Industries Increasing Efficiency through Improved Software
Commission staff convened a technical conference on June 25-27, 2012, to discuss opportunities for increasing real-time and day-ahead market efficiency through improved software. The technical conference was held in the Commission Meeting Room and Meeting Room 3M-3 at the Federal Energy Regulatory Commission, 888 First Street, NE, Washington, DC 20426.
This conference brought together diverse experts from ISOs/RTOs, decentralized market utilities, the software industry, government, research centers and academia for the purposes of stimulating discussion and sharing of information about the technical aspects of these issues and identifying fruitful avenues for research. This conference was intended to build on the discussions initiated in the Commissionís June 2010 and June 2011 staff technical conferences on increasing market and planning efficiency through improved software.
The conferences held in June 2010 and June 2011 produced presentations on several advanced approaches to market modeling which appear to have significant promise for potential efficiency improvements: stochastic modeling; optimal transmission switching; AC optimal power flow modeling; and use of active and dynamic transmission ratings. Significant computational impediments to efficiently and reliably implement these approaches must be understood and overcome before benefits can be realized. In this conference, we explored research and technical steps that would be needed to implement these and other advanced technologies in the future.
In particular we solicited proposals for presentations on topics and questions such as the following:
- Stochastic modeling for unit commitment and operating reserves:
Given the difficulty in formulating and solving full-scale stochastic unit- commitment problems, what interim steps might be taken to more intelligently incorporate information about uncertainty into unit- commitment and dispatch? Specifically:
- How can uncertainty be described in a manageable set of scenarios or constraints that improve unit-commitment and dispatch while allowing good solutions to be achieved in the required timeframe?
- If a stochastic unit-commitment model is used, how should day-ahead prices be calculated, given that the stochastic formulation no longer produces as part of its solution a single set of deterministic shadow prices for power at each location?
- How would a stochastic day-ahead unit commitment mechanism alter current market software for other processes (for example, reliability unit-commitment processes)?
- What steps toward better incorporation of uncertainty into unit- commitment might be taken over the next 5 to 10 years?
What methods can be used to calculate requirements for
contingency reserves and regulating reserves?
- How can reserves calculations more completely capture the uncertainty and variability of the system, including forecast error?
- How can outage probability be captured in contingency reserve calculations, and how good is the available data?
- What methods can be used to determine reserve zones?
Optimal transmission switching:
- Simple optimal DC transmission switching appears to represent a potentially solvable technical problem using existing computational resources if transmission operators optimize only a small number of transmission switch positions. It is less clear whether transmission switching model formulations that include realistic representations of reliability requirements are solvable. What is the performance of these more complex model formulations?
- What additional computational impediments, if any, exist to implementing optimal transmission switching over a small number of switches while maintaining reliability?
- What steps toward optimal transmission switching might be taken over the next 5 to 10 years?
- AC optimal power flow modeling:
- What is the current state of computational capability with respect to dependably solving AC optimal power flow problems, including analysis of power system reliability?
- Discussions during previous conferences have centered on concerns that current system data quality might not allow for an AC optimal power flow model to be properly formulated and solved. What are the specific data concerns, and what needs to be done to address them?
- What steps toward use of AC optimal power flow modeling might be taken over the next 5 to 10 years?
- Adaptive and dynamic transmission ratings:
- Previous presentations examined the use of post-contingency analysis when determining transmission ratings, including consideration of availability of ramping capability. How can (or have) adaptive transmission ratings been implemented?
- Previous presentations also examined how transmission ratings might be updated in real time in response to ambient conditions. How have such dynamic transmission ratings been implemented?
- What are the data or computational challenges associated with implementing adaptive or dynamic transmission ratings?
While the topics suggested above were largely forward-looking, we also encouraged proposals for presentations on best practices and other analyses of current operations with respect to these and related topics.