SPACE WEATHER AND THE INCIDENCE OF REACTIVE “OFF-COST” OPERATIONS IN THE PJM POWER GRID 7.6 Kevin F. Forbes * The Catholic University of America O. C. St. Cyr The Catholic University of America and NASA-Goddard Space Flight Center * Corresponding author address: Kevin F. Forbes, The Catholic University of America, Department of Business and Economics, Washington D.C. 20064. Email: Forbes@cua.edu 1. Introduction Forbes and St. Cyr (2004) have provided empirical evidence that adverse space weather conditions have affected the price of electricity in the PJM power grid. In this paper we examine one of the mechanisms by which space weather impacts the electricity market in the PJM power grid. The starting point of this paper is that there are two types of power on alternating current systems: real power and reactive power (Sauer, 2003). Real power is the power that consumers need to light their lamps and run their computers and refrigerators. In contrast, reactive power maintains the voltages required for system stability and thus is critical to the delivery of real power to consumers. With respect to space weather, it is well-known that there is an increase in reactive power consumption when GICs pass through a transformer (Kappenman, 2003, p 4). Generators in PJM are normally dispatched based on their cost with the lowest cost generators being dispatched first. One major exception to this “economic merit’ or “on-cost” method of dispatch is when reactive power conditions warrant an “out of economic merit” order dispatch. In PJM, this is known as a reactive “off-cost” operation (PJM, 2005). During a reactive off-cost operation, generators are redispatched so to reduce power flows across transmission lines vulnerable to voltage collapse. In this paper, we examine the effect of GICs on the incidence of these reactive “off-cost” operations in PJM using a nonparametric statistical analysis as well as an econometric model. Before proceeding, it is worth noting that there was an abnormally high incidence of reactive off-cost operations in PJM during the “Halloween Storms” of late 2003. Specifically, over the period 29-31 October 2003, reactive off-cost operations were implemented in 18 of the 72 hours, a rate more than twice the average rate of incidence. Moreover, inspection of PJM’s emergency logs and its postings of off-cost events reveal that one four hour long reactive off-cost operation on 29 October 2003 was recognized by PJM as being space weather in origin. Interestingly, Pulkkinen, Viljanen, and Pirjola. (2005) have observed that these same storms led to significant operational problems for the Swedish power grid. 2. The PJM Electricity Grid PJM Interconnection is a regional transmission organization (RTO) that as of 30 April 2004 coordinated the dispatch of 76,000 megawatts (MW) of generating capacity over 20,000 miles of transmission lines in all or parts of Delaware, Maryland, New Jersey, Ohio, Pennsylvania, Virginia, West Virginia and the District of Columbia (PJM, 2004). PJM operates both real-time and day-ahead markets for energy. Prices in these markets are reported hourly. The prices are location based which means the prices will be equal across locations when the transmission system is not congested, but can vary substantially from one location to another when there are transmission constraints. The sample period for this study is 1 April 2002 through 30 April 2004. The starting date of the sample period represents the first day of PJM’s functional control of Allegheny Power’s five state transmission system(PJM, 2002). The ending date of the sample period is one day prior to the integration of Commonwealth Edison’s control area in northern Illinois into PJM (PJM, 2004). PJM has a number of trading hubs whose economic function is to facilitate electricity trading. The two most important hubs over the course of the study period are its Eastern Hub and the Western Hub These two hubs are located in eastern and central Pennsylvania, respectively. The difference in the day- ahead Hub prices reflects expected transmission constraints. The difference in the real-time Hub prices reflects actual transmission constraints. On average, because of transmission congestion largely related to terrestrial supply and demand considerations, the real-time price at the Eastern Hub over the sample period was $1.80 per MWh higher than at the Western Hub. The differential was significantly above average during the “Halloween Storms” of October