Handling uncertainty in quantitative estimates in integrated resource planning

Technical Report · Sun Jan 01 04:00:00 EST 1995

DOI:https://doi.org/10.2172/28313· OSTI ID:28313

Tonn, B E ^[1]; Wagner, C G ^[2]

---

1. Oak Ridge National Lab., TN (United States)
2. Univ. of Tennessee, Knoxville, TN (United States). Dept. of Mathematics

This report addresses uncertainty in Integrated Resource Planning (IRP). IRP is a planning and decisionmaking process employed by utilities, usually at the behest of Public Utility Commissions (PUCs), to develop plans to ensure that utilities have resources necessary to meet consumer demand at reasonable cost. IRP has been used to assist utilities in developing plans that include not only traditional electricity supply options but also demand-side management (DSM) options. Uncertainty is a major issue for IRP. Future values for numerous important variables (e.g., future fuel prices, future electricity demand, stringency of future environmental regulations) cannot ever be known with certainty. Many economically significant decisions are so unique that statistically-based probabilities cannot even be calculated. The entire utility strategic planning process, including IRP, encompasses different types of decisions that are made with different time horizons and at different points in time. Because of fundamental pressures for change in the industry, including competition in generation, gone is the time when utilities could easily predict increases in demand, enjoy long lead times to bring on new capacity, and bank on steady profits. The purpose of this report is to address in detail one aspect of uncertainty in IRP: Dealing with Uncertainty in Quantitative Estimates, such as the future demand for electricity or the cost to produce a mega-watt (MW) of power. A theme which runs throughout the report is that every effort must be made to honestly represent what is known about a variable that can be used to estimate its value, what cannot be known, and what is not known due to operational constraints. Applying this philosophy to the representation of uncertainty in quantitative estimates, it is argued that imprecise probabilities are superior to classical probabilities for IRP.

• Research Organization: Oak Ridge National Lab., TN (United States)
• Sponsoring Organization: USDOE, Washington, DC (United States)
• DOE Contract Number: AC05-84OR21400
• OSTI ID: 28313
• Report Number(s): ORNL/CON--388; ON: DE95007292
• Country of Publication: United States
• Language: English

Related Subjects

29 ENERGY PLANNING, POLICY, AND ECONOMY
99 GENERAL AND MISCELLANEOUS
CAPACITY
DATA COVARIANCES
ELECTRIC UTILITIES
LOAD MANAGEMENT
PLANNING
POWER DEMAND
STATISTICS