ToxicologyLetters, 64/65 (19921535-545 0 1992 Elsevier Science Publishers B.V., All rights reserved 03784274/92&i 5.00
535
Environmental hazard classification of chemicals Alf Lundgren National Chemicals Inspeetorate, Solna (Sweden) Key words: Classification and labelling; Environmental hazard labelling; Harmonisation; Classification criteria
SUMMARY Information on the environmentally dangerous properties of chemicals is of fundamental importance if the users of chemicals are to take account of the risks to the environment in their choice of chemical products, and to organize the use of chemicals and waste disposal in an environmentally safe manner. In order to assist suppliers of chemicals in setting out their product information and to ensure its consistency, it is important to have rules on classification and labelling. Such rules must be based on criteria for environmental hazard classification. During the 1980s specific classification criteria for the aquatic environment were developed in the Nordic countries and by the Commission of the European Communities. These criteria now form the basis for the regulations on classification and labelling of chemical substances in both the EC and EFTA countries. The regulations entered into force in the EC on 1 July 1992 and will be enforced in the EFTA countries as soon as possible. Besides labelling there are several other areas, e.g. transport regulations, where classification criteria are or will be applied on a national or international scale. International harmonization of criteria at this early stage is highly desirable as it will prevent the development of divergent systems which will require harmonization at a later stage. To this end, the OECD initiated a Clearing House activity on the feasibility of elaborating harmonized approaches to environmental hazard classification. The Clearing House recommended adoption of the current classification criteria, joint development of criteria for other sections of the environment and joint development of procedures for updating and methodologies for classification of chemical preparations. Further international work in this area will be coordinated by IPCS. Harmonization of classification is also a programme area under Chemical Forum with IPCS as coordinator as agreed at UNCED 92,
INTRODUCTION
Information on the environmentally dangerous properties of chemicals is of fundamental importance if users of chemicals are to take account of the Correspondence to: A. Lundgren, National Chemicals Inspectorate, P.O. Box 1384, S-17127 Solna, Sweden.
536
risks to the environment in their choice of chemical products and to organize their use of chemicals and waste disposal in an environmentally safe manner. In order to assist manufacturers/importers of chemicals in setting out their product information and to ensure its consistency, it is important to have rules on classification and labelling. Rules on environmental hazard labelling are now being implemented in both the EEC and the EFTA countries. These rules are based on criteria for environmental hazard classification, which give the quantitative content of the concept hazardous or dangerous for the environment. CRITERIA FOR ENVIRONMENTAL CLASSIFICATION
Scientifically-sound criteria is a necessary basis for, e.g., administrative systems for classification and labelling of chemicals. Such criteria have been in operation for a long time with regard to health hazards, flammability, etc. In the 1980s the development of criteria also for environmental hazards was started, both at the Commission of the European Communities (CEC) and as a joint project of the Nordic countries (Denmark, Finland, Iceland, Norway and Sweden). In this paper the issue is approached from a Nordic viewpoint. The objective for the joint Nordic project was to develop scientific criteria for the classification of environmentally hazardous chemicals, which can form the basis for administrative systems, e.g. classification and labelling chemical of products. The work was carried out by representatives from the competent authorities in the Nordic countries with Sweden as the lead country and in co-operation with a Scientific Advisory Group (SAG) with experts from the different countries. Three basic prerequisites were formulated for the work: 1. The criteria should be based on the intrinsic properties of chemicals and on the results from a limited number of simple and internationally recognized tests (e.g. OECD’s Minimum Premarketing set of Data (MPD) or the EEC Base Set). 2. The criteria should be “transparent”, i.e. they should be given a form that could be readily understood also by an inexperienced user. 3. The criteria should cover all relevant compartments of the environment. The SAG concluded at an early stage that a complete hazard assessment was neither possible nor necessary as a basis for classification. Instead the criteria should reflect the minimum information on which a meaningful first evaluation could be made. But, of course, if a more comprehensive assessment is at hand it will supersede a simpler one.
537
It was also recognized that the criteria should not be restricted to information on acute environmental effects giving criteria on “eco-toxicity”, but also to take into consideration that many of the classical environmental poisons are not characterized by their acute toxicity, but by their persistence in the environment and their bioaccumulating properties leading to longterm effects. It was also found to be very difficult to combine the first and the third prerequisite, as the tests on fate and effects in the aquatic environment were much better established than the corresponding tests for the air and the terrestrial compartments. In the Nordic proposals, however, mammalian acute toxicity was included in order to cover a wider range of organisms and to address the terrestrial compartment to some extent. No criteria could be formulated for the air compartment since there were no generally recognized tests available at that time. In 1987 the framework for a classification scheme or model was elaborated by the Nordic group. There was, however, much uncertainty as to how this model would work in practice. Therefore it was decided that the Nordic model with a number of modifications should be tested, together with the corresponding models currently discussed at the CEC, on at least one hundred chemical substances and differences in the outcome should be analyzed. Figure 1 gives the general outline of the different models. The exercise also included different cut-off values for toxicity and bioaccumulation, inclusion or exclusion of algal and mammalian toxicity, etc., giving a total of 18 different models. The results and details of the procedure are reported elsewhere [ll. The different models arrived at surprisingly different numbers of substances classified for posing a potential hazard to the environment considering the great similarities between the models. With the most stringent model 79 substances out of 121 were classified, while 22 were classified with the least stringent model. The results were scrutinized by the SAG and the following conclusions were drawn: 1. The exclusion criterion for not bioavailable substances (MW > 1000, Fig. la) was excluded as it could not be evaluated. It was anticipated that this criterion would apply mainly to polymers, which for several reasons would need special consideration. 2. A strong tendency for a chemical to partition to the air compartment (PED,i, > 90%), calculated according to the level I fugacity model of Mackay and Paterson [21,was proposed to be an exclusion criterion in the Nordic model (Fig la). This was rejected as it was found that high volume chemicals, known to occur in the aquatic environment, would be excluded from further examination.
538
W
r---------t Mw>1ooo?
-cyesj
iL--------,4
I 1 i
0
1
rl
I
Ready
biodegr.?
I.
JNO
Ilwh4fhi
cl U-e
fbsfbssified
Very toxic?
yes+ Not Classified as Hazardous I
NO
J. I!??, insuff. 2T!Gzl1Bioacc?l’+2f3
r-l I
Classified as
verytoxic? -yes--_)Classified
as Hazardous
NO
Uxksikdw/
Ready biodegr.?l
NO
1
L&ass#Z
as Hazardous
+hssi&3
as Hazardous
Fig. 1. General outline of the tested models: (a>basic Nordic model; (b) modified Nordic model; (c) the CEC model; (d) the model suggested by Germany.
3. Mammalian acute toxicity added relatively little to the overall rate of classi~cation and can be excluded if only the aquatic com~ment is addressed.
539
4. The algal toxicity was shown to be a very sensitive component of an aquatic toxicity test battery and should be included. Besides it sensitivity it should be included as it will serve as an indicator of effects on photosynthetic organisms/primary producers. 5. Changing the aquatic toxicity threshold by a factor of 10 was the single most important modification of the models. The SAG found the more protective level to be appropriate considering the frequency of classification of substances of environmental concern. Even more important than the toxicity threshold was how the information on toxicity, biodegradability and bioaccumulation were combined. In all models, toxicity alone is regarded as a sufficient basis for classification if the substance is very toxic. In all other cases a combination of toxicity, biodegradability and bioaccumulation forms the basis for classification. Biodegradability is given much weight in many of the models as only not-readily biodegradable substances are considered for classification (see Figs. la, c and d). The reasoning is that readily-biodegradable substances will have a limited impact in time and space. This can, however, be questioned as it has been shown that other parameters (e.g. logK,,J can be more important predictors for occurrence in the environment [3,41. The model in Figure lb compensates for this to some extent as it gives equal weight to all three parameters. There are four possible combinations of two or more of the three properties: toxic, biodegradable and bioaccumulating. Only the model in Figure lb includes all four combinations. A unique combination for that model is “not-readily biodegradable+bioaccumulating”. Even if this combination did not substantially contribute to the rate of classification, the SAG stressed its importance in a classification scheme as it could serve as an indicator of possible long-term effects. Based on these conclusions and on the knowledge on environmental behaviour and effects of the substances selected for this comparison, the SAG recommended in December 1988 the model in Figure 2 for environmental classification of chemical substances. The consequences for the universe of chemicals of applying this classification model could not estimated as the selection of substances was not randomly made from all chemicals but was severely biassed towards environmentally dangerous substances [ll. Nevertheless, some idea of how frequent a classification will be can be given from a comparison with the frequency of classification for health hazards. This comparison showed that for the selected substances the rate of classification for environmental hazards was slightly lower than that for health hazards. The SAG found this rate satisfactory and in agreement with their general understanding of this concept.
540
Criteria: I---
Very toxic?
/YES_, Classified as Hazardous
Not classified as Hazardous
Threshold
vatues: 11Bioaccumu-/I Not ready bioii degradable jj lating
I
Oral LDm
~25 mgtkg
s 200 mg/kg 11 jj BCFrlOO
lnhal I-C,,
d 03 g/m3
-
1Fish LC&,
5 1,O mg/l
5 100 mg/i
!:
~~Asdefinedin 1;OECD Guidelines /or /i BODJCOD < 03
100 mgfl
ij ii I/
s 100 mg/l
/I
5
Fig. 2. The classification
91000
I
I j/
\ i 1
1 I
/
;I
/ ! I
model recommended by the Nordic Scientific Advisory Group.
Denmark, being a member of the EC, brought the recommended model to the working group at the CEC as a Danish proposal. This proposal was further developed in the CEC working group and in 1989 the working group concluded their work by presenting classification criteria that could form the basis for a system for classification and labelling of chemical substances (see Fig. 3). The major differences from the Nordic recommendation were: 1. Mammalian toxicity was excluded and only the aquatic en~ronment was addressed. 2. A third (intermediate) cut-off level for aquatic toxicity was introduced. 3. Substances that are “Harmful+ Bioaccumulating” were not regarded as hazardous to the environment. It should be noted that although the classification is composed of several components and combinations there is only one class for environmental
541
hazard: “Dangerous for the Environment”. It should also be noted that the criteria were developed for chemical substances and are not applicable to chemical preparations in their present form. SYSTEMS FOR CLASSIFICATION AND LABELLING
Once the basis was set in the form of specific classification criteria for the aquatic environment, the work continued with applying these criteria in administrative rules. Again, this was done both as a joint Nordic project and in a working group at the CEC. The final report from the Nordic project was published in July 1990 [51 and was formulated as recommendations on a system for classification, labelling and safety data sheets. The report covered four main areas: 1. Classification of substances: The classification of substances should be based on the classification criteria discussed above, but as the specific criteria do not cover all aspects of environmental hazard the specific criteria were supplemented by a general criterion: “Substances which are not classified as dangerous for the environment in accordance with the specific criteria but which, in the light of current knowledge regarding their toxicity, persistence and ability to accumulate and their calculated or observed behaviour and final fate in the environment, may constitute an immediate or long-term danger to the structure and/or function of a natural ecosystem should be classified as dangerous for the environment.” This general criterion has essentially the same wording as the general criterion elaborated by the CEC working group at that time. A list of examples on classified substances, continuously updated by the competent authority, would give further guidance to the manufacturers/importers of chemicals. 2. Classification of preparations: The classification of preparations should be determined by the properties of the single substances that the preparation contains. As a main rule, a preparation would be classified as dangerous for the environment if it contains 1% or more of a substance dangerous for the environment. Furthermore, a preparation would be classified if the preparation is very toxic, even if it does not contain 1% or more of a dangerous substance. 3. Labelling: Detailed proposals for how the information should be presented on the label were given in the report, as well as proposals for relevant risk- and safety-phrases. Regarding the hazard symbol a reference was given to the symbol proposed by the CEC. 4. Safety data sheets: The information on environmental, health and fire hazards should be coordinated. Besides the basic environmental hazard information on the label the safety data sheet should contain information on
542
risks and safety measures in free text and detailed data on LC&C~Ovalues, etc. The corresponding work in the CEC resulted in rules on environmental classification and labelling of chemical substances set out in the Commission Directive of 1 March 1991 adapting to technical progress for the twelfth time Council Directive 67/548/EEC [61. These rules cover classification and labelling of chemical substances (items 1 and 3 in the Nordic report), but do not address preparations or safety data sheets. The specific aquatic criteria form the basis for classification together with two general criteria, one for the aquatic environment and one for the non-aquatic environment. The directive also gives rules for the use of specified risk- and safety-phrases. In the latter part of March 1991 a seminar on environmental classification and labelling of chemicals was held in Uppsala, Sweden. The seminar was jointly sponsored by the CEC and the EFTA countries. Representatives from 17 of the 18 EC/EFTA countries and representatives from the CEC, the EFTA Office and the industry shared their experiences in this field and discussed different options for the future work. It was concluded 171that: - the experts from the EFTA and EC countries have a common understanding on the need for further elaboration of criteria, classification and labelling rules and safety data sheets on chemicals hazardous to the environment; - the present EC-regulations make a good basis for further work as well as e.g. the present Nordic proposals on different issues; - further work should be done in close cooperation between the EC and EFTA - the EFTA chemicals group and the CEC directorates DG III and XI should be prepared to organize the discussions needed for future cooperation subject to the outcome of the EES negotiations The situation today is that the regulations on environmental classification and labelling of chemical substances dangerous for the environment will enter into force no later than on 1 July 1992 in the EC countries and the EFTA countries will do the same as soon as possible. The same rules as in the EC countries will, e.g., be in force in Sweden from 1 January 1993. INTERNATIONAL HARMONIZATION OF CLASSIFICATION CRITERIA
The Uppsala seminar might be regarded as a first step towards international harmonization. The classification/labelling systems for chemical substances were discussed and coordinated although no formal decisions were taken. There are, however, several other areas, e.g. transport regulations, where classification criteria are or will be applied both on a national and international scale. Criteria for the aquatic environment have been
developed in some administrative systems, but have not yet been widely used. Criteria for other environmental compartments (soil, air, etc) are used even less. International harmonization at this early stage will help prevent the development of divergent systems which will require harmonization at a later stage. Along these lines, the OECD arranged a workshop on “Certain Aspects of the Classification of Chemicals” in Paris, November 1991. The workshop addressed two specific endpoints for classification: Acute oral toxicity, which is one of the oldest criteria for classification, and environmental hazard, which is one of the newest. It was concluded that international harmonization was desirable because it would increase the health and safety of chemical users, it would facilitate trade in chemicals, and it would reduce duplicative animal testing, and other work related to classification. The workshop recommended an ongoing OECD activity on harmonization of classification and labelling, starting with classification of acute oral toxicity and environmental hazard. At the Joint Meeting of the OECD Chemicals Group and the Management Committee the CEC, Sweden and the USA were given a mandate to initiate a Clearing House activity to examine existing systems and to report to the Joint Meeting on the feasibility of elaborating harmonized approaches with regard to these endpoints and to make recommendations to the Joint Meeting on possible options for the harmonization of existing schemes. To promote an efficient sharing of the burden with other UN organizations and UNCED 92 with respect to harmonization of classification schemes, it was agreed that OECD would participate in a proposed IPCS Coordinating Committee which also would involve ILO, WHO, UNEP, UNTDG and FAO. In February 1992 the leading countries discussed different classification criteria and drafted the proposal to the Joint Meeting at a meeting in Washington. This draft was open for discussions at an extended Clearing House meeting in Brussels in April, 1992. Experts from all OECD countries were invited to the Brussels meeting, together with representatives from relevant international organizations. Participants discussed the proposal and other issues related to harmonization. In particular it was noted that the IPCS Coordinating Committee in its work plan had given highest priority to harmonization of classification criteria for acute oral toxicity and hazard to the environment. The final proposal to the Joint Meeting was based upon the draft proposal of the Clearing House, but was modified to meet the concerns raised at the Brussels meeting. With regard to environmental hazard the Clearing House proposed: 1. Adoption of the current classification criteria as developed in the EC and The Nordic countries (see Fig. 3) as a starting point for common criteria for the aquatic environment.
classsfied
aa
Dangerous
clasaad
~nge~e
for the EnviroMnrwrt
the water solubiity of the substance i.skssthanlmg/L
Threshold values:
Fig. 3, Final classification scheme for the aquatic environment as developed in the EC and the Nordic countries and recommended for adoption by the OECD Clearing House.
2. Joint development of common criteria for other en~ro~en~l compartments and possibly for ozone depleting substances. 3. Joint development of procedures for updating and methodologies to take account of the cl~si~cation of preparation and those situations where additional information is available on the aquatic environment. In May 1992 the Joint Meeting agreed to the Clearing House proposal and urged the Clearing House to continue its work as proposed.
545
Harmonization of classification and labelling of chemicals is also a programme area under Chemical Forum that was agreed upon at UNCED 92. This will probably provide the necessary political pressure to bring these issues to a successful end within a reasonable period of time. REFERENCES Lundgren, A. (1989) Comparison of different models for environmental hazard classification of chemicals. Kern1 Report No 9/89. Swedish National Chemicals Inspectorate. Stockholm, 40 pp. Mackay, D. and Paterson, S. (1982) Fugacity revisited. Environ. Sci. Technol. l6,654A660A. Ikeda, M., Koizumi, A., Kasahara, M., Watanabe, T., Nakatsuka, H. and Sekita, Y. (1987) A statistical approach to the prediction of possible presence of pollutant chemicals in the environment. Regulatory Toxicology and Pharmacology 7(3), 321536. ECETOC (1988) Concentrations of industrial organic chemicals measured in the environment: The influence of physico-chemical properties, tonnage and use pattern. ECETOC. Technical Report No 29. Brussels. 105 pp. Gustafsson, L. and Ljung, E. (1990) Substances and preparations dangerous for the environment: A system for classification, labelling and safety data sheets. Miljerapport 1990:lOE. Nordic Council of Ministers. Copenhagen. 65 pp. Commission of the European Communities (1991) Commission Directive of 1 March 1991 adapting to technical progress for the twelfth time Council Directive 67/548/EEC (91/325/EEC). Official Journal of the European Communities. No. L 180 of 8.7.91 Freij, L. (Ed.) (1991) Seminar on Environmental Classification and Labelling of Chemicals. EC-EFTA Meeting, Uppsala, Sweden. March 20-21,199l. Swedish National Chemicals Inspectorate. Stockholm. 85 pp.
535
Environmental hazard classification of chemicals Alf Lundgren National Chemicals Inspeetorate, Solna (Sweden) Key words: Classification and labelling; Environmental hazard labelling; Harmonisation; Classification criteria
SUMMARY Information on the environmentally dangerous properties of chemicals is of fundamental importance if the users of chemicals are to take account of the risks to the environment in their choice of chemical products, and to organize the use of chemicals and waste disposal in an environmentally safe manner. In order to assist suppliers of chemicals in setting out their product information and to ensure its consistency, it is important to have rules on classification and labelling. Such rules must be based on criteria for environmental hazard classification. During the 1980s specific classification criteria for the aquatic environment were developed in the Nordic countries and by the Commission of the European Communities. These criteria now form the basis for the regulations on classification and labelling of chemical substances in both the EC and EFTA countries. The regulations entered into force in the EC on 1 July 1992 and will be enforced in the EFTA countries as soon as possible. Besides labelling there are several other areas, e.g. transport regulations, where classification criteria are or will be applied on a national or international scale. International harmonization of criteria at this early stage is highly desirable as it will prevent the development of divergent systems which will require harmonization at a later stage. To this end, the OECD initiated a Clearing House activity on the feasibility of elaborating harmonized approaches to environmental hazard classification. The Clearing House recommended adoption of the current classification criteria, joint development of criteria for other sections of the environment and joint development of procedures for updating and methodologies for classification of chemical preparations. Further international work in this area will be coordinated by IPCS. Harmonization of classification is also a programme area under Chemical Forum with IPCS as coordinator as agreed at UNCED 92,
INTRODUCTION
Information on the environmentally dangerous properties of chemicals is of fundamental importance if users of chemicals are to take account of the Correspondence to: A. Lundgren, National Chemicals Inspectorate, P.O. Box 1384, S-17127 Solna, Sweden.
536
risks to the environment in their choice of chemical products and to organize their use of chemicals and waste disposal in an environmentally safe manner. In order to assist manufacturers/importers of chemicals in setting out their product information and to ensure its consistency, it is important to have rules on classification and labelling. Rules on environmental hazard labelling are now being implemented in both the EEC and the EFTA countries. These rules are based on criteria for environmental hazard classification, which give the quantitative content of the concept hazardous or dangerous for the environment. CRITERIA FOR ENVIRONMENTAL CLASSIFICATION
Scientifically-sound criteria is a necessary basis for, e.g., administrative systems for classification and labelling of chemicals. Such criteria have been in operation for a long time with regard to health hazards, flammability, etc. In the 1980s the development of criteria also for environmental hazards was started, both at the Commission of the European Communities (CEC) and as a joint project of the Nordic countries (Denmark, Finland, Iceland, Norway and Sweden). In this paper the issue is approached from a Nordic viewpoint. The objective for the joint Nordic project was to develop scientific criteria for the classification of environmentally hazardous chemicals, which can form the basis for administrative systems, e.g. classification and labelling chemical of products. The work was carried out by representatives from the competent authorities in the Nordic countries with Sweden as the lead country and in co-operation with a Scientific Advisory Group (SAG) with experts from the different countries. Three basic prerequisites were formulated for the work: 1. The criteria should be based on the intrinsic properties of chemicals and on the results from a limited number of simple and internationally recognized tests (e.g. OECD’s Minimum Premarketing set of Data (MPD) or the EEC Base Set). 2. The criteria should be “transparent”, i.e. they should be given a form that could be readily understood also by an inexperienced user. 3. The criteria should cover all relevant compartments of the environment. The SAG concluded at an early stage that a complete hazard assessment was neither possible nor necessary as a basis for classification. Instead the criteria should reflect the minimum information on which a meaningful first evaluation could be made. But, of course, if a more comprehensive assessment is at hand it will supersede a simpler one.
537
It was also recognized that the criteria should not be restricted to information on acute environmental effects giving criteria on “eco-toxicity”, but also to take into consideration that many of the classical environmental poisons are not characterized by their acute toxicity, but by their persistence in the environment and their bioaccumulating properties leading to longterm effects. It was also found to be very difficult to combine the first and the third prerequisite, as the tests on fate and effects in the aquatic environment were much better established than the corresponding tests for the air and the terrestrial compartments. In the Nordic proposals, however, mammalian acute toxicity was included in order to cover a wider range of organisms and to address the terrestrial compartment to some extent. No criteria could be formulated for the air compartment since there were no generally recognized tests available at that time. In 1987 the framework for a classification scheme or model was elaborated by the Nordic group. There was, however, much uncertainty as to how this model would work in practice. Therefore it was decided that the Nordic model with a number of modifications should be tested, together with the corresponding models currently discussed at the CEC, on at least one hundred chemical substances and differences in the outcome should be analyzed. Figure 1 gives the general outline of the different models. The exercise also included different cut-off values for toxicity and bioaccumulation, inclusion or exclusion of algal and mammalian toxicity, etc., giving a total of 18 different models. The results and details of the procedure are reported elsewhere [ll. The different models arrived at surprisingly different numbers of substances classified for posing a potential hazard to the environment considering the great similarities between the models. With the most stringent model 79 substances out of 121 were classified, while 22 were classified with the least stringent model. The results were scrutinized by the SAG and the following conclusions were drawn: 1. The exclusion criterion for not bioavailable substances (MW > 1000, Fig. la) was excluded as it could not be evaluated. It was anticipated that this criterion would apply mainly to polymers, which for several reasons would need special consideration. 2. A strong tendency for a chemical to partition to the air compartment (PED,i, > 90%), calculated according to the level I fugacity model of Mackay and Paterson [21,was proposed to be an exclusion criterion in the Nordic model (Fig la). This was rejected as it was found that high volume chemicals, known to occur in the aquatic environment, would be excluded from further examination.
538
W
r---------t Mw>1ooo?
-cyesj
iL--------,4
I 1 i
0
1
rl
I
Ready
biodegr.?
I.
JNO
Ilwh4fhi
cl U-e
fbsfbssified
Very toxic?
yes+ Not Classified as Hazardous I
NO
J. I!??, insuff. 2T!Gzl1Bioacc?l’+2f3
r-l I
Classified as
verytoxic? -yes--_)Classified
as Hazardous
NO
Uxksikdw/
Ready biodegr.?l
NO
1
L&ass#Z
as Hazardous
+hssi&3
as Hazardous
Fig. 1. General outline of the tested models: (a>basic Nordic model; (b) modified Nordic model; (c) the CEC model; (d) the model suggested by Germany.
3. Mammalian acute toxicity added relatively little to the overall rate of classi~cation and can be excluded if only the aquatic com~ment is addressed.
539
4. The algal toxicity was shown to be a very sensitive component of an aquatic toxicity test battery and should be included. Besides it sensitivity it should be included as it will serve as an indicator of effects on photosynthetic organisms/primary producers. 5. Changing the aquatic toxicity threshold by a factor of 10 was the single most important modification of the models. The SAG found the more protective level to be appropriate considering the frequency of classification of substances of environmental concern. Even more important than the toxicity threshold was how the information on toxicity, biodegradability and bioaccumulation were combined. In all models, toxicity alone is regarded as a sufficient basis for classification if the substance is very toxic. In all other cases a combination of toxicity, biodegradability and bioaccumulation forms the basis for classification. Biodegradability is given much weight in many of the models as only not-readily biodegradable substances are considered for classification (see Figs. la, c and d). The reasoning is that readily-biodegradable substances will have a limited impact in time and space. This can, however, be questioned as it has been shown that other parameters (e.g. logK,,J can be more important predictors for occurrence in the environment [3,41. The model in Figure lb compensates for this to some extent as it gives equal weight to all three parameters. There are four possible combinations of two or more of the three properties: toxic, biodegradable and bioaccumulating. Only the model in Figure lb includes all four combinations. A unique combination for that model is “not-readily biodegradable+bioaccumulating”. Even if this combination did not substantially contribute to the rate of classification, the SAG stressed its importance in a classification scheme as it could serve as an indicator of possible long-term effects. Based on these conclusions and on the knowledge on environmental behaviour and effects of the substances selected for this comparison, the SAG recommended in December 1988 the model in Figure 2 for environmental classification of chemical substances. The consequences for the universe of chemicals of applying this classification model could not estimated as the selection of substances was not randomly made from all chemicals but was severely biassed towards environmentally dangerous substances [ll. Nevertheless, some idea of how frequent a classification will be can be given from a comparison with the frequency of classification for health hazards. This comparison showed that for the selected substances the rate of classification for environmental hazards was slightly lower than that for health hazards. The SAG found this rate satisfactory and in agreement with their general understanding of this concept.
540
Criteria: I---
Very toxic?
/YES_, Classified as Hazardous
Not classified as Hazardous
Threshold
vatues: 11Bioaccumu-/I Not ready bioii degradable jj lating
I
Oral LDm
~25 mgtkg
s 200 mg/kg 11 jj BCFrlOO
lnhal I-C,,
d 03 g/m3
-
1Fish LC&,
5 1,O mg/l
5 100 mg/i
!:
~~Asdefinedin 1;OECD Guidelines /or /i BODJCOD < 03
100 mgfl
ij ii I/
s 100 mg/l
/I
5
Fig. 2. The classification
91000
I
I j/
\ i 1
1 I
/
;I
/ ! I
model recommended by the Nordic Scientific Advisory Group.
Denmark, being a member of the EC, brought the recommended model to the working group at the CEC as a Danish proposal. This proposal was further developed in the CEC working group and in 1989 the working group concluded their work by presenting classification criteria that could form the basis for a system for classification and labelling of chemical substances (see Fig. 3). The major differences from the Nordic recommendation were: 1. Mammalian toxicity was excluded and only the aquatic en~ronment was addressed. 2. A third (intermediate) cut-off level for aquatic toxicity was introduced. 3. Substances that are “Harmful+ Bioaccumulating” were not regarded as hazardous to the environment. It should be noted that although the classification is composed of several components and combinations there is only one class for environmental
541
hazard: “Dangerous for the Environment”. It should also be noted that the criteria were developed for chemical substances and are not applicable to chemical preparations in their present form. SYSTEMS FOR CLASSIFICATION AND LABELLING
Once the basis was set in the form of specific classification criteria for the aquatic environment, the work continued with applying these criteria in administrative rules. Again, this was done both as a joint Nordic project and in a working group at the CEC. The final report from the Nordic project was published in July 1990 [51 and was formulated as recommendations on a system for classification, labelling and safety data sheets. The report covered four main areas: 1. Classification of substances: The classification of substances should be based on the classification criteria discussed above, but as the specific criteria do not cover all aspects of environmental hazard the specific criteria were supplemented by a general criterion: “Substances which are not classified as dangerous for the environment in accordance with the specific criteria but which, in the light of current knowledge regarding their toxicity, persistence and ability to accumulate and their calculated or observed behaviour and final fate in the environment, may constitute an immediate or long-term danger to the structure and/or function of a natural ecosystem should be classified as dangerous for the environment.” This general criterion has essentially the same wording as the general criterion elaborated by the CEC working group at that time. A list of examples on classified substances, continuously updated by the competent authority, would give further guidance to the manufacturers/importers of chemicals. 2. Classification of preparations: The classification of preparations should be determined by the properties of the single substances that the preparation contains. As a main rule, a preparation would be classified as dangerous for the environment if it contains 1% or more of a substance dangerous for the environment. Furthermore, a preparation would be classified if the preparation is very toxic, even if it does not contain 1% or more of a dangerous substance. 3. Labelling: Detailed proposals for how the information should be presented on the label were given in the report, as well as proposals for relevant risk- and safety-phrases. Regarding the hazard symbol a reference was given to the symbol proposed by the CEC. 4. Safety data sheets: The information on environmental, health and fire hazards should be coordinated. Besides the basic environmental hazard information on the label the safety data sheet should contain information on
542
risks and safety measures in free text and detailed data on LC&C~Ovalues, etc. The corresponding work in the CEC resulted in rules on environmental classification and labelling of chemical substances set out in the Commission Directive of 1 March 1991 adapting to technical progress for the twelfth time Council Directive 67/548/EEC [61. These rules cover classification and labelling of chemical substances (items 1 and 3 in the Nordic report), but do not address preparations or safety data sheets. The specific aquatic criteria form the basis for classification together with two general criteria, one for the aquatic environment and one for the non-aquatic environment. The directive also gives rules for the use of specified risk- and safety-phrases. In the latter part of March 1991 a seminar on environmental classification and labelling of chemicals was held in Uppsala, Sweden. The seminar was jointly sponsored by the CEC and the EFTA countries. Representatives from 17 of the 18 EC/EFTA countries and representatives from the CEC, the EFTA Office and the industry shared their experiences in this field and discussed different options for the future work. It was concluded 171that: - the experts from the EFTA and EC countries have a common understanding on the need for further elaboration of criteria, classification and labelling rules and safety data sheets on chemicals hazardous to the environment; - the present EC-regulations make a good basis for further work as well as e.g. the present Nordic proposals on different issues; - further work should be done in close cooperation between the EC and EFTA - the EFTA chemicals group and the CEC directorates DG III and XI should be prepared to organize the discussions needed for future cooperation subject to the outcome of the EES negotiations The situation today is that the regulations on environmental classification and labelling of chemical substances dangerous for the environment will enter into force no later than on 1 July 1992 in the EC countries and the EFTA countries will do the same as soon as possible. The same rules as in the EC countries will, e.g., be in force in Sweden from 1 January 1993. INTERNATIONAL HARMONIZATION OF CLASSIFICATION CRITERIA
The Uppsala seminar might be regarded as a first step towards international harmonization. The classification/labelling systems for chemical substances were discussed and coordinated although no formal decisions were taken. There are, however, several other areas, e.g. transport regulations, where classification criteria are or will be applied both on a national and international scale. Criteria for the aquatic environment have been
developed in some administrative systems, but have not yet been widely used. Criteria for other environmental compartments (soil, air, etc) are used even less. International harmonization at this early stage will help prevent the development of divergent systems which will require harmonization at a later stage. Along these lines, the OECD arranged a workshop on “Certain Aspects of the Classification of Chemicals” in Paris, November 1991. The workshop addressed two specific endpoints for classification: Acute oral toxicity, which is one of the oldest criteria for classification, and environmental hazard, which is one of the newest. It was concluded that international harmonization was desirable because it would increase the health and safety of chemical users, it would facilitate trade in chemicals, and it would reduce duplicative animal testing, and other work related to classification. The workshop recommended an ongoing OECD activity on harmonization of classification and labelling, starting with classification of acute oral toxicity and environmental hazard. At the Joint Meeting of the OECD Chemicals Group and the Management Committee the CEC, Sweden and the USA were given a mandate to initiate a Clearing House activity to examine existing systems and to report to the Joint Meeting on the feasibility of elaborating harmonized approaches with regard to these endpoints and to make recommendations to the Joint Meeting on possible options for the harmonization of existing schemes. To promote an efficient sharing of the burden with other UN organizations and UNCED 92 with respect to harmonization of classification schemes, it was agreed that OECD would participate in a proposed IPCS Coordinating Committee which also would involve ILO, WHO, UNEP, UNTDG and FAO. In February 1992 the leading countries discussed different classification criteria and drafted the proposal to the Joint Meeting at a meeting in Washington. This draft was open for discussions at an extended Clearing House meeting in Brussels in April, 1992. Experts from all OECD countries were invited to the Brussels meeting, together with representatives from relevant international organizations. Participants discussed the proposal and other issues related to harmonization. In particular it was noted that the IPCS Coordinating Committee in its work plan had given highest priority to harmonization of classification criteria for acute oral toxicity and hazard to the environment. The final proposal to the Joint Meeting was based upon the draft proposal of the Clearing House, but was modified to meet the concerns raised at the Brussels meeting. With regard to environmental hazard the Clearing House proposed: 1. Adoption of the current classification criteria as developed in the EC and The Nordic countries (see Fig. 3) as a starting point for common criteria for the aquatic environment.
classsfied
aa
Dangerous
clasaad
~nge~e
for the EnviroMnrwrt
the water solubiity of the substance i.skssthanlmg/L
Threshold values:
Fig. 3, Final classification scheme for the aquatic environment as developed in the EC and the Nordic countries and recommended for adoption by the OECD Clearing House.
2. Joint development of common criteria for other en~ro~en~l compartments and possibly for ozone depleting substances. 3. Joint development of procedures for updating and methodologies to take account of the cl~si~cation of preparation and those situations where additional information is available on the aquatic environment. In May 1992 the Joint Meeting agreed to the Clearing House proposal and urged the Clearing House to continue its work as proposed.
545
Harmonization of classification and labelling of chemicals is also a programme area under Chemical Forum that was agreed upon at UNCED 92. This will probably provide the necessary political pressure to bring these issues to a successful end within a reasonable period of time. REFERENCES Lundgren, A. (1989) Comparison of different models for environmental hazard classification of chemicals. Kern1 Report No 9/89. Swedish National Chemicals Inspectorate. Stockholm, 40 pp. Mackay, D. and Paterson, S. (1982) Fugacity revisited. Environ. Sci. Technol. l6,654A660A. Ikeda, M., Koizumi, A., Kasahara, M., Watanabe, T., Nakatsuka, H. and Sekita, Y. (1987) A statistical approach to the prediction of possible presence of pollutant chemicals in the environment. Regulatory Toxicology and Pharmacology 7(3), 321536. ECETOC (1988) Concentrations of industrial organic chemicals measured in the environment: The influence of physico-chemical properties, tonnage and use pattern. ECETOC. Technical Report No 29. Brussels. 105 pp. Gustafsson, L. and Ljung, E. (1990) Substances and preparations dangerous for the environment: A system for classification, labelling and safety data sheets. Miljerapport 1990:lOE. Nordic Council of Ministers. Copenhagen. 65 pp. Commission of the European Communities (1991) Commission Directive of 1 March 1991 adapting to technical progress for the twelfth time Council Directive 67/548/EEC (91/325/EEC). Official Journal of the European Communities. No. L 180 of 8.7.91 Freij, L. (Ed.) (1991) Seminar on Environmental Classification and Labelling of Chemicals. EC-EFTA Meeting, Uppsala, Sweden. March 20-21,199l. Swedish National Chemicals Inspectorate. Stockholm. 85 pp.
