In article GhWXc.328298$%_6.328292@attbi_s01,
Scott Moore wrote:
Bruce L. Bergman wrote:
On 27 Aug 2004 16:43:32 -0700, (ken) wrote:

Otherwise you might have to install your own primary engine
generator plant (Diesel, Gasoline, Natural Gas, Propane) to run the
CNC mill from, to get the reliability you need to never stop in the
middle of a cut - that way the power will be rock steady unless the
generator breaks down, or the fuel tank runs dry.

[ ... ]

For electronics, we used (I forget the term), an "in line" or "on line"
generator. Its a very heavy motor/generator combination with desel drive.
The idea is that the power comes from the generator, but it is driven by
a power line motor. The inertia of the large motor/generator rotors smooth
the power out through dips in the line, and also will give enough time
for the desel engine to start up if the power fails completely.

Hmm ... there was a emergency power backup for a building in
which I used to work. The primary reason for its existence was to make
sure that the fume hoods in the chem lab continued to work, as they were
sometimes working with some rather nasty gasses. (Nothing intended as
weapons, but simply as stages in creating experimental IR detectors.)

Anyway -- the backup system consisted of a big synchronous
three-phase motor spinning a flywheel (about 4' diameter by about 8"
thick, IIRC) monted in a pair of self-aligning roller bearing races in
pillowblocks.

Beyond that was a flexible coupling made of stacks of steel
shims as flexible links, a magnetic clutch made from the
electro-magnetic landing-gear brake from a B-58 (IIRC), followed by a
big healthy diesel.

There was a cabinet of electronics against the right-hand wall
as you entered the shed in which it lived.

The electronics monitored both the power line voltage and its
frequency -- and if either went out of spec, it would engage the mag
clutch, instantly spinning up the diesel, and that would generate power
by continuing to spin the synchronous motor at the same speed (1800 RPM,
IIRC).

Circuit breakers to this were 400 A three phase. One at the
synchronous motor, and three back in the nearby room where the air
conditioning and heating equipment lived.

If well maintained, it worked perfectly, even kicking in to add
support when the hydraulic freight elevator would start up in hot
weather during near brownouts. It supplied power to about half of the
building.

However -- there are still things which could go wrong. And
when they did go wrong, they tended to be spectacular.

First -- remember the big dual-row self-aligning roller bearings
which I mentioned? They were mounted in pillow blocks, fed oil from
individual tanks on the side rails of the machine. There was a rubber
hose from each tank to the corresponding pillow block. There was also a
temperature sensor screwed into the pillow block to monitor the oil
temperature, to shut things down gracefully if the bearing starts to
overheat.

Well ... the hoses got old and brittle, and one broke from
vibration -- apparently late on Friday of a three-day weekend. This
drained the oil out of the tank -- and out of the pillow block. After a
while, the bearing started to overheat, and generate interesting noises,
which got the attention of the guard force. However, they did not have
a key to the shed where it lived, and somehow had lost the home phone
number of the person whose responsibility it was -- so it was allowed to
continue to run.

Remember the temperature sensor? Well, it monitored the *oil*
temperature, and since the oil was on the floor in a puddle, there was
nothing to monitor, so the sensor thought that things were fine.

On Tuesday (remember the three-day weekend?), when the fellow in
charge came in and heard the noise, he went into the shed, and saw the
bearing spitting bright yellow sparks, so he shut it down and bypassed
it. If things had gone a bit longer, it might have seized and ripped
the pillow blocks loose from the frame and gone galloping across the
lawn to the next building -- about 150 feet away.

It was out of service for a couple of months while the
replacement bearings were procured (through an Army procurement system),
and the one fellow who maintained these around the country (and in
Alaska for sure, since that is where he was just before coming to our
place one time when I talked to him.) (We were in Virginia.)

Both bearings were replaced as a precaution, though there was no
doubt that one of them was beyond hope. It had rollers about the size
of 35mm fill cassettes. At least one of the rollers came out of it
*looking* like a film cassette, as it had metal smeared to form the lip
where the film exits. :-)

The second time involved the shims used as a flexible coupling.
The fellow who did maintenance said "The shims are starting to fail, it
is time to order some more. Here is the part number and a source. It
should cost about $90.00. I'll install them in six months when I am
back.

Well ... remember that I said that this was a US Army site? (A
R&D lab, actually.) There are always priorities, and some other things
took all the money during that period, so they were not ordered.

End of next six months -- same situation.

Sometime before the next six months runs out, someone presses
the call button on the freight elevator, and there is a horrendous noise
from outside. It turns out that the final shim in one of the three
stacks broke, applying a serious side load to the magnetic brake. It
was destroyed beyond repair.

Another several months of downtime -- more than six, this time,
as those brake assemblies are hard to find.

Money saved: $90.00

Cost of new brake assembly: $30K+

This is how the Army saves money. :-)


Final one involved someone testing the backup system. Remember
me mentioning that there were three 400 A three phase breakers? They
took enough force to set so there was a moulded phenolic helper handle
to slip on over the normal toggle.

Anyway -- testing involved turning off the breaker controlling
input to the backup generator. As expected, it took up the load without
a glitch. Now -- when power returns, the controller drifts the power
frequency until the local power is in phase with the power line before
switching back.

There were three breakers there. One disconnects power from the
input. One disconnects the building load from the output. The third
bypasses the backup unit, accepting power directly from the line. The
person doing the testing switched the wrong breaker back on -- the one
which connected the line directly to the building -- and the backup unit
was about 180 degrees out of phase with the line power. Result was that
the motor/generator was instantly stalled (with a loud grunt), and at
least two (and I think all three) of the fuses at the power pig cluster
on the pole blew -- blacking out not just that building, but several
others nearby as well. (It may have taken out some fuses upstream as
well -- I only know what I saw. :-)

Oh yes -- the backup system survived, but there was no way to
re-start it until the line power came back. :-)

And as it turned out -- nobody was using any of the nasty gasses
during any of those failures.

However -- something like this would probably be an excellent
way to assure clean uninterrupted power for a few CNC machines. Just
make sure that nobody short-cuts the maintenance. :-)

Enjoy,
DoN.
--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---