-
-
Notifications
You must be signed in to change notification settings - Fork 105
Description
I noticed #3544 and #3545 have been opened recently on load predictions but I think my issue is different but happy to move this to one of those if you wish.
This has been happening on and off since I started using Load ML a month or so ago. However, I have been having a few inverter issues so I wanted to make sure I have at least a week of clean load history before reporting.
This is my plan from about 15 minutes ago...
As you can see, around midday it has some ridiculously high load numbers that in my system aren't even physically possible. The rest of the day the numbers are too low but not so much to make the plan unworkable.
I recently turned on the compare feature to try and decide whether to switch to agile outgoing now the fixed export rates on Octopus have dropped. Here is the plan from the compare tab for my current tariff for the same time period...
I realise the compare run was done at minight and the current plan about eight hours later but how can the predictions be so wildly different using the same source data?