Would like some feedback on my radiant design "plans"

kenjohnson

Hello All,

I'm a long-time lurker, first-time poster.

I'm planning an underfloor radiant heat install in an ~1700 square foot well-insulated remodel. Total heat loss is ~18,000 BTUs at 0 deg F outdoor temp (I'm capable of and very confident of the calculation). Calc is room-by-room. Floors are either 1x6 pine topped with fir strip floors (and maybe 40% coverage with area rugs - average R-value with rugs of ~2.9) or 1x6 pine topped with 1/2" subfloor topped with linoleum (the real stuff, average R-value 1.9). I've used a variety of charts and tools from Healthyheating.org to calculate heat dissipation from fir strip floors as 12 BTU/hr and for the linoleum floors of 16-17.5 BTU/hr (depending on location and exact construction) with 110 degree water and 9" PEX spacing with good quality (thick) plates.

The heating system is a GSHP wtih a large water storage tank. I have an 8 manifold system set up. There is a Grundfos Alpha II pump to push water through the zones.

I've done complete calculations to understand how much heating I'll get at different water temperatures (I'm completely capable of this). Other calculations show an ~0.5 to 1 gpm flow with 0.5" inner diameter PEX (need 0.5" to keep head loss down). My longest runs are 400' total 0.5" PEX (head loss ~15 feet). Downstairs, I'm planning on one zone with a master thermostat (the floor plan is very open) and upstairs (three bedrooms and a bathroom) individual zones with a thermostat in each room. Two of the larger downstairs rooms with larger, older windows will be augmented with cast iron radiators (this is necessary to supply the required heat at cold temperatures). I can get the heat I need with 120 deg F water temperature at 0 deg F outdoor temperature (low water T with outdoor reset at higher outdoor Ts).

Here are my basic questions I'd like to get answered and then I'm happy to listen to any other suggestions/ideas as well:

1. If I reduce the PEX spacing to 6" from 9" (or 8") (effectively putting three 4" plates in one joist bay instead of two 4" plates), I don't seem to get much more radiated heat off the floor (I see maybe a factor of 1.125 more heat). Is that the case?

2. The pump I have already installed is a Grundfos Alpha II. If the head requirement at 1 gpm is for the longest loop only, then that has what it takes, from what I understand. But if all loops are calling for heat, will it just divide it's total ~4.25 gpm pumping capacity (at 15 feet head) by all the loops (assuming equivalent frictional resistance) or will it fall short for some reason? At 0.5 gpm, head loss of 400' of 1/2" PEX is a lot lower...which sounds like it works in my favor as well...

3. I calculate a delta T of ~9 degrees F at 0 degree outdoor T (0.5 gpm per loop) and 5 degrees when at 1 gpm (fewer loops calling for heat). Everything I read says that is reasonable. Can anyone confirm?

4. What would be recommended for balancing flow on the first floor loops, seeing that they are all on one T-stat controlled zone? I am designing for (mostly) balanced temps, but I want to plan for some miscalculations or adjustment after we've lived in the house for a bit.

5. There is 100 year old felt paper between the 1x6 pine subfloor and the fir wood strip floor. Given that my water T will be 120 deg F (130 deg F worst case at -30 deg F outdoor T), everything I've read says that this shouldn't be an issue. Any comments?

Thanks in advance. I've learned a lot already from reading the posts and plan on learning even more now.

Henry

As a rule, we don't connect or install DYI or internet designs. We have architects and engineers asking us to connect radiant systems. NO! We use plans that are designed by the radiant manufacturer. There is no if I do that etc.. And it works every time. You are not saving any money trying to DYI but only risking a system that does not work well!

kenjohnson

I have not been successful getting my basic inquiries from radiant manufacturers answered, despite sending them detailed floor plans and room-by-room heat losses (which is what they request). Perhaps you could suggest some radiant manufacturers I should be contacting?

There is nothing more I would rather do than just buy a design from a competent and capable supplier.

Brewbeer

A couple of thoughts.
Can you install on top of the sub floor instead of under?
Can the loop lengths be reduced? 15 ft of head is a lot. Try to keep loop lengths the same.
Look for other sources of information for dissipation assumptions to confirm what you are getting from your healthyheating sources . If you are good with a spreadsheet and understand the concepts, you can do the design yourself if you are using the right tools.
Or you can buy the design software ...

Zman

It sounds like an interesting setup. I don't see a need for a 3rd row in each bay.
What are the lengths of all the loops? Post the loop lengths and I will run the flow rates for you.
Your delta t numbers look about right.

kenjohnson

Thanks for the comments and thoughts. I'm laying out the floor plate location on a floor plan in Visio and I'll post it with the floor plans, likely over the Thanksgiving holiday.

The install is a renovation to an older home, so I have to install underneath already finished strip flooring.

My assumption of 12-16 BTU per square foot from the finished floor seems pretty consistent with "rule of thumbs" I've seen elsewhere. The healthyheating.org site provided a nice chart so that I could derive it myself for the specific floor.

I understand the need to keep loop lengths nearly the same. If I skip the 3rd plate, that should bring the largest loop length down by 125 feet or so. It would be nice to skip the third plate.

15 feet head was calculated at 1 gpm - would be a lot less at 0.5 gpm (4 feet or so). Would get even less with 2 plates per bay (125 feet less loop length).

Rich_49

Ken ,

I applaud your endeavour but have a couple questions .

1. Is your heat load calc a room by room calculation ?

2. Why would you limit or design your new system around a circ that you have on hand whether it is the right or wrong circ ?

3. Zone it properly , regardless of how many zones it is now , make it right this time around . This is a complete reno correct ?

4. Where are you located , city , state , climate zone ?

5. Dis regard anything negative or questioning of HealthyHeating as some apparently have no idea whose site that is and that most manufacturer info actually comes from there or had input from there .

6. The fact that you found healthyheating tells me that you have a real desire to do it right , are there issues that make doing it right difficult or prevent you from doing so ?

keyote

Consider radiant ceilings using the routed foam with graphite plates. you can do higher swt w/o fear of wood floors and no worry about floors being covered.
400' really long loop Yeah i know what they say its BS keep them under 300 250 better still. this ll also reduce head, ease pump tighten dt. guessing bed loops will be closer to 250 ea so if all on one manifold balance easier too, just split into two zones
your spacing down to 6" is probably reducing your tubing to 3/8 the min rad of 1/2'' is 8" the foam board btw is 8" loops and 1'' thick.

BTW Im not a pro just someone who did this and in the biz airside

leonz

Installing hot water radiators sized for the heating load would be faster to install if you use PEX and the installation would be cleaner. They make beautiful radiator covers for them.

kenjohnson

1. Is your heat load calc a room by room calculation ? Yes, I've done a very complete and thorough calc for each room.

3. Why would you limit or design your new system around a circ that you have on hand whether it is the right or wrong circ ? The circulator pump installed (Grundfos Alpha II) was originally installed as I was planning on cast iron radiators with TRVs, so would have been a great option for that. I can change this out, if needed, no problem, if there is something more suitable for constant circulation (design goal).

3. Zone it properly , regardless of how many zones it is now , make it right this time around . This is a complete reno correct ? Yes, complete renovation, access to all walls and ceilings. Plan is one zone on the first floor (very open plan) and one zone per room upstairs (three bedrooms and one bathroom). Will have to have a good Q&A on the right thermostats for downstairs and upstairs zones.

4. Where are you located , city , state , climate zone ? Location is in Central NY, about 50 miles southeast of Syracuse. Climate zone 4.

5. Dis regard anything negative or questioning of HealthyHeating as some apparently have no idea whose site that is and that most manufacturer info actually comes from there or had input from there . It's hard to argue with an ASHRAE document on the HealthyHeating site...

6. The fact that you found healthyheating tells me that you have a real desire to do it right , are there issues that make doing it right difficult or prevent you from doing so ? Going slow and doing it right. No money or schedule issues. I believe in doing it once and doing it right.

kenjohnson

Consider radiant ceilings using the routed foam with graphite plates. you can do higher swt w/o fear of wood floors and no worry about floors being covered. I won't rule out radiant ceilings, but my heat loss will only be 18 kBTUs at 0 deg F (~11 BTUs per square foot). My supply temperatures will be pretty low (115 degrees F for this load, as calculated today) even accounting for the R-value of wood floors with 40% throw rug coverage. My heating system is GSHP with outdoor reset and I'll be setting the ODR curve to optimize temps to be as low as possible with constant circulation.

kenjohnson

Thanks for all the feedback. I've used it to optimize the floor loop size/length and created detailed drawings of the loop layouts. I've also indicated my room-by-room heat loss per hour, floor R-values (averaged) and PEX loop lengths and PEX and radiant plate locations. Would appreciate any further feedback and conversation. Thanks in advance.

kenjohnson

I'll assume (for a moment) that my plan looks acceptable. Does anyone have any feedback on the right ODR controller and pump to get a new constant circulation?

kenjohnson

I'm just about completed with installing my loops and am looking for feedback on manifolds and controls.

Just a summary from before - I have 9 loops and all are 1/2" Uponor He-PEX driven by a Grundfos Alpha2. Each loop is approximately 250 feet in length, including supply and return. Four of the five downstairs loops are on one zone, and all the upstairs zones (four total) are zoned separately.

For a manifold, I'm thinking of this https://s3.amazonaws.com/s3.supplyhouse.com/product_files/Uponor-A2701002-Product-Overview.pdf. Would I need one manifold "loop" for each zone or one for each loop?

Then, I would need a thermal actuator and a thermostat, so I'm thinking of https://www.supplyhouse.com/Uponor-Wirsbo-A3023522-Thermal-Actuator-Four-Wire and (for each zone) https://www.supplyhouse.com/Uponor-Wirsbo-A3100101-Heat-only-Thermostat-with-Touchscreen.

Any thoughts/comments?

Zman

For the upstairs, that manifold with 4 actuators makes sense.IT will make for a nice clean installation.

Downstairs might work better with zone valves. You could use a 3/4" Taco zone sentry to operate a 4 hole manifold a 1/2" to take care of the single loop.

If you kept your loops +/- 10%, you should not need the balancing. You would save a bit of $ going with a copper manifold, especially downstairs.

I have not tried that thermostat. I really like the model it replaces with similar features. I wish the new one let you lock the temp range. That was a nice feature.

hot_rod

Another great manifold with unique user friendly features

kenjohnson

Hi Zman, it sounds like you are suggesting a two manifold design - 1) a four loop manifold with actuators and thermostats for each loop for the upstairs and 2) a five loop manifold for the downstairs with a zone valve for the four loops that are one on thermostat and an actuator for the one loop that is on the second thermostat. Is that the case? Or would I put all the loops with thermostats and actuators on one 5 loop manifold?

I believe I am plumbed for a 1-1/4" copper pipe from the output of the storage tank, so I would just get the 1-1/4" manifold (or the 1" if that is what I have plumbed).

Hi HotRod - thanks for your other suggestions. I'll be sure to check them out.

Zman

The 4 upstairs with actuators would work fine.
Downstairs, if you use a 5 port, you would need 4 actuators tied to the same t-stat. That is less than ideal.
I would suggest 4 upstairs with actuators on the manifold then a zone valve feeding a 4 port manifold (maybe simple copper) and a zone valve feeding the single loop with no manifold.
I would size the piping for 4 ft/sec from this chart. Your manifolds would be better sized for 2-3 ft/sec
http://s3.supplyhouse.com/product_category_files/11448-Flow-Rate-Chart.pdf
You can figure your loops at about .75 GPM each

kenjohnson

Hi Zman, thanks for the advice.

Right now, I've got a custom-built 6 port copper pipe manifold, which I think has 3/4" outlets on each port (I'll be at the house this weekend to measure it). So it sounds like I could have a 3/4" port feed a 4 port manifold (upstairs) and a separate 3/4" port feed a 4 port manifold (downstairs) and a separate 3/4" port feed the single remaining zone.

The 3/4" pipe at a little more than 2 feet/second would deliver 4 gallons/minute. This flow split into 4 different 1/2" PEX pipes would result in 0.75 gallons/minute each at 4 feet/second. Correct?

Would I need to do anything special with the single zoned PEX pipe at the outlet of the 3/4" copper supply?

Zman

That sounds like a good plan.

kenjohnson

Hi Zman, thanks for the advice and comments.

kenjohnson

Thanks to all for your advice and suggestions on reviewing my plans and helping me get this system going. I am now able to report success!

First manifold is connected to the supply and I (finally) have hot water moving through the loops. Right now, I only have the four manifold downstairs loop connected, and hot water is pumping through three of the four loops in the manifold. One loop has no flow, so I am guessing that there is an air bubble in the loop. Otherwise, each loop has about 0.5 gpm through it and a delta T of about 6-7 degrees F - pretty much as predicted. The floors are noticeably warmer and everything seems to be working as predicted.

One question - the Grundfos Alpha II pump is pulling about 30 watts @ 2 gpm - that seems a little high to me. Is that because Auto Adapt mode hasn't fully adapted yet, or is it because the one loop (with the probable bubble) has too much back pressure, forcing the pump to work harder?

kenjohnson

That Feb 17th update (above) is actually from about 3 weeks ago - wrote it but never formally posted it.

Now, the system is fully connected. Initially, with all loops connected and pump running 100% of the time (with outdoor reset control of water temperature and no thermostats), house was at 70 degrees F throughout with 123 degree tank water temperature and -5 degree outdoor temperature overnight (a couple of weeks ago) - everything seems to be working exactly as planned (maybe a little better...). Thermostats have since been installed - all working good so far.

Pump is drawing 48 watts pushing water through all loops at the same time. While this is higher than I expected, I guess that is what it uses with smaller PEX pipe (as opposed to 1/2" copper or something like that), and the power usage still seems half as much power or less compared to a conventional pump. It is, however, at the very upper end of what it is rated for (45 Watts, from my recollection). Should I be concerned?

kenjohnson

Finally had a chance to inspect the system performance myself, fully installed with thermostats, instead of getting performance via phone from the installer. All is still as expected.

Overnight outdoor temperature was 6 degrees F. Controller was driving the heat pump to produce 120 degree F water (based on outdoor reset temperature). With all loops open, the Grundfos Alpha was drawing 18 watts and pushing 4 GPM (from its readout) through a 1" pipe. Each 1/2" PEX loop is getting 0.6 to 0.8 GPM (nicely balanced, per the built-in flow meter on the Uponor manifolds) and the Delta T in each manifold is about 10-15 degrees F. I turned the thermostats up by 5 degrees overnight and 11 hours later the room temperature was 4 degrees higher. We drilled big holes in the joists for the PEX and used the PEX isolators - I can't hear a thing when the system is warming up - no pops or sounds at all.

My system design temperature was for 120 degree water at 0 degree outdoor temperature, so it is performing at design (or slightly better) - before installation of storm windows, last 1/3 of insulation in the attic, and no insulation in the radiant joist bays (yet). Basically, I'm heating my basement right now too...it's as warm as the rest of the house.

To say I am thrilled with the performance thus far is an understatement.

The fact that the Grundfos pump is only using 18 W under full load is amazing to me. Most people (not on this site) told me that pump wouldn't work, that it was too small, but all the calcs showed it would work. And after it "adapted" it settled in nicely. 18 watts to just do constant circulation through all loops (and as low as 8 watts with one loop) all day long is 13 kWh/month - practically nothing. Glad I went with the 1/2" PEX to get this efficiency, despite it being a little tougher to handle during the install.

Shout outs to @Zman for his critically important advice along the way, to @Brewbeer and @Rich_49 for their encouragement, and to @hot_rod who (I believe) has written all of those great Caleffi technical papers that I learned so much from - I didn't use Caleffi parts but those are some great materials to learn from, and @hot_rod has been a helpful contributor in other forums I use as well.

A shout out to my installer, Dan Cochrane at Renaissance HVAC in Verona, NY for gently guiding me down the path to a geothermal heat pump, doing that install, and connecting that install up to my radiant loops. The last part of the project was not something that is part of his core interest, and he had to listen to all my great ideas along way, but he was a willing and very capable partner and did a great job. Not every installer has the patience or interest to implement someone else's design...

A shout out to Bennett Sandler at Equity Energy LLC in Fly Creek, NY. He asked a simple question - "if your heat loss is so low, why aren't you using radiant heating?" It got me thinking...And if I knew (at the time) that Bennett's firm did some (or all) of the heat loss engineering work I did, I would have involved him more. Time is money, after all...See you at Saturday hockey, Bennett...

A shout out to my long-time plumber Don Searles who has done a few projects like this over the years, looked at my plans and said "yep, that will work". When most people say "won't work" but someone with 40+ years of experience says "will work", that adds a lot of confidence to keep moving forward.

Lastly, a shout out to two guys - Philip Kane and Wayne Congden - that materialized at the perfect time to help me with this project. Philip did a lot of self-study, outside of the time I paid him, to learn about radiant installs and had some critical experience in pressure testing PEX to make sure that all was well with the loops, despite a few minor kinks. Wayne was battle-tested and intuitively knew the right way to get this project done. It was great to use their combined knowledge on this project and also great fun to work with these guys as a team to get this project done. Running a half a mile of PEX through a few hundred holes is no fun, but with the right crew you can get through it fine with a lot of laughs along the way.

What would I do differently the next time? Not too much...but:

1. 1) I'd avoid the bother of a 3rd plate in a few of the joist bays - that made running the loops a lot more complicated for not much extra heat.
   
2. 2) I think I would also have bought the PEX coil unroller tool for $500. I probably spent that much money on time with my crew pulling 300' of PEX out the door of my house into my big yard and back into the house again to keep it from getting tangled and kinked.
   
3. 3) I would definitely pre-drill the holes through the joists before installing the plates. We probably had to uninstall and re-install 1/3 of the plates once we realized that we couldn't get the minimum bend radius we needed for the PEX where the hole could be drilled.
   
4. 4) I would have bought the big DeWalt right-angle cordless drill a lot sooner rather than use the electrician's drill. That DeWalt is an animal that just chewed through those tough old larch joists.
   
5. 5) Instead of saving money and buying 1000' rolls of PEX, I would just buy the 300' rolls to avoid the bother of cutting 250' off a 1000' roll. Also, we ended up 10 or 20' short on two loops, despite my best calculations as to loop length.
   
6. 6) I might not also have bothered with the big Stiebel Eltron 180 gallon storage tank that everyone tried to talk me out of...but my long term vision is that I will be able to take advantage of that much storage to:
   • a) heat my house for 5 hours in the event of a power outage without using the woodstove
   • b) take advantage of (future) demand-based electric rates offered at a future time and/or let the utility shut off my heat pump (for a fee) when demand gets too high.
     
   • c) run my cold water inlet for my DHW through the top heat exchanger of this tank to pre-heat the cold water inlet in the winter before heating it with the normal DHW heater.
     

kenjohnson

Thermostats are fully connected and I had a good opportunity to really observe the system on Thursday night. I am very pleased with the operation.

The Grundfos Alpha pump now seems completely "adapted" - it is drawing 18 watts at 4 GPM with all loads open. Each zone is between 0.6 and 0.8 GPM (as read on the Uponor manifold) and Delta T is 10-15 degrees. Zone valves are all working nicely. At 6 degrees overnight, system was generating 120 degree water. I increased the thermostats from 60 to 65 degrees and 8 hours later the house was at 64 degrees, so the output is in line with what I modeled, maybe a bit higher. The system is quiet (no pops) during warmup - glad I drilled the big holes and used the plastic isolators. I still have to insulate under the radiant plates, so the system is keeping the basement at a toasty temperature right now - not a bad thing in an unoccupied house.

I can't believe the Grundfos Alpha uses as little power as it does...about 13 kWh/month maximum - pretty amazing! I'm glad I used the 1/2" PEX to reduce the head loss and get this efficiency.

What would I do differently if I did it again?

1. I would install the plates after I drilled the holes. We had to remove and re-install half the plates to get the bend radius on the PEX we needed - not fun.
   
2. I would never use three plates in a joist bay. It added complexity to the runs and wasn't worth it for the extra 10% or so of heat delivered. Luckily, I didn't have many of these.
   
3. I would buy the $500 uncoiler machine to manage the loops during installation instead of running them out the door into the big yard and back inside again.
   
4. I would have bought the DeWalt professional battery-powered right-angle drill earlier instead of borrowing my electricians drill. That DeWalt drill is an animal that just plowed holes right through tough larch framing.
   

A shout out to @Brewbeer and @Rich_49 for their encouragement and @Zman for his great feedback at the critical junctures. Also to @hot_rod for all those great Caleffi publications (pretty sure you wrote them) - I learned a lot from all of them.

A shout out to my installer Dan Cochrane at Renaissance HVAC in Verona, NY. Dan gently guided me towards a ground-sourced heat pump, and then connected it to my radiant loops. He was patient and understanding all the time, and the system works great due to his help. Thanks Dan.

A shout out to my long-time plumber Don Searles, who looked at my plans and said "that will work" when everyone else was saying it wouldn't work. That gave me the confidence to move forward.

A shout out to Bennett Sandler at Equity Energy. Bennett asked a simple question "if you are going to have that little heat loss, why wouldn't you install radiant heat?" He also consulted on other aspects of my project. Thanks Bennett.

Shout outs to Philip Kane and Wayne Congden who magically appeared at just the right time to help me get this done. Philip studied up on radiant heat in his off hours and Wayne had great wisdom and practical experience to help make this happen. They both contributed in far more ways than simple "helping hands". Running a half a mile of PEX through a few hundred holes is not anybody's idea of fun, but if you can joke around and have a few laughs along the way, it is a lot easier. Thanks guys.

Lastly, a shout out to my wife Laura for all of her patience during all the weekends I worked on learning about and planning for this.

hcpatel78

Hi Kenjohnson... do you mind to share of your hydronic setup pictures here or my personal email;? I live in Northern NJ and planning to do DIY radiant under joist installation. I am a mech.engr and can perform this kind of job very well. I have done heat load calculations and temporary design.
email:- hcpatel78@gmail.com

kenjohnson

I can share whatever you'd like whichever way you'd like. Sharing on this forum would benefit others in the future. Let me know what kind of pictures you are looking for. I can get them up in about a week's time or so (I am not near that house right now).

hcpatel78

Hi Kenjohnson, Thank you... I am planning to use the Noritz brand combi boiler (NG operated). I have some basic design done through the watts radiant software which is giving me 6 loop manifold with 1/2"m hep-ex piping. I bought and start to install joist plates now. I am interested to see how you set up whole piping including boiler, circulator, air eliminator, exp. tank, manifold and return and all gages with valves.