A friend in Northern Virginia reports that there were two particularly bad things about the outage that left him and some 750,000 other people without power last week. “The temperature hovered around 100 degrees with high humidity all six days; and for the first couple of days, all of the nearby stores, bars/restaurants, gas stations, etc. were powerless and therefore closed.“
What and how did he eat? For the first couple of days, he “carefully grazed off of the food in the fridge, hoping that the undisturbed freezer would retain its low temperature until the power came back on. But by the end of the weekend, we declared the food a total loss.” His wife having had the good fortune to be out of town, he sought refuge in his D.C. office much of the time.
On a larger scale, one of Amazon’s major server facilities went down, leaving cloud customers like Netflix, Instagram, Perest, and Heroku without access to their databases.
How can we do better in the future?
Smart grid technologies surely will help. For example, in a big pilot project in the area about Harrisburg, Pa., improved distribution management tools, automated control devices, and high-speed communications will enable managers to identify in a matter of seconds what parts of the grid are in trouble and crews need to be sent.
“At a high level,” writes Steve Gelatko, distribution asset manager for PPL Electric Utilities in Allentown, “we will be tracking the number of interrupted customer minutes we will be able to avoid each year. As we roll out new phases and capabilities of the distribution management system, we will be able to reduce the number of minutes interrupted. When the system is fully operational, we expect the number to be 1.5 million smaller per year in the Harrisburg area.”
Gelatko’s article appears in the July issue of IEEE’s monthly smart grid e-newsletter. In another article in the same issue, authors connected with Georgia Tech and Southwire describe how software they have developed could provide much better real-time readings of temperature and sag on transmission lines. Traditionally temperatures have been estimated by means of ampacity tables that are calculated from expected average weather conditions and values that have been considered valid for a whole season. But in reality, weather conditions vary over much shorter time intervals. The new software will provide more accurate and timely estimates of a cable’s position and thermal state.
However timely and accurate the information coming into control centers is, in the final analysis, management will only be able to address problems to the extent well-trained technicians are available. That is an acute problem today and will remain so for the foreseeable future. During the Mid-Atlantic outages last week, the local utilities recruited crews from as far away as Canada to help out. New York City’s Con Ed technicians were unavailable because of a lockout that had them marching in picket lines rather than working in cherry pickers.