Air to Water Heat Pumps for Radiant Heating And Cooling

One of the things about humidity control is that humidity is really hard to model. Heat loss is easy, sensible gains are easy although solar can be tricky, but I'm not aware of any methodology for modeling humidity.

Counter to the folk wisdom, the hottest days aren't the most humid. At least around here, the most humid days are when it's not super hot. We had a day this summer when the dew point hit 77F: outdoor temperature of 78F, raining, outdoor humidity 99%.

If your thermostat is set at 77F, on a day like that your AC is hardly going to run at all. It's rainy out so there's hardly any solar gain, and the outdoor temp is right around the thermostat setpoint. There's almost zero sensible load, so the AC doesn't run. But the humidity is thick.

Hence the popularity of dehumidifiers. But the thing about dehumidifiers is they are quite efficient space heaters, they return to the space all of the latent heat that is extracted by removing humidity, in addition to the heat from the energy required to run them. So not only do they extract humidity, they also provide a sensible load that encourages the house AC to run more.

Now, from an energy perspective running a dehumidifier just so that you can run a 100% SHR cooling source like radiant cooling makes no sense. It's like leaving your windows open in the winter to get your radiant heat to run more so you can have toasty toes.

I don’t like standards based on features and not performance. So I reject the European method. Moreover, what makes this different than air-to-air?

@John Ruhnke

I put together a sample of what I think is the most comprehensive way of looking at performance. It's based on the performance data for the Arctic 050ZA that you sent me, with the temperature distribution data for Washington, DC from NEEP.org.

It's based on a theoretical house with a 30,000 BTU/hr heating load at 22F, which is what NEEP has for the DC heating load.

For each degree of temperature I estimated heat pump capacity and COP. I assumed outdoor reset, water temperature of 95F at 5F and 86F at 44.6F. I just extrapolated from the full-load data give; NEEP has performance at minimum output which allows you to estimate part-load performance.

The model shows a weighted average COP of 4.6 for the heating season.

You can see it here:

https://docs.google.com/spreadsheets/d/1uxIOKJJKmOCMtxbWPZXU2WQRnSwHColipUuQ5uPa_9M/edit? usp=sharing

I hope it's self-explanatory, if not please ask.

I'm with Kaos here. I spend an unnatural amount of time looking at heat pump performance curves. For the most part they're straight lines. I know the Mitsubishi vapor injection models have curves with a kink, it's really two lines depending on whether they're in vapor injection mode or not. This is true of both the capacity curve and the COP curve. In vapor injection mode the capacity curve is flatter, more capacity is retained. But the COP curve is steeper, COP suffers.

I see there are now a number of Air to water HPs available including Mitsubishi.

I have a radiant floor heat system (staple up) and low temp baseboards that operate well with a max HWS temp of about 130. This is now within the range of some of these HPs.

My remaining issue is water quality. Some iron in the water and ph 6. Also the radiant tubing is effective but has no O2 barrier. With water like this, a plate Heat Exchanger in the HP is going to clog up quickly. I may be able to fill the closed loop with pristine water and even keep topping it up with same. If I can do that, is the lack of an O2 barrier going to still cause problems for the plate to plate Heat Exchangers?

If so, are there some HPs that use a coaxial or other more forgiving Heat Exchanger?

Any recommendations of how to proceed? BTW probably need 32 - 40K heat at design temp.

thanks hot_rod,

"Protector F1" — It sounds like you think this will keep the water clean enough to avoid a problem with clogging a plate heat exchanger, correct?

Also, the Ecodan has apparently been recently introduced for sale in U.S. I expect to be quoted on it shortly. From the curves I've seen on other systems, I agree that COP and capacity do look better at 120 hws. From the econdan calculator I can't see COP but I do see the 4 ton unit is supposed to put out 29K BTU at 130 but jumps to 42K at 120 under certain conditions at 2 def F outside DB.

It would be possible to increase emitter area to probably make it all work with 120 deg HW. Encouraging!

thanks Bob, that is is very helpful.

So we would clean the whole system with detergent, get out as much crap as possible.

Then run better water through it to flush it out as much as possible.

Then remove all that water and put in clean, purchased water, then treat that water with F1 or similar. TEST.

Ideally, then we would have some way of adding the small amounts of required make up water from a small tank of clean water as well. Then keep testing annually or on some schedule.

thanks bJohnhy

I looked over the filters in my case I assume I’ll be looking for possible products of corrosion caused by the no O2 barrier tubing. Plus any old junk that wasn’t loosened at the initial cleaning.Do you think the 140 mesh would be the best size to use?

Thanks again, this is a great add to the above system.

Yes, this looks very good!

BTW could I have a significant amount of water stored in a large expansion tank that could be my reserve to add more clean water as needed? This could expensive feeder tank.

If so, how much water might be stored here —more than one or 2 gallons?

yes, that’s the idea. The system still has the existing pressure reduction valve for the original make up water connection piped to the house system.

I guess I would just need to take it off of the house water and feed a new expansion tank into the existing pressure reducing valve at 12–15 psi.

Sounds like this part could be a fairly inexpensive change, great!

just to be clear, this would be a split system with water only inside the house and refrigerant piping going to the outdoor unit, like a typical air to air heat pump.In that case would we still be OK sticking with treated but not glycol water ?

Thanks for any ideas.

Are you doing cooling too? Here's the problem I've run into: you can't let the heat exchanger freeze or it will crack. On my unit, a Chiltrix CX34, the default lower limit is 35F, if it goes below that it throws an error and shuts off. I found that if I set the aquastat below 50F the temperature would occasionally dip below the lower limit and the unit would shut off. But with the aquastat that high the fan coils weren't as cold as I wanted to get as much dehumidification as I wanted.

The only safe way to set the lower limit lower is to have a glycol mix.

When you're doing cooling condensation is a major issue, every pipe, fitting, valve, etc., has to be insulated. And you have to do a much better job than with heating, the insulation has to be 100% and airtight.

So if you think you'll be doing cooling in the future, think about which parts of your system will be carrying cold water, and make sure you'll have the ability to insulate them. You'll probably want to separate the hot-only part of the system from the hot-and-cold part close to the heat pump, and have the ability to isolate both with valves on both the supply and the return.

thanks!

@hot_rod outlined it well above. I would go with distilled water for your final filling of the system. Although the spring water could be good water relative to some city supply, go with distilled if your going through that effort.

Nice concept article below, although subscription website, I believe you get to read 1-2 articles for free. Distilled water is even better than the "demineralized" water referenced in artical.

https://www.supplyht.com/articles/99832-its-time-to-get-serious-about-water-quality-in-hydronic-systems