And over what period of time you're averaging Reloadron, for accuracy bellow 1hz the native clock is important, and the power grid averages every single cycle over hours if not days of time. You'd need a meter than can provide a calibrated .01 hertz or so accuracy to investigate power line frequency. Local voltage fluctuations are perfectly normal, it's not funny at all, I think the typical voltage allowance is 10%. Frequency much lower.
I understand how the US power grids work (actually the US has three power grids). Texas is unto a grid all its own with an east and west grid for a total of three US power grids. Corrections to the grid are done on a periodic basis which could be days or weeks.
When I answered the OPs question initially I was not aware if his clocks were deriving their accuracy from mains power or an internal oscillator and if mains frequency I had no clue as to the country of origin. While I understand the uncertainty of US Mains Power Frequency I have no clue as to the uncertainty of the mains frequencies around the world? I can only speak somewhat for Italy where I lived for 3 years and mains power frequency was nothing to write home about.
As to measuring US Power Mains Frequency. Resolving 60 hertz down to .001 ( 1 mHZ) is not that difficult at all. At home I use an old HP 5334B Frequency Counter with the 10 MHz TCXO oven option(5 parts in 10 ^ -10). While an older counter it will measure 60 Hz with .001 Hz resolution to an uncertainty of +/- about 30 uHZ. If I want better I can use an external time base. Anyway measuring 60 Hz mains power with .01 Hz resolution and doing it accurately really does not require much. Measuring the period for .001 Hz resolution would require a gate time of 100 seconds so a ten period average would be about 1,000 seconds roughly 17 min.
Back to US mains uncertainty. Yes, it is accurate but there seems to be change in the air. Remember those periodic corrections? Earlier in this thread I linked to
this forum thread. Which links to proposed changes to the US power grids and how they are maintained as to frequency. For those unaware
this is what is going on with the US power grids as to frequency.
All of this being a moot point as when the OP returned they stated they had battery powered clocks which brings us to crutschow's watch.
The idea at this point will be to see how much the clocks deviate (drift) over specified periods of time.
<EDIT> @ Hans:
It's funny looking at an analog clock when time is changed from summer to winter time. The clocks make good one hour within a few seconds.
Hans, isn't it Spring Ahead transitioning from winter to summer and Fall Back going from summer to winter? So when an atomic synchronized clock goes from winter to summer it would advance an hour real fast. When they go from summer to winter what does it do? Advance a quick 11 hours? That has to be amusing to watch. We have them here in the US but I have never bothered to get one. </EDIT>
Ron