[time-nuts] Yukon Power

Dean Weiten dmw at weiten.com
Thu Apr 8 22:33:40 UTC 2010

Actually, these days, in the widely-connected continental area of US &
Canada (with exceptions like Texas), the frequency will be much closer
than +/- 0.5 Hz. Power systems will start "islanding" and shedding load
if frequency goes outside say +/- 0.1 Hz. It's very important for the
power system frequency to stay stable - energy flows and therefore
revenue flows depend on it. The power utilities are absolutely anal
about tracking frequency and maintaining it precisely.

It turns out that an AC power line is almost entirely inductive in
nature. This means that the power flow has little to do with the
voltage, more or less all about the phase angle difference between the
ends. Actually, a voltage difference generally means a reactive power
flow, which is counter-productive, so voltage is maintained very
constant as well. At the transmission line between cities, you see the
bird's eye view aggregate of power consumption - and utilities will put
systemic voltage support at various points to ensure that voltage is
well matched.

In order to keep the system stable, power utilities must maintain strict
frequency control.

In the case of Whitehorse YT, they are not connected to the continental
grid and therefore are not subject to the same controls. The power is
probably generated by diesel, coal, or perhaps hydro power. It is much
more difficult to maintain frequency control under these conditions - a
single big load (like a mill) "dumping" because of some disturbance, can
change your load base substantially, throwing the frequency way out of
whack. You can't instantaneously change the speed of the motor or
turbine shaft, so you have to slowly bring the system back in line.

Here in Manitoba where I live, we have hydro-electric generation way up
at the north end of the province supplying something like 90% of our
power, with most of the load in the south, separated by, hmm, something
like 800 km. Plus, we actually export *a lot* of power into the
north-central states. In the 1960s, Manitoba Hydro built an AC-to-DC
converter up north and a DC-to-AC converter down south, and ran the 800
km with a 500 kV DC line. There are lots of benefits to this, but one of
the biggest ones is that the phase & frequency of the output of the
DC-to-AC converter can be controlled on a cycle-by-cycle basis - it's
done with thyristors (huge SCRs, used to be mercury-arc but now almost
all solid state), so there are no rotating shafts. This gives our power
system excellent stability. In fact, Manitoba Hydro can modulate the
output of the DC-to-AC converter to help stabilize the system from
other, external disturbances. The possibilities are limitless. Of
course, the downside is that the loss of that DC line to the north
leaves us up a creek, and that happened about a decade ago when a
cyclone went through and took it down. What an exciting time that was!

The frequency of the AC system up north from the generators to the
AC-to-DC converter station is not well controlled. If we dump a big load
in the south, the DC system changes almost instantly to adjust, but then
there's nowhere for the energy to go, so the northern AC system goes way
up in frequency. Conversely, if we add a big load in the south, it can
go down, but then that's generally planned, and more controlled. They
say you don't run your clock on the frequency up north. In fact, nobody
does, because it's high voltage transmission, and there isn't enough
population to justify putting a distribution substation in there - some
folks complain that they can SEE the transmission line, but they are
forced to run on diesel generated power. It isn't a conspiracy, it's
pure practicality and economics - you can't just drop a wire and plug in
your stove! OK, maybe I digress, sorry.

What I do find interesting is that the Yukon folks don't have multiple
cross-correlated timing systems checking each other. If the frequency
were to go very low (say 50 Hz) it will cause significant dissipation in
power transformers and the like, perhaps even damaging them. It turns
out that flux in the core of said transformers is related to V/f, so if
the frequency drops over 10%, you get a substantial increase in flux,
and if you are close to the saturation of the core, well, let's say that
makes for not-so-happy customers.

Heavens, designing such systems is what I (used to) do. I could close my
eyes and see the design - it wouldn't be that difficult! Maybe I should
give them a call.

So when I see derogatory comments directed at the Yukon power folks, be
careful - they might be guilty of not having a backup time tracking
system, perhaps even of not tracking it automatically. But, it's much
more difficult when you are not connected to the continental grid!


Dean Weiten.

More information about the time-nuts mailing list