Reading Assignment # 9

Article Reference:
Jenq-Tzong Shiau, Fu Chun Wu, Compromise Programming Methodology for Determining Instream Flow Under Multi-Objective Water Allocation Criteria, 2006, Journal of American Water Resource Association

Summary:
The paper by Jenq Tzong and Fu chun determines a quantitative assessment for determining in-stream flow under multi objective water allocation criteria. They explain the concept of compromise programming which is used to optimize water allocation scheme by minimizing hydrologic alterations and water supply shortages. This methodology was applied to a case study of Kaoping diversion weir in Taiwan. The paper focuses on evaluating the hydrologic alterations of the weir and the optimal operation schemes.

The Kaoping diversion weir was constructed to meet the increasing municipal water demands. The design diversion capacity for the municipal use is about 35 cumec. However prior to the construction there was long history of agricultural water withdrawals from the downstream of the Kaoping creek. From the data it is observed that there has been no flow diversion during dry season for meeting municipal demand due to insufficient amount of water. The operation model of weir has three components – meeting the reserved agricultural water demand, providing projected municipal water demand and instream flow release. Based on the environmental protection, they have set different priorities, first being instream flow release, second priority being registered agricultural demand and third priority being projection flow for municipal demand.

Range of variability of approach has been used to evaluate the hydrologic alterations caused by the flow diversions of the weir. RVA assesses the hydrologic regime with respect to 32 ecological relevant indicators of hydrologic alteration (IHA). As suggested by Richter et al. (1998) RVA range of variation for each IHA is taken as 25th and 75th percentile pre diversion values. Weir operations are aimed such that pre diversion flow conditions reach the targeted RVA ranges. Overall degree of hydrologic alteration (D) is estimated to measure the deviation between post impact flow regime and pre impact flow regime. Richter et al. (1998) classified the overall degree of alterations as low, medium and high alteration. Therefore value of D for a particular IHA quantifies the effect of flow diversion on the flow regime. For including these overall hydrologic alteration index needs to be estimated. Richter et al. suggested the method of averaging the 32 IHA degrees so as to have an overall impact index. By using this averaged index, high impact alterations would be offset by the low degree alterations. This paper considers the method proposed by Shiau and Wu (2004b) which calculates the single index for hydrologic alteration in categories of high and medium alterations.

A water supply deficit is estimated when the actual demand is greater than the registered or projected demand. In the optimizing model, water supply objective is to minimize the shortage ratio (total deficit / total demand) of agricultural water and municipal demand. Minimizing both hydrologic alterations and supply shortages is the goal of Kaoping diversion weir which is formulated as a multi objective programming.
Min {SRW, SRD, D}
SRW = supply shortage of registered agricultural withdrawals
SRD = supply shortage of projected municipal uses
D = overall degree of hydrologic alteration
All these supply shortages values are the function of the instream flow value (i.e. decision variable). There are many techniques used to deal with the multi objective problems, this study considers using of compromising programming.

Compromising programming estimates the optimal solution which has the least distance from the ideal point (i.e. point where multiple objectives reach the optimal solution). This compromise programming is carried out in two steps –
1. First the best and worst solutions of both the objective functions needs to be found which is within the computation domain (which is found out based on the decision variable- instream flow)
2. Seeking the optimal solution by calculating the shortest distance to the best estimated point

In the current operation of the weir, it releases a minimum of 9.5 cumec instream flow. The overall hydrologic alteration is about 69.3% i.e categorized as highly altered. Though agricultural withdrawals are the highest priority 18% shortage of water exists when compared to 6% of municipal water shortage. With the highest percentage of alteration it is evident that the present instream release is not sufficient to restore the natural flow variability. To evaluate effect of different instream flows, Shiau and Wu studied different flows ranging from o to 100 cumec. Corresponding SRW, SRD, D and value of L were estimated. This study showed that SRW increased rapidly with smaller instream than that of larger instream flow. SRD increased linearly. Value of D decreased with the increase in the instream flow. Two drops were observed – one (26 cumec) to convert from highly altered to medium alteration and the second drop (93 cumec) to convert from medium alteration to low alteration. For a given instream flow the SRW, SRD and D values are fixed, but the value of L changes with respect to the weighing factors. Therefore a study of various weighting factors was considered to see how the instream flow and the Minimum distance L is changed. The proposed optimal instream flow 26 cumec would not be able to restore the altered hydrologic regime to pre diversion condition but then it helps in mitigating the adverse effects of the highly and moderately altered IHA’s.

Discussion:
The paper was very informative about how we can optimize a model considering multi objectives i.e. minimizing water shortages and hydrologic alterations. With the increase of municipal demands, many reservoirs have been affecting the riverine environment by reduced instream flow. By using the quantitative assessment of 32 relevant IHA’s it would be helpful to check the hydrologic alterations caused due to this reduced instream flow. Future research would obviously be considering the biological component into the RVA. It would be interesting to note the changes in the optimized results with the added biological component. Compromising programming is a useful technique which can be used in many models to optimize two conflicting objectives for example flood management vs. downstream eco system,