Use and emissions of hazardous substances in Norway, 2002-2007

Increased awareness of the effects of hazardous substances on human health and the environment has led to a growing demand for information regarding use and release patterns of such substances. For the past few years, Statistics Norway has been working on new statistics on the consumption and emission of hazardous substances in Norway. At present, the statistics cover the emissions of hazardous substances caused by the use of products that are subject to the duty of declaration to the Norwegian Product Register. The statistics include about 450 substances that are categorised as being CMR (may cause cancer, be mutagenic or generate reproductive disorders), chronically toxic, sensitising and/or dangerous for the environment.

The emission of hazardous substances to the surroundings (including air, soil and water) is calculated by multiplying the amount consumed of a substance by an emission factor. The emission estimates include diffuse emissions. This means that not only emissions from production of products containing the substance, but also emissions from the subsequent use of these products are included. The consumption of a substance is expressed as the sum of production and import, minus export, declared to the Product Register. The emission factor represents the fraction of the substance that is not incorporated into a new product, transformed into a new compound, or handled in some other way (e.g. as waste). The fraction of a hazardous substance emitted is assumed to be dependent both on the type of product used and the sector in which it is used.

The model and its emission factors have been gradually improved, incorporating an increasing amount of information. In the present model, the majority of the emission factors are specific for the combination of product type and industrial sector. Some are product-specific factors, while a few emission factors are specified for the combination of product type, industrial sector and substance. The first two groups of emission factors are obtained from two Swedish studies, while the substance-specific factors are based on consultations with relevant industry.

Each year approximately 6 million tonnes of around 450 selected hazardous substances is consumed. Roughly 0.3 per cent of these substances is emitted to the surroundings. In other words, around 19 000 tonnes of a long list of substances classified as CMR, chronically toxic, sensitising and/or dangerous for the environment find their way into the environment every year.

According to present estimates, emissions of CMR- and chronically toxic substances decreased from 2002-2004, but has since then increased again and are now roughly at the 2002 level. The emissions of sensitising substances have been relatively stable in the time period, while substances dangerous for the environment have shown an increase in emissions from 2002 to 2005, followed by a considerable decrease due to reduced consumption in the manufacturing industries.

While the overall emission levels are comparable in 2002 and 2007 for all hazard categories, there has been much variation at the more detailed level. There is substantial between-year variation in the number of substances being declared within each hazard category. Generally, for each hazard category, more substances entered than left the activity data between 2002 and 2007. In addition to the variability in number of substances included in the analysis, there is also variation in the emissions of each substance. More than twelve per cent of the CMR substances with maximum emissions above 50 tonnes had a variation of more than 50 per cent between maximum and minimum emissions (of years present). This percentage is lower for the other hazard categories, but there is still a considerable proportion of the substances showing large variation in emissions between years.

Trends in emissions of hazardous substances may be considered a rough indicator of the development of the risk of damage to human health and the environment caused by these substances. However, estimates of emissions are not measures of the actual risk. Firstly, not all hazardous substances or emission sources are covered. Secondly, a risk assessment would require the integration of more information, both about the recipient and about various properties of the substances. Thus, one of the first steps towards an approximate quantification of risk, is to give emission figures for different substances weights according to their hazardous properties.

A first attempt at such a hazard weighing has been performed for CMR substances and substances dangerous for the environment. The most hazardous CMR substances, evaluated on the basis of their R-sentences, were given weights ranging from one to ten, while the remaining received the weight one. For the substances dangerous for the environment, the risk phrases contain no information for evaluating hazard. For this hazard category, substances that are persistent (P), bioaccumulative (B) and toxic (T) (so-called PBT-substances) were given hazard weights ranging from one to ten, while the remaining substances dangerous for the environment received the hazard weighting one.

The hazard weighing is a first approach towards an assessment of hazardousness within each hazard category, and the weighing with a range of hazard weights gives an indication of how sensitive the trends are to the weighing procedure. The results should thus be interpreted with caution.

Depending on the size of the weight, the hazard weighing resulted in trends quite opposite of the emission time series. With a hazard weight of ten, the hazard weighted CMR substances showed a steady decrease from 2002 to 2007, while the hazard weighted substances dangerous for the environment showed a considerable increase from 2002 to 2005, followed by a slight decrease. This indicates that there may have been a shift from very hazardous to less hazardous substances within the CMR substances.

For the substances dangerous for the environment, however, there seems to be an increased emission of PBT substances. An uncertainty analysis has been performed for the emission estimates. This uncertainty analysis evaluates uncertainties in activity data and emission factors. The results show that the CMR- and chronically toxic substances have by far the highest uncertainties. This is a result of high consumption of energy goods, where the uncertainty in emission factors is rather high. An improvement of both activity data and emission factors for energy goods will thus result in a large improvement of the emission estimates.