A little mental puzzle when reading an electricity meter

JohnB47

OK, a little mental exercise for you.

If you have an electricity meter (not a smart one), that provides readings in whole Kwh, what is the range of possible actual usage between any two readings?

For example, if you take a reading of 00500 at one point in time and later that reading is 00501, the actual usage might just be 1Kwh but it could be less or more.

I've had my brain cell working on this and I believe the actual usage could be anywhere between just over 0Kwh and just under 2Kwh. So when I take a reading, the actual usage could be the difference between the readings plus or minus 1 Kwh.

Have I got that right?

[Deleted User]

Mathematically it depends if the numbers are rounded or truncated.

Meters will truncate (or that’s what you do manually by ignoring numbers after the decimal) - so 00500 could mean anywhere from 00500.000 to 00500.999 (and the same for the next number up)

QrizB

JohnB47 said:

Have I got that right?

Yes, I think you do.

DullGreyGuy

JohnB47 said:

So when I take a reading, the actual usage could be the difference between the readings plus or minus 1 Kwh.

Have I got that right?

Unlikely, but what the actual answer is, as said already, depends on if its rounding or truncating (or doing something very weird)

When you read a normal meter with decimals you are instructed to truncate and so if your meter worked the same way 00499.99999999 would still read as 00499 and only when it gets to 00500.0000 does it tick over, then when it actually is 00500.999999 it still shows 00500 and therefore the potential deviation is +1 not ±1

If it has natural rounding then it could be anywhere from 00499.5 to 00500.4999999 and so ±0.5

If is "round up" then its anywhere between 00499.0000000001 and 00500.0000 and so max deviation of -1

The total spread therefore is only ever going to be 1 but how it sits relative to the read can change. Only if you introduced a silly type of rounding such as round up on odd numbers and round down on even numbers could you get the ±1 of one and therefore spread of 2

QrizB

There are some well-meaning people in this thread who seem to be making a meal of the question.

JohnB47 said:

If you have an electricity meter (not a smart one), that provides readings in whole Kwh, what is the range of possible actual usage between any two readings?

Yes.

For example, if you take a reading of 00500 at one point in time and later that reading is 00501, the actual usage might just be 1Kwh but it could be less or more.

Yes.

Whatever the mechanism - rounding, rounding up, rounding down, truncating - there will be a value that registers as 500 and a value that registers as 501.

• If the meter reading 500 is nearly-but-not-quite reading 501, then a tiny increase in use (small enough to be effectively zero) will flip it over to 501.
• If the meter reading 500 has only-just changed over from 499, then a full kWh will take it to 501 and nearly-but-not-quite another kWh will still read 501 without incrementing it to 502.

So yes, in practice 500 -> 501 could be only-slightly-over zero kWh, or could be very-nearly-but-not-quite 2kWh.

[Deleted User]

And for billing, it doesn’t really matter because whatever you miss (or gain) on one pair of readings - which isn’t really very much anyway - the opposite (ish) will have happened either in the pair before or the pair after.

DullGreyGuy

QrizB said:

There are some well-meaning people in this thread who seem to be making a meal of the question.

I admit I misread the question so yes my answer is on a wrong tangent.

So yes, the maximum consumed between the two values is 2kw (given that 0.9999999999-> = 1)

JohnB47

QrizB said:

There are some well-meaning people in this thread who seem to be making a meal of the question.

JohnB47 said:

If you have an electricity meter (not a smart one), that provides readings in whole Kwh, what is the range of possible actual usage between any two readings?

Yes.

For example, if you take a reading of 00500 at one point in time and later that reading is 00501, the actual usage might just be 1Kwh but it could be less or more.

Yes.

Whatever the mechanism - rounding, rounding up, rounding down, truncating - there will be a value that registers as 500 and a value that registers as 501.

• If the meter reading 500 is nearly-but-not-quite reading 501, then a tiny increase in use (small enough to be effectively zero) will flip it over to 501.
• If the meter reading 500 has only-just changed over from 499, then a full kWh will take it to 501 and nearly-but-not-quite another kWh will still read 501 without incrementing it to 502.

So yes, in practice 500 -> 501 could be only-slightly-over zero kWh, or could be very-nearly-but-not-quite 2kWh.

Thanks to everyone for replying.

This answer is probably word for word how I would have described my thinking. Thanks for saving me the trouble of typing it out!

MalMonroe

Hi,

I am no mathematician but I would think that in the end, it doesn't really matter because everything gets evened out in the end and the energy company will send you a bill and you have to pay it and there's no way to wriggle out of it. Just my own experience.

If you don't pay your bill then nasty things could happen. Or so I've heard!

JohnB47

MalMonroe said:

Hi,

I am no mathematician but I would think that in the end, it doesn't really matter because everything gets evened out in the end and the energy company will send you a bill and you have to pay it and there's no way to wriggle out of it. Just my own experience.

If you don't pay your bill then nasty things could happen. Or so I've heard!

Thank you. Of course you are correct but I wasn't remotely thinking of wriggling out of anything or not paying a bill. How did you get that idea?

It was purely a question about actual energy usage when you have an electricity meter that shows whole Kwh. Just a little mental arithmetic puzzle.

BooJewels

I've been following a similar thought myself @JohnB47.  I have just got a couple of the Tapo P110 energy monitors and have one on my freezer and one on the washing machine. 

Disappointingly they only record consumption in 100Wh increments - which means the data is less useful over smaller loads and shorter time periods.  0.1kWh could be 100Wh or 199Wh.  So I've decided - for my purposes - to assume it's half way between the reading and the next increment and count it as 150Wh etc. hoping it will balance out  Although I did just look at the freezer one, before I put some of my food delivery away and between screen views it did flip over from 0.9kWh to 1kWh - so the time of that is a useful data point - but that was pure chance.

You'd think, if they can display power draw to a tenth of a watt, they could record consumption to a watt hour, or even 10.