August 08
Comments of the International Antimony Association (i2a) on the INERIS report concerning
the implementation of requirements on priority substances
within the context of the water framework directive
1. High priority for water
The classification as high priority substance for antimony for water was based on the fact that
PEC2 /90th percentile of background > 10
AND PEC1/90th percentile of background > 1
AND number of analyses > DL / total number of analyses ≥ 10%
AND PEC1 ≥ PEC2
The third and fourth requirements were only fulfilled when using the PECs based on total antimony concentration measurements.
We do not agree with this approach and are of the opinion that antimony should not be classified in any of the classes used in the prioritization report for the following reasons:
1. When comparing both PEC1 and PEC2 (no matter whether based on total or dissolved antimony concentrations) with the PNECwater, which was taken from the EU risk assessment of diantimony trioxide, PEC/PNEC ratios are far below 1. Based on these calculations, antimony should not be classified in any of the classes used in the prioritization report. We agree that the PEC/background approach may be used as an indication for substances of which the effects on living organisms have not been (satisfactorily) assessed yet. However, when a reliable PNEC is available, the PEC/background approach should no longer be considered. Only the PEC/PNEC approach (risk-based approach) has ecological relevance.
2. In the EU risk assessment report on diantimony trioxide, the PECregional was calculated to be 0.72 µg Sb/L. This concentration was calculated using ambient measured Sb concentrations from eight European countries. The prioritization exercise of INERIS leads to much higher PECs. This is probably due to the inclusion of measurements near point sources (this is for instance clear from Annex II.2, mentioning a maximum measured concentration of 1000 µg Sb/L). Such measurements should be excluded from PEC calculation. The EU risk assessment concluded on no regional risks. The inclusion of point source measurements is probably also the reason why PEC1 > PEC2 (for PECs based on total antimony concentrations).
3. In the EU risk assessment report on diantimony trioxide, the PEClocal varied between 0.72 and 69.4 µg Sb/L, the highest PEClocal being calculated using conservative default values in a generic scenario for application of textile back-coating. No local risks were identified. Note that the highest PEClocal is much lower than the highest ‘ambient’ concentration given in Annex II.2 (1000 µg/L). Clearly, measurements near point sources (with very high emissions, suggesting measurement directly in industrial effluent) have been included for ‘ambient’ PEC calculation in the INERIS report.
We can agree with the proposed background concentrations (based on the FOREGS database), although it should be recommended not to include the two outliers (up to 2.91 µg Sb/L) since these have a large influence on the 90th percentile.
Overall, it should be avoided to compare PECs based on total concentrations to PNECs or background concentrations based on dissolved concentrations.
2. Medium priority for sediment
The classification as medium priority substance for Sb for sediment was based on the fact that
PEC2 /90th percentile of background > 1
AND PEC1/90th percentile of background > 1
AND number of analyses > DL / total number of analyses ≥ 10%
AND PEC1 ≥ PEC2
These requirements were only fulfilled when using the PECs for the 20 µm fraction and the background concentrations for the 63 µm fraction. No PECs were calculated for the 63 µm fraction.
We do not agree with this classification and are of the opinion that antimony should not be classified in any of the classes used in the prioritization report for the following reasons:
1. No PNECsediment was mentioned in the INERIS report. However, in the EU risk assessment of diantimony trioxide, a PNECsediment of 11200 µg Sb/kg dw was calculated based on results from reliable studies. For reasons of consistency, we propose using the PNEC value for sediment of the EU Risk Assessment in the INERIS report. None of the PECs calculated by INERIS is higher than this PNEC value (no matter which sediment fraction is considered). Based on the PEC/PNEC ratios, antimony should not be classified in any of the classes used in the prioritization report, since only the PEC/PNEC approach has ecological relevance (see remark on the PEC/background approach given under section 1).
2. In the EU risk assessment report on diantimony trioxide, the PECregional was calculated to be 3000 µg Sb/kg dw. This concentration was calculated using ambient measured Sb concentrations from five European countries. The prioritization exercise of INERIS leads to higher PECs. This is probably due to the inclusion of measurements near point sources (this is for instance clear from Annex II.8, mentioning a maximum measured concentration of 34400 µg Sb/L, which is in the range of the PEClocals calculated in the EU risk assessment using conservative default values in the generic scenarios for formulation as flame retardant in textiles and application of textile back-coating ). Such measurements should be excluded from an ‘ambient’ PEC calculation. No regional risks were identified for sediment in the EU risk assessment.
The inclusion of point source measurements is probably also the reason why PEC1 > PEC2 (for the 20 µm fraction).
We can agree with the proposed background concentrations (based on the FOREGS database).
Overall, it should be avoided to compare PECs for certain sediment fractions to PNECs or background concentrations for other sediment fractions.
3. No evaluation for biota
Apparently, no concentration measurements in biota were provided by the member states. Consequently, no evaluation was performed.
An overview of antimony concentrations measured in biota and corresponding BCFs is given in the EU risk assessment of diantimony trioxide. PECs for fish were calculated using both a BCF of 40 and 15000. The PNECoral for secondary poisoning has been calculated to be 374.8 mg Sb/kg food. This value was based on the results from a 90-d repeated dose study on rats and the assessment factors suggested in the TGD (for more detailed explanation see the EU risk assessment report on diantimony trioxide). All PEC/PNEC ratios were below 1, indicating no risk for secondary poisoning.
Although no PNECoral or ADI is mentioned in Annex III of the INERIS report, Annex VIII reports a PNECoral or ADI of 0.25 mg Sb/kg food. It is not clear on which study this value is based. In case calculations for biota will be performed in the near future, it is strongly recommended to use the PNECoral mentioned in the EU risk assessment.
4. Drinking water
Annex III of the report mentions a drinking water standard of 5 µg Sb/L (Dir. 98/83/EC). However, the WHO guideline (revised in 2003) for antimony in drinking water is 20 µg Sb/L. Apparently, in case several standards are available, the lowest has been retained in the report. However, we do not agree with the retained value, since the drinking water directive is currently under revision, and adjustment of the EU standard to the more recently revised WHO standard has been proposed.
The WHO standard was adjusted in 2003. The adjustment was based on the NOAEL of a subchronic drinking water study in rats (Poon et al., 1998). Lynch et al. (1999) suggested a NOAEL for this study of 6.0 mg/kg bw/d based on decreased body weight and reduced food and water intake. Applying an uncertainty factor of 1000 (100 for intra- and interspecies variation and 10 for the use of a subchronic study), a TDI of 6 µg/kg bw could be determined. The guideline value of 20 µg/L was derived from this TDI by assuming a 60-kg adult drinking 2 L of water per day and allocating 10% of the TDI to drinking water. The WHO report mentions that this value could be highly conservative because of the nature of the endpoints and the large uncertainty factor.
INERIS did not take the evaluation for drinking water into account for classifying substances in prioritization classes. However, we strongly recommend using the WHO guideline (which will most likely be adopted by the revised EU drinking water directive) instead of the standard mentioned in the current EU drinking water directive. In Annex VIII, comparison of PEC1 (based on total measured concentrations) with the current EU standard results in a ratio > 1, whereas this would not be the case when using the current WHO drinking water guideline for antimony.
5. Summary
1. When a reliable PNEC is available, the PEC/background approach should no longer be considered.
2. Measurements near point sources should be excluded from PEC calculation (for both water and sediment).
3. For water, it should be avoided to compare PECs based on total concentrations to PNECs or background concentrations based on dissolved concentrations.
4. The PNEC value for sediment of the ATO EU Risk Assessment should be used in the INERIS report. The field was currently left blank in the INERIS report.
5. It should be avoided to compare PECs for certain sediment fractions to PNECs or background concentrations for other sediment fractions.
6. PECs for biota and the PNEC sec poisoning from the EU Risk Assessment report should be used for the evaluation for biota.
7. Please explain how the ADI’s have been calculated.
8. More recent and lower WHO guideline value for drinking water should be preferred over the older and higher 98/83/EC value
Taking the above comments into account, antimony should not be classified in any of the prioritization classes proposed by INERIS.
Literature
European Union Risk Assessment Report – Diantimony trioxide. 2007. Rapporteur: Swedish Chemicals Agency. 545 pp.
World Health Organization. 2003. Antimony in drinking water – Background document for development of WHO guidelines for drinking water quality. WHO/SDE/WSH/03.04/74. 14 pp.
For the references of Poon et al. (1998) and Lynch et al. (1999) see WHO report.
International Antimony Association VZW – Av. De Broqueville 12 – 1150 Brussels – Belgium
Tel: +32 (0) 3 297 60 92 Fax: +32 (0) 3 297 60 92 WWW.IAOIA.ORG VAT : BE 0894.536.176 Page 1