[time-nuts] Heathkit clock available
bill.iaxs at pobox.com
Fri Aug 12 21:58:50 EDT 2016
There are two systems that affect line frequency anywhere in the world.
One is the use of multiple power producers generating steam for turbines
that turn huge generators. The generators are synchronized by the
distribution networks that connect them. A generator rotates at the
frequency determined by all of the other generators, If its turbine is
receiving less energy than is required to keep up, the generator will
take the balance of power from the network. If the turbine produces more
energy than is required, it will cause the line frequency to increase a
very little bit. Change 'steam' to 'water' for hydro-electric plants.
A networked connection of generators and loads requires a perfect
balance of power produced with power consumed to maintain constant
frequency. This can only be done for very small networks. You cannot put
a PID controller on each turbine and set it for a GPS derived frequency.
The control actions would fight each other and destabilize the network.
In the early days, each power station had a clock driven by the
generators and a reference clock. An operator would increase steam to
the turbines a bit if the station clock fell behind the reference clock,
or decrease steam slightly if it was gaining time. There was (is) no way
to predict the behavior of the loads, except from general experience of
the effects of weather and holidays.
Today's networks are much too large for control by station clocks. A
large region has a central power dispatching station. A dispatcher
tracks the difference between network time and GPS time. If the network
is losing time, the dispatcher calls the necessary number of generating
plants and asks them to increase power. It is common to lose time as the
loads of the manufacturing day increase and to make it up after 4 AM or
so. (During the 50s, the Air Force determined that 4:30 AM was the time
of minimum human activity, and so a probable time for an enemy to strike
with missiles.) I visited the Pennsylvania, New Jersey, and Delaware
(PennJerDel) region's dispatch center in the seventies. Very impressive
wall maps of major generation stations and load centers with their data.
It is expensive for a plant to change power, and so they hatched the
plan to stop trying to hold the time difference to zero over a day.
There was sufficient outcry to abandon the plan.
The other system comes from the use of high voltage DC tie lines to
exchange power between networks isolated by geography, such as the West
Coast and Texas. The DC lines use high voltage, high power solid state
inverters to convert DC to AC or reverse the direction when power could
flow the other way. The inverter frequency can be precisely controlled,
but it is controlled to balance the power flow, not hold the line
frequency. A network pays for the tie line power it produces or
Different regions can have different phase behavior. I have only seen
West Coast plots on this list. When I did some work with this in
Minnesota in the eighties, the phase variation was only about 6 seconds
during a day and zero from day to day.
This is my understanding of the system. People with more knowledge,
please correct my misconceptions.
More information about the time-nuts