Smart grid simulations – SPGTC 2011
May 12, 2011, smart power grid technology conference, Santa Clara, CA—Amit Narayan from Stanford University is looking to develop tools to model and simulate the increasing system-level complexity of the electrical grid. His current projects are focusing on interactions of the energy systems and economics.
The electrical grid is adding new technologies in areas like recoverable or distributed generation, demand elasticity, and storage while the market is looking at wholesale and retail market deregulation and retail competition. The pace of change is causing design issues driving the need for tools for analysis.
Like SPICE (simulation program integrated circuit emphasis), the program simulates multivariate analyses through the use of detailed models to investigate issues like the coupling between generation and distribution grid. The growing volume of data related to use, generation, and other functional parameters of the grid, coupled with the size of investments needed to upgrade various portions of the grid, calls for much more intensive modeling for design and optimization.
The current tool works on steady-state analyses and future work will include dynamic characteristics. The objectives for this work are to develop a framework to evaluate indicators, operational and business optimization, maximizing return on investment, all while increasing reliability and efficiency. This overlap of physical and financial characteristics is necessary for the industry to move forward.
Software simulation of complex systems has proven its worth in many other areas, most notably in integrated circuits. Now utilities and other participants in the grid can use agent-based modeling to analyze the market processes, data flows, and power delivery issues. The major inputs are from market and distribution network data as well as characteristics of the transmission network, and the outputs are spot prices and indicators for bringing generators on or off-line.
They are also working on modeling dynamic pricing issues and incorporating machine learning for end-user modeling. The simulations are hierarchical and include economics and data mining capabilities. Demand response ties into customer billing and distribution automation integrates voltage variability options. Conservative voltage reduction and voltage variation are simulated integrated to identify hotspots or areas were there are changes in voltage across the loads. One interesting capability of the tool is to detect instances of collusion in the market.
