Resistance Heat vs. Wall-mounted Air Handler
I am looking at converting the upstairs of my house from an oil-fired hot water system to a mini-split heat pump system. The upstairs effectively has 5 areas, 4 of which are currently heated:
1. Master bedroom (~400 sq ft)
2. Office (~225 sq ft)
3. Kid’s Bedroom (~225 sq ft)
4. Bathroom (~45 sq ft)
5. Hallway/Landing/Stairs area (not currently heated, I don’t see any need to heat it).
The walls are all about 7.5′ tall from floor to ceiling. The floor has 10″ high joists underneath it, followed by the downstairs, which is heated by a pellet stove. As far as I know, there is no insulation in the floor.
Ceiling opens to the unconditioned attic, with fiberglass batts in between the 10″ high joists.
Looking at some rough calculations, I can add nominal 6 KBTU/hr cooling (~8.7 KBTU/hr heating) units to the Kid’s Bedroom and office, with a 12K BTU/hr cooling (~15 KBTU/hr heating) unit to the master bedroom and be just fine. However, I am worries about not getting enough heat into the bathroom then (see attached sketch). As I see it, my options are to either:
1. Add straight resistance heat to the bathroom (about .6-.7 kW of demand on a cold day). This is cheap and easy to do, but in-efficient. Something like the cadet in-wall electric heater. Couple of hundred bucks plus a little bit of wiring.
2. Instead, replace the planned Mitsubishi model FS 6K wall-mounted air handler in the office with a 9K model SEZ concealed duct unit. Then run duct through the attic so that that unit heats both the office and the bathroom. This would presumably be more efficient, but:
a. It’s a lot more costly
b. A lot more work to run the duct
c. While I don’t have a rating for the full system with the two different units plugged in in this location, if I look at each option as a single zone system, ecomfort.com shows the efficiency dropping from 24.6 SEER with the FS unit to 17.3 SEER with the SEZ unit. Plus, I will be running duct through the unheated attic. And I will lose attic space.
So, my questions would be:
1. Does anyone have a way to estimate the actual differential electricity usage between the two options? In other words, what is the equation used to go from SEER to actual energy usage.
2. Are there other options here that I haven’t thought of?
3. Unrelated, but I am most of the way through finishing my basement. Currently the basement is just heated by the waste heat from the oil boiler. Also, during the basement renovation, I disconnected the hot water heating loop in the boiler, and instead installed a heat pump water heater (A.O. Smith 55 gallon). The basement is noticeably colder already – once I remove the boiler (and thus the waste heat generation), any thoughts on if I will need to stick a 6K BTU/hr heat pump unit in the basement to keep it warm in the winter? Or at that point, would I actually save energy by just using the back-up resistance elements in the water heater during the winter (instead of first pumping the heat into the basement, and then pumping it again into the water). At the moment, I am thinking I won’t install anything down there, but will leave room in the system to add one if I have to.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
The units of SEER are BTU/hr/Watt. A resistance heater has a SEER of around 4. See: https://en.wikipedia.org/wiki/Seasonal_energy_efficiency_ratio
To your second question, I think it's more efficient to run a heat pump and a heat pump water heater than a resistance water heater. The formula for combined COP of a two-stage process is 1/(1/COP1+1/COP2). It all depends where on the performance curve your outside heat pump is. If it's at COP=1 -- ie resistance backup heat -- then yeah, you're better off running the HPWH on resistance. But if the outside unit COP is 2, so long as the HPWH has a COP over 2 you're ahead.
You are proposing the one wall mount per bedroom setup. Unless your bedrooms are 1000sqft each, the chances of getting it to work efficiently is pretty low. A number of threads on this site with people trying to make it work ie:
https://www.greenbuildingadvisor.com/question/was-my-system-designed-right
The issue is that the heat load (and cooling load) for a typical bedroom is much smaller than the smallest wallmount you can get. Combine this with the limited modulation of a multi zone system, the outdoor unit will be cycling and bypassing refrigerant through the zones that are off which results in dismal heating efficiency and almost non-existent dehumidification in the summer.
First figure out what your actual heating and cooling loads are for each space than install a single right size slim ducted unit to feed all the spaces. This will get you much better efficiency and comfort than what you are proposing. It is more invasive to install but well worth the effort. Plus multi-splits are not cheap, the extra cost of a couple of zones can pay for a lot of ducting.
>"Looking at some rough calculations, I can add nominal 6 KBTU/hr cooling (~8.7 KBTU/hr heating) units to the Kid’s Bedroom and office, with a 12K BTU/hr cooling (~15 KBTU/hr heating) unit to the master bedroom and be just fine."
That all feels RIDICULOUSLY overestimated/oversized (possibly ludicrously so). Show us the rough calculations!
Then compare your whole house calculations using the same methodology to the wintertime only oil consumption numbers as a sanity check using these methods:
https://www.greenbuildingadvisor.com/article/out-with-the-old-in-with-the-new
Then run the room by room Manual-Js on the space using LoadCalc.net or CoolCalc.com using AGGRESSIVE (per the instructions in the Manual) assumptions on all inputs. Done correctly a Manual-J will usually deliver numbers less than 20% higher than what the fuel use indicates, at which point you'll know the input assumptions are close enough to size the equipment correctly for both heating and cooling.
You are talking only about a total of ~1000 square feet of conditioned space. Unless the windows have no glass and there isn't a hint of insulation in it the 99% design heat load for the whole shebang should be under 25,000 BTU/hr, probably somewhere in the 15-20,000 BTU/hr range if it's insulated 2x4 construction with clear glass (not low-E) double panes or clear storms over single panes. The 1% design cooling load for the entire space is likely to be less than 12,000 BTU/hr unless it's less than R10 under a hot mopped black tar roof, or there is an enormous amount of unshaded west facing glass.
Unless the one exterior wall of the bathroom is 100% west facing glass it would likely stay within a few degrees of the rest of the house with NO direct heating. A low speed continuous exhaust fan would provide enough latent cooling to keep it comfortable during the cooling season if there is an adequate transom-vent or door cut to guarantee the air is drawn from conditioned space.
I personally prefer Fujitsu's -xxRLFCD series for these applications than the lower-static Mitsubishis, since the Fujitsu units can be mounted vertically in a mechanical cabinet/closet taking up less than 10 square feet of conditioned space. The top picture on this page is a 1.5 ton Fujitsu mounted above a large common return grille:
https://www.greenbuildingadvisor.com/article/getting-the-right-minisplit
The 1.5 tonner is good for about 20,000 BTU/hr @ +5F, and over 20,000 BTU/hr @ +95F:
https://ashp.neep.org/#!/product/32101
You can probably do fine with a 1-ton (run some more realisitic load numbers!) which is even smaller:
https://ashp.neep.org/#!/product/25312
Drop down to comment #3 for links to related pictures and discussions of that system.
Even with 7.5' ceilings should be possible to run ducts in soffits below ceiling level, which is more work, but far preferable to sticking them in the attic. With 3"-3.25" (designed to fit into 2x4 framing bays) or 4" x hard piped rectangular ducts (rather than flex) in a sub-ceiling soffit there is only 5" of headroom loss on the duct runs. I situations where every fractional inch counts you can even pull the ceiling gypsum or plaster/lath to gain some space, but there is usually a way to get there.
Thank you all for the information and corrections. Sorry that it took me a long time to reply. For my load basis, I had mostly been basing my assumptions on:
1. I have a pellet stove rated for 36,000 BTU/Hr to heat the downstairs that maxed out at about a dT of 60-65 F, so I was initially assuming that the two floors would take about the same duty (since they are roughly the same size). Apparently, that was a very bad assumption. I should also note that I have since gotten about 85% down with insulating the basement, and the pellet stove appears to be doing far less work since then, but I haven't had any true cold snaps to test it.
2. The existing hot water radiators on the top floor are rated for around 40,000 BTU/hr, but of course I should have realized that those would likely be oversized.
I ran my house's dimensions through the loadcalc.net page that Dana (thank you so much), and came up with 20,365 BTU/hr @ 5 F external. Until 3 months ago, the oil furnace was also providing domestic hot water, so I don't have good oil usage numbers to compare to the loadcalc results. However, I ran the loadcalc for my downstairs characteristics from before I insulated the basement, and came up with 38,000 BTU/hr at 5 F, which agrees pretty well with the observed performance of the pellet stove, so that seems reasonable.
Unfortunately, the 20,365 BTU/Hr requirement is enough higher than the 19,756 BTU/hr unit recommended by Dana, that I am leaning more towards the AOU30RGLX (https://ashp.neep.org/#!/product/25351). The AOU24RGLX would be big enough, but has really low COP's as compared to the 30. I am also considering the Haier 1U36LP2VHA (https://ashp.neep.org/#!/product/30022)
My house was built in 1951 and the only insulation is fiberglass batts in the attic (and now EPS on the basement floor and PIR board on the basement walls). Most windows are double-pane anyways, and the 3 that are still single pane are getting replaced this spring. I ran the numbers to see what the performance would look like if I insulated the wall cavities with blown cellulose or blown fiberglass, and got about R13 insulation from that. That would drop my upstairs heating duty to about 12,250 BTU/hr, which would let me go for a much smaller unit. So, I guess I will be reaching out to some insulation contractors as well, although I am nervous about what the pricing for that will look like.
In terms of ducting, I am thinking of having the following done:
1. Run supply and return ducts to the master bedroom, office, and kids bedroom.
2. Also run a return duct from the downstairs L.R. (has the pellet stove in it), with a supply duct to the other side of the downstairs (Kitchen/Dining Room).
3. Put air zone dampers to divide into three zones: Master, Downstairs, and office/other bedroom.
My thought is that then I could use the unit to provide AC to the whole house on the hottest days of summer (preferably just let the house stay at ambient on not-super-hot summer days). On spring/fall days when I need some heat but not a lot (e.g. when it's 50 out), I could choose to just run the fan and circulate the hot air from the pellet stove through the whole house instead of running the heat pump (or, run the heat pump to the whole house and shutdown the pellet stove).
Any thoughts/comments/"don't do THAT" warnings on this approach?
Thanks again for everybody's comments.
Two things stand out to me. If you need 20k BTU/h at 5 F, then you'll need about 7k at 47 F. The 30k unit modulates down to 9k at 47 F. It will cycle. The 24 k unit modulates down to 5.4 k. More comfort, less electricity.
If you go with a zoned system, you have to make sure that your duct work is designed so that any single zone can handle the total output of the unit without exceeding the allowable pressure drop. This may be challenging with a slim duct unit, especially with your limited space.