TLDR (too long, didn't read): It has been strongly suggested to me to run a continuous flowing system and utilize a outdoor reset control to control the heating system. In this design I was told not to use thermostats inside the house and to utilize the balancing valves to insure the home is heated evenly. I have read plenty and come to understand how a boiler reset system works (in theory: the colder it is outside, the hotter the water the boiler makes, flows through your house and keeps you warm). I'm confused how and when the boiler fires if the thermostats are not calling for heat? Does the reset control request the hot water? Do I need to use some type of mixing valve with an electronic control to separate the boiler loop from the radiant loop?


The longer version:

I've undertaken a DIY radiant heat install in my new home. I have previous experience doing this at my parents house when I was much younger, however, I was not involved in the control planning or implementation stages- I got the grunt work of running pipes for a 4k square foot home...

Anyway the facts:

I am converting from oil to gas and having a Navient natural gas burner installed. This burner will also handle my DHW.

My house is ~1,800 sq/ft ranch with a main above grade floor of ~1,00 sq/ft and a half below grade, half above grade (house is on a hill) finished lower level of 800 sq/ft. I designed the radiant heat system using a trial version of loop-cad then did my own measurements as I went.

I gutted the home and on my main floor I've installed 6 radiant heat circuits (or loops?) between 250 to almost 290 feet each. I did a below sub-floor install utilizing pex clips and 1" spacers. 8" on center spacing between pipes when possible. The flooring is 3/4" hardwood (no carpets) and 3/4" sub-floor. The pex is Uponor 1/2" with oxygen barrier. After pressure testing I plan to staple in reflective bubble wrap about 1" below the tube and then spray insulate the remainder of the joists.

Unfortunately, my (now unfinished) lower level is a slab and I do not have the ceiling height to raise the floor or the funds to break up the slab and install foam / pour a new slab. It was explained to me that this is an expensive process. I'm having the entire lower level and half of the second floor spray insulated (walls and vaulted ceiling) with 3" of closed cell foam. My plan is to install a pair of radiators and feed them with Pex from the same manifold I use for the radiant or possibly a different one (still planning). I was told with the tightly sealed insulation and a few radiators I should be OK in this space.

I'm in the phase of trying to decide how to control the radiant heat system. I'm an IT guy so naturally I gravitate to advanced controls and utilizing actuators on every circuit / thermostats in every room. I received the advice of the continuous flow system from a contractor whom did this in his own home with great results. After reading about it online it appears that this design is popular in Europe and energy efficient, but I do not understand how it works and when the boiler fires -- or what controls I should look into buying?

Any help is appreciated!

Do both - outdoor reset for determination of the water temperature and interior thermostat(s) for control of valves/circulator on/off and interior air temperature. Adjusted properly, it will operate close to continuously. Or with bypasses on the valves, continuously.

Outdoor reset and boiler on/off are probably built-in to the boiler.

Yes, the most efficient way to operate a HR system is continuous circulation. This has been done in Europe for many years. In short, it involves using a PID controller to modulate water temperature using three or four way mixing valves and using a control curve that is based on actual building heat loss accounting for outdoor temperature and solar heat gain. We use PLCs to accomplish this in our passive solar and integrated HR building design/builds. When your floor slab is being heated by the sun and this area of the building has a surplus heat gain, you don't want to stop circulation like a conventional HR system would do, you want to maintain the circulation but at a lower temperature so as to convey heat gain to other building areas that need heat gain. I wrote a treatise on this subject several years ago that should be in the GBTF archives.

Thank you for the advice. I'm searching for your post on the subject, but coming up blank. Do you recall the title of the post? If you had a link that would work too. Thanks!

I switched from a PLC with temp sensors in every room and outdoor temp to an Ecobee3 and couldn't be happier. I am selling the house and wanted to take the PLC with me.....now I am junking it. The Ecobee3 is soooo much better. It doesn't over shoot or under shoot.

Sorry, I initially had a hard time finding these discussions too! Anyhow, here they are:

PID Controller & Sensors

Western WA Sunroom & Passive Solar Performance

Any controller using PID control logic will never over or undershoot if you set the feedback gains properly. The key is knowing how to accomplish this for the given controller and the given system that will be controlled. I spent several years in graduate school learning all the nuances of both classical and optimal control theory...so this is not a trivial pursuit.

"I plan to staple in reflective bubble wrap about 1" below the tube and then spray insulate the remainder of the joists"

Don't! You only need about R10 of fluff to achieve adequate zone separation between the basement & first floor. Aluminized bubblepack is pretty useless and expensive stuff compared to "contractor rolls" of R13 or R19 batts.

"I'm having the entire lower level and half of the second floor spray insulated (walls and vaulted ceiling) with 3" of closed cell foam."

And don't waste expensive high R/inch foam here either! For foundation walls a continuous 1-1.5" shot of closed cell foam on the concrete with a 2x4/R13 studwall on which to hang the wallboard works at about the same thermal performance of 3" of 2lb foam, but rigid polyiso is often cheaper & easier. The amount of foam you need for wintertime dew point control at the foam/fiber boundary depends on local climate. Where are you?

You can do a full fill of 2x6 framing with half pound foam for less than half the amount of polymer 1.5" of closed cell foam (2.75 lbs per square foot vs. 3lbs per square foot), and it will outperform 3" of closed cell foam. Installing 3" of 2lb foam between studs to hit a code-min R20 dramatically underperforms R20 fluff or half-pound foam between 2x6s, due to the much shorter thermal bridge through 3" of framing.

The HFC245fa blowing agent used for closed cell foam is a powerful greenhouse gas. The water used for blowing half-pound foam is not.

In cooler climates with open cell foam you may need to add an interior side vapor retarder such as half-perm "vapor barrier" latex primer, or (better) 2-mil nylon to protect the structural sheathing from wintertime moisture drives. If you install 3" of 2lb foam you would be at a bit less than a half-perm too, but the cost delta between that and 2-mil nylon (Certainteed MemBrain) is huge. The material cost of the nylon is less than 15 cents per square foot. Better still, nylon is a "smart" vapor retarder, and will become vapor open if the sheathing ever needs to dry out, but it otherwise less than 1-perm under typical wintertime conditions.

Even with the cost of the vapor retarder, 5.5" of open cell foam will be about a buck a square foot cheaper than the 3" of closed cell foam that it outperforms (when installed between framing.)

Paragraph breaks make it easier to read- if they're not showing up, try a different web browser (FireFox & Chrome both work well enough on this site.)

Posted By sailawayrb on 31 Oct 2016 11:10 AM
Sorry, I initially had a hard time finding these discussions too! Anyhow, here they are:

PID Controller & Sensors

Western WA Sunroom & Passive Solar Performance

Any controller using PID control logic will never over or undershoot if you set the feedback gains properly. The key is knowing how to accomplish this for the given controller and the given system that will be controlled. I spent several years in graduate school learning all the nuances of both classical and optimal control theory...so this is not a trivial pursuit.

With all do respect, I am sure you have put a ton of time into PLC's and PID loops and are trying to sell them as the fix all. But I would rather have a simple solution for the masses and let people know of the option. For under $200 you can have a remotely controlled thermostat that learns your house, factors in the local weather and is very configurable. I have my one thermostat controlling my geo forced air unit for air conditioning and my heated floors by turning on and off the pump through a 24VAC relay. Down and dirty simple. As far as PID never over/under shooting....Is that plus and minus 2 degrees? :) PID is very hard if not impossible to do with so many variables in a house and not over/under shooting. Works great for constant systems like a motor controller lifting a load. In a house you have multiple varying loads...Sun, Wind, Outdoor Temp, Humidity, Cooking, Cloths Washing, Fans, Season/ground temp.....

Posted By newbostonconst on 31 Oct 2016 11:09 AM
I switched from a PLC with temp sensors in every room and outdoor temp to an Ecobee3 and couldn't be happier. I am selling the house and wanted to take the PLC with me.....now I am junking it. The Ecobee3 is soooo much better. It doesn't over shoot or under shoot.

Exactly. I have been designing radiant systems for decades and came to the same conclusion. Though I typical only use one per home, with simply ambient thermostats for the rest, an Ecobee 3 is the answer. All the fancy talk is a waste of hot air, that is better used for space heating.

This post is about HR continuous circulation systems which require PID control logic and high level of multi-sensor feedback. Does the Ecobee use PID sensor control logic? Does the Ecobee utilize solar irradiance and appliance/occupancy sensors? Have you actually used the Ecobee for HR continuous circulation system control? Somehow...I don’t think so...

If you don’t know what a HR continuous circulation system is, you might find this basic information useful:

Contractor - Continuous circulation means greater efficiency

Heating Help - Considerations when using constant circulation

If a simple controller works adequately for your building and heating system, than that is what you should use. Simplicity is always best...if it works adequately.

I used to try to make my systems more complicated back when I tried to impress-for-success. Then I grew up.

Now I often use an ECM circulator to drive radiant panels, what the Contractor article actually alludes to, but this is not necessarily full-time and is rarely relevant for radiant floor application.

If your house is very large, sprawling, multi-story and features disparate room loads, you may be in the market for a more sophisticated system.

https://www.uponor.co.uk/~/media/countryspecific/uk/support/manuals/c56-manual.pdf?version=1

Most of us can be perfectly comfortable with less control and fewer headaches.

Posted By Dana1 on 31 Oct 2016 02:38 PM
"I plan to staple in reflective bubble wrap about 1" below the tube and then spray insulate the remainder of the joists"

Don't! You only need about R10 of fluff to achieve adequate zone separation between the basement & first floor. Aluminized bubblepack is pretty useless and expensive stuff compared to "contractor rolls" of R13 or R19 batts.

I was told I needed a reflective backing to make the radiant heat effective. Is this not true?

Posted By Dana1 on 31 Oct 2016 02:38 PM
"I'm having the entire lower level and half of the second floor spray insulated (walls and vaulted ceiling) with 3" of closed cell foam."

And don't waste expensive high R/inch foam here either! For foundation walls a continuous 1-1.5" shot of closed cell foam on the concrete with a 2x4/R13 studwall on which to hang the wallboard works at about the same thermal performance of 3" of 2lb foam, but rigid polyiso is often cheaper & easier. The amount of foam you need for wintertime dew point control at the foam/fiber boundary depends on local climate. Where are you?

I'm located in NY- specifically a Long Island town called Centerport. I live at the end of a peninsula so it gets REALLY windy (60+ mph frequently). My home is a 1954 ranch with poor (in some places no insulation). I started to look at closed cell spray foam because I vaulted the ceilings in my main LR/DR/EIK. I only have 2x6 rafters so in order to make a non-vented attic I choose to use spray foam.

My official spray foam measurement plan is as follows (all closed cell):
Great Room:
-Walls: 2"
-Roof rafters: 3"
Garage (below great room) & Basement:
-Above grade exterior walls: 1"
-Rim joists: Fill
-Below grade exterior walls with cinder block foundation I will build 2x4 stud walls and use conventional batting insulation

The spray work to seal in the radiant heat I was considering doing as a noise barrier as well. I know it will be more expensive, but I do a lot of work in the garage so it might make my small home feel like we have space from one another.

Please let me know if there is a better way to do this as I've schedule the spray foam company for a week from today.

Posted By Dana1 on 31 Oct 2016 02:38 PM


Paragraph breaks make it easier to read- if they're not showing up, try a different web browser (FireFox & Chrome both work well enough on this site.)

Sorry! I didn't realize this site required manual entry of HTML breaks. I thought it was a form error at first.

Job site pictures:

At this time my control strategy is to utilize a Taco i or IR series 4 way mixing valve and a Taco zone control valve.

The first floor is radiant heat, while the basement is radiators.

Utilize the thermostats (currently have two Nest thermostats) to trigger the mixing valve or zone valve dependant on which calls for heat.

Please let me know if there are better products or a better solution. I'm trying to follow the KISS method here and not purchase individual loop actuators and expensive controls.

One of my main concerns is that the temperature of water the radiant heat requires (design temp 115-120) will not be enough to heat the basement that utilizes radiators.

First, a room-by-room Manual 'J' heat load on dedicated software that will calculate the heat load and determine which, if any, radiant panel system will satisfy the load.

This is a suspended tube system, all-be-it a little different, and will heat the floor. Whether it will heat the room is anyone's guess. I have seen hundreds but would use my software anyway. Why guess?

Reflective insulation is over-rated, particularly for sub-floor application on retrofit systems. Any radiant reflector will be well dusted after the first season and any reflectivity greatly reduced. Better to make sure that each joist space it tightly sealed so convection does not rob the potential output of the floor or turn into a radiant ceiling for the space below. Something a the proper software can predict with great accuracy.

Basement walls should be insulated with XPS or Iso and then 2x4 with batts.

http://www.badgerradiantdesigns.com/images/SampleHeatLoadAnalysis.pdf

Doooglasss, your control strategy should work okay. One normally does a room-by-room heat loss analysis BEFORE installing the PEX to optimize the PEX spacing, heat zoning and size the heat source. If you don’t have a copy already, you will find John Siegenthaler’s “Modern Hydronic Heating” to be invaluable. We have several free DIY HR analysis/design software tools on our website that you may also find useful:

Borst Heat Loss Analysis Software

Borst Hydronic Radiant Floor Heating Design Software

Borst Buffer Tank Design Software

Borst Expansion Tank Design Software

If you want to invest in professional HR design software, Avenir’s HeatCAD and LoopCAD can accomplish both the dated/simplified ACCA Manual J8 and the more accurate ASHRAE analysis methodologies. Either would be adequate for your project and we have samples on our website. Dana has provided you with great guidance with regard to the insulation.

Posted By doooglasss on 03 Nov 2016 11:04 AM

Posted By Dana1 on 31 Oct 2016 02:38 PM
"I plan to staple in reflective bubble wrap about 1" below the tube and then spray insulate the remainder of the joists"

Don't! You only need about R10 of fluff to achieve adequate zone separation between the basement & first floor. Aluminized bubblepack is pretty useless and expensive stuff compared to "contractor rolls" of R13 or R19 batts.

I was told I needed a reflective backing to make the radiant heat effective. Is this not true?

Absolute rubbish. Radiant heat does not require a reflective backing. The heat transfer via contact with batts is primarily conducted, even low density batts aren't translucent enough in the deep IR to matter at those low temperature differences.

"I'm having the entire lower level and half of the second floor spray insulated (walls and vaulted ceiling) with 3" of closed cell foam."

I'm located in NY- specifically a Long Island town called Centerport. I live at the end of a peninsula so it gets REALLY windy (60+ mph frequently). My home is a 1954 ranch with poor (in some places no insulation). I started to look at closed cell spray foam because I vaulted the ceilings in my main LR/DR/EIK. I only have 2x6 rafters so in order to make a non-vented attic I choose to use spray foam.

My official spray foam measurement plan is as follows (all closed cell):
Great Room:
-Walls: 2"
-Roof rafters: 3"
Garage (below great room) & Basement:
-Above grade exterior walls: 1"
-Rim joists: Fill
-Below grade exterior walls with cinder block foundation I will build 2x4 stud walls and use conventional batting insulation

The spray work to seal in the radiant heat I was considering doing as a noise barrier as well. I know it will be more expensive, but I do a lot of work in the garage so it might make my small home feel like we have space from one another.

Please let me know if there is a better way to do this as I've schedule the spray foam company for a week from today.

L.I. is in US climate zone 4A. In zone 4A with unvented ceilings you only need 30% of the total R to be low permeance closed cell foam, with the rest being fiber. In walls it takes less than 15%. The thermal bridging of the rafters & studs robs the high R/inch foam of its true value. 0n the ceilings you can put 2" of foam (R12) on the underside of the roof deck and have 3.5" of depth for R15 rock wool or fiberglass and have HUGE dew point margin. (R12 foam /R27 total = 44%). If you can give up 2" of ceiling height you can install cut strips of 2" polyiso on the stud edges and install R23 rock wool batts and STILL have adequate dew point margin (R12/R35= 34% ), and it would have more than double the performance of 3" of closed cell foam thermally bridged by rafters.

Batts work as well or better than closed cell foam as a sound barrier. Just make sure there are no edge or end gaps, and make the gypsum air tight.

On the walls 2" of closed cell foam buys you nothing that 3.5" of open cell wouldn't, at less than half the price. The only cost adder would be a smart interior side vapor retarder such as 2-mil nylon (eg Certainteed MemBrain), which may not be needed, depending on the type of siding. MemBrain runs about 12 cents per square foot, compared to a buck a square foot for just 1" of closed cell foam. If it's 2x6 studs, a full fill of open cell foam would even meet current code min, outperforming thermally bridged 2" of closed cell foam by twice, and would still be cheaper than 2" of closed cell foam.

In the garage, 3" of open cell costs the same as 1" of closed cell, and would outperform it, both on air-tightness and whole-wall R.

On the foundation wall with an interior batt-insulated studewall you need at least R2.5 between the studwall & CMU for wintertime dew point control where it extends above grade, but you also need some for summertime dew point control below grade. As little as a half inch of foil-faced polyiso (R3) or 3/4 inch of foil or vinyl faced EPS with the seams taped works, at which point you don't need interior side vapor retarders tighter than standard latex paint- unfaced R15s would be fine, or kraft faced R13s, just no foil faced batts, and no interior side polyethylene, or wall finishes more vapor tight than standard latex paints. Put an inch of EPS (but not polyiso) under the bottom plate of the 2x4s, extending to the CMU wall. That gives the bottom plate summertime dew point control as well as a capillary break against ground moisture. Don't put cut edges of polyiso in contact with the slab, or the CMU wall, since it could take on moisture over time. Don't sweat a few nicks in the facer where it contacts the CMU, just don't rest the bottom edge on the slab.

BTW: If you go with just 2" of roof foam + R15 batts, without edge strips for higher-R/lower ceiling, when it's time to re-roof you an get to code-minimum performance on a U-factor basis by putting 4" of continuous polyiso above the roof deck. If you did the 2" polyiso edge strips and R21 fiberglass or R23 rock wool you could get there with 2" of exterior polyiso.

Going with suspended-tube instead of aluminum heat spreaders (even the cheap sheet metal type) is probably a mistake, since your water temps will likely be above the condensing zone most of the heating season, limiting you to no more than high 80s efficiency. Even stapling it in direct contact with the subfloor without the heat spreaders would be an improvement. If you're going to keep as-is you don't want the batts to be indirect contact with the suspended tubing, since it would impede the convection that's supplying some of the heat transfer. But if you rework it as a full-contact staple up (with or without heat spreaders) snugging the batts up to the tubing & subfloor is the right thing to do.

Dare I ask which model Navien you're installing? Hopefully not one of the combi boilers, most of which don't have sufficient modulating range at your likely heat load, unless you go down in size to the point where domestic hot water performance suffers badly. A tightened up 1800' rancher with an insulated foundation is going to have a heat load of about 22-27,000 BTU/hr @ +15F (the 99% outside design temp for most of Long Island), and ideally you'd have something that modulates down to something less than 10,000 BTU/hr. The smallest Navien combi (the NCB 180) only modulates down to 14K-in/13K-out, but would have pretty wimpy domestic hot water performance in winter. The NHB-55 or NHB-80 both modulate down to about 8K in/7.5K out. The NHB-80 would have fairly fast recovery times for an indirect fired hot water tank, and the NHB-55 would still beat the typical 50 gallon standalone.

One of my main concerns is that the temperature of water the radiant heat requires (design temp 115-120) will not be enough to heat the basement that utilizes radiators.

Balance is a valid concern in that your outdoor reset water temperature has to be set to the highest of all the requirements - this reduces efficiency. Maybe you want a little more downward heat leakage from the radiant floor to the basement. Less load on the radiators could allow a lower supply and reset temperature.

I suspect that your plumbing can be simplified (note the use of closely spaced Ts). Perhaps with as little as two pumps (plus the pump inside the boiler), a manual 4-way mixing valve (proportional reset) and no zone valve. Less if you can balance for a single temperature.

The heat loads of code-min insulated air-tight basements in climate zone 4A are pretty tiny, even if you don't insulate the slab. A 2x4/R13 wall with enough foam for dew point control would beat code-min performance. Assuming its 1800' of insulated basement mostly below grade (and not a walk-out) we're probably looking at less than 10,000 BTU/hr after it's insulated, maybe even as low as 5K. I'd worry more about suspended tube being able to deliver design-day heat with 120F water than the basement. (Of course you COULD run the heat load calculations rather than relying on a WAG from members of the peanut gallery here. :-) )

With 120F water it only takes ~150-200 square feet equivalent direct radiation ( EDR) of radiator to deliver 10,000 BTU/hr. Do you have pictures & dimensions for the radiators, to estimate the EDR?

Posted By Dana1 on 03 Nov 2016 01:37 PM
Going with suspended-tube instead of aluminum heat spreaders (even the cheap sheet metal type) is probably a mistake, since your water temps will likely be above the condensing zone most of the heating season, limiting you to no more than high 80s efficiency. Even stapling it in direct contact with the subfloor without the heat spreaders would be an improvement.

Dana- I have to re-read your insulation post and ask some questions next, however, I wanted to touch on this topic as I can go through and change the install type if required.

Originally I wanted to grab cheap eBay aluminum heat spreaders and that was going to be my insulation method. I was discouraged from this due to noise they create and having the pipe directly on the sub floor I was told would create "striping" of heat. Either of these facts true? Since I have 2k feet of tube up there the cost would be about $1400 for the cheap plates.

I do have a design in LoopCAD currently with as close to correct specs plugged in. I can post it up later when I'm home. I know that the home showed as "supplemental required" until I raised the design temp to 120'.