By Guy Waterman, PointServe
The electric utility industry is in the midst of an historic restructuring process, which is intended to create greater competition in the marketplace. Regional Transmission Organizations (RTOs) are entities that manage or operate the transmission of facilities of all or part of an electric supply system for geographic regions. Though typically a for-profit entity, RTOs operate on a revenue-neutral basis. The current trend for consolidating operations is driving RTOs to expand control areas and markets. RTOs are interconnected with surrounding/adjacent RTOs and ISOs. The power flows into, out of, and through the control areas, and each region monitors and meters the net interchange. Most RTOs have retail markets that operate (very well) in their regions and this model has become the standard for how the Federal Energy Regulatory Commission (FERC) wants the new RTOs to operate.
There are several needs facing RTOs at the present time. Currently, energy markets are extremely volatile at times because of limited or inefficient tools and because of market drivers beyond their control. The general business objective is to develop software tools to assist RTOs in managing their control area with their associated markets. These tools will forecast volatility, provide insight into market drivers, and measurably improve operational efficiencies. Accurate and reliable tools are needed to solve the following problems:
- Short-Term Forecasting (STF):One-to-seven day forecasts of energy load demand. Accuracy needs to be on the level of 500-1000 MW, or about 1-3 percent of the average daily load.
- Very-Short-Term Forecasting (VSTF):Very short-term adjustments to the load forecast for prediction ahead from one minute to four hours. Accuracy needs to be on the level of 50-100 MW on average, and no more than 300 MW for 15 minutes out. A precise tracking filter of real-time measurements and the one-day forecast model (from STF above) is required to achieve this objective.
- Net Interchange Modeling:Modeling and forecasting the net interchange schedule of energy from the various ISOs that are connected together is a major business objective. Systems calculate the economic dispatch for next four hours by using the net interchange schedule. But the net interchange schedule can be updated by the market players. The reason for update may depend on location marginal price (LMP), external prices, outages and constraints, forecasted demand, and/or dispatch control signal. RTOs need tools to help manage and forecast of net interchange schedule for next four hours every 15 minutes considering any possible updates. While the challenges of the problem of forecasting energy load demand are well understood and can be handled by a number of different vendors and approaches, there are no readily available off-the-shelf products that offer net interchange modeling and optimization. The optimization challenge for any model will most likely be non-linear and require advance techniques.
PointServe has a history with using genetic algorithms to solve highly non-linear optimization problems, and it is hoped that this technology can be adapted to the net interchange problem. Before an approach can be decided upon, fact-finding and domain knowledge gathering must be done. The fundamental challenge is essentially a supply-demand time series problem from the adjacent control areas. Data gathering can help explain the short-term behavior of the adjacent areas, and this will feed into model construction and methodology.
PointServe is helping to create a system that has the potential to predict short-term price movements in the commodity markets. For the VSTF, PointServe will provide software tools that achieve the defined performance requirements based on actual load readings and an input STF model. In addition, the software will become a product platform for advanced modeling and optimization of the net interchange. This software platform could be used by RTOs to manage larger areas with fewer resources, to effectively train new operators and dispatchers, and to enable advanced study of the volatility of the grid management problem.
The PointServe Optimization Platform is software that models the service provider’s service supply chain and enables users to drive their service supply chain to greater levels of performance. These models are used by the Optimization Platform to improve operations by decreasing costs and increasing customer satisfaction.
The robust, feature-rich modeling system enhances many aspects of service-supply chain operations by providing service providers the ability to create operational business models based on customer, appointment, and resource data supplied by the provider. These models interpret the state of the data and actively integrate it with specified business processes to arrive at a desired execution plan. During this process users can adjust schedule parameters to meet their business objectives.
Updated/current trend information is fed back into the models to help service providers accurately forecast market demand. Accurately predicting demand enables resource capacity plans to be adjusted to continuously provide better service solutions over time.