The Role of Demand Side Management
- Written by Pedram Samadi, Hamed Mohsenian-Rad, Vincent W.S. Wong, and Robert Schober
Whether in the form of direct load control or real-time pricing, demand-side management should be an essential ingredient of the smart grid. To induce residential consumers to participate, energy companies may need to offer billing discounts and be sensitive to how electricity consumption preferences vary with time of day.
To achieve a high level of reliability and robustness in power systems, the grid is designed for peak demand rather than the average load. This can result in under-utilized power generation and distribution systems and a waste of natural resources. Moreover, most of the fast-responding generators that are used to meet the peak load, such as gas and coal units, are expensive and have large carbon footprints.
To overcome these problems, different programs have been proposed to shape users' energy consumption profiles. Such programs allow the available generation capacity to be employed more efficiently so that new generation and transmission infrastructure do not have to be installed. These programs, generally known as demand side management (DSM), aim either at reducing consumption or shifting consumption.
One option in DSM is direct load control (DLC), where, based on an agreement between the utility and the customers, the utility remotely controls the operation of certain appliances in a household. Smart pricing is an alternative where elaborately designed pricing rules are adopted to encourage users to individually and voluntarily manage their loads in order to reduce their own energy cost.
The successfulness of DSM programs mainly depends on how big a portion of the total load is controllable. The emergence of new types of equipment on the demand side, such as plug-in hybrid electric vehicles, is expected to increase the percentage of controllable load over the next few years. They will make large demand on the grid, especially as they will be charged at different times of the day and at different rates.
In addition, the electricity storage capacity of electric vehicles will provide new opportunities to economically store the electricity in the batteries of parked plug-in hybrids so that the stored electricity can be made available at peak hours. Thus, hybrids may make DSM an even more powerful tool for balancing power supply and demand.
Several pricing schemes have already been proposed for DSM, such as peak load pricing and adaptive pricing. The two-way digital communications capabilities of future smart grid systems are expected to enable DSM programs to even use real-time pricing to more efficiently pass on the fluctuations of wholesale prices to retail users.
In real-time pricing schemes, the operation cycle is divided into several periods. The exact price for each period is selected in real-time, and random events and the reactions of users to the previous prices influence the price set in upcoming operation periods. That kind of system may also help the grid to integrate renewable energy sources. In fact, utilities can tackle the intermittency and the inherently stochastic nature of renewable energy sources by changing the price.
Most of the current load control decisions in existing DSM systems are made manually, which makes it difficult for the participants to monitor the real-time prices and to use other advanced pricing methods. In fact, lack of knowledge among end users about how to respond to time-varying prices is currently a main barrier for fully utilizing the benefits of real-time pricing methods and DSM in general. This problem can be resolved by equipping users with home automation systems and by implementing automated energy consumption scheduling units that can draw on pricing information to schedule the operation of various residential appliances on behalf of customers.
Smart pricing should provide appropriate incentives for individual users to participate in DSM programs, for example by discounting their electricity bills. In general, users may have different energy consumption preferences at different times of day. To quantify the level of satisfaction of each user as a function of load at different times of a day, the concept of the utility function has recently been adopted from microeconomics. In this case, "utility" refers to usefulness and satisfaction of goals not energy organizations in the power industry.
Utility theory can particularly be used to achieve social fairness among participating users in DSM programs. Other applicable economic tools include game theory and mechanism design, which can help to analyze users’ behavior and to align the load objective of users with the social goals of energy companies and the power grid. Finding appropriate utility functions—that is to say, mathematical functions—that accurately capture the energy consumption behavior of end users remains an open problem in this line of research.
While automated DSM systems can significantly enhance efficiency and reliability, they may also create new vulnerabilities in power infrastructures if not accompanied by appropriate security measures. For example, recently identified load altering attacks against smart grids may target controllable load by compromising load commands and price signals. This can be done by simultaneously switching on several remotely controllable appliances of an adversary to cause a major spike in the aggregate load demand. Therefore, DSM programs for the future smart grid not only have to be efficient and fair, but also secure against potential cyber attack.