Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Optimisation of physical and financial power purchase portfolios

: Spangardt, G.; Lucht, M.; Althaus, W.

Fulltext (PDF; )

Optimization Online, 2002, 17 pp.
Electronic Publication
Fraunhofer UMSICHT Oberhausen ()
electric utility; small and medium-sized business; demand for electricity; power trading; portfolio management; risk management; stochastic optimization (SO); multicriteria optimization; modelling

The deregulation of the European power market brings new sales prospects for the power-suppliers as well as an appreciable increase of entrepreneurial risks. In order to handle the novel price- and volume-risks, the optimisation of decisionmaking under uncertain boundary conditions is of essential interest. The former task of resource management in energy-supply was the minimisation of costs for the fulfilment of a foreseeable power-request at long-ranging conditions of pricing. Now a multicriterial optimisation problem arises: simultaneous minimisation of cost and risk.
In the last years a number of power-exchanges have been established where power is physically traded day-ahead and derivatively as future contracts. Furthermore, financial power-derivatives like Options, Caps, Floors or Swaps are traded bilateral in the so called over-the-counter market. A serious question is how to use different physical contracts and financial derivatives in an optimal way to protect a power purchase portfolio against market risks.
Facing this question, a multicriterial linear stochastic optimisation model has been developed. It is based on scenarios for the market price, generated by Monte-Carlo-simulation that uses a mean-reversion market model calibrated for the German power spot-market. The different optimisation criteria are merged into a single objective by a weighted summation. Individual risk aversion is considered by the coefficients of the weighted sum. The model is adapted to resolvability by Benders-Decomposition in a way that even a larger number of optimisations with different coefficients of the weighted sum can be solved in acceptable time giving an idea of the shape of the efficiency-frontier. Nevertheless all important microeconomic features have been taken into account.