Sensitivity Analysis and Water Quality Modeling of a Tidal River Using a Modified Streeter–Phelps Equation with HEC-RAS-Calculated Hydraulic Characteristics

Abstract

A modified Streeter–Phelps equation and the Hydrological Engineering Centers River Analysis System (HEC-RAS) were combined to assess water quality of the Tan-Sui River and its tributaries. The Tan-Sui River is the main source of water supply for northern Taiwan, and the water quality of its stream is significantly affected by tides. In this study, HEC-RAS was employed to assess the impact of tides on water quality and to calculate reoxygenation coefficients. The modified Streeter–Phelps equation was used to calculate water quality in terms of contaminant degradation and reoxygenation. Biochemical oxygen demand (BOD) and ammonia nitrogen (NH3–N), the most important identified sources of water pollution in the rivers investigated, were evaluated. Dissolved oxygen (DO) was also simulated, since it is often used as a staple of water quality. Results showed that employing HEC-RAS for hydraulic calculations improves the modified Streeter–Phelps simulation. In river sections without tidal influence, water quality was sensitive to the BOD and NH3–N degradation constants. Downstream of Chin-Mei Creek, while the BOD degradation constant decreased by 80%, BOD and DO concentrations increased from 7.1 mg/L to 10.7 mg/L and 5.0 mg/L to 7.2 mg/L, respectively, indicating that water quality was not as sensitive to variations of the BOD degradation constant as expected. The concentrations of DO and BOD at the river mouth had a significant impact on water quality for the tidal sections of the investigated rivers due to mixing of tidal and river waters. The BOD and NH3–N degradation constants in the tidal sections had little impact on water quality simulations. This study demonstrated the innovative combination of the modified Streeter–Phelps equation and HEC-RAS to assess the impact of tidal variation and to simulate the water quality of a tidal river when available data is rather limited.