<sorry @Barbara, was not able to add the beta tag?>
In beta VRM the forecast contains the heatpump element but the history seems to change its mind often. And even for the first screenshot, the numbers don’t add up. Quite some hours are missing.
Screenshot from around noon:
Later this afternoon (after having refreshed, yesterday-today’ed etc etc):
Actual recorded consumption by the meter and the proper registration in VRM:
Not sure if it’s related but I seem to have had a problematic Shelly migrations from acload to heatpump causing a ghost device in VRM.
With ghost I mean I do see them on the Advanced page:
But I cannot remove these devices from the Device list page as they simply do not exist (nor do they exist in the VRM id section of Venus):
Please note that the same question has been asked in a Venus beta section for Shelly:
Shelly naming and VRM issues problem #4
@Barbara Not sure but if you have a moment, I’m fairly sure it’s not a generic issue. I added a new Shelly heatpump device this week and for that device the past consumption is tracking fine.
It therefore seems highly likely that only my other 3 shelly that have moved from acload to heatpump have this problem. Is there anything Victron can do to cleanup my stale/ghost devices and repair this in VRM ? On Venus it’s all working fine.
Hi @kudos50 , you can delete the stale devices yourself from the device list page.
Unfortunately we have a policy to not clean up / delete database entries.
Thanks for responding @Barbara but that’s the problem. I cannot delete the two devices from the device list page as they do not exist. They do not exist in the device list page and they do exist in the advanced page for statistics.
In the screen shots, please find the red 7 and 11 days last reported device. Scrol down and look at the device list. Both do not exist. They “counterparts” for heatpump with new VRM ids do exist.
Oh now I see, actually this is a feature: for as long as we have your historic data for removed devices, they will show on the advanced page. Let’s say you replaced an energy meter, but at the end of the year you still want to see the total energy consumed, you can plot both in a graph to get that visibility.
They will automatically hide from the advanced page as soon as we do not have data stored for them anymore, which is 6 months after their last data update.
Would you have your site ID, so I can check why the heatpump recorded consumption on the dashboard adjusts after the hour has passed?
I never noticed that… should have payed more attention, it’s 290452. Thanks.
To add to the confusion. The 51 and 52 heatpump devices that were previously known as 51 and 52 acload devices never showed up in the device list EVEN when they were still acload. 50 did not have this problem and behaved as expected.
Please note that the consumption disappearing problem is for all three shelly devices that have migrated from ac load to heatpump (50, 51 and 52). The new device that was added last week seems to record just fine (the Nuos heatpump boiler, 53))
Hi @Barbara, did you find some time to investigate the issue ?
Somehow it seems to have been a mix and merge of problems. After the update yesterday, the old disconnected ac load VRM ids for the “heatpumps that changed type from ac load to heatpump” finally appeared in the device list in VRM with a last seen ages ago. I was able to delete them.
Not sure if this will also fix the missing consumption data for the 3 heat pumps. Probably not. The most recent example is yesterday. The ATAG heatpump had about 6kW of consumption over night (± 800W) according to the VRM advanced tile. But during those hours, the VRM graph is not showing it as heatpump.
Hi @Barbara it’s been a while since you requested my site id. Have you had chance to look at this ? Problem is still there and somehow I think the forecast is taking all heat pumps into account whereas the consumption results only seem to show the single (1 out of 4) heat pump devices.
Other than that, really like the new ability to change EV charger forecast. Fixes the system from charging the EV from battery due to incorrect forecast 
It would be even better if you can change/edit the current hour as well. If forecast is wrong, worse case, the EV is charging from battery for a full hour before your first “corrected forecast” starts.
