Marine ecotoxic effect of pulse emissions in life cycle impact assessment

Characterization factors for ecotoxicity in life cycle impact assessment are traditionally calculated as the product of effect and fate factors. Steady-state multiple compartment models are used to calculate the fate factor, while effect factors are derived from species sensitivity distributions (SSDs) for multiple substances using average or marginal gradients. Others have shown that steady-state multicompartment models can be used to calculate characterization factors if linear dose-response functions are used. Average gradients are linear dose-response functions per definition. Marginal gradients are first-order Taylor approximations of the effect function and require marginal exposure at all points of the compartment. Instantaneous mixing, giving marginal exposure within compartments, is an implicit assumption of the multicompartment model. This paper investigates if the assumption of marginal exposure results in significant errors for the characterization factor. Ecotoxic effect of pulse emissions is simulated in a transient three-dimensional single compartment model of the marine aquatic environment. Results show that the error in characterization factors for the Taylor approximation is less than a factor of two for multisubstance SSDs assuming concentration addition only in the aggregation of toxic effect of substances. Assuming a combination of response and concentration addition may result in a deviation of several orders of magnitude.