Optimal TMDL Allocations: Incorporating Costs into TMDL Calculations for Complex Urban Watersheds Mohammad Tufail 1 , P.E., PhD and Lindell E. Ormsbee 2 , PhD, P.E., P.H., D.WRE 1 Postdoctoral Research Associate, Kentucky Water Resources Research Institute, 233 Mining and Minerals Building, University of Kentucky, Lexington, Kentucky 40506- 0107; PH (859) 257-1299; FAX (859) 323-1049; email: mtufail@engr.uky.edu 2 Director, Kentucky Water Resources Research Institute and Professor of Civil Engineering, 233 Mining and Minerals Building, University of Kentucky, Lexington, Kentucky 40506-0107; PH (859) 257-6329; FAX (859) 323-1049; email: lormsbee@engr.uky.edu ABSTRACT This paper will highlight the importance of economic considerations in the determination of load allocation leading to the development of a TMDL for a highly complex urban watershed impaired by point sources (CSOs and SSOs), non-point sources (urban runoff), and other unknown suspected sources of pollution. Developing a TMDL for such a complex urban watershed requires the use of a variety of computer-based hydrologic, hydraulic, and water quality models. These simulation models are used to perform load allocations leading to the development of a TMDL. The goal in most cases is to select a strategy (solution) that meets target water quality objectives as mandated by the current EPA guidelines for TMDL development. However, there often exists multiple feasible solutions for which the cost differential is significant and there is a need to select the optimal strategy that satisfies water quality objectives and at the same time is the most cost effective. For such cases where multiple feasible scenarios need evaluation, the use of simulation models alone can be cumbersome, time consuming, and cost prohibitive. In such situations, an optimization model coupled with simulation model(s) can be used to identify optimal solutions. The work described in this paper will present a case study for Beargrass Creek watershed in Louisville, Kentucky, in which an optimal management model is developed for the complex watershed leading to effective management strategies for water quality control. Such an optimal management model can be used to develop a TMDL for the watershed and can also be useful for prioritizing and phasing capital improvement projects. Key Words: Decision support system, Optimization, TMDL allocation, Water Quality modeling, Watershed management Copyright ©2007 Water Environment Federation. All Rights Reserved TMDL 2007 1139