Replies: 7 comments 2 replies
-
I should add that my Libbi automation also kicked in and started to boost charge from grid, however it turned off as soon as the Dispatch sensor turned off. This automation does not depend on a target sensor, just on the dispatch sensor. In any case I had 4 minutes of boost at the high rate and would also like to avoid this from occuring. |
Beta Was this translation helpful? Give feedback.
-
This is unfortunately the consequence of OE only providing polling for this data, and would probably have been an issue if you configured everything manually based on the schedule you originally had. Maybe this is there way to stop people using the cheap data for anything but your car. If you know when your car is plugged in etc with your Libby sensor, you could take advantage of the manual polling service - https://bottlecapdave.github.io/HomeAssistant-OctopusEnergy/services/#octopus_energyrefresh_intelligent_dispatches. With this method, you could call the service just before your target sensor comes on to force a refresh which should then be re-evaluated. You could also set an offset on the target sensor so it turns on a couple of minutes after the period starts where the dispatching sensor should have updated, and so you can then check the dispatching sensor as well. You'd need to check the end time of the sensor as well as the target sensor to determine when it should turn off, as the offset shifts the entire time period not just the start time. |
Beta Was this translation helpful? Give feedback.
-
Thanks for that @BottlecapDave . It looks like I would have to control intelligent refreshes completely with this method so something I need to get comfortable with.
How do I do that? At present my automation is triggered by the state of target sensor, I'm unsure how I can trigger the service just before the target sensor comes on? I guess an option would be a time-based trigger based on the attributes of the target sensor - "next_time:" maybe? |
Beta Was this translation helpful? Give feedback.
-
Thanks - something for me to try over the next few days. Also I'll continue to monitor OE's cheap slot changes - maybe what I saw was a one-off. My worry with going down this rabbit hole is that OE continue to evolve the way they allocate slots so I could end up piling workarounds on top of workarounds over time. If this is is a <5% scenario I might just live with it... |
Beta Was this translation helpful? Give feedback.
-
Just as an update, I tried a simpler method to check if the cheap slot really exists (thanks for the suggestion @BottlecapDave ). Using 2 triggers, one based on off-peak sensor and the other on the dispatching sensor but with a 5 minute delay. The 5 minute delay should allow the Octopus integration (3 minute refresh as default) to update the dispatching sensor in case the switch was made last minute.
Then a condition for the action that performs an AND against the trigger id and the dispatching sensor. i.e. no need to honour the 5-minute delay if it is in the off-peak period.
So far so good... |
Beta Was this translation helpful? Give feedback.
-
Hi @nickwaring just to try to address your queries. Yes they are part of the same automation - I did not add the code for the action in the second snippet, I only showed that I performed an additional check : if triggered "on" and off_peak sensor is on then run the action straight away - i.e. I do not need the 5 minute delay which is there in case Octopus change the planned slot, which is monitored via the dispatch sensor. My premise here is that the dispatch sensor is always at risk, however if the off_peak sensor is on then the dispatch can be ignored. The action would be a switch action against entity switch.myenergi_libbi_charge_from_grid. I don't know if this helps - I'm no coder so I would not use my example as good practice - I'm sure someone with better skills could make it simpler and more robust, however it works for me. |
Beta Was this translation helpful? Give feedback.
-
thanks @ragg987 - really appreciate you taking the time to reply. I'll have a go and post results later :-) |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
I'm on IOG and in the last few weeks Octopus have become increasingly "aggressive" in the way the allocate and continuously re-allocate cheap slots. I would appreciate any input on how I can handle this situation.
Example, tonight they offered a cheap slot at 22:00 to 22:30. I saw it on the Octopus app up until around 22:00 and then they pulled it and turned that into an expensive slot.
My target sensor for heating (heat-pump) is set with the option "Existing target rates haven't started or finished". As I understand it this option will re-evaluate the slot in the minute prior to planned start to make sure it is still the optimal slot, however in tonight's example I believe the change from Octopus came too late. So my heat-pump started in the (now) expensive period at 22:00 and the target sensor did not re-evaluate as the target period had started.
Would appreciate any tips into how I can improve the robustness of this so that the target sensor and hence the heat-pump do not react until the slot is confirmed by Octopus.
I had thought that bringing in the Dispatch sensor might help, however if I look at the sensor history it looks like the sensor also got fooled by this last minute switch from Octopus.
Another option I thought of was to see if my Zappi has been set to boosting by Octopus, but in this instance I notice that the sensor was stuck on boosting for some 24hrs so that would not have helped either. I think it got stuck on boosting as my EV was full at that point, however Octopus did not cancel the unused boost.
Beta Was this translation helpful? Give feedback.
All reactions