Application of Genetic Algorithm Optimization for Collection Systems to Johnson County's Turkey Creek Sewer-Shed

Collection system utilities are facing increased costs due to population growth, aging infrastructure and new regulations. At the same time, many utilities are facing a decrease in the funding available to meet their changing needs. As a result, utilities are searching for new, more powerful problem-solving approaches to identify the best solution while continuing to provide a low cost, safe, and reliable service to their customers.The application of Genetic Algorithm (GA) optimization is a powerful approach that many water supply utilities are currently embracing. Cities such as Phoenix, Las Vegas and San Diego are using this technology to reduce costs and enhance system operating performance. GA optimization has been successfully applied to water supply system planning, design and operations since 1995. Until recently, the technology had not been successfully adapted for Collection System Utilities to reduce their system improvement costs through GA optimization.GA optimization can now be applied to stormwater, wastewater, and combined collection systems. This application gives utilities a systematic approach to evaluate millions of possible solutions to resolve complex, multi-objective infrastructure problems and identify the most cost-effective capital improvement program (CIP) or operational strategy. Utilities may customize the infrastructure improvement options, cost estimating procedures and variables, and penalties associated with alternatives to address the specific requirements and needs of their system.Building upon the past success of GA optimization for water distribution systems, CH2M HILL and Optimatics developed a similar GA optimization application for collection systems. This application works with the USEPA Stormwater Management Model (SWMM). SWMM has the capability to simulate gravity and pressure sewer systems and is commonly used to analyze sanitary, combined, and separate storm sewer systems.Johnson County, KS is facing many of the same sanitary sewer collection system challenges common to wastewater collection system agencies throughout the United States. These challenges include basement backups, sanitary sewer overflows (SSOs), aging infrastructure with increasing maintenance needs and regulatory compliance challenges. CH2M HILL and Optimatics worked with Johnson County, KS to apply GA optimization and identify cost effective solutions to collection system problems. These solutions offered significant savings over solutions identified through traditional master planning approaches. This paper presents the GA optimization approach and results for Johnson County's Turkey Creek sewer-shed. Overview of Optimization Process Collection system GA applications use an intelligent decision-making process to simulate a complete range of possible system components and configurations. This includes pumps, gates, real time control, weirs, pipe segment improvements, storage, and inflow reductions. GA optimization works with SWMM to create and evaluate successive generations of trial solutions as the GA search narrows in on a number of near-optimal solutions for consideration by the decision-makers. This technique differs from the traditional simulation analysis approach in which a modeler uses trial-and-error to evaluate a small number of trial solutions. Although a hydraulically feasible solution can usually be found with this technique, the hydraulic performance of the system may not be the most efficient possible. Additionally, the cost of the proposed solution may be considerably higher than the cost determined through the use of GA optimization. Project Application and Results GA optimization was applied to Johnson County's Turkey Creek sewershed to establish a well documented approach to facility planning, exhaust the evaluation of all possible system improvement alternatives and identify cost savings over traditional master planning approaches. A Design Data Summary (DDS) document was prepared as the first step in GA Optimization to establish system performance criteria (maximum water elevations, minimum and maximum velocity, etc.), system alternative options and capital and O&M cost estimation procedures. This DDS was then used to program the GA optimization application for simulation. The results of the GA optimization for the Turkey Creek sewershed were a preliminary capital improvement plan that considered all possible options identified with the County, met system performance criteria and cost significantly less than results produced using traditional facility planning methods.The recommended solution for the Turkey Creek sewershed included a combination of infiltration and inflow (I&I) measures, improvements to sewer system capacity in select areas, system storage and pump station improvements. CH2M HILL worked with Johnson County to develop an innovative approach for the GA optimization to consider various levels of I&I reduction through cost per I&I removal curves derived from pilot I&I programs completed in Johnson County. Sewer system capacity improvements and storage systems were selected from options identified throughout the sewershed where improved facilities were possible. Pump station improvements were selected at the outfall of the sewershed, where flows are pumped to the Nelson Wastewater Treatment Plant and treatment plant expansion was considered as an option. The recommended solution included a combination of all possible options presented to the GA optimization application with the exception of satellite treatment facilities.