DIRECTORATE-GENERAL
ENVIRONMENT
Directorate D - Water, Chemicals & Biotechnology
ENV.D.1 - Water
12 February 2010
Expert Group on Review of WFD Priority Substances List (EG-R)
Sub-group of the Working Group E on Priority Substances
DG Environment, 26-27 January 2010
Draft Minutes
First day of the meeting: 26 January 2010
Participants: Jorge Rodriguez Romero (JRR), Madalina DaviD (MD), Helen CLAYTON (HC), Karola Grodzki (KG), Ana Paya Perez (APP), Klaus Daginnus(KD), Alice JAMES (AJ), Benoît FRIBOURG-BLANC (BFB), Manfred CLARA (MCl), Flemming INGERSLEV (FI), Susan LONDESBOROUGH (SL), Raphaël DEMOULIERE (RD), Dieter SCHUDOMA (DS), Mario CARERE (MCa), Dorien TEN HULSCHER (DTH), Eric VERBRUGGEN (EV), Concepcion SANZ MARTIN (CSM), Niklas JOHANSSON (NJ), John BATTY (JB), Helen WILKINSON (HW), Ann DIERCKX (AD), André LECLOUX (AL), Dolf VAN WIJK - Euro Chlor (DVW), Klaas DEN HAAN (KDH), Mick HAMER - Syngenta (MH), Thierry SCHOONEJANS - Dow (TS), Richard ALLEN - Bayer Cropscience (RA), Frank VAN ASSCHE (FVA), Ismene JAEGER – EEB (IJ), Stefan SCHEUER – Greenpeace (SS).
1. Welcome and adoption of the Agenda
John BATTY (JB) and Ana PAYA-PEREZ (APP) welcome everybody. No comment being made, the agenda is adopted and JB reminds the first session will not focus on substances specific comments.
2. Introduction
Jorge RODRIGUEZ-ROMERO (JRR) welcomes all participants and introduces the “new name” of the “group” which is not an “expert group” anymore but a “sub-group” of WG E (even if it is still a group set up of experts). 2010 is a challenging year and the WD agreed this group needs enough resources to meet the timescale: WD conclusions especially stating this point are part of the documents for the present meeting.
A special WG E meeting was organised on the 12th of January: the EQS guidance was discussed, some changes were made, and the guidance was endorsed and supported by the working group with changes agreed. The TGD EQS will be ready by the end of the current week and final version will be distributed beginning of next week and presented at the next SCG meeting on 23rd of February. After this meeting, two additional consultations are planned: an inter-service internal COM review (peer review) and a review by the SCHER. The TGD EQS will also be presented to WD meeting in May.
JRR stresses that the current meeting is a key meeting with objective to conclude on the list of substances on which preparation of dossiers will be needed, as support in the delineation of the final list. The objective is to select a number of substances as low as possible and comprised between 30 and 50 substances. The objective is also to share the work between Member States (MS) and the Commission as leaders for the preparation of dossiers, with a consultation with stakeholders and other MS before dissemination.
3. Review of the results of monitoring-based prioritisation
Benoît FRIBOURG-BLANC (BFB) presents the two products provided on the quality check of the database. They are complementary to the first quality check made with the monitoring-based prioritisation report. These products were asked and agreed by the EG-R at the September meeting:
- the first product provides summary statistics and information on substances ranking high or very high through monitoring-based approach and substances ranking high through modelling (score 1). It is provided in the form of tables and graphs presenting quantified and non-quantified analyses. It is available in the form of an information sheet for each substance selected, a downloadable pdf file and accessible through the substances website in a dedicated section.
- the second product is an excel file gathering summary statistics on 102 pesticides and more specifically on the highest concentrations (peak measurements) found in the dataset for these substances.
Mick HAMER (MH) asks what are PEC1 and PEC2.
BFB explains these are figures calculated as reported in the monitoring-based prioritisation report, i.e. both values are 90th percentile of mean concentrations by stations, PEC1 being based on quantified concentrations only while PEC2 is based on all concentrations by stations, with values below DL being replaced by DL/2.
Klaas DEN HAAN (KdH) asks if the peak calculation could be made available for substances finally selected during this meeting.
JRR answers the peak calculations were asked for by EG-R and its extension can be envisaged if the group requires it.
4. Results of the modelling-based prioritisation
Jose ZALDIVAR COMENGES (JZC) presents the approach used for this part of the prioritisation exercise: the screening including JRC proposal and additional lists proposed by various stakeholder lead to 2034 substances. Each criteria used consist in deriving a score. The first screening criterion is the hazard assessment using PBT (Persistent, Bioaccumulative and Toxic) and ED (endocrine disruptors/disrupting) properties. The second criterion is exposure, based on total production and use index. The combination of both resulting scores is used to define a list of substances, and 78 substances are identified in this first step. It is then necessary to make a PEC and PNEC estimation and calculate the associated risk ratio to rank the substances, based on databases and available information. Some results are presented on the top 20 ranked substances. The results are complementary to the monitoring based approach with some uncertainties related to lack of (reliable) data on some criteria.
Eric VERBRUGGEN (EV) mentions the first screening tier is based on “hazardous” properties and then the second tier is based on PEC / PNEC ratio. Nothing further is done with the hazardous properties of the substances in the final ranking. He asks thus why there is so much emphasise on these properties in the screening phase.
Ismene JAEGER (IJ) warns that taking on board hazard and exposure the approach embeds many substances but there is a need to take into account also the so-called emerging chemicals such as pharmaceuticals and ED that are not with the current scheme, due to lack of data. She adds that the substance-by-substance approach is not the only one, with the need to consider toxicity of mixtures.
JZC answers this stems from the need to take into account both hazard and exposure, and from the fact that the overall objective is to rank substances. He agrees the substance by substance approach is insufficient and not protective enough and that this drawback needs to be explored further in the future.
André LECLOUX (AL) says ECETOC TRA database used in prioritisation predicts high concentrations, which is normal because this tool intends to be very protective to avoid additional toxicity testing. He mentions it was said that production figures are not always available and when several figures on use are found, choice is made to use the highest emission value. He then asks what are the production figures used for the worst case because if overall figures are difficult to find, how could production figures for a specific use be found more easily?
JZC answers they used the worst-case scenario which may affect the results when data are scarce.
MH requests for substances ranking high in the modelling-based approach that website fact sheets are also made available with what is in the monitoring database.
Dorien TEN HULSCHER (DtH) asks if it would be possible to have an intermediate step between final selection of substances on the shortlist and preparation of the dossier: some substances selected could be found at a later stage as being not relevant and preparing a dossier is a heavy work that needs to be correctly targeted.
JCZ mentions that this could be, in fact, the case where the substance has a RAR which concludes on no risk for water, or the case where the substance is not anymore produced.
JRR recalls the decision taken at the September meeting where it was agreed modelling-based prioritisation results were used as they are provided, with the set of information used to derive the substances, combined with the other lists and refined with the preparation of dossiers. The list for which the group will prepare dossiers will be defined in the current meeting, it is then necessary to find more technical criteria to select/deselect substances from the list to reduce the overall effort needed. However, in the current meeting or when preparing the dossier, if relevant technical arguments show that a substance is not to be considered, then it can be the conclusion.
5. List of candidate substances for prioritisation and identification of the shortlist of 30-50 substances for dossiers preparation
5.1. Proposals and comments received on the list of candidate substances since September 2009
JRR recalls the sequence of steps within the overall prioritisation process and the work done since EG-R September 2009 meeting. He then presents the list of candidate substances; the comments and proposal received and recalls the content of dossiers required. He presents the final short list of substances selected for preparation of dossiers and the current list of countries who proposed to take the lead on dossiers before the current meeting and which should be finalised.
EV reminds the September meeting, where it was raised that some substances of annex I of Directive 91/414 were not selected and says that in the current proposal they will not be further investigated, whereas monitoring in the Netherlands downstream of big rivers show that they are of concern.
DtH asks what the state of play of the use of peak exposure measurements is within the overall process.
JRR answers that elements were provided for the MS and stakeholders (EG-R) to look at and use to propose additional identification criteria and support their views. He then presents the summary of comments and additional substances proposed.
Alice JAMES (AJ) then presents the essential data needed in the dossier so that the dossiers are a strong basis for the derivation of EQS.
Discussion:
-Flemming INGERSLEV (FI) + MH: do all data need to be reported (e.g. REACH data)? Because this represents quiet a huge amount of work.
AJ answers the aim is not to report the whole set of available data but to refer as much as possible to already validated/compiled data and to report only core data to determine QS.
-KdH argues the most secure toxicological data is not a good term to use because we shall not choose systematically the lowest data
AJ precise “the most secure” does not necessarily mean “the lowest data”, but “realistic worst case” based on relevant validated data, it may be replaced by “the most conservative”, if too controversial.
-DtH says good quality data are the most relevant ones but that this implies high costs and the need to be sure the selected substances are relevant in a first step.
-Dieter Schudoma (DS) asks if all data should be presented or only those relevant for standard derivation.
AJ mentions that the overall objective is to refer as much as possible to the work done already and validated at a certain level.
JRR adds it is not a scientific validation work but an identification of the most available recent and relevant information, peer reviewed or validated at MS or other level. Evidence of risk will lead to inclusion and provide elements for EQS derivation. Preparation of dossiers will be uneven, with some needing 2-3 days and others with many data requiring more. This is under the remit of the sub-group on Review, completed by extensive consultation during the process.
JRR presents then document 5.2a with the information sources used to define the sub-lists of substances finally selected for the prioritisation exercise:
-the monitoring-based list, with de-selection of 14 substances only selected through monitoring and for which rank is due to a fraction with less than 4 countries reporting,
-the modelling-based list of substances with the highest risk ratio,
-the list of substances based on final Risk Assessment Reports with risk for aquatic environment other than local involving few sites,
-the list of substances based on final risk assessment reports for the remaining 11 pesticides. This number is defined after de-selection of a) all authorised substances with no other information source and b) those not ranking high in any prioritisation, and ranked with the Toxicity Exposure Ratio (TER). 2 substances are proposed for de-selection due to monitoring-based ranking.
-the list of substances based on final risk assessment reports for the remaining biocide, after de-selection of a) all authorised with no other information source and b) those not ranking high in any prioritisation process (nor monitoring-based, nor modelling-based),
-the substances selected through other sources of information:
- PBT/ vPvB properties, criteria associated high rank of substances in monitoring and/or modelling approaches. Three substances are only identified with this criterion and not in monitoring/modelling, and are proposed for de-selection.
- Substances of Very High Concern in the context of REACH (SVHC): 12 substances selected.
- POP with no addition based on these sole properties
- Citations in Environmental Quality Standards Directive (2008/105/EC) annex III: some substances not already selected and added.
-proposals from Sweden (Irgarol) and the European Environmental Bureau (pharmaceuticals and ED) added to candidates for discussion
DtH mentions Chromium and Arsenic are on the list of Rhine relevant substances.
EV adds deselection of these substances is based on monitoring only.
JRR confirms that these substances were not considered in the modelling-based prioritisation process and adds there deselection is due to the so-called a-posteriori checks on monitoring datasets. He reminds that even if other datasets exist, the agreed process is to rely on available European monitoring data gathered for the exercise.
5.2. Identification of the 30-50 shortlist of substances to prepare dossiers / 5.3. Discussion on the proposal for a shortlist of 30-50 substances to prepare dossiers
Warning: for more details on the process, the dossier content or any comment, please refer to document 5-Shortlist PS Sub-Group Jan 2010.ppt available on Circa
JRR briefly introduces document 5.2d that gathers general and substance specific comments and issues received since the EG-R September meeting and proposes to focus the afternoon discussion on each comment, to conclude.
Various general comments are made on the overall process and its associated reliability:
-Mario CARRERE (MCa) and Frank VAN ASSCHE (FvA) feel concerned by the equilibrium between the various pathways considered, with more emphasise on monitoring and modelling approaches. They say that when a substance is selected only on the basis of RAR or PBT properties it should be completed by other sources of prioritisation because it could be based on old/outdated data.
-AL adds risk reduction strategies should lead to reduction of concentrations.
-Karola Grodzki (KG) reminds the sub-group that substances not on the list should not simply be forgotten.
-Raphaël DEMOULIERE (RD) proposes conversion of concentrations between biota and water to take into account all the substances that are ranking high/very high for biota but not enough countries were represented so they failed the “EU representative criteria”.
-DtH and EV suggest exploring this in future prioritisation for improvement of the process.
-MCa highlights that in the document PAH are mentioned as already on current PS list. He reminds PS list contains only 6 indicative PAHs for PAH group and pyrene is not part of them. Therefore, de-selection cannot be based on that criteria and he suggests changing this formulation. EV supports this proposal.
JB answers the problems highlighted have to be taken into account together with a balanced effort and ads that dossiers will also identify major gaps in knowledge.
JRR says the approach chosen is the best possible weighted approach and an iterative process, which comes from extensive shallow analysis to a more in-depth one for a limited number of substances. He adds that dossiers will be used to conclude on the selection of substances. On the issue of already existing priority substances and their EQS revision, a separate process is planned.
JRR then presents the general comments in detail.
-The Netherlands comment on the use of BLM to derive PNEC that poses a risk of underestimation.[1]
Katrien DELBEKE (KD) answers that using the realistic worst case provides the same conclusion with corrected or non-corrected PEC and PNEC values.
-DtH argues on the fact that some data from pesticides dossiers have been used for prioritisation process while they should not have (because they take on board some WFD irrelevant criteria such as recovery, etc.)
DS supports EV and DtH and suggests a review of pesticides selection.
JRR mentions the annex of monitoring-based prioritisation report dealing with choices made for the use of mesocosms data and choice made for derivation of PNECs. He also mentions the online factsheets to check when a substance specific question arises.