Octopus Tracker incorrect billing during power cut

Qyburn

Has anyone seen this? Our power was off from 0630 on 17th, until 1330 on the 18th. As a result we only used 2.339kWh on the 17th, and 7.075 on the 18th. Those figures are from my Octopus online account, and match those shown by Bright.  The Tracker bill charges 9.88 and 10.30 for those two days.

That was the first thing that stood out making me think Octopus couldn't cope with power cuts and had estimated a full day in the absence of data for those times. But the following day is wrong as well, we used 14.219 but were billed 10.70.

Should the bill be correct day by day, or do Octopus "spread out" your usage over the billing period? If they do then that kind of reduces the point of carrying out your high usage activities on the cheaper days. 

I'll need to download the data and do a full comparison.

Alnat1

If any of the data is missing, even 1 day, then Octopus will average out usage over the month. I've been on Tracker for 8 months and 2 bills were averaged, others were correct daily readings.

If you actually work out the difference, it's a matter of pennies. Tracker is still way cheaper than being on SVT.

Qyburn

Thanks. They haven't averaged it over the period, each day bills a different number of kWh, but what they seem to have done is more complex, increasing the low use days and reducing the high use. 

Alnat1

Does use gradually increase over the month? That is the usual pattern. Although my 2 averaged bills were as we headed into winter and the weather was getting colder. Maybe at this time of year it starts high and gradually decreases.

Qyburn

It increases a little but is up and down by not very much.  Those two highlighted days included the 36 hour power cut so actual use was 9.41 over the two days.  It looks like they've used estimated figures to fill in what they see as "missing" data.  In this case it's not actually missing, it's a complete record of the time power was available.  As for why they've billed for that period I have no idea, they've never used the 20th as a billing date before, but just seem to like issuing bills at odd times of the month,

Qyburn

Another thought.  When they download Smart Meter data do they get the actual cumulative meter reading, or do they just get the half-hourly consumption figures and hope that these add up to the actual total?

Alnat1

If you've got all the figures you need, work out the cost difference of an accurate bill compared to the average bill you received. Does it really make much difference?

Would it be possible that Octopus have considered the power cut to be "missing data" as no use was recorded?

MWT

Qyburn said:

Another thought.  When they download Smart Meter data do they get the actual cumulative meter reading, or do they just get the half-hourly consumption figures and hope that these add up to the actual total?

They get both.

If there are missing periods within the half-hour data then they use an industry standard profile to allocate the total consumption over the month.

So with your power-cut causing there to be missing periods on the HH data, you should find that the total energy you have been billed for the month is correct, but the split by day will not match the HH data.

They are looking at improving the handling of power-cuts which would mean only using the reallocation method if the total of the HH data doesn't match (within a reasonable tolerance) the amount used shown by the index readings.

Qyburn

Thanks. I don't know if the data for those HH slots is missing, or zero. Octopus download has them missing, but of course they may skip zero entries. Bright includes all the entries as zero usage, but that could be filled in by them.

I'll need to take a timed meter reading, and compare with the sum of Octopus usage data since last reading on  6th. Not 100% accurate as I don't have a time for that reading. I normally do them first thing, but that was a requirement when applying for Tracker, so some time late morning.

Spoonie_Turtle

MWT said:

Qyburn said:

Another thought.  When they download Smart Meter data do they get the actual cumulative meter reading, or do they just get the half-hourly consumption figures and hope that these add up to the actual total?

If there are missing periods within the half-hour data then they use an industry standard profile to allocate the total consumption over the month.

They do this with our electricity (no smart meter), it's never a flat average even within weekdays.

Qyburn

Just realised I should have grabbed the meter reading last night, now I'll need to wait till tomorrow to download Octopus data again.

I hope they get better at dealing with power cuts as everyone expects them to be more frequent nowadays than we've been used to for the last few decades. In the case of a power cut the total of all the HH figures should match the cumulative total.  In contrast if data was missing then it won't match.