Given that sizing of the equipment is the most important issue to get right to maximize efficiency, what are the factors an installer should take into account to determine the correct size? Is it possible for me to calculate it?

For example, this homeguide describes two formula for sizing a furnace -- one based on square footage (understood to be too crude) and one based on heating degree-days:

50 x heating therms x furnace efficiency = load in BTUs/hour

Oddly, this formula suggests the required BTUs/hour increases as furnace efficiency (AFUE) increases. Surely that can't be right, since the furnace size should not need to increase as the furnace gets more efficient.

So if these formula are not correct, what is a good method of calculation?

Correction: In the formula, as furnace efficiency increases, actual heating therms used would decrease. So I'm wrong; the formula does, at least, pass the smell test.

  • 6
    The Air Conditioning Contractors of America Association's (ACCA) Manual J and Manual S procedures are very thorough. If you're looking for maximum efficiency, it's difficult to beat them. – Tester101 Nov 14 '16 at 15:08
  • 2
    Although those do not take into account percieved comfort very well. Worked on a project where the mechanical engineer spec'd the air volumes and unit sizes, but the main entrance had a wall of windows west facing that would make the people feel warmer due to radiant energy from the sun beaming in even though the room would be "correctly" serviced per the calcs. Upsizing the A/C unit servicing that area was against the the results of the calcs but a necessary change with very positive results. Point is, todays calcs do not tell the whole story. – Damon Nov 15 '16 at 6:27
  • 1
    @Tester101 is right though, if you aare not an expert you generally start with something like the manual J or S, then adjust based on experience. Even experts start with those many times. – Damon Nov 15 '16 at 6:33

With a Manual J heat loss calculation. The number depends on your house's size, insulation, and air sealing, among other things. So you (or the installer) need to know those things first. You can do a rough, ballpark calculation yourself at http://loadcalc.net/ (no affiliation). It won't be 100% accurate, but it's a heck of a lot better than using antiquated rules of thumb. Dismiss anyone who tries to use rules of thumb or blindly recommends replacing equipment with something of identical size. They don't know that they don't know what they're doing.

| improve this answer | |

Commercially sizing is done on the basis of 'worst case scenario'. Contractors hate to be called back or to deal with complaints, "The furnace you installed doesn't keep my house warm"

The problem: To be efficient, the unit should run flat out most of the time. There aren't many systems that work well at partial settings.

To have sufficient capacity for the worst case, it has to be drastically oversized most of the time.


You have to run it infrequently, and store the output.

You don't have to take their advice.

Option 1. Add buffering to the system.

An oversize unit wastes energy short cycling. If you have some form of storage, the unit can run for longer periods efficiently, and turn off for a longer period of time.

For hot water systems this is easy to do: You put in a large insulated buffer tank, so when the boiler runs, it has to get several times its own volume hot. If your system will work at the right temp, and you use the boiler to also heat your DHW, you can use your hot water tank as the buffer.

You are still buying a unit that is too big.

For forced air, it's much more difficult. I can come up with schemes that involve running all the air through a room full of water filled pop bottles, but this is either a lot of aggravation or a lot of expense.

Option 2. Use multiple units in parallel.

These are turned on successively as the demand increases.

E.g. If you are using hydronic heating, buy a single high efficiency hot water heater about 1/4 the size that the contractor recommends but leave room to add a second one if one isn't enough. I wish I had done that. I have a 90,000 BTU unit that has never run for more than 3 minutes at a time. I should have gotten a 20,000 BTU hot water heater instead. (My house is 2500 square feet. I live in a 10,000 degree day (F) climate. Worst temp to date is -47 F)

A school where I worked maintenance, replace a single 3 million BTU boiler (hot water and radiator) with 5 150,000 btu boilers. Two were nominally for DHW, and 3 were nominally for radiators (they had different set points.)

So, for example, in the radiator setup one boiler would come on when circulating hot water dropped below 140, #2 would come on if it dropped below 135, and #3 would come on at 130. This reduced on/off cycling by a bunch.

The units were interchangeable, and had quick disconnects on them, plugs fitted to the termostats. The service guys could uncouple one, and take it back to the shop to fix in less time than it took to drive from the shop.

Option 3. Use a modulated fire unit.

This is a unit that doesn't have to be full on or full off. Some have just high/low, some have continuous fire control.

Option 4 Accept being chilly at times.

The heating unit to first approximation will maintain a given temperature differential between inside and outside. If your normal winter is -20 and you want your house at 70 then the unit is maintaining a 90 degree differential.

In addition a house built to current standards (R-2000) has enough thermal mass that it will take about 24 hours to respond to a change in temperature. So an overnight excursion to -40 will likely produce a 10 degree drop in internal temperatures with a unit sized to -20.

Addition tricks you can use under these circumstances: Partially block ducts in seldom used portions of the house. Tune your house to be comfortable in the parts you live in, and occasionally chilly in the other parts.

With natural gas prices right now, the return on investment is lengthy. I use about 70 GJ (~700 therms) a year, which costs me about $600/year. Of that $400 is energy, and the rest fixed fees. We pay 3-7 bucks a GJ. (All currency is canadian)

| improve this answer | |
  • Interesting story, but most of this is irrelevant or inapplicable. – iLikeDirt Jan 28 '17 at 2:46
  • We'll see what others think. I thought that the first two paragraphs addressed the problem with using the standard methods, and the rest gave principles that could be used to deal with the resulting inefficiencies. – Sherwood Botsford Jan 28 '17 at 2:59
  • 1
    Unfortunately, this is probably true a good deal of the time. Instead of properly sizing the equipment, contractors use an inaccurate rule of thumb, and err on the larger side at every opportunity. This results in oversized equipment, which is exactly what the OP is trying to avoid. While this approach may be used in many cases, it's exactly the wrong answer to the question. The correct answer is to use Manual J as Tester and iLikeDirt suggest. – Mark Jan 28 '17 at 4:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.