In another thread:

http://www.greenbuildingforum.co.uk/forum114/comments.php?DiscussionID=12791&page=1#Comment_213286

FosterTom wrote: “There must be a crossover, when with increasing insolation ST begins to produce 'something' while PV produces more, and at some point they equalise even for same collector area, before ST races away to produce much more for same collector area. Any 'feel' of where that equalisation point lies?”

This is my page with the closest relevant results: http://edavies.me.uk/2012/01/solar-per-area/ The bottom graph is slightly less horrible to look at.

Taking just about producing heat at 25 °C above ambient the relevant lines are the <s>darker blue</s> green for flat plate and light blue/cyan for evacuated tube. PV is pretty much temperature independent so its single muddy-brown line is the one which applies.

Going on rough insolation read off the graph:

[Edit 2014-11-18: corrected to use the appropriate line for flat plates. More detailed and output-temperature-specific graph added. Oops.]

< 40 W/m²: nothing works.
40 to 75 W/m² the PV produces a trickle of output.
75 to 95 W/m² the PV produces a bit more and the ETs begin to produce a bit, but less than the PV.
95 to 125 W/m² the ETs produce steadily more than the PV.
125 to 145 W/m² the ETs then PV are still winning but the flat plates are beginning to produce.
145 to 185 W/m² the order is now ETs, FP, PV.
> 185 W/m² the order is FP, ET, PV.

Looked at like that it seems that the region where PV wins is very narrow. There are a few things to bear in mind:

1) That's an important and common region at the times of year when the performance of space heating is critical; any old system will do in the summer but if you want to extend well into the heating season then you need to tune things more carefully.

2) This is based on area. Solar thermal is more than twice as expensive per m² compared to PV. When you go on price the break points are different (http://www.edavies.me.uk/2012/01/pv-et-flat/).

3) This is about producing relatively low temperature outputs for space heating. Solar thermal is penalised significantly compared to PV when higher temperatures for DHW are required.

PV does not need risky heat dump systems in the summer when sized for 9 months usage. The excess summer output of PV can also be sold, unlike thermal systems.

Yes, the flexibility of use of PV output is a big point in its favour, as is its robustness.

Posted By: ringiThe excess summer output of PV can also be sold, unlike thermal systems.

That does affect any business case for one or the other.

Oddly enough I bath more in the summer, party because of my job, and partly because I sunbath before work. Not unusual to have 2 or 3 showers a day in the summer.

And some real world graphs.

120 tube evacuated tube solar thermal, 3.8kW PV with SB3800 inverter.

Daily output from 2013 in WHr/sq.m of gross panel area, ordered by PV output, first for every day of the year.

Thanks billt - very interesting. Nice way of showing it - days ordered by PV output. Took me a few minutes to read what you actually wrote properly while wondering if it was cumulative output or something.

The blue “poly” line, that's a polynomial fit to the solar thermal, is it?

Any idea of the temperatures your solar thermal was harvesting at?

In general, quite encouraging as it at least confirms the idea that PV wins on the rubbish days, solar thermal on the better days.

Is there anything between the two in terms of installation and maintenance costs though?

I've been toying with the idea of a DIY solar thermal rig whereas I'd fear it would be more difficult to do PV without paying through the nose on installation.

As for ongoing costs, in the one case you've got the inverter etc. but I suppose there's a pump and so on for thermal (and of course, the fat subsidy)