A smart Monte Carlo procedure for production costing and uncertainty analysis
- Decision Focus Inc., Mountain View, CA (United States)
Electric utilities using chronological production costing models to decide whether to buy or sell power over the next week or next few weeks need to determine potential profits or losses under a number of uncertainties. A large amount of money can be at stake--often $100,000 a day or more--and one party of the sale must always take on the risk. In the case of fixed price ($/MWh) contracts, the seller accepts the risk. In the case of cost plus contracts, the buyer must accept the risk. So, modeling uncertainty and understanding the risk accurately can improve the competitive edge of the user. This paper investigates an efficient procedure for representing risks and costs from capacity outages. Typically, production costing models use an algorithm based on some form of random number generator to select resources as available or on outage. These algorithms allow experiments to be repeated and gains and losses to be observed in a short time. The authors perform several experiments to examine the capability of three unit outage selection methods and measures their results. Specifically, a brute force Monte Carlo procedure, a Monte Carlo procedure with Latin Hypercube sampling, and a Smart Monte Carlo procedure with cost stratification and directed sampling are examined.
- OSTI ID:
- 389864
- Report Number(s):
- CONF-960426-; TRN: IM9646%%149
- Resource Relation:
- Conference: 58. annual meeting of the American power conference, Chicago, IL (United States), 9-11 Apr 1996; Other Information: PBD: 1996; Related Information: Is Part Of Proceedings of the American Power Conference. Volume 58-II; McBride, A.E. [ed.]; PB: 886 p.
- Country of Publication:
- United States
- Language:
- English
Similar Records
Quantifying uncertainty in environmental risk assessment
An Integrated Paradigm for the Management of Delivery Risk in Electricity Markets: From Batteries to Insurance and Beyond [Slides]