SETTLING FOR 80% UPTIME ON A UPS
The lack of lightning and transient protection ahead of a UPS can expose its components to damage, resulting in possible
periodic transfers to the bypass mode.
A HEALTH insurance company was experiencing chronic problems with its UPS and, as a result, invited us to examine the installation
to pinpoint the source of these problems. While in the process of conducting our survey, the facility manager inquired as to how much
uptime he should expect on his UPS. We explained that he should not tolerate much less than 100% uptime, since the product was an
uninterruptible piece of equipment! He replied, "I thought you would say that. We get 80%!"
After proceeding with the rest of our survey, we installed a power disturbance analyzer at the input to the UPS and its bypass line.
As we were performing our monitoring, an electrical storm was raging outside. As the facility was on an elevated plane, it was ideally
suited to be the first strike area for storms in the vicinity. As we finished connecting the leads to the analyzer, the UPS transferred
to bypass with a loud bang! An examination of the recorded levels on the analyzer indicated that the facility had received an impulse of
2300V line-to-line, and 1340V line-to-ground. Obviously, some UPS component had failed at the time of the recorded impulses, causing the
UPS to transfer to its bypass line so that power would be maintained to the computer load.
The maintenance person at the site arrived shortly after the transfer alarm sounded and began to systematically troubleshoot the
UPS components. He restored the UPS to working order in 45 minutes, while we watched. When we commented on how thoroughly and professionally
he had cared for the system, he exclaimed, "It's no big deal; I've done it 27 times before!" Astonished, we asked to review the
maintenance documentation, which consisted of over four years of reports on restoring the UPS to operating status under the same
circumstances. In each instance, a severe lightning storm preceded the transfer.
At this point we had an idea. Why not compare the maintenance records with the log in the computer room? Sure enough, we found that the
only time the computer room went down in the last 4½ years was within 24 hours of every event listed in the UPS maintenance records. When
we shared this information with the operating personnel in the computer, room, they attested to the fact that the only'' time the computers
had a problem was when the UPS "did those funny things." The reason why "those funny things" happened was because the
UPS was doing its job, which was to transfer to bypass when one of its components had been damaged, causing the UPS to no longer be able to
produce output power.
The flaw in this installation was that the bypass line was the same line that was carrying the high impulse strikes from outside the
facility. There was no lightning or subpanel protection equipment ahead of the UPS. As a result, those lightning-induced impulses, after
damaging the UPS components and causing transfer to bypass, were going right around the UPS and straight to the computer room load. What
started out as a concern that this UPS was not operating properly (80% uptime), turned out to be a case of no transient protection.
The first recommendation was to install proper lightning and transient protection ahead of the UPS for future protection. In this
way, the owner would not only protect his UPS investment, but also the bypass line to his computer equipment. The second was to have the
existing UPS printed circuit boards rebuilt by the UPS manufacturer, since they had been exposed to transients.
Additional information and site survey information can be obtained by contacting
POC: Mike Helms mike@lightningmike.com
|