Hydronic Heating System Used for Cooling
I’ve been reading and posting here for a while, and a question that comes up a lot is whether a radiant heating system can be used for cooling. John Siegenthaler has written extensively on the subject (see : https://www.pmmag.com/ext/resources/PM/2014/June/016-020-0614-Siggy.pdf for instance) but he admits that the control of such systems is problematic. I’m particularly interested in this because I’m working on a house right now where hydronic cooling would fit really well. I’ve come up with an idea for a simple control mechanism that uses off-the-shelf parts and I’d like to bounce it off of the community.
Air conditioning is called “air conditioning” and not “cooling” because it does two things, it cools the air but it also removes humidity. Depending on the occupant load, outside humidity, and tightness of the envelope a house can need three to five gallons of liquid water removed from the air each day. The way that water is removed is by running the air over a cool surface, which causes the water to condense. The key problem with radiant cooling is humidity removal, (or “latent heat removal” as those in the biz say). A radiant tubing system has no way of dealing with the condensate, so it cannot be used to provide humidity removal.
Humidity removal can be done with a hydronic coil with a blower fan and a drip tray to catch condensate. Examples are the Carrier 42WKN (https://www.carrier.com/commercial/en/us/products/airside/fan-coils/42wkn/) and the line from PHNIX of ultrathin fan coil units (https://nlsolarheating.solartubs.com/ultra-thin-hydronic-fan-coil-heating-cooling-water-radiator-100v-p-357.html) which are available under a variety of different names. Similar units are also available from Myson and SpacePak. (A word on nomenclature: A hydronic system is one that uses water as the heat conducting medium. I’m going to use the term “hydronic blower” to refer to units like these, and the term “radiant loop” to refer loops of PEX embedded in the building. I hope that avoids confusion.)
When air is run over a cold surface to remove humidity, the air itself is chilled as well. The fraction of the cooling power that goes into cooling the air – sensible cooling – as opposed to humidity removal – latent cooling — is known as the Sensible Heat Ratio, or SHR. The SHR of a hydronic blower is determined by the water temperature; the colder the water the lower the SHR (ie the more humidity is removed). If the water is at or above the dewpoint of the incoming air no humidity is removed, the SHR is 100%. As a practical matter you can’t run the water any colder than about 35F without ice forming; at 35F the SHR might reach 65%. A radiant loop has no humidity removal capacity – there’s no place for the water to go – so it has a SHR of 100%, and has to be run at a warm enough temperature to prevent condensation from forming (ie at or above the dew point of the surrounding air).
When calculating the cooling loads for a building both the sensible and latent loads are calculated. The principal sources of humidity are the occupants – breathing, bathing and cooking – and infiltration, hot humid outside air leaking in. The SHR for a residence might range from 85% to 95% depending on the number of occupants and the tightness of the construction. My idea is to combine a hydronic blower and a radiant loop in each zone with an appropriate control system to provide the proper balance of sensible and latent cooling.
The hydronic blower is sized to provide all of the latent cooling that is required for the zone. It will also provide some sensible cooling. The radiant loop will be sized to provide the remainder of the sensible cooling. They also need to be sized to meet the heating load as well. Hydronic blowers are much more forgiving of oversizing than conventional air handlers so it’s OK to build in some margin. In the house I am working on the overall SHR is 93% and the cooling load is slightly higher than the heating load. I plan to run the water for my blowers at 45F, which gives a SHR of 70%. To meet the overall SHR the hydronic blowers have to be about 25% of the capacity of each zone. The PHNIX hydronic blower units come in capacities as small as a quarter ton so in zones under one ton they will account for more than that, which is OK.
Note that the hydronic blower and the radiant loop require different water temperatures. The blower should be as cold as practical to maximize humidity removal, while the radiant loop needs to be warm enough to avoid condensation at the set point. It would be possible to use mixing valves to provide two different water temperatures. However, in my house I need to have two heat pumps anyway, so it’s just easier to run them at different temperatures and have the blowers and the coils on completely separate circuits.
Control is provided by a two-stage thermostat with humidistat control. “Two-stage” means it is designed to control a system with two levels of cooling. It has two outputs, one for primary cooling and a second set a degree or two above for secondary cooling. If the primary cooling is unable to maintain the set point the temperature rises and the secondary cooling is called. The humidistat has an output which turns on a dehumidifier if a humidity set point is exceeded. The control circuit is set up this way:
· Radiant loop: On when primary cooling on and dehumidifier off
· Blower unit: On when primary cooling on and dehumidifier on; OR secondary cooling on
The radiant loop cools when cooling is needed and the indoor humidity is low enough for it to run without condensation. The blower unit runs under two circumstances, when either the indoor humidity is too high for the radiant loop to run, or if the cooling load is too high for the radiant loop to satisfy on its own.
In heating mode the system is configured as a standard two-stage system, where the radiant loop is the primary heat and the blower unit is the secondary heat.
Please let me know your thoughts. Thanks.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
So two hydronic systems, one for radiant only (higher water temps), one for fan coils (lower water temps).
That would work for heating with radiant as primary, blower as secondary adding heat only when needed.
The question is whether a two stage controller can be adapted for independent control of humidity. Seems like you have more logic here (two parameters, temp and humidity) than a regular controller would allow as inputs? I think this is the domain of controllers like Tekmar 406 radiant controller (with mixing valve and dew point sensor), which I know that Chiltrix recommends.
Since you are going with separate systems, and presumably will also have ventilation, have you considered doing your dehumidification using something like the UltraAire SD12? Needs ducts, of course, but only one drainage location and filter to change. And gives you fresh air ventilation as well (although not ERV style).
Seems there the control trick is having a way for a humidistat to turn off (or raise the temperate of the radiant cooling system above the dewpoint), were dehumidification to fail.
An Ecobee thermostat does dual-zone heating and cooling and humidistat control.
My initial thought was to try and use the ventilation system for humidity control. But it didn't pencil out. The problem is you never get latent cooling without sensible, so the humidity control ducts have to be sized for at least 25% of the load. That's a lot bigger than would be required to meet the ventilation needs. In fact you could argue that a 25% system costs almost as much as a 100% system.
John Siegenthaler has an article where he muses about running a dehumidifier and running the output through the ventilation system. From an airflow perspective that would work. But it would mean cooling air and then reheating it, and then cooling it again. Doesn't seem very efficient.
>An Ecobee thermostat does dual-zone heating and cooling and humidistat control.
I should add that it would need a few relays to give the logic I describe. I've attached a schematic that shows a four-relay circuit. This is conservative that ensures full isolation. If the thermostat outputs don't mind being connected to each other it could be done with a single SPDT relay on the humidistat output.
Interesting. Doesn’t a non-split dehumidifier e.g., 98H do latent cooling but with sensible heat production? So that wouldn’t require larger ducts than ventilation. Seems to me that is what the Messana “Neutral temperature dehumidification” is doing as well.
But I’m high jacking your thread. Afraid I can’t read the relay circuit. Put in in Python code and I could :)
Python version of the diagram:
if(SecondHeat || (FirstHeat && Humidistat))
if(FirstHeat && !Humidistat)
A dedicated dehumidifier like the 98H removes both latent heat and sensible heat, then dumps the sensible heat plus the mechanical heat generated in the process back into the exhaust air. From an efficiency standpoint it doesn't seem very green to remove that heat and then add it right back instead of dumping it outside the building. That heat is going to be something like a quarter of the total sensible heat load so it's pretty significant.
Let me also add that the hydronic blower units could be replaced with minisplit heads. Since it's a separate system it would in fact just be a drop-in switch. Cost seems to be about the same. I prefer the hydronic blowers for two reasons. First, they come in sizes as small as a quarter ton, about half of the smallest minisplit I head I could find. Second, if your goal is to minimize SHR, it's easier with hydronics, just drop the water temperature.
One option to simplify the setup is to feed your floor cooling loop with the outlet water from the hydronic coil. Provided you size the coil right and have the correct flow rate, the outlet water from the coil will be at house dewpoint.
You can than take this water and mix it with say 15% return water from your floor loop and it will always be above dewpoint. Not electronics or controls required.
P.S. If you are using a mini split for dehumidification, might as well use it for cooling. Any unit that hits the SHR required for humidity control of a well sealed house already provides enough sensible cooling. No need for extra sensible cooling.
Thanks Akos. I did think a bit about running the water through the hydronic coil to temper it before running through the loop, but I couldn't get it to pencil out, maybe I need to think about it more.
One of the things they talk about on the Chiltrix website is that the heat pump efficiency is better the warmer the water temp. So they recommend against using mixing valves, as it's less inefficient to cool the water and then warm it up than to have it only cooled to the temperature you want. Hence my design where the colder water is only used when dehumidification is needed and warmer water is used for sensible cooling.
But I will think about it.
You might to look at FanHandler, It looks like they are modifying the blower controls to control the air movement over coils. Hi velocity.com has their hydronics air handler with a programmable controller WEG where you can modify the ladder logic to run at a lower setting. I am looking at installing sensors in the supply as part of a feedback loop.
The FanHandler looks interesting. It reads both temperature and humidity and tries to adjust the fan speed accordingly. I believe that this is the future, soon it will become the norm for control systems to react both to temperature and humidity.
It seems like it is only for conventional HVAC systems and only modulates the air flow. Chiltrix makes a similar controller they call the "Psychrologix" which controls both water temperature and indoor fan speed to match the need for sensible and latent cooling. Link is at: https://www.chiltrix.com/documents/Chiltrix-Psychrologix-TS.pdf
I had initially thought of using a Psychrologix controller on my fan coil units. However, it was pointed out to me that I only want to be using the hydronic blowers when humidity removal is needed, and that if I set them for maximum humidity removal and controlled them with the humidistat I would achieve the same goal at less complexity (and cost).
> recommend against using mixing valves
If need two temperatures (three if you want DHW), you can do it most efficiently with a single heat pump that switches between two different temperature buffer tanks.
Designing an optimal system is a good exercise, but beware of cost/complexity issues making it non cost effective.
That's how the Chiltrix DHW option works. In cooling mode it switches to heating mode and mechanized valves switch the output from the hydronic system to the DHW tank.
I have radiant cooling in my personal residence in western WA. It's been in place for about 6 years and has worked well. The system mainly uses the loops running through the floor and we have loop panels in one ceiling. All of the floors are polished concrete though there are area rugs. Heating and cooling of the fluid is from a Daikin Altherma and the all the controls are Tekmar. Humidity is monitored by the Tekmar controls so that the surfaces remain above the dewpoint to prevent condensation. The conditioning also includes an ERV.
In the last couple of years, we've run dehumidification simply because we like inside humidity ≤ 50%. The dehumidifcation allows having even cooler temps inside if desired.
How is your dehumidification done? Is it a hydronic fan coil controlled by the Tekmar?
The only integrated dehumidification is a coil chiller on the ERV intake that runs only when the cooling is on. That’s how we ran the system for several years.
We now run a standalone dehumidifier to get the indoor humidity below 50%.
I probably should have added to my write-up that if you live in a climate where humidity control isn't necessary then this is all superfluous. Where I live the outdoor dew point on the design-day is in the low 70's. Lots of humidity removal is necessary to keep it from raining indoors.
DC, the guys at Positive Energy in Austin have done a radiant cooling system for their shop.
Doing radiant cooling in the DC area would indeed present extra (humidity) challenges. Why not have a separate system for controlling humidity, which is needed much of the year in DC even when cooling isn’t required.
I do have a separate system -- the fan blower system. In fact, what I really like about this design is that it separate latent cooling from sensible cooling. In a conventional HVAC the thermostat controls sensible cooling and the latent cooling you get is whatever the system happens to produce.
If you're asking why I don't have a dehumidifier, unless you're in the rare situation where you need dehumidification and heat, dehumidifiers are a problem because they produce a lot of heat. They're actually very efficient space heaters, with a COP in the 2-3 range. All that heat then needs to be removed by the AC. Better to just remove it with the AC in the first place.
I have thought that if the sensible heat ratio I need were to dip lower than the fan unit can provide -- say if it's 70F and 100% RH -- I could run cooling through the fan units and heating through the radiant tubing at the same time. I could effectively drive the SHR to zero.
So what is the projected additional cost to do latent removal with the dual hydronic system and how many $/year will it save over conventional dehumidifiers?
I guess the question is compared to what.
The system I've penciled out has about $3000 worth of hydronic blowers. They replace the heating/cooling capacity of 1500 square feet of radiant tubing. There's no way I could get it installed for $2 a square foot so I'm actually saving money. I have to add in a few more control elements like valves and relays but I think I'm still ahead.
On my design day a dehumidifier has to remove 1.7 pints per hour or 40 pints per day. A 40-pint portable dehumidifier that is Energy Star rated uses 580 watts. In addition to the cost of the electricity to run the dehumidifier that 580 watts is added to the heating load, which increases it by about 2,000 BTU/hr, or about 6% in my case.
The efficiency of air-source heat pumps is dependent on the water temperature. Having to run one circuit cold so it can dehumidify imposes a performance penalty. However, the other circuit can be run warmer, so I'm not sure how large the combined effect is. I would be surprised if it is more than the 6% penalty of running a dehumidifier.
Overall I believe this approach either saves money or breaks even.
Hi there @DCContrarian I am also considering radiant cooling versus traditional HVAC/forced air conditioning for my project in the DC area. I'm having a hard time finding installers or projects to take a look at and learn more about this option. Was wondering if you could share your experience. Thanks!
I haven't installed it yet. I did get the design approved by the city which was a big step. I've been working with a friend who is an ME and specializes in commercial hydronics but is doing this as a favor so I don't have any referrals for you.
I have to update my above post to include new knowledge that I have acquired here on GBA since I posted it.
A dehumidifier introduces more heat into the environment than the electricity it consumes. It also exhausts the latent heat of the water it removes. So in my example, the dehumidifier that removes 1.7 pints per hour and uses 580 watts releases 2,000 BTU/hr from the electricity, but also 1700 BTU/hr from the latent heat of the removed water. That increases the sensible load not by 6% as I originally said, but by 10%.
DC, I wonder if you could update us. I was doing some reading and came across this thread. Did you ever end up doing any Radiant cooling?
It's installed but I haven't lived through a full summer yet with it.
You installed it as described here? One hydronic coil and then some radiant ceiling panels? I like it! Let us know in a couple months if it's like a rainforest inside or pleasantly cool.
What did you end up using for your controller?
One hydronic coil per floor.
I had intended to use a 2-stage thermostat with dehumidifier control, but I couldn't find anything that I liked so I made my own using an arduino and DHT22 temperature/humidity sensors located in every zone.
It worked great in the winter, but summer is the real test.
Awesome! I love the capability of Arduino. I have a variant of one doing the CANbus integration on my project car. It's still a little surprising how expensive the cooling controls are otherwise.
I'd pay $0.00 + a thumbs up for a few pictures of the setup.