display the forecast production for the day

It's nice to have the forecast graphs for the day to see when it's likely to be sunny, but it would actually be nice to see a figure for say "Solcast Forecast Today: 14.52kWh"

I'm pretty sure this is coming in an update, it was a feature request I submitted not long after the Solcast card first arrived.

Agreed - at the moment my question is "is there going to be enough sunshine to charge my battery" (or "how much solar will I be able to put into the battery"), and whilst I can eyeball the area under the curve, having an actual value would make it much easier to work out if / how much to charge overnight.

would that number be what you really need?

For me its the difference between home base load and PV generation, which is the 'spare' power. Thats what energy is availalble to charge the battery, Having say 500w all day would result in zero charge in my battery even though the total PV may be 4kW. No PV all day apart from 3kW for an hour md day may result in 2kWh charge to the battery. So less PV for the day but more charging.
I succesfully use this algorithm in my own charging automation to optimise overnight charging.

I would propose not only showing the total PV forecast for the day, but also the forecasted "Spare"

Since our panels were installed in mid-November, we've been curious just how good generation will be in the Spring/Summer, but given we're already making around 10kWh per day, it's looking positive. We've currently got one 8.2kWh battery, but will have a second installed later this month. Although the cost of filling up the existing battery overnight is negligible (at least on our current Octopus Go tariff), I'm keen to optimise the system around maximising the use of the harvested energy.

I've been looking at the Solcast API, with a view to doing precisely what you have; having an automated peek at the forecast for the next day to determine how much the battery should be charged. The battery should be sufficiently charged to get us through to when the panels start producing again, but leave as much room to store the daytime sun (less the expected daytime usage) as possible.

Looking at what the API returns, I think GivEnergy is using the pv_estimate attribute, rather than the pv_estimate10/pv_estimate90 attributes (forecasts endpoint docs: https://docs.solcast.com.au/#forecasts-rooftop-site).

{"pv_estimate":0.5344,"pv_estimate10":0.0708,"pv_estimate90":2.3854,"period_end":"2022-01-18T10:30:00.0000000Z","period":"PT30M"},{"pv_estimate":0.6667,"pv_estimate10":0.0845,"pv_estimate90":2.4845,"period_end":"2022-01-18T11:00:00.0000000Z","period":"PT30M"},

Also, are the actual observed PV readings from the inverter being sent back to Solcast for tuning the site (https://docs.solcast.com.au/#estimated-actuals-rooftop-site)? (my 'Estimated Actual PV' and 'Actual PV' graphs line up very well, so I suspect this is already being done.)

adamdoran the Solcast tuning option is defunct sadly, details on their site.

The base Solcast forecast is very good normally, it's one of the reasons they dumped the tuning. I know mine was very close when I don't get the shading problem I mentioned at the start of this thread.

You must have a good sized array to get 10kwh, is that a sunshine day or cloudy?

Sorry for the delayed reply.

Thankfully no shading here, and I quite agree that the estimates are very good already, so additional optimisation could be overkill.

We've been managing 10kWh even on partially cloudy days. We have a 5.475kWp array (15 x 365w JA Solar), feeding the Hybrid 3.6. Our installer told us the array being slightly "over provisioned" should result in a smoother generation curve throughout the year.

This week has been pretty disappointing though, so fingers crossed the sun makes a return soon!

adamdoran be interesting to see if your array is 'clipped' by the inverter over summer resulting in some lost power.

My array is 1kw larger than the inverter can manage but 500w under the maximum it can handle. I believe some of that excess can be fed to the DC aspect of the battery system though (might be wrong there).

As there are occasions where the UK sees solar radiance levels above the baseline used to calculate panel performance (1000 something's per m2) which might make it worse - that said panel efficiency drops with increases in temperature too.

Just found an interesting site for sun position depending on time of year and time of day. https://www.suncalc.org/
Has about 4 mapping options including satellite.
This seemed an appropriate place for this!