Smart Meter Replacement - what happens with the HH data?

Qyburn

Phones4Chris said:

I don't think there will be any gap on a new meter. 

The HH data for the  day of the meter replacement won't be complete. Even in the unlikely event that data from the old meter is bang up to date when the fuse is pulled, there'll still be a gap before data from the new meter starts.

I phoned Octopus this morning about billing date, which they've screwed up again, and while i was on the phone I asked them to give the meter a prod. That seems to have worked as Octprice now shows complete data for yesterday.

MWT

Qyburn said:

I phoned Octopus this morning about billing date, which they've screwed up again, and while i was on the phone I asked them to give the meter a prod. That seems to have worked as Octprice now shows complete data for yesterday.

Unfortunately Kraken does not offer an option for a fixed billing date, so it will wander from time to time.

The CS staff can nudge it in the right direction, but they can't fix it in place.

Qyburn

MWT said:

Unfortunately Kraken does not offer an option for a fixed billing date, so it will wander from time to time.

The CS staff can nudge it in the right direction, but they can't fix it in place.

That doesn't really match what I've experienced. To start with I controlled the billing date by submitting readings on the 1st, that continued once the smart meter was fitted.

Then when I changed tariff to Tracker they changed to bill on the 20th. A few times I was told this was because that was set on my account (apart from the guy who said it was set on the meter). After a bit of pressure they changed back to the 1st and continued billing on that date for two years. Even where I submitted a manual reading mid-month I still got the bill on the first taking it to the end of the month.

Now immediately after I changed tariff this March it's gone back to billing on the 20th. Octopus can't tell me why but again explicitly stated it was set on the account, and that he had changed it back to the 1st.

MWT

Qyburn said:

Now immediately after I changed tariff this March it's gone back to billing on the 20th. Octopus can't tell me why but again explicitly stated it was set on the account, and that he had changed it back to the 1st.

That will still happen from time to time, it might be a year or two but it can still change.

CS staff will always try to help, but as a rule there are good reasons why the supplier does not want all their customers billing on the 1st of the month, so the CS staff are fighting the system to some extent.

Phones4Chris

Qyburn said:

Phones4Chris said:

I don't think there will be any gap on a new meter. 

The HH data for the  day of the meter replacement won't be complete. Even in the unlikely event that data from the old meter is bang up to date when the fuse is pulled, there'll still be a gap before data from the new meter starts.

I phoned Octopus this morning about billing date, which they've screwed up again, and while i was on the phone I asked them to give the meter a prod. That seems to have worked as Octprice now shows complete data for yesterday.

I think you may have misunderstood. The new meter won't have any gaps, it will have data as soon as it's active, that's my experience across a number of new smart meter installs. Yes, there may be a gap from the old meter if the data hasn't been pulled - which is why I suggested try pulling it yourself last thing via n3rgy if only to reassure yourself that any estimates that may be made are fair.

Phones4Chris

Ildhund said:

Phones4Chris said:

AFAIK suppliers read meter rate registers (defo R01, R02 etc., Export?) daily just after midnight (if they've got permission for more frequent than monthly) and smart meters apparently store 31 days of these register readings - 

I think it might be more accurate to say that the meter takes a snapshot of the registers daily precisely at midnight GMT and stores it for 31 days. Suppliers can then retrieve the snapshot at their leisure, although my own has been taking its time about it recently. There have been days when OVO have first retrieved readings more than 24 hours after their time-stamp. 

That's not my experience based on a number of situations where manual meter readings were also available, it appeared to be just after midnight but BEFORE night rate kicked in.
I think if it was precisely midnight then there could well be disagreements about the DATE of the reading. I was going to say that maybe this is something in-built to Kraken as it was encountered with suppliers using Kraken, but then I remembered that I personally had a row with EDF about their readings on the last day before a rate change (31st of the month) where the rate on the 1st went up. Their readings for the 31st did NOT include the off-peak usage on the 31st which they then tried charging at the new rate - this was before they switched my own account to Kraken.
EDF also said the readings were just after midnight.

MWT

Phones4Chris said:

That's not my experience based on a number of situations where manual meter readings were also available, it appeared to be just after midnight but BEFORE night rate kicked in.

Agreed, the daily snapshot is never guaranteed to be exactly at midnight as the meter is also handling the recording of the HH period that ends at the same time and that is a higher priority as it is time sensitive which the snapshot is not.

Ildhund

@Phones4Chris @MWT Interesting. This is what the manual for my Aclara meter says:

Daily Billing Data

The SGM1400-B Series meter stores 31 daily billing data sets before overwriting occurs. The snapshot will automatically occur on the crossover of the midnight boundary.

The daily billing data includes the following:

• Tariff TOU Register Matrix
• Tariff Block Counter Matrix
• Total Active Import Register
• Time and date stamp in the Daily Read Log
• Total Active Export Register

I'd be surprised if other SMETS2 meters behaved differently. "After midnight but BEFORE night rate kicked in" is surely the effect of the randomized offset, which applies both to the tariff table and to the ALCS calendar when appropriate. That's why you can never align the sum of nominal offpeak Hh usage figures to the advance of the offpeak register, which will exclude a few minutes after the start of the first Hh slot and include a few after the end of the last. It would be very difficult to establish just how much was used during those two few-minute periods. 

Have you ever watched the IHD's meter reading screen at midnight in order to be able to compare it to what the supplier eventually retrieves? I've never been able to fault mine, even with ± 1Wh precision.

It's just coming up to 23:59 GMT on 31 March, so time to take a reading to submit tomorrow morning ...


[ETA a few minutes later: 1677.019]

Phones4Chris

@Ildhund Thanks for that, interesting! However I would not interpret "on the crossover of the midnight boundary" as meaning exactly midnight, I'd say that would be one second past, but it's a mute point.
My remark "After midnight but BEFORE night rate kicked in" was not intended to have any significance to do with the  ALCS or randomised offset, it was to do with the fact that the supplier was attempting to charge me for the night rate on the last day of the old tariff, at the new tariff rate! When you are in the middle of winter (Dec/Jan) with 5 storage heaters running, the usage is significant (never mind the following day usage) and when the new rate is higher, that adds up to a few quid!
The solution is easy, you use the reading taken at midnight/one second past (which ever) - 1st Jan as the readings for the last day of the old tariff (31st Dec.). The usage between one second before midnight to one second after is insignificant.

MWT

Ildhund said:

Have you ever watched the IHD's meter reading screen at midnight in order to be able to compare it to what the supplier eventually retrieves? I've never been able to fault mine, even with ± 1Wh precision.

It's just coming up to 23:59 GMT on 31 March, so time to take a reading to submit tomorrow morning ...


[ETA a few minutes later: 1677.019]

Just a small note of caution, the time-base of the meter is not constantly synced to any external reference, small questions like at midnight or as it passes midnight are a little pointless as for example my current meter is about 5 seconds fast right now and I've seen it up to 2-3 minutes out in the past.

The suppliers can request the meter to be re-synced but it does not appear to be a routing event.

I first spotted this with small amounts of use drifting from one HH slot to the next around the point when my EV tariff changed from high to low rate.