from Annex XIII and confirmed by the chapter of the guidance relevant to the assessment of PBT and vPvB substances, is not called into question by the argument of ACC and of the applicants based on Section R.4.4 of the ECHA Guidance of December 2011 entitled ‘Evaluation and integration of all available information including weight of evidence’, according to which weight should be assigned to each piece of information and those weights or values may be assigned either in an objective way by using a formalised procedure or by using expert judgement. ECHA has indicated that that section of the guidance referred to by ACC is not intended to be used for the weight-of-evidence determination in accordance with Annex XIII. In any event, it is less specific and does not explicitly describe the situation provided for in Annex XIII, in which it may prove necessary to compare very different types of information.

96      Similarly, the applicants’ reference, in the fifth part of the first plea which also concerns the requirements of the weight-of-evidence determination, to Section 1.1.1.3 of Annex I to Regulation (EC) No 1272/2008 of the European Parliament and of the Council of 16 December 2008 on the classification, labelling and packaging of substances and mixtures, amending and repealing Directives 67/548/EEC and 1999/45/EC and amending Regulation No 1907/2006 (OJ 2008 L 353, p. 1), is irrelevant. That section applies to the hazard assessment for the classification of dangerous substances and mixtures in hazard classes and there is no indication that the rules laid down therein should also apply, mutatis mutandis, to the identification of PBT and vPvB substances in accordance with Annex XIII. The applicants have not put forward any argument to justify such application.

97      Consequently, contrary to what is argued by ACC, the weight-of-evidence determination based on expert judgement as provided for in Annex XIII does not require a quantitative weight to be given to each piece of evidence.

98      Nevertheless, it follows from the third paragraph of the preamble to Annex XIII that appropriate weight must be given to the quality and consistency of the data which must be taken into consideration in a single weight-of-evidence determination. In the absence of a specific example provided in the present part of the first plea, for which ECHA did not comply with those requirements in the present case, and in the light of what has been stated in paragraph 51 et seq. above, it must be held that the arguments raised by the applicants in the present part of the first plea are not capable of demonstrating that ECHA committed a manifest error in its assessment of D4 and D5 and in concluding that those two substances were vB substances.

 Second part of the first plea: commission by ECHA of a manifest error of assessment in concluding that D4, D5 and D6 met the PBT and vPvB criteria set out in Annex XIII even though persistence (P property) and bioaccumulation (B property) were not established for the same compartment

99      The second part of the first plea comprises, in essence, three complaints.

100    In the first complaint, the applicants claim that ECHA committed a manifest error of assessment in concluding that D4, D5 and D6 met the criteria for PBT and vPvB substances set out in Annex XIII even though the P and B properties were not established for the same compartment. ECHA concluded that those three substances were vP substances in sediment, but not in water, and that they were vB substances based on bioconcentration in fish from water, but not in sediment.

101    The criteria set out in Annex XIII should, according to the applicants, be interpreted as requiring that the P and B properties manifest themselves in the same compartment. It is apparent from Annex XIII that ‘information used for the purposes of assessment of the PBT/vPvB properties shall be based on data obtained under relevant conditions’, which refers to the relevant compartment. The principle that the P and B properties must manifest themselves in the same compartment is also established in Section 4.1 of Annex I to Regulation No 1272/2008. The Court has already recognised in its judgment of 7 March 2013, Rütgers Germany and Others v ECHA (T‑96/10, EU:T:2013:109, paragraph 110), that the principles established under Regulation No 1272/2008 may be relevant for assessments carried out under Regulation No 1907/2006.

102    In the second complaint, first, the applicants claim that the presence of D4, D5 and D6 in the environment is much shorter than that of the ‘classical vP chemicals’ and would be easily reversible if emissions were to cease, as indicated in the study by Kim et al. (2018). On account of their unique properties mainly due to their inorganic backbone chain of Si-O-Si units, their high air/water partitioning constant and relatively low octanol (lipid)/air partitioning constant, D4, D5 and D6 readily partition from other compartments to air where they degrade, which has a considerable impact on sediment concentrations in the environment. ECHA did not consider multimedia modelling, which would have allowed incorporation of all those factors in order to understand the reversibility of the substances at issue.

103    Second, as regards the environmental fate of the substances at issue, the applicants refer to the Quantitative Water Air Sediment Interaction fugacity model (‘the QWASI model’) and to several environmental multimedia models which use the concept of fugacity as a criterion to assess thermodynamic equilibrium, which provide mathematical benefits to describe chemical equilibrium, partitioning and transportation of substances between environmental compartments. Since this is a standard and realistic scientific practice, the applicants maintain that ECHA should have taken account of the information on the volatility of the substances at issue as part of a weight-of-evidence approach, instead of dismissing such information.

104    Third, the applicants state that D4, D5 and D6 are not found in water. The study by Kozerski and Durham (2006) and that by Xu and Kropscott (2012) demonstrate that the substances at issue undergo hydrolysis or volatilise in water. The study by Mackay et al. (2015) and that by Wang et al. (2013a) showed that D4, D5 and D6 are not released from sediment into water above the level of solubility. The Krogseth et al. (2017) study found that the lake-water concentrations of those substances were below the limits of quantification, even when the sediment was considered to be highly contaminated.

105    In the third complaint, raised at the stage of the reply, the applicants claim that ECHA based its identification of D4 as having T properties on effects observed in unrealistic laboratory tests, since monitoring data demonstrates that D4 is not detected in surface waters.

106    ECHA, supported by the Federal Republic of Germany and the Commission, disputes those arguments.

107    As regards the first complaint, first, it should be noted that only the criteria for the identification of the P property, set out in Sections 1.1.1 and 1.2.1 of Annex XIII, expressly refer to specific compartments and define different and alternative criteria for different compartments. Sections 1.1.2, 1.1.3 and 1.2.2 of Annex XIII, which define the criteria for the identification of substances with B or T properties, make no express reference to a specific compartment and, above all, do not establish a link between such a compartment and the compartment in which the P property has been identified.

108    Second, as ECHA and the Commission rightly point out, the interpretation relied on by the applicants does not take account of transfer between compartments with the result that a substance which is persistent in one compartment, for example sediment, may be transferred to another compartment, for example sea water, where it may bioaccumulate or have toxic effects.

109    Third, according to the second paragraph of the preamble to Annex XIII, the weight-of-evidence determination using expert judgement allows a comparison of ‘all relevant and available information referred to in Section 3.2 [of that annex] with the criteria set out in Section 1’. Clearly, information based on data obtained in different compartments or which concerns animals living in different compartments, such as aquatic species and terrestrial species, is thus covered. Similarly, Section 3 of Annex XIII refers to factors which in themselves may concern several compartments, such as the biomagnification factor.

110    Consequently, it must be held that Annex XIII does not require that all the criteria be met in one and the same compartment.

111    That finding is not called into question by the applicants’ argument based on Section 4.1 of Annex I to Regulation No 1272/2008.

112    First, Annex XIII does not establish a direct link with Section 4.1 of Annex I to that regulation in the sense that the provisions of that section or the alleged principles which are apparent from them are applicable to the identification of PBT and vPvB substances in accordance with Annex XIII.

113    Second, as the Commission rightly submits, that section of Annex I to Regulation No 1272/2008 deals with substances that are ‘hazardous to the aquatic environment’. No general principle applicable to the assessment of PBT and vPvB substances under Annex XIII can be inferred from that very specific section which concerns substances hazardous to the aquatic environment. Rather, Annex XIII sets out rules which are not limited to the aquatic environment but also concern other environments.

114    Thus, the circumstances of the present case differ from those which gave rise to the judgment of 7 March 2013, Rütgers Germany and Others v ECHA (T‑96/10, EU:T:2013:109). In that case, the Court held that it was apparent from Article 10(1) of Regulation No 1272/2008 that the legislature had recognised the principle that the fact that a substance with certain properties was present in another substance could ground the conclusion that the latter substance possessed those same properties. The Court also held that several factors in Regulation No 1907/2006 confirmed that that principle was applicable to the procedure for the identification of a substance as a substance of very high concern. In the case at hand, however, the applicants have not invoked any such factors. On the contrary, the applicants’ line of argument, were it to be accepted, would have the effect of limiting the scope of the analysis, which is not expressly provided for in Annex XIII.

115    Accordingly, the first complaint of the second part of the first plea must be rejected as unfounded.

116    As regards the second complaint, it must be recalled, once again, that it is apparent from the case-law, and in particular from the judgment of 15 March 2017, Polynt v ECHA, C‑323/15 P (EU:C:2017:207, paragraphs 24 and 25), that the identification of a substance as a substance of very high concern under Articles 57 and 59 of Regulation No 1907/2006 is based on an assessment of the hazards arising from the intrinsic properties of that substance. A risk assessment is therefore not necessary for that purpose. Accordingly, the applicants’ arguments that the substances at issue are not found in water and are not released from sediment into water above the level of solubility are ineffective for the purpose of determining whether a substance, on account of its intrinsic properties, is a PBT or vPvB substance in accordance with the criteria set out in Annex XIII.

117    In any event, the applicants’ assertion that D4, D5 and D6 are not found in water is contradicted by the elements in the file. The support documents contain numerous data on the concentrations of D4, D5 and D6 in organisms in the field, in particular aquatic organisms. The applicants do not explain how those substances can be present in those organisms if, as they claim, those substances should not be considered present in water. Nor have the applicants provided specific evidence suggesting that the measured data on the presence of D4 and D5 in the environment, as are set out in the file proposing the restriction of those two substances, were incorrect.

118    In addition, even if it were true that the substances at issue are not released from sediment into water above the level of solubility, as the applicants claim, that does not exclude that there may be bioaccumulation, given that bioaccumulation could also occur directly from sediment through organisms living in sediment, as ECHA rightly indicates. Similarly, low water solubility does not mean that a substance is not at all present in water or that there is no bioconcentration in aquatic species.

119    Consequently, the applicants’ arguments that any bioconcentration is excluded by the fact that the substances at issue are not found in water must be rejected, as must their argument that, for those reasons, no ‘hazard [can] be demonstrated “under relevant conditions”’.

120    In addition, it is apparent from the file, and in particular from the Annex XV dossiers and the support documents, that ECHA took into account, when identifying D4, D5 and D6 as substances of very high concern, the volatility of those three substances and the Kim et al. (2018) study.

121    As regards the latter study in particular, it is apparent from the support documents that ECHA acknowledged that that modelling studies demonstrated a rapid initial reduction in concentrations in the event of the cessation of emissions caused by the degradation of airborne siloxanes. However, those models also showed that concentrations in sediment persisted and that degradation half-lives in sediment were higher than what is provided for by the criteria set out in Annex XIII. The applicants merely claim that the Kim et al. (2018) study demonstrates that the reversibility of the substances at issue is easier than that of PBT and vPvB substances. However, they do not dispute that the degradation half-lives in sediment identified were higher than what is provided for by the criteria set out in Annex XIII.

122    In that regard, it should also be noted that Annex XIII lays down the criteria for P and vP properties in relation to the degradation half-life in a specific compartment. Furthermore, that annex requires neither an overall determination of the ‘difficult or impossible reversibility in the event of emission ceasing’ nor that the persistence or bioaccumulation of a substance be assessed by taking into account its behaviour in a number of compartments. Nor does it lay down specific rules for volatile substances.

123    The applicants have also failed to demonstrate in their argument relating to the fugacity and QWASI models that ECHA should have assessed the volatility of substances differently or that it should have used such a model. As ECHA has rightly pointed out, it is apparent from Section 4.0.1 of Annex I that the estimation of long-term exposure of humans and the environment cannot be carried out with sufficient reliability for PBT and vPvB substances. Thus, quantitative fate models, in the context of exposure estimation, do not provide sufficiently reliable results for PBT and vPvB substances. ECHA also states that the fugacity and QWASI models were not accepted for regulatory purposes for the forecast of intrinsic properties such as bioaccumulation potential. The applicants have not put forward a single argument capable of dispelling those reliability concerns.

124    In any event, in their replies to the Court’s questions, the applicants stated that the QWASI model demonstrated that the substances at issue would be found in sediment, but not in water and, consequently, would not be available in water for bioconcentration. As has already been stated in paragraph 117 et seq., first, that argument is contradicted by numerous elements in the file and, second, it is irrelevant because bioaccumulation can also occur directly from sediment through organisms living in sediment.

125    It follows from the foregoing and from what has been stated above in paragraph 51 et seq. that the applicants’ line of argument is not capable of demonstrating that ECHA made a manifest error in its assessment of those data.

126    The applicants’ claim that, if the substances were taken up by fish, they would metabolise and would therefore not bioaccumulate is also incapable, in the absence of sufficiently concrete and tangible evidence, of demonstrating a manifest error of assessment on the part of ECHA. That is also true of their argument that the Selck et al. (2018) study indicates that sediment-dwelling organisms would metabolise the substances without showing any toxicity. Contrary to the tests referred to in the Selck et al. (2018) study, tests generally carried out in laboratories on samples of sediment do not involve those organisms.

127    First, ECHA rightly points out that, if the substances at issue were metabolised by fish, as the applicants assert without further substantiation, the BCF values would be low, which is not the case. Even if such metabolism did occur, it would not be sufficiently strong to exclude bioaccumulation by bioconcentration.

128    Second, it is apparent from the file that the Selck et al. (2018) study was taken into account in the procedure which led to the adoption of the contested decision. The RCOM documents state, for example, that the available results of the degradation tests took into account the potential effects of metabolism by sediment-dwelling organisms relevant to the degradation half-life, and that the Selck et al. (2018) study provided no evidence that metabolism had a significant effect on the degradation half-life in sediment. In addition, ECHA observes that the MSC, at its meeting which took place from 12 to 14 June 2018, examined the Selck et al. (2018) study, a point which is not disputed by the applicants. The MSC considered that that study confirmed the potential absorption of the substances at issue by sediment through which they entered the food chain.

129    The applicants have not demonstrated how or to what extent that assessment by ECHA is manifestly incorrect. In particular, they have not argued that metabolism has such an impact on the half-life of substances in sediment that D4, D5 and D6 no longer satisfied the criteria for vP properties set out in Annex XIII. The applicants merely repeat that there is metabolism by organisms living in sediment. In their replies to the Court’s questions, they state that the Selck et al. (2018) study showed that there was no risk of bioaccumulation in sediment and no bioaccumulation at the primary level of the food chain. Even if that were the case for the organisms examined as part of the Selck et al. (2018) study, that argument is not capable of dispelling the concerns relating to bioconcentration resulting from the high BCF values. Bioconcentration directly from water does not require the substances to be found in organisms living in sediment.

130    Third, as is described in paragraph 73 et seq. above, ECHA committed no manifest error in finding that the absence of biomagnification in a food chain did not contradict the high BCF values. Similarly, ECHA’s finding in the support document for D5 that a high bioaccumulation in only part of the food chain is of concern is not vitiated by a manifest error of assessment. That concern is not dispelled by the existence of metabolism in certain organisms, assuming it were established.

131    The second complaint of the second part of the first plea must therefore be rejected as unfounded.

132    Lastly, without it being necessary to determine whether the third complaint, raised only at the stage of the reply, is out of time and therefore inadmissible under Article 84(1) of the Rules of Procedure, it is also necessary to reject it given that the applicants merely assert that the identification of D4 as a T substance was based on unrealistic laboratory tests since D4 is not detected in surface water. As has been established in paragraph 117 above, that assertion is contradicted by the elements in the file.

133    In any event, and as has already been pointed out in paragraph 116 et seq. above, that argument of the applicants is not relevant for the purpose of determining whether a substance, because of its intrinsic properties, is a PBT or vPvB substance in accordance with the criteria set out in Annex XIII, given that the identification of a substance as a substance of very high concern under Articles 57 and 59 of Regulation No 1907/2006 is to be based on an assessment of the hazards arising from its intrinsic properties and that a risk assessment is not necessary for that purpose.

134    Accordingly, the second part of the first plea must be rejected in its entirety.

 Third part of the first plea: commission by ECHA of a manifest error of assessment in not taking into consideration the specific (hybrid) nature of D4, D5 and D6

135    In the third part of the first plea, the applicants claim that ECHA committed a manifest error of assessment in not taking into consideration the hybrid nature of D4, D5 and D6. First, those substances possess an inorganic chain of silicon and oxygen atoms and siloxanes are not organic substances or organo-metals. The applicants note that the International Union of Pure and Applied Chemistry (IUPAC)’s Compendium of Chemical Terminology (‘the Gold Book’) defines ‘inorganic polymer’ as a polymer or polymer network with a skeletal structure that does not include carbon atoms and lists polysiloxanes and polysilanes as examples of inorganic polymers.

136    The applicants claim that ECHA dismissed any discussion on that matter even though the hybrid nature of D4, D5 and D6 was relevant information to be taken into consideration in ‘how the Annex XIII criteria should be applied’, given that Annex XIII, according to its preamble, applies ‘to all organic substances, including organo-metals’.

137    Second, siloxanes such as D4, D5 and D6 are different from organic substances as regards their solubility and partitioning properties which influence their distribution and fate in the environment. According to the applicants, when the substances in question are released into the air they remain in the air without being transferred to other compartments. When those substances are released into other compartments, they volatilise into air where they are broken down. Owing to their appreciable Soil Organic Carbon-Water Partitioning Coefficient, the substances in question bind to organic carbon and hydrolyse in water. For this reason, D4, D5, and D6 are not found in water. Moreover, on account of the high octanol-water partitioning coefficient of those substances, the dietary exposure via the food chain should have been taken into consideration in order to assess the bioaccumulation potential of those substances.

138    Third, D4, D5 and D6 biodilute and undergo metabolism in certain benthic, pelagic and terrestrial species. Consequently, they do not build up food webs and do not biomagnify.

139    ECHA, supported by the Federal Republic of Germany, disputes those arguments. It took into account all the relevant properties of D4, D5 and D6. Their ‘hybrid’ nature is not a relevant factor under Annex XIII.

140    In that regard, it should be noted that the sixth paragraph of the preamble to Annex XIII states that that annex ‘shall apply to all organic substances, including organo-metals’. Unlike its earlier version, the version of Annex XIII applicable in the present case no longer contains any explicit reference to its non-application to inorganic substances.

141    Consequently, even a substance with a ‘hybrid’ structure, namely ‘neither fully inorganic nor fully organic’, which, according to the applicants, is the case with D4, D5 and D6, is not necessarily excluded from the scope of that annex.

142    Furthermore, it should be noted that the applicants do not dispute ECHA’s argument that siloxanes, such as D4, D5 and D6, are listed in the volume published by IUPAC on organic substances (‘the Blue Book’) and not in the volume on inorganic substances (‘the Red Book’). Nor do they dispute ECHA’s argument that, apart from a few specifically defined exceptions, all chemical compounds which contain carbon, like D4, D5 and D6, constitute ‘organic chemicals’.

143    As a result, and considering that the Blue Book and the Red Book constitute the standard reference in chemistry and that IUPAC’s task is, in particular, to establish a generally recognised nomenclature and terminology, it must be held that the classification of D4, D5 and D6 as organic substances is consistent with internationally recognised terminology.

144    None of the arguments raised by the applicants is capable of demonstrating that the application of Annex XIII to those substances is vitiated by a manifest error of assessment despite such conformity with internationally recognised terminology.

145    First, it is necessary to examine the applicants’ arguments alleging that the Gold Book considers polysiloxanes and polysilanes to be examples of inorganic polymers. The applicants, in their observations on the statement in intervention submitted by the Federal Republic of Germany, add, making reference to the Gold Book, that siloxanes were considered in the IUPAC nomenclature to be organic substances in the 1990s, but that, since 2007, IUPAC has identified polysiloxanes as examples of inorganic polymers.

146    As the Federal Republic of Germany submits in its replies to the Court’s questions, D4, D5 and D6 are not polymers according to the definition in Article 3(5) of Regulation No 1907/2006 or the IUPAC nomenclature. Furthermore, those substances are registered by the registrants as monomers. In their observations on the Federal Republic of Germany’s replies, the applicants acknowledged that the substances at issue are not polymers.

147    Consequently, it must be found that, according to that nomenclature, the substances at issue are not polysiloxanes. Similarly, the fact that polysiloxanes are inorganic polymers does not mean that D4, D5 and D6 should have been considered to be hybrid, or even inorganic, substances.

148    That finding is not called into question by the applicants’ argument, in their observations on the replies of the Federal Republic of Germany, that the substances at issue are siloxanes and that, because they have ‘more than one Si-O union’ and because ‘poly’ means ‘more than one’, they are, therefore, polysiloxanes. That assertion is not supported by any evidence, in particular by any internationally recognised source, and contradicts the Gold Book entry relating to inorganic polymers.

149    Consequently, the applicants’ argument relating to polysiloxanes must be rejected as irrelevant.

150    Given that, as has been found above, ECHA did not make a manifest error of assessment in considering D4, D5 and D6 to be organic substances, it must be held that, consequently, Annex XIII was in fact applicable in the present case, as is provided for in the sixth paragraph of the preamble to that annex. That annex does not provide for any exception or specific rules for substances possessing properties such as those referred to by the applicants.

151    Nevertheless, according to the second paragraph of the preamble to Annex XIII, ‘all relevant and available information listed in Section 3.2’ is to be compared with the criteria set out in Section 1 of that annex. According to the third paragraph of that preamble, ‘all available information bearing on the identification of a PBT or a vPvB substance [must be] considered together’ in the weight-of-evidence determination. Certain specific properties of a substance may, in principle, constitute such information. However, in the present case, the applicants do not put forward any evidence capable of demonstrating that ECHA failed to take such information into account and that such an alleged omission would have had a bearing on the identification of D4, D5 and D6 as vB substances.

152    Second, as regards the applicants’ arguments that D4, D5 and D6 are not found in the environment and their arguments relating to the behaviour of those substances in the environment and between different compartments in particular, it must be noted that they have already been examined and rejected in the second part of the first plea, reference to which is therefore made.

153    Third, as regards the applicants’ assertion that the substances at issue undergo metabolism in benthic, pelagic and terrestrial species, it should be noted that that assertion, in so far as it refers to metabolism in fish and in organisms living in sediment, has also been examined and rejected in the second part of the first plea, reference to which is therefore made.

154    In so far as that assertion by the applicants concerns the metabolism of the substances at issue in other organisms, it must be stated that that is a general assertion which is in no way substantiated. In that regard, ECHA argues that it is impossible to make such a claim on the basis of the facts available. In such a situation, it must be concluded that the applicants have not demonstrated that ECHA’s assessment in that regard was manifestly erroneous.

155    In any event, even if the substances at issue were fully metabolised in certain organisms, thereby excluding, according to the applicants, any biomagnification and any trophic magnification in certain food chains or in certain parts of certain food chains, that would not call into question ECHA’s conclusion concerning the vB properties of D4, D5 and D6 based on high BCF values and supported by multiple sources of data on biomagnification drawn from dietary studies. The arguments raised by the applicants in that regard appear to be based on their misinterpretation of Annex XIII and in particular of the relationship between the criteria laid down in Sections 1.1.2 and 1.2.2 of that annex in relation to the BCF, on the one hand, and the data on biomagnification and trophic magnification to which Section 3.2.2 refers inter alia, on the other. In that regard, reference is made to the considerations set out in the fourth part of the first plea.

156    As regards, moreover, the applicants’ line of argument according to which dietary exposure via the food chain should have been taken into consideration in order to assess the bioaccumulation potential of the substances at issue on account of their high octanol-water partition coefficients, it must be stated that it is apparent from the file, and in particular from the support documents, that ECHA did take into consideration dietary exposure via the food chain. As regards ECHA’s assessment of those data and its comparison of those data with the bioconcentration data and the fact that the applicants have failed to demonstrate the existence of a manifest error of assessment in that regard, reference is made once again to the considerations set out in the fourth part of the first plea.

157    Accordingly, the third part of the first plea must also be rejected as unfounded.

 Fifth part of the first plea: commission by ECHA of a manifest error of assessment in failing to assess the new evidence on bioaccumulation (B and vB properties) for D4 and D5 available to ECHA after the MSC and the RAC’s opinions had been issued

158    In the fifth part of the first plea, the applicants, supported by ACC, claim that ECHA failed to include new data in an overall weight-of-evidence assessment of the data, as is required by Annex XIII and by Section 1.1.1.3 of Annex I to Regulation No 1272/2008.

159    ECHA presented several studies in the support documents for D4 and D5 as new studies or as studies published after the MSC’s opinion had been accepted.

160    With regard to D4, those studies comprise the following: Domoradzki et al. (2017); Bridges and Solomon (2016); Jia et al. (2015); Gobas et al. (2015); Powell et al. (2017); Powell et al. (2018); Woodburn K. (2017); Ruus et al. (2016); Ruus et al. (2017); Krogseth (2014); Krogseth et al. (2017a); Krogseth et al. (2017b); Sanchis et al. (2015b); Wang et al. (2017); McGoldrick et al. (2014); Sanchis et al. (2016); Huber et al. (2015); Lu et al. (2017); Lucia et al. (2016), and Fjeld et al. (2016).

161    With regard to D5, those studies comprise the following: Domoradzki et al. (2017); Bridges and Solomon (2016); Jia et al. (2015); Gobas et al. (2015a); Gobas et al. (2015b); Powell et al. (2017); Powell et al. (2018); Woodburn K. (2017); Ruus et al. (2016); Ruus et al. (2017); Krogseth (2014); Krogseth et al. (2017a); Krogseth et al. (2017b); Sanchis et al. (2015b); Wang et al. (2017); McGoldrick et al. (2014); Sanchis et al. (2016); Huber et al. (2015); Lu et al. (2017); Lucia et al. (2016), and Fjeld et al. (2016).

162    Those studies also included certain studies which, according to ECHA, had already been assessed in the context of the MSC’s opinion of 22 April 2015. The applicants claim that the findings of the Bridges and Solomon (2016) study were not assessed and that the Jia et al. (2015) study was mentioned but not assessed.

163    The applicants also claim that ECHA merely referred to those new studies without assessing them in the context of a weight-of-evidence determination, despite that being required by Annex XIII.

164    ACC adds that a weight-of-evidence determination requires that specific weight be given to each piece of information, as is confirmed by Sections R.4.4 and R.11.4.1 of the ECHA Guidance. However, ECHA failed to assign specific, quantitative or qualitative weight to each item of evidence, which constitutes an error of law and makes it impossible to assess the manner in which ECHA exercised its discretion in identifying the substances at issue as PBT or vPvB substances.

165    ECHA, supported by the Federal Republic of Germany, disputes those arguments.

166    As a preliminary point, it should be borne in mind that, as has been pointed out in the fourth part of the first plea, the weight-of-evidence determination using expert judgement under Annex XIII does not require that each item of evidence be given a quantitative weight.

167    Nevertheless, it follows from the third paragraph of the preamble to Annex XIII that appropriate weight must be given to the quality and consistency of the data and that those data must be taken into consideration in a single weight-of-evidence determination.

168    Consequently, it is necessary to examine, in the light of the line of argument raised by the applicants in the present part of the first plea, whether ECHA, having regard to those requirements, committed a manifest error in its assessment, in particular in the weighting of the new studies.

169    It should be noted in that regard that it is apparent from the file, and in particular from the support documents for D4 and D5, that all the relevant studies to which the applicants refer were taken into consideration. Those support documents indicate the information which affected the overall conclusion on bioaccumulation and the impact they had. It is also apparent from the file, and in particular from the support documents for D4 and D5 and from the RCOM documents, that ECHA did explain why some information had not been considered relevant or had to be assessed with caution.

170    The following studies are inter alia dealt with in the support document for D4: Domoradzki et al. (2017) (p. 17); Bridges and Solomon (2016) (p. 17); Jia et al. (2015) (p. 18); Gobas et al. (2015) (p. 18); Powell et al. (2017) (p. 18); Powell et al. (2018) (p. 18); Woodburn K. (2017) (pp. 18 and 19); Ruus et al. (2016) (p. 19); Ruus et al. (2017) (p. 19); Krogseth (2014) (p. 19); Krogseth et al. (2017a) (p. 19); Krogseth et al. (2017b) (p. 19); Sanchis et al. (2015b) (p. 19); Wang et al. (2017) (p. 20); McGoldrick et al. (2014) (p. 20); Sanchis et al. (2016) (p. 20); Huber et al. (2015) (p. 20); Lu et al. (2017) (p. 20); Lucia et al. (2016) (p. 20), and Fjeld et al. (2016) (p. 20).

171    The following studies are inter alia dealt with in the support document for D5: Domoradzki et al. (2017) (p. 17); Gobas et al. (2015a) (pp. 17, 18 and 19); Bridges and Solomon (2016) (pp. 17 and 18); Gobas et al. (2015b) (p. 18); Jia et al. (2015) (pp. 18 and 19); Powell et al. (2017) (p. 19); Powell et al. (2018) (p. 19); Woodburn K. (2017) (p. 19); Ruus et al. (2016) (p. 19); Ruus et al. (2017) (p. 19); Krogseth (2014) (pp. 19 and 20); Krogseth et al. (2017a) (pp. 19 and 20); Krogseth et al. (2017b) (pp. 19 and 20); Sanchis et al. (2015b) (p. 20); Fjeld et al. (2016) (p. 20); Wang et al. (2017) (p. 20); McGoldrick et al. (2014) (p. 20); Sanchis et al. (2016) (p. 20); Huber et al. (2015) (p. 21); Lu et al. (2017) (p. 21), and Lucia et al. (2016) (p. 21).

172    In addition, the RCOM documents for D4 and D5 deal, for example, deal with the approach followed in the Bridges and Solomon study (response to comment No 5119 on D4 and response to comment No 5122 on D5), the Gobas et al. (2015a) study (response to comment No 5119 on D4 and response to comment No 5122 on D5), the Gobas et al. (2015b) study (response to comment No 5119 on D4 and response to comment No 5122 on D5), the Powell et al. (2017) study (p. 16 of each RCOM document), the Powell et al. (2018) study (p. 16 of each RCOM document) and the Krogseth et al. (2017) study (p. 16 of each RCOM document).

173    As a result, the general assertion of the applicants, supported in that regard by ACC, that the quantitative or qualitative weight accorded to each study is unclear, cannot be accepted.

174    That finding is not called into question by the applicants’ argument, supported by ACC, in the reply, that the mere citation of the studies is not sufficient to demonstrate that ECHA took new information into account. According to the applicants, there should have been a process of assessment of the merits and demerits of each piece of information as well as a process for assessing all the information as part of a weight-of-evidence approach. The overall approach adopted by ECHA makes it almost entirely impossible to determine whether an element of information was actually subject to the weight-of-evidence assessment. In assessing only whether the new data differed from the earlier information, ECHA failed to examine critically whether the weight of the new information led to a different overall weighting of the available evidence.

175    It has already been noted in the first part of the first plea that the mere fact that ECHA has used results obtained from previous work, for example by referring to or citing the MSC’s opinion of 2015, does not mean that the contested decision is vitiated by a manifest error of assessment. The applicants seem to infer that there were manifest errors of assessment as regards the format chosen to present the assessment data in the procedure which led to the adoption of the contested decision. Yet Annex XIII does not lay down specific requirements relating to presentation as are advocated by the applicants.

176    The applicants do not state how, or why, the conclusion on vB properties and, consequently, the decision on the identification of D4 and D5 as substances of very high concern would have been different had the assessment been presented in a different manner.

177    Similarly, the manner in which the data and the assessment were presented in the present case is not in itself capable of demonstrating that the new data and their impact on the conclusion were not examined ‘critically’.

178    Furthermore, the applicants cite no specific examples of data for which it would have been impossible or indeed ‘near impossible’ for them to know the manner in which they had been assessed or the extent to which those data had had an impact on the conclusion relating to the vB properties of D4 and D5.

179    It is also necessary to examine whether the other arguments raised by the applicants demonstrate that a manifest error of assessment was committed by ECHA in its assessment of the studies in question.

180    In the first place, the applicants provide a number of examples allegedly demonstrating that ECHA did not carefully examine new information relating to certain field studies on trophic magnification and biodilution, for example the research project of Krogseth and co-workers on the fate of cyclic volatile methyl siloxanes in Lake Storvannet in Hammerfest (Norway), and the publications of Powell et al. (2017) and Powell et al. (2018) with regard to Tokyo Bay and Oslofjord, respectively. The applicants claim that ECHA dismissed those studies on a generic basis or examined them superficially. They add that ECHA did not mention that the study relating to Lake Storvannet supported the biodilution of the three substances at issue.

181    Moreover, they claim that ECHA’s assertion that it had already assessed the latter two studies in the MSC’s opinion of 2015 is unclear. Assessing the drafts of those studies – which were the only versions available at the time and had not been peer reviewed – is not equivalent to assessing those studies.

182    In addition, they claim that ECHA has likewise not demonstrated that it assessed why some studies indicating a trophic magnification factor (‘TMF’) value of greater than 1 were followed by a number of new peer-reviewed studies that arrived at a TMF value of less than 1. They claim that ECHA also committed an error of assessment in not giving greater weight to the studies examining the merits and demerits of past study design. Similarly, ECHA ruled out the possibility of more weight being given to new data.

183    Lastly, the applicants reiterate certain arguments raised in the fourth part of this plea. The MSC’s assertion in its opinion of 22 April 2015, according to which the absence of biomagnification should not be used to conclude that a substance does not fulfil the criteria for vB properties, is not consistent with Annex XIII which, in Section 3.2.2(c) thereof, recognises the importance of such data. The fact that ECHA based its conclusion as to the B or vB properties of the substances at issue on their high BCF values ‘supported by multiple lines of evidence on biomagnification in dietary studies’ constitutes an arbitrary selection of data on the part of ECHA.

184    ECHA disputes those arguments.

185    In that regard, it should be stated that it is apparent from the support documents for D4 and D5 that ECHA did examine the Krogseth (2014) study, the Krogseth et al. (2017a) study and the Krogseth et al. (2017b) study. It is also apparent from the support documents that ECHA examined the results of the two Powell et al. (2017 and 2018) studies, which had been submitted before publication and which therefore were available to the competent authority of the United Kingdom before the MSC submitted its opinion on 22 April 2015 and which are described in the annexes to that opinion. The Krogseth (2014) study also appears in the RCOM documents.

186    It is also apparent from the support documents what information from those studies was taken into account, how that information on trophic magnification was considered in relation to existing data and the way in which they were taken into account in reaching the conclusion on bioaccumulation. The support documents for D4 and D5 after all indicate that none of the new data on trophic magnification differs from the diverse data that had been available before the MSC’s opinion.

187    The applicants have not put forward any evidence suggesting that the data in the studies in question were interpreted in a manifestly erroneous manner or that that conclusion on trophic magnification was manifestly incorrect. More particularly, their argument that the support documents do not indicate that the Krogseth et al. studies supported the existence of biodilution must be rejected. Biodilution is the reverse phenomenon of trophic magnification and the support documents for D4 and D5 indicate that those studies do not demonstrate trophic magnification.

188    It is also necessary to reject the argument that ECHA should have assessed, in its weighting of the studies, the fact that certain studies had been peer reviewed. In particular, the applicants have not shown that the peer review of the preliminary drafts of the studies, such as those of Powell et al., between the MSC’s opinion of 22 April 2015 and the identification of the substances at issue as PBT and vPvB substances, resulted in relevant amendments to those studies, or that ECHA did not take those amendments into account or took them into account incorrectly. The mere fact that a study is peer reviewed is not a decisive factor in a weight-of-evidence determination using expert judgement.

189    Similarly, it is necessary to reject the applicants’ arguments that ECHA should have given greater weight to new studies examining the merits and demerits of past study design and that ECHA ruled out the possibility of a new study being given more weight than an earlier study.

190    ECHA has rightly argued that a new study is not given greater weight than a previous study solely because it has been published subsequently. Once again, the applicants merely make a general assertion without stating how the new studies indicated new evidence which ECHA assessed in a manifestly incorrect manner.

191    In any event, the applicants’ argument is irrelevant. The applicants have not stated, for example, how a new study indicating the existence of biodilution in certain food networks could have called into question the conclusion on the bioaccumulation of the substances at issue merely because that study is more recent or has, in the meantime, been peer reviewed. The data resulting from field studies indicating the absence of trophic magnification in certain food webs have already been taken into consideration. The mere fact that another study has been published after the other studies or that it has in the meantime been peer reviewed does not alter the general finding as to the existence of data indicating that there is no trophic magnification in certain food webs. That is also true of the MSC’s conclusion that those data did not invalidate the other evidence supporting the conclusion on bioaccumulation, particularly not the data relating to the BCF.

192    In the second place, the applicants submit that ECHA erred in dismissing the Gobas et al. (2015a) study and the Bridges and Solomon (2016) study according to which D4, D5 and D6 are not bioaccumulative, ‘due to differences in the underlying bioaccumulation assessment approach’, without explaining why those studies were not relevant.

193    According to the applicants, the explanation that the Gobas et al. (2015a) study was dismissed because it was based on the Canadian assessment process, the bioaccumulation criteria of which differ from those of Regulation No 1907/2006, is incorrect. The criteria are the same. The Canadian assessment process looks beyond the BCF, which is also expressly required by Section 3.2 of Annex XIII.

194    They claim that the explanation that the Bridges and Solomon (2016) study was dismissed because it had based its conclusion with regard to bioaccumulation solely on biomagnification and had entirely disregarded bioconcentration is also incorrect. The authors of that study disregarded the BCF because it is not decisive for assessing the hazards linked to the substances in question for the environment.

195    ECHA disputes those arguments.

196    In that regard, the Court finds that the applicants’ arguments are contradictory. On the one hand, they state that ECHA has not provided the reasons for which it dismissed those studies. On the other hand, they state that the reasons provided by ECHA are incorrect.

197    In addition, it should be noted that it is apparent from the documents in the file that ECHA did explain why those two studies were not taken into consideration. The support document for D5 states that the Gobas et al. (2015a) study is based on a Canadian bioaccumulation assessment process which is different from the system under Regulation No 1907/2006, in particular in that biotransformation is given significant weight.

198    The support documents for D4 and D5 indicate that the Bridges and Solomon (2016) study considered BCF values to be reliable but of little relevance, and that the conclusion on bioaccumulation completely neglected bioconcentration. The RCOM documents on D4 and D5 indicate that that approach was thus not in line with Annex XIII.

199    That explanation provided by ECHA is consistent with Annex XIII, in which the criteria for identifying B and vB substances are established in relation to the BCF in aquatic species. By contrast, the arguments raised by the applicants calling that explanation by ECHA into question are based on their misinterpretation of Annex XIII, already put forward and rejected in the fourth part of the first plea.

200    At the hearing, the applicants submitted that, in the present part of the first plea relating to D4 and D5, they also claimed ‘indirectly’, by referring to the Gobas et al. (2015) study in paragraph 114 of the application and in paragraph 59 of the reply, that the BCF values of D4 and D5 had been incorrectly growth corrected, as they had argued in respect of D6 in the sixth part of the first plea. In that regard, it is true that those paragraphs refer to the Gobas et al. (2015a) study and that paragraph 59 of the reply refers also to the Gobas et al. (2015) study. However, those paragraphs mention neither that a growth correction was made nor that that correction constitutes a manifest error of assessment, or even the BCF values reported for those two substances. Thus, it is not apparent either from the application or from the reply that the applicants intended to challenge, in the fifth part relating to D4 and D5, an alleged growth correction of the BCF values relating to those substances. That argument is therefore out of time and, consequently, inadmissible.

201    Also irrelevant in that regard is the applicants’ argument, raised in paragraph 67 of the reply in the sixth part of the first plea relating to D6, that ‘the assessment of D4, D5 and D6 was not carried out [by ECHA] by using the BCF factor as defined in Annex XIII, i.e. as the tool to determine whether the vB criterion is met’ and that, rather, ‘instead, a growth corrected BCF factor and a 5%-lipid-normalised BCF factors were referred to as bioaccumulation metrics for the assessment’. First, this argument is found under the heading ‘By failing to take into account all relevant information in concluding on the bioaccumulation (vB) of D6’ and it seems to provide clarification as regards the arguments in the application on the bioaccumulation of D6. Second, that argument was raised only belatedly – at the stage of the reply – and does not provide any clarification as regards the argument on the assessment of the bioaccumulation of D4 or of D5 which appears in the application. Third, the applicants do not mention any details relating to the BCF values of D4 and D5 before or after the alleged correction. Fourth, the applicants’ general reference in footnote 107 of the reply to the Annex XV dossiers relating to those three substances cannot replace an argument put forward by the applicants in that regard. It is not for the Court to seek and identify in the annexes the pleas and arguments on which it may consider the action to be based, since the annexes have a purely evidential and instrumental function (see judgment of 17 September 2007, Microsoft v Commission, T‑201/04, EU:T:2007:289, paragraph 94).

202    Consequently, the applicants’ argument relating to an alleged growth correction of D4 and D5, raised at the hearing, must be rejected as out of time and, in any event, as unsubstantiated.

203    The applicants’ line of argument relating to the Gobas et al. (2015a) study and to the Bridges and Solomon (2016) study must therefore be rejected.

204    In the third place, the applicants claim that ECHA stated that new studies indicated the existence of background contamination without assessing the reliability and accuracy of those studies. However, the presence of the substances at issue in the environment is not necessarily indicative of bioaccumulation. It indicates that ECHA selected studies to support its claims on bioaccumulation.

205    The applicants submit that several experts have pointed out that the analytics of siloxanes and certain environmental findings were clearly influenced by point sources. Consequently, it was necessary to assess each study for potential sample contamination or issues arising during handling and analysis.

206    ECHA disputes those arguments.

207    As regards the applicants’ argument that the presence of a substance in the environment is not necessarily an indicator of bioaccumulation, it should be noted, as ECHA rightly submits, that the studies at issue are presented in the support documents as available information. The conclusion drawn from those studies in the support documents is that they indicate the existence of background contamination. The applicants have not proved that that conclusion is incorrect. It is not apparent from the support documents that those studies were regarded as indicating the existence of bioaccumulation.

208    As regards the potential sample contamination, it must be stated that the applicants have not established a sufficient link between the analytics of siloxanes and some environmental findings, on the one hand, and ECHA’s alleged obligation to assess, in respect of each study, whether the samples might have been contaminated or whether problems arose during their handling and analysis, on the other. ECHA is not under such a general obligation since the laboratory procedures and test standards applied by scientists are designed to avoid as far as possible the problems to which the applicants refer. Moreover, the applicants have not demonstrated that those rules and standards were not complied with or that results should be considered to have been distorted for such reasons. Furthermore, it is apparent from the support documents for D4 and D5 that ECHA acknowledged and took into consideration the fact that certain results had been influenced by point sources.

209    Thus, the applicants’ argument relating to the studies which, according to the support documents, indicate the existence of background contamination must be rejected.

210    In the fourth place, the applicants claim that ECHA failed to take into account certain data that had been provided to it during the public consultation relating to the risk-management option analysis, from which it is apparent that D4 is highly metabolised by benthic organisms, which prevents any bioaccumulation or any toxicity to those biota.

211    The Court does not consider it necessary to rule on whether or not ECHA was obliged to take such information into consideration, as the applicants claim, even though that information had not been submitted as part of the procedure leading to the adoption of the contested decision but as part of a different and, moreover, voluntary procedure. In any event, it has already been noted in the fourth part of this plea that information relating to metabolism by benthic organisms was taken into consideration during the procedure leading to the adoption of the contested decision. The applicants have not shown, in that context, that the assessment of those observations on metabolism in the RCOM document was manifestly incorrect, nor have they indicated how the data submitted during the public consultation on the risk-management option analysis concerning the same subject were different from the information taken into consideration.

212    Accordingly, it must be held that none of the arguments raised by the applicants is capable of demonstrating that ECHA committed a manifest error in its assessment of the new studies.

213    The fifth part of the first plea must therefore be rejected as unfounded.

 Sixth part of the first plea: commission by ECHA of a manifest error of assessment in failing to take into account all the relevant information in concluding on the bioaccumulation of D6

214    The arguments raised by the applicants in this part of the first plea can be divided into four complaints.

215    By the first complaint, first of all, the applicants maintain that the support document for D6 does not explain why the information available on the BMF and the TMF did not invalidate the other lines of evidence, or why certain laboratory bioaccumulation tests cited as a basis for the BCF were ‘more reliable’. By those mere assertions, ECHA did not apply a weight-of-evidence determination such as that provided for in Annex XIII and Section 1.1.1.3 of Annex I to Regulation No 1272/2008.

216    Next, the applicants, supported by ACC, claim that ECHA merely cited the new evidence without indicating what weight, quantitative or qualitative, had been granted to it or how it had been assessed in relation to the data assessed previously.

217    Lastly, they claim that ECHA attached disproportionate importance to the field studies showing biomagnification and to the higher values, such as a BCF value of 5 000 or more, to support the conclusion that D6 was a vB substance.

218    By the second complaint, the applicants claim that the CERI (2010) study shows a BCF value in common carp (Cyprinus carpio) below the BCF of 5 000 which is required to meet the vB threshold. ECHA referred to a growth-corrected BCF value and to 5%-lipid-normalised BCF values. That practice, as the Gobas et al. (2015) study noted, is subject to scientific errors and regulatory errors.

219    The applicants do not object to the normalisation of the BCF values, but claim that normalisation in the present case is vitiated by errors, with the result that only growth (the denominator) was normalised, whereas, in their view, air and food intake (the nominator) should also have been normalised. Thus, the only approach which ECHA should have followed was to apply the 5 000 threshold for the BCF without normalisation. Similarly, ECHA also failed to take account of the fact that the CERI (2010) study applied the methodology originally intended for the BCF calculation, without applying normalisation.

220    The applicants add in their reply that ‘Gobas [noted] that [the Annex XV dossiers] [do] not include an appropriate error analysis or a statistical test to determine that the reported BCF [was] higher than the criterion value of 5 000’.

221    By the third complaint, the applicants claim that ECHA imported the errors made in the assessment of D4 and D5 into the assessment of D6. Given that the identification of D4 as a PBT substance and of D5 as a vPvB substance was incorrect, the identification of D6 as a PBT and vPvB substance since it contains D4 or D5 is also incorrect. ECHA also failed to assess independently the alleged PBT and vPvB properties of D6, merely referring, on several occasions, to the earlier MSC and RAC opinions on D4 and D5.

222    By the fourth complaint, the applicants assert that ECHA failed to take into account all the available information and did not evaluate the evidence ‘in any meaningful manner’.

223    ECHA disputes those arguments.

224    As a preliminary point, it should be noted that the arguments raised by the applicants in the first and third complaints of this sixth part are closely linked to the arguments raised in the first and fourth parts of this first plea.

225    As regards the first complaint of the sixth part of the first plea, it should be borne in mind that the applicants’ argument relating to Regulation No 1272/2008 has already been rejected in the fourth part of this first plea. Given that Section 1.1.1.3 of Annex I to Regulation No 1272/2008 applies to the classification of hazardous substances and mixtures into hazard classes according to that annex and that there is no indication that the rules laid down therein should also apply to the identification of PBT and vPvB substances under Annex XIII, that argument must also be rejected in this sixth part.

226    In addition, it must be stated that, contrary to what the applicants claim, the support document for D6 does indeed state why the data establishing that biodilution could occur in some food chains or in parts of some food chains did not invalidate the other lines of evidence. That support document lists, inter alia, the various results of the TMF data recorded in the field and notes that certain studies which had established a median TMF value of less than 1 had measured BMF values above 1 in certain predator-prey relationships. Additionally, that support document states that the criteria laid down in Annex XIII do not contain a threshold for TMF values or BMF values, but that values above 1 could be sound evidence that the substance meets the B property criteria of Annex XIII. The applicants have not demonstrated that manifest errors were committed in that reasoning or as regards the weight given to higher BCF values. The reasoning set out in the support document for D6 is, moreover, consistent with Annex XIII, in accordance with the considerations already set out in the fourth part of the first plea.

227    The applicants’ argument concerning the laboratory bioaccumulation tests cited as the basis for the BCF, which were considered to be ‘more reliable’, is linked to the arguments raised in the second complaint of the sixth part relating to the interpretation of the results of the CERI (2010) study, reference to which is therefore made.

228    The first complaint must therefore be rejected.

229    Next, the third complaint must be rejected since the applicants have not demonstrated that errors were committed in the assessment leading to the identification of D4 and D5 as PBT and vPvB substances, with the result that the identification of D6 as such a substance since it contains D4 or D5 is not vitiated by such an error having been transposed.

230    Furthermore, it is apparent from the file, and in particular from the support document for D6, that the MSC assessed the information on the bioaccumulation of D6 in detail and in approximately 70 pages. As has been indicated in the first part of the first plea, the fact that ECHA made specific references to results obtained in other procedures does not mean that it did not carry out an independent assessment and does not constitute a manifest error of assessment. Similarly, the comparisons made between the results relating to D6, on the one hand, and the results relating to D4 and D5, which are similar substances, on the other, do not in themselves constitute such an error. The applicants have not demonstrated that a specific reference or a specific comparison made in the support document for D6 constituted a manifest error of assessment on the part of ECHA.

231    As regards the second complaint of the sixth part, relating to the application of normalisation of lipid content and growth correction of the results of the CERI (2010) study, it must be held that the applicants’ argument, raised only at the stage of the hearing, according to which normalisation of lipid content should have been carried out, is out of time and must, therefore, be rejected as inadmissible.

232    In any event, as ECHA stated at the hearing and as is apparent from the support document for D6, normalisation of lipid content was not applied to the BCF values obtained from the CERI (2010) study because the deviation from the standard content of 5% was considered negligible. More specifically, it is apparent from the support document for D6 that the application of lipid normalisation would have led, in the present case, to BCF values that were 94% of those values. The applicants have not claimed that the assessment according to which the deviation was negligible was manifestly incorrect. Nor have they argued that the application of normalisation would have had a decisive impact on the BCF values in that the non-normalised values exceeded the threshold for vB properties laid down in Section 1.2.2 of Annex XIII, whereas normalisation would have led to values below that threshold.

233    Consequently, that argument of the applicants must, in any event, be rejected as ineffective.

234    As regards growth correction, it must be noted that ECHA claims that the kinetic BCF results from the CERI (2010) study were normalised in accordance with OECD Guideline No 305, OECD Guidance Document No 264 and Chapter R.11 of the ECHA Guidance, as updated in 2017. That is not disputed by the applicants, which rather call into question that methodology recognised by ECHA and by the OECD, since the formula used is, in their view, unrealistic in that the denominator (substances released) is adapted to correspond to a fish which does not grow whereas the numerator (substances taken in) is not adapted and corresponds to a fish which is growing. In their replies to the Court’s questions, they state that a specific critique relating to the calculation of the data in question, such as that expressed in the Gobas et al. (2015) study, should have taken precedence over a general guideline.

235    In that regard, it is necessary to reject the argument raised by the applicants in their replies to the Court’s questions that a divergent scientific opinion should take precedence over a general guideline. If that were the case, the usefulness of such ECHA-recognised guidelines would be called into question, as would ECHA’s broad discretion, given that disagreement in the scientific community on a given issue is not uncommon. Thus, the mere existence of a divergent opinion, even if it specifically concerns the substance at issue, cannot oblige ECHA no longer to follow a guideline which it has accepted as being appropriate.

236    In addition, it must be stated that it is apparent from the file that the specific criticism to which the applicants refer, namely that expressed in the Gobas et al. (2015) study, was addressed during the administrative procedure, as ECHA claims, by the MSC during its 60th meeting and in the response to comment No 5058 in the RCOM document for D6 (the ‘response to comment No 5058’).

237    As regards the criticism in the Gobas and al. (2015) study relating to growth correction, the reply to comment No 5058 sets out the reasons why the values were growth corrected, including in order to prevent rapid fish growth from creating the false impression of rapid depuration. While acknowledging certain uncertainties relating to growth correction, that reply states that the alternative to that correction would have been to reject the studies in which excessive fish growth had been found and to request further vertebrate testing, which is not consistent with the objectives of Regulation No 1907/2006.

238    The applicants have not shown that that response to comment No 5058 establishes the existence of a manifest error of assessment. In that regard, it should be noted that the growth correction provided for in OECD Guideline No 305 and in the ECHA Guidance, both updated in 2017, seeks to reduce scientific bias, as ECHA claims. ECHA’s choice to apply that recognised and bias-reducing method, even if there were some uncertainties in relation to it, instead of requiring further testing on vertebrates or using non-standardised results, does not constitute a manifest error of assessment.

239    Lastly, the applicants’ argument that the CERI (2010) study initially followed another methodology is irrelevant, in particular because the data in that study were sufficiently detailed to allow the 2017 updates to be applied, as ECHA contends without being contradicted by the applicants on that point.

240    As regards the other criticisms raised in the Gobas et al. (2015) study referred to by the applicants, namely those relating to the lack of analysis of the errors and of statistical criteria, it must be held that the applicants have also failed to demonstrate that inter alia the detailed response to comment No 5058, which also deals with those criticisms, reveals a manifest error of assessment in that regard.

241    First, contrary to what the applicants suggest in their comments on ECHA’s replies to the Court’s questions, the response to comment No 5058 does not contain a mere reference to the criticism, but rather a detailed response to that criticism given over several pages. Second, it is correctly stated in that response that neither Annex XIII nor the ECHA Guidance requires a statistical analysis or an error analysis, as are proposed in that comment. Third, that response indicates that a statistical analysis of the data relating to the BCF should be carried out only if there were several BCF values resulting from different tests carried out under the same conditions with the same species. The applicants, in their replies to the Court’s questions, merely assert that ‘there were additional studies which showed different results as opposed to the BCF study on which [ECHA] relied exclusively’. The applicants do not specify whether those alleged studies met the criteria mentioned in the response to comment No 5058, in particular whether they were carried out under the same conditions and with the same species. Nor do they put forward any argument calling into question those criteria applied by ECHA. ECHA, in its observations on the applicants’ replies, disputes the existence of such studies. It must be stated that there is nothing in the file to indicate the existence of such studies.

242    Consequently, the second complaint of the sixth part must be rejected.

243    Lastly, the fourth complaint must also be rejected on the ground that the applicants provide no specific example of relevant information that was not taken into account in the assessment of the bioaccumulation of D6.

244    Accordingly, the sixth part of the first plea must be rejected in its entirety.

 Seventh and eighth parts of the first plea: commission by ECHA of a manifest error of assessment in failing to take into consideration the information on the toxicity of D5 and D6 on their own, in identifying D5 and D6 as PBT substances based on the presence of D4 as an impurity, and without taking account of the specific limits relating to the quantity of D4 agreed upon by the MSC

245    The seventh and eighth parts of the first plea are each divided into two complaints.

246    In the first complaint of the seventh and eighth parts of the first plea, the applicants, supported by ACC, claim that ECHA committed a manifest error of assessment in failing to take into account in its assessment information relating to the toxicity of D5 and D6 on their own in order to establish whether those two substances met the criteria for T substances set out in Annex XIII, and in relying instead solely on the properties of D4 as an impurity of D5 and D6.

247    In the first place, the fifth paragraph of the preamble to Annex XIII provides that the identification ‘shall also take account’ of the PBT and vPvB properties of relevant constituents of a substance. In addition, the fourth paragraph of the preamble to Annex XIII, according to which the information used for the purposes of assessment of the PBT and vPvB properties are to be based on ‘data obtained under relevant conditions’, requires that the data be obtained on ‘the substance itself’.

248    The applicants submit that the RAC, in its opinion of 10 March 2016, stated that ‘D5 [did] not meet the REACH Annex XIII criteria for toxicity on the basis of the available evidence’. According to them, that conclusion is based on data obtained through tests performed with D5 containing D4 as an impurity.

249    Moreover, D5 and D6 are substances made up of a single main constituent, namely D5 and D6, respectively. There exists testing aimed at identifying whether D5 may be identified as ‘toxic’.

250    The support document for D6 indicates that human-health toxicity and ecotoxicity properties were not assessed. Given that information on tests on the human-health toxicity and ecotoxicity of D6, including its impurities, was available and submitted to ECHA, the latter was required to classify D6 on the basis of those tests or at least to take that information into account.

251    Accordingly, the present case differs from that which gave rise to the judgment of 7 March 2013, Rütgers Germany and Others v ECHA (T‑96/10, EU:T:2013:109). In paragraph 114 of that judgment, the Court assessed whether the ‘study of a [substance of unknown or variable composition, complex reaction products or biological materials] as a whole’ was possible and established that ‘most of the testing methods available for determining the inherent properties of those substances [were] only suitable for the study of substances made up of a single main constituent’. As a result, ECHA was required to identify D5 and D6 on the basis of tests conducted with D5 and D6 on their own or, at the very least, to take into consideration the information relating to the toxicity of D5 and of D6.

252    In the second place, the applicants, supported by ACC, refer to settled case-law of the EU Courts, and more specifically to paragraph 29 of the judgment of 7 October 2015, Bilbaína de Alquitranes and Others v Commission (T‑689/13, not published, EU:T:2015:767), according to which ‘it cannot be held that, merely because a constituent of a substance has a certain number of properties, the substance itself also has them, but, rather, the proportion in which that constituent is present and the chemical effects of such presence must be considered’. ECHA did not take into account the chemical effects of D4’s presence as an impurity in D5 and D6.

253    In the second complaint of the seventh and eighth parts, the applicants observe that the contested decision includes D5 and D6, respectively, in the Candidate List as such. In the applicants’ view, that decision does not specify that those substances meet the criteria laid down in Article 57(d) of Regulation No 1907/2006 only when they contain D4 in quantities equal to or greater than 0.1% (weight by weight). Thus, the contested decision is neither consistent with the MSC’s agreement nor supported by the evidence.

254    ECHA, supported by the Commission, disputes those arguments.

255    As regards the first complaint of the seventh and eighth parts of this plea, in the first place, it should be recalled that Annex XIII, in the version applicable to the present case, expressly provides in the fifth paragraph of its preamble that ‘identification shall also take account of the PBT/vPvB-properties of relevant constituents of a substance’.

256    First, contrary to what the applicants and ACC submit, the word ‘also’ does not indicate that ECHA should have assessed the toxicity of D5 and D6 as such when it knew that the data were not complete in that regard or that the toxicity of those two substances was already apparent from other available data, in particular the data relating to D4 present as an impurity in D5 and D6. Such an assessment would have been of no purpose and have had no bearing on the conclusion.

257    Moreover, such an interpretation would not be compatible with one of the objectives of Regulation No 1907/2006, namely the objective to ensure a high level of protection of human health and the environment, as laid down in Article 1(1) of that regulation. It is consistent with that objective that, even in situations in which the assessment of one or more of the PBT and vPvB properties of a substance is neither possible nor useful, the properties of the constituents of that substance be taken into account.

258    Second, it is necessary to examine whether impurities of a substance such as D4 contained in D5 and D6 in the present case fall under the term ‘constituent’.

259    That term is not defined in Regulation No 1907/2006. According to the ECHA Guidance, it refers to the main constituents, impurities and additives of substances the composition of which is well defined, as well as the constituents of substances of unknown or variable composition, complex reaction products or biological materials (‘UVCB substances’).

260    The relevance of the taking into account of impurities for the characterisation of a substance is confirmed by other provisions of Regulation No 1907/2006. For example, the chemical safety assessment covers, according to Section 0.3 of Annex I, the ‘use of the substance on its own (including any major impurities and additives)’. Similarly, according to Section 3 of Annex II to Regulation No 1907/2006, Section 3 of the safety data sheet is to describe the ‘chemical identity of the ingredient(s) of the substance or mixture, including impurities’.

261    Moreover, an interpretation such as that, which allows the properties of impurities to be taken into account when assessing the PBT or vPvB properties of a substance, is consistent with the objective of ensuring a high level of protection of human health and the environment, as laid down in Article 1(1) of Regulation No 1907/2006. It would not be convincing to make a distinction between the various components of a substance in such a way that the constituents of a UVCB substance would be considered relevant for the identification of PBT or vPvB substances whereas impurities would not, regardless of their concentration in that substance and their effects.

262    Third, the applicants’ line of argument relating to the judgment of 7 March 2013, Rütgers Germany and Others v ECHA (T‑96/10, EU:T:2013:109, paragraph 114), must be rejected.

263    To begin with, the legal framework applicable in the present case is different from that applicable in the case which gave rise to the judgment of 7 March 2013, Rütgers Germany and Others v ECHA (T‑96/10, EU:T:2013:109). In particular, the fifth paragraph of the preamble to Annex XIII was not applicable when that judgment was delivered. As a result, that paragraph could not be interpreted in that case.

264    In addition, the judgment of 7 March 2013, Rütgers Germany and Others v ECHA (T‑96/10, EU:T:2013:109), concerned UVCB substances. There is no indication in that judgment that constituents of other substances, including impurities, should not be taken into account in an assessment of PBT or vPvB properties.

265    In the second place, it should be recalled that the Court found, in a case concerning classification under Regulation No 1272/2008, that it could not be held that, merely because a constituent of a substance had a certain number of properties, the substance itself also has them and that the proportion in which that constituent is present and the chemical effects of such presence must be considered (judgment of 7 October 2015, Bilbaína de Alquitranes and Others v Commission, T‑689/13, not published, EU:T:2015:767, paragraph 29).

266    Moreover, the Court held in the judgment of 7 March 2013, Rütgers Germany and Others v ECHA (T‑96/10, EU:T:2013:109, paragraph 113 et seq.), that the reasoning given by ECHA according to which it was ‘important to assess a UVCB substance on the basis of its constituents because, once in the environment the individual constituents of such a substance will behave as independent substances’ and according to which ‘the substances in question will release several [polycyclic aromatic hydrocarbons] with PBT or vPvB properties during use, for example by heating during processing or by leaching upon contact with water’ was not vitiated by an error of assessment.

267    In the present case, it is apparent from the RCOM document for D5, and in particular from the response to comment No 5122, as well as from the RCOM document for D6, and in particular from the response to comment No 5123, that ECHA found that the D4 present as an impurity in D5 and in D6 could be released into the environment over a shorter or longer period and behave as a PBT substance there.

268    Consequently, it must be held that ECHA did consider the behaviour of the D4 contained as an impurity in the substances at issue.

269    The applicants’ argument based on the fact that, during toxicity tests, D5 and D6 containing D4 as an impurity did not reveal toxicity is not capable of demonstrating that that finding is vitiated by a manifest error of assessment. In response to the Court’s questions, ECHA gave as an explanation for that observation that the concentration of D4 contained in D5 or D6 during the tests might not have been high enough or the toxic effects might have been too subtle. ECHA also stated that, if D5 and D6 were tested up to the limits of their water solubility (approximately 17 μg/l and 5 μg/l), the resulting concentration from a 1% D4 impurity in the test systems would still be far below the ‘no observed effect concentration’ for aquatic toxicity of D4, which is between 4 and 8 μg/l.

270    Contrary to what the applicants suggest in their comments on ECHA’s replies, it was not for ECHA to prove why the tests in question had not revealed toxicity. Without it being necessary to determine whether such evidence could have been adduced, it must be stated that the relevant question is whether ECHA, in failing to take into account the toxicity data of tests conducted with D5 and D6 on their own, committed a manifest error of assessment. In that regard, it must be held that the explanations provided by ECHA, according to which (i) it concluded that D5 and D6 were toxic on account of the presence of D4 in those substances as an impurity and (ii) that conclusion was not called into question by the results of the tests carried out with D5 and D6 on their own, are plausible.

271    Similarly, the argument raised by the applicants in their observations on ECHA’s replies, according to which the concentration of D4 in water would never reach a toxic level because of the low water solubility of the substances at issue and the low concentration of D4 in D5 and D6, is unconvincing and does not dispel the concern expressed by ECHA. That concern was that the D4 contained in D5 and D6 could be released and, subsequently, behave as a PBT on account of its intrinsic properties. Once released, however, the initial concentration of D4 in D5 or D6 is no longer decisive. The D4 will find itself in the environment where it may accumulate and ultimately have toxic effects.

272    Furthermore, and unlike the applicants in the case which gave rise to the judgment of 7 October 2015, Bilbaína de Alquitranes and Others v Commission (T‑689/13, not published, EU:T:2015:767), the applicants in the present case do not put forward any substantial evidence demonstrating that the reasons relied on by ECHA were incorrect and, in particular, that the D4 present as an impurity in D5 and D6 would not release itself into the environment.

273    Lastly, as regards the 0.1% threshold of D4 in D5 and D6 set in the contested decision as a factor entailing the identification of D5 and D6 as substances of very high concern, it must be pointed out that the applicants have not put forward any specific argument calling that choice into question. Consequently, it is sufficient to state, in that regard, that it is apparent from the case-law that the choice of a threshold of 0.1% as a factor entailing the identification of a substance on the basis of its constituents does not in itself constitute a manifest error (see, to that effect, judgments of 7 March 2013, Bilbaína de Alquitranes and Others v ECHA, T‑93/10, EU:T:2013:106, paragraph 96 et seq.; of 7 March 2013, Rütgers Germany and Others v ECHA, T‑94/10, EU:T:2013:107, paragraph 117 et seq.; and of 7 March 2013, Cindu Chemicals and Others v ECHA, T‑95/10, EU:T:2013:108, paragraph 124 et seq.). That finding applies not only to the constituents of a UVCB substance such as that at issue in the cases which gave rise to those judgments, but also to the impurities present in a substance, such as D4 in D5 and D6.

274    It follows from all of the foregoing that the applicants, in the first complaint of the seventh and eighth parts of the first plea, have not demonstrated that ECHA committed a manifest error of assessment in not taking into consideration the information relating to the toxicity of D5 and D6 as such, and in identifying those two substances as PBT substances due to the presence of D4 as an impurity in them.

275    Accordingly, those complaints must be rejected as unfounded.

276    As regards the second complaint of the seventh and eighth parts, it is appropriate first of all to clarify their subject matter. It is apparent from the line of argument set out by the applicants in the context of these complaints that they consider it possible to challenge, in isolation, Decision ED/61/2018 of the Executive Director of ECHA including D4, D5 and D6 in the Candidate List. However, such reasoning ignores the clarifications contained in the MSC’s opinion, as elaborated in the support documents for D5 and D6 and in the entries relating to those two substances in the Candidate List published on ECHA’s website.

277    In that regard, it should be noted that it is apparent from settled case-law that it is the act of identifying a substance resulting from the procedure referred to in Article 59 of Regulation No 1907/2006 which is intended to produce binding legal effects vis-à-vis third parties within the meaning of the second sentence of the first paragraph of Article 263 TFEU (judgments of 7 March 2013, Bilbaína de Alquitranes and Others v ECHA, T‑93/10, EU:T:2013:106, paragraph 32; of 7 March 2013, Rütgers Germany and Others v ECHA, T‑94/10, EU:T:2013:107, paragraph 33; and of 7 March 2013, Cindu Chemicals and Others v ECHA, T‑95/10, EU:T:2013:108, paragraph 40).

278    The Court thus held in cases also concerning the identification of a substance as being of very high concern, that the subject matter of the dispute had been identified by the applicants by reference to the MSC’s agreement on the identification of substances (pursuant to Article 59(8) of Regulation No 1907/2006), to the decision of the Executive Director ordering the inclusion of the substance in the Candidate List (under the same provision) and to the publication and updating of the Candidate List on the ECHA website (pursuant to Article 59(10) of Regulation No 1907/2006) (judgments of 7 March 2013, Bilbaína de Alquitranes and Others v ECHA, T‑93/10, EU:T:2013:106, paragraph 31; of 7 March 2013, Rütgers Germany and Others v ECHA, T‑94/10, EU:T:2013:107, paragraph 32; and of 7 March 2013, Cindu Chemicals and Others v ECHA, T‑95/10, EU:T:2013:108, paragraph 39).

279    In the present case, it is apparent from the application that the applicants seek the annulment of the contested decision in so far as that act is intended to produce binding legal effects. More specifically, the applicants’ claims relate to ECHA’s decision in so far as it includes the three substances at issue in the Candidate List ‘as substances of very high concern’. Thus, it is not convincing that the applicants, solely in the context of the present complaints, do not take account of the other elements of the act identifying D4, D5 and D6 as substances of very high concern.

280    Moreover, Decision ED/61/2018 expressly refers to the unanimous agreement reached by the MSC at its 60th meeting and orders the publication and updating of the Candidate List. It is true that that decision merely indicates the names of the substances, their EC and CAS numbers and their intrinsic properties, such as those provided for in Article 57(d) and (e) of Regulation No 1907/2006. Nevertheless, it must be stated that nothing in Decision ED/61/2018 – especially not the omission of certain details from the text of that decision itself – indicates that the inclusion of D5 and D6 in the candidate list of substances on ECHA’s website should depart from the MSC’s agreement, as reasoned in the support documents for D5 and D6.

281    Those latter documents, which incidentally are also published on ECHA’s website, and the entries relating to D5 and D6 in the Candidate List, each contain a consistent clarification that D5 and D6, respectively, meet the PBT criteria when they contain 0.1% or more of D4. It is unmistakably clear from this that ECHA identified both D5 and D6 as PBT substances when they contain 0.1% or more of D4.

282    It follows from the foregoing that the mere fact that Decision ED/61/2018 does not expressly mention that D5 and D6 meet the PBT criteria only when they contain 0.1% or more of D4 does not mean that D5 and D6 were identified as having PBT when they contain D4 in lower concentrations, or that that decision is not in conformity with the MSC’s unanimous agreement.

283    Consequently, the second complaint of the seventh and eighth parts of the first plea must be rejected, as must, therefore, those two parts and the first plea in its entirety.

 Second plea in law:breach of the principle of proportionality

284    The applicants, supported by ACC, claim that the contested decision breaches the principle of proportionality.

285    In the first place, they argue that identifying D4, D5 and D6 as substances of very high concern was neither appropriate nor necessary to attain the objectives pursued.

286    They claim that ECHA’s conclusions were based on a number of uncertainties, as ECHA itself acknowledged. Consequently, ECHA was required to request further information by means of the evaluation procedure before identifying the substances at issue as substances of very high concern. Moreover, it was disproportionate to disregard all uncertain information only when it pointed to the conclusion that D6 was not a vB substance.

287    They also claim that the contested decision exceeds the limits of what is necessary to attain the objective pursued. That decision, after all, identifies D5 and D6 as substances of very high concern on account of their PBT properties even when they contain D4 in quantities below 0.1%. However, it is apparent from the support documents for D5 and D6 that those substances meet the PBT criteria only when they contain D4 in quantities equal to or greater than 0.1%.

288    In the second place, the applicants submit that the identification of D4, D5 and D6 as substances of very high concern was not the least onerous measure to which ECHA could have had recourse. According to the applicants, ECHA could have used the substance evaluation procedure to gather the additional information needed to reach a conclusion on the alleged PBT properties of the substances. In that case, the applicants would have had to bear only part of the costs of the studies required.

289    They claim that the identification of the substances at issue as being of very high concern causes direct losses to the suppliers of those substances. In addition, that identification ‘triggers significant adverse market attention from [environmental protection] organisations’.

290    Furthermore, the applicants submit that identifying D5 and D6 only when they contain D4 in quantities equal to or greater than 0.1% would have been a less onerous measure.

291    ECHA, supported by the Federal Republic of Germany and the Commission, contends that the second plea must be rejected as unfounded.

292    It is necessary to reject the applicants’ line of argument according to which the contested decision exceeds the limits of what is necessary to achieve the objective pursued and is not the least onerous measure on the ground that ECHA allegedly failed to specify that D5 and D6 met the PBT criteria only when they contained D4 in quantities equal to or greater than 0.1%. Those arguments have been raised and rejected in the second complaint of the seventh and eighth parts as being based on a misinterpretation of the facts. Consequently, those arguments must also be rejected in the context of the second plea.

293    Moreover, it should be recalled that, according to settled case-law, the principle of proportionality, which is one of the general principles of EU law, requires that measures adopted by EU institutions should not exceed the limits of what is appropriate and necessary in order to attain the legitimate objectives pursued by the legislation in question, and where there is a choice between several appropriate measures, recourse must be had to the least onerous, and the disadvantages caused must not be disproportionate to the aims pursued (see judgments of 21 July 2011, Etimine, C‑15/10, EU:C:2011:504, paragraph 124 and the case-law cited, and of 30 April 2015, Polynt and Sitre v ECHA, T‑134/13, not published, EU:T:2015:254, paragraph 104).

294    With regard to judicial review of the conditions referred to in the previous paragraph, ECHA must be allowed a broad discretion in a sphere which entails political, economic and social choices on its part and in which it is called upon to undertake complex assessments. The legality of a measure adopted in that sphere can be affected only if the measure is manifestly inappropriate having regard to the objective which the legislature is seeking to pursue (see, to that effect, judgments of 21 July 2011, Etimine, C‑15/10, EU:C:2011:504, paragraph 125 and the case-law cited, and of 30 April 2015, Polynt and Sitre v ECHA, T‑134/13, not published, EU:T:2015:254, paragraph 105 and the case-law cited).

295    It is apparent from Article 1(1) of Regulation No 1907/2006 that that regulation seeks to ensure a high level of protection of human health and the environment, including the promotion of alternative methods for assessment of hazards of substances, as well as the free circulation of substances on the internal market while enhancing competitiveness and innovation. To that end, that regulation introduces an integrated system for monitoring chemical substances, including registration, evaluation and authorisation, together with possible restrictions on their use (see judgment of 13 July 2017, VECCO and Others v Commission, C‑651/15 P, EU:C:2017:543, paragraph 27 and the case-law cited).

296    Having regard to recital 16 of that regulation, it must be stated that the legislature set as the main purpose the first of those three objectives, namely to ensure a high level of protection of human health and the environment (see judgment of 30 April 2015, Polynt and Sitre v ECHA, T‑134/13, not published, EU:T:2015:254, paragraph 106 and the case-law cited).

297    As regards, more specifically, the aim of the authorisation procedure, of which the identification procedure set out in Article 59 of Regulation No 1907/2006 forms part, Article 55 of that regulation states that its aim is essentially to ensure the proper functioning of the internal market while assuring that the risks from substances of very high concern are properly controlled and that these substances are progressively replaced by suitable alternative substances or technologies where these are economically and technically viable (see judgments of 13 July 2017, VECCO and Others v Commission, C‑651/15 P, EU:C:2017:543, paragraph 28 and the case-law cited, and of 30 April 2015, Polynt and Sitre v ECHA, T‑134/13, not published, EU:T:2015:254, paragraph 106 and the case-law cited).

298    As for the objective of protecting human health and the environment, it must be stated from the outset that the identification of a substance as being of very high concern serves to improve information for the public and professionals as to the risks and hazards incurred and that, consequently, such identification must be regarded as a means of enhancing that protection (see judgment of 30 April 2015, Polynt and Sitre v ECHA, T‑134/13, not published, EU:T:2015:254, paragraph 108 and the case-law cited).

299    With regard to the applicants’ line of argument relating to the uncertainties regarding certain data identified during the assessment of the substances at issue, it must be stated that the recognition of some uncertainties in relation to certain data did not, as such, render the identification of D4, D5 and D6 as substances of very high concern inappropriate. The recognition and assessment of scientific uncertainties, where they exist, is an essential part of a weight-of-evidence determination, as ECHA correctly submits. Indeed, Annex XIII states that appropriate weight must be attached to the quality and consistency of data. In the present case, ECHA, while acknowledging certain uncertainties relating to certain data, considered the overall data to be conclusive. It has already been found in the first plea that the applicants have failed to demonstrate that ECHA selected certain studies in order to support its conclusion or that it committed any other manifest error in the assessment of the data or in its conclusion.

300    Consequently, the applicants’ line of argument relating to the allegedly inappropriate nature of the contested decision must be rejected.

301    In the light of the line of argument put forward by the applicants according to which ECHA should have carried out an evaluation, which would have constituted a less onerous measure, it should be recalled that, in a proportionality test, consideration may be given to possible less restrictive means than the measure adopted by the Union institution only if they are equally suitable for achieving the objective pursued by the EU measure in question (see, to that effect, judgments of 14 December 2004, Arnold André, C‑434/02, EU:C:2004:800, paragraph 55; of 14 December 2004, Swedish Match, C‑210/03, EU:C:2004:802, paragraph 56; and of 4 May 2016, Philip Morris Brands and Others, C‑547/14, EU:C:2016:325, paragraph 180).

302    An evaluation, however, would not have been as appropriate for achieving the objective pursued by the identification of the substances at issue as substances of very high concern as described in paragraph 295 et seq. above. It is true that the evaluation procedure, which is intended as a follow-up to registration, according to recital 20 of Regulation No 1907/2006, also serves to improve information for the public and professionals as to the hazards and risks of a substance, as is apparent from recitals 19 and 21 of that regulation. However, while registered substances should be allowed to circulate on the internal market, as is apparent from recital 19 of Regulation No 1907/2006, the objective of the authorisation procedure, of which the identification procedure set out in Article 59 of that regulation forms part, is, inter alia, progressively to replace substances of very high concern by suitable alternative substances or technologies, where these are economically and technically viable (see paragraph 297 above) (judgment of 30 April 2015, Polynt and Sitre v ECHA, T‑134/13, not published, EU:T:2015:254, paragraph 114).

303    Moreover, it is not apparent from Regulation No 1907/2006 that the legislature intended to make the identification procedure carried out pursuant to Article 59 of that regulation, which forms part of the authorisation procedure for a substance set out in Title VII of that regulation, subject to the registration procedure laid down in Articles 44 to 48 of that same regulation (see, to that effect, judgment of 30 April 2015, Polynt and Sitre v ECHA, T‑134/13, not published, EU:T:2015:254, paragraph 114).

304    As a result, the applicants’ argument that an evaluation would have constituted a less onerous measure must also be rejected.

305    Lastly, it is necessary to examine whether the identification of D4, D5 and D6 caused disadvantages disproportionate to the aims pursued, as the applicants claim, on the ground that suppliers will suffer commercial effects and will be subject to adverse market attention from environmental protection organisations.

306    The applicants do not specify in what those commercial effects and that adverse attention consist. Thus, it is not possible to assess the extent of those alleged disadvantages and, therefore, to conclude that there are disadvantages disproportionate to the aims pursued.

307    In any event, it has already been established that the identification of a substance as a substance of very high concern serves to improve information for the public and professionals as to the risks and dangers incurred (paragraph 298 above). Moreover, as is apparent from recital 69 of Regulation No 1907/2006, the legislature wished to pay careful attention to substances of very high concern.

308    It is clear that the public and professionals may adapt their decisions and actions when they obtain better information following the identification of a substance as a substance of very high concern. Thus, the provisions of Articles 57 to 59 of Regulation No 1907/2006 would be rendered redundant if their application had to be deemed disproportionate merely because the public and professionals could take that information into account in their future actions and decisions. Therefore, and given that the applicants have not alleged the existence of more specific disadvantages which go beyond the disadvantages mentioned above, it cannot be concluded that the contested decision causes disadvantages which are disproportionate to the aims pursued.

309    Consequently, the second plea must be rejected as unfounded and, therefore, the action must be dismissed in its entirety.

 Costs

310    Under Article 134(1) of the Rules of Procedure, the unsuccessful party is to be ordered to pay the costs if they have been applied for in the successful party’s pleadings. Since the applicants have been unsuccessful, they must be ordered to bear their own costs and to pay those incurred by ECHA, in accordance with the form of order sought by ECHA.

311    Article 138(1) of the Rules of Procedure provides that the Member States and institutions which have intervened in the proceedings are to bear their own costs. The Federal Republic of Germany and the Commission shall therefore bear their own costs.

312    Under Article 138(3) of the Rules of Procedure, the Court may order an intervener other than those referred to in Article 138(1) and (2) of those rules to bear its own costs. In the circumstances of the present dispute, ACC shall bear its own costs in connection with the proceedings before the Court.

On those grounds,

THE GENERAL COURT (Eighth Chamber, Extended Composition)

hereby:

1.      Dismisses the action;

2.      Orders Global Silicones Council and the other parties listed in the annex to bear their own costs and to pay those incurred by the European Chemicals Agency (ECHA);

3.      Orders the Federal Republic of Germany, the European Commission and American Chemistry Council, Inc. (ACC) each to bear their own costs.

Papasavvas

Svenningsen

Barents

Pynnä

 

Laitenberger

Delivered in open court in Luxembourg on 30 June 2021.

E. Coulon

 

S. Papasavvas

Registrar

 

President


Table of contents


Legal framework

Background to the dispute

Procedure and forms of order sought

Law

First plea in law: commission by ECHA of a manifest error of assessment in the assessment of the B properties of D4, D5 and D6 and in the assessment of the T properties of D5 and D6, exceedance of its powers and infringement of Article 59 of Regulation No 1907/2006

First part of the first plea: commission by ECHA of a manifest error of assessment in relying on the opinions of the MSC and the RAC without conducting its own assessment of the information available to it, thereby ‘importing’ the errors vitiating those opinions

Fourth part of the first plea: commission by ECHA of a manifest error of assessment in reaching conclusions on the bioaccumulation of D4 and D5 which the evidence relied upon is incapable of supporting

Second part of the first plea: commission by ECHA of a manifest error of assessment in concluding that D4, D5 and D6 met the PBT and vPvB criteria set out in Annex XIII even though persistence (P property) and bioaccumulation (B property) were not established for the same compartment

Third part of the first plea: commission by ECHA of a manifest error of assessment in not taking into consideration the specific (hybrid) nature of D4, D5 and D6

Fifth part of the first plea: commission by ECHA of a manifest error of assessment in failing to assess the new evidence on bioaccumulation (B and vB properties) for D4 and D5 available to ECHA after the MSC and the RAC’s opinions had been issued

Sixth part of the first plea: commission by ECHA of a manifest error of assessment in failing to take into account all the relevant information in concluding on the bioaccumulation of D6

Seventh and eighth parts of the first plea: commission by ECHA of a manifest error of assessment in failing to take into consideration the information on the toxicity of D5 and D6 on their own, in identifying D5 and D6 as PBT substances based on the presence of D4 as an impurity, and without taking account of the specific limits relating to the quantity of D4 agreed upon by the MSC

Second plea in law: breach of the principle of proportionality

Costs


*      Language of the case: English.


1      The list of the other applicants is annexed only to the version sent to the parties.