Historical usage data after physical Smets2 upgrade

Howdy!

You’ll probably still have access to it under the Usage tab. But the Meter Readings tab would be reset to zero.

Thanks for the prompt reply - “Probably” suggests that I’m going to have to manually copy them across to a spreadsheet, as without them it would not be possible to assess whether a “storage battery project” would be cost-effective.

Are you having your physical meter swapped or is this a software upgrade? 

Physical meter swap

In that case, the Meter Readings will definitely be reset to zero.

If memory serves, I still had historical usage after a physical swap two years ago

Nice workaround, @Firedog!

@Firedog

I wonder if the 30min data on the usage page may be useful to assess battery storage solutions? 

I can't remember if you posted about looking at something similar for the usage page before to extract the data to a csv file?

Interesting the data on the usage page goes back further than that on the readings page where the older historical data looks like it is stripped out.

OVO are working on a csv download function, hopefully in 2023

  
@Jeffus @AlanW The 30-minute data are returned one day at a time, so it would take some automation to retrieve them for a period of days. They’re there, though, from time immemorial:
smartpaymapi.ovoenergy.com/usage/api/half-hourly/{account}?date=2017-03-10

… {account} being the account number, as usual. That was the day after my smart meter was installed! 

It looks as if usage data are stored indefinitely, but readings only for the 400 days of mandatory meter storage; I’d be happy to be proved wrong. Of course, if you had the time and inclination, readings could be reconstructed from the usage figures. There’s a caveat to that: the GMT/BST switch twice a year might introduce inaccuracies*, depending (I think) on the type of meter. You’d have to look at a few sample days to establish whether the usage recorded for, say, 00:00-01:00 belonged to today or yesterday according to the readings, and correspondingly for 23:00-00:00. Once the pattern is established, the appropriate adjustments could be made to the collated data. 

*  These could be significant in situations where, for example, a power-hungry appliance is set to switch on at midnight on some days only.
It’s interesting that the DCC data (as retrieved by Bright, for example) don’t always tally with OVO’s, although the variations are rarely more than 0.2kWh/h and balance out over a few hours.