can you try again - i fixed a couple things!
My fork should appear in the HACS store along side Nathan's original.
i just downloaded it fine from the hacs store.
It Worked!
Thank you so much for your hard efforts! It's great to finally have real data without relying on the cloud.
I've connected my Evo 10 with a LAN cable (for those of you who are interested).
Thank you again!
I've connected my Evo 10 with a LAN cable (for those of you who are interested).
Thank you again!
no problem.
if you do see bugs, register them on github and i'll endeavour to work on them
if you do see bugs, register them on github and i'll endeavour to work on them
Thank you Adam, your post inspired me to add two more automations.AdamNewberry wrote: ↑Tue Jan 13, 2026 8:32 pm Managed to get a full control loop going now!
Checks solar forecast tomorrow - works out a reasonable Max SOC at 7pm and writes it to the register.
Each day at 05:30 it will discharge if the current predicted solar output is above 12 kWh (I can adjust this later once I get used to it) and then it will force discharge until 06:30, when another automation switches it to Feed-in.
This is so I can mop up the solar above my 3.9kW export limit, as I have 6.16kWp and clipping occurs.
And it shouldn't work if the day is dull, so I keep my battery SoC higher.
Yeah via HA, and built in solar forecast tool
I will see in the morning if it works, but on paper it does at least.
Forecast is for almost 20kWh, and I set the threshold to work above 12kWh.
I will be back crying if it doesn't work
Forecast is for almost 20kWh, and I set the threshold to work above 12kWh.
I will be back crying if it doesn't work
I signed up for solcast. It’s free for single residential owners. I may have to compare it to the stock solar forecast tool!
Much more accurate from what I've seen. In particular it also gives figures for the 90% percentile and 10% percentile which gives a good grasp of the variability in the forecast.AdamNewberry wrote: ↑Tue Feb 24, 2026 6:22 pm ... I may have to compare it to the stock solar forecast tool!
The built in HA solar forecast said 9.3kWh
Solcast said for today 13.4kWh
Inverter reported 12.9kWh
I will need more long term data, but have to see how it pans out. The built in tool doesn't seem that accurate for today, but it's only 1 set of data points.
Solcast said for today 13.4kWh
Inverter reported 12.9kWh
I will need more long term data, but have to see how it pans out. The built in tool doesn't seem that accurate for today, but it's only 1 set of data points.
I have node red fill a csv file each day after sunset that has actual and predicted on it. I just can’t find it on my HA VM 
Very envious of you all, especially when the sun is out!
Our install date came and went after our installer got the response from National Grid stating they were prohibited to complete the install as "this customer is at the end of a long run of LV network, the volt rise figures fall outside of tolerance for the install and export limits.".
I find this incredible and find it hard to understand the apparent lack of planning and want to futureproof the network by National Grid. I presume it also means that no one in the village will be permitted to install a PV array at this point, as most if not all will be fed from the same "long run LV network".
National Grid have until the end of March to design a solution but, I'm led to believe they need to find and procure a site for a new sub-station, bulid said sub-station and perform all the plumbing before this situation is likely to be resolved.
I was so incredulous that I've raised with our local MP, in the blind hope that they can throw the governments weight behind speeding up the solution.
Paul
Our install date came and went after our installer got the response from National Grid stating they were prohibited to complete the install as "this customer is at the end of a long run of LV network, the volt rise figures fall outside of tolerance for the install and export limits.".
I find this incredible and find it hard to understand the apparent lack of planning and want to futureproof the network by National Grid. I presume it also means that no one in the village will be permitted to install a PV array at this point, as most if not all will be fed from the same "long run LV network".
National Grid have until the end of March to design a solution but, I'm led to believe they need to find and procure a site for a new sub-station, bulid said sub-station and perform all the plumbing before this situation is likely to be resolved.
I was so incredulous that I've raised with our local MP, in the blind hope that they can throw the governments weight behind speeding up the solution.
Paul
It is frustrating, Paul. I work in renewables (mostly commercial and public sector projects).
Despite the grid being obliged to deliver grid infrastructure upgrades, they are extremely slow and end user demand is way ahead of any planned upgrades.
On commercial projects the DNO use projects to pay for their required upgrades. I have killed projects due to grid capacity and DNO killing the financials with 500k worth of non contestable works.
Despite the grid being obliged to deliver grid infrastructure upgrades, they are extremely slow and end user demand is way ahead of any planned upgrades.
On commercial projects the DNO use projects to pay for their required upgrades. I have killed projects due to grid capacity and DNO killing the financials with 500k worth of non contestable works.
I cannot wrap my head around this logic at all. The export I can understand them saying no to, but the Import side of it makes little sense.youngsyp wrote: ↑Fri Mar 06, 2026 10:57 am Very envious of you all, especially when the sun is out!
Our install date came and went after our installer got the response from National Grid stating they were prohibited to complete the install as "this customer is at the end of a long run of LV network, the volt rise figures fall outside of tolerance for the install and export limits.".
If they have given you an 80Amp fused supply, you can import whatever you like up to that limit no?
I have heard of export limits being set to 0 on a G99 application.
Thanks Adam, that's useful information.
To add to my frustration, I have at least 3 neighbours who have installed a PV array in the last couple of years or more recently. So I kick myself for not doing this sooner.
Paul
To add to my frustration, I have at least 3 neighbours who have installed a PV array in the last couple of years or more recently. So I kick myself for not doing this sooner.
Paul
Yes, that was essentially an argument I tried to make. I said to the installed, "just turn off export". Export isn't a factor in my decision to go PV so I couldn't careless for that aspect. A further frustration.MaterialBarracuda48 wrote: ↑Fri Mar 06, 2026 11:07 amI cannot wrap my head around this logic at all. The export I can understand them saying no to, but the Import side of it makes little sense.youngsyp wrote: ↑Fri Mar 06, 2026 10:57 am Very envious of you all, especially when the sun is out!
Our install date came and went after our installer got the response from National Grid stating they were prohibited to complete the install as "this customer is at the end of a long run of LV network, the volt rise figures fall outside of tolerance for the install and export limits.".
If they have given you an 80Amp fused supply, you can import whatever you like up to that limit no?
I have heard of export limits being set to 0 on a G99 application.
Paul
I might be wrong, but I thought if you fitted a 3.68kW inverter that you notify your grid supplier afterwards via G98 notification? That would be without PV, just battery storage for buying in cheaper energy.
Either way, it's total madness.
Possibly. We're looking to install a 10kW inverter.MaterialBarracuda48 wrote: ↑Fri Mar 06, 2026 11:16 amI might be wrong, but I thought if you fitted a 3.68kW inverter that you notify your grid supplier afterwards via G98 notification? That would be without PV, just battery storage for buying in cheaper energy.
Either way, it's total madness.
The installer asked NG whether they could just reduce the export rate, as they've done that previously. But, it was a flat no install whatsoever from NG.
Paul
It's probably a wait and see for you, seems a waste of money to install a 3.68kW and then swap later to a 10kW (and still face a possible rejection)
My 'fun' with National Grid was odd to me, G99 submitted with an EP11 battery, I realised that there was a heated choice that just came out, so they had no choice to but to submit G99 again. 1st one I had unlimited export, days later it was 3.9kW export for just a swapped battery
Boggles the mind.MaterialBarracuda48 wrote: ↑Fri Mar 06, 2026 11:25 amIt's probably a wait and see for you, seems a waste of money to install a 3.68kW and then swap later to a 10kW (and still face a possible rejection)
My 'fun' with National Grid was odd to me, G99 submitted with an EP11 battery, I realised that there was a heated choice that just came out, so they had no choice to but to submit G99 again. 1st one I had unlimited export, days later it was 3.9kW export for just a swapped battery![]()
When I think about it, I don't think we'd ever export anyway. Checking our consumption to this point today and we're already at 7.25kWh. Yesterday must have been a quiet day, as we only went through 12.08kWh.
Paul
That's very short sighted by your distributor, because distributed battery storage reduces the need to upgrade infrastructure especially when you factor in that unlike solar only systems, batteries can export to the grid when it's needed most.
The primary reason we wanted a battery is due to frequent outages because we're toward the end of a 30+ km low voltage transmission line with no redundancy. So when summer storms blow past the grid can go down for hours at a time, and at least once not that long ago it went down for more than 24hrs. For most of that we had lots of sunshine but no grid so no solar power either.
Here's the unexpected upside though, after decades of declines in solar feed in tariffs, where some people are no longer compensated for any solar production, it's lucrative enough to export in the evening instead.
Our distributor now seems to want more battery systems by rebating power companies 11.1c/kWh for all exports between 5pm and 8pm, at the same time they've introduced a 2 tiered system where customers are allowed to export up to 7.5kWh/day between 10am and 3pm and then the distributor charges power companies 0.7c/kWh above that to supposedly encourage more self use and they now tell people to point their panels west.
Australia got here because solar systems were seen as a way to lower power bills, but because it was too successful especially in milder spring/autumn months without demand for it, it makes the wholesale price go negative and the energy regulator is worried the grid could even become unstable.
Batteries are great at soaking up all the excess solar and then sending it back to the grid when desperately needed.
I built an app that does some analysis on forecast accuracy for me... only got a months worth more or less, but its interesting data. my system is consistently over producing vs forecast from Solcast!MaterialBarracuda48 wrote: ↑Tue Feb 24, 2026 7:29 pm The built in HA solar forecast said 9.3kWh
Solcast said for today 13.4kWh
Inverter reported 12.9kWh
I will need more long term data, but have to see how it pans out. The built in tool doesn't seem that accurate for today, but it's only 1 set of data points.
whacked in my claude API key to see if it could give some recommendations too!
Attachments:
I haven't looked at my forecast tool built into HA too much, I just needed a rough figure to program a discharge if it thinks there's going to be enough solar. I do this as my export limit is 3.9kW on 6.16kWp, and I clip in Summer naturally.AdamNewberry wrote: ↑Sat Mar 07, 2026 11:35 am I built an app that does some analysis on forecast accuracy for me... only got a months worth more or less, but its interesting data. my system is consistently over producing vs forecast from Solcast!
whacked in my claude API key to see if it could give some recommendations too!
I have it dump power if it forecasts above 12kWh for the day, seems to work and I can adjust this figure up and down.
That does look mighty good, and even more impressive is getting better than expect performance.
Took a few tweaks before I got solcast forecasts that actually made sense, it didn't help that solcast and forecast.solar use different ways of expressing azimuth.AdamNewberry wrote: ↑Sat Mar 07, 2026 11:35 am my system is consistently over producing vs forecast from Solcast!
In terms of usability the mind boggles why can't we just enter a degree from true north using a compass app on our phones.
The trick I found to both is fresh data, things really can change within half an hour.
Also solcast blindly assumes 20% degradation, unless you enter an install date or similar, and forecast.solar assumes no degradation.