Revisiting Pipe Component Design

Under recent economic pressures, the nuclear power industry has explored options to reduce costs and increase electricity generation. One overlooked area where lost power output can be recaptured and attendant problems (e.g., flow-induced vibration) addressed is the main steam piping system. The technology underlying this system is essentially over a 100 years old, and incentives to improve pipe components have been limited by costs (development, retrofitting and unforeseen risks and problems) and analysis tools capable of assessing candidate designs. More recently, these conditions have changed. Access to sophisticated computational fluid dynamics methods has significantly advanced design assessment capabilities. Also, smaller profit margins have increased willingness by the industry to squeeze out moderate power generation increases (e.g., 1% via measurement uncertainty recovery). Finally, power uprates have resulted in higher steam flow speeds that increase head losses and in some cases exacerbate the potential for flow-induced vibration. Taken together, these developments have brought renewed attention to addressing losses and vibration in nuclear power plant steam piping systems. Current approaches to redesign piping components to reduce vibration and improve performance (i.e., reduce head loss) use a combination of experimental testing and flow modeling, together with human intuition regarding the geometrical modifications anticipated to improve flow and reduce unsteadiness. What is lacking, however, is a formal design approach that automates both the analysis process and the design search and is able to factor in test data for improved prediction. This effort will develop such a design code by leveraging prior work in nuclear flow analysis and testing, a proven Cartesian grid-based flow solver that eliminates user involvement in the mesh generation process, and adjoint methods that are widely used in the aerospace industry for both steady state and transient flows. Experimental testing remains a critical element in nuclear piping assessment, particularly regarding unsteady flow effects. A key emphasis in the effort will be integrating the flow analysis software and experimental testing components into a comprehensive design program. Thus results from the optimization will direct experimental testing (and thus aim to minimize the number of tests required to identify and validate a new design). Conversely, test results would be fed back into the flow analysis to improve the applied boundary conditions and validate the computational analysis.A successful effort would produce a robust, validated, and easy-to-operate computational tool for the geometrical design of fluid transport components that minimize head loss and dynamic structural loads. Given the need for and lack of such a capability together with prior experience in addressing internal flow problems in the nuclear industry CDI believes that with modest market entry, the combined sales, and associated service work, could generate ~$10M in sales over several years, with major cost savings for customers attributed to lower design and maintenance costs, improved power generation, and longer component service life.