PV For Space Heating

An Eccentric Anomaly: Ed Davies's Blog

Using photovoltaic panels to generate electricity to heat your house is nutty. Still, as the price of PV steadily drops (mainstream panels are now available at around £1/watt, including VAT) it's instructive to keep an eye on just how nutty it is.

Take a small house with a surface area (roof, walls, floor) of 400 m² with good insulation U-values around 0.1 W/m²·K so with a fabric heat loss of 40 W/K (i.e., losses of 40 watts for every degree Celsius difference between the indoor and outdoor temperature). An internal volume of 120 m³ combined with ventilation at the rate of 0.5 air changes per hour would give an additional loss of 22 W/K but with heat recovery that might be reduced to, say, 5 W/K giving a total loss of 45 W/K.

Using the data from www.degreedays.net for Wick airport (EGPC) using a base temperature of 15.5 °C last winter (2010-12, 2011-01, 2011-02) required just under 1120 heating degree days or 97 MK·s (million kelvin-seconds) of heating.

We can multiply that by the heat loss of 45 W/K to get the total energy required: 4365 MJ (1212.5 kWh).

Those three months contain 90 days. Looking at PVOutput confirms that the equivalent of half an hour's generation per day from panels in Scotland in December might not be a ridiculous average so that's 45 hours or 162 ks (kiloseconds) of production.

Dividing the 4365 MJ of energy needed by the 162 ks available to generate it would indicate we'd need 26.94 kW of panels.

Yup, still nutty.

But, using PV with a heat pump in combination with solar thermal panels might begin to make sense. I've long thought that using PV to drive a heat pump to raise some heat to DHW (domestic hot water) temperatures from lower values only suitable for space heating would be necessary but using the same mechanism for space heating has to be considered, too. More later.