6

I have a submersible pump for which I'm trying to determine the variability in flow rate, to high precision. Since this is just an impeller driven by an AC motor, what is the % error in frequency of a US wall outlet? Is there a commonly accepted 'ballpark' value, such as 60 hz +/-0.5 hz?

My suspicion is that for most electrical grids, this largely depends on the citywide usage throughout the day.

I apologize because I have an inkling that my question is poorly presented -- that being said, if anyone knows of where I can find real-time data regarding the West Los Angeles electrical grid, I'm all ears. Thanks!

Chris Knudsen
  • 3,323
  • 12
  • 20
Austin Prater
  • 129
  • 8
  • 2
    Related: [mains frequency accuracy worldwide](https://electronics.stackexchange.com/q/83597/6334). Answers cover Australia and North America. – The Photon Sep 21 '18 at 01:33
  • 2
    Also, there is no "West Los Angeles grid". Pretty much the entire western US and Canada from the Rockies west are a single synchronized grid. – The Photon Sep 21 '18 at 01:37
  • The whole grid is connected together, you cannot have variations in frequency without large circulating current (which no supplier wants. Typically the grid frequency will not vary by more than 1mHz (somewhat old but this may help: http://www.leapsecond.com/pages/mains/). The time is adjusted so mains synced clocks read the correct time in every 24hr period, so constant small adjustments are negotiated over the suppliers. – Jack Creasey Sep 21 '18 at 01:39
  • Also related: [How precise is the frequency of the AC electricity network?](https://electronics.stackexchange.com/q/57878/6334). – The Photon Sep 21 '18 at 01:41
  • 10
    You'd be better off *measuring* the flow rate directly, rather than trying to predict it from things like mains frequency and motor characteristics. – Dave Tweed Sep 21 '18 at 01:51
  • Clocks which are synchronized to the grid never need to be adjusted. That tells you everything you need to know. – user57037 Sep 21 '18 at 02:34
  • I don't know much about measuring flow rate directly. But you could certainly get a pressure transducer and measure pressure directly. Depending on what the pump is doing, maybe pressure is good enough. You could also possibly measure pump RPM. Put a sensor on the pump shaft. Measure pump RPM, liquid temperature and output pressure of the pump. Use all that to calculate flow. – user57037 Sep 21 '18 at 02:39
  • 1
    "an impeller driven by an AC load" - what do you mean by that? – Ale..chenski Sep 21 '18 at 03:50
  • About every country somewhere has a website measuring that frequency, you might want to search – PlasmaHH Sep 21 '18 at 05:49
  • 2
    Related: in Europe, there recently was [a situation where the frequency of the grid deviated by 0.004Hz](https://electronics.stackexchange.com/questions/360328/serbia-kosovo-power-grid-row-delays-european-clocks-why) for a few months, resulting in mains-referenced clocks running up to 6 minutes late. Eventually, the frequency of the grid was raised by 0.01Hz for a period of time, allowing clocks to catch back up. – Tom van der Zanden Sep 21 '18 at 07:50
  • 1
    From the number of easy duplicates, did you even search for this? – pipe Sep 21 '18 at 09:33

4 Answers4

15

Typical accuracy of the AC Mains in North America is extremely accurate. Over a long period of time (hours), total error is such that an electric clock with synchronous motor has exactly zero error.

The instantaneous frequency may vary by some tiny fraction of one Hertz but the long term error is actively corrected.

I'm assuming that your impeller pump uses an induction motor. The slip in that motor will introduce far more error in your measurement than frequency changes of the AC Mains.

Dwayne Reid
  • 23,390
  • 2
  • 35
  • 67
  • Is the error held to zero for timekeeping (as wikipedia implies), or is there any other reason? And would you assume every country has zero long term error, or is it a very hard task which only more advanced countries can accomplish? (https://en.wikipedia.org/wiki/Utility_frequency) – piojo Sep 21 '18 at 03:57
  • 1
    @piojo check out the info in the other comments, but you may be better, if you are certain it has not been asked before, to ask your own new question. – Solar Mike Sep 21 '18 at 04:58
  • 2
    Your answer would be better if you explain WHY it is so stable - that it is actually the best way to make sure that all grids have enough power. If more power is consumed than produced, the frequency will dip untill consumption meets production. And opposite, if less power is consumed than produced, frequency will increase. Power consuption will increase b/c frequency, and again we are at equality. (But the utilities and the power companies will work to reverse any changes to frequency) – Stian Sep 21 '18 at 07:13
  • @piojo Have you actually read the wiki article you've referenced? Because it clearly says: "The primary reason for accurate frequency control is to allow the flow of alternating current power from multiple generators through the network to be controlled" – Maple Sep 21 '18 at 07:35
  • @Maple Yet that doesn't imply zero-error over time is needed for that application. Not every self-balancing system has zero average error. – piojo Sep 21 '18 at 07:52
  • 1
    @piojo You are right, the same could be achieved with synchronization only. Well, timekeeping is reason enough for me – Maple Sep 21 '18 at 08:00
  • 1
    To add some support to the answer, I had an alarm clock made in the early 1980's that kept time almost as well as my cell phone. I don't remember ever really adjusting the time. The clock was basic and ran off the freq of the mains. I used it until it recently (mere months ago) had a component fail, causing it to gain hours a day. On the other side of things, my 3 year old microwave has all kinds of features and gains about a minute a month. – computercarguy Sep 21 '18 at 14:28
14

Frequency of AC power grid in North America is monitored by a network of Universities. You can see the live gradient map at FNET here in real time. It is maintained to about 0.1% across the continent.

enter image description here

However, this doesn't look like having any relevance to accuracy of impeller-based water pump, since the flow will depend on pressure differences and viscosity. Maybe a peristaltic-based pump might give you some accuracy proportional to AC mains frequency, but it will give you a constant flow rate, so it is totally unclear what you are trying to measure.

Ale..chenski
  • 38,845
  • 3
  • 38
  • 103
13

If your submersible pump uses an induction motor - more than likely - then its rotational speed will be mains frequency minus slip frequency. The slip frequency will vary with load, including motor/pump friction, pump head and viscosity of water which varies with temperature.

Mains frequency drift won't be your problem.

Transistor
  • 168,990
  • 12
  • 186
  • 385
5

The answer is that a impeller pump is almost certainly not a positive displacement pump. Thus, the flow will vary upon back pressure, viscosity and inlet pressure.

In addition, the motor is not a synchronous motor, so the speed of the motor will vary upon load (=flow rate).

In short the frequency deviation of the mains network is the least of your problems when you try to calculate flow rate from rated motor speed.

vidarlo
  • 991
  • 7
  • 15