Congress Resources: Papers, posters and presentations

< Return to abstract list

Integrated Operation Optimization Of Cascade Hydropower Projects

Congress: 2015
Author(s): Emrah Yalcin, Sahnaz Tigrek
Middle East Technical University1

Keyword(s): Sub-theme 10: Management of water resources,
AbstractEmrah Yalcina,*, Sahnaz Tigreka

a Department of Civil Engineering, Middle East Technical University, 06800 Ankara, Turkey

* Corresponding author Emrah Yalcin

Introduction Growing external energy dependence and rising oil prices are encouraging Turkey to turn to renewable energy. In this context, the Electricity Market Law No. 4628 and the revised establishment law of General Directorate of State Hydraulic Works No. 6200 opened a new era in the Turkish energy market by transferring operational rights of existing, under construction and planned hydropower plants to private sector and allocating water right licenses to develop new projects for producing electricity. Generation companies can sell their electricity through bilateral contracts, the renewable energy sources support mechanism or the day-ahead market operated by Market Financial Reconciliation Centre (MFRC). Companies have to report their sales method choices to Energy Market Regulatory Authority each year. The day-ahead market is the main structure of energy trade. Producers which prefer to sell electricity in the day-ahead market report their hourly expected production plans to MFRC. Appropriate predictions made for short-term productions of power plants contribute to not only system energy balance but also profits of companies. However, in most of cascade hydropower systems in the country, a single-reservoir simulation model is employed in operation of each of system reservoirs with a limited knowledge about long-and short-term operation strategies of upstream schemes. This causes energy imbalances and also widens range of energy prices. This study presents a catchment-based optimization model for integrated operation of hydropower plants under different sales methods.
Methods/Materials The key components of the system are database management, inflow modelling and forecasting, optimization and real-time operation. The assigned system integrates a database with basic hydrological and technical information to run the optimization algorithm. The optimization model is formulated in terms of nonlinear programming (NLP), which offers the most general formulation [2]. The objective function of the model is maximization of income, which is the product of produced energy and energy price. The constraint set includes flow continuity, turbine capacity, minimum release, minimum storage and reservoir capacity. Power releases are divided as firm and secondary considering different sales alternatives. The equations of volume-area, volume-elevation and turbine efficiency curves are expressed by high order polynomials in the NLP model. The system is established on a monthly basis for one-year period to assess production strategies for that year. To simulate real-time operations, inflow forecasts are utilized with frequent updating. The state of system reservoirs is updated at the beginning of each month according to observed inflow values of the previous month. This procedure can be continued with daily and hourly optimizations due to the floating energy prices in the day-ahead market. The proposed model was tested on the Garzan hydropower system as a case study. Garzan Creek is a branch of the Tigris River and flows in South-eastern Anatolia Region of Turkey. The hydropower system consists of the Aysehatun Dam and HEPP Project with Mutki Derivation, the Kor Dam and HEPP Project, the Garzan Dam and HEPP Project and the Garzan irrigation scheme that covers an area of 60000 ha [1, 3, 4, 6]. Mean, historical and forecasted flow values of dry and rainy seasons were presented to the system as input to analyse the limits and effectiveness of the NLP model in real-time operations. The results of the optimizations with historical and mean flow values demonstrate the range of income that can be derived for the period under consideration. In order to investigate how close results could be obtained to the upper bound, forecasted flows were generated with seasonal autoregressive integrated moving average (ARIMA) models using historical flow values [5, 8, 9]. In order to increase the forecasting performance of the ARIMA models, other hydroclimatic data including precipitation, temperature and evaporation can be integrated as independent variables. MINOS was employed to solve this nonlinear optimization problem in the General Algebraic Modelling System (GAMS) package [7]. The projects were analysed under different hydrological scenarios for the purpose of delaying or advancing the schedule of a power plant, expanding the capacity of existing plants or changing the normal and minimum operating levels of the system reservoirs. Moreover, in order to verify the efficiency of integrated operation, the same process was applied to the system reservoirs sequentially.
Results and Discussion It was seen in this study that improvement in accuracy of forecasts brings about economic benefits as a consequence of optimum reservoir operation. When the performance of the integrated algorithm is checked against the sequential optimization of the system reservoirs, the catchment-based optimization model produces more energy by maximizing head and by minimizing spill. Even a small percentage increase in energy production is in reality quite substantial.
Conclusion Integrated operation of cascade hydropower projects will contribute to avoid energy imbalances and enormous price differences. Instead of optimizing projects in themselves, basin scale operation models have to be applied to use hydropower potential more efficiently. 1. Aksa (Aksa Energy Generation Incorporated Company). (2004). Kor Baraji ve HES fizibilite raporu [Kor Dam and HEPP feasibility report]. Aksa Energy Generation Incorporated Company.
2. Barros, M.L.T., Tsai, F.T., Yang, S., Lopez, J.E.G., & Yeh, W.W. (2003). Optimization of large-scale hydropower system operations. Journal of Water Resources Planning Management, 129(3), 178-188. doi: 10.1061/(ASCE)0733-9496(2003)129:3(178)
3. DSI (XVIIth Regional Directorate of State Hydraulic Works). (1987). Bitlis-Garzan Projesi, Aysehatun Baraji ve HES planlama raporu [The Bitlis-Garzan Project, Aysehatun Dam and HEPP planning report]. Van: XVIIth Regional Directorate of State Hydraulic Works.
4. Enersu (Enersu Engineering Consultancy Construction Industry and Trade Limited Company). (2008). Garzan Baraji ve HES revize fizibilite raporu [Garzan Dam and HEPP revised feasibility report]. Ankara: Enersu Engineering Consultancy Construction Industry and Trade Limited Company.
5. Ghanbarpour, M.R., Abbaspour, K.C., & Hipel, K.W. (2009). A comparative study in long-term river flow forecasting models. International Journal of River Basin Management, 7(4), 403-413. doi: 10.1080/15715124.2009.9635398
6. Jemas-Su (Jemas-Su Groundwater Survey and Engineering Limited Company). (2001). Bitlis-Garzan Projesi, 1987 yilinda hazirlanan Aysehatun Baraji ve HES planlama raporunun revize bolumleri [The Bitlis-Garzan Project, revised sections of Aysehatun Dam and HEPP planning report prepared in 1987]. Jemas-Su Groundwater Survey and Engineering Limited Company.
7. Murtagh, B.A., Saunders, M.A., Murray, W., & Gill, W.E. (2014). MINOS solver manual. Retrieved from (accessed 12 October 2014).
8. Shabri, A., & Suhartono. (2012). Streamflow forecasting using least-squares support vector machines. Hydrological Sciences Journal, 57(7), 1275-1293. doi: 10.1080/02626667.2012.714468
9. Valipour, M., Banihabib M.E., & Behbahani, S.M.R. (2012). Comparison of the ARMA, ARIMA, and the autoregressive artificial neural network models in forecasting the monthly inflow of Dez dam reservoir. Journal of Hydrology, 476, 433-441. doi: 10.1016/j.hydrol.2012.11.017
2011 IWRA - International Water Resources Association - - Admin