On 6/1/2011 5:47 AM, Robert Green wrote:
wrote in message ...
....

... he started, at least,
asking three specific questions:

1) when coming up with a total current measurement, do I include the
current flowing on the neutral line?

2) He was looking how to calculate the "worst case" (all loads operating
simultaneously I assume)

3) He wanted to know if there was an error in how he intended to calculate a
maximum load.

That's at least why I recommended the Kil-O-Watt. He described a large
number of easily tested, non-hardwired loads on his premises.
....
From the numbers on usage he posted before, it appears to be a fairly
consistent usage w/ a couple of months extreme outliers. Both of those
are early in different years; my hypothesis is that the other months
are, for the most part, estimated rather than actually read and the one
annual blip is the catchup because they've not updated the average usage
to reflect actual since it was set up (probably before OP bought the
building).

Because I've had my own tangles with bad meters and something the phone
company called a "left on" connection (that actually turned into a national
security matter), I pay attention to news reports of billing errors. This
could just as easily be a screw-up with his electric company's billing
systems. ...

Indeed it could...


While you may be perfectly correct in why the readings show the way they do,
there are a lot of other explanations. But it terms of learning, I think we
(and perhaps he) have learned some critical things. You can't meter the
neutral of a three-phase feeder and expect to get information about total
usage, only the level of imbalance of the downstream circuits. We've
learned that it's unusual not to be able to see your meter whenever you
choose to, but apparently not universal. That's something for the OP to
work out between himself, the power company and his landlord. As nearly
everyone has said, to solve the puzzle more information is needed.

But I don't think it could ever hurt him to spend $25 on a Kil-O-Watt meter.
He indicated from the start he wanted to measure total wattage, and the KOW
will do that far better for the gear he has than any tong meter I know of.
....

I never said (and hopefully didn't imply by poor wording) my idea was
the only possible explanation; simply a theory (apparently refuted,
eventually).

OP is the (new, but not terribly so iirc 6yr) owner of the building so
I'd presume he has some pull w/ the landlord... It would seem
anything about the lock on the cabinet/meter that he would be allowed to
do by whatever limitations placed by the utility/city could have been
resolved in that time. If the limitation is only inertia on his part,
that's another kettle...there's insufficient information to know _why_
it is currently as it is.

It's a commercial building w/ 3-phase service and the loads he mentioned
were a few computers, basically. The Kil-O-Watt meter can monitor them
but it'll do nothing for what is most likely by far the majority of the
load which will be the 3-phase lighting, possibly water heater and other
service loads. It'll be measuring the noise around the edges.

The characteristics of the peculiarities aren't likely to be explained
by a loading issue anyway imo. The doubling of the readings in months
that are roughly a year apart is just not credible as an ordinary event
of somebody left the lights on over a weekend. It's either an
accounting issue, an error in the reading or the like that is
artificially being induced somewhat like your above suggestion (or mine
of a different yet similar mechanism) or there's a _major_ intermittent
load or fault somewhere on the system.

Either of those isn't going to be found by a minimal one-time
guesstimate of the maximum possible monthly usage even if he measures an
instantaneous 3-phase usage correctly and it certainly won't be found by
poking around on one computer supply at a time.

$0.02, imo, ymmv, etc., etc., etc., ...

--