Introduction to Smart Grid Concepts

Nate Bogdanowicz
November 16, 2011

Submitted as coursework for PH240, Stanford University, Fall 2011

Fig. 1: Characteristic shape of a typical demand curve, showing base, intermediate, and peak-level power demands.

It is estimated that the total US energy consumption in 2010 was 28.7 PWh (petawatt-hours), with electricity generation accounting for 11.8 PWh or about 41% of this total. [1] From this 11.8 PWh input, a net 4.1 PWh of electricity was generated, a conversion efficiency of 35%. Of this net electricity produced, 0.3 PWh or about 7% was lost during transmission and distribution (between generation by the plant and consumption by the customer). While transmission and distribution can be improved, these numbers seem to indicate that perhaps even greater gains can be had by increasing energy generation efficiency.

The so-called "smart grid" is a collection of approaches using modern (i.e. "smart") technologies to improve efficiency in the power grid. These all generally center around the idea of introducing new sensors, controls and high-speed communications to dynamically adjust parameters of the grid on time scales not currently possible. Some of these approaches leverage a greater use of more efficient generation sources to improve on the current average mark of 35%.

One primary goal of the smart grid is to help level the electrical load via "dynamic demand management". Much of difficulty and cost of managing the grid stems from the variability of demand. The typical shape of such a demand curve is shown in Figure 1. For the most part, electrical energy from the grid cannot be effectively stored, meaning that production must follow demand. To account for this variable demand, utility companies generally use base-load power plants for constant demand and supplement them with peak load power plants. [2,3] The base-load plants are typically coal-burning, nuclear, or hydroelectric. Base-load plants are expensive to construct but inexpensive to operate, meaning they are most cost-effective if running constantly and at full-capacity. In addition, they generally have long start-up times and are not designed to be turned on and off frequently, so they are not good for dynamically adjusting to peak demand.

So-called peak-load plants, on the other hand, are able to be turned on quickly during times of peak demand. Why not use peak-load plants to satisfy the entire demand, if they respond so well to variable loads? Peak-load plants are inexpensive to build, but when compared to base load plants, cost more to operate, and thus are most economically useful when used to supplement a more inexpensive base-load supply. Gas turbines are commonly used for such purposes. There also exist intermediate plants that have characteristics in between those of base- and peak-load plants.

From this, we can see that it is clearly more efficient to have a more even demand curve so that inexpensive energy from base-load plants can be utilized more. Thus, it's desirable to try to flatten this demand curve somehow, which is called "load leveling". To do this, utility companies ultimately must provide incentives for their customers to reduce usage during periods of peak demand. To provide such an incentive, utilities have begun rolling out "smart" meters that are able to bill consumers using a variable rate that changes throughout the day rather than billing all consumption at one flat rate. [4] High prices during peak hours encourage consumers to shift energy consumption to off-peak hours, thereby flattening the demand curve, allowing for less expensive overall energy production and (ideally) lower total prices.

In addition to allowing customers to manually reduce usage during peak hours, a smart grid allows for automatic control of large appliances such as water heaters and clothes dryers in order to reduce cost and improve grid stability. [5] For instance, upon detecting a lowering of the grid's frequency (which is an indication that grid production is overmatched by consumption), the appliance could shed its power load to preserve grid stability. In a trial run by Pacific Northwest National Laboratory, clothes dryers were programmed to turn off their heating elements under such conditions. Despite this, the disturbances were short enough in duration that the participants never noticed the difference, indicating the promise of this technique. Conceivably, customers who opt to use such control devices could be rewarded by the utility company with a credit or discount of some sort.

In short, a great deal of stress is placed upon the electrical power generation sector to provide for a highly variable demand. Some aspects of an overall "smart" grid approach, including variable metering and load-control switches, have potential to alleviate some of this stress, allowing for more efficient overall electricity production and consumption.

© Nathan Z. Bogdanowicz. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.


[1] "Annual Energy Review 2010," U.S. Energy Information Administration, DOE/EIA-0384(2010), October 2011, pp. 233

[2] R. J. Daniels, Ontario Hydro at the Millenium, (McGill-Queens U. Press, 1996), p. 77.

[3] G. M. Masters, Renewable and Efficient Electric Power Systems, (Wiley-IEEE Press, 2004), pp. 135-137

[4] H. Wallop, "Every House to Have Smart Meter in £bn Scheme," The Telegraph, 11 May 09.

[5] D. J. Hammerstrom et al., "Pacific Northwest GridWise Testbed Demonstration Projects - Part II. Grid Friendly Appliance Project," Pacific Northwest National Laboratory, PNNL-17079, October 2007.