Corrosion-included metal runoff from external constructions and its environmental interaction : a combined field and laboratory investigation of Zn, Cu, Cr and Ni for risk assessment

Sammanfattning: The 1990s has seen an increased awareness of possible environmental effects of corrosion-induced metal release from outdoor constructions. Considerable efforts have been initiated to perform critical assessments of possible risks of selected metals. Gaps of knowledge have been identified and research investigations started. This doctoral thesis is the result of an interdisciplinary research effort in which scientific insight into corrosion, soil chemistry and ecotoxicology has been integrated. The work comprises atmospheric exposure of pure metals and commercial materials for outdoor use. The focus is on release of four metals, copper, zinc, chromium and nickel. Their chemical speciation and bioavailable fraction in metal runoff were determined, both at the release moment and after environmental interaction with, e.g., soil and limestone. Total metal concentrations in runoff are influenced both by material properties (e.g., corrosion product solubility, and specific surface area) and by exposure parameters (e.g., rain volume, intensity, contact time and pollutants). Long-term runoff rates of copper, zinc, chromium and nickel were based on exposures (4-8 years) at standardized conditions (45o inclination facing south) in Stockholm, Sweden. Runoff rates for pure copper range from 1.2 to 1.5 g m-2 yr-1, depending on year. At the copper release moment the potential environmental effect was evaluated using 72 hours growth inhibition test with the green algae Raphidocelis subcapitata. This resulted in a mean value of 15 μg L-1 causing a 50% growth reduction (EC50). Long-term runoff rates for pure zinc range from 1.9 to 2.5 g m-2 yr-1. A considerable variation in average annual runoff rates (0.07-2.5 mg zinc m2yr-1) was observed between different investigated commercial zinc-based materials. An average 72 hour (EC50) value of 69 μg L-1 towards Raphidocelis subcapitata was found for runoff water from zinc-based materials. Long-term runoff rates from stainless steel of grade 304 and 316 range from 0.23 to 0.30 chromium and 0.28 to 0.52 nickel mg m-2 yr-1, with corresponding concentrations in the runoff at the release moment far below reported ecotoxic concentrations for chromium and nickel. Two predictive runoff rate models were successfully developed for transforming copper runoff rate data from Stockholm to other exposure sites. One model is based on rain pH, yearly precipitation and building geometry, and the other on average annual SO2 concentration, yearly precipitation and building geometry. In addition to total metal concentration, adequate effect assessments also require information on chemical speciation of the released metal and its bioavailability. Metal chemical speciation in runoff was determined experimentally through an ion selective electrode (for copper), and also modelled with the Windermere Humic Aquatic model (WHAM (V)). Bioavailability assessments were generated through bioassay tests. At the moment of metal release, all methods show that the majority (60-99%) of the metal in runoff exists in its most bioavailable form, the hydrated metal ion. During subsequent environmental entry the metal undergoes major reductions in concentration and bioavailability. This was evidenced by model column studies of the capacity of soil to retain and immobilize the metal in runoff water, and by model and field column studies of the capacity of limestone to retain copper. The retention by soil of all metals investigated is very high (96-99.8%) until each materials retention capacity is reached. Limestone also exhibits a substantial capacity (5- 47%) to retain copper. The capacity is significantly increased by increased amount and decreased fraction of limestone particles. Any outer or inner surface with significant retention ability and with low possibility of subsequent mobilization is an excellent candidate for neutralizing metal release and its potential ecotoxic effects. This was demonstrated through computer modelling (WHAM(V)) and biosensor tests (Biomet™), which showed the most bioavailable and ecotoxic metal species to be reduced during passage through soil and limestone. Predictions based on the computer model HYDRUS-1D suggest a time-period of between 4 and 8000 years, depending on runoff water and soil characteristics, before saturation in soil retention capacity of copper and zinc is reached. A significant fraction of the retained metal is extractable towards the strong complexing agent EDTA, indicating possible future mobilisation. It is also available for plant uptake, as shown by DGT- (Diffuse Gradients in Thin films-) analysis of copper and zinc in soil. The data generated, presented and discussed are all believed to be important for risk assessment work related to corrosion-induced metal release from outdoor constructions. As evidenced from this doctoral thesis, such work requires a complete set of data on annual runoff rates, concentrations, chemical speciation and bioavailability and its changes during environmental entry, together with knowledge on, e.g., type of material, service life of coating, building geometry, and dewatering system.