Project Background
Northern Virginia Electric Cooperative (NOVEC) is headquartered in Manassas, Virginia and is an electricity reseller serving residential, commercial, and industrial customers. NOVEC owns an electricity distribution network but does not currently own any electric generation. Consequently, NOVEC must purchase power from the wholesale power market which must be delivered through the PJM Interconnection (PJM). PJM is a regional transmission organization (RTO) that coordinates the movement of wholesale electricity in all or parts of 13 states, including Virginia and the District of Columbia.

Generating capacity must be sized and acquired for the worst case demand which, for NOVEC, occurs during the summer period that spans June 1 through September 30 of each year. It benefits NOVEC customers when NOVEC is able to control its customer peak hour usage and thus minimize its annual summer peak capacity requirements and energy purchases. An advance purchase ahead of the summer is necessary because NOVEC must be sure that its meets its PJM assigned capacity obligation and has sufficient capacity to serve its customersí energy needs during summer peak hours.

Energy is purchased using two methods that reflect both the average energy consumption (base load) and the deviations from the average (intermediate and peak loads). Base load consumption is satisfied by energy purchased using negotiated, firm price, bilateral contracts. These contracts are purchased one month to three years in advance of consumption and make up a majority of NOVECís energy purchases and sales.

Peak load consumption is satisfied by spot market energy purchases made either on the same day, or up to one day before, the energy is consumed. The quantity of these purchases is determined by the number of kilowatt-hours used during peak electricity-use times. Spot market prices are volatile and high, particularly during periods of high energy consumption, due to the competition that occurs between utilities seeking purchases on the PJM marketplace.

NOVEC has instituted the voluntary Load Management Program (LMP) for its residential customers as a way to reduce peak power demand and thus reduce the purchase of expensive spot market purchases of the energy to satisfy the peak power demand. Customers in this program allow NOVEC to install remote controlled switches on their hot water heaters and/or air conditioning compressor units. Customers then agree to allow NOVEC to switch off these appliances for a fraction of an hour during peak energy use times. Participation in the program gives NOVEC the opportunity to better control the residential demand for peak power and potentially reduce NOVECís energy purchase costs by a few million dollars each year.

Problem Statement
Currently, the Load Management Program has a Load Management System (LMS) comprising approximately 43,000 remote control switches installed on residential hot water and/or air conditioning units within NOVECís service region. The 43,000 switches are controlled in blocks of approximately 5,000 units each such that all the switches within a block are commanded to switch on or off in unison at any time.

NOVEC would like to determine how the operation of LMS can be optimized to maximize the benefit of reducing peak electricity demand. In particular, NOVEC has interest in the daily on/off scheduling of the load management switches as well as discovering if the current switch block size of 5,000 units is optimal. A successful optimization approach is one that simultaneously reduces peak electricity demand, reduces peak energy consumption but minimizes reductions in overall energy sales, and maintains customer satisfaction in the provided electricity service.

NOVEC has indicated the operation of the LMS is guided by an associated load management policy. As such, NOVEC seeks a recommended load management policy or set of policies that:
• Accommodate multiple peak load scenarios driven by seasonal and time of day influences
• Provide potential cost savings for NOVEC customers.

Project Objective
The scope of this project is the development and demonstration of an algorithm and associated load management policies that more consistently reduce peak energy demand while maintaining customer satisfaction with NOVEC's service. Demonstration of the effectiveness of the algorithm and associated policy will be accomplished by the creation of a Load Management Director (LMD) prototype computer model and the assessment of this prototype using historical data sets of power demand and weather data provided by NOVEC. The project will provide a technical report that describes the research approach, the experiment design, obtained results, and conclusions. Furthermore, the developed simulations and load management policies will be made available to NOVEC.