Impact of Partitioning on Chemical Risk Assessment

Abstract As Risk Based Approach (RBA) becomes increasingly prevalent as a risk assessment method for chemical discharges, it stands to reason that it can be used as a chemical selection tool for either new chemical applications or as a method of assessing alternative chemicals for existing applications. Use of this tool goes beyond the classification of chemicals based on eco-toxicological parameters utilised on both the Norwegian and UK continental shelves. It also raises some intriguing possibilities when it comes to the assessment of production chemicals. One of the key eco-toxicological properties tested as part of the established OSPAR regulatory scheme is how a particular chemical partitions between octanol and water. This data is used as a proxy for how likely that chemical is to bioaccumulate in the fatty tissue of an organism and is expressed as the logarithm of the partitioning coefficient between those two phases. However, this research uses this data to calculate how much of the chemical will partition between the hydrocarbon and aqueous phases prior to discharge. Factors prior to discharge are not considered in the current environmental risk assessment model and due to the prescriptive regulatory process cannot be accounted for. However, by going beyond regulatory compliance and incorporating process conditions, the criticality of including partitioning calculations in the risk assessment process becomes apparent. By using partitioning data and derived equations, a discharge dosage of each added chemical can be calculated thus allowing accurate determination of the environmental impact factor. A chemical that preferentially partitions into the produced hydrocarbon phase will be exported with that fluid and will not be discharged to the marine environment. To perform these calculations a number of different data points must be used. Initially, an equation had to be derived that would allow the partitioning behaviour to be applied to any production system with any ratio of hydrocarbon to aqueous fluids. This equation was derived from the equation for the partitioning coefficient. In conjunction with metered/calculated process data and component level chemical data was used to determine a discharge dosage for each chemical which was then modelled. The results from this model were then compared against results generated using dosages taken from the regulatory chemical permit without accounting for any substance level data or partitioning calculations and results produced using just substance level data. From the results of each modelling run, it was clear that without taking into account both the component level data and the partitioning calculations erroneous conclusions can be drawn about which chemicals pose the highest risk to the environment. It is also clear that chemical preference for the hydrocarbon phase, long seen as an issue within the environmental regulatory schemes, is actually a benefit when it comes to production chemicals, as the partitioning behavior of the chemical (prior to discharge) prevents some of the chemical from being discharged to the marine environment, thus reducing the risk posed by oil soluble chemicals.