Retrofit heat pump sizing
I am in the process of adding a solar array to a 4 unit building that I own. As part of that, I am planning to convert one or two of the 4 units to heat pumps (currently hydronic oil). It’s an old row house, but in 1900, insulation isn’t great, nor are the windows (30 y/o vinyl). The units are two story town house style, about 1300 sq ft each.
One of the companies I have quoting the solar also does HVAC. They initially spec’d a 2-ton ducted system for the second floor and a 2-ton, 2 head system non-ducted system for the first floor. The only justification they can offer for that spec is square footage.
I thought this spec sounded way oversize, so I started digging into utility bills from when I lived there. I am concerned with heating load, not cooling in general, given the property is in Massachusetts. I used oil consumption from a 5 week period in Jan-Feb 2014 (avg temp of about 16F) to calculate the actual heating load. Working through efficiencies, oil-btu conversions etc I came to about 2 tons for the whole unit (this includes hot water).
As has been suggested in other threads here, real data is better than a calculation. I proposed a Mitsu 29k 3 zone system (MXZ-3C30) with 2 6k units on the first floor and one 12k ducted air handler on the second. Of course the contractor balked and assured me it wouldn’t keep up.
All of this brings me to another question, can I rely on the max capacity rating at 5f? Per the submittal it’s still 28,600 at that temp, but the rated capacity at 17f is only 18k (there is no rate capacity at 5f).
Thanks in advance,
Anthony
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Anthony,
Four tons sounds like a lot for 1,900 square feet. Dana Dorsett has a 15-minute method for calculating accurate building loads. It's available here: https://www.greenbuildingadvisor.com/article/out-with-the-old-in-with-the-new
Look into the efficiency differences between zone systems and individual systems (ie 1-to-1).
It sounds like you've done something at least similar to Dana's 15-min. method, but do his again and see if it holds up. I would be much more confident in that that something pushed by the contractor.
>"They initially spec’d a 2-ton ducted system for the second floor and a 2-ton, 2 head system non-ducted system for the first floor. The only justification they can offer for that spec is square footage."
OK, don't be coy- how many square feet on each of those floors? Is that 1300' total, or 1300' per floor?
Are these end units, or in the middle, with a heated unit on each side?
Double wythe brick, framed walls or...????
Run yourself a quick & dirty I=B=R load calculation as a sanity check on the fuel-use calc, or an AGGRESSIVE manual-J-ish calculation using LoadCalc.net or CoolCalc.com. (These tend to overshoot reality by quite a bit, but are at least in the realm of sanity, unlike most freebie online load tools.) The fuel use numbers usually (but not always) come in a bit lower than an IBR type calculation, but are often within 10% of a competent Manual-J using pro tools.
The -NAHZ version of the -3C30 is good for 28,600 BTU/hr @ +5F, an probably /WILL keep up if you were estimating a 24,000 BTU/hr load based on oil use:
https://ashp.neep.org/#!/product/26173
If the oil burner is in a semi-conditioned uninsulated basement the standby & distribution losses are most likely well into a double-digits percentage, so the actual load is probably smaller.
More specifically...it is an end unit and 1300 sq ft is the total for the unit. It's all very rectilinear, so its on the order of 650 sqft per floor (I don't have exact measurements). Townhouse style, 3 outside walls. The basement is also walkout, unheated, and generally quite cold. The washer was right next to the outside wall and it would freeze when temps dropped below zero for any appreciable time (we used to suck the water out of it with a shop-vac during cold snaps). So the heat loss from the boiler is certainly a relevant point. [edit: Now that I think about it, losing that boiler heat loss might be a real problem. I wonder if I should put a head/vent in the basement. I would rather put a 6k heat pump head in than a 1500w resistive strip.] Framed walls, plaster and lathe, cellulose in the attic, pretty sure there is nothing (or close to it) in the walls.
I did a quick I=B=R and got 26.4k without the basement and 28k with it (it's only one outside wall and a temp delta of 25 deg).
If I scale the oil use calc linearly down to the 5 deg design temp I get 29k, so that's pretty good agreement there. I noticed in your other fuel use estimation post that someone linked you recommend a factor of 1.4x, which puts me at about 41k.
This additional safety factor is kind of a big deal, as it would move me from a 3-zone single unit to separate systems for each floor. I say this because, from what I can find, once you get over 30k the systems just don't throttle down low enough to make sense 90% of the time. For example, the 3C30 goes as low as 7.2k heating while the next size up 4C36 will only go as low as 22.5k. Maybe I am overestimating the efficiency hit here, but it feels wrong to spec something that has minimum output of 75% of my max heating load.
The 2 system setup (probably 2x 18k units, maybe a 12k + 18k?) is likely more efficient but costlier upfront.
My other thought is to go with the smaller system and add some baseboards as backup. I figure this is cheaper in the long run than the oversize 3 zone or the 2 system setup.
Let me know what you think, and thanks for the time.
-Anthony
As an update, I think I have made a decision about what to do.
I am going to put an 18k AH in the attic to feed the 4 upstairs rooms. This will connect with a 12k outdoor unit. The Mitsu 12k actually behaves more like an 18k in reality, so this should work fine. On the first floor I am going to do 2x 9k units, and 1x 6k in the basement. These will connect to a -3C24.
This will give me 42k-ish capacity, as suggested by the 1.4x sizing factor and a heat load of 29k. The only spot I have some over sizing is when the basement is running all by itself, since it will be a 6k on an outdoor unit that only throttles to 7.2k. The 2x 9k's on the first floor is probably a bit large, but the system should be able to throttle to make this reasonable.
I think this is sane, and the cheapest/most efficient way out. Interested if anyone thinks otherwise.