[time-nuts] Power lines and time
bill at iaxs.net
Mon Aug 22 15:14:31 EDT 2005
Poul-Henning Kamp wrote, quoting me,
">Power companies bill on time-integrated power - watt-hour
>meters in the US. Watt-hour meters are still mostly driven
>by electric clocks, in a way. The frequency does matter.
Uhm, sorry, that is just plain wrong. The Ferrantis (sp?) power
meter which is the most widely used meter in the world is not
frequency sensitive within a band of +/- 10% or more."
You are correct, and I sit corrected. I was looking for a
simple example of why frequency mattered. I did not feel good
abut it as I wrote, but I didn't look it up. I knew that
crossed-coil wattmeters were not sensitive to frequency as
long as the coil inductance didn't matter.
">1. It is unlikely that any power network just lets itself go,
>with no standard time/frequency to hold. The under-frequency
>relays would make that hazardous.
Does not follow."
Well, it doesn't follow from the watt-hour meter, but look at
the larger network picture. The most certain way to tell that
the network supplies and loads are not balanced is to measure
the frequency. The frequency reflects the speed of the generators.
Generator speed is determined by the balance between steam power
to the turbine and load on the generator. Steam power determines
If the frequency is used to determine power balance then it
follows that all users of the network must agree on a nominal
frequency. Given a nominal frequency and the ability to detect
overload by dropping frequency then it is possible to protect
the network before the generators come to a stop. This is not
linear because the load goes up as frequency drops and transformer
As an example of the relationship between steam power and frequency,
there was a paper mill with three generators driven by co-generation
turbines. That is, the local boilerhouse steam pressure had to be
let down anyway, so turbines were used to drop 400 PSI steam to 30
PSI steam for low pressure equipment. It happened that an oil leak
froze the governor for one of the turbines. The operator wanted to
shut the turbine down, but forgot his training. Instead of closing
the manual steam valve, he tripped the generator's circuit breaker.
This left the generator and turbine with no load and full steam.
The turbine and generator rapidly accelerated beyond their rated
speeds. The generator disintegrated and threw cubic foot chunks of
metal through the roof. No one was killed, somehow, but the operator
was badly burned by escaping steam.
Bottom line, it is within reason that some networks do not synchronize
clocks with something standard. It is not reasonable that the network
dispatchers do not care about frequency and do not work to regulate it.
It's not regulated with any accuracy that we'd spend much time discussing
on this list but it is regulated.
When I toured the Pennsylvania, New Jersey and Delaware dispatch center
some years ago, I thought they had talked about keeping a daily cycle
count. The count had economic importance, in that a high count meant
that sources gave power away and a low count meant that users paid too
much. I can't justify that in terms of the fact that power is VxI on
the real axis. It has no frequency component. (Plain old VxI measured
with separate meters is volt-amps and VxI on the imaginary axis is
volt-amps reactive or VARS.)
Too bad that there's no one that understands power dispatching that
also has an interest in precision time.
How many of you have collected a radio clock that was meant to compare
time on a local power network with WWV or something more recent? Are
they still being made?
More information about the time-nuts