Should I make measurements during a cold snap to better size my boiler?

MikeInMassachusetts

I've got a 2800 square foot house and a 175,000 BTU gas boiler (a Triangle Tube Prestige Solo 175). I'm guessing it's probably oversized, especially after I replaced some drafty old windows.

Where I live in Massachusetts, it gets down to minus 10 degrees Fahrenheit only about once per winter, and that one time seems like it's coming up next weekend.

On the boiler screen, I can see when the boiler is firing, when it turns off, and the temperature of the water entering and leaving the boiler. I also have an outdoor thermometer and of course I can see which of the three zones is calling for heat.

Is this kind of cold snap the best time to record that kind of information, to help determine how big a boiler for this house should really be? Or does that kind of thing not really need to be done during a cold snap?

Thanks for any suggestions!

STEVEusaPA

It's up to you. I'm always listening to my boiler at/below design temp, hoping it runs continuously, but it doesn't...lol
Design temp helps you dial in your outdoor reset.
As far as sizing, you can do that any time if you know the amount of elapsed degree days, boiler input/efficiency/output.

Hot_water_fan

https://www.greenbuildingadvisor.com/article/replacing-a-furnace-or-boiler

Basically describing this method - give it a go!

pecmsg

Accurate data is always helpful.

With that said if that boiler cycles at all during the cold snap its oversized. By how much is to be determined.

EBEBRATT-Ed

@MikeInMassachusetts

Just so you know, you don't size a boiler based on the coldest temp you will ever see. If you do it will be oversized 99%of the time. See the attached You use the 99% heating outside design temp. Mass is on pg 44 I think.

Jamie Hall

That's a good reference -- but needs to be sort of looked at a bit sideways. In some areas the selected weather station is a good reference. In some other areas it can be as much as 10 degrees off, particularly in hilly or mountainous areas.

STEVEusaPA

Hot_water_fan said:

https://www.greenbuildingadvisor.com/article/replacing-a-furnace-or-boiler

Basically describing this method - give it a go!

Wonder where they stole that from?
https://forum.heatinghelp.com/discussion/187373/heat-loss-calculation-accuracy

EBEBRATT-Ed

The boiler selection seldom exactly hits the heat loss so you usually end up with a boiler that's a little larger which gives you a little fudge.

Plus, we could debate the piping and pick up all day. (1.15 water) (1.33 steam) so there is always that.

My theory is in the spring and fall and the boiler sits for maybe hours between calls for heat you need some pickup but the boiler is oversized for the load then anyhow. Although on steam the load is about the same at all times

When the outdoor temp is at design the piping will usually be hot or at least warm so there is little need for
pick up so now it is available to help with any minor mistakes in the heat loss

Hot_water_fan

You can use other base temps for HDD. All you have to do is select a different number from a list on degreedays.net. Subtracting summer gas usage can roughly approximate DHW 

exqheat

You can also consider indoor reset for precise adjustment of source production of BTUs

Jamie Hall

A couple of thoughts. First, heating degree days is a good way of getting a handle on how much energy you will need for a given heating season, or period within a season. It will tell you exactly nothing about how big a boiler you need. Useless for that.

Now... back to how big a boiler. The question really should be -- how to design the whole system. However, the boiler power is part of the question, so let's contemplate that.

One needs two numbers. First, what is the heat loss rate of the structure at various temperatures and design conditions (particularly wind) and, second, what is the condition for which one wants to design?

There are several approaches to the question of heat loss rate, and there comparing the energy usage to the heating degree days is one quite valid approach. Properly flipping the numbers around, one can get a pretty close estimate on the relationship between exterior conditions and power required under those conditions, provided one uses averages over a long enough period of time (a few years, for residential structures, is probably long enough).

One can also use a Manual J or similar calculation and arrive at a different, but probably closely similar, relationship.

OK. Got that far, Now. What is the design condition? And this is where I get puzzled. Some would say to use the 99% design day low temperature. Some would say 95%, since it doesn't get down to the other temperature that often. But what is the rationale around deliberately designing something which you know will not do the job (in engineering terms, "fail") at least once, every year (using the 95% figure)(statistics can be weird -- the probability of there being at least one day in any heating season with the temperature below the 95% figure is 0.995 -- about as close to a dead cert. as one can get)? What other engineering discipline does that, except when the cost/benefit ratio rises dramatically (the exception is flood protection, where increasing the flood protection from a 100 year to 500 year return can be eyewatering). Do you design a beam to hold you up 95% of the time? The brakes on your car to stop you 95% of the time? Your airplane to fly safely 95% of the time? I certainly hope not. The additional cost to go from the 95% design day to the 99% day is not large. So -- what's the excuse?

Hot_water_fan

OK. Got that far, Now. What is the design condition? And this is where I get puzzled. Some would say to use the 99% design day low temperature. Some would say 95%, since it doesn't get down to the other temperature that often. But what is the rationale around deliberately designing something which you know will not do the job (in engineering terms, "fail") at least once, every year (using the 95% figure)(statistics can be weird -- the probability of there being at least one day in any heating season with the temperature below the 95% figure is 0.995 -- about as close to a dead cert. as one can get)? What other engineering discipline does that, except when the cost/benefit ratio rises dramatically (the exception is flood protection, where increasing the flood protection from a 100 year to 500 year return can be eyewatering). Do you design a beam to hold you up 95% of the time? The brakes on your car to stop you 95% of the time? Your airplane to fly safely 95% of the time? I certainly hope not. The additional cost to go from the 95% design day to the 99% day is not large. So -- what's the excuse?

I've never seen anyone advocate for the 95% temperature - 99% vs. 99.5% is usually the debate in my experience. I agree, 5% is too common. I think we all do!

mattmia2

Where is wind factored in? It may be that the chances of the worst wind conditions and the 99% temp at the same time are very small. (Although I suspect in most cases infiltration is the biggest guess of the inputs to the equation). Here in Ann Arbor when it was -3 and very windy for a couple days I suspect the load was bigger than a couple years ago when it was -15 for a couple days. Then there is the available sizes of equipment that almost always puts you 20%+ over the calculation.

STEVEusaPA

Jamie Hall said:

A couple of thoughts. First, heating degree days is a good way of getting a handle on how much energy you will need for a given heating season, or period within a season. It will tell you exactly nothing about how big a boiler you need. Useless for that.

I disagree and I, and others, have been using this method for a little while now with success.

https://forum.heatinghelp.com/discussion/187373/heat-loss-calculation-accuracy

Jamie Hall

So you have, @STEVEusaPA , and while I can see without difficulty how you could derive a usable figure for the heat loss rate as a function of delta T interior to exterior (in fact, I have done exactly that for Cedric -- the answer is 2277 BTUh per degree F delta T), I've yet to figure out how you reliably convert that rate into a boiler power output without using some sort of design delta T. I must be missing something?