Hydro inflow forecasting and virtual power plant pricing in the Turkish electricity market

---

Tezin Türü: Yüksek Lisans

Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Uygulamalı Matematik Enstitüsü, Aktüerya Bilimleri Anabilim Dalı, Türkiye

Tezin Onay Tarihi: 2016

Öğrenci: SEZER ÇABUK

Danışman: SEVTAP AYŞE KESTEL

Özet:

Hydro inflow forecasting with most accurate quantitative models is a very crucial subject for effective hydro optimization, virtual power plant pricing, volume risk management and weather derivatives pricing in the Turkish electricity market. Predicting increase or decrease in hydro inflow, seasonal characteristics of hydrological years such as wet, dry or normal, allow the decision makers to economically use water for optimal periods, quantify of volume risk and determine effective portfolio management strategies. In this study, we focus on defining and pricing a hydroelectricity power plant as a Virtual Power Plant (VPP). For pricing of this non-standard option, we worked on inflow and price scenarios and optimization model with the possible real world constraints. For the hydro inflow forecasting that will be used in optimization model, we applied Seasonal Autoregressive Integrated Moving Average model with Exogenous Variable (SARIMAX), whereas lagged indexed precipitation data, having the highest correlation with historical inflow data, is included as exogenous variable. In addition to point forecast of hydro inflow, we generated various inflow scenarios by using the distribution of model fit residuals as stochastic processes for defined VPP. Moreover, we worked on hydro optimization problem where objective function is maximizing the expected value of generation by tracing to generated inflow and price scenarios. Price scenarios are simulated by using the hourly shape of historical Day Ahead Market (DAM) prices. As a result, we could analyze the optimization outputs according to different price and inflow levels. For defined VPP, Volume at Risk measure is expressed to explain the meaning of risky volume for the valuation of VPP. Furthermore, in the last section of the study, by the help of the flexibility of optimization model, by using different constraints, we worked on the constructions, solutions and evaluations of different optimization cases as a significant contribution in academic literature and common practice in current electricity markets.