Air to Water Heat Pumps

Thanks, yes, I was thinking 2 large AWHP's, like 40kBTU each.

Individual air-to-air units would be less costly, but the way the house is built would make it difficult to run ducts or mini-splits. So the only practical heating solution would be AWHP's.

Maybe the most cost effective solution would be to install just one AWHP sized for 50% of design heating load, and keep one of the two existing boilers as supplemental heat.

The two things to look out for are 1. low temperature capacity and 2. water temperature. Although you said you thought you could heat the place with 120F water.

Is there an obvious way to split the system into two zones? Two heat pumps on the same zone would fight each other for control. If you could split it up I think a pair of five-ton nominal heat pumps would do the job.

The heat pump efficiency makes sense, but one of our hurdles as a 4-unit condo association is the up-front cost. Owners are not going to make a huge up-front investment when they may not be planning to own here long-term enough to recoup their investment in energy savings. So a less-expensive first step that still saves some energy is an easier sell.

Also, "half" a heat pump may end up providing more than half the BTU's in a heating season, since a 40kBTU heat pump could provide 100% of the heat on days above 35 degrees, and 50% of the heat on zero degree days. So a seasonal average could end up giving us 75% or so of our heat from that half a heat pump.

Yes, the house is already split into 2 zones, one per side of the house, each zone serving 2 units. Hence the 2 boilers, one per zone. So my thinking was one 40kBTU heat pump per zone instead of the oil boilers. But I don't think the owners would go for the up-front cost, which is why I'm wondering if it makes more sense to try and start with a single heat pump that could probably provide 75% of the BTU's to both zones in a season.

"With an HBX controller you can set up the heat pump to run when it is warmer outside and with lower water temps. Then run the boiler at higher temps when it is colder outside. During those conditions the SCOP will be much higher on the HP. You could average 4 or 5. That means 400 to 500% efficient."

Right, that's pretty easy to do. But that's not what Jesmed1 was talking about. He wants the heat pump to run even when it's cold, but if it doesn't have the capacity then the boiler "helps" it out.

My point is that the only reason to do that would be if at the cold temperatures the heat pump is cheaper to operate than the boiler, but it just doesn't have the capacity. I argue if that's the case you're better off just having a heat pump that has the capacity you need and running it all the time.

There are heat pumps that have a boiler control output, to run the boiler only when the heat pump can't keep up, I know US Boiler makes one. What I don't know is whether they truly run the boiler only as much as needed and let the heat pump run as much as possible.

@John Ruhnke said:

With the HBX I can set up the system to do what Jesmed1 wants just a little differently. 75% or so of the heat from the Heat Pump. 

How does the plumbing work to get the boiler and the HP working together simultaneously on a design day to heat the same zone? I'm a beginner here, so I'm just learning, but I was imagining something like a primary/secondary with both the boiler and the heat pump on the same primary. On a milder day, the controller would have the HP only feeding the primary. On a colder day, the controller would have both HP and boiler feeding the primary.

But I suspect there's going to be some sort of water temperature mismatch problem with that concept. I imagine that's the difficulty @DCContrarian was referring to. Maybe the answer is some sort of buffer tank arrangement with a heat exchanger.

I'd like to offer an analysis, the data is in a spreadsheet at

https://docs.google.com/spreadsheets/d/1KmnOjq_MEQ2fNqj5AZvLSXHkNx3dwDPvn_SNBV2iHMA/edit?usp=sharing

@John Ruhnke had sent me a performance sheet for the 5-ton Arctic (050ZA(BE)) I used that for performance curve of COP and output vs. outdoor temperature. I used weather statistics from Boston Logan Airport for the hours per year at each temperature. I assumed an electricity cost of $.27 per kWh and an oil cost of $3.50/gallon, maybe someone closer to Boston can check those for me. I assumed 80% efficiency for the oil burner.

I modeled a building with a heating load of 35,000 BTU/hr at a design temp of 13F, based upon the discussion above.

I calculate that this building needs 80,115,978 BTU of heat for the winter.

To get that 80 million BTU using fuel oil would take 731 gallons and cost $2558 per year.

To get that 80 million BTU using only the heat pump, and supplementing with resistance electric when the capacity of the heat pump is insufficient would take 8471 kWh and cost $2287 per year — savings of $271 or 11% over oil.

I then modeled three other scenarios:

Scenario 1 is the oil burner is only used to supplement the heat pump at times when the heat pump output is insufficient, basically the oil burner is used instead of resistance electric. This saves $24.33 per year compared to the all-electric scenario.

Scenario 2 is the oil burner is used for all heat when the temperature is so low that the heat pump costs more to run than a boiler. In this scenario, the break-even temperature is 10F. This saves $34.18 per year compared to the all-electric scenario.

Scenario 3 is the oil burner is used for all heat below a certain temperature where the heat pump is assumed to be unable to produce hot enough water to meet the heating need. I used 25F because that was what John had assumed in his hypothetical. This saves $21.92 per year compared to all-electric and $293 per year compared to all-oil.

Some conclusions: in this climate, for this building, a heat pump will save money compared to an oil burner. But it's modest, 11% of annual fuel cost. It's probably only worth considering when the boiler is due for replacement anyway. Other improvements, like weather-sealing and regular maintenance of the boiler, may be more cost-effective ways of seeing savings.

The payback period of any scheme to combine a boiler and heat pump is going to be decades if not centuries. If the existing radiation is insufficient to meet the heating load with the temperature water that a heat pump is capable of producing, it's hard to justify a heat pump for part of the heating load, you're better off sticking with the boiler. Alternately, it may be worthwhile either to increase the radiation or improve the weather-sealing in the building to the point that a heat pump becomes effective.

I welcome criticisms of my methodology. Some notes on that. First, I came up with 731 gallons of fuel oil used per season for my hypothetical building, which is significantly different from the 1200 gallons that was reported. So something isn't being accounted for. Is the boiler also used for hot water?

I am probably over-estimating the amount of electricity used by the heat pump. I set up my spreadsheet for air-to-air heat pumps, which provide part-load capacity and efficiency numbers. Arctic only provides full-load numbers, and heat pumps tend to have higher COP at lower loads so actual COP is probably higher than what I'm seeing. Also, I'm assuming a constant water temperature of 122F, in reality outdoor reset would be used which would give better performance at warmer outdoor temperatures.

But I don't feel that any of those issues affect the conclusions.

"What type of heat emitters would be use with 122 SWT."

That comes from @jesmed1 in the first response in this thread: "Because we have so much cast iron radiation, I believe we could heat the building on a design day with 110-120 degree supply water." I'm just assuming that he's correct.

"How would you get to the 4 plus COP that Heat Geek claims?"

From what I've read, there really isn't much more to it than having lots of radiation and really low water temperatures.

Note that they're claiming SCOP, seasonal COP, which is normalized to exclude the effect of climate and not directly comparable to the calculated COP I'm showing.

Apparently guys who are really good at it get SCOP's in the 5.5 range. This article has some details:

https://arstechnica.com/science/2024/07/the-hunt-for-the-most-efficient-heat-pump-in-the-world/?fbclid=IwY2xjawH2VwlleHRuA2FlbQIxMQABHfKALOmCjmEXQAGBtiNnsnbf-oHDPA7H6uxXmaPQyAuTPi4CPuwMTkgKpw_aem_Oq4kXweijwakCFct0PrJFA

Thank you, that is fabulous. You just did a way better job of what I should have done to start with!

I think your analysis is excellent and offers me the surprising result that heat pumps would not save us nearly as much as I expected. The key parameter is the "break even" COP of 2.48, which is much higher than I would have expected. I figured we'd get a seasonal COP average of somwhere in the 2-3 range, and your analysis shows that, in fact, a seasonal COP in that range is not going to save us much. Which is surprising to me.

Just to clear up any ambiguities in our numbers, for our 4-unit condo building, we have:

4800 sq ft

1200 gallons oil/yr

Design heat loss at zero degrees: 80,000-100,000 BTU/hr

(I like to use zero degrees which is lower than the standard 9 degrees in my area, just to be conservative).

The design heat loss numbers are based on actual oil consumption from boiler run times at various outside temps, the Green Building Advisor method for calculating heat loss based on actual oil consumption and recorded HDD's for a cold month, and various other methods. All methods end up somewhere in the 80-100kBTU/hr range.

None of that changes your conclusions. The house is split down the middle by a demising wall, with each half heated by its own boiler. So the 35,000 BTU/hr heat loss number you used is about what one half of the house uses, and the 700 or so gallons of oil you derived is a bit more than what one half the house uses. So we would just double your numbers to get the results for the entire house.

But your bottom line finding that a heat pump would not save us as much as I expected is, I think, correct. So thank you for that excellent analysis.

"the 700 or so gallons of oil you derived is a bit more than what one half the house uses."

DUH! I had completely forgotten that I was only doing half the house!

You are actually quite close. We were paying $3.50/gal until last week, when the price started to trend up. Not sure what our current electric rate is, but Google AI says the current average in MA where I live is 33 cents/kWh.

Ummm….wow. Well, that saved me a lot of wasted effort. Thanks!

It's not a simple calculation, it's not something you can apply a rule of thumb to.

It depends on your location, both for the climate and for the local cost of energy. It depends on the house, both for the heating load of the house and also for the existing radiation. And it depends upon the choice of heat pump.

It's probably best to start a new thread, but if you gather:

1. your zip code
2. The cost of electricity where you are in cents per kWh
3. The cost of your existing fuel, whatever it is (or your preferred fuel)
4. An estimate of your house's heating load, using this method:

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

I can create a spreadsheet for you comparing like the one above. I'd do it for the Arctic because that's what I already have the performance curve for, I don't believe that any of the heat pumps on the market right now have radically different performance curves.

The other practical question would be whether your existing radiation could support your heating load with the kind of water temperature a heat pump is going to provide, around 120F. If not you would have to add more radiation.

This weekend ill try to find time to gather all the info up.