The Utah Water Research Laboratory, in conjunction with the State of Utah, initiated a study to determine the accuracy of a wide variety of flow measurement devices in Utah. The project selected 70 sites with flow measurement devices throughout the state. During the assessment each device had its physical condition and flow measurement accuracy documented. Although a wide variety of flow measurement devices were tested, the majority were Parshall flumes. Many of the assessed Parshall flumes were not measuring flow to the specified ±5 percent design accuracy. Problems in flow measurement were due to issues with the staff gauge location and incorrect entrance geometry. Laboratory tests were conducted at the Utah Water Research Laboratory in an attempt to provide accurate flow measurement from flumes with these issues. The tests simulated incorrect locations for measuring upstream head with different entrance geometries on a 2-ft-wide Parshall flume. The flume was tested with three different entrance wingwall configurations, eighteen stilling wells, and two point gauges, allowing water surface profiles to be collected throughout the flume. Corrections for incorrect head measurement locations and entrance geometries were created. The objective of this thesis is to provide water users and regulators with the information necessary to help improve open-channel flow measurement accuracies. An overview of design accuracies and flow measurement devices is given. In addition, a method to correct for incorrect head readings in Parshall flumes, a widely used flow measurement device in Utah, is presented. It is expected that this information will help water users and regulators monitor their water with the understanding necessary to ensure that water is more accurately measured.

Studying the behavior of hydraulic structures under various extreme conditions is far beyond the reach of traditional build-test experimental methods. Following the typical method, it is necessary to provide the downscaled model to be used in the laboratory and determine various structural parameters against unforeseen scenarios, which should be mimicked in the laboratory. Usually, human and instrument errors as well as scale effects are some of the causes of inaccurate results; therefore, substitute methods have always been sought to determine the stability and efficiency of various hydraulic devices. The implementation of computer models, also referred to as numerical simulation, is one of the most efficient ways to reduce time and cost, and at the same time, add to the degree of confidence in the design process. Improvements in computational power of supercomputers in recent decades have led researchers and engineers to become familiar with these numerical models and implement them in various studies. One of the basic hydraulic structures that is widely used to measure the flow for open channels is the Parshall flume. Although the Parshall flume is simple to use, the application of various rating equations for different sizes highly affects the output value, which is the flowrate. To avoid this, appropriate rating equation must be developed for various sizes that are not listed in the standard Parshall flume size chart. With the help of the Computational Fluid Dynamic (CFD) techniques, numerous turbulence models i.e., standard k-ε, RNG k-ε, realizable k-ε, k-ω, k-ω SST, k-ω SST DES, Smagorinsky and Dynamic k equation, have been used to simulate different geometric setups for different sizes of Parshall flumes. The result from various families of turbulence models, i.e., Reynold Average Navier-Stokes (RANS), Large Eddy Simulation (LES) and Detached Eddy Simulation (DES), used in this study, provide promising values with acceptable margins of error, which were found to be less than 3% in all cases except one. The application of numerical modeling to simulate the flow in Parshall flumes is used to verify the reliability of applying OpenFOAM as the open-source CFD used for all the simulations in this study. The data obtained from the numerical simulations are considered a reliable source to adjust the rating equation for any future non-standard Parshall flume. Overall, it should be pointed out that the quality of non-linear turbulence models, i.e., Shih-Q, LC, and v²-f, were considerably higher than those obtained using linear turbulence models.