My Inspiration for this column came from a discussion on a Facebook nostalgia group. The question was “Who remembers the Great Northeast Blackout [of 1965]?” I scanned the various posts about where people were and how they coped. A couple of posts stated unequivocally that the cause was never determined, but I knew that was not true. Early in my career in the insurance industry, I learned from an experienced engineer about a federal committee that developed a detailed report with recommendations for the president.

Although I was only 12 at the time, I remember the beginning of the blackout well. My family had gone to the opening of an impressive, brightly illuminated, new store. When we got home, we noticed that, by contrast to where we had just been, the kitchen was very dark and seemed to be getting darker. Soon, the overhead light flickered and went out.

Little did we know that this was not a normal, local power failure. In fact, it began in Ontario, as a large portion of power needs in the northeastern United States were met by Ontario Hydro. Most people were unaware that the U.S. and Canadian grids were interconnected.

According to the official report “Northeast Power Failure, A Report to the President by the Federal Power Commission,” the outage covered approximately 80,000 square miles and affected some 30 million North Americans. Twenty-eight utilities were affected, and all were connected to the Canada-United States Eastern Interconnection.

The event began at 5:16 p.m. on Nov. 9, 1965, with the operation of a protective relay for one of five 230-kilovolt (kV) transmission lines. Oddly, the initiation was not caused by a fault. The relay was set well below the overload point. When that load was exceeded, the first relay operated to trip the breaker. The load from that transmission line was transferred to the remaining four transmission lines, which were overloaded by the additional load. They tripped offline in 2½ seconds. It became a cascading outage as the load was transferred to a decreasing number of generating facilities, causing instability.

The outage spread across New York, Connecticut, Massachusetts, Rhode Island and Vermont. It affected parts of New Hampshire, New Jersey, Pennsylvania and a large portion of Ontario. It took 14 minutes from the first trip in Canada to the last trip at the Long Island Lighting Co. Restoration took from a few minutes to as long as 13 hours. Some islands were created where power was not lost.

Since the 1965 blackout, utilities have improved communications and planning with each other. An entity, now the North American Electric Reliability Corp., was created to administrate the national reliability standards. Several regional power pools were created, and many utilities consolidated into larger ones.

A significant difference is that utilities are no longer the only suppliers of power. Many commercial, industrial and residential customers now supply power to the grid. Photovoltaic panels and wind turbines on the grid supply power through interactive inverters. When grid power is lost, the inverter drops offline. Multimode inverters will drop off the grid and island to supply power to the premises.

Is the grid reliable enough?

Large-scale blackouts are extremely rare. One occurred on Aug. 14, 2003, affecting some 55 million people in the Northeast, Midwest and Ontario. It began when trees contacted some 345-kV transmission lines in northeastern Ohio, causing blackout areas and islanding some utilities. This outage resulted in standards to address system restoration, especially as it relates to coordination among utilities.

While a major fault can be felt elsewhere on the grid and disrupt power over a wide area, most outages happen because of weather-related damage to the distribution system and motor vehicle accident damage. Major disruptions and weather-related outages have led to the development of the smart grid. The smart grid relies on enhanced communications technology to manage generating sources and loads while detecting and isolating outages before they can affect larger areas. It also improves resiliency by providing the means to reroute power around isolated faults.

Who needs a backup?

Do you need a backup power supply? That is an individual decision. What is the risk if you don’t have one? Even with a more resilient grid, there will always be a risk of an outage.

One of my jobs was at an engineering and research firm that worked for the commercial and industrial insurance industry. One day, a car collided with a roadside utility pole, resulting in a power outage. Common practice is for the affected utilities to bill the driver for the replacement of the pole, including the use of vehicles, parts and labor.

However, the outage interrupted power to a pharmaceutical plant, resulting in the loss of $9 million worth of antibiotics. The insurance company tried to recover what they could from the vehicle owner’s insurance company. Interruption of many continuous industrial processes can result in production loss and equipment damage. Replacing equipment takes time, resulting in further losses. Storm-related outages can persist for many days as crews repair the affected parts of the grid.

Having a backup supply is no guarantee that it will work. After a recent major hurricane, there were news stories about generators that didn’t start. Regular starting tests and those under full load are essential to ensure they will operate. If it fails under test conditions, there may be time to correct the situation before the generator is needed for an actual outage.

Fortunately, most generators now have automatic testing capability. Equipment that can be adversely affected by testing should be supplied through an uninterruptible power supply.

The automatic testing feature is only valuable if it is used, of course. Some facility owners don’t leave generators in self-test mode. Monitoring the fuel level is also critical because self-tests use it. Fuel deteriorates over time, so using some and refreshing it with new fuel improves the generator’s reliability, another advantage to self-tests.

I once worked in a communications facility for a medium-sized municipal fire department. It had a diesel generator for backup power that supplied the entire load for the building during outages. Unfortunately, although it was an unwritten rule that we were responsible for testing the generator, no one established a test frequency, nor were records kept. One way to test a generator was with a test switch; the other was to open the service disconnect. I always opened the service disconnect because it simulated a power failure. The telephone system and the fire alarm system were supplied by float-charged batteries, so this brief event would have no impact on other critical activities, other than a momentary loss of radio communications.

Over time, generator tests became infrequent. One night there was a power failure, and the generator didn’t start. Radio communications were lost, but the phones still worked. The police were informed by phone that the communications facility was out of service and to have an engine company respond to the communications facility so the dispatchers could use the vehicle radio to dispatch and communicate with other fire and EMS apparatus. This incident put many people at risk. The generator was repaired, but no longer trusted, so a temporary generator was used until a new generator was installed.

Major utility outages are rare. Smaller outages resulting from weather and other causes should be expected to occur occasionally. Smart grid improvements will improve the reliability of the grid; however, a risk assessment may still indicate a need for a reliable standby power source.