Radiant Heat Floor and Heat Pump AC
I’m working on AC/Heat for a 2190 sq ft house and had a Manual J load calc completed that had the following summary info.:
1. Total building supply: 669 CFM
2. Volume ft cubed: 19,710
3. CFM Per Sq Ft: 0.306
4. Total Heating Required Including ventilation Air: 18,017 Btuh
5. Total Sensible Gain: 13,893 Btuh
6. Total Latent Gain: -1,133
7. Total Cooling Required Including ventilation Air: 13,893 (1.16 Tons ~ sensible+latent)
A local HVAC team is recommending:
Heat Pump: Mitsubishi SUZKA18NA2
Air Handler: Mitsubishi SVZKP18NA
1. If I plan to install radiant hydronic heated floors using boiler; does it make sense to use a heat pump that heats and cools? It seams like getting something that focuses on cooling would be more efficient. A little back up heat in the system might be helpful, but not full heating capabilities.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Looking at this from the other direction, if you have a heat pump then the boiler/hydronic in floor system is pointless and a waste of money.
You shouldn’t need both a heat pump and a boiler, I agree with this_page. The easiest option would be a heat pump that can deliver both heating and cooling. If you are set on in-floor radiant heat you will be disappointed in its cooling capabilities.
I will add one thought without any personal experience here: if you are open to considering a radiant panel ceiling instead of flooring, there is some evidence that a radiant ceiling panel would be capable of both cooling and heating. It would require an air-to-water heat pump though which are a bit “experimental” and are far from mainstream in North America. See the quote and link below from Tom Tesmar:
“Just as the overhead rays of the sun are absorbed by the beach sand, radiant ceilings warm the floor. There are no cold floors in radiant ceiling projects.”
The reason they call it "air conditioning" and not "cooling" is that it needs to dehumidify too. Radiant panels have no ability to remove condensation. You need some sort of air handler to remove humidity.
The Manual J says latent load of 1133 BTU/hr. It takes about 1000 BTU to remove a pint of water from the air. So that's about 26 pints per day, over three gallons. That's a lot of water to be dumping into your house.
DC - I'm not 100% on HVAC ins and outs, but that was a negative (-) 1,133 Btuh. Does that make a difference. They did say that my area/city has very dry air.
I'm trying to wrap my head around what a negative latent load would even mean.
It may be that it is the result of the sum of a few factors (?), one of which is a credit of some sort, whose sum ends up negative. Attached is the excerpt from the Man J.
I think all it really means is to hit the desired indoor RH you have to run a humidifier.
If the exterior air was 25% humidity and the interior desired humidity at 50%, any air leakage would be a negative latent load.
Based on that - you may be in a really ideal area to use radiant cooling. Not sure if Idaho ever gets much humidity, so not much to worry about. You could add a single 6,000 btu mini split in a central part of your house if needed for a little bit of humidity removal. Then your floors are used for a distribution of sensible cooling.
You might want to look at Small Planet Supply for Radiant Panels for ceilings https://www.smallplanetsupply.com/messana-radiant-heating-and-cooling.
Consider designing a Heat-Pump for Stage 1 and a Boiler for Stage 2. The boiler should only be used where the heat pump can't keep up to the demand. Be sure to calculate in DHW loads and ancillary loads (Garages) when doing the load calculations.
For the ceiling panel to be radiant, it needs to be high temperature. This is fine with electric and it works great.
With low temperature water, which you need for an air to water or condensing boiler, the ceiling panel will never get hot enough to radiate. It still provides heat to the house but don't expect magical comfort.
Ceiling panels also put out half the heat of the equivalent area of floor heat. This means much larger surface area or running them hotter.
Electric ceiling panels do work. Smaller ones over a big window can heat the local area and help offset night time radiation losses.
There is also no benefit with radiant ceiling cooling. The cooling issue in north America is humidity removal, not sensible cooling in most cases. Floor/wall/ceiling panels can't provide any latent cooling so you always need an air handler. The air handler required to deal with the moisture, even a low SHR unit, still provides a fair bit of sensible cooling. The amount of sensible cooling left over is so small that adding in the complexity of radiant cooling makes no sense.
Of course that is a lot of hand wavy arguments from my side but I welcome you to crunch the numbers in any A or B area and post them.
Okay, I’ll bite.
> “Ceiling panels also put out half the heat of the equivalent area of floor heat.”
I don’t have personal experience with either of these products, but according to their spec sheets this Messana radiant ceiling panel outputs 26.9 BTU/hr/ft^2 with 100 degF water and 70 degF room temperature. The spec sheet for Warmboard S at the same 100/70 degF shows only 25 Btu/hr/ft^2 assuming hardwood floors with an R-value of about 0.75.
I think it is also worth pointing out how many more obstructions a typical home has along the floor than it does on the ceiling. Besides lights and ceiling fans, the ceiling will have much more available surface area for both heating and cooling. Think about your bedroom. Laying in your bed at night which radiant heat source do you think would be more effective?
Warmboard: https://www.warmboard.com/wp-content/uploads/2022/03/InstallGuide_Warmboard-S.pdf (skip to chart on page 37).
Seigenthaler says slightly under a third for ceilings -- 0.71 BTU/hr/sf/degree, vs 2 for floors. I don't recall a source being cited by him.
I have radiant ceilings in my house and measurements of water flow and temperature delta tend to agree with what Seigenthaler would predict. I will also say it is just the most unobtrusive heat. You can't tell it's on, the room is just comfortable.
The heat loads in a well insulated and sealed room are so low that no matter what you heat with it will be comfortable. It might as well be heated door trim, would be just as unobtrusive and comfortable.
Degrees Fahrenheit or Celsius?
2.0 BTU/hr/ft^2/degree would mean 60 BTU/hr/ft^2 at a 30 degree (Fahrenheit) delta?? Why would Warmboard intentionally publish a value far, far less than that if it weren’t true? This 2.0 factor seems fishy.
Fahrenheit. You'd go crazy using BTU with anything else.
A 30F delta would be floors over 100F, they'd be unwalkable. I'm talking floor surface temperature, which will be quite a bit lower than water temperature.
What does WarmBoard publish?
I agree with Akos. I am Zone 6A - Minnesota. The heat loss modeling that I am doing on a new house design shows a CERV2 with a dehumidifier handles most of the comfort issues. Could also use a Minotair. For the great room and porch, a Jaga Designs Briza (Hydronic Mini-Split) handles any cooling load for the house - without major bucks on ductwork.
I would look at it this way. A heat pumps costs a couple of extra dollars to install, compared to overall house build cost, this is noise.
This gives you the option of either heat source down the road. Depending on how energy prices go, you can select the cheaper one to run.
With a heat pump you can always install a PV array later to offset the cost. Can't do that with gas/propane.
I generally find that minimal floor heat (say a couple of degrees above room temperature) plus a heat pump for bulk of the space heating tends to be most efficient option. The bit of floor heat takes the edge off the colder floor without adding a lot of BUTs to the space.
First, congrats on such a low load home!
I’d rethink the boiler plan. You’re looking at an extremely expensive and complicated heating system when a heat pump is essentially free. Also, since the heat loss is so low, the floor won’t be warm!
A good solution might be electric radiant heat in the bathroom tile and kitchen tile.
Thanks paul_w, we've made every effort to design with cost/efficiency in mind. It has been a long road, with much input from this forum. Now, we just have to get it built. :)
Good ideas. What about something like a variable-speed, standard AC wouldn't that work in the summer to cool the house with the radiant handling heating in the winter. That seems like it would work better.
Akos - I would agree that it might be 'noise' (depending on cost); especially if I need some back up heating capacity. I think that I will ask them to quote a variable-speed AC and compare them with a known cost difference.
There’s just such a small delta between a variable speed HP and a variable speed AC that if you asked for the AC and installer only had a HP on hand they’d probably just charge the same.
I have both heat pumps and floor heat at home.
More than ones I've used the heat pump to bump up the house heat as the floor heat is slow. In my case the heat pump VS AC only was about $200 on the equipment, it was money well spent.
Your comment here is bang on accurate.
I monitored data over a 4500 square foot commercial space with radiant and air handlers for AC/HRV and upper floor heating/cooling. (9000 sq/ft total over 2 floors). 9 Ecobee stats were installed in total to cover the 5 zones of radiant and 5 air handlers in the building. The retrofit included continuous interior insulation and insulation/over-pour of the existing slab. It is a very efficient building. With all the data available from the Ecobee stats, a few things were pretty obvious after five years.
In retrospect the radiant on the main floor was a pretty massive waste of cash as we needed the air handlers running for fresh air in winter anyway. It was very easy to overheat the space, even in winter at -20C, as staff activity and sunlight would create a thermal bump around 1-2 pm. It was more efficient to bump the temps in the morning via the air handlers, and leave the thermal mass in the slab a bit cooler to moderate temps later in the day. I experimented with aggressive setbacks at night (with demonstrated reduction in gas use vs radiant) which again is not possible with radiant over a short period.
Would I use radiant again? No. Insulate an existing slab (with no insulation)? For sure.
When radiant floors first hit the scene it was believed that "thermal mass" was a good thing and concrete was pushed as the ideal floor material. It turns out that the opposite is true, you want the radiating surface to have as little heat capacity as possible so that it can respond quickly to changes in the demand for heat. More modern systems like Warmboard pride themselves on being fast responding/low heat capacity.
We only had 2” of concrete over 2” foam on the over pour which did react relatively quickly. It was still a very large expense in the mechanical budget that I would skip time around. The hydronic air handlers provided heat, AC and fresh air and were very efficient combined with the triangle tube gas unit that provided hot water.
There is no reason you couldn't use radiant heat in the winter and a heat pump or "regular" A/C for cooling in the summer. The only issue is that at least with a heat pump, you're essentially putting in heating twice, since the heat pump could do heating too, but you wouldn't be using that function if you also have a radiant system as your primary source of heat.
My personal opionion here is that you if you're set on a radiant system (and they are comfy!), go for it. You can put in a heat pump for cooling and keep the heating side as a secondary heat source, or play around with the two for best balance of operating cost vs comfort. There is no reason you can't set things up that way if you're willing to spend the money to have both systems installed. The heat pump won't really hurt you in terms of operating costs when it's not being used to heat the home while the radiant system is running. The downside is all in upfront install costs.
An A/C only system is probably a little cheaper to install than a heat pump, but I'd price that out -- if the cost difference isn't too bad, having a heat pump instead of a basic A/C only system would give you more operating flexibility during the shoulder seasons.
Thanks Bill. That all makes sense. I think I'll wait to make a final decision until I have the standard AC cost quote back. Then, I can make a more informed decision about whether the heat pump w. radiant floors is a viable options.
This is a question that comes up a fair bit. The basic strategy is to install a heat pump sized for your cooling load, and then use radiant to make up the difference between the heat it can produce and what you need.
The answer depends a little bit on how cold it gets where you are. The NEEP database has ouput numbers for 1300+ heat pumps at 17F and 5F. The median unit on the list delivers 99% of its rated cooling output as heat at 17F, but only 80% at 5F. So if you were to get a heat pump that was sized for your cooling load (15K BTU), if your design temperature is 17F it would be 3K short of your heating load (18K). But if your design temperature is 5F it would be more like 6K short of your heating load.
Then you put radiant wherever you want warm floors. You need to add enough to make up the gap, but you can add more if you want. I'd start with bathrooms. The back-of-the-envelope formula is heat output =(temperature difference between floor and room)*(floor area in square feet)*2 . Floor temperatures above about 85F become uncomfortable so figure a maximum of about 30 BTU/sf. One thing you'll run into is that in a tight house large areas of toasty-warm floors will overheat the house, you end up turning down the temperature which doesn't give the toasty effect you were looking for.
What you'll want to do is use a thermostat that prioritizes the floors, so the heat pump only runs when the floors full-out aren't enough to warm the house.
With that said, here's the catch:
There currently isn't any technology for delivering heat through floors that is low-cost to run and reasonable cost to install.
There are basically three ways to heat water: electrical resistance, electrical heat pumps, and burning fossil fuels.
Electrical resistance is always going to be the most expensive of those three to operate. How much more expensive depends on how much electricity and other fuels cost where you live. Note that if you configure your heat as I recommend in post #18, the floor heat will be your primary heat source and the heat pump will back it up, which negates quite a bit of the efficiency advantage of a heat pump. If you decide that the cost of resistive heating is acceptable to you, I would recommend strongly looking at just putting the resistive wiring in the floor, it's cheaper and easier to control than any form of hydronics. The disadvantage is that the in-floor resistive is probably less durable, and it's not as future-proof -- if you have piping in the floor you can change your mind in the future about how it's powered.
If you're looking at fossil-burners, a load of 3K, 6K, or even the full 18K is really small for a conventional boiler. There are gas water heaters, both tank and tankless, that are rated for space heating, I'd probably recommend plumbing in with the hot water for the floor heat.
In terms of heat pumps, there are air-to-water heat pumps that are well-suited to your application. The drawbacks are that they're kind of pricy, and it's an emerging technology, none of the companies supplying them are really established companies. It's going to be hard to find someone to install it. If you go that way there are air handler units that essentially work like minisplit heads but get hot or cold water from the heat pump for heating or cooling. So you install enough air handlers to meet your cooling needs, as much radiant floor heat as you want.
Both SpacePak (Mestek) and Taco System M are well established. If Veissman heat pump become available, with purchase by Carrier, A2WHP becomes much more viable.
I think we're still a bit away, but I believe that we're on the cusp of an explosion of AWHP availability in this country. From what I've read they're huge in both Europe and China. The Taco System M doesn't have a track record in this country but is an import of a German model by the German firm Glen Dimplex.
Hi WD: another alternative could be a two zone system, using two Carrier Infinity heat pump mini splits indoor/outdoor 40MPHAQ09XA3/38MPRAQ09AA3 units. High efficiency at 42 SEER cooling and 15 HSPF heating. 9K BTU/Hr cooling rated capacity each. 13K BTU/Hr heating capacity at 5F and 8917 BTU/Hr at -13F each, should be adequate plus there is a reasonable turndown to 3100 BTU/Hr. The heating COP remains a good 2.78 at 17F.
Thanks for the input jj1,
The biggest problem for mini-split systems for my house is that even though there is a low load, it is spread across 2100 sq ft. That makes placement and air circulation very difficult to achieve without the use of a central air handler.
Currently, I'm considering the equipment that I initially listed and using the radiant heated floors to augment where necessary--for code compliance. In reality, we'll likely be able to set the radiant floor as the primary system (since the heating requirements are so low for MOST days) and use set the heat pump/air-handler as backup.