A recent windstorm which left much of western Washington groping around in the dark without electrical power also taught a lot of people an important lesson about back-up power supplies.
In Washington, a substantial number of the batteries installed as part of emergency power systems didn't operate, potentially leaving inoperable substations, telecommunications switching equipment, cellular and microwave networks, backup power generators, and a wide variety of microprocessor based equipment thought to be protected by uninterruptible power sources (UPS).
Those responsible for equipment, building or office operations thought they were protected where necessary. What happened?
The answer is that potential for major problems has increased over the last ten years. Valve regulated lead acid technology (VRLA) uses see-through containers allowing for a number of different and easy tests.
New equipment has been sold as "bullet proof" and "maintenance free" by manufacturers eager for a major share of the burgeoning of a highly competitive backup power market.
This has lead to a general complacency regarding testing and maintenance of a wide variety of critical emergency power supplies. That complacency is enhanced by manufacturers equipment instructions which leave testing and maintenance issues with a brief mention at the end of manuals.
All this makes it difficult to get testing funds budgeted and lulls those responsible into a false sense of security. Many people tend to equate all batteries with car batteries. In the case of standby batteries, this is a critical mistake. Standby batteries are on charge 24 hours a day, year in and year out and are only used when a power outage occurs. An automobile battery generally is used daily, providing quick regular feedback as to its health and is only on charge when the vehicle is being driven.
Seventy-five percent of the system failures are due to batteries not operating correctly, problems which could be avoided with periodic testing.
VRLA's are responsible for supplying varying amounts of direct current. Their cells have a valve which acts as a safety valve. Explosive gases produced by discharging and recharging are contained within the cell by the valve under normal conditions.
Abnormal usage can cause the valve to open and release the excessive internal gas pressure. If the valve doesn't reseat properly or continues to vent gases, the cell dries out, destroying it's storage capability. Too much heat (generally over 90 degrees) or overcharging causes the battery to develop a high interior pressure, opening the valve.
Many of today's most popular cheaper VRLA storage batteries have a "design life" of five years, but a real world life expectancy of two to three years. In the rush to be competitive, components have been made less heavy duty and battery use under less than ideal conditions can shorten life further.
With the old flooded battery design, we can measure specific gravity, open cell voltage, float voltage and temperature as well as doing discharge testing.
Today, depending on the application, up to 90 percent of the storage systems use VRLA technology because it is smaller and safer and can be used in a greater variety of surroundings. However, their maintenance requirements are different from the flooded technology systems.
Since we can neither see nor get inside, testing options are limited to float and open cell voltage measurement, discharge testing and to a new conductance/impedance measurement test. Conductance impedance testing has only been available over the last three years and requires special equipment.
If the internal impedance of the battery is too high, it can't deliver the required current during a discharge.
To be sure, testing standby power generally requires knowledgeable technicians and costs need to be compared to direct and indirect losses in case they fail to deliver.
Generally, however, it's safe to say that if the system was important enough to back up in the first place, the potential loss is too great to leave standby power equipment maintenance to chance.
As they say in the tire business, you need to assess what is resting on those tires, the direct and indirect costs associated with failure.
Furthermore, battery warranties generally are prorated. We recently were asked to test a five-year-old standby power installation. It was defective and the warranty provided only five percent of the replacement cost. It's more cost effective to find problems early.
Since so-called five year batteries generally begin to fail after two or three years, a maintenance program probably should kick in during the second year.
Some sophisticated battery plants can have hundreds of batteries. In these cases, you may want to install some monitoring equipment and train your staff in regular testing procedures.
We have developed some simple tests which can be performed on many of the more commonly available standby batteries. In addition to that, however, it'll still make sense to contract for regular maintenance.
In many cases, the results of backup power failure can be embarrassing, as in the case of telephone switching or utility substation equipment.
In some cases, however, the failure of a battery to provide the electrical start-up needed to fire a generator can be disastrous, as in the case of a hospital.
Steve Gomes is a power technician with Portland General Energy Systems, a non-regulated division of Portland General Electric, in Portland, Ore.