Filling existing radiant system with antifreeze mixture

Smyth

Hi, hoping to get some advice on my project.

I have an in-slab radiant system in my house with two zones, the basement and attached garage. I am in the process of adding a snowmelt system for the driveway. My original plan was to use a heat exchanger and only use glycol in the driveway, but when I started adding up the equipment (another pump, expansion tank, HE) plus the heat lost through the exchanger I decided to just run glycol through the whole system.

My question is what is the best way to fill the part of the system that already has water? I have purge valves on both the supply and return side of each zone and the manifold. When I filled the system the first time i just opened those until water came out with no sputtering.

Should I do the same thing and just run it until blue water starts to come out? Or should I blow it out with air and then refill? I don't really want to add air to the system, but also want to keep the right mix of glycol.

Thanks

hot_rod

I would first add a cleaner run it up to high temperature for a day if possible, flush it out and refill. It is a bit of a risk pushing glycol in as water comes out. If the mix is to low you end up purging out good glycol to get the mix percentage correct. Ideally drain or blow it out with air and start from empty with the correct premix. Check that all your pumps are up sized if you plan on a 40% mix, isolate any auto fill valves also. Label the system as to what is in it and %

Smyth

Sounds good, I'll blow it out instead of pushing the glycol, thank you.

For the cleaning, the system has only been operating for about 18months would you still recommend a cleaner?

hot_rod

Yeah, the cleaner will cut the flux, pipe dope, manufacturing oils, etc. Those "left over" chemicals can destroy the glycol. Really no harm in squirting in a cleaner for a few days to be sure the system is glycol clean. Rhomar and Fernox have aerosol cans with a hose connection to easily add the cleaner

HomerJSmith

It sounds reasonable, however, glyco sys need to be checked yearly and sometimes replaced on a 5 yr basis. You might have to run a 40% concentration to satisfy a SIM sys. Do you have an indirect water heater connected to the boiler?

I would opt with a heat exchanger and pump and EX tank tridicator gauge, etc. What makes you think that your current pump will do the job of your SIM sys. Heat transfer is less and pumping requirements greater with glycol. The tube size and length in the driveway has a lot to do with the pumping requirements. You could use a booster pump to give you the flow requirements to the SIM sys. If you decide to go with the heat exchanger, I will tell you how I do it.

The SIM sys has to have a manual or automatic control. You have to separate the SIM section from the house section because you are not heating the driveway all the time.

It makes a difference if your driveway is concrete or pavers as to temperature of the SIM sys. In concrete you don't want to shock the concrete and bring it up to temp slowly.

mattmia2

Glycol is also a lot more corrosive than water. It would be better if those components were only a HX, circulatior, expansion tank, relief valve, tridactor, air separator, and maybe a flow check if not part of the circulator instead of subjecting the boiler to the glycol.

Smyth

Thanks for the input, I'll grab a cleaner it seems simple enough. Also open to changing if it makes more sense. Is corrosion to the boiler a concern in boilers rated for it? The manual allows up to 50%. Otherwise it seems nearly a wash to have to replace all of the items listed on the driveway side of the hx.

To give a few more details about the system. The layout has the boiler and it's circulator on one side of a hydraulic separator the zones are on the other side each with its own pump. The headers are sized for the total flow/btu requirements of all three zones but the plan is to run the snowmelt as priority as the other loops are largely supplemental. I believe this should allow me to isolate the poured concrete driveway.

The pipe size/layout and subsequently the pump for the snowmelt was sized based on 50% glycol in loopcad, though I wasn't planning on running that high of a concentration.

HomerJSmith

How cold do you expect the outside temp to be before you get slush in the pex. What is your coldest design day? 50%? Don't do more antifreeze than you absolutely need to.

On priority? It can snow for 24hrs or more. Can you do without space heating for that long or are you going to switch back and forth? What's your control strategy?

JohnNY

My opinion is this:
Maintaining a glycol mixture is a nightmare. Use as little of it as you can. If that means isolating the portion the system that needs it then so be it. Spend the money now and save your sanity and wallet for other things.

HVACNUT

What size is the boiler? Is it enough for snow melt? 

KC_Jones

HVACNUT said:

What size is the boiler? Is it enough for snow melt? 

I was wondering the same thing. If it's enough for snow melt it must be ridiculously over sized for the house, if it's sized properly for the house I can't imagine it has even close to enough guts for snowmelt.

Smyth

Appreciate the interest, trying not to get too far off topic with the boiler. It seems like the snowmelt heat requirement will always be much greater than the home. Unless the house is huge and as drafty as the lawn.

HomerJSmith said:

How cold do you expect the outside temp to be before you get slush in the pex. What is your coldest design day? 50%? Don't do more antifreeze than you absolutely need to.

On priority? It can snow for 24hrs or more. Can you do without space heating for that long or are you going to switch back and forth? What's your control strategy?

I used 50% to size the pump I'll be running less, a local installer told me to plan for ~ -14F. I'm in Michigan just west of Detroit

Controls I need help with but will start another post. Basically the other zones are not required for space heat, though in the case of 24hrs of snow I'm not sure I'd want the thing running that long.

JohnNY said:

My opinion is this:
Maintaining a glycol mixture is a nightmare. Use as little of it as you can. If that means isolating the portion the system that needs it then so be it. Spend the money now and save your sanity and wallet for other things.

I have not maintained a glycol system before, is doing it for a single loop that much different than the whole system?

hot_rod

What size is your driveway, that would be a good starting point. Take the square footage times at least 100 btu/ sq. ft. Do you have that much boiler? If not maybe a separate boiler just for the snowmelt would be a better option.
SIM loads are typically 5 times a heating load, so unless you have a large home, you may not have enough boiler to even consider SIM. Plus when it is in melt mode, you may not be covering the house load.

As others have mentioned it is rare to see a home boiler with enough capacity do do much melting, perhaps a small sidewalk.

Smyth

I didn't word that particularly well, but I think we are saying the same thing. You mention snowmelt is typically 5 times the heat load. That would guarantee the boiler was oversized for the house, which is the same thing KC_Jones was saying and I agree.

JohnNY

JohnNY said:

My opinion is this:
Maintaining a glycol mixture is a nightmare. Use as little of it as you can. If that means isolating the portion the system that needs it then so be it. Spend the money now and save your sanity and wallet for other things.

I have not maintained a glycol system before, is doing it for a single loop that much different than the whole system?


Yes. Maintaining a percentage is tricky and expensive especially if you need to purge the system of air from time to time. Glycol is sticky and has an odor. It gets on your clothes and hands....it just sucks.

GroundUp

While I disagree that glycol is "difficult to maintain" in a closed hydronic system, I wholeheartedly agree that mixing snowmelt into a space heating system is a terrible idea. Please spend the extra $100 and do it right with a HX.

Smyth

I am a little surprised at the pushback on dropping the hx, I spoke with two installers somewhat locally and both steered me away from the HX.

The reasons for going away from the HX wasn't really to save money, though I don't follow how it'd only cost $100. A 250k btu+ HX alone is more than that. One reason was to get heat to driveway efficiently. The HX doesn't transfer at 100% so I am losing some heat to the drive and adding it to my return temp possibly impacting the condensing boiler efficiency.

In addition the garage really should be on glycol anyway so I don't have to keep heating it when I have no plans to be out there. With the garage and drive both on glycol 2/3 of the system is antifreeze. Finally, removing the HX somewhat simplifies the install.

Having said of all of that, if I switch it and run HX for the garage and driveway how is that plumbed?

Just focusing on the drive for now. The model is showing 25gpm @ 14 ft of head. I have a Magna 3 32-60 for the drive loop. That would go on the glycol side of the HX. The pump on the water side would need supply that same flow rate but at a much lower head, basically whatever the 2" header ends up at. Right?

Then to add the garage I would either have to add a second HX or tie into the snowmelt one somehow. A short supply and return header off the glycol side of the HX with the snowmelt and garage zone pumps pulling from that maybe?

So part wise, in addition to what I have now; I'd need a HX with at least 1 1/2 ports if I only run one zone at a time, a 14 gal exp, 1 1/2 air separator, and something like a 26-99 to provide the water side of the HX. Then the ancillary stuff, flanges, gauges/wells, etc.

hot_rod

Use one if the heat exchanger simulation programs available at most of the manufacturers site. You should be able to size for high 90% HX efficiency. It’s just a surface area game. Yes glycol will change the HX sizing, 40% if possible to keep the pumping penalty reasonable. So around a 250,000 SIM load? The house and garage adds how much?

Plenty of homes in ski country are completely glycoled, it is possible. You need an absolutely tight system. Glycol will find any small path to get out, even some of the crimp fittings struggle with glycol. Assume the home has a 40 gallon or more capacity, factor that into the glycol costs vs the HX costs

Smyth

The SIM load is 286k according to the design. The house is less than 60k also according to design, but is conditioned with a forced air furnace. The radiant in the basement is for comfort I didn't break out the heat loss for it or the garage. The heat output is sized to meet the snowmelt only, not run everything at once.

I guess the part I am struggling with is what the HX buys me. That is the better part of 1k in parts, not including the time to put it together. Again price isn't really the point, but there should a decent benefit. Any downside to running glycol is still going to be present in the parts of the system that are running it. I would be saving possible leak paths in the near boiler piping, while adding them back by duplicating hardware on the glycol side of the HX. What am I missing?

hot_rod

In your case I'd glycol the whole thing. In cases where the majority of the components, piping and fluid volume is inside and doesn't need freeze protection a HX usually makes sense.
It really is your choice, the pros and cons have been explained. The best feature of glycol is that it doesn't freeze, it brings some baggage along.

I'd be more concerned how that size boiler will behave with just a basement garage load? How will you address short cycling?

Smyth

Appreciate the time talking through this. I wasn't aware of some of the downsides with glycol mentioned here, I am reconsidering the HX to accommodate possible expansion later. I had no experience with radiant floor before this, the difference in comfort it made in the basement was eye opening.

I have a modulating boiler running the two zones now and haven't had a problem with short cycling. I am adding second modulating boiler to make up the difference to the SIM demand and plan on staging the two. I am still trying to figure the best way to control everything. Open to suggestions there as well, ha.

hot_rod

IF you plan on dumping all your boiler horsepower to the slab, that leaves you without adequate btus for the heat load. So you need some way to assure the loads can be operated concurrently or control it to always have adequate heating btus. Via the HX and pump sizing you could limit what is available to the SIM.

when that SIM load hits, 30 degree water will be returning to the boiler, or HX

For this and the lopsided load many installers prefer to use a boiler dedicated to the SIM.

Snoop around the tekmar site possibly the 667 could be your SIM control and interface with the boiler staging control, or add a 264 staging control to accomplish all this.  Assuming you want an automated SIM control?

Smyth

My plan was to prioritize the slab and dump everything into that.

The house is fairly tight and well insulated. I had the boiler and relay panel shut-off a couple of days ago to verify the natural gas pressure. I forgot to turn the relay back on for just over 24 hours and the temp in the basement was only a couple of degrees below the set point with out side temps in the teens or low 20s.

I did look into getting a dedicated boiler but the price difference was pretty steep to get something close to 300k output vs 150-200k. Especially since it only snows a few days out of the year and would be idle the rest of the time, staging them I can tell myself it's doing something useful. Also, with the SIM running full tilt and the current boiler/furnace/dryer/fireplace I'd most likely have to upsize the gas supply according to the table.

I will check out those staging controls. I read a little about the snowmelt controls, and installed a socket for the tekmar snow/ice sensor before the drive was poured. I have to get a better understanding of how they all work together.