(173f) Modeling Landfills As Large-Scale Bioreactors. A Tool For Predicting Landfill Gas Production And Addressing Safety Issues

Landfills can be viewed as large-scale bioreactors that produce significant amounts of landfill gas (LFG), that consists primarily of CH4 and CO2, smaller amounts of N2 and O2, and traces of other organic compounds, collectively known as NMOC. In many landfills across the U.S. the LFG is utilized in turbines and engines to generate electricity, and in boilers and furnaces to produce energy. In others, the LFG is collected to be flared, often posing a variety of environmental challenges. Accurately predicting the LFG production is important in order to determine the amount of CH4 and CO2 that can be expected to be extracted from the site during its operational lifetime. A detailed model has been developed which accounts for all the key phenomena that take place in a landfill. Using a novel genetic algorithm and large-scale parallel computations this model has been utilized to fit experimental gas generation data for a major Los Angeles landfill. The model is shown to be successful in predicting the LFG generation rates at the landfill.

Despite its success in predicting the LFG generation rates and its usefulness for forecasting the long-term performance of a landfill, the detailed model is still way too complex to be utilized for on line control of every day landfill operations. Of particular concern in these operations is the risk of fires and explosions when the ratio of CH4 and O2 in LFG reaches its flammability limits, since underground fires often do occur. Of equal concern is the issue of hot spots throughout the site where the temperature can often be in excess of 150 F. Such high-temperatures can damage or destroy wells and pose a risk for the stability of the landfill structure. Having a predictive tool is, therefore, important so that the proper actions can be taken to insure that these problems are handled properly. In this paper a forecasting technique is also described that efficiently and quickly predicts the CH4, CO2, O2, and temperature profiles in a landfill. The technique is tested against historical temperature and composition data at the site, but also against the predictions of the detailed model. Combining accurate detailed models with simple forecasting methods provides a powerful tool not only for the long-term planning and design, but also the every day operation and control of landfills