Jump to content

Modelling and environmental pollutants

Foto: Guttorm Christensen, Akvaplan-niva

Storvannet is a small lake close to Hammerfest. It looks idyllic, but the reason why NILU scientist Dr. Ingjerd Sunde Krogseth and colleagues are so interested in it is what lies below the lake’s surface. At the lake’s bottom, scientists have found that sediment and aquatic animals contain large amounts of chemical ingredients from cosmetic products called siloxanes.

Siloxanes are chemical substances used in various industrial applications. However, their dominant application is within personal care and cosmetic products such as soaps, deodorants, hair gels and skin creams. Emissions to the environment occur primarily during use of such products, where siloxanes either vaporize to the air or are rinsed off, ending up in the sewage.

Catching a ride with the sewage

Until the 1970s, Lake Storvannet received sewage discharge from the surrounding residential areas. The sewage is now directed to the sea. However, despite attempts from the municipality to improve the wastewater system, sewage release into Storvannet still occurs from time to time due to leakages from drainage pipes and overflow events during the snowmelt season and heavy rainfalls. – Siloxanes have been found in fish and sediments in many fjords and lakes affected by sewage discharges, even in Svalbard, explains Dr. Krogseth, – but there is still uncertainty about how siloxanes behave in the environment.

In order to carry out a rational risk assessment of possible new environmental pollutants, it is crucial to have a good understanding of how they behave in the environment. – Through a project in the Norwegian Research Council’s MILJØ2015 program (NORDIC-LACS – Nordic Lake Exposure to Cyclic Siloxanes), we wanted to dig deeper into this by using Storvannet as an example and look at the whole picture; from when the siloxanes are released with the sewage to when they end up in the fish. To do this, actual samples from the lake were combined with environmental model predictions, explains Dr. Nicholas Warner, project leader and senior scientist at NILU – Norwegian Institute for Air Research.

What is an environmental model?

An environmental model is based on existing knowledge and theory of how a chemical moves within particular environments. It serves as a tool to help us simplify the real environment to better understand what processes are important in controlling pollutant exposure. NILU, under the direction of senior scientist Dr. Knut Breivik, has for many years worked to develop mathematical models to understand how organic environmental pollutants distribute in the environment (see fact box), in cooperation with partners in both Sweden, England and Canada.

– These kinds of models are very useful, explains Dr. Krogseth. – They can be used to estimate what is in the environment before we go out to sample, they can calculate processes that are impossible to measure in the field, and help us put measurement results into a larger context that we can’t do on the basis of measurements alone. At the same time, we are completely dependent on comparing the models with actual measurements to confirm that what the models calculate is correct.

In the case of siloxanes, the models have been used to predict that siloxanes can fly through the air to the Arctic, which was later confirmed by measurements, and to reveal new siloxane compounds present in the environment. In recent years, NILU scientists have joined forces with partners at the Universities in Stockholm and Leicester and Akvaplan-niva AS, a collaboration strengthened through funding from the Fram Centre in Tromsø, and turned their attention to the aquatic environment.

Siloxanes in Storvannet

In 2014, the scientists collected a number of samples from Storvannet. The samples included sewage, water and sediment from the lake, as well as muscle and liver samples from char and trout. In addition, they took samples of the fish’s food; sticklebacks, mosquito larvae, and tiny clams living in the bottom of the lake.

– We found siloxanes throughout the entire system, says Dr. Warner, – except in the water samples from the lake. Siloxanes do not like being in water due to their inherent properties. They prefer to be in the sediments. Hammerfest is far north, and with a short and cool summer, the plankton season is short and intense. Thus, the animals living in the sediments are an important food source for the fish for much of the year, and the fish get exposed to the siloxanes primarily by eating these bottom-dwellers.

Model vs. samples

Fish from Storvannet, sorted by size. The scientists took muscle and liver samples from the fish, in addition to sewage, water and sediment samples samples of the fish’s food. Photo: Ingjerd S. Krogseth/NILU
– The samples we collected provided a snapshot of just how high the concentrations of siloxanes are in those exact samples, at the moment of collection, Dr. Krogseth explains. – Then we tested the model against these results. We used a previously developed food chain model and incorporated bottom-dwelling animals into the model as we knew they played an important role in the Storvannet food web. Then we fed the model with the properties of the lake, the food chain, and the siloxanes. Based on this, the model can predict both concentrations in animals, as well as the processes controlling movement of siloxanes throughout the food chain. If what the model predicts is similar to what the samples show, it’s great, because it means we have a good understanding of what’s going on. If the model and test results do not match, there’s still something we have not understood, and need to look further into.

The model performed well in recreating what the scientists had measured in Storvannet. It confirmed that the bottom-dwellers were the most important source to siloxane exposure for the fish. In addition, it showed that the fish can actually get rid of siloxanes through ventilation to the water through their gills.

– This was a little surprising, says Dr. Krogseth – as siloxanes do not like being in water. However, this is explained by the special conditions in Storvannet, where siloxanes are predominantly found in the sediments while the water itself is relatively free of siloxanes. This is because sewage only leaks into the lake from time to time, and the siloxanes can be removed relatively quickly from the lake water itself as the water flows rapidly through the lake. Thus the fish, which ingest siloxanes from the sediments and bottom-dwellers, has much higher concentrations than the water they swim in, and therefore they can “lose” siloxanes to the water. In this case, the models helped us understand more about what we measure, and what actually happens.

NILU-forsker Ingjerd Sunde Krogseth vannprøver ved bredden av Storvannet like ved Hammerfest.
The models must be compared with actual measurements to confirm that they are dependable. Here, Dr Ingjerd Sunde Krogseth takes water samples at the banks of Storvannet, a lake close to Hammerfest. Photo: Guttorm Christensen/Akvaplan-niva

Restriction on the way

The fish, bottom-dwellers, and sediments in Storvannet contain siloxanes. The siloxanes known as D4 and D5 (see fact box) are on the Norwegian Environment Agency’s “List of Priority Substances ” of chemicals that pose a serious threat to health and the environment. For substances on this list, the goal is to stop emissions by 2020. Until now there have been no regulations on the use of siloxanes, but in Europe, it will now probably be forbidden to add more than 0.1% of D4 and D5 in personal care products that are washed off during normal use, such as shampoo, conditioners and soaps. – It will be interesting to see what will happen in the future with more limited use. Especially regarding how quickly the environment will react to lower emissions, says Dr. Warner.

The road ahead

While the scientists will continue to monitor the siloxanes, they will also turn their focus and models to other potential new environmental pollutants, challenges and food chains. – The model we have developed in cooperation with Stockholm University and used in Storvannet will now be used in two major projects from the Norwegian Research Council’s programs; ØKOSYSTEM (NEM: Development, Evaluation and Application of a Nested Exposure Assessment Model for Organic Contaminants in the Nordic and Arctic Region) and MILJØFORSK (SERA: Source-Exposure Relationships for Airborne Organic Contaminants of Emerging Concern in Northern Terrestrial and Freshwater Ecosystems), explains Knut Breivik, head of these new projects.

– In SERA, we will continue the cooperation with Akvaplan-niva, and look at the connection between long-range transport of environmental pollutants in air and the occurrence of the same pollutants in lakes that do not have any local emissions of these substances. In NEM, the model will be part of a larger model framework, aiming to model the entire relationship between global emissions of pollutants and what we measure in Nordic and Arctic food chains.

– A good overall understanding of the relationship between emissions and exposure is crucial in order to implement rational regulatory measures where necessary, something siloxanes are a good example of, says Dr. Krogseth. – The more we test our models, the more trustworthy they become, and thus the more useful they are in learning more about this relationship for new environmental pollutants, she concludes.

What are multimedia models?

Multimedia models are mathematical models developed by Professor Don Mackay at the University of Toronto during the late 1970s.NILU has been working with such models since the early 1990s.

The models can calculate how chemical substances behave in the environment, where they accumulate, how long it takes before they disappear, and how they accumulate in food chains and ultimately lead to human exposure.

The models are based on the principle of mass balance. By connecting this with knowledge of the distribution, transport and degradation of chemical substances, one can achieve a quantitative and mechanistic understanding of how they behave.

Multimedia models have been crucial to our understanding of why and how organic pollutants are present in high concentrations in the Arctic, which has again helped form the basis for international agreements on environmental pollutants.


Krogseth, I. S.; Undeman, E.M.; Evenset, A.; Christensen, G. N.; Whelan, M. J.; Breivik, K.; Warner, N. A. Elucidating the behavior of cyclic volatile methylsiloxanes in a subarctic freshwater food web: A modeled and measured approach. Environ. Sci. Technol. DOI: 10.1021/acs.est.7b03083

Krogseth, I. S.; Whelan, M. J.; Christensen, G. N.; Breivik, K.; Evenset, A.; Warner, N. A. Understanding of cyclic volatile methyl siloxane fate in a high latitude lake is constrained by uncertainty in organic carbon–water partitioning. Environ. Sci. Technol. 2017, 51, 401-409.