MARKOV DISCRETE WATER SOURCE MODEL AND THE OPTIMIZATION OF HYDROELECTRIC EXPLOITATION OF CITARUM CASCADE RESERVOIR FOR DOWNSTREAM RAW WATER NEEDS

• Main Author: Iskandar Muda Nasution, Eddy
• Format: Dissertations
• Language: Indonesia
• Subjects: Teknik saniter dan perkotaan; teknik perlindungan lingkungan
• Online Access: https://digilib.itb.ac.id/gdl/view/66926
• Institution: Institut Teknologi Bandung

The Citarum River is one of the strategic rivers in West Java Province, has a watershed area of about 6,600 km2 with its headwaters at Gunung Wayang, Bandung Regency and empties into the Java Sea in Bekasi Regency, flowing along 297 km with an average annual flow volume of 5.5 billion m3. Citarum cascade reservoirs, starting from upstream to downstream, namely the Saguling, Cirata and Jatiluhur reservoirs which function mainly as power plants with increasing demand reaching 265,051 TWH in 2030. Especially for the Jatiluhur reservoir, it has other functions as irrigation, flood control and water sources. especially for the fulfillment of DKI Jakarta with an estimated demand of 31,875 m3/second in 2030. The characteristics of the monthly average regional rainfall for the 1995-2019 period in the Citarum cascade reservoir show a monsoon-type rain pattern, namely one peak of rain (unimodal) which occurs in March (Saguling reservoir and Cirata reservoir) and February (Jatiluhur reservoir). This study area is also included in the Tropical Climate Zone, with the characteristics of high rainfall variations in the rainy season and low rainfall in the dry season. The problem in the Citarum cascade reservoir is the deviation of the realization of the reservoir trajectory from the planned rule curve. This is because the operation of the reservoir cannot predict the year of the season that will occur. Another problem is when cascade reservoirs have different functions and are operated by different managements, this will create a conflict of interest. To overcome this, it is necessary to operate the right reservoir with the objective of maximizing hydroelectric benefits, meeting the needs of raw water downstream and no water being wasted through the spillway. The Citarum cascade reservoir has unique characteristics, which are a function of space and time. Therefore, its management as an integrated management unit from upstream to downstream which is dependent with the concept of one river one management. Each reservoir is treated as having a local watershed with a different hydrological regime, so that it is dependent on the reservoir upstream. Thus, the conceptual model of the hydrological regime in the Citarum cascade reservoir, namely the reservoir input is the reservoir output upstream plus local discharge. The discharge sensibility in the Citarum cascade reservoir tends to increase, this is an indication of the extremity of discharge as a result of changes in the hydrological regime due to land changes with reduced cover area which can increase run off. Markov's discrete model as a stochastic model is proven to be able to anticipate random and stochastic input discharges to the reservoir with dry, normal and wet years, resulting in a guide trajectory close to the actual trajectory. The results of the calculation of the allocation of raw water for drinking water (R20 dry) in the Jatiluhur reservoir (81.43 m3/second) are sufficient to meet the needs of DKI Jakarta in 2030 (31.87 m3/second), so that the calculation (R5 dry) in the reservoir Jatiluhur (123.79 m3/second), there is still an allocation for agricultural irrigation (R5 dry - R20 dry) of 42.36 m3/second. The optimization model for the hydroelectric management of the Citarum cascade reservoir using an iterative du couloir model on changing electricity prices produces a better correlation value than fixed electricity prices, amounting to 0.792 (Saguling); 0.864 (Cirata) and 0.633 (Jatiluhur). This conceptual model shows strong results - so strong, that the actual trajectory is close to the guideline trajectory, which represents the concept of SPAM development. The results of the normal discharge hydroelectric comparison, using the du couloir iterative method, obtained that the gain in electricity prices changed was greater than the fixed electricity prices (16.72% in Saguling; 22.45% in Cirata and 49.23% in Jatiluhur). For renewable energy sustainability and the increasing demand for raw water in the downstream, the optimization model uses a dynamic program with changing electricity prices in order to integrate the management of 3 (three) reservoirs in the Citarum cascade. The Bellman dynamic program has succeeded in simulating a conceptual model for optimizing reservoir management that is integrated within the one river one management frame. This program still has shortcomings in overcoming the curse of dimensionality, so the iterative du couloir method is applied. This method is able to produce volume discretization to infinity (~0) with computation time exponentially much faster 66 (sixty six) times than the Bellman dynamic model at volume discretization of 10 million m3. With the application of the du couloir iterative method, it is able to produce the most optimal reservoir management, so as to increase electricity production at peak hours to meet electricity needs in the Java-Bali system in a sustainable manner.