It's about 8:30am with light cloud.  We're generating 350W at the moment, which is more than we're using.  It was already running around 200W when Richard got up at 7am.

Last night, we hit the point where we were using all we generated around 5:30 (apart from cooking tea which is obviously a higher load) and power generation went to zero around 6pm.

I imagine the real test of the system will be on days with heavy cloud.  And days are much shorter in winter, of course.

We had 13 panels installed, which is the maximum the roof can take.  The economics are very clear on this.  If you're going to have panels at all, you should have as many as you can take.  There's a large fixed cost for things like erecting scaffolding and installing the inverter - to spread the fixed costs over as many panels as possible makes clear sense.  It would not be easy to add more panels at a later date as you'd have to put scaffolding up again.

The interesting question is how much of our electricity usage can be shifted into the daytime.   Washing and vacuuming are obvious ones, but I'm wondering about cooking - which is the really heavy drain on any electricity system (apart from kettles and electric showers).

We don't have an electric shower, but if we did, I'd be showering in the daytime from now on.

I see some possibilities with regard to cooking.  It may be possible to cook things like potatoes earlier in the day and reduce the amount that needs to be cooked in the evening.  (though anyone can save a lot on cooking this kind of thing by using towels on the hay box principle.  Bring your spuds/stew/whatever to the boil and simmer for a few minutes, then turn off the ring, cover the pan/casserole with some thick, fluffy towels and come back when you want to eat it and it will be cooked though, still warm, and never burnt)