Heat load calculation from gas bill data
Distilling information from various sources here, with my head spinning, I would appreciate confirmation that the following formula is correct. I have three years of gas usage, and the company was kind enough to include heating degree days in the billing report.
There are 1,037,000 BTU in a gas MCF.
Thus gas usage (MCF) x 1,037,000 = monthly usage in BTU.
Divide the result by total monthly HDD to get BTU per HDD.
Divide the result by 24 to get BTU per degree hour.
Multiply by delta T at design temperature to get theoretical heating load in BTU/hr at design temp.
Thinking ahead to electrification, I presume that the old boiler probably sent at least 20% of those BTUs up the chimney, so is it reasonable to peg the actual heating load at about 80% of the above gas calculation?
What are other factors that may have non-trivial effects on the real-world usage?
Thanks so much! Appreciate the wisdom here!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Your math looks right.
One thing that the formula fails to take into account is solar gain. If your house is absorbing significant amounts of solar warming the actual usage would be less than the HDD model would suggest. So a calculation based on actual usage will underestimate your actual BTU/HDD.
And it doesn't average out. The design load is based on maximum heat demand, not average.
I don't know any way to compensate for it, but it's helpful to know that your actual load is going to be somewhat larger that what the usage suggests.
Thanks, DC. Didn't thiink about solar gain, which for this house will be somewhat of a factor. However, since my goal is to size new equipment for the same space, whatever amount is contributed by solar gain will remain constant, so I'm fine with NOT factoring it in.
This is what I was trying to get to in the third paragraph. It doesn't average out.
You're sizing the system for the coldest hours of the year, which is going to be 3am in January on a full moon. There will be no solar gain at that time. The fact that your heating is sometimes offset on sunny afternoons by solar gain does you absolutely no good in the middle of the night.
Oh, I think I get it. My calculation starts with some broad numbers, produces some averages, and then extrapolates from those averages into an extreme. So my calculated extreme will not be accurate given that at least one embedded factor--solar gain--is not as constant as the calculation makes it and in fact skews my final load number somewhat lower than it actually is. "somewhat".
Right. And the hard part is estimating what that factor is.
How much solar gain can a house in Ohio reasonably expect in January or December? You could use months with shorter daylight and that’s probably fine.
Cleveland averages just over 2 kWh/m2/day in January.
Let's say a house has a footprint of 1000 square feet, that's about 100 m2, so 200 kWh/day or about 680,000 BTU/day. The question is, what percent of that goes into the house? Probably not much -- a highly efficient solar collector collects around 20% of the energy hitting it. If you assume it only comes through the windows, you could estimate the area of south-facing windows and multiply by their solar gain coefficient. Just for fun let's say it's 2%. That's about 13,500 BTU/day, if you assume the sun is out for 8 hours a day that's under 2,000 BTU/hour for those hours.
I think that's a fair estimate of the magnitude.
In my corner of Ohio, we rarely get 8 hours of sunlight on ANY January day. My estimate for solar gain on the south facing windows would be that it eases boiler work for a few hours a few times in a month. Non-trivial, on the order of a few percentage points at most, and given the confidence of my final extrapolated calculation, I will choose not to worry about it. Other factors: water heater (mentioned above), adult children home for pandemic, gas stove, longstanding marital disagreements over setpoint, climate change. Only the first and last are non-trivial but even those I will still choose to ignore.
The 2 kWh/m2/day takes all that into effect. It's the equivalent of about two hours of sunshine on an average day.
"Thinking ahead to electrification, I presume that the old boiler probably sent at least 20% of those BTUs up the chimney, so is it reasonable to peg the actual heating load at about 80% of the above gas calculation?"
Yes. It might be less than 80% if the boiler is very oversized.
Some portion of the MCF will go to domestic hot water if you have an indirect or direct gas water heater.
Unless solar gain is going to change because the heat pumps will block the windows, it is not really relevant.
BTU 'use' is a theoretically going to be the same
How much of a furnaces BTU's get used is the question
An 80 percent furnace can be very much less than that in the shoulder seasons. A direct vent with outside air less so
It really is about estimating cost, and that is not going to be that accurate
Sizing is, as mentioned about the coldest night of the year
Heat pumps frequently end up oversized when you start putting heads in bedrooms anyway
I've already said this twice in the thread, but I'm going to say it a third time because it's important.
If what you were interested in was projecting seasonal energy usage, you'd be absolutely right. But we're talking about sizing equipment.
When you're sizing heating equipment, you're sizing it for the maximum load. Which is almost certainly going to be the middle of the night when there's no sun. Calculating using the heating bill is taking the average and extrapolating to estimate the maximum. That extrapolation is not valid, because the average includes a lot of time when the sun is shining.
It's like one of those high school math problems: "A house uses 500 MCF of gas in a month. How long does the boiler run in an hour if it's 6 degrees outside?" The right answer is we don't have enough information, and have to make a bunch of assumptions and introduce some outside data to help. The debate about how to best do that is what makes GBA so useful.
Yeah I wouldn’t stress about the solar gain. If the indoor temp is 1 degree under set point once per year while you’re asleep anyway, who cares! Share your data if you want, we’ll be happy to help!
I would say it's useful for setting a lower bound and getting an approximate value.