The effect of applying river environmental requirements on the design and operation characteristics of storage reservoirs using the Monte Carlo method (case study: Nazlou reservoir dam)

Introduction

In recent years, one of the most important challenges in planning and operating reservoir systems is the estimation of actual (sufficient) environmental flows in rivers (downstream) during operation years and evaluating its effect on the performance of such reservoir systems. Failure to comply with the environmental flows of the rivers downstream has had long-term destructive consequences for socio-ecological systems. Therefore, accurate and timely estimation of environmental flow and its application in the operation of existing and new reservoir systems is of particular importance.

Several studies have been conducted to assess environmental flows and the resulting changes in river ecology after the construction of reservoir dams. All studies have investigated and evaluated the flow regime before and after dam construction on rivers and have emphasized the need for timely allocation of environmental flows and changes in reservoir system operation patterns. Studies of the environmental flows of rivers have mostly focused on estimating environmental flows using various methods and comparing the river flow regime before and after dam construction in both current and ecosystem-compatible conditions in the historical period. However, the effect of timely release of environmental demand compatible with the ecosystem on the long-term behavior of various important characteristics of the storage reservoir system (such as useful volume, command curve, hydrological behavior, critical period, evaporation losses, etc.) has not been considered in previous studies, and this could be an innovative approach in this field, which has been carried out in the present study. For this purpose, in this study, a comprehensive Monte Carlo simulation approach is carried out to compare several important common methods for estimating the environmental demand of the river (dam downstream) and evaluate their effects on important design characteristics (such as useful volume) and operation (such as the command curve) in Nazlou reservoir system. The Nazlou river is one of the important rivers of the Lake Urmia basin. After providing drinking and agricultural water to the lands under its coverage, it is discharged into Lake Urmia.



Materials and Methods

In this study, the Nazlou Reservoir Dam and the Nazlou River in Urmia were selected as a case study. The Nazlou River is one of the important rivers of the Lake Urmia basin. After providing drinking and agricultural water to the lands under its coverage, it is discharged into Lake Urmia. Methods for determining environmental demands in the river downstream are classified into three groups: hydrological, hydraulic, and habitat simulation, but hydrological methods, with the availability of the required data, are used by researchers as reliable and common methods. In this study, five common hydrological methods, which are preferred by researchers, namely Desktop Reserve Model, flow duration curve, Smakhtin, Tessman, and Tennant methods, were used to estimate environmental demands. This comprehensive study is based on a Monte Carlo simulation approach with a very large process (924,000 times) of stochastic simulation of the storage system, i.e. 1,000 time series of production flow, 7 demands (DAnnual=0.1 to 0.4 MAF, step 0.05), 11 time reliability indices (Rel=0.9 to 1.0 step 0.01), 2 vulnerability indices (Vul=0.0 and 0.3), and 6 methods of determining environmental demand (the 5 above-mentioned methods + 1 Method without applying environmental demands). The aforementioned Monte Carlo simulation approach is based on a hybrid model combining the Valencia-Shaake disaggregation stochastic model and the Modified Sequent Peak Algorithm (SPA) simulation model.



Results and Discussion

The main results of this study are summarized as follows: a) Comparison of the environmental demand values obtained from the five applied methods showed that the environmental demand based on the Tennant and Tessman method were 20% (as minimum value) and 53% (as maximum value) of average annual flow, respectively, and the obtained values of the other three methods are approximately close to each other and are about 28% of the average annual flow. The monthly variation of the obtained environmental flow in river also indicated that the results of the three methods Tessman, FDC shifting (C) and DRM (C) had a similar behavior with the mean monthly historical flow values. b) The results showed that the performance of the used stochastic model in reproducing the statistical parameters of historical flow data at both annual and monthly levels and reproducing the correlation structure between the flows of different months and the correlation between the annual and monthly flows is quite desirable, and this plays a key role for a reservoir system simulation in real conditions. c) The results indicate a systematic long-term behavior of main reservoir storage characteristics (i.e., active storage, resiliency index, evaporation loss, and critical period) against demand for two condition, without and with including environmental flow, with the exception of the Tessman method.



Conclusion

The main outcome of investigation could be summarized as follows: a) The ecological system of rivers often adapts to its monthly flow regime over time, therefore, using a method of estimating the ecological demands of a river in accordance with the monthly flow regime and with an appropriate amount (30-40% of the average monthly flow) has a more relative advantage than other methods. So, FDC Shifting (C) method, with an average of 32 percent of the average monthly flow and the minimum coefficient of variation (CV=0.20), was the most appropriate method for estimating the environmental demand in tested reservoir dam. b) The combination of the annual AR(1) model and the monthly Valencia-Shaake disaggregation model has the ability to reproduce the statistical characteristics of the historical streamflow values at both annual and monthly levels sufficiently, and this provides a reliable framework with high accuracy in the stochastic analysis of a reservoir system. c) The long-term storage-demand-performance relationships of the reservoir system due to environmental demand, with the exception of the Tessman method, show a similar systematic behavior change for different methods of estimating environmental demand. d) By applying environmental demands to the analysis of storage reservoir systems, the active storage, evaporation loss, and critical period increase by 2 to 10 times.