I think they meant that it was practical to replace half the electricity usage in the average house. Their web page quotes the average electricity usage of 3300kWhr/year and says that a 2kW rated array of panels will give about 1600kWhr/year.

This is reasonable but 3300kWhr/year is nowhere near the average total energy usage per house. It is nearer 8200kWhr/year the extra being mainly gas. It would take a 5 kW system to give 50% of that even on an ideally orientated site. I am not sure what Sharp panels they are using but Sharp's best gives 127W/m² To get 5kW output would need over 39m² of roof area. I doubt the percentage of UK homes with that area facing close to south is very high.

To get the area available on most UK house roofs to supply 50% of the total household energy would require extensive insulation and a ground sourced heat pump to amplify the electrical energy up to 3 or more times that in thermal energy.

Well, if you need heating, why would you try to address that with solar electric panels?  That's just silly, as there are much more efficient methods of using solar energy for heat.  Solar air or water heating can be something like 60% effective, compared to only about 15% effectiveness for electrical panels, and then the gross efficiency loss of converting that back to heat.  

Their site is focusing on electricity which is what the PV panels supply.  I don't understand the logical jump from electricity to energy here, given the acknowledged efficiency limitations of using PV for heating.  

The differences in efficiencies are not quite that big.

Solar water heating is something like 45% efficient despite what some vendors of solar thermal equipment claim. Have a look at the data on the CPC OEM on the Ritter Solar site. These are some of the best vacuum collectors going. Although the top line gives a eta zero value of 66%, that is the efficiency at zero temperature difference and 800W/m² insolation. The coefficients C1 & C2 show how much that degrades with temperature difference and lower insolation. Plug in ambient 15°C  and water 55°C and  700W/m²  and this drops to 60%. In addition the efficiency is quoted relative to the aperture area which is just the blackened area of the tubes despite mirrors that collect sunlight over the gross area. Quote the result relative to the gross area and it drops 48%  Furthermore the heat output is measured at the panel outlet and ignores losses in the pipework, heat exchanger and the  pump and controller energy.  90% efficient heat transfer from the output of the solar panel to the tank would be very good and would drop the overall efficiency of transfer of solar energy hitting the solar panel to the hot water tank to 43%

Solar panels  suitable for domestic use can achieve 18% efficiency at DC panel level (see Sunpower's SPR220 which will still give 17% efficiency at mains AC level with the best of inverters

However the real reason solar thermal cannot be used for heating in the UK for existing buildings is that most of the solar energy falls in the summer while heating is needed in the winter and there is no easy way of storing heat from season to season. This table shows that only 23% of London's solar energy falls in the six months Oct-Mar and 77% in the other six months. There have been schemes using very large underground heat stores and carefully designed passive solar heating buildings have been built but these need to be built from scratch and there is a very slow turnover of the UK housing stock so that such housing will not spread very fast if it is built at all.

However with a grid connected photovoltaic system you can sell excess energy to the power company in summer and buy it back in the winter when you need it. In effect the grid becomes like a vast battery.

If you use a ground sourced heat pump you can multiply up the electrical energy. Gains of three are reasonably obtainable and with a well insulated house and underfloor heating it is possible to heat the house with 35°C water and get an energy gain of 3.5  If the house has a large thermal mass then at the start and end of the heating season it is possible to use mainly off-peak electricity at night and buy it back  cheaper that you sold it for.

Thus 1kWhr of summer sun gives 0.17kWhr of summer electricity the sale of which will pay for 0.15kWhr of winter electricity which will give you 0.45kWhr of winter heat. This is 45% overall efficiency with season to season storage thrown in. It is however not a cheap solution as I know personally.

Nick,

If I would like to heat my house in the winter with solar thermal power, there is a problem like you said. Little sunlight in the winter.

However, most solar thermal installations are rather small. 1-2 m^2.

What about just installing a thermal unit twice its size? The cost of the unit itself (that part at least) would be about 35% of the total, so I could double the capacity for little money.

Hello Richard
The amount of solar energy falling on 1m² of optimally inclined surface in the UK is from 1200 to 900 kWhr/year depending where you are. There is a useful map here. A solar thermal system will convert 40% to 50% of that to useful heat in your system. Say 45% of 1100kWhr equals about 500kWhr/m².year

However only 23% of that solar energy will fall in the winter six months. That is 115kWhr/m² of output spread over 6 months or 0.6kWhr/m² per day on average. This ignores the fact that vacuum tube systems are somewhat less efficient in cold weather and flat plate systems very much less efficient. That is bad enough but averaged over a shorter period it is much worse. I have both a solar thermal system (2.7m² aperture area) and a 17m² photovoltaic array (and a 10.8kW heat pump) Although I  have fitted flow gauges and thermocouples to the solar thermal system I have not yet connected them up. However the relative response of the two systems over time will fairly similar and the photovoltaic system is instrumented. I average about 46kWhr /week of electrical generation but in the first of June this year I generated 100kWhr while in the first week of January this year I generated 4kWhr all week. A solar thermal system would be expected to have the same sort of ratios.

Thus the 500kWhr/m² per year may be an average of  9.6kWhr/m² per week but you must expect 21kWhr/m² in the best week of summer and only 0.83kWhr/m² to last the whole  of the worst week of the winter. It depends on your house but you would be lucky to get through the worst week of winter with less than 200kWhr of heating (1.2kW continuously to heat the whole house and 500kWhr would not keep many less well insulated houses warm for a week in the depths of winter.

You would thus require 12m² of aperture area solar thermal system to make a 5% contribution to your 200kWhr week's heating. (or a 2% contribution to a 500kWhr heating requirement).

Even if you did invest in such a system and you had enough room for it (with a vacuum system this about 24m² gross area) you would find that in summer you would have weeks that gave you 250kWhr and that is more than you are likely to need for domestic hot water. You would have to spill some of that heat in a radiator or some such, wasting the output of your fairly expensive system.

Solar thermal systems in this country are a good way of providing almost all of your domestic hot water in the summer with useful contributions to some other form of heating in spring and autumn and a token contribution in winter. My 2.7m² system does just that and I installed myself for about £1900. Those that sell solar thermal system  for central heating in this country (and I have come across those that do) are plain dishonest.