Appalling error by Octopus Energy.

Update: Within 24 hours of publishing this, Octopus sorted out the issue. Of course it should never have happened but mistakes do happen and I am grateful that they (eventually) sorted it out.

Apparently the error arose from a firm ware update to the smart meter which clashed with a meter reading, and Octopus’s system didn’t catch the error. 

Worryingly, several other readers of this blog and followers on Twitter have reported the same problem at exactly the same timeAnyway. I am older and wiser!

Friends, I am a customer of Octopus Energy who seem like a decent company with an agenda to push forward the energy transition.

But everyone makes mistakes. And I have just been sent a bill for a single month which is in error by more than £1,600. Yes, it is approximately one thousand six hundred pounds too much on a bill expected to be around £100.

This is an obvious error. Octopus’s own numbers do not add up. And yet they have still billed me for this incorrect amount.

I am sure this will be corrected eventually but I still feel frightened simply to see that amount indicated as an amount that I owe. Indeed, I am shaking as I type this.

The important question here is not my liability: it is this: If Octopus can make such a gross error and not notice it, how can we have have confidence that they – and other energy companies – are not making errors routinely but at a level at which people just might not notice?

What’s the problem?

Here are some excerpts from my bill for 12th December 2022 to 11th January 2023.

Click on Image for a larger version. Excerpt from Octopus Energy bill claiming that I used £1,873.30 of electricity between 12th December 2022 and 11th January 2023.

The excerpt above compellingly suggests that I owe Octopus energy £1,873.30 for electricity used between 12th December 2022 and 11th January 2023.

The excerpt below emphasises this. It shows the breakdown of that consumption between peak and off-peak hours. Most of the consumption arose because I apparently consumed 4,458.6 kWh of electricity during peak hours during this single month.

To understand how ridiculous that is, last year I used 3,323 kWh for the whole year! It’s an obvious error – but it gets worse.

Click on Image for a larger version. Excerpt from Octopus Energy bill suggesting that I used 4,458.6 kWh of electricity during peak hours between 12th December 2022 and 11th January 2023. For comparison , last year I used 3,323 kWh for the whole year!

Pleasingly, Octopus send a detailed breakdown of electricity showing consumption every 30 minutes during the entire month. the 31 pages each look something like the page below:

Click on Image for a larger version. Octopus send details of energy consumption for every 30-minutes throughout the month. Here is the page showing consumption on 12th December 2022.

From this you can see that on 12th December 2022, we downloaded energy to charge our battery using off-peak electricity. But on this very cold day, the battery had discharged by around 2:00 p.m. and we then consumed full price electricity for the rest of the day.

This cold day involved a relatively high consumption of electricity – 37.4 kWh/day – to operate the heat pump. But even if we did this every day it would still only amount to only 1,160 kWh for the whole month.

I went through each page and added up the energy used: It looks like this:

Click on Image for a larger version. Adding up the amount of electricity consumed through the month teh answer comes to 750 kWh at a cost of £104.94 – an average cost per unit of around 14.0 p/kWh.

Looking at the days individually and adding up the cost on each day, Octopus’s analysis suggest that I used around 750 kWh of energy in this period, or around 24 kWh/day. This is very much in line with what my manual meter readings suggested and what I was expecting.

It is not in agreement with the statement on the bill that I used 4,458.6 kWh of electricity during this period at a cost of £1,730. So Octopus’s own bill simply does not add up.

Octopus Response

I wrote to Octopus on 13th January pointing out this error and a representative replied to tell me that Octopus were aware of the issue and I would not be billed.

However it is now 24th January and I now have been billed! I am now alarmed and distressed.

Click on Image for a larger version. Octopus told me they were aware of the issue and I would not be billed.

How has this come about?

I think I know how this has occurred. I think it arises from a single erroneous reading of my meter.

I subscribe for £12 a year to the Powershaper service which gives me access to my own half-hourly electricity readings! I think these data are read from the same database used by the electricity companies for billing.

Looking through these readings I noticed an anomaly on 11th January 2023. In a single half hour between 7:00 and 7:30 a.m. I apparently used 4,295 kWh.

This is more than my entire annual usage last year in a single half-hour period! It corresponds to electricity consumption at a rate of 8.59 MW – megawatts! – for the entire half hour. This is not physically possible, and is obviously an erroneous reading.

Click on Image for a larger version. Octopus told me they were aware of the issue and I would not be billed.

I think Octopus have overridden their automated systems to correct this error on the day-by-day analysis, but failed to correct the consequental error on the monthly bill.

So how can Octopus have made such an error?

I just don’t know how sophisticated billing systems which must surely be audited (?) could possibly make such an error.

I am sure that Octopus will eventually correct this. But I am frankly appalled that it is even technically possible to issue a bill which is so grossly in error.

So check your bills!


33 Responses to “Appalling error by Octopus Energy.”

  1. Ross Mason Says:

    My Mum lived just down the road. At some point she wanted to sort out the Power Bill. She said that they come round every cuppla months and read the meter. Yes. Read the meter. One month she gets a bill. The next month she gets a credit. This had been going on a wee while. Over Winter thankfully. 🙂
    Anyhoo. She asked us to sort it. Looking over the last year of bills, sure enough, debit then credit. Silly duffs we thought lets check the meter reading…… It hadn’t changed for something like 9 months. Close look at the bill showed they had been diligently writing down the meter reading and not noticing it failing to move on in life.
    Pleading widowed pensioner, she managed to get a new meter and no bill for the period. Winner.
    I had a win a few years back. The Power Co decided they would make the bill…clearer. They enthusiastically said we used this much energy and here is your bigger bill. I pointed out that looking at the bill it was clear the energy cost was significantly less than the total. The line charge was included in the energy part. They ended up reprinting the bill to correct their marketing ploy.

    • protonsforbreakfast Says:

      Ross, Lovely to hear your ‘voice’. It seems from your comments – and others – that this is actually fairly common. It really makes me wonder about small mistakes – ones that are not obvious? Are there millions of them?

      All the best


  2. Simon Duane Says:

    Hi Michael

    That’s distressing – fwiw, I’m also with Octopus, and switched to them because trust in my previous supplier had completely evaporated. In the months leading up to my switch just over a year ago, I spent an unreasonable amount of time learning about and checking smart meter data …

    The meter registers (cumulative) consumption, and a meter “reading” is the value in this register (at a particular moment), and is generated either manually, by pressing a button on the meter, or in response to a request received electronically. As far as I can tell, only the supplier is authorised to make the electronic request for this kind of reading (and I thought this was because it’s what bills are based on, and only the supplier prepares bills).

    Try as I might, I was never able to establish what meter readings, if any, are actually stored in a smart meter. My best guess is that just 5 register values might be stored, one each for the start of the current year, month, week, day and half-hour.

    Almost all of what’s saved is half-hourly increments in the register value: “consumption” data. These are generated autonomously by the meter (using its own clock, which is corrected re UTC from time to time), and saved internally: the meter stores up to 13 months worth. Crucially, it is only the increment that is stored, not the register value itself, and these stored increments are what Loop Energy, Hildebrand, et al. may be authorised to request (as can the supplier, of course).

    Sometimes there is a failure in the system, and smart meter data gets lost along the way. To their credit, Loop Energy make it explicit when this happens, in the way they show the data. Others, including Hildebrand, replace missing values with zeros. At one time, Ovo would replace each missing half hourly increment in my gas consumption with a particular value ~180 MWh that just happened to precisely equal 2^24 litres of gas (using CV=39.5). But it didn’t matter, because Ovo didn’t (and still don’t) offer a “time of use” tariff, and these nonsense incremental values were never used in calculating bills.

    Some, including both Loop Energy and Hildebrand, will make repeated attempts to retrieve missing data. (Ovo may do this now, they never had much success in this when I was with them.) But if the value was never stored by the meter in the first place, those attempts are doomed to fail, forever.

    The point, of course, is that the odd missing cumulative value is of little consequence. The unknown value is somewhere in the interval bounded by the last reading before, and the first reading after the one that is missing.

    (These registers only record positive values – if energy might be fed back to the grid, the flow in the other direction is recorded in a separate register. As far as I know – I don’t have solar PV myself.)

    But a missing incremental value? Well, it’s not negative but, unless the link is made to the cumulative values, there’s no limit to how large it could be.

    And it’s only the supplier who has those cumulative values (so far as I can tell – my appetite for reading the regulations re smart meters has its limits).

    If Octopus is offering a time of use tariff (I’m assuming you’re on one of these, because your bill has such detailed info) without automatically making a robust link between the incremental and cumulative data, I’m genuinely surprised – I share your generally favourable impression of Octopus.

    Best wishes Simon

    • protonsforbreakfast Says:

      Simon, thanks for that information on the data structure. And yes, I am on a time of use tariff. And I am grateful to have the detailed information. Without it, it would be their word against mine, even though the consumption is preposterous!

      All the best


  3. edhui Says:

    It occurs to me that a retired physicist secretly experimenting with miniature fusion devices might have a cover story of working on a zero carbon home while, once in a while, he might have a momentary enormous spike in power consumption. I used to know a guy called Larry Livermore who did this.

  4. Terry Robson Says:

    Blimey! I used to think EdF were bad when they repeatedly “forgot” to repay the excess which would build up between my fixed monthly payments based on their estimate of a year’s consumption, and the actual readings. However, they readily coughed up when I phoned to complain. In those days there were humans around!
    I think you should take this up with the Regulator, as it could cause real distress to a customer unable to understand the charges. A fine is the only way to bring an an energy giant to heel.
    Good luck

    • protonsforbreakfast Says:

      Terry, Hi. Yes, I had that same experience with EDF, but they did eventually pay up.

      I am sure Octopus will eventually sort this out as it is obviously an error. It just worries me that it can happen at all – it seems there is no ‘sense-checking’ before bills are issued. And if big errors can get through, surely small errors would be almost undetectable?

      All the best


  5. Simon Duane Says:

    edhui’s comment reminds me of the case of, aka “the radioactive boy scout” – I was very entertained reading the book (that’s the title), almost wishing my own childhood had been so adventurous. It’s a delightful reminder of what used to be possible…

  6. Gordon Rowlands Says:

    I am surprised that the power companies do not use their historical data to undertake exception reporting that would quaratine bills that are way out of forecast before they are released.

    I don’t know if you have a mobile app for your energy consumption – that would give you an early warning if its the same data that goes into the billing system? It would be useful if such apps could automatically notify an alert for any excessive energy use – mine doesn’t.

    Also, the inverter for your PV system if similar to mine will have actual historical data for imported energy. That is based on an induction clamp so it is actual measured data.

    None of the above makes up for the shock of a big unexpected bill and the all the hassle of dealing with something that is obviously wrong. 🙁

    • protonsforbreakfast Says:


      Good Morning. I mainly monitor my daily usage through the Tesla App which is more detailed than the Octopus App. So I didn’t notice until I looked on line a few days later. And I have weekly manual meter readings which all make complete sense.

      So I am confident that I have enough evidence to argue my case, but as you say, it is simply shocking to receive a demand for such a large amount of money. It just shouldn’t be possible.

      All the best


  7. Paul Rudman Says:

    Thanks for the info on One of the (many) problems of living in a flat is that the meter is too far away from the flat for the home monitor to work, so they didn’t leave me one, so I can’t get the code number from it, so I can’t sign up to powershaper 😦 Hope Octopus sort things out for you quickly.

  8. Bob Pugh Says:

    Hi Michael,

    Sorry to here you have had this happen to you. I had a similar issue last April.
    This never showed up on my bills I only found out when I opened a loop account to view the smart meter data. I can only assume that Good Energy spotted that this was a spurious reading and worked around it.
    There is an interesting article from a university study in the Netherlands looking at smart meter accuracy:
    There was a rash of these problems reported coincident with the implementation of a price cap round in October :
    It looks like there is a problem with the smart meters and some energy companies have not implemented basic sanity checking on readings to pick it up. If only we had sane government perhaps they would engage your friends at the NPL to investigate.


  9. David Edwards Says:

    I was thinking of getting a smart meter!

    • protonsforbreakfast Says:

      David, overall I would recommend it. And similar screw ups happen with manual meters. I knew someone who had a reading in which there was a transcription – two digits were swapped – an British Gas insisted that the obviously erroneous reading was correct.

      We all just need to be careful!

      All the best


  10. Marcin Says:

    On a side note, would you recommend households with batteries and solar/cheap night tarrifs to prioritize going off-grid by switching to the battery power during the early evening peak (5-8pm)?

    I wonder what range of arguments there may be for/against(?) such prioritization.

    • protonsforbreakfast Says:

      Marcin, Good afternoon.

      It’s complex!

      There are many reasons to avoid those hours. Using the new ‘savings hours’ it can be possible to make money ! The reason companies will pay people to not use grid electricity is because it is MUCH cheaper to pay people not to use electricity during that time than it is to pay generators to operate ‘weaker’ plants. These are inefficient and dirty.

      But doing so is complicated even for people with batteries. First of all these savings events are most important in winter. If one has a gigantically large battery able to meet an entire day of winter demand then it is possible to avoid these hours with ease. But with small and medium size batteries, it can be difficult to plan – one would need to hold back battery usage earlier in the day – use the battery during those hours – and then switch back. Programming that stuff is quite complicated.

      For people who are still heating with gas this might be easy (easier), but heating with gas is terrible from a carbon dioxide emissions perspective. The best thing one can do from an emissions perspective is to stop using gas for heating and use a heat pump – but that makes it awkward during the cold days of winter.

      My recommendation is to do what you feel able to do. But to stay warm.

      Best wishes


      • Marcin Says:

        A household with a battery and a heat pump can be a very impactful `demand response’ resource in winter because by taking a heat pump off the grid during peak hours (and powering it from a battery instead) it can reduce the peak by a much bigger chunk than a household without a heat pump.

        Your comment about the difficulty of programming when a battery and when the grid are used by a household is very interesting! Given the great potential `demand response’ value of domestic battery systems for `peak shaving’, it seems to me that there should be a policy/innovation push for solutions that make it easy and attractive for home battery systems.

      • protonsforbreakfast Says:

        Yes! Agree with you absolutely. It needs to be available automatically otherwise people won’t do it!

        Best wishes


      • Marcin Says:

        What is the `programmability’ of your Tesla Powerwall battery like?

        Do you have views on who should idealy/by default do the `programming’ to significantly improve the demand response utilisation of home battery systems?
        – The householder (with the help of a suitably user-friendly `programming’ interface of the home battery system)?
        – The household grid power supplier?
        – The battery manufacturer?
        – Other?
        Of course, no matter who controls the power flows, the householder will need to find it (financially!?) attractive in a way that is easy to understand.

      • protonsforbreakfast Says:

        The Powerwall allows you specify the cost of electricity in 3 bands: peak, shoulder bands on either side of peak, and off-peak.

        The Tesla App is very clear, but the controls are limited. You have just allow the Powerwall to decide what it is going to do in one of two modes ‘money-saving’ or ‘self-powered’. You can’t programme a time when you want the battery to feed back to the grid, or a time to fill up from the grid, or to only fill up to a particular level. It makes its own decisions and it mostly does smart things, but it is sometimes difficult to understand why it has decided to do what it does! And you need to do that in order to benefit from the Tesla guarantee of degradation.

        I honestly don’t know who should look after this stuff. I think that appropriate solutions will emerge over the next few years.It’s all just beginning.


  11. David Edwards Says:

    More misery here:

    • protonsforbreakfast Says:

      David, that is really distressing. And really that’s the reason I wrote the article: this kind of thing should be just *impossible*. And somehow the errors seem to go in favour of the billing companies more often than they go in favour of customers. Mmmmm…

  12. Simon Duane Says:

    Hello again Michael

    I’m not sure why it took me quite so long to look this up but, by a remarkable coincidence, Octopus Energy assert that I managed to consume 8590.072 kWh betwen 05:30 and 06:00 in the morning of that same day, 11 Jan: [image: image.png] (that’s a screenshot from my account viewed on the web, but their app shows the same numbers. I’m not on a time of use tariff, so my charges are based on meter readings rather than consumption data. I shan’t bother to hound them for a correction, unless this error turns into a demand for payment. I”m not expecting that to happen, but I’ll see what next month’s statement brings. BW

    • protonsforbreakfast Says:

      Simon: How absolutely fascinating. I guess it’s likely that quite a lot of people have been affected by this. Hopefully Octopus will screen all the bills a little better than they did mine.

  13. Gareth Manton Says:

    Hi just checking my bill today I have also been billed for nearly £2000 I have looked back through the days and the error is on January 11 where in half an hour I used 4333.36 kw of electricity!!

    • protonsforbreakfast Says:

      Gareth, Good Evening. So that is several people to whom this has happened – just among that small elite group that read this blog! So I think we can conclude it must have been pretty widespread – arising from the firmware update on our smart meters.

      The good news is that it looks like Octopus are onto this. The worrying news is that it happened at all!

      Do let me know if you think I can help in any way.

      And good luck!


  14. Rose Rosinidi Says:

    My latest bill has an error of around + £1000. All down to an entry on 12th Jan ’23 when I was billed for 12,885 units at off-peak. all in one 30minute period. Should be around 0.05kWh! Now I have to wait from my email to be replied. Just hoping they dint try to take it out my bank.

    • protonsforbreakfast Says:

      Rose, Good Morning.

      This reading error – caused by a ‘firmware update’ from the meter manufacturer – seems to have been quite widespread.

      Octopus are aware of it and do seem to be dealing with it sensibly, but nonetheless it is worrying that such a thing can happen at all!

      Best wishes


  15. Neil Kitching Says:

    Sorry to hear about Octopus. I have found them to be very good. I am puzzled by my Vaillant Arotherm. I have the flow temp set to 45 degrees and heat curve of 0.6. I am on Octopus cosy tariff so want to do most of the heating in the twice daily 3 hourly cheap tariffs. The house heats up rather slowly. Should I increase the target flow temperature or increase the heat curve. It seems like they might both do the same job? In fact, given I can set the heating curve, why does it offer me the option to change the target flow temp?

    Any ideas? Neil

    • protonsforbreakfast Says:

      Neil, Good Afternoon.

      Unfortunately it is difficult because the Vaillant Controls are undocumented!

      You mention a flow temperature AND a weather compensation (WC) setting. These are two different ways to control the internal temperature.

      In Weather Compensation, the system looks at the outside temperature ONLY. Your thermostat setting of room temperature to (say) 20 °C is irrelevant. The system uses the WC curve setting to determine the flow temperature and that determines the temperature of the house. If the house is at the right temperature then your WC curve is correct. But as a consequence the system is limited to only just balance heat losses at the working point and so it will always be slow to warm up. I would use this if you keep the house warm 24/7.

      But there is then a Room Temperature Modulation setting which can be off (pure WC) active or expanded. I don’t know how to describe what these things do – Vaillant provide no documentation – but they definitely change the way the system works. It may be that one of these setting will make the system operate in a more responsive mode.

      I hope some of that helps.

      Good Luck: M

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: