This study provides a comprehensive assessment of the state-of-the-art MOGA - ε- NSGAII's effectiveness in yielding optimal design of air valves in cooling water systems. In testing this optimization algorithm for a cooling water system in a real refinery project, problem formulation with 4-objective functions is evaluated. The power failure and pump switching scenarios, individually and in combination are considered in the assessment. In summary, the multi-objective optimization tool is effective in conducting the global search for optimal solutions. The tool finds the solutions with significantly fewer air valves, while maintaining transient pressures and load within acceptable ranges, compared to the original design proposed by the hydraulic transient analysis. In about 10,000 function evaluations, the algorithm can find a set of optimal solutions automatically within a single optimization run. These solutions are non-dominated because none exceed the performance in all objective functions. With each solution as a design, these tradeoff solutions provide decision makers with more options in the design process. Narrowing down the candidate solutions for final decision making can be achieved by defining the preferred ranges of objective values. Based on the decision maker's preference for different objectives or through a risk-based assessment, one solution may be selected as the final design. The optimization algorithm identifies not only the locations but also the sizes of the air valves. Such an effort would be very tedious and time consuming without the automatic optimization tool. Further, it is impractical to test all possible combinations manually. This study shows that the power failure scenario is more critical compared to the pump switching scenario in air valve design. However, other scenarios including pump switching, that could incur down-surge and column separation and rejoining should not be neglected. Given that the transient analysis should cover all possible critical cases to provide a safe engineering design, the optimization problem formulation should combine the transient scenarios during objective function evaluations.
Advanced surge mitigation design in pipeline systems by integrating the hydraulic transient analysis with a state-of-the-art multi-objective genetic algorithm
2012
15 Seiten, 7 Bilder, 3 Tabellen, 9 Quellen
Conference paper
English
Multi-objective optimization of large pipeline networks using genetic algorithm
Tema Archive | 2004
|Surge Event Mitigation in Propulsion Systems
AIAA | 2017
|Surge, Transient & EMI Avoidance Systems
British Library Conference Proceedings | 1993
|