A Study on the Variations of Model Performance between QUAL2K and WASP for a River with High Ammonia and Organic Matters

• Main Authors: Ping-Kuei Tang, 湯平貴
• Other Authors: Yi-Chu Huang
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/98285900468289540267

博士 === 國立屏東科技大學 === 環境工程與科學系所 === 101 === Two water quality models, QUAL2K (River and Stream Water Quality Model) and WASP (Water Quality Analysis Simulation Program), were employed to calibrate and verify the biochemical parameters spatially and temporally based on the data collected from 1965 to 2012, four surveys sampled on 25, 27 of November and 4, 8 of December in 2012, and three surveys in March 13 of 2008, October 9 of 2010, January 7, 2012. Furthermore, sensitivity evaluation, allowable discharge amounts, and cost analysis were also conducted. Through the calibration, validation, and re-validation processes, two selected water quality models are identified as applicable to the Tungkang River basin. The model results were reasonable evaluated by the mean absolute percentage error (MAPE) method. Several feasible preventive measures such as promotion of clean pig farming practices, wastewater treatment facilities, shallow bed gravels contact oxidization process as well as constructed wetlands were proposed to improve the water quality of the Tunggang River by cutting down the pollutant loading. Through the sensitivity diagram analysis, the selected biochemical parameters of water quality can be applied to various modes of reaches or segmentation and to long-term or short-term water quality variations. QUAL2K possessed a better tolerance to dispersion coefficient within the non-tidal reach. The results showed that the allowable discharge amount significantly varied between the wet and dry seasons. Allowable discharge amount during the dry season was only one third of that during the wet season. Dissolved oxygen during the dry season was frequently not to meet the Class B water quality standards. The variations of allowable discharge amount were around 23% between WASP and QUAL2K to achieve the Class B water quality standards regulated by Taiwan EPA. In this study, biochemical coefficients were collected from field sampling and laboratory tests as inputs to the QUAL2K and WASP/EUTRO models. Except dispersion coefficients were obtained using the conservative electric conductivity (EC) via model calibration, the other four biochemical coefficients were calibrated and validated by QUAL2K, and then transferred to WASP model with graph method. Graph transformation is applicable in the water quality simulation of the Tunggang River with 3% error which is within a reasonable range. The error may attribute to the lag effects of the drainage pollutant loads transporting into various grids. Overall, the model results including calibration, validation, re-validation, sensitivity evaluation, allowable discharge amount, cost analysis of WASP works better than those of QUAL2K which possesses the advantage of easier operation.