Low-flow criteria for water quality standards for selected rivers

Abstract

Low flow quantiles are commonly used to describe stream characteristics and are used for establishing water quality standards. The objective of this study is to investigate the surface water quality in terms of parameter concentrations and specified standards as well as to develop low flow criteria for water quality standards. Water quality and stream flow data for Brahmaputra at Bahdurabad, Surma at Sylhet and Karotoa at Panchagahr were collected and analyzed. The Department of Environment collects surface water quality samples without any information and reference to the water discharge values at the time of sampling. Analyses performed in this study began with obtaining concurrent discharge values from BWDB database for the dates on which water quality data were available. To gain an insight of the discharge-water quality relationships, first the variation of water quality parameter concentrations with discharge was investigated by visual inspection of plots and linear regression. For comparison, water quality standards for irrigation use, fish culture and drinking water were superimposed on these plots. In case of Brahmaputra river measured BOD concentrations exceeded both the drinking water and fish culture standard, whereas for Surma river about a half of the measured values exceeded the standard, and for Karotoa river concentrations were found to be below the standards for all uses. Plots for DO and pH showed that their amounts were well above the standard values for all of the three rivers. All measured concentrations of coliform in Brahmaputra exceeded the drinking water standard, several values exceeded irrigation standard, and fish standard was generally met. It is seen that the discharge water-quality plots showed substantial scatter although a decreasing trend in parameter is expected with increasing discharge. It may be due to the fact that all the available water might not serve to dilute the polluting substances. In rivers it is not uncommon for waste effiuents to remain segregated along a shore line for many miles below the outfall. Regression equations relating water quality with discharge were not found to be useful in predicting parameter concentration as a function of discharge in term of such criteria as correlation coefficients and other test statistics. One of the critical elements of any water quality standard is the low-flow criterion to be applied. Usually low flows corresponding to 7-day, lO-year return period or lO-day, 25- year return is used. These flow values were computed using best-fit probability distributions. Then to assess the exceedence probabilities of 7Q 10 and lOQ25 flow values, flow duration curves were obtained for each of 7-day and lO-day annual low flow series. Computed 7Q 10 and lOQ25 flows were found to exceed 99.99 percent of the time. Estimates of such exceedence probabilities provide guidance in setting level of effiuent treatment required for protection of water quality. For example if the 7-day, lO-yr flow is equalled or exceeded 99% of the time on a given river, it would be required to provide effiuent treatment which should be adequate 99% of the time.