Adapting the EU Seveso II Directive for the

Transcription

1 Adapting the EU Seveso II Directive for the Gobay Harmonised System of Cassification and Labeing of Chemicas (ghs) in Terms of acute Toxicity to Peope: Initia Study into POTENTIAL Effects on UK Industry Mary Trainor 1, David Bosworth 2, Anna Rowbotham 1, Ji Widay 1, Susan Fraser 1 and Ju Lynne Saw 1 1 Heath and Safety Laboratory, Harpur Hi, Buxton, Derbyshire, SK17 9JN, UK 2 Heath and Safety Executive, Redgrave Court, Merton Road, Boote, L20 7HS, UK Crown Copyright This artice is pubished with the permission of the Controer of HMSO and the Queen s Printer for Scotand Within the EU, the risks of major accidents from chemica instaations are reguated under the Seveso II Directive. This paper describes an initia study into potentia impications for reguation of UK instaations arising from changes to the cassification of acute toxicity to peope when the EU adopts the Gobay Harmonised System of Cassification and Labeing of Chemicas (GHS). The study s aim was to identify a means of adapting the Seveso II Directive for GHS that: wi not increase the Directive s scope and attention uness this increases safety from major accidents; wi not increase the risk of a major accident by creating gaps in the reguation of instaations; and wi be transparent and straightforward for industry to appy. The outcome was to identify a possibe option, the Simpe Aignment, whereby references to the EU cassifications Very Toxic and Toxic are repaced by GHS acute toxicity hazard Category 1 and Category 2 respectivey for a exposure routes and physica states. To prevent reguatory gaps, the adapted Seveso II Directive woud incude further Named Substances, such as the ower moecuar weight gases ammonia and suphur dioxide, which have a ess severe GHS acute toxicity category but are currenty in the Seveso II regime and correspond to instaations with major accident hazard potentia that woud not otherwise fa within the scope of the Directive. These substances woud be identified using Technica Criteria that coud, for exampe, be used by an EU Technica Committee to incude further Named Substances in future. The other options considered were rejected either because of cost or because of the potentia to significanty increase the scope of Seveso II. The outcome of this initia study, together with work by the German and Dutch Seveso II reguatory authorities, is being taken forward through an EU Technica Working Group. Keywords: GHS, Seveso II Directive, acute toxicity, major accident

2 Introduction: the Seveso II Directive and the Gobay Harmonised System of Cassification and Labeing of Chemicas In the EU, the risks of major accidents from chemica instaations are reguated through the Seveso II Directive (96/82/EC as amended) for the Contro of Major Accident Hazards Invoving Dangerous Substances [ECC, 1997 & 2003]. The Directive covers accident prevention and mitigation. Seveso II appies to estabishments where dangerous substances may be present or generated in quantities in excess of specified threshod tonnages the Quaifying Quantities. The status of reguated estabishments is either ower-tier (Directive Artices 6 and 7 appy) or the more highy reguated top-tier (Artice 9 additionay appies) depending on whether ower or higher Quaifying Quantities appy. The Quaifying Quantities differ according to which of the Seveso II Dangerous Categories the dangerous substances fa into on the basis of their cassification and whether they are Seveso II Named Substances. There are ten Dangerous Categories, they reate either to substances : physico-chemica properties such as fammabiity and exposivity, toxicity to peope, or toxicity to the aqueous environment. This simpe threshod tonnage approach operates as an approximate screen to determine the appropriate degree of reguation of estabishments under Seveso II. The screen is approximate since off-site risk in the vicinity of any specific instaation depends on factors such as: a substance s packaging or containment and inherent physica properties such as vapour pressure; the process and storage conditions; and the geography of the oca area. The approximate nature of the screen is expicity recognised in Seveso II in so far as an instaation may be granted a derogation exempting the operator from preparing a fu Seveso II safety report if there is no major accident hazard potentia. At present, the basis of cassification of substances and mixtures (preparations) is the EU s cassification system according to the provisions of The Cassification, Packaging and Labeing of Dangerous Substances Directive, CPL (67/548/EEC as amended) and The Cassification, Packaging and Labeing of Dangerous Preparations Directive (99/45/ EC as amended) [ECC, 1967 & 1999]. Approximatey five thousand substances are isted in Annex 1 of CPL with a harmonised cassification that is egay binding in the EU. Other substances must be sef-cassified by the suppier or Seveso II instaation operator. The EU is repacing this cassification system by the Gobay Harmonised System of Cassification and Labeing of Chemicas (GHS) [ECC 2007], [UN, 2005]. At an internationa eve, it is anticipated that major benefits of adopting GHS wi incude: Major accident `sha mean an occurrence such as a major emission, fire, or exposion resuting from uncontroed deveopments in the course of the operation of any estabishment covered by this Directive, and eading to serious danger to human heath and/or the environment, immediate or deayed, inside or outside the estabishment, and invoving one or more dangerous substances [ECC, 1997]. 2

3 reducing cassification costs to industry by having a singe system in use; increasing the consistency and transparency of those pubic protection eves that are based on cassification of chemicas [ECC, 2006a]; and reducing anima testing [UN, 2005]. The adoption of GHS at EU eve is a major endeavour because there is not a oneto-one correspondence between GHS and the current EU cassification system. Seveso II is ony one of over twenty reguations that wi potentiay be affected. A proposed reguation on `Cassification and Labeing of Substances and Mixtures based on the Gobay Harmonised System was pubished in August 2006 [EEC, 2006a] and updated in June 2007 [EEC, 2007] foowing stakehoder consutation. The currenty proposed EU timescaes for the adoption of GHS are for cassifications of substances to be mandatory from Dec 2010 and of mixtures from June 2015 [Bierman, 2007]. The timescaes for GHS are being coordinated with those for the introduction of the new EU reguatory framework for chemicas caed REACH (Registration, Evauation and Authorisation of Chemicas) under which enterprises that manufacture or import more than one tonne of a substance per year wi be required to register it in a centra database [European Commission, EIDG & EGD, 2007]. The registration process wi incude reguatory scrutiny of the substance cassifications submitted by enterprises. An EU ad hoc Technica Working Group on Seveso II and GHS (TWG) is considering the impications for Seveso II when GHS is adopted. Essentiay, the differences between the EU and GHS cassification systems mean that if GHS cassifications are used there is a potentia for changes to: the scope of Seveso II where estabishments move between being reguated under Seveso II and not being reguated under Seveso II, or vice versa; and the reguatory attention of Seveso II where estabishments move between ower-tier and the more highy reguated top-tier status or vice versa. Aims of Initia Study into Impications for UK industry of the options for Seveso II in terms of Acute Toxicity to Peope when GHS is adopted This paper describes an initia study into the impications for UK industry of the options for Seveso II in adopting GHS substance cassifications for acute toxicity to peope. (The study s remit did not incude cassification of mixtures or cassification for toxicity to the aquatic environment and physico-chemica properties.) The study was carried out by the Heath and Safety Laboratory, HSL, working with, and on behaf of, the Heath and Safety Executive, HSE, which is the ead UK Competent Authority (reguator) for those aspects of Seveso II which reate to harm to peope. The objective was to inform the considerations of a group of interested EU Member State Competent Authorities drawn from the EU TWG. The overa aim of this initia study was to identify a means of adopting GHS for the Seveso II Directive in terms of acute toxicity to peope that: 3

4 wi not increase the Directive s scope and reguatory attention uness this increases safety from major accidents since this woud pose a needess cost burden on the chemica industry 2 and diute the UK reguatory effort; wi not increase the risk of a major accident by creating gaps in the reguation of instaations; and wi be transparent and straightforward for industry to appy. The Options for seveso II when GHS is adopted and ApproacheS used to Study Them At EU eve, two approaches are under consideration for Seveso II in terms of acute toxicity to peope when GHS is adopted: The first approach, the Dua Cassification Option, is to continue to use the current EU cassification system to determine the Seveso II Dangerous Category of a substance, whist requiring industry to cassify by GHS under other EU egisation. This option was proposed in [ECC, 2006b]. It is of interest because there woud be no change to the reguation of instaations under Seveso II. The second approach is to repace the EU cassification system by GHS using one of the possibe Aignment Options whereby references in Seveso II to EU cassifications are repaced by references to specified GHS cassifications. HSE s view is that the above ong-term Dua Cassification Option is not acceptabe for Seveso II because using two cassification systems in parae woud present an additiona cost to both industry and EU Member State reguators compared to using GHS cassifications aone. To assess the possibe Aignment Options, a two-part approach was used. The primary approach was to anayse the impications for UK instaations based on consideration of: the operation of the Seveso II Aggregation Rue and Quaifying Quantities; and the differences between the EU and GHS cassification systems for acute toxicity to peope. The second, suppementary, approach was an initia study into the impications at a substance-by-substance eve; it is imited by two confounding factors: 1. ack of knowedge of the GHS cassifications that wi be in use for individua substances within the EU since these wi be made, at a future date by industry 3 ; and 2 The costs to UK industry of compying with Seveso II (which is impemented in the UK through the COMAH Reguations) are estimated in [Brazier, 2003]. For exampe: the cost of anaysis and safety report writing starts at about 35k (approximatey 50k euros) for storage and warehouse instaations, rising to about 220k (approximatey 300k euros) for petroeum refineries; and for a fifth of companies considered, Seveso II safety report preparation diverts resources away from other safety activities. 3 Under the proposed GHS reguation [ECC, 2007] enterprises woud cassify substances on the market by the end of the transitiona period for substances; this is currenty proposed to be Dec 2010 [Bierman, 2007]. This GHS cassification does not require substance testing: it woud be notified to the EU Chemicas Agency 4

5 2. the extreme difficuty of identifying substances that are not currenty within the Seveso II regime 4 in terms of acute toxicity to peope, but may be brought in depending on how Seveso II is adapted for GHS, since at present they have no Seveso II reguatory significance and are therefore not isted in any Seveso II reated database. Instaations Faing in the Scope of Seveso II in Terms of Substance Cassifications for Acute Toxicity to Peope Two of the ten Seveso II Categories of Dangerous Substance reate to acute toxicity to peope: the Toxic and Very Toxic Categories. A substance fas in these categories if its overa EU cassification (the most severe of the cassifications for the ora, derma and inhaation exposure routes) is Toxic (T) or Very Toxic (T + ). Tabe 1 shows the Seveso II Quaifying Quantities for these Categories of Dangerous Substances. Dangerous substances present at an estabishment in quantities greater than 2% of the reevant Quaifying Quantity need to be considered - the Aggregation Rue. The Aggregation Rue aso appies to Seveso II Named Substances for the reevant Categories of Dangerous Substances. Tabe 1 aso shows the Quaifying Quantities for two exampes of Named Substances that are acutey toxic. In practice, in the UK many of the Seveso II instaations that meet the Quaifying Quantity conditions for the Toxic or Very Toxic Categories of Dangerous Substances, do so on the basis of the Aggregation Rue. For exampe, instaations manufacturing pharmaceuticas or agrochemicas tend to produce a range of substances with overa cassification as T or T +. Simiary, some instaations aso fa within the scope of Seveso II because they meet the Quaifying Quantity conditions for one or more Categories of Dangerous Substances reating to physico-chemica properties such as fammabiity: exampes incude refineries, and some manufacturing pants using toxic substances as intermediates. Hence, for many UK Seveso II instaations, any potentia for a reduction in reguatory status and attention arising from a reduction in the severity of the acute toxicity cassification of some substances, wi in practice be offset by the operation of the Aggregation Rue and their Seveso II status for Dangerous Categories reating to physico-chemica properties. However, the converse is not the case: increases in the reguatory status of instaations may arise from an increase in the severity of the acute toxicity cassification uness a substance has aready been registered through the REACH egisation. Thereafter, the cassification may change, for exampe when a substance is newy registered through REACH (when the cassification wi be subject to reguatory scrutiny). `It is anticipated that for some substances the cassifications wi vary. Over time, it is expected that notifiers and registrants wi agree on a singe entry [cassification] [ECC, 2007]. 4 For ease of reference, we refer to any substances that either fa into one of the Seveso II Dangerous Categories, or are named in Seveso II, as faing within the Seveso II regime whether or not they ead to instaations faing within the scope of Seveso II. 5

6 Tabe 1. Seveso II quaifying quantities for: the toxic and very toxic categories of dangerous substances not named in the directive, and two exampes of named substances Dangerous substances Quaifying quantity in tonnes of dangerous substances Lower-tier (Seveso II Artices 6 and 7 appy) Top-tier (Seveso II Artice 9 additionay appies) Very Toxic Category 5 20 Toxic Category Phosgene (a Very Toxic Named Substance) Chorine (a Toxic Named Substance) of some substances, and in some instances this effect may be ampified by the operation of the Aggregation Rue. Overview of the EU and GHS Cassification of acute toxicity to peope Conceptuay, the EU and GHS cassification systems are broady simiar in terms of acute toxicity to peope except in the treatment of inhaation exposures to substances cassified as gases under GHS. Both systems assign substance cassifications based on their acute toxicity foowing exposure via the ora, derma and inhaation routes. For inhaation exposures, the physica state of a substance is taken into account: that is to say whether it is cassified as a gas or vapour, or as an aeroso or particuate. Both systems rank the acute inhaation toxicity of substances cassified as vapours, and as aerosos or particuates, in terms of the mass inhaed in a given voume. However, unike the EU system, the GHS system makes a cassification distinction between vapours and gases. For inhaation exposures to those substances cassified as gases under GHS 5, the EU and GHS systems are fundamentay different: the EU system is set up to rank acute inhaation toxicity in terms of the mass inhaed in a given voume, whereas the GHS system is set up to rank acute inhaation toxicity in terms of the number of moecues inhaed in a given voume. 5 Substances cassified as gases under GHS are those for which the test atmosphere is a gas or a vapour near the gaseous state [UN, 2005]. A `vapour near the gaseous state is not defined. No reason is given for cassifying gases and vapours (gases in contact with the iquid or soid state) differenty. 6

7 Tabe 2. 4hr LC 50 and LD 50 acute toxicity cassification boundaries used under the GHS and EU cassification systems for each exposure route/ physica state combination Exposure route/physica state Ora Derma Inhaation Aerosos & Particuates (GHS terminoogy mists & dusts) Vapours Gases (EU as for vapours) Definitions of 4hr LC 50 or LD 50 (with units) used to set acute toxicity cassification boundaries LD 50 mass fraction (mg/kg) LD 50 mass fraction (mg/kg) 4hr LC 50 mass fraction (mg/) 4hr LC 50 mass fraction (mg/) 4hr LC 50 mass fraction (mg/) EU ony 4hr LC 50 voume fraction (ppmv)* GHS ony Cassification boundaries EU cassification GHS cassification <5 T to to 50 T 50 to to 300 Xn 300 to <50 T T Xn <0.05 T T Xn 4 <0.5 T T Xn <0.5 T T 2 20 Xn < * For individua substances the conversion factor is: 4hr LC 50 mg/ = 4hr LC 50 ppmv Moecuar Weight g/mo 24,450. Tabe 2 shows the LD 50 and 4hr LC 50 6 boundaries used to cassify a substance under the EU system as T +, T, or the ess severe Harmfu (Xn), and under GHS as Category 6 For a particuar species, the LD 50 is the dose that wi ki 50% of the exposed popuation whist the LC 50 is the equivaent airborne concentration for a specified exposure period. 7

8 (Cat) 1, 2, 3 or 4 of which Cat 1 is the most severe. It can be seen that the correspondences between the boundaries fa into three groups: 1. For inhaation exposures to substances cassified as vapours, the boundaries for the EU T and T + and GHS Cat 1 and Cat 2 cassifications are identica. 2. There is a straightforward shift in some boundaries for derma and ora exposures and for inhaation exposures to aerosos. Therefore, for these exposures, an aignment can either be chosen for which substances may move to a ess severe cassification but not a more severe one, or aternativey where substances may move to a more severe cassification but not a ess severe one. 3. There is a correspondence depending on moecuar weight for inhaation exposures to substances cassified under GHS as gases but as vapours/gases under the EU system; this is iustrated in Figure 1. It can be seen that some ower moecuar weight substances that have inhaation cassification as T or T + in the EU system, wi not have a severe GHS cassification (GHS Cat 1 or 2). Exampes are the industriay important substances ammonia, suphur dioxide, and ethyene oxide (a Seveso II Named Substance). Conversey, some higher moecuar weight gases that are not cassified as T or T + in the EU system wi have a reativey severe GHS cassification (GHS Cat 2). Moecuar weight P Cat 1 HF T+ T Xn 4hr LC50 = 100 ppmv EO SD A hr LC50 (mg/) Figure 1. The 4hr LC 50 boundaries used to define acute toxicity inhaation cassifications for substances cassified in the EU system as gases/ vapours but as gases under GHS: shown by dashed ines for EU T +, T and Xn, and by fu ines for GHS Cat 1, 2, 3 or 4. Aso shown are the 4hr LC 50 vaues of four exampe ower moecuar weight substances 8 A = ammonia EO = ethyene oxide HF = hydrogen fuoride P = phosphine SD = suphur dioxide 4hr LC50 = 500 ppmv 4hr LC50 = 2500 ppmv Cat 2 > Cat 3 Cat 3

9 OUTCOME OF PRIMARY ANALYSIS: THE SIMPLE ALIGNMENT OPTION WITH TECHNICAL CRITERIA Based on the above, we considered the suitabiity of various aignments incuding the CPL and Precautionary Aignments discussed beow. We identified a possibe option that meets the study s aims, the: Simpe Aignment where references to the EU T + and T cassifications are repaced by GHS acute toxicity Cat 1 and Cat 2 for a exposure routes and physica states. We refer to this as GHS Cat 1 being GHS T + -equivaent, and GHS Cat 2 being GHS T-equivaent. This is suppemented by Technica Criteria to be used to retain other substances with a ess severe GHS category that are currenty within the Seveso II regime and correspond to instaations with major accident hazard potentia that woud otherwise fa outside the scope of the Directive. This coud, for exampe, be impemented by the addition of extra Named Substances in the Directive, and the use of an EU Technica Committee to incude further Named Substances thereafter. Our reasoning for this option, in order to best meet the UK aims stated above for adapting Seveso II for GHS, and assuming the EU uses the GHS cassification distinction between gases and vapours, is as foows: 1. This aignment is the same for a physica states for inhaation exposures. Therefore, knowedge of the physica state used for cassification is not needed thus maximising both the ease of use and transparency of this aignment. 2. This aignment minimises the potentia for substances to move to a more severe equivaent cassification, hence minimising the potentia for increases in scope and oversight of Seveso II. This potentia arises ony for some higher moecuar weight substances cassified under GHS as gases where changes from T to GHS T + -equivaent, or Xn to GHS T-equivaent are possibe. (We are considering whether further technica criteria coud be used to address this.) 3. With the exception of these higher moecuar weight gases, this aignment means that substances may move to a ess severe equivaent cassification but not to a more severe one. Hence, with this exception, the reguatory scope and attention of Seveso II cannot increase. The opportunity for gaps in reguation to arise is imited by the operation of the Aggregation Rue and Dangerous Categories of Substances reating to physicochemica properties. The use of the Technica Criteria as described above woud act as a safety net to ensure that such gaps cannot arise. PRIMARY ANALYSIS OF THE CPL AND PRECAUTIONARY ALIGNMENTS We aso considered the suitabiity of two other Aignments Options that have been of interest at EU eve: the CPL Aignment and the Precautionary Aignment. The CPL Aignment Option was the initia EU proposa for cassification and abeing purposes [ECC, 2006a]. It differs from the Simpe Aignment in aigning T with 9

10 GHS Cat 2 and 3 for the ora route and inhaation exposures to aerosos. It has the drawback of requiring knowedge of the physica state used for inhaation cassifications. It aso has significant potentia to increase the reguatory scope and oversight of Seveso without increasing safety from major accidents, by bringing Xn substances into the Seveso II regime through GHS cassifications for the ora route. The proposa has now been dropped pending review [ECC, 2007]; we do not consider it further. The Precautionary Aignment Option aigns Seveso T + with GHS Cat 1 and Cat 2, and Seveso T with GHS Cat 3 for a exposure routes and physica states. It is of interest because it is the most straightforward aignment that woud not resut in a reduction in the scope of Seveso II because substances coud not, in practice, move out of the Seveso II regime. Like the Simpe Aignment, it has the advantage that the aignment is independent of the physica state for inhaation exposures. However, this aignment does not meet the UK s aims because it has significant potentia to increase the reguatory scope and oversight of Seveso without increasing safety from major accidents. For exampe, from Tabe 2 and Figure 1, it can be seen that Xn substances may be newy brought within the Seveso II regime as GHS-T equivaent through GHS cassifications for the ora and derma routes and inhaation exposures to vapours and higher moecuar weight gases, or as GHS T + -equivaent through GHS cassifications for higher moecuar weight gases. INITIAL SUPPLEMENTARY STUDY INTO IMPACT ON UK INDUSTRY: SUBSTANCES, CLASSIFICATIONS, AND ASSESSMENT OF IMPACT ON INSTALLATIONS FORMING BASIS OF STUDY The initia suppementary study into potentia impact on UK instaations considered two groups of substances: 1) Substances currenty cassified as T or T + that are important in the UK in terms of Seveso II - the UK Seveso T and T + substances. 2) Substances that are not currenty cassified as T or T + but coud newy be brought into the Seveso II regime under the Simpe or Precautionary Aignment Options. To attempt to identify candidates, we trawed the EU High Production Voume Chemicas, HPVCs. These are the approximatey 2,500 chemicas that were on the European market before September 1981 and are produced or imported in quantities exceeding 1,000 te per year. Data provided by manufacturers and importers on HPVCs such as tonnage and toxicity is hed on the IUCLID database [Hansen, 1999]. HPVCs do not incude a the substances of interest (exampes are ow production substances such as intermediates and reagents, or substances that have ony been high production voume since 1981). We trawed approximatey 1,300 HPVCs: a those with an EU harmonised cassification that means they may potentiay be brought within the Seveso II regime; and the approximatey 40% of highest production voume 7. 7 We used a ist provided by the European Chemicas Bureau, ISPRA, in October 2006 of HPVCs ordered in groups of descending tonnage voume using atest reported voumes from manufacturers. 10

11 A confounding factor in identifying the second group of substances, and in using both groups for the study, is ack of knowedge of the GHS cassifications that wi be made by EU industry subject. We assigned reativey rapid informa GHS substance cassifications (see Appendix) which without doubt wi differ from the detaied industry GHS cassifications for some substances. The informa GHS cassifications were assigned to: 71 (30%) of the UK Seveso T, T + substances. This was done using readiy avaiabe toxicoogica data. (See Appendix for the data sources used and the constraints on substances that coud be incuded.) 29 candidate substances that may be newy brought into the Seveso regime under the Precautionary Aignment. This was done using industry toxicity data from the IUCLID database. Substances with an EU harmonised cassification that is inconsistent with this data were excuded from the study. No candidates were found that woud be brought in under the Simpe Aignment. The reguatory impact on UK instaations arising from cassification changes to these substances was considered in a series of three meetings with HSE speciaists. These were informed by Internet information on industria use of the substances, together with HSE information on tonnages at specific UK instaations for some substances. The aim was to identify changes to the scope or reguatory attention of Seveso II. We did not aim to, and coud not, identify borderine estabishments such as those which are borderine top-tier and woud become borderine ower-tier. We do not consider that such cases have reguatory significance given that the Seveso II Quaifying Quantities act as an approximate screen ony. INITIAL STUDY INTO IMPACT ON UK INDUSTRY OF PRECAUTIONARY ALIGNMENT OPTION: OUTCOME Two of the candidate Xn substances that may be newy brought into the Seveso II regime under the Precautionary Aignment based on industry toxicity data in IUCLID were found to have UK reguatory significance: sodium dodecy suphate [CAS ] which is used in detergents and foamy persona hygiene products ike shampoo, shaving foam and bubbe bath; cacium diproprionate [CAS ] which is used as a moud inhibitor in processed foods such as cheeses, non-acohoic drinks, confectionaries and some meat products, as we as in ivestock and poutry feeds. As a resut, some UK formuators of processed foods, anima feeds, detergents or frothy persona hygiene products might become Seveso II sites. (A formuator bends ingredients to make fina products: therefore a site may have stock tanks or other reativey arge storage of ingredients.) 8 See ist at HSE uses this in connection with: the assessment of Seveso safety reports, and the provision of advice on and-use panning in the vicinity of instaations. 11

12 Our traw of candidate substances from IUCLID wi ony have identified a sma fraction of the substances that might be newy brought into the scope of Seveso. Hence, whist the GHS cassification assigned by industry in future may differ for these two exampe substances, our view is that this initia study confirms the potentia for the Precautionary Aignment to widen the scope of the Seveso II Directive without giving an increase in safety from major accidents. INITIAL STUDY INTO IMPACT ON UK INDUSTRY OF SIMPLE ALIGNMENT OPTION: OUTCOME For the 71 UK T, T + substances considered, the effect of the Simpe Aignment on overa cassification is that: between 29% and 43% of overa T + cassifications drop to overa GHS T-equivaent (between 15 and 22 substances out of 51); and between 30% and 45% of overa T cassifications drop to overa GHS-equivaent cassification ess than T we refer to this as GHS sub-t equivaent (between 6 and 9 out of 20 substances). We quote a range because for some substances the overa cassification is dependent on the physica state assumed for inhaation exposures. No exampes were found of substances moving from T to GHS T + -equivaent. Tabe 3 gives the cassifications for exampe substances and their current proposed EU harmonised GHS cassification 9. The cassification changes were ony found to have reguatory significance for two substances: the ower moecuar weight gases ammonia and suphur dioxide. As a resut of their overa cassification change from T to GHS sub-t equivaent, UK instaations with major accident hazard potentia woud fa outside the scope of Seveso. To address this, they woud be made Named Substances. No other adverse reguatory impact was found. For one substance, it was found that there woud be a reduction in the scope of Seveso but that this woud be beneficia as the reduction appies to instaations with no major accident hazard potentia to peope 10. For one substance, information on inventories at UK instaations is imited but there is none to suggest that there woud be an unacceptabe reguatory impact. For the remaining 9 The current proposa [ECC, 2007] is that EU harmonised acute toxicity cassifications woud be transated to minimum GHS cassifications except for T + for derma exposures which transates directy to GHS Cat 1. Industry woud increase the cassification over this minimum where appropriate. An aternative proposa of interest at EU eve is maximum harmonised GHS cassifications that industry woud decrease where it can be demonstrated that this is appropriate. 10 Potassium dichromate: some UK surface engineering industry estabishments are brought into the scope of Seveso II soey on the basis of this substance s inventory. Under the Simpe Aignment they woud be out of scope. gives an HSE anaysis showing that there is no off-site major accident hazard potentia in terms of acute toxicity to peope for the quantities of potassium dichromate typicay stored. 12

13 Tabe 3. Effect of simpe aignment on the overa cassification for exampe substances with breakdown by Ora (O), Derma (D), and Inhaation (I) routes. Listed are the: current EU cassification; informa GHS cassification foowed by its EU equivaence under the simpe aignment; and proposed harmonised (H) GHS cassification marked * where a minimum Substance name Anhydrous ammonia Ammonium dichromate as CrVI Methyene dithio-cyanate Substance CAS No. Cassification O D I Over-a Effect of simpe aignment on overa cassification EU T T Change: EU T to GHS Xn- equivaent based on informa GHS cassification GHS 3 3 GHS-equiv Sub-T Sub-T H-GHS 3* 3* EU T Xn T + T + Change: EU T + to GHS T- equivaent based on informa GHS cassification GHS GHS-equiv Sub-T Sub-T T T H-GHS 3* 4* 2* 2* EU T T + T + No change: EU T + to GHS T + - equivaent based on informa GHS cassification GHS GHS-equiv Sub-T T + T + H-GHS 3* 2* 2* 13

14 substances, there is no change in reguatory attention or scope of corresponding instaations due to the operation of the Aggregation Rue and the Dangerous Categories reated to physico-chemica properties. No substances were found that woud be newy within the Seveso remit. As described above, the potentia for this to occur is imited, arising ony through GHS inhaation cassifications for some higher moecuar weight gases. Our view is that this initia suppementary study supports the concusion that the Simpe Aignment with Technica Criteria is a suitabe option and identifies ammonia and suphur dioxide as exampes of substances that woud need to be Named Substances in the adapted Seveso II Directive. Further work is underway at HSL to deveop the Technica Criteria. STATUS OF STUDY AND NEXT STEPS In March 2007, this initia UK study was disseminated to the group of interested Member States drawn from the EU TWG. Together with studies from the Netherands and Germany it formed the basis of discussion on acute toxicity at the group s 2 nd meeting in September It is anticipated that at the 3 rd meeting in November 2007, an agenda wi be mapped out for technica sub-groups to consider the issues for toxicity to peope, the aqueous environment and physico-chemica properties in order to faciitate data sharing and pooing of expertise between Member States. ACKNOWLEDGMENTS The study described here was carried out working cosey with HSE experts in the fieds of process safety, toxicoogy, and Major Hazards poicy. We are particuary gratefu to Peter Ridgway for his vauabe advice and insights into substance cassification, and to Sandra Ashcroft, Tim Beas, Andrea Caitens, Dave Carter, Richard Cary, Steve Porter, Raph Rowands and Kirstin Wattie. We aso thank Oe Nørage of the European Chemicas Unit, ISPRA, for hepfu information from IUCLID on HPVCs ranked by tonnage. REFERENCES Bierman T., 2007, Seveso and GHS, Presentation on behaf of European Commission Environment Directorate Genera at Informa GHS and Seveso Meeting, HSL, Buxton, UK, Brazier A. and Waite. P, 2003, Safety Report Regime Evauating the Impact on New Entrants to COMAH, HSE Research Report 092, ISBN , HSE Books. ECC (European Communities. Commission), 1967, Counci Directive 67/548/EEC on the Approximation of Laws, Reguations and Administrative Provisions Reating to the 14

15 Cassification, Packaging and Labeing of Dangerous Substances, O.J. European Communities , L196: ECC (European Communities. Commission), 1991, Counci Directive 91/414/EEC of 15 Juy 1991 Concerning the Pacing of Pant Protection Products on the Market, O.J. European Communities , L230: ECC (European Communities. Commission), 1993, Counci Directive (EEC) 793/93 of 23 March 1993 on the Evauation and Contro of the Risks of Existing Substances, O.J. European Communities , L84: 1 8. ECC (European Communities. Commission), , Counci Directive 96/82/EC of 9 December 1996 on the Contro of Major Accident Hazards Invoving Dangerous Substances, O.J. European Communities, L10: ECC (European Communities. Commission), 1999, Directive 99/45/EC of the European Pariament and of the Counci of 31 May 1999 Concerning the Approximation of the Laws, Reguations and Administrative Provisions of the Member States Reating to the Cassification, Packaging and Labeing of Dangerous Preparations, O.J. European Communities L200:1 68. ECC (European Communities. Commission), , Directive 2003/105/EC of the European Pariament and of the Counci of 16 December 2003 Amending Counci Directive 96/82/EC on the Contro of Major-Accident Hazards Invoving Dangerous Substances, O.J. European Union, L345: ECC (European Communities. Commission), 2006a, Proposa for a Reguation of the European Pariament and of the Counci on Cassification and Labeing of Substances and Mixtures Based on the Gobay Harmonised System, August ECC (European Communities. Commission), 2006b, Anaysis of the Potentia Effects of the Proposed GHS Reguation on its EU Downstream Legisation, August ECC (European Communities. Commission), 2007, Proposa for a Reguation of the European Pariament and of the Counci on Cassification and Labeing of Substances and Mixtures Based on the Gobay Harmonised System, June European Commission. EIDG & EDG (Enterprise and Industry Directorate Genera and Environment Directorate Genera), Feb 2007, REACH in Brief, enterprise/reach/docs/reach_intro.htm. Hansen B.G., van Haest A.G., van Leeuwen K. and van der Zandt P., 1999, Priority Setting for Existing Chemicas: European Union Risk Ranking Method, Environ. Tox. and Chem. 18 (4): Ruden C. and Hansson S.O, 2003, How Accurate are the European Union s Cassifications of Chemica Substances, Toxicoogy Letters 144: UN, 2005, Gobay Harmonized System of Cassification and Labeing of Chemicas (GHS) First Revised Edition, ST/SG/AC.10/30/Rev. 1, UN. Wood M., Pichard A., Gundert-Remy U., de Rooij C. and Tissot S., 2006, The AETL Methodoogy as a Potentia Soution to Current Chaenges Associated with the Deveopment and Use of Acute Exposure Leves in Seveso II Appications, J. Haz. Mat. A133:

16 gossary Aeroso (mist): iquid dropets of a substance or mixture suspended in a gas (usuay air) [UN, 2005]. Particuate (dust): soid partices of a substance or mixture suspended in a gas (usuay air) [UN, 2005]. ppmv: parts per miion by voume (cm 3 /m 3 ). Vapour: the gaseous form of a substance or mixture reeased from its iquid or soid state [UN, 2005]. Appendix: Assignment of Informa GHS Cassifications For the purposes of this study, informa GHS cassifications were assigned without carrying out the detaied checks and data gathering that woud, for instance, form part of the work of EU harmonised cassification assignment. For exampe: we did not check that data came from vaid we-performed tests; where information was incompete we did not request further detais from the source; and we ony considered data on the notiona preferred test species rather than carrying out a fu evauation of a avaiabe experimenta anima data. For those substances that have an EU harmonised cassification, we generay coud not use the corresponding LC 50 and LD 50 as this information was not retained in the earier years of the programme. (The need to pubish the scientific motivations of cassifications was proposed in [Ruden, 2003] which discusses the accuracy of harmonised cassifications.) For inhaation exposures, assigning the physica state of the test atmosphere appropriatey is non-trivia for substances that are iquids at ambient conditions: it may not be specified in the avaiabe account of the test, and testing may have been conducted using a mixture of physica states. Therefore, we considered a possibe physica states. For instance, based on the manufacturer s toxicity data in IUCLID, tridemorph (CAS ), has overa GHS acute toxicity Cat 4, Cat 3, or Cat 2 according to whether the aeroso, vapour or gas state is assumed. For the UK T, T + substances, the data sources used were toxicoogica reviews incuding: CICADS 11, ATSDR 12, OECD SIDS 13, EU Risk Assessment Reports on HPVCs under reguation 793/93/EEC [ECC, 1993], WHO Environmenta Heath Criteria reports 14, draft Technica Support Documents prepared by HSE for the EU ACUTEX project [Wood, 2006]; and UK Pesticide Safety Directorate substance evauations carried 11 CICADS: `Concise Internationa Chemica Assessment Documents Internationa Programme on Chemica Safety IPCS Co-operative Programme of WHO/ILO/UNEP. See 12 ATSDR: US `Agency for Toxic Substances and Disease Registry, US Department of Heath and Human Services. See 13 SIDS: `Screening Information Dataset for High Production Voume Chemicas, Organisation for Economic Co-operation and Deveopment (OECD). See: 14 See: 16

17 out 15 under the EU Pesticides Directive 91/41/EEC on Pant Protection Products [ECC, 1991]. Additionay, confidentia HSE records were accessed, discussions were hed with HSE toxicoogists invoved in the EU harmonised cassification process to carify the basis of decisions taken for some substances and, where no other information was avaiabe, Manufacturer s Materia Safety Data Sheets were used. Ony 30% of the UK T, T + substances were considered. For the remainder either: data were not avaiabe for a exposure routes or were not in a format that woud aow a comparison with cassification criteria; the overa EU harmonised cassification is not supported by the data regardess of the physica state assumed for inhaation cassifications; or the EU harmonised cassification is corrosive but not T or T + athough there are data to support the atter. (Inconsistencies with corrosive substances can arise because of priorities, unreated to Seveso II, within the administrative system for agreeing the EU harmonised cassifications.) 15 Evauations carried out by the UK Pesticide Safety Directorate Advisory Committee on Pesticides, see: http// 17