ORNL Develops the Oak Ridge Competitive Electric Dispatch (ORCED) Model for Simulating the Operations and Costs of Bulk Electric Power Markets
The Oak Ridge Competitive Electricity Dispatch (ORCED) model dispatches the power plants in a region to meet the electricity demands for any given year up to 2030. It uses public sources of data describing electric power units from the National Energy Modeling System (NEMS) or other sources and hourly demands from utility submittals to FERC that are projected to a future year. The model simulates a single region of the country for a given year, matching generation to demands, assuming no transmission constraints within the region and limited transmission in and out of the region. ORCED can calculate a number of key financial and operating parameters for generating units, including average and marginal prices, air emissions, and generation adequacy. By running the model with and without demand changes such as plug-in hybrids or distributed generation, the marginal impact of these technologies can be found.
In the mid-1990s the electric utility industry was faced with major changes in how they would operate. Restructuring would cause utilities to buy and sell most of their power through the wholesale market, and many utilities would no longer receive their expected return on investment. Instead, prices would be based on the market and not on cost of service. The transition could mean stranded costs on their expensive plants or long-term contracts. To evaluate the impacts, we developed a model called ORFIN (Oak Ridge Financial) model. (Hadley 1996) It calculated a utility's costs and revenues over a multi-year time period and allowed a financial comparison between a regulated market and market-based pricing. Among its most notable analyses was an examination of the stranded costs for each utility in North Carolina.
While ORFIN could examine a single utility over multiple years, it only roughly modeled the production and sales in a regional wholesale market. The ORCED model was developed to meet that need. Its first big test was analysis of the impact of different technologies and carbon reduction strategies on the nation (Inter-Laboratory Working Group 1996). Since that time, the model has been used in a variety of studies including:
- Impact of restructuring on power prices in California and the Pacific Northwest (Hadley and Hirst 1998)
- Effect of carbon taxes on power production in Ohio and the ECAR region (Hadley 1998)
- Market incentives for adequate generation capacity in a restructured electricity market (Hirst and Hadley 1999)
- Effect of NOX emission control implementation plans on system reliability
- Impacts of hydropower relicensing on carbon emissions in each NERC region (Sale and Hadley 2002)
- Impacts of restructuring on prices and transmission in Oklahoma (Hadley et al 2001a) (Hadley et al 2001b)
- Benefits of multiple emission controls strategies
- Benefits of distributed generation to utilities, customers, and society (Hadley and Van Dyke 2003) (Hadley, Van Dyke, Poore and Stovall 2003) (Hadley, Van Dyke, and Stovall 2003)
- Potential for economic biomass cofiring on a state and regional basis (English 2004)
- Air pollutant concentration changes across the Southeast due to demand reductions (O'Neal, Imhoff, Condrey, and Hadley 2006)
- Impact of plug-in hybrid electric vehicles on electric generation in individual regions across the country (Hadley 2006)(Hadley and Tsvetkova 2008)
The original ORCED model was a single Excel spreadsheet that dispatched 25 power plants against two seasons using simple 3-segment demand curves. The current version pulls power plant data from a database of over 25,000 units plus other data sources, segregates plants by region and aggregates them into 200 plant groups, converts hourly load data from over 100 utilities into three seasons with 11-segment regional demand curves, and calculates market-based and cost-based prices, air emissions, and full financial statements for each power plant.
The overall flow of information is shown in Figure 1. On the left, the demand information is gathered and converted. On the right, the supply information is also gathered and converted for use in the dispatch section. At the bottom, the demand and supply are brought together so that supply is dispatched to meet the demands. Lastly, the results for the scenario are stored for comparison to other scenarios.
Raw data is gathered from independent sources, such as the Federal Energy Regulatory Commission (FERC), Energy Information Administration (EIA), Environmental Protection Agency (EPA), North American Electric Reliability Council (NERC), state utility commissions, independent system operators, non-government organizations, and utilities themselves. Sufficient data to operate the model can be found from open sources, although some studies have used purchased, proprietary information on power plant statistics.
The data is typically collected into spreadsheets for further manipulation. While the model could be developed in another computer language or architecture, spreadsheets offer the flexibility that lends itself well to the varied tasks used for the model. Some of the processes involved, including hydropower capacity allocation and probabilistic dispatch, involve complex calculations that use Visual Basic routines that have also been translated to FORTRAN for testing. Other techniques used, such as histogram calculations and Solver optimization for the load duration curves, rely on built-in functions of Microsoft Excel. The formulas within the spreadsheets can be quite intricate, which makes documentation and error-checking more difficult than with other languages.
A set of spreadsheets is used sequentially for the various steps in the process. Data can be either linked between spreadsheets to ensure consistency or manually copied to ease calculation time. Various macros are used to ease the calculations and connections between steps.
Figure 1. ORCED Flow diagram
A discussion of ORCED's operation is provided in the ORNL report, The Oak Ridge Competitive Electricity Dispatch (ORCED) Model. (A paper describing the model from 1998 can be found at the ORNL report entitled ORCED: A Model to Simulate the Operations and Costs of Bulk-Power Markets.)
- UnitsV4b.xls - A listing of the power plant units for each of 13 regions of the country in 2020 and the aggregated plant groups, fuel costs, and emission credit prices used in Dispatch
- SupplyV4b.xls - The spreadsheet used to aggregate a region's plants into the 200 used in Dispatch
- DemandV4b.xls - The spreadsheet that adds the plug-in hybrid charging to the hourly loads for a region and the calculations to convert the hourly loads to a load duration curve
- DispatchV4e.xls - The spreadsheet that matches the region's supplies to demands to determine the financial and environmental results ResultsV4b.xls - The spreadsheet the collects the summary data from multiple scenarios for comparison.
If a user wishes to just examine the workings of Dispatch, the Supply and Demand inputs for all regions and seven different plug-in hybrid scenarios (including no hybrids) are already stored in the spreadsheet. Users can select them by changing the values in cells Input!H1:H2.
The webpage also has two earlier versions of ORCED: one used for the Oklahoma Restructuring study, and a fifty-unit version from 1998. Also on this page is a link to a separate page with the 1998 versions of the model for each of the regions of the country.
For more information contact Stanton Hadley.