"They all suck!" Re: UPS failure modes (was: fire at NAC)
Robert Boyle
robert at tellurian.com
Thu May 29 21:55:26 UTC 2003
At 04:39 PM 5/29/2003 -0400, you wrote:
>A much cheaper and easier to implement external maintenance
>make-before-break bypass will accomplish the same thing.
>
>I've heard many a story of the paralleling gear causing the problem in the
>first place, as well...
We have two MGE 150KVA UPSes at our Newton facility. When I designed the
electrical system, I originally specified a make-before-break 208V 450A 3PH
switch. This would enable us to isolate our internal and customer load
panels from the two UPSes and their respective static transfer switches and
output transformers - so we would have completely redundant dynamic A/B
switching. The ONLY vendor at any price I could find was for a switch made
for the FAA for their regional approach control centers critical power
systems. It cost $250,000 wholesale with a six month lead time!!! I decided
I would take my chances with the integrated MGE static transfer switches
and use one output transformer. For that amount of money, I can buy two
more UPS units when my load grows to 75% (to minimize the stress on the
circuit components and to control the transformer temperature by running it
well within limits) and still have $90k in the bank. Since we designed and
installed everything ourselves, I know we can bypass one UPS if we ever
have a catastrophic failure of the internal static transfer switch. In
retrospect, I think it may have been better to have a split output for UPS
#2 one feeding the bypass for UPS #1 and one feeding a break-before-make
manual switch which normally connects the output of #1 to our circuit
panels for quick service restoration in the event that UPS #1 fails in the
critical path such as the output transformer or the static transfer switch.
The comments about two MGE units failing due to component failure is a bit
scary since our serial numbers are actually the same and we just have
different suffixes due to being purchased on the same PO! Whatever
components are defective on one are probably defective on the other also.
So far, 16 months and zero problems... (knock on wood)
-Robert
Tellurian Networks - The Ultimate Internet Connection
http://www.tellurian.com | 888-TELLURIAN | 973-300-9211
"Good will, like a good name, is got by many actions, and lost by one." -
Francis Jeffrey
More information about the NANOG
mailing list