Stochastic Approach to Waste Treatment Planning

A methodology for the sizing and staging of infrastructure in rural regions facing uncertainty is developed and applied to wastewater treatment planning in Utah’s Ashley Valley. The methodology involves the construction of a first-order Markov chain. The transition matrix of the Markov chain contains subjective probability estimates relating to the occurrence (or nonoccurrence) of future regional events. The subjective probability estimates are obtained using a series of questionnaires sent to knowledgeable government officials, industrial managers, and members of the academic community. The transition matrix is used to compute an expected population for the region. It is also employed to compute the necessary probability estimates for an expected loss function for a stochastic mixed-integer programming model. The model derives a least cost sequence of projects consistent with regional probabilistic population growth.