DIYbanter - View Single Post - Estimating KWh electicity billing using clamp-on amp meter

Home Guy

"hr(bob) " wrote:

You may be lucky and only have an error of 10 - 20%, but
I wouldn't be surprised if your error was as much as 25%.

When you speak of this error, do you mean

BOTH

a) an error between (what the billing meter is reading in terms of
KWh for some arbitrarily short time span) AND (what my amp-meter
derived watt-measurement would give for the same arbitrarily short
time span)

That error should be no higher than 10%. The billing meter is measuring
power (watts) that is phase-corrected. My method (of just measuring
current) is not phase corrected. So my calculation of watts will always
be higher by 5 or 10%. If all the loads in the building were purely
resistive, then my method should exactly match the billing meter.

b) an error between (the monthly power use of the office derived
from several different measurements with various devices and
appliances turned on or off as per time-of day and day-of-week)
AND (the actual or real pattern of device usage over the course
of a real month).

The variability or error in actual device or appliance usage is (I would
think) quite low, given that this is an office environment where the
duty cycle (ie the on/off power cycle pattern) of all of the 15 or so
computers is very regular and known, as are the lights.

The lights and computers probably account for 50 to 75% of the
electricity usage, the rest being fixed electronic infrastructure
(networking routers / switches / DSL modem, multi-function printer-fax,
multi-line phone system) a few low-power exit lights, a couple out-door
75 watt out-door mercury vapor lights (on a mechanical timer), an 18 CF
refrigerator, small microwave, a small capacity water distiller /
chiller (instead of bottled water), a toaster, a couple of small
bathroom ventillation fans wired to the bathroom light switches, the
furnace fan (granted it's a large motor, 220 VAC and probably 1 hp) but
it's not running all the time, a couple of radios.