Computational Modeling of A Williams Cross Flow Turbine

Hydropower is not only the most used renewable energy source in the United States, but in the world. While it is well known that large hydropower facilities, like the Hoover Dam, provide large amounts of electrical power, there is also a tremendous opportunity for hydroelectric power generation from small scale facilities that has largely been overlooked. The work being presented here studies a new cross flow turbine called the Williams Cross Flow Turbine (WCFT), which was designed to extract electric energy from low head, run-of-the-river, small hydropower sites. To spur the implementation of the WCFT in small hydropower applications, and thus to spur the development of small hydropower, the work here is focused on developing a detailed computational fluid dynamics (CFD) model of the WCFT. The computational model produced as part of this work was developed in the commercial software ANSYS Fluent. This CFD model solves the incompressible Navier-Stokes equations in their three-dimensional, unsteady form including the effects of turbulence, using detailed numerical routines. The multiphase fluid flow of air and liquid water in the turbine is simulated using a volume of fluid technique (VOF). A very detailed geometric representation of the WCFT turbine is imported into ANSYS from the computer aided design software SOLIDWORKS. Using commercial software made the development of this detailed CFD model possible within a Master’s thesis time frame. Coupled to the computational work done here, is some experimental work. Having access to the WCFT experimental facility at Central State University was beneficial to the computational work. A good deal of knowledge about the computer modeling was gained by undertaking a small amount of experimental work. In addition, an experimental result was used to verify the average power predicted by the computational model. This comparison showed a difference of 18.1%, which is deemed reasonable given the complexities of the CFD modeling undertaken. To further verify the computational results, independence of the results on the meshing and time step utilized was demonstrated. The primary computational results presented are plots of turbine power versus shaft rotational speed for twelve and nine bladed WCFTs. These results indicate that a nine bladed WCFT turbine performs better than the currently used twelve bladed, lab-scale WCFT. While these results are specific to the input operating conditions used in the analysis, they indicate the usefulness of the computational tool developed here. Additional results presented for the nine bladed, lab-scale WCFT are field plots of water volume fractions, many types of velocity vector plots, and field plots of the fluid pressure. Histogram plots of some of the interesting quantities are given to show the distribution of certain quantities throughout the turbine