I think my meter is clocking too fast - How can I test if my high usage is due to a faulty meter?

Not unrelated to @fatbloke88‘s questions, I’ve been looking in more detail at my own IHD recently. I’m in the fortunate position that, aside from only about half the nightly gas meter readings being retrieved successfully by Ovo, everything else about my smart meters seems to be working ok. In particular, almost all the HH usage data get through successfully. (It’s a shame, therefore, that Ovo seem not to make as much use of the HH data as they might.)

So, the info that has caught my attention is the usage so far data, which is available

so far today, so far this week, so far this month, and so far this year.

From my own observations (of an IHD and meter that seem to be working fine) I’ve concluded that something (either the IHD, or the meter, or the Comms Hub, I can’t say which) is taking the difference between the current meter reading and a stored meter reading. I can explain what I see no other way: these usage so far values are live and update as one would expect them to, given how the current meter reading changes.

With care, one can work out what the stored meter readings are (with a precision of 0.001 kWh, despite the usage so far values only being shown as a whole number of kWh when they are 100 kWh or more).

I believe these readings were taken within a minute or so of midnight at the start of 1 Jan this year, and the start of the 1st of the current month, and the start of Monday this week, and  last night.

I’m still reviewing the data I’ve recorded from my own meter, but it would be fascinating to do this with @fatbloke88 ‘s IHD to see what those old meter readings are.

BW

PS in the case of a new account, I suppose that the new supplier resets these old meter readings (especially if they’re stored in the IHD - I’m unclear to what extent the supplier can go trampling over data stored in the meter/comms hub, and kind of hope that they’re not able to do that but, because it relates to billing, I fear that DCC will let them do exactly that) to ones that relate to the day the account opens, as best they can, anyway.

Sheesh!

It’s only 10 o’clock on a Monday morning and I’m already getting fresh homework from @fatbloke88 

I’m confused: Can anyone offer insights into this at all please? It looks totally random to me. Yes I can see usage spikes for cooking I can account for some variation with laundry days (washing and drying). Apart from that I cant see anyhing other than a weird suspension bridge! Is this the best kind of chart for this type of data.

Note: this is only from 12:00pm midday onwards for the period in question. The last day in the above sequence is missing 11:00 and 11:30PM readings for some reason @Tim_OVO 

 

This is the summary table data from the above chart and this is the bit that confuses me most;

The difference between the reads I take and submit and the HH reads allegedly given by the SM is astonishing. 34.59kWh units over 15 days!!

Am I doing something wrong in my collection and analysis or is there a fault somewhere in the Ovo system?

 

Thoughts and suggestions please, all welcome @Blastoise186@Simon1D@Transparent 

So, let me test my understanding by describing how you’ve processed the data, @fatbloke88 … (please put me straight where needed)

You’ve got 15 days’ worth of HH usage data (afternoon values only), copied and pasted (I assume) from the Electricity, Usage, Daily pages of My Ovo (the web portal) and stacked up in the plot so that we can see the total broken down not by day (as we’re more used to seeing) but by time of day.

The first thing to keep in mind is that the HH values are only ever kWh per half-hour. What you get when you read the meter is the accumulation of these since the meter was installed (or whenever it was reset to zero).

To relate the two, you have to take the difference between any two of your meter readings and that should be close to the sum of all the HH usage values between the times when you took your readings.

Somehow, to make sense of the comparison, you’d need to be taking readings at midday (and midnight, presumably) each day, to pin down how much of the daily usage is happening am, and how much pm.

My turn to pause, see if that makes sense to you, and allow you to correct any misunderstandings there might be on my side.

Cheers

PS It might be ok to just observe what your IHD says around midday, every day - it will show you the meter reading, and the usage “kWh so far today”. Just to save you the trouble of actually reading the meter at midnight, I mean. If you’ve got both numbers for each day, that should be helpful in making sense of things, going forwards from now, anyway.

PPS The missing values from yesterday evening are to do with the clocks being put back, and the web portal My Ovo does a really poor job in handling that. If you have the smartphone app, that is much better, and worth looking at for yesterday’s data, even though the copy/paste trick no longer works… (I’m not sure if you can see what’s in the Treehouse area, but I posted something on this earlier today, here)

Thanks for the update, @fatbloke88.

Ok, so the read at xx time in HH data set is the kWh used sinced the last HH completed, yes?

Yes indeed.

Why would it be close rather than exact? Is there something that would allow or generate a discrepancy and if so would you expect it to be of the size as per the data set?

The complication is that the HH usage data are increments in the meter reading, i.e. the amounts by which the reading would have increased, every 30 minutes, if we’d been taking readings at exactly the right moments, but we take our readings when we take them, and that’s at slightly different moments in time.

So, “close to” really means with the few tenths of a kWh that might plausibly have been consumed in the intervening time. This is definitely not potentially a several kWh discrepancy. (I’m just cautious in the way I express things. Or try to remember to be...)

I do have the HH readings for all of the above throughout the midnight - midday period showing circa 0.2kWh/HH. For this reason I did not provide it to make the chart more readable during the active period, I can provide that for comparison purposes if required.

As you do have all the HH usage figures, I would suggest the following. First, I’ll identify the key data:

1. You have your 15 readings, taken around 9:30am daily (meter readings, not usage values)
2. Ovo has its readings, taken nightly, around midnight (as far as I know). Also meter readings, not usage values. The only place I know to get these is in the smartphone app (where they’re truncated to whole kWh). Of course, it may be that the smartphone app doesn’t list many daily readings at all, because they had to be estimated. (My app only ever shows the “smart” readings. If Ovo had to estimate, that date is just missing from the list. If very many are missing, that would be revealing in itself...)
3. Ovo does present what it thinks the daily usage is. These are usage data, not readings. (They are increments in the meter reading, from one day to the next - sorry for the repetition, but I’d like others reading this to understand things as well as we do.)
4. You have the HH usage data, already in a spreadsheet.

Next, the analysis I would make

1. subtract your earliest meter reading from all the others. The result of this subtraction is a value which is the total usage since the moment of that first reading. It’s as if the meter had been reset to zero when you took that first reading.
2. Do the same thing with Ovo’s readings. Their first reading will have been taken within a few hours of your first reading (first in this set)
3. add up Ovo’s daily usage figures, starting with the day you took your first reading.
4. add up the HH data, also starting at a moment close to the time of your first reading.

You now have four versions of what should be the same thing, i.e. the answer to the question, how much energy (kWh) did I use from date x to any point, N days later than x. where N is anything from 1 to 15.

Since they’re the same thing, they should agree pretty closely.

Good luck

A miracle the smart meter lives!!!

Now what  I have to do is follow the readings it gives on my phone and lets see what it says, figure out with Ovo why the massive discrepancy for last year +£500 settlement, get it refunded, (no evidence that we did actually use it) as I believe data supports us….

What do you think our chances are??

@Blastoise186@Simon1D@Transparent 

Nice. I wonder if your meters have just completed the migration to DCC perhaps? Enrolment & Adoption can take a while and the ability to communicate with meters undergoing the process is somewhat limited between the point where the migration starts and the meter reconnects again via DCC.

I wouldn’t know for sure if that’s what your meters have been doing, but OVO can check the system and find out for you. Now that you’re (hopefully) getting Smart readings in, it might be worth thinking about whether you want Half-Hourly, Daily or Monthly readings again. I’d definitely recommend considering either Half-Hourly or Daily mode if you’d like to help prevent issues like this again - especially since you’d be able to spot them far more quickly if they do reoccur.

It’s possible that OVO might have been able to grab and rebuild your usage data from your meters, since the meters keep around 13 months worth of data before it starts to be overwritten. This data can be retrieved at anytime by the supplier if needed, even if they’ve previously retrieved it. Comes in rather useful sometimes. :)

Obviously, you won’t be expected to pay for any energy you didn’t actually use. If OVO can verify what your actual meter readings should have been all this time, it’s possible to do some heavy number crunching to recalculate and regenerate the bills that might have been affected. If that number crunching spits out a result that indicates a refund is due, OVO will be happy to make the arrangements. I think the Support Team can either have the refund credited to your account balance or fire up BACS to throw the cash directly back into your current vault at the bank via a Giro Bank Credit. Either way, there’d be an audit trail so you’d be able to find out where it goes to.

Definitely worth asking the Support Team to take a look into this. It might take a while, but they should be able to help.

Tim

Thanks and as I think you can see I’ve been round the ring a few times with this issue, despite the complaints team allegedly handling my complaint, it was closed.I In fact at least two have been closed without notifying me or advising me of the outcome. In essence they were the same complaint. I’m having to spend a lot of time on this issue, far more than I would have expected and it’s draining and distracting from my job.

I will follow up on @Blastoise186 suggestion about historical data but as I think we all agree it’s not likely a faulty meter or theft issue I’m going to be left looking at data suggesting we consumed a large amount of power with no actual evidence as to what, how or why. Very frustrating!

I’d say, not just you, @fatbloke88.

If you have the smartphone app, look at the list of what should be daily meter readings, under “Account”. My guess is that the reading for 14 Nov (which is the one taken near midnight on the 13/14 Nov, and which Ovo uses to arrive at its daily usage on both those days, 13 Nov and 14 Nov) will be missing from the list, but by now there should be a reading for 15 Nov. Here's what my app shows at the moment, for example:

My list for electricity is complete but, just for comparison, here's today's list for gas:

which, you'll see, is missing the reading from 12 Nov last week.

Looking at my daily gas usages from last week:

I know that those usages on 11 Nov and 12 Nov are Ovo estimates. I can compare them against the HH usages, added up over each day:

I’ve used the Bright app to get the daily totals of HH usages - the HH data are the same, they’re all retrieved via DCC, but the Bright app does the sum for me ;-)

You’ll notice that the daily figures match for 10, 13, 14 Nov (the days for which Ovo have both opening and closing readings) but not on 11, 12 Nov. Ovo made estimates, 45.44 and 46.69, which are different from the Bright app figures, 45.0 and 42.9.

But, remarkably, this is an occasion when Ovo made almost the best estimates it could because, for unrelated reasons, there is a gap in the HH data on 11 Nov:

and I happen to know (me? obsessive?? perish the thought...) that the consumption in that missing hour was a gas volume 0.380 m^3 which converts to 4.21 kWh.

So, no-one knows what that usage was - aside from me, who saw 4.20 kWh appear on my IHD as the first non-zero “so far today” figure on 11 Nov ;-)

But, because Ovo use opening and closing meter readings to get daily usage, they can fiddle their estimate for consumption on 12 Nov so that the total for the two days is correct:

Ovo’s total is 92.13 kWh

Bright app total is 87.9 kWh

The difference, 4.23 kWh, is within a gnat’s crotchet of the amount that I saw on the IHD, first thing on 11 Nov.

Ovo’s estimate for that day, is 0.44 kWh more than the sum of the available HH usages, and that really is an extremely plausible estimate for what might have be consumed during that missing hour. At that rate, I’d use 10 kW in a day and that is, in fact my typical minimum daily consumption (heating water only, in the summer, with the old boiler, now replaced).

But the thing that amazed me was that the estimate for 12 Nov has been adjusted (obviously?) so that the total over the two days is correct.

This is (I think, and perhaps @Transparent will confirm) the way Ovo’s estimates always used to work, but that stopped working sometime this summer just gone, and it was still not working last month, when I have two examples of the “sum” rule not being respected (documented elsewhere on this forum).

To be quite honest, I was amazed that Ovo seems to be doing the right thing again, for me, and for gas. But they haven’t gone back and corrected the estimates that they made late last month, and I wonder if they will do that before the end of the current billing month is up.

I know your questions are all about electricity usage, and the relation between HH and daily usage doesn’t work quite as perfectly as it does for gas (at least, when all the data are there), but it’s pretty close, and I suspect that Ovo’s cr*ppy estimation process is still being used there, now and back when your mystery consumption may or may not have happened.

Finally - have you tried looking at the other values on your IHD?

Usage so far today/this week/this month/this year??

I’m really interested to know if there is useful info there…

BW

PS Tagging in @Tim_OVO explicitly, even tho I’m sure Tim will see this anyway.

PPS I only say that Ovo made almost the best estimates that it could because, had it looked more carefully, it would have seen that the HH data for 12 Nov is complete, so that missing usage really should be all “apportioned” to 11 Nov (to be precise, that missing usage, I mean the usage inferred from the two meter readings they have, 48 hours apart, minus the usage that they have evidence for in the form of HH data, should be apportioned to that hour, nominally labelled 06:00 and 06:30 in Ovo’s list for 11 Nov). And no-one on earth can pin it down better than that because, as far as I can tell, it was at the moment separating those two HH intervals that the gas meter sent a reading, but the attention of the Comms Hub must have been elsewhere because that reading was never received. Without that reading, the Comms Hub couldn’t work out what the HH usages were for the interval before and the interval after that reading. It’s a little bit like Ovo’s experience with the missing meter reading at midnight, but on a smaller scale. [I’m really not suggesting that these two missing readings are in anyway linked. That would be superstitious mumbo jumbo.]

(sorry for the overlong post. But sometimes I have to get these things off my chest)

@fatbloke88 wrote:

as I think we all agree it’s not likely a faulty meter or theft issue

Erm… I haven’t concluded that you didn’t have a faulty meter when you had to pay OVO hundreds of pounds extra.

Just because us customers haven’t managed to arrive at an explanation for such a sudden jump in consumption doesn’t mean that the meter readings must’ve been correct.

The errors which @Simon1D and I have investigated/evaluated have been accumulative, based on lots of little mistakes and estimates.

That doesn’t match the experience you reported.

Polite words fail me at this point, @fatbloke88 

Sorry to see the discussion is still continuing here and that you haven’t got this resolved, @fatbloke88 - to state the obvious, no one here has access to your OVO account or meter data, so the forum really isn’t well placed to offer anything beyond what has already been said. 

I just need to quickly step in, as there’s some misadvice unfortunately. If that complaint is with a resolution specialist, that means it's already been escalated to a manager, and has been escalated to the next (and final) point internally. I’m confident this team of experts will offer a fair resolution to your complaint. If you’re not happy with the resolution when it’s submitted to you, the next stage is a deadlock letter and the Ombudsman. Full process outlined here. 

More generally, I’d like to remind everyone that we don’t know the issue here, or if Simon’s observations around usage estimation (not billing) relate to this. I’m going to suggest that it’s not accurate to suggest otherwise without having access to anything other than forum comments. Our guidelines are in place to try and avoid potentially misleading advice (full T&Cs here), but we’d prefer not to have to remove posts given that I’m sure the comments were made with the best intentions. 

Let’s keep this online space full of facts, and everyone benefits. 

Latest extract from email;

This will allow us to use actual reads automatically coming from the meters. The readings are coming through, the billing platform was not able to use them, hence needing to make some changes.

Why??