Optimization of integrated urban wastewater systems using multi-objective evolution strategies

Abstract

The present paper deals with the development of a new multi-objective evolution strategy in combination with an integrated pollution-load and water-quality model to optimise the performance of an urban wastewater system. The optimisation algorithm combines the advantages of the non-dominated sorting genetic algorithm and self-adaptive evolution strategies and contains further development improving convergence and diversity. The identification of a good spread of solutions on the Pareto-optimum front and the optimisation of a large number of decision variables equally demands numerous simulation runs. In addition, evaluation of criteria with regard to the frequency of critical concentrations in the river and peak discharges to the receiving water requires continuous long-term simulations. Therefore, a fast operating integrated simulation model is needed providing the required precision of results. For this purpose, a hydrological deterministic pollution-load model has been coupled with a river water-quality and a rainfall-runoff model. Wastewater treatment plants (WWTP) are simulated in a simplified way. The integrated simulation model and the multi-objective optimisation algorithm were implemented in a modular common software shell. The functionality of the optimisation and simulation tool has been validated by analysing a real catchment area including sewer system, WWTP, water body and natural river basin. For the optimisation/rehabilitation of the urban drainage system, both innovative and approved measures have been examined and used as decision variables. As objective functions, investment costs and river water quality criteria have been used.

Keywords:

• best management practice
• evolutionary algorithms
• integrated modelling
• multi-objective optimisation
• urban wastewater systems