Open Collections

Measuring and modeling sediment dispersion in small streams

University of British Columbia

Turbidity-based events in multiple-use watersheds can potentially lead to negative impacts on water quality. This is of particular concern in small streams that serve as drinking water sources in the Province of British Columbia. The one-dimensional advection-dispersion equation (1D ADE) is the most common approach to modeling the transport of substances in flowing water. However, relatively little is known about its applicability to suspended sediment, especially regarding the sink term that accounts for sediment settling. The aim of this study was to assess the degree to which the 1D ADE accurately predicts suspended sediment dispersion in small channels. In addition, an evaluation of the applicability of predictive formulas for the longitudinal dispersion coefficient to small channels was undertaken. Tracer experiments were conducted in three different channels: (1) a concrete channel; (2) a semi-natural channel; and (3) a natural channel. Sodium chloride and suspended sediment were injected simultaneously in the channels. The sediment particle sizes ranged from <0.075 mm to 1 mm. Sodium chloride was treated as a conservative tracer (i.e., no losses or gains during transport), and the sodium chloride plumes were modeled first to obtain best-fit estimates for the longitudinal dispersion coefficients. Suspended sediment plumes were modeled subsequently using the best-fit longitudinal dispersion coefficients from the sodium chloride plumes with an additional settling rate parameter to account for the sediment loss. The 1D ADE was capable of reproducing the observed curves with ±50% relative error. The settling rate term was found to be essential to properly simulate the suspended sediment plumes. The commonly used formula for settling rate (settling velocity/depth) overestimated the loss of particles and it was not applicable to the observed data. Twenty-six predictive formulas for the longitudinal dispersion coefficient were evaluated on their ability to reproduce the observed plumes. None of the predictive formulas were able to predict the dispersion process in the small channels with less than ±50% error. The formulas from Sattar and Gharabaghi (2015) had the best performance overall. Findings from this thesis can serve as a guideline for engineers and scientists working with tracer data and water quality models in small streams.

Text

2020-03-10

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/73706

Earth and Environmental Sciences

University of British Columbia

UBCO

http://creativecommons.org/licenses/by-nc-nd/4.0/