Air to water heat pump, Northern NJ
littlegojira
| Posted in General Questions on
We are doing a deep energy retrofit in Northern New Jersey. This includes air sealing as well as exterior insulation (3” on walls/4” on roof). Is an Air to Water Heat Pump system the right choice. We are converting from a gas powered steam boiler and traditional ac system. The home is approximately 3400 square feet. With the change in refrigerant and potential future changes, I wonder if it makes sense to switch to a monobloc now. My concern is performance in cold weather. Minimizing the use of heat strips is the main concern/wish to avoid.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Search and download construction details
Replies
Read this post for general background on sizing and modeling heat pumps:
https://www.greenbuildingadvisor.com/question/how-to-model-and-predict-electricity-usage-for-a-heat-pump
The NEEP database doesn't contain information on air-to-water performance, you have to get that from the seller. My understanding is that NEEP only accepts listings from the manufacturer, and none of the air-to-water heat pumps sold in the US are distributed by the manufacturer.
A list of air-to-water heat pumps currently being sold in the US is at:
https://www.reddit.com/r/AirToWaterHeatPumps/comments/1jj8kvu/air_to_water_heat_pumps_currently_being_sold_in/
Great, thank you!
You need to start with your building load first. Anything high performance in milder climate like yours should be under 10btu/sqft, so well within the reach of most heat pumps. I'm in much colder climate and air to water monoblock works without issues. I even run plain water out to the unit (there are a number of freeze protection backups though).
Usually the decision for heat delivery comes down to the amount of cooling and summer humidity. Since you are in an area with a lot of humidity, radiant cooling won't work. This means air handlers and ducts. Since those ducts can also be used for heat, radiant heat starts to become a redundant system that at least doubles the BOM cost never mind installed cost.
If you must have radiant, the in-between option is to size the air to air to handle your cooling load and air to water to make up for the difference in your heat load.
Great, thank you. Do you use the air to water for domestic hot water as well? The only hydronic radiant is in the mudroom and we will keep that in place because it’s embedded in an insulated slab. It was used to heat the room but switching to hot air (via heat pump), we will use it to warm the floors.
What brand of monobloc did you use? Pros/cons/recs?
If you have a hydronic floor then you have a hot water boiler, not steam.
Air-to-water is generally a poor fit for domestic hot water. You're better off using a heat pump water heater.
Thanks. Just to clarify, there is hydronic in-floor heating in an addition we did a few years ago. This system is on a separate gas furnace. The majority of the house (95%) is on the steam system that is run on a gas boiler. The steam system is one zone. We have two condensers and two air handlers for the cooling system (one on the first floor and one on the second).
Curious why an air to water system would be a poor fit for domestic hot water. My understanding is that you would have a very well insulated storage tank downstream from a warming tank. I am basing this off the latest TOH series. They are installing an air to water system for both heating/cooling as well as domestic hot water. The storage tank is huge.
With a heat pump, if you're mixing, you're doing it wrong.
The efficiency of a heat pump is determined by the difference in temperature between the hot side and the cold side. If you're heating water to a higher temperature and then mixing it down to a cooler temperature, you're giving away efficiency.
Air to water heat pumps do need a buffer tank because the compressor needs to be protected from short-cycling. But that buffer tank is a necessary evil, it should be as small as possible to do the job of protecting the compressor.
That's a round-about way of saying that TOH is giving very, very bad advice.
Heat pumps are typically sized close to the heating load of the house. Heating domestic hot water typically creates high instantaneous loads, so most heat pumps turn off heat while the hot water call is serviced. This means you can go an hour or more without any heat. In the summer, you have to disable cooling while heating hot water happens, so you get an hour or more without cooling. At the same time, the air-to-water heat pump is typically less efficient than an interior heat pump water heater, so there's no operating savings. You need a large storage tank, so there's no space savings in the utility room. And the bill of materials for the hot water option is going to be several times the cost of a stand-alone heater.
Yes I do run an indirect but it doesn't work well for the issues DC Contrarian mentioned. You are in warmer climate so recovery will be better but it is still not worth the cost/complexity. This summer I'll have to add in an element to the tank and hook it to the monoblock to speed up recovery during deep freeze. Loosing heat for 1.5h or so in the middle of the winter does not work.
Stand alone heat pump water heater is the way to go.
For smaller existing heated floor areas, the simplest is to get a small resistance tank heater with a larger element. You can also get a resistance boiler.
I got a Dettson/Gree unit. It was reasonably priced with local support which is the most important part. Some of the smart programmable features on it are not well though out making them kind of useless. It does work very well for basic hot water supply with outdoor reset which is really all I need.
Does work for cooling as well but I found cooled slab to be uncomfortable especially with smaller kids at home where you spend a lot of time on the floor. Won't be using it for cooling.
I see, thank you, much appreciated. Do you think the heat pump water heater is as efficient as a gas powered tankless? We are considering two WHs for the home. One would be in the basement to service the kitchen and two powder rooms and one would be in the conditioned attic for the three full baths on the second floor.
Thanks for the background on the hydronic system. Yes, we have small children as well and the cool floors are not ideal. We have ducts run to that room as well for cooling.
By "efficient" I assume you mean economical to operate. That's going to depend entirely on what electricity and gas cost where you are. Prices vary so dramatically around the country you really can't say.
The existing comments are (unsurprisingly) quite good.
I'm in the 4A portion of NJ, but came from the same place (steam heat + AC, high velocity in my case). Because the scope of the project increased rapidly as I was running it, I did not decisively decide to rip out all the heating/cooling plant during the initial demo. I thought some things could potentially be reused. That was a mistake.
As was said, determine your new actual loads. On the heating side, my boiler was dramatically oversized, even with a substantial addition. Not to mention, some of my existing radiators were "sun rads" where the radiator was inside the wall cavity, and therefore basically 50% of its output was going out the exterior wall.
Starting fresh, I think heat pumps are best in our climate. One set of equipment. One set of transport.
I appreciate it. We have the same steam rads. They are sunk into the wall. It is a 1938 home and the system works well for most parts of the home but it is temperamental. With new natural gas delivery fees it is quite expensive to run the steam system. It is one reason why switching to electric (for all intents and purposes) makes sense. I do not believe we will save annually on an apple to apple basis but it is greener and the steam system will fail eventually. To your point: one set of equipment.
Before you jump on the air to water band wagon let’s consider an A to A system in the existing AC ductwork that would heat the home 90% of the winter and keep the steam for the coldest few days of the year. This would get you 90% of the way to your goal for 25% of the cost.
From what I understand your part of the world has some very high electric rates. It seems unlikely that a fuel switch to electric could save you money period no less enough money to recover the very high cost of a A to W HP.
Boilers tend to simply just keep working with minor maintenance almost forever.
Walta
Valid points- thank you! Unfortunately the cost of using natural gas has skyrocketed given an uptick in the delivery cost. I do not believe there is a cheap utility anymore (gas/electric/solar/etc).
The issue is not maintaining the steam boiler, it is maintaining the supply lines for it. Corrosion plays into keeping it running appropriately. Balancing the system is very difficult and it is a whole house system. We are removing our fireplace and chimney as part of the energy retrofit so venting the boiler is now an issue.
We need to replace our second floor ac unit anyway given it’s on r22. Our decision comes down to air to water or air to air.
It's hard to beat air to air unless there's just no room for ductwork or other extenuating circumstances.
Appreciate it. The Air to Water than I am referring to is a monobloc system heat pump where all the refrigerant is located at the outside unit. The system dumps heat or removes heat from water with glycol in it. The water is then brought to an air handler that blows air over the heated or cooled water.
Since air to water is a niche product, you will find that the parts for such a system will cost two to three times as much as the parts for a similar air-to-air system that gives very similar performance: you have an air handler, a chilled medium flowing through it and air blowing over it goes in warm and comes out cool. With air-to-air you'll find a much better choice of sizes and configurations of air handlers. It will also be much easier to find an installer who is familiar with the system.
It's not that one technology is inherently better than the other, it's that one is more popular.
These are great points- thank you. What has drawn me to the monobloc system is mainly two reasons. First, all the refrigerant is located on the exterior of a tightly sealed home. My view is that leaks in a system are detected after the fact (not cooling/heating properly, etc). Second, the amount of refrigerant used is much less than a traditional heat pump. It is difficult to determine where heating and cooling equipment will settle out but towards electrification is obvious. That said, the constant changes of refrigerant used can make some equipment obsolete and when a unit breaks, both components may need to be replaced if a different refrigerant has been mandated in the market.
It is a reason why we may go with a split system heat pump for a WH as well. Some units use CO2 as the refrigerant which has a much lower GWP than all other refrigerants.
Air to water for cooling sounds good until you dig into the details. You can't just wing it, you need somebody to design it, outside of commercial world there are very few folks that do.
The 2nd part is installation. Again, not an issue in commercial world, but residential installers are not used to the attention to detail required. Main problem is any chilled part will sweat. This is not just you lines but valves, pumps and fittings. The insulation has to be perfect not just haphazardly installed like you see over heated pipes.
The cost DC mentioned above is real. I can get a a full air to air ducted heat pump for less than the cost of an FCU.
There is also an extra conversion as you are now going from refrigerant to water to air instead of refrigerant to air directly. This reduces the efficiency of the setup, oversizing can get it closer but will always be a bit less.
It looks like a number of industry players are going to R32 but that hasn't fully settled out yet. Overseas the market seems to have standardized on R32 for split systems, so there is a good chance that will be the norm.
Thanks and fair points to consider. Our contractor we are using does commercial as well as residential. He just completed a commercial job that included geo thermal units.
That said, would it be better to go with air to air but future proof with roughing in, or at least pre plan, for an air to water system?
The refrigerant is my biggest concern. We are doing a complete mechanical overall. Right now, our r22 outside condenser broke and we had to replace it with an off brand to just keep the ac running. It is something I do not want to bother with down the road. Again, we are not keeping the r22 unit.
I believe that air-to-water is fundamentally technically superior to air-to-air, I wrote a piece about it over two years ago:
https://www.greenbuildingadvisor.com/question/are-air-to-water-heat-pumps-about-to-boom
The problem is it's still a niche market, and the cost of equipment is high and the selection is poor.
You see a similar dynamic with ground source heat pumps. No one disputes that theoretically they are capable of higher efficiencies than air source. But since it's a niche market it's a backwater, innovation comes to air source first and the result is in real world products the air source efficiency is close to ground source.
As examples, modulating compressors have been available in air source for decades, we're just starting to see them in ground source. Similarly vapor injection is an established technology in air to air and it's still rare in air to water.
I see, thank you. After a quick Google search, there is not much of a delta between, say, a 3.5 ton Mitsubishi Hyper Heat Pump system and an Apollo (for example) unit. I could be missing something. That does not include the buffer tank, AH and smart housing (water distribution system).
It is difficult to determine where the bigger players (Mitsubishi and the like) are in terms of air to water systems for residential. Mitsubishi offers it for commercial use so my gut tells me residential is not too far behind. Obviously, the current political climate plays a role with government incentives (or lack thereof) for greener methods of heating and cooling as well. That may hurt adoption and marketing efforts.
It is a difficult decision.
A 3.5T Mitsubishi Hyperheat is $8,813 at HVACDirect (https://hvacdirect.com/mitsubishi-42-000-btu-19-seer-tri-zone-heat-pump-system-12-12-12-id7771.html )
I see the 3.5T Apollo as $4249 from MBTEK (https://www.mbtek.com/collections/heat-pumps-air-to-water/products/apollo-heatpump-42k-btu)
Now, to do a comparison, you have to include all the things that Mitsubishi includes that the Apollo doesn't. First off, it has three heads, so you need three fan coils, which are going to be at least $750 each and possibly more. Plus circulators, a buffer tank, zone valves, piping, insulation, purge valves.
So maybe they come out even. But there are two big difference. The Mitsubishi Hyperheat is vapor injection, it has much better cold-weather performance, specifically in the amount of capacity it retains. At 10F that nominal 42,000 BTU/hr Apollo only produces 22,000 BTU/hr of output. The same nominal size Mistubishi produces 44,000 BTU/hr at 5F.
Also, the Apollo is a bargain brand. If you look at a more established brand like Chiltrix -- which still doesn't have the reputation of Mitsubishi, and also is not vapor injection -- the CX50, also 3.5 tons, is $8999 for just the outdoor unit.
I appreciate it. I know the argument is to stick with the traditional heat pump. My climate change concerns are, I think, causing me to want to go with the air to water. That said, you’ve made a compelling argument for the traditional route. I think going traditional with future planning is probably the right call. The deep energy retrofit is reducing our carbon footprint to begin with.
DC when we look at the COP charts for every unit one of the constants is the greater the difference between the coolest and warmest point in any system the lower the COP for that system will be at that time.
This fact puts the A to W at no less than a 20 degree disadvantage on the COP charts because one needs the water to be warmer than the air in the house will be. About .5 COP across the board.
The ground source systems W to A use this same fact to get higher COPs as the coldest point in their system is now about +30F instead of -15F.
Walta
Just looked up the Mitsubishi Hyperheat, which I consider to be a premium air-to-air (https://ashp.neep.org/#!/product/112110/7/25000/95/7500/0///0 )
At max output, at 47F, COP is 3.69. At 17F COP is 2.02
Compare to a similar-size Chiltrix (https://www.chiltrix.com/CX50-air-to-water-heat-pump/ )
At 47F, COP is 4.6 with 95F output water, 3.7 with 113F output water.
At 17F, COP is 3.2 with 95F output water, 2.5 with 113F output water.
It's not uncommon for these systems to use 95F water under floors, 113F water in fan coils.
Now, there's a lot of reasons why this isn't necessarily a fair comparison. First, the Mitsubishi is vapor injection, which trades lower COP at cold temperatures for higher capacity. And Chiltrix doesn't publish part-load COP numbers, which is where the Mitsubishi really shines. But at first glance the numbers don't support the conclusion that air-to-water automatically means lower COP.
I'll see if I can dig up numbers for Arctic, which is a vapor injection air to water.
OK, the Arctic website (https://www.arcticheatpumps.com/specifications.html ) only has ratings at 7C, -7C and -20C -- 45F, 19F and -4F. So it's hard to do a direct comparison. But the COP's they give for those three temperatures are 3.1, 3.1 and 1.7. That's with water temperature of 45F (113F).