Deterministic And Stochastic Water Balance Modeling Of The Gilt Edge Mine Superfund Site, South Dakota

A water balance model with deterministic and stochastic components was developed and used for the Gilt Edge Mine Superfund Site, South Dakota, to evaluate risks of contaminated runoff discharging to surface waters based on assumed water treatment plant (WTP) rates for different remedial alternatives. The objectives of the modeling were to use a stochastic approach for analysis of precipitation based on historic records, and for probabilistic estimation of annual and monthly runoff from primary disturbed/mining activity areas that currently requires, or will need, treatment at the WTP, water storage and changes in a major pit, and risk of contaminated water bypass of the WTP to a downstream drainage. Subareas/source areas generating contaminated runoff requiring treatment were delineated. A simple water balance model was used based on precipitation inputs and evapotranspiration and groundwater losses. Annual and monthly precipitation records for Lead, South Dakota were used to estimate precipitation at the mine site. A lognormal distribution was estimated for monthly precipitation. The stochastic software package/Excel add-in “@Risk” was used to generate random monthly precipitation values using a Monte Carlo method based on the distributions. The average annual and monthly distribution of pan and lake evaporation, and groundwater losses were also estimated for the site. Stochastic outputs from the model for the end of each month and year included contaminated runoff requiring treatment, water storage in mine pits and the risk of WTP bypass given different assumed WTP rates. The probability of the time required to dewater the pits was also modeled.