Probabilistic Risk Assessment for dairy waste management systems

Abstract

Probabilistic Risk Assessment (PRA) techniques were used to evaluate the risk of contamination of surface and ground water with wastewater from an open lot dairy in Erath County, Texas. The dairy supported a complex waste management system for the treatment and utilization of milking parlor wastewater and open lot runoff. Wastewater and runoff were routed through settling basins and anaerobic treatment lagoons before irrigation onto cropland. Specific objectives of the study were to integrate a stochastic hydrologic model with event tree and fault tree analyses and to evaluate the use of PRA techniques in agriculture. Event trees were used to identify pathways of water and wastewater movement through the dairy and environmental hazards associated with the system. A simulation model was developed to perform water balance calculations for the lagoon and to estimate annual lagoon discharge and irrigation. The model required inputs of stochastically generated precipitation, average monthly lake evaporation, irrigation and wastewater loading rates, and dairy site characteristics. The frequency and magnitude of nitrogen release to surface and ground water at various points in the system were estimated using model output and other information about the dairy. Preliminary failure scenarios included surface infiltration, underground pipeline rupture, channel overflow, lagoon overflow, and lagoon seepage. For existing conditions, lagoon seepage presented the greatest risk with an estimated release of 125 kg of nitrogen per year. All other preliminary failure scenarios yielded risks of less than 15 kg of released nitrogen per year, indicating that risk of surface water contamination associated with the dairy was minimal. When the wastewater loading rate in the model was increased by 50%, however, the risk of surface water contamination from lagoon overflow increased to I 100 kg of released nitrogen per year. The risk associated with wastewater irrigation was not examined in detail. It appeared that the potential for crop uptake of nutrients in the irrigated wastewater was sufficient to prevent contamination of groundwater with proper management practices. Although difficulties were encountered in using event trees to determine failure frequencies, the diagrams served as an effective conceptual aid for identifying hazards and presenting results. The simulation model proved more useful for quantifying failure frequencies. Integration of the simulation model with PRA techniques yielded a valuable tool for estimating risk and for promoting communication between producers, scientists, regulatory agencies, and the public. With further development, the PRA methodology will help in the identification and implementation of effective solutions to environmental problems in a number of agricultural situations.