It is one of the laws of the universe that, generally, a turbine

thermodynamic cycle is only about 36% efficient.

First item, I don't think we'll be spinning a furnace fan at 10,000

rpm, so we're not talking about turbines here.

More like a piston.

Second thing, we're probably talking very low displacement. We're

only trying to generate 1/4 to 1/2 horsepower.

1 HP = 750 watts, so 1/3 hp = 243 watts

1 BTU per hour = 0.3 watts, 243 watts = 810 btu's per hour

Assuming a 15% efficiency, then we need 5400 btu's per hour. Natural

gas has about 1200 btu's per cubic foot. So we need 4.5 cubic feed of

NG per hour.

I'm currently paying $2.45 (CDN) per cubic meter (which includes a

variety of taxes and charges besides the actual cost for the natural

gas). $2.45 CDN = $2.07 US = $1.05 GBP.

1 cubic meter = 34.3 cubic feet = $2.45.

4.5 cubic feet = 0.131 cubic meters = $0.32 per hour.

Since most small electric motors are typically 50% efficient, our

original 243 watts needs to double to 500 watts for the actual

electrical energy required. That is actually a bit on the low side -

this web site:

http://www.dbkpowergenerators.com/Sizing Generators.htm
puts the actual wattage needed for 1/3 hp to be more than 700 watts.

So I'll assumne 600 watts.

I'm paying 11.2 cents (CDN) per kwatt-hour (all taxes and other

charges included). For comparison, 11.2 (CDN) = 4.8 (pence?) = 9.5

cents (US).

600 watts = 0.6 kwatt = 6.72 cents per hour.

So, a small NG-powered motor with a theoretical efficiency of 15%

would cost 32 cents per hour to operate, vs 6.7 cents per hour for a

comparible electric motor.

But if the waste heat from the motor were input into a heat exchanger,

perhaps as a return-air pre-heater for the furnace, then the cost of

operation might make this more attractive. Such a motor with

waste-heat capture could be used as the sole heat source during the

low-demand months of the heating season, with the main furnace burners

operating during the middle of the season.