Meritorious Applications (continued)

Interactive Real Time Electric Power Allocation

The electric power industry is rapidly getting immersed in the information age as a result of congressional legislation and a notice of proposed rulemaking issued by the Federal Energy Regulatory Commission, all aimed at promoting wholesale competition through non-discriminatory open transmission access, i.e., deregulation. Much of the impact of deregulation takes on a regional or even international dimension as utilities seek to operate large scale power grids in the most efficient engineering and economic mode possible. Many of the leading electric utilities are already involved in transactions based on real time information transmitted through an electronic information network over the Internet. The network, called the transmission services information network, or TSIN, allows access by all authorized users anywhere in the world and provides current information on the available transmission capacity (ATC), cost and tariffs for transmission capacity, and cost of ancillary services, such as spinning reserves and reactive support. The ATC is calculated as the difference between the total transfer capability (TTC) and the firm commitments along the transmission corridor. Although the firm commitments can be easily found, the TTC is a quantity that can not be calculated easily. The latter is defined as being equivalent to the first-contingency total transfer capability and is usually limited by thermal or stability limits on the transmission interface. The term "first contingency" refers to outage of a single element, such as a generator, a transmission line, a transformer, and so forth which are integral parts of a power grid. Since the power system is dynamic, the system will continuously settle into new equilibria based on various post-outage conditions. On most occasions, this post-disturbance equilibrium point will be stable, meaning that the system frequency will be retained at or close to nominal. However, occasionally, the system can reach an unstable equilibrium point because of inadequate transfer capability or inadequate var support. Two most recent examples of system-wide blackout were the July and August 1996 disturbances in the western US grid. Fig. 1 shows the layout of the interconnected system in the Western United States and Canada comprised of more than 20 different electric utility companies in 10 states and two Canadian provinces.

Figure 1. A map of the 500 kV main transmission grid in Western US

Identifying these severe and harmful contingencies, by exhaustively going through a list of contingencies, even on an off-line basis, is very time-consuming and computationally intensive. Each contingency analysis (out of a list of thousands) emphasizing system dynamics that require solving non-linear differential equations may take CPU times in excess of a minute on UW’s fastest computers. Naturally, performing full-scale contingency analysis, on-line at even close to real time, for power grids containing more than 5,000 nodes and 10,000 branches is still not possible without resorting to supercomputers. Therefore, for a massively interconnected network such as the Western Systems Coordinating Council (WSCC) grid of Fig. 1, performing accurate, on-line ATC calculations and posting those on the Internet would not be possible without high-speed connections to supercomputing centers, coupled with low-latency feedback across the network. Such a capability will maximize the economics of power transfers between two or more control areas of a power system. An illustration of a lack of fast, secure communication among the control area entities is in the August 1996 outage of the western grid. The initiating fault event was merely a tree touching a high voltage transmission line in Wyoming that caused a short circuit in that line. However, this event cascaded into a series of outages because of excessive power demand in load areas such as California and Idaho. Long before a "point of no return" was reached, two major power lines in the Pacific Northwest had been lost. Yet, system operators in California were unaware of the growing problem and kept on importing power from northern western hydro plants without increasing backup generation in local areas. Therefore, ways to provide real time data sharing among control centers automatically is a necessity in the face of this ever-growing complexity in interconnected power grid operation.

ATC calculation is only one example of a need for high-speed Internet connection in the area of power systems operation. Many progressive electrical utilities are now developing comprehensive data acquisition facilities. The emerging critical path challenge is to extract essential information from this data and to distribute the pertinent information where and when it is needed over very wide spread geographical locations. The uses of a high performance Internet connection could range from the basic tasks of remote data storage and shared retrieval to the remote near real time applications of power system monitoring and control. Information sharing between utilities accomplished in near real time on a high speed Internet connection could have a tremendous impact on the efficient and reliable use of the system. Again, these issues were brought into sharp focus by the power outages experienced by the Western US grid in the summer of 1997. Post analysis of monitoring records from these disturbances [2] provided many indications of potential problems. If this analysis could have been performed in near real time and distributed to the appropriate user, the system's vulnerability could have been realized prior to the power outage.

Currently the University of Wyoming (Department of Electrical Engineering) is working with Bonneville Power Administration (BPA), Pacific Northwest National Laboratory (PNNL), and the Electric Power Research Institute (EPRI) on developing methods for monitoring the stability of an interconnected power system in near real time [3]. For example, the Western United States and parts of Canada have a large interconnected power system with data acquisition systems spread at several key locations throughout the system [4]. The amount of data and the great distances involved between monitoring stations will require a high performance network to manage and share all the data. For instance, the Dittmer Control Center, Vancouver, Washington collects over 96 channels of data. With fast Internet access this data could be shared among users throughout the system. The DOE-funded Wide Area Measurement System (WAMS) program foresees use of a high performance connection to the Internet in this manner. Possible future uses of this monitoring network involves the area of automated control technology. Development of prototype models, such as ADAPT at UW, as well as software, and technologies which address the problems described above require additional bandwidth and low-latency connections for "real world" testing of systems being perfected at the University of Wyoming.

previous  |  next