Use of genetic algorithms to optimize in-line storage of a storm system and improve water quality

Mohamed Abdel-Latif, Hossam Mahmoued and John Ortli

ABSTRACT

United Parcel Service Inc. is expanding its Kentucky, Louisville hub. The $1.1 billion expansion will speed parcel processing by 40 percent at the company’s main air hub. The seven-year project, the most expensive in UPS history, more than doubles the Louisville hub’s sorting complex to 4 million square feet - the equivalent of more than 80 football fields.

As a result of the airport expansion, the storm sewer network will be extended and upgraded significantly. A study of the UPS Worldport storm water drainage facilities is conducted to determine the necessary storm water management requirements associated with the expansion plans for the Worldport operation. The Worldport drainage system has the following components:

The “dirty system”, which is the part of the storm water system that collects and transports storm water runoff from areas involving aircraft operations, fueling facilities, and de-icing activities. The “clean system”, which is the portion of the Worldport storm water system that collects and conveys storm water from roofed areas and ground areas not affected by aircraft de-icing or refueling activities. The weir structure, which directs the contaminated waters to existing Louisville and Jefferson County Metropolitan Sewer District (MSD) lines for transportation to MSD treatment facilities.The size of storm sewer and the setup and operation of the weir structure are intended to capture all dirty storm water for the design storm.

In this paper, the PCSWMM Genetic Algorithm (GA) routine is utilized to optimize the in-line storage of the proposed improvement of the storm sewer network to minimize weir by-pass to the clean storm system. The proposed sizes of improvement to the existing storm system and the setup of the weir operation are taken as the decision variables.


Permanent link: