|
Abstract:
|
Gas leaks represent a major risk for gas utilities that can lead to disastrous human, financial and publicity consequences. Managing the response to possible gas leaks safely, effectively and efficiently requires balancing the often conflicting objectives of regulators, management, labor unions and the public. This case study analyzed a year of data from a major metropolitan area, representing over 48,000 reported gas leaks, in order to develop a stochastic simulation model for the response to possible gas leaks and allow the utility to identify and test potential response strategies to more effectively manage gas leaks.
This talk demonstrates the application of business analytics and incorporates methodologies from multiple disciplines such as statistics (data analysis and stochastic models) and operations research (simulation models and optimization). The complex pattern of temporal, seasonal and geographic variation in gas leaks was incorporated into an agent-based stochastic simulation model. The model permits ease of communication and the ability to evaluate multiple response strategies.
|
