Distribution of heavy metals in sediment pore water due to offshore discharges: an ecological risk assessment

Barite is used as a weighting agent in synthetic and oil based drilling fluids (SBFs and OBFs) to maintain bore hole pressure during offshore oil drilling operations. Substitution of OBFs by SBFs has reduced the risk of ecological impacts. Barite makes up approximately 33% by weight of an SBF and contains traces of heavy metals, which significantly contribute to the toxicity of drilling waste. Due to the hydrophobic nature of SBFs, drilling wastes are not dispersive in the water column. Arsenic, copper and lead are the three important toxic heavy metals, amongst others, found in the drilling waste. The concentrations of heavy metals are determined using a steady state aquivalence-based fate model in a probabilistic mode. Monte Carlo (MC) simulations using Latin Hypercube Sampling (LHS) are employed to determine pore water concentrations at known pollutant loading rates (E) and impact area (AW) conditions. This paper considers a hypothetical case study to determine the water quality impacts for 4 and 10% attached SBFs which correspond to proposed best available technology (BAT) option and current discharge practice in US offshore. The exposure concentration (CE) is a predicted environmental concentration (PEC), which is adjusted for exposure probability (p) and bioavailable fraction (BF) of heavy metals. The probabilistic response of an ecosystem is defined by developing an empirical distribution function (EDF) of predicted-no-effect-concentration (PNEC) derived from LC50 and NOEC data. The lowest 10th percentile value of an EDF is defined as a representative response of an ecosystem, RE. The ratio of exposure concentration (CE) to the representative response (RE) is defined risk quotient (RQ). The ratio CE /RE 1, is an indicator of the aquatic system’s safety. The pollutants’ pore water concentrations within the radius (R) of 815 m are estimated and cumulative distributions of RQ are developed to determine the probability, p(RQ 1).