Uncertainties abounded 10 years ago in early debates about exactly how much the proposed REACH legislation would cost the chemical industry. Now, after the first batch of REACH registrations are in for 25,000 high volume chemicals, as well as the most hazardous substances, there are still questions as to just how much REACH will cost the industry.
As a result, companies, mostly specialist chemical producers, having to register substances with an annual output of 100 t or more by mid-2013 and remaining chemicals at or above 1 t/year by mid-2018, still cannot be certain about the costs of registration, in particular the expense of testing substances.
It is still difficult to assess the accuracy of early predictions of the costs of registration and testing required by REACH because of continued doubts about the longer term trends in these costs. The European Chemical Agency (ECHA), which runs REACH, still has to assess the first batch of registration proposals to carry out around 1500 animal tests. These could make up a significant chunk of the total costs of the registrations as animal tests are very expensive. The initial predictions made around five to seven years ago on total registration and testing costs ranged from less than €1bn to over €15bn. At the time that was equivalent to 0.25% to more than 3% of the European chemical industry’s total annual revenue, although far less in terms of total revenue over the 11 years of REACH’s entire registration process.
After the draft REACH legislation was amended to reduce costs, the European Commission calculated in 2003 that total testing and registration costs would amount to €2.3bn over the 11 years. It estimated that the costs of testing the average chemical would be €85,000, while chemicals requiring more stringent testing, like high volume products, could be as high as €325,000.
BASF, the world’s largest chemical company, now predicts that its total REACH costs over 11 years will be €500-550m, equivalent to about 1% of its annual global chemical sales, excluding revenue from oil and gas, but around 0.1% of chemical revenue over the full period of the legislation. ‘The costs include internal processing of available information on the basis of legal requirements and compilation of dossiers and assessments and experimental investigations in accordance with the list of [test] requirements,’ says a BASF spokesman.
The European Chemical Industry Council (Cefic) believes that costs are too substance specific to create an industry-wide perspective on the issue. ‘It’s impossible to provide figures [for the costs of REACH to the whole industry] at this stage,’ says one Cefic official. Nonetheless, some REACH specialists feel that there is enough evidence to show that registration and testing costs so far have been higher than initially expected.
‘If you look back at some of the earlier calculations, it is clear that some REACH costs have been underestimated,’ says Jo Lloyd, technical director at ReachReady, the consultancy set up by the UK Chemical Industries Association. ‘Administrative costs in many cases have been higher than expected because the people setting up and running consortia and substance information exchange forums (SIEFs) for collecting data and drawing up registration dossiers have had to go through a learning curve,’ she adds. ‘That should mean... these administrative costs should be lower for the 2013 and 2018 registrations.’ Animal tests make up the bulk of REACH costs and they have been the hardest to budget for, Lloyd points out.
The registration fees charged by ECHA, which range from €120 for a joint registration of a lowvolume substance to over €30,000 for a large company registering a high volume chemical, account for a relatively small proportion of the costs. Instead, much of REACH expenditure stems from the fees for collecting safety data and putting registration dossiers together which SIEFs charge for (C&I, 2009, 11, 8). SIEFs are networks set up by companies to share the expense of gathering and preparing data for individual substances.
Trade associations and consultancies are reporting that their own inquiries show wide variations in the fees charged by SIEFs for safety data and drawing up dossiers. A lot depends on the safety of the chemical. The SIEF costs of registration dossiers for a non-hazardous substance, requiring little safety information, will be a lot cheaper than those for toxic chemicals. The SIEF costs for non-hazardous chemicals could be only a few thousand euros, while toxic substances could run to more than €500,000.
For SMEs and other companies with little safety data of their own, the big ticket item of REACH expenditure is payments they have to make to other companies within SIEFs. Safety information tends to rest in the hands of larger producers and these fees give smaller companies access to this data.
A lot of the data may come from tests carried out over the last few decades. But under REACH companies are allowed to charge other SIEF members access fees equivalent to the cost of conducting the tests today. Although companies are complaining about high data-access charges, few details are emerging about the exact level of fees because of confidentiality clauses in SIEF membership agreements.
‘There is plenty of transparency within SIEFs as the REACH regulations require, but little or none from outside them,’ says Douglas Leech, technical director at the UK Chemical Business Association (CBA), which represents SMEs and distributors. ‘Our members are keeping very quiet about how much they are paying because they are worried that if they breach the confidentiality rules their right of data access might be declared null and void.’
‘We do know that there are big differences in SIEF costs and that some of them can be extremely expensive,’ Leech adds. ‘Tests on a carcinogenic or environmentally toxic chemical can amount to €3–4m. The sharing of costs like these within large SIEFs may be less painful than in SIEFs with only a few members.’
Confidentiality has been a key concern within SIEFs and not just over disclosure of costs. Under REACH, companies are allowed to withhold information about their chemicals from other SIEF members, particularly when they are worried about the content of their product safety files leaking out to competitors, particularly those outside Europe.
‘We welcome the principle of sharing animal testing data and the option of sharing other substance-specific data,’ says a BASF spokesman. ‘However, in practice this has elicited extensive discussions on confidentiality issues and on consequences for non-EU programmes on chemicals. Confidentiality agreements and other contracts have consumed significant resources.’
In 2007, Cefic commissioned the Berlinbased Social Science Research Centre (WZB) to carry out a major study to investigate the cost of toxicological tests across Europe. The study has since become a benchmark for the costs charged by SIEFs. The WZB found large disparities in costs across European laboratories, particularly between small and big operators, even when carrying out identical tests. As a result it leaves room for flexibility when SIEFs calculate data access fees.
According to the study, acute aquatic toxicity studies on Daphnia cost €2330 from small labs and €4900 from large labs. The cheapest labs carried out toxicokinetic behaviour studies for €25,818, while the more expensive labs charged €74,803, on average. For aquatic bioconcentration studies, the cheaper labs charged €43,873 against €87,082 for the more expensive labs.
‘The huge variations in the costs for similar types of tests seem to be linked to the amount of knowledge of the properties of substances,’ explains Manfred Fleischer, an economist, who conducted the WZB study. ‘The more knowledge of the properties of chemicals the less likely there is to be variations in costs.’ Fleischer says that costs are unlikely to have gone up much since the study – despite reports that some Western European laboratories are booked out for months.
‘In the survey we were not able to take account of recent increases in laboratory capacity in Eastern Europe nor the greater use of Chinese and Indian laboratories for testing by chemical producers and importers in Europe,’ he says. ‘I would expect that, as a result of this rise in capacity outside Western Europe, costs would not have gone up among GLP (good laboratory practice) European laboratories over the last few years.’
The study verified that by far the most expensive tests are for chemicals that are highly toxic to humans and the environment. Testing for carcinogens in rats can cost as much as €790,000, the study found. These high-cost tests will need ECHA approval before they are carried out.
The first set of registrations demonstrates that a lot of companies have, apparently, been using cheaper alternatives to animal tests. These include quantitative structure-activity relationships (QSARs), which use computer modelling to predict adverse effects on human health or the environment and a weight-of evidence strategy applying data from a range of sources to waive the need for animal tests. ‘My impression is that very little animal testing has actually been done so far,’ says Thomas Hartung, professor of pharmacology and toxicology at Konstanz University, Germany, and director of the Center for Alternatives to Animal Testing (CAAT), John Hopkins University, Baltimore. ‘Most companies and SIEFs have been trying to waive testing. Of critical importance will be the extent to which [ECHA] agrees to this alternative approach.’
ECHA can also tell SIEFs, which have asked for approvals for animal tests, to do a risk assessment on their substances using less expensive animal alternatives. Tests permitted by the agency could use as many as 890 animals each, on average, according to the European Coalition to End Animal Experiments (ECEAE). So decisions still to be made by ECHA could have a huge influence on the final cost of REACH.
Sean Milmo is a freelance science writer based in Essex, UK.