Ducted Heat Pumps
I have a house in MA with a 2018 Mitsubishi heat pump downstairs, a large ductless unit. It is not used very heavily because the whole house is still heated using gas, from the furnace in the attic. I am considering replacement of that 18 year old gas furnace, using another heat pump. It does not have to Mitsubishi, but if a Mitsubishi does the job it may make maintenance a lot easier to procure. The furnace is way oversized and creates drafts when it runs, it is like a tornado. So, the house is ready to be updated with a less powerful attic unit. (Currently I use ~90 therms of gas in the coldest month of the year). I can adjust the duct work to direct more of the heat to the upstairs from my attic unit. The downstairs unit is 18K, the MSZ-FH18NA2. https://www.ecomfort.com/manuals/mitsubishi-708136bd9a2949a926c55242eb7574b3.pdf
Final bits of info: My zone is 5 in central Mass. and the HPSF of my current heat pump is an impressive 12.0 Looking at the spec sheets online I see the following from Mitsubishi: SV-KP24NA (24K unit), SV-KP30NA (30K) Any experience with these systems? It does appear that they have an HSPF of only 10.4, lower than my current downstairs unit. Are there other systems coming out from Mitsubishi and other vendors that push the envelope of HSPF rating, or that are cheap to maintain on an annual basis?
My house needs very little cooling so I only care about the heat ratings. Also in my climate zone (it rarely hits 0F and hits -13F only every 30 years or so) do I need to keep any backup heating source like a wood stove or can I just go all electric? In about 18 months I plan to add more solar and a battery backup. Thank you.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
I have the SVZ-KP24NA (indoor, airhandler) and SUZ-KA24NAHZ (outdoor, hyper heat) and I highly recommend it. The HSPF is a very difficult thing to measure and therefore to compare for modulating heat pump systems. For instance: this combination has a seemingly awful COP of 1.66 at Max 23,000 btus/h at 5 degrees. But if you're at 13,927 btu/h at 5 degrees outdoor, its COP is 3.2 (at 5 degrees!), which is outstanding. With a heat loss so low (90 therms/ winter month - does that include cooking and hot water too?), you'll be modulating nearly the whole time.
I don't understand the calculation you provided. Are you saying the unit peforms better than spec, or what? In my house each floor is about 1000 sq. ft.
If you are suggesting I could go with an 18K unit instead I will explore that. When I was considering a gas furnace replacement the tech suggested swapping the 5ton system for a 3 ton, so those HVAC guys will be shocked if we can go down to 1.5 ton. But it is doable as we have spent $11K on insulation. See my GBA posts from 10 years ago and you will get the idea.
fyi my usage did not include hot water and the cooking load is negligible. I honestly don't know what the monthly power usage is on my Mitsubishi, so I should try to measure that, and report back. My water is heated with electricity, a GE unit that still works and keeps the basement cool and dry.
I'm saying that "spec" isn't really meaningful here. A low efficiency furnace is roughly 80% efficient at all temps and all loads which is pretty well captured by AFUE, but HPSF is harder to nail down. Basically, no need to wait on a future super heat pump, a part load COP of 3 at 0 degrees outdoor is hard to beat. 24kbtu seems well sized for you.
ok, that is helpful. So oversizing is still helpful. My only remaining concern is the size of the duct work, did you run into any issue putting the heat pump air handler into existing ductwork which expects a powerful blower motor? I imagine that there will be some redesign of ductwork or additional heads involved if I want the smaller rooms in the downstairs to have any working registers. Ultimately the HVAC guys will tell me what I should do but I want to be educated before I talk to them.
The Mitsubishi has a powerful blower. The most noticeable difference is that since the heat pump modulates, it's a consistent heat. The furnace was on for maybe 10 minutes per hour, so had to move almost the same quantity of air in 1/6th the time. Slow and steady is better.
To make an informed decision you need a more detailed analysis. This will give you a more refined estimate of your heating requirements, which of course will reflect your climate (zone, or preferably more detailed data), square footage, insulation package, windows and glazing, air leakage, etc. If you're heating primarily with a gas furnace, and it's used mostly for heating (or you can estimate cooking/hot water), and you can estimate the efficiency of your furnace, then you could calculate energy use (and thus sizing) and if desired relate this to temperature (given climate changes and warming trends). There are MANY posts on this site about sizing HVAC systems - see e.g. https://www.greenbuildingadvisor.com/article/when-do-i-need-to-perform-a-load-calculation
What Paul was noting is that heat pumps are much more efficient when they're not running at full capacity. In his example, the efficiency of the unit is almost twice as great when it's running at 60% (14,000/23,000) btu/hr than at 'full speed'. You need to size the system so it meets your maximum load requirement, but it's very rarely going to be running at that full capacity.
Question: based on some visits from contractors it looks like I may need a system that is around 45,000 to 48,000 BTU. If I were to pursue that, does anyone recommend any units that would operate well with existing ductwork, with the main unit installed horizontally?
I am waiting on a quote from a contractor who said he could handle this requirement, but another said that they would need to go with an older tech heat pump that had "heat strips" that would kick in when it got cold.
In general, how do I find a contractor who understands that I am willing to go all out with insulating (already I have spent about $11,000 before rebates on insulation) and even add insulation on my exterior siding to make the house more compatible with heat pumps that are not of the 4 to 5 ton variety.
Latest estimate: the house uses about 260 therms of natural gas per year plus the downstairs unit consumes about 1300 to 1600 kilowatt hours of electricity (it is an HPSF 12.0 mini split, the hyper heat variety).
I don't mind spending for a larger system than absolutely required but I don't want a large system that ends up being very inefficient.
Square footage is officially 1780 but we have some added attic space, a bonus room, that has almost no heat vents but stays pretty warm just because it gets heated by what is below it.
What's the typical thermostat set point in your home? A heat loss of even the low end estimate of 40,000 btu/hr in Central MA is extremely out of line with a usage of 263 therms/year. Unless you set the thermostat really low, using 263 therms/year implies a very low heat loss. I wouldn't proceed until the contractors sort out why their sizing is so much larger than reality.
Try this: https://www.greenbuildingadvisor.com/article/out-with-the-old-in-with-the-new
Contractors are not in the business of right sizing equipment. When it comes to heat pumps in colder climate, it is even worse. They probably look at the label on the furance and will want to replace a 60k BTU unit with a 5 ton heat pump.
Run through the calculations in the link Paul posted and see where you are. My guess it will be under 2 tons.
Thermostat set point is usually 68 degrees. At night the house is usually set to 64 degrees. Perhaps the contractor is assuming that future occupant would want 72 degrees and sizing accordingly.
You bring up an interesting question on set point. To avoid a sudden warm up, it might be nice to have a thermostat we can ramp down and ramp up to keep night temps good for sleeping. So at 9pm, drop to 67, then to 65 at 11pm, then back to 66 at 4am, 67 at 5:30, 68 at 7am. Not all thermostats allow you to have so many changes and that could be a factor in deciding which brand or which external thermostat to use.
Take to heart, oversizing is not good.
Quick calculations (hope they're right): 90 therms is 2637 kWh, with an estimated 0.9 efficiency of your gas furnace, the 2637 kWh gas in is down to 2373 kWh delivered heat (heat demand/load for your coldest month). If you had baseboard heat, 2373 kWh in from the grid would give you close to 2373 kWh heat out. With a COP of 2 for a heat pump, you can knock that down to 1186 kWh in from the grid and 2373 kWh out (heat to your home). Again, a heat pump with a COP of 2 will deliver 2 units of heat to your home for 1 unit of power to the heat pump.
My heat pump rated at 28.6 kBTU/h sucks about 50 kWh/day for 2100+ sq ft in CZ6 when the average monthly temperature is 20 F (hitting - 7 F outside during the night) for about 1500 kWh in from the grid for the entire month. But most of the time, we're not even close to the design temperature of -1 F. And when we are sub-zero outside, inside is at 68 F (no problem). Heat pumps are OK in sub-zero temps (e.g., even as low as -20 F or lower).
As Akos mentioned, you're probably closer to 24 kBTU/h than 45 kBTU/h or 48 kBTU/h. As to efficiency, here's a plot, note that if you oversize, the heat pump will turn on/off (duty cycle) for low demand, which is hard on the heat pump and quite inefficient. Most of the year, you'll be at low demand - so, oversizing is not good. It's best to have a heat pump that can ramp down to a very low minimum output and meet your target design temperature heat load when running at 100% on those rare days.
There are too many stories on GBA about oversized equipment mistakes. I think people are too afraid they'll freeze in winter, if you're worried, all you need to do is buy a couple of space heaters for the day or two every few years that it drops below -20 F where the heat pump may have a problem keeping the house at 65 F.
My old 1.8 L turbocharged station wagon can go over 120 mph and accelerate fine (with less than 200 HP). Bigger is not better, why would I need more HP? I guess there may come a day when I need to drive at 150 mph . . . like the dealer mentions when I bring it in for service ;-)
This is helpful. It seems that our heat pump is doing more than just the downstairs, after I examined the numbers. Our gas usage is down to a worst case of 80 therms per month and this winter we were well below that.
For electricity usage, we have a baseline in the winter of around 450 kwH and the winter usage pushed it up to 750kwH, mainly to keep our water heater working in cold conditions, as it is a Geospring unit. The air source heat pump pushes that usage up another 550KwH to 1300KwH. In a cold January, as before I am assuming 23% higher usage and that would push us up to 675KwH.
Since the Mitsubishi has a COP close to 3.0 let' s multiply. So that works out to the equivalent of 2025kwH of electric resistance heating (it feeds downstairs but some of it makes its way into the air handler, the way our home is laid out).
So you are saying that this 80 therms is equivalent to 2109 kwH of electric resistance heating. Which means our total one month electric resistance load would be 4134 KwH in a bad month. That would be 23% more degree days than what we experienced in the most recent peak month, which lasted from January 20 to February 19, 2022.
So if we size our unit to work alongside the downstairs unit it appears we may be able to get away with a second 18000 BTU unit. But it is a mystery how much air flow that 18000 BTU unit would give us, and that air flow needs to be large enough to push air into some floor registers on the first floor, all the way from the attic.
So our coldest month we are talking about 3309 kwH in total load. If we get a second heat pump, ducted, with 2.5 COP the electricity actually needed to run the load in a colder than normal winter is:
675kwH (mentioned earlier)
2109/2.5 = 844 KwH (upstairs unit)
TOTAL = 1519 KwH
So it sounds like, to be safe we would might need the 36000 BTU unit since it is twice what we have in the downstairs now. But as you say, you are running a larger load with a 28600BTU as your units are running more than mine is, apparently.
This is very interesting. I don't know who would be able to tell us that a 30000BTU unit would qualify for a whole house rebate, as there is a verification procedure and the people doing the verifying may include some of the contractor types that make money off of selling oversized systems.
The issue outstanding is that on a very windy day some of the wind blows cold air through our walls and thus peak day load is greater than peak month load. It would be nice if our state law in MA allowed a consumer not to design a system for a peak day, like a nor'easter with 50mph winds, that may occur only once a year. I told the contractor that we planned to address some of the deficiencies in the walls, mainly the portion that is very old and doesn't seem to have modern sheathing.
The rebates might not be aligned with efficiency or comfort unfortunately.
If you used 80 therms from 1/20/2022 to 2/19/2022, that was 8,000,000 Btu input into your home during a period where Boston had 1092 HDD_65, which is 305 Btu/HDD/hr. So for a 60HDD day (average daily temperature of 5 degrees), that implies the furnace provided about 18313 btu/hr input, or ~16,500 btu/hr output if the furnace is 90%. Even with a 40% oversize, that puts you in the 24,000 range if you use the existing heat pump as well.
Thanks. Our actual therms were 65 and I used 80 for the calculations above. So I was adding 23% beyond this winter's peak.
Just curious, do you have some HVAC credentials? If I say to the next company I talk to that I consulted a random guy on the Internet it won't get me very far! I know you are a homeowner with an actual story but it will be necessary for me to have someone in the field verify that your sizing recommendations are good ones.
The calculation that Paul did is sizing based on actual fuel use. As such, you don't need to oversize, so even his 40% add on is not needed.
What I would do is take the actual output here (18k ductless+16.5k furnace) and set up your place in CoolCalc. Adjust the house parameters, usually the big one is air tightness, until it matches this actual load.
From there you have a real Man J you can provide to the tech as well as any agency that requires it for sizing.
Nope, the contractors are extremely off base here unless there’s some other explanation. Do you have any other heat sources? Wood burning stove maybe? Long periods of low set points (like months long vacations)?
We do have vacations but it has only been about 7 days this winter when the heat was lowered to around 59 degrees. We have no wood burning stove or pellet stove, but we but would consider one as a backup in case of a general power outage.
“Just curious, do you have some HVAC credentials?”
Zero! But flip it around: have their calculations ever been verified in the field? They get call backs if a system is too small - do they get call backs if it’s too big, loud and inefficient? Unlikely. Ask for some load calcs they’ve done and the same fuel usage.
To Rich, I have a PhD in chem eng and did work for National Renewable Energy Labs, Sandia, Lawrence Livermore, etc. If something stood out to me as odd, I'd post (like oversizing). That said, take it with a grain of salt as I'm unlikely to take more than a couple of minutes for a cursory review.
Akos' posts are quite informative (expert member) and Paul too. I've learned a lot from reading the Q&A, a pretty good community (wisdom from an experienced crowd). Akos at #22 is good advice :-)
there should be a community of HVAC guys who have figured out how not to oversize systems (if there is not one, it needs to be created). Otherwise the next generation of HVAC people is going to be trained to oversize some of these systems.
Of course it is also true that many people with almost uninsulated homes are trying to get these HVAC rebates so they have to deal with that.
Given that a building temperature will not respond instantaneously to set-point changes (whether coasting down toward a dropped set point or ramping up to an increased set point) I don't think you need anywhere near that many stepped adjustments. Probably one on each end would suffice. But also keep in mind that many recommend not to set-back too deep with heat pumps.
In the Mitsubishi, the system is programmed to be aggressive if there is a difference of 3 degrees from the set point, but it is more relaxed in its heat output if there is a difference of 1 or 2 degrees. So for best efficiency my understanding is it is best to keep the thermostat relatively close to actual temperature. If this sounds like I am using the heat pump thermostat like a manual transmission, maybe that is accurate. Other brands might by default be less aggressive when trying to ramp up (sort of an "eco mode"). I don't know anything beyond what I have.
Hmm, yeah I could see that making sense on the ramp up perhaps (I'm definitely not familiar with the program and performance specifics). Not sure about the ramp down where it seems like you might just end up in 'neutral' anyways, depending on the characteristics of heat loss in your structure and the timing of your programmed step downs... but by programming it in small steps, I suppose you are guaranteeing to stay tighter to the set-point for the eventual turn back up.
It does seem like it would be much simpler if there was simply a slow ramp eco option like you say.
yes the step down could be abrupt and I don't think it would matter.
Deleted
Just a quick update: I had a difficult time getting anyone to quote non-oversized unit and by the time I got any company to provide this, it was too late to get the outdoor "hyper-heat" compressor, it was out of stock and about to be superseded by a new model.
But, thanks to a supplier in Maryland and a second installer who did our 2018 unit, we we able to have the hyper heat compressor (Mitsubishi SUZ-KA30NAHZ) we wanted shipped to our house in early September. We then completed our install of a 30000 BTU ducted air handler (SVZ-KP30NA). This hyper heat unit doesn't have the best HPSF rating on the market so I am not positive it will be super efficient but it is working extremely well.
In fact it is almost silent compared to our old unit. We elected to get a simple Mitsubishi thermostat, not the internet based one that is often installed with these units.
I have attached photos of the old unit and the new unit. Sorry that the photo of the new unit is blurry but you should get the idea.
Once we get statistics on our energy usage I can post those. We also opened some walls to make improvements to the building envelope so I am guessing we will may run up the kwh a bit until new insulation is put into them.
Do follow up with your experience after the winter, this is something many people may be considering
Just curious whether you have an update how things have been working out this winter?
Here is a follow up. As you might be able to tell, the particular combo of Mitsubishi we used is not great. Even so, after the install plus insulation work we are spending less on electricity than we previously spent on power plus natural gas plus gasoline.
If we had to do it again we would probably not go for the cold climate version of the 30K unit since we already had a cold climate downstairs heat pump that can deliver heat during the extreme -13 cold. Even so, we will be able to keep the new whole house system off for 8 to 9 months out of the year and heat the whole house with the more efficient 18K unit starting next year due to our improved building envelope. Also, by the time you read this, Mitsubishi might have a newer model of the 30K (with air handler) that can both meet the new federal HSPF2 standards and perform at peak level when the temperature gets down below 10 degrees Farenheight.
Here is our Net kwH usage for the whole house for the entire 22-23 heating season:
668 month ending may 19 2023
931 month ending april 19
1438 month ending march 19
1537 month ending february 19
1583 month ending january 19
1780 month ending december 19
947 month ending october 19
Here is our Net kwH increase year over year for each month:
123 month ending may 19 2023
219 month ending april 19
447 month ending march 19
235 month ending february 19
507 month ending january 19
1031 month ending december 19
398 month ending november 19
This may look alarming, but it is actually fine because you have to take into account the power usage by the new electric car, in use from october 18 2023. It got 4 miles per kw and traveled almost exactly 4000 miles. So that is 1000 kw divided by seven months we have had the car. So 143 kilowatts per month is the energy usage of the new electric car. Plus we needed to replace the gas stove with induction. Let's assume it is 7 kilowatts per month (I have no idea actually) but that gives us 150 total kilowatts of increased consumption not related to space heating, per month. Therefore here is our year over year change in consumption with the new electric car & electric range usage taken out:
(amounts in kwH)
-27 - month ending may 19 2023, a decrease of 4.4% overall
69 - month ending april 19, an increase of 7.4%
297 - month ending march 19, an increase of 20.6%
185 - month ending february 19, an increase of 12.0%
357 - month ending january 19, an increase of 22.6%
881 - month ending december 19, an increase of 49.5%
248 - month ending november 19, an increase of 26.5%
You can see that November and December were bad, and February was quite good. This is because we stripped off many layers of siding to air seal the house we kept making improvements over time. Work was done in November, December, February, and April. We had a little heat wave in February before our solar panels got covered in snow in early March. And by May, we were able to finally turn off the heat after a couple of days in the 80s warmed up the house for good.
It's likely that when we are all done, the additional energy consumption due to the heat pump will come in below 250 kw a month in the 3 coldest months and below 100 kw a month in the shoulder seasons. In July and August we are likely to see a DECREASE in energy consumption because the house should require almost no air conditioning now. We typically ran our heat pump downstairs for A/C no more than 15 days a year the last few years. This year probably run the AC a bit less, but also turn the whole house fan on using the upstairs unit to bring the cooler air into the 2nd floor for the first time.