140 likes | 300 Views
ÖGOR Energy Workshop (EW06) Pricing Models for Electricity Markets Dr.-Ing. Hagen K. Schmöller Wien, 17.03.2006. From Interconnected Hydro Power Plants to Trading Recommendations: Possible Applications for Stochastic Programming. Introduction. stochastic programming
E N D
ÖGOR Energy Workshop (EW06)Pricing Models for Electricity MarketsDr.-Ing. Hagen K. SchmöllerWien, 17.03.2006 From Interconnected Hydro Power Plantsto Trading Recommendations:Possible Applications for Stochastic Programming
Applications for Stochastic Programming | ÖGOR Energy Workshop | 17.03.2006 | Hagen Schmöller Introduction stochastic programming so far not in operational use stochastic pricing models and Monte-Carlo-Simulations for individual investigations cooperation with FH Vorarlberg and University of St. Gallen project duration: June 2005 – May 2008 improvement of portfolio management system forecast and modeling stochastic optimization risk management IT integration three workshops presentation of project partners analysis of requirements definition of different work packages first results improving the deterministic model design
Applications for Stochastic Programming | ÖGOR Energy Workshop | 17.03.2006 | Hagen Schmöller Outline the real system of interconnected hydro power plants modeling challenges: reflection of contractual shares application Iexercise price: stochastic optimization of a fictive model application IItrading recommendations: optimized supply curve conclusions system of interconnected hydro power plants EEX spot market: trading recommendations
Applications for Stochastic Programming | ÖGOR Energy Workshop | 17.03.2006 | Hagen Schmöller Interconnected Hydro Power Plants (ii) highly interconnected reservoir system partner: VKW AG TIWAG AG EnBW AG several agreements from 1922 through 2000 determine the shares of the partner functions: peak load energy pump storage energy balancing power black start power supply guarantee for Vorarlberg flood protection
Applications for Stochastic Programming | ÖGOR Energy Workshop | 17.03.2006 | Hagen Schmöller Modeling Challenges (i) dispatch of individual machines by different partners + precise arrangement - economical optimum not achievable fictive and real dispatcher one partner determines the real dispatch of the system other partners determine the dispatch of a fictive model reflecting their shares requirement: fictive model as simple as possible, but economical equivalent +economical optimum achievable + not all details have to be modeled - difficult to consider changing general conditions market rules and prices inflow situation partner structure system structure
Applications for Stochastic Programming | ÖGOR Energy Workshop | 17.03.2006 | Hagen Schmöller Modeling Challenges (ii) different levels of abstraction over the history fictive model of the whole system including all constraints simplified 3-reservoir-model 2 yearly reservoirs 1 daily reservoir current fictive model 3 reservoir accounts energy limits percentage of natural inflow capacity differentiation scheduled capacity (declaration day before) variable capacity conversion factor for pumped energy
Applications for Stochastic Programming | ÖGOR Energy Workshop | 17.03.2006 | Hagen Schmöller Application I: Exercise Price (i) dispatch of the fictive system objective: maximization of profit margin and risk management (e.g. value at risk) application I coordination with short term optimization (weekly update) exercise prices associated capacities capacity turbine operation price exercise prices pumping operation
Applications for Stochastic Programming | ÖGOR Energy Workshop | 17.03.2006 | Hagen Schmöller Application I: Exercise Price (ii) time pattern hour time horizon remaining + succeeding year (considering current reservoir account levels) uncertainties (up to now calculation of different scenarios) inflows prices forecast and stochastic models needed constraints contractual aggreements of the fictive model restrictions due to maintenance
Applications for Stochastic Programming | ÖGOR Energy Workshop | 17.03.2006 | Hagen Schmöller Application I: Exercise Price (iii) • repeating optimization • stochastic ... 1 2 weeks current year succeeding year peak price +80% over 100% increase of exercise price in 2005 quarter IV of 2005 inflows50% below standard
Applications for Stochastic Programming | ÖGOR Energy Workshop | 17.03.2006 | Hagen Schmöller Application II: Trading Recommendations (i) application II trading recommendation for the EEX spot market coupling with commitment of other power plants open positions no feedback of the market due to small market share reserve provision water management requirements spot price forecast comparable historic data weather forecast (fog, cloud cover, ...) market assessment broker prices
Applications for Stochastic Programming | ÖGOR Energy Workshop | 17.03.2006 | Hagen Schmöller Application II: Trading Recommendations (ii) day ahead realized price exercise price price price forecast energy limits? capacity turbine operation pumping operation
Applications for Stochastic Programming | ÖGOR Energy Workshop | 17.03.2006 | Hagen Schmöller Conclusions different optimization tasks concerning time horizon pricing models constraints complexity requirements for operational use transparency adjustability output of parameters: e.g. volatility, … changing market conditions integration in IT-environment integration in work flow