Page 102 - Uninterruptible Power Supplies

Additional Information Relating to the Standby Supply Installation
100 Chapter Three

arrangements, the probabilities for success are found to be and 0.9999,
0.9997, 0.9994, and 0.99902, respectively. The corresponding probabili-
ties for failure are the differences between the probabilities for success
and unity, which are found to be 0.0001, 0.0003, 0.0006, and 0.00098,
respectively. (These are also the sums of the figures for failure in
columns 4, 5, 6, and 7.)

Routine Test Runs
It is most important that standby sets be regularly tested; despite this
well-known fact many installations are either not tested or are tested
in an unsatisfactory manner. Ideally, there should be a test run every
2 weeks, the period between tests may be extended but should not
exceed a month.
The best test is to switch off the normal power supply and to allow
the set (or sets) to run up to speed and supply the load as would be
expected in a real mains failure situation. For many installations this
is not allowed, sometimes the user does not have confidence in the
standby power or in the uninterruptible power supplies due to prob-
lems or equipment failures in the past.
If real mains failure testing is not allowed, some other procedure has
to be devised which proves the operation of the three subsequences,
mains failure detection, engine starting sequence, and load acceptance.
The engine control module may include a Test Start facility which proves
the mains failure detection, if it does not it may be possible to prove oper-
ation by manual intervention such as removing a control fuse.
If testing on the real load is not allowed the installation should include
a permanently installed test load which will provide the standby supply
with a load not less than the minimum recommended by the engine
manufacturer, probably 30 or 40 percent. This will prove all components
in the sequence, including voltage regulation, speed regulation, and load
sharing if applicable. This procedure introduces three small risks—
mains failure detection may not be tested, any loading sequence is not
tested, and the engine is not tested on full load. It may be that occa-
sional testing can be arranged at weekends to prove all three of these
functions.
Once started the set (or sets) should be left running until all temper-
atures have reached a steady state or for 1 h, whichever is longer. At
the end of the test check the fuel remaining in the daily service tank,
and if necessary and if it is not automatically done, top it up.
The test procedure described above assumes that the set will be
attended while running. If there is a supply failure during a test run the
control logic should ensure that the standby supply will supply the essen-
tial load, and in order to avoid an overload situation, the attendant must

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2004 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.