What is the allowed range of the mains frequency in the US? It's pretty easy to find a norm for the EU, but for the US the various specifications I can find are contradictory. So:
Is there any formal act which defines that? If yes, what are the limits?
If not, is 69Hz within the sanity range (the upper limit of my NCO if somebody wants to know the application)?
I used to work for a large electrical utility. Typically in the daytime, we'd see the frequency drop to 59.98, but I don't think I saw it go any lower. Every night, they would then overgenerate slightly to make the daily average frequency work out to 60.00 Hz. I don't know about a specific document, but we members of the NEPCC, which set the standards.
If you are using a portable generator due to loss of power, you may be very glad to have 69 Hz!
I agree that 69 Hz is rather high, but it may only affect your clock. If you have critical equipment, even 61 Hz may be too far off.
A portable generator that can run at 69 Hz should be easily adjustable, but you would need some way to measure frequency. One handy tool to have for generator measurement, is a "P4400 Kill A Watt" meter. You should be able to find one for around $20. Reference
This shows how good it is, I'd like to know how bad it can be. In the EU it is 50Hz -6/+4% (47..52Hz), but certainly in most cases UCTE guarantees it to be 50+-150mHz.
What does the power company do if one day they accidentally exceed the "guarantee" for X seconds, which is more than your application can tolerate? Cut you a check for a thousand dollars/euros?
I think the only answer to the question of how bad it _can_ be is "nobody knows". I mean anything could happen. Finally all you actually have is the historical record and whatever inferences on average behavior you can make from that.
Half of Continental Europe ran on over frequency (overproduction mainly due to wind power) , while the other on under frequency (overload).
On the overload side, the frequency dropped to 49.0 Hz, when loads were disconnected, cause country wide blackouts.
Generators also have frequency protection relays, which will trip when the frequency is far too away from nominal frequency. This often an indication that the generator is no longer connected to a healthy grid.
Microchip's _AN954 Transformerless Power Supplies_ [1] explores capacitive drop supplies. Example 1 in the app note suggests 59.5 Hz as a lower bound. Example 2 suggests 60.1 Hz as an upper bound.
Caveat emptor: My math works for Example 2. It does not work for Example
So either there's a typo in Example 1 or my math's wrong.
Impossible, the inertial mass of all the rotating gear prevents it from happening instantly. The only problem are the portable generators mentioned by John, but even them need to comply to some norms.
The application will then switch to battery power mode, so it is safe. I just don't want it to happen within the allowed range.
This is exactly how it's specified: during a week-long period of measurement the freq must be within +/-1% for 99.5% of that time and -6/+4% for 100%. So, if that 1% is crucial for some reason, your downtime should not exceed 0.5%.
Locally it can go to 0Hz and 0V in an instant if a circuit breaker goes.
And if you are on a UPS at that point you are dependent on how well the switchover to local generator mains shapes up. Ours was never anything like as well frequency regulated as the true mains supply.
From an astronomers point of view we seldom saw UK mains drift much beyond +/- 0.2Hz except on a handful of really cold winters days.
Realtime UK frequency chart with 15s sampling online at:
formatting link
I'd expect the USA to maintain something similar most of the time.
I'd be very surprised if they ever had excursions of more than 1Hz.
Before cheap mass-produced quartz clocks became widely available, I used to see mention of electric clocks that used the mains frequency as the time base. I often wondered how that was acceptable.
45 years ago, my boss at the first place I worked at told me that the frequency of public power supplies in the US were kept within
+/-0.1Hz. That's a possible error of about 2.5 minutes per day.
Even if control has been tightened by a factor of ten, that's still a quarter of a minute per day.
After the network was split and this region was seriously overloaded, the frequency fell rapidly to 49 Hz, After much load was shedded, the frequency stabilized at 49 Hz in more than 5 minutes.
When extra capacity came on-line, the frequency slowly climbed to 50 Hz, making it possible to connect the other area with surplus capacity.
Much less long term since it runs slow during the peak load in daytime and faster at night when the load is lower. The indicated clock time on a mains clock might in the worst case drift around +/- 30s over the course of a bad day relative to a precise reference time standard.
But they also guarantee to hold the frequency when averaged over a 24 hour period to a very much tighter tolerance so that clocks do run true. ISTR the actual long term specification in the UK was classified. (maybe still is)
It was the bane of astronomers back in the days when mains synchronous motors were used for some telescope sidereal drives since in mid winter with the low night load they would always be running fast to catch up.
Today they are all quartz regulated steppers or servo drive often with closed loop real time star tracking ability.
The first paper covers Time Error Correction (line-based electric clock long-term accuracy), Frequency Relay Limits (for load shedding where load > demand)), and Frequency Trigger Limits (notifications only).
TEC uses small "tweaks" that slightly increase or decrease grid frequency by 0.020 Hz above or below nominal 60.000 Hz one the error has accumulated above a given threshold (10 seconds in Eastern US, 5 seconds in Western US, and by operator discretion in most of Texas). Under TEC, synchronous clocks will gain or lose about 1.2 seconds/hour until the time correction is complete.
There are also operational limits to protect equipment. Larger deviations from 60 Hz indicate a load-demand imbalance that may require capacity adjustments or load shedding. Under-frequency (Load > supply) is a much larger problem than over-frequency. Frequency Trigger Limits (FTL) of 59.95 Hz and 60.05 Hz over 5 minutes trigger notification that the grid demand and load are out of balance so that generating capacity can be appropriately adjusted. A larger under-frequency deviation of 300 mHz (59.7 Hz) can trigger Under-frequency Load Shedding (UFLS). If you happen to be in an affected region, your frequency will drop to zero Hz.
Large power generators sense under- and over-frequency using wider deviations (5% or so) than the power grid. These limits are mainly for turbine and generator protection. Typical under-frequency trips for a turbine-generator are 57.7 Hz (instantaneous trip) and 58.5 Hz (60 second delay). Smaller frequency deviations of +/- 0.036 Hz are autonomously managed via generator governor controls.
Also of possible interest, California recently adopted a separate set of frequency and voltage trip limits for green energy power inverters (Distributed Energy Resources). Under the new regulation, grid-connected power inverters must continue to operate at grid frequencies between 57
- 62 Hz to help stabilize the grid under abnormal conditions.
formatting link
So, at the grid level, the normal lower limit would appear to be 59.70 Hz. The upper limit does not seem to be well defined, but would appear to be less than 60.036 Hz. However, the new California regulations implies that worst-case short-term excursions as large as 57 Hz to 62 Hz may occur under abnormal conditions.
With the number of synchronous clocks dropping off I've read that the utilities are considering loosening the long term stability requirements. So my oven timer/clock may not run true much longer. :(
I used to have one of those flip-paddle digital clocks, but it appears the synchronous motor has gone to the big clock in the sky.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.