My off peak and peak tariff charges do not match the half hourly readings?

The quantity labelled 00:00 is - in the OVOverse - the quantity measured in the period 00:00-00:30, with due notice taken of GMT vis-à-vis BST. The OVO website uses, I think, local time everywhere, even though the underlying data are based on GMT timings. This explains why you may see just two bars on the Day usage page for today. They represent the first two half-hours of today during BST, but were in fact measured during the last two half-hours of yesterday GMT. Coincidentally, I just has cause to post an image illustrating the importance of noting the time stamps on usage data.

I hope you’re going to pin down the delay in your meter’s tariff switching. The possible 10-minute shift can prove very expensive if high loads are set to switch on at the wrong time each night.

I am going to take a careful look at the prices timings - I can get these from the app.

Getting there.

Having looked at the peri-midnight data for the period Feb6 to Mar5,  I had a strong suspicion that the first 30 minutes is being divvied up between the half hour and the half hour after midnight. 

In the 1130 to 0000 period there is barely a unit of use.

The kWh use between midnight and 0030 is significant (100kWh). That is way bigger than the apparent discrepancy between ~30 (half hr data) and 60kWh (bill) for the peak use.

It would make my peak use 133 instead of ~30kWh. If I halve it to 50 (ie 50% before and after midnight), then my peak use would be 84kWh - now higher than the 60kWh I am being charged for. Dont know what to make of it all!

I have now set my system to start off peak at 0030-0700 winter and 0130-0800 for BST. Hopefully that will fix it. So much for the 7 in Economy 7.

Please keep us posted with how you get on ​@Luppitt.

I take it that’s the aggregate of the 00:00-00:30 buckets for all 28 days in the billing period, so averaging ~3.5kWh each day, representing a power draw of ~7kW. That is a normal draw for a heavy domestic load, like an EV charger or an immersion heater and a storage heater charging at the same time. In your case, probably the battery being charged, possibly along with some other stuff. 

Assuming that there’s a delay in switching tariff at midnight, it’s possible that a third (10/30 minutes) of the 100kWh usage (i.e. ~33kWh) is being recorded on the peak register before the switch to the offpeak rate takes place. There will be a concomitant shift at the end of the offpeak period, so the first few minutes’ usage after 07:00 will have been recorded as offpeak. Regardless of what the delay is, there will still be 7 hours of usage billed at the offpeak rate, as promised by ‘Economy 7’. The one-hour shift during summertime only confuses these calculations still further.

Ok I get that. So my policy of not starting my battery charge until 0130 should ensure I am not charged peak at all for this. Likewise dishwasher etc

I am determined tot be mean and cheap!!

I have had zero understanding or even willingness from OVO complaints team to either look at he data I provided or to explain how discrepeancies might arise. You are far better informed than the OVO employees!

Incidentally, I have not seen on my register any “compensatory” difference at the end of the offpeak period. That “compensation” effect would also be affected by the relative amounts of power drawn at the start and end of the period.

So the total amount registered as offpeak is still less than that registered by the half hourly measurements even though the total energy over each month is in good agreement between the meter and the amount shown by half hourly reads. This of course suits OVO nicely (I am being really cynical here) if they are paying generators using accurate half hourly figures but charging consumers using less reliable monthly meter offpeak reads, then OVO are making a significant profit from those who are not aware of the switching problem.

I would like to know whether OVO now pays their generators using half hour reads.

I have raised the problem with the Ombudsman so will see what they say. Mean time I am continuing to avoid the half hour intervals at start and end of the offpeak to charge my battery. etc.

I have to repeat that this isn’t a problem! You are expecting two different quantities measured in different ways to be the same, when they are and always will be different.

Your meter sees to recording offpeak usage on one register and peak usage on another. It switches from one register to the other at times set by a ‘tariff switching table’ stored in the meter. The stored times are ‘nominal’ ones, because for the reason I gave many weeks ago the meter applies a delay of a few minutes to them. The meter takes a snapshot of all register readings each day at midnight (GMT) and stores it. OVO then collect the snapshot each day, so long as you haven’t opted for monthly data collection frequency. 

In parallel with this, the meter also records how much electricity has been used in each half-hour period, 48 times each day starting each day precisely at midnight (GMT). The meter’s bult-in randomized delay doesn’t apply to the timings of these half-hour periods, so you can’t use the quantities recorded to establish how much was at peak and how much at offpeak rates - that’s determined solely by the meter peak and offpeak registers.

If you have heavy loads controlled by the meter’s Auxiliary Load Control Switch (ALCS), the circuit they’re on will (ideally*) be activated and deactivated at the same instant as the tariff changes. 

You wrote ‘I have not seen on my register any “compensatory” difference at the end of the offpeak period.’ At the end of the period, the first few minutes of the randomized delay will be at the offpeak rate. Only if the load at the time is the same as it was in the first few minutes of the period will the quantities cancel each other out - the ‘compensation’ you’re expecting to see.  

Can you now see that there is no problem, no overcharge and no extra profit earned by a greedy supplier?

You raise an interesting question about how OVO pays generators. Trials are currently in progress to help develop a Market-Wide Half-Hourly Settlement (MHHS) scheme which should be introduced for domestic customers within the next year or so. I’m sure there will be an immensely complicated mechanism for avoiding settlement imbalances resulting from the discrepancy you’re not happy about.   
  

*   Ideally, because it’s quite possible for the switching times of the tariff table on the one hand and the load control switch on the other to be different. In an ordinary domestic situation, they should be identical. 

OK thank you. I am clearly a bit of a numpty when it comes to this stuff. I dont have ALCS by the way.

I can fully see that the meter is recording only 1 full unit (1kWh) each time it clocks up a notch and a snashis taken at 00:00 and 07:00. Whereas the half hourly record picks up much smaller amounts at each interval. I can also see that at switch over time there will be some variability as to which period any use is allocated owing to factors such as level of demand in the area, or whether a full unit has been reached on the meter.

What I cant get my head around is this.

Both methods give the same total amounts of consumption over 30 days. So the meter works and both methods give the same result despite the different methods of recording.

I would also expect some random variation in “allocation” of tariff around the start and finish times of offpeak/peak for the reasons you explain.

What I can’t understand is how over 30 days the meter clocks up 30 extra units of peak time which according to half hrly measures these should be offpeak units. The data looks to me a non-random error (~1 unit/day), especially as when inspection of the half hourly data shows minimal use over the day until 00:00 and nothing or tiny amounts after 07:00 every single day. If that non-random amount is clocked over 12 months that could amount to 360 kWh charged at peak rather than off peak. Not huge but significant. Say ~ £90/year vs £61. Three good bottles of wine depending on your standards. Could be more, could be less as I dont have a years worth of data (my data was wiped by OVO when a new meter was installed). But, if there are100,000 customers £6m/yr is a nice little earner for OVO. (call me cynical)

If this was down to random events at start and end of the offpeak (for the reasons you say), I would expect that difference to be nullified over 30 days. 

Maybe I am being really stupid and I am missing something. If so please accept my apologies.

No, no external factors will have any bearing on which half-hour bucket or meter register usage is recorded in.

No again, there is no ‘random variation’. The delay I’ve mentioned repeatedly is ‘randomized’. This means that each meter has a unique offset; it’s decided by a random number baked into the meter in the factory and another one entered by the engineer on installation. Once the meter is commissioned and up and running, that offset will never change. That’s why I’ve asked you to find out just how big the delay is for your meter, because it means you can decide just when - to the nearest minute - to switch a load on or off to be sure that it’s billed at the cheaper rate. The delay can in theory be up to 30 minutes, but in practice seems usually to be below 15 minutes.

I’m sorry I’ve confused you again by talking about ALCS. I forgot that you have a four-port meter, which is unusual for Economy 7 customers. The effect of the randomized offset is still extant, though, so you really must take it into account.