PARENT SESSION
60 - Probabilistic Methods in Risk Assessment
8:30 AM to 12:30 PM, Wednesday, 15 May 2002
Session Chair: Hart, Andy ¹, Hendley, Paul ², van den Brink, Paul ³, Blake, Naomi ⁴, ^{1 2 3 4} .
Lehar B

(60-01) Improving risk management by probabilistic risk assessment of agricultural pesticides with USES.

Rikken, Mathieu^*,1, Van Wijnen, Harm¹, Linders, Jan¹, Jager, Tjalling¹, ¹ RIVM Postbus 1, Bilthoven, The Netherlands

ABSTRACT- For the registration of pesticides in the Netherlands, the risk assessments follow the EU procedure according to directive 91/414/EEC. Risk assessments are carried out for active substances and formulations (products) with the computerised system Uniform System for the Evaluation of Substances (USES). The measure of risk used in USES is a single-point or deterministic estimate. A deterministic approach can be useful for a first survey, but an uncertainty analysis accounts for the variation in the input values by incorporating their probability distributions, which leads to a probability distribution of the risk. The main goal of this study is to make decision-makers see the added value of probabilistic analysis versus the traditional deterministic approach, which hopefully can lead to a better and more defensible decision-making. The example calculations only account for chemical-specific uncertainties. Other sources of uncertainty and variability are ignored. It is discussed how a probabilistic analysis can be interpreted. The example calculations show that decision-makers are presented with more information to base their decisions on. With a probabilistic approach the full range of the resulting risk and the main sources of uncertainty are available. Another advantage is that it also shows the information needed to refine the risk assessment. The examples show that the registration of pesticides can also be influenced by a probabilistic analysis. The uncertainties in the results of USES depend on the chemical properties of a pesticide. Generally, the degradation is the main source of uncertainty. Nevertheless, for the surface water end point Kom, vapour pressure and water solubility are also important. For the leaching to groundwater, not only degradation but also Kom is important. The result is less sensitive for degradation when the uncertainty in the toxicity values is included. Furthermore, an option is presented for a more detailed uncertainty analysis of the effect assessment.

Key words: probabilistic, uncertainty, risk, pesticide