View Single Post
  #14   Report Post  
Anthony
 
Posts: n/a
Default

keith bowers wrote in
:

Anthony wrote:

"john johnson" wrote in
u:



Hi,
I work for a power company, and in one of my earlier roles
worked in the
test section, the section where all the difficult power quality
issues ended up. One that I recall involved a CNC machine shop, the
owner was very angy that we were supplying him with poor quality
power and damaging his machinery. We installed a dranetz power line
analyser, and boy was there some crap in his power systems, it was a
wonder anything worked at all.

The thing is though, we don't supply crap power, well mostly
anyway. By
crap power, I mean power that is still on, but full of dips and
spikes. We do on occasion fail to supply any power, but it's
excellent quality grin We can on occasion, have failures in our
voltage regulation equipment , and supply the wrong voltage, and
while low voltages can cause motors to burn, it's still very rare on
our network. It's also very rare for our equipment to mess up the
quality of the power supply. The only thing we can do to pollute the
nice clean stuff we get from the generators is burning connections,
and they dont usually last very long once they begin to burn. Where
99 percent of the crap comes from is customer loads, either the
customer themselves is doing it, or a close neighbor.



Wholeheartedly agree John....It isn't the power company's fault in
our case, and we know it. It's all the crap we generate in-house.
Induction furnaces, huge 400 hp air compressors (4x), big-assed
anodizing operations (some serious power use there), all the CNC
machines, etc. We pull enough power in our facility to run a small
city.



Is your CNC equipment fed directly from the main switchgear where the
power company feeds the facility? If it isn't, you might be plesantly
surpirsed at the improvement you would get by setting up a separate
distribution AND GROUNDING system inside your building just for the
CNC equipment; feed it right off the main power switchgear. Yse
larger wire than code requires; keep the non-CNC equipment on the old
feeds.

Another thing you cand try almost for free is to trace the lines from
the CNC equipment all the way back to the main building feed and check
that every connection is properly tightened. Over time this stuff WILL
WORK LOOSE. Several years ago I was in charge of powering a major
computing facility. Every Christmas we shut the whole place down for a
day and tightened connections. We always found one or two 8o(. It is
truly amazing the stunts some industrial electricians will pull;
especially with grounds. Ever see 50 volts of trash between neutral
and green on a 120 circuit? It does WONDERS for networked computers.
In this case the ground was connected to a different transformer than
the one feeding the power.


Keith,
There are 3 buildings being fed off of two 13.8K feeds a primary feed and
a secondary, the secondary feeds a limited area of the plant, unless the
primary has issues. There are 4 substations within the plant. Almost
all the equipment is fed from busways, and they are load-balanced through
the substations. The equipment not busway'd would be the very high power
users (induction furnaces, holding pots, the compressors and anodizers).
There is *no* way to segregate the CNC equipment from the rest of the
stuff, since the CNC equipment makes up probably 80% of the equipment in
the plants, and is located within all three buildings. There is over 1000
CNC machines at our facility. The cost would be prohibitive to re-
configure the entire facility.
However, as it stands now, we are basically maxed out on the in-plant
infrastructure. If we add much more, a new sub will have to be
installed, along with upsizing the incoming transformers, etc to handle
the additional load. This may be a time to segregate some equipment.

All connections for infrastructure are checked regularly on a PM basis,
all the way to the machines.



--
Anthony

You can't 'idiot proof' anything....every time you try, they just make
better idiots.

Remove sp to reply via email