Hi, I have a 2 year old home/office that I'm trying to work some kinks out of the radiant system.  The home is heated via a 5-ton W-W geo unit, and there a 5-ton W-A unit for forced air backup (primarily for cooling). 

I had planned to be able to get all my heating from the radiant, but it appears I have no choice but to use the forced air for makeup when it gets cold out.  The tubing is 8" on center in joist for normal ceilings, and 5" on center in vaulted ceilings.  It's stapled up underneath the sub-floor with tube talons, then has 1/4" double bubble (about 1.25" airspace), then has about 4" of expanded polystyrene underneath (free recycled sheets).  The temperature of the radiant is modulated by an outdoor reset on the geo, but is typically 87-102 degrees.  I try to keep the water as cool as possible for efficiency, but also our wood floor is relatively unstable wood.  The floor covering is mostly brazillian cherry 3/4" hardwood with some tile.  My deltas are around 8-10 deg at about .30/gal/loop flow.  It is not a pressurized system, rather is pulls water from a non-pressurized buffer tank.  In normal conditions (25-30 deg), my home is losing about 7 BTU's/foot, or around 28K BTU's/hour.  When its in the teens, I find I'm losing about 10 BTU's/ft, or 40K BTU's.  I've got about 2500' of first story floor space with just over 4500' of tubing on 17 loops (17 loops about 275') on 2 manifolds.  The basement/workshop has no problem getting to temperature via slab heating.

* Would adding additional tubing be beneficial (I have a few extra loops on my manifolds)?

* What is the typical temperature is takes to push radiant through the subfloor/finished floor via double bubble method?

* I was of the impression you wanted to maintain a 10 deg delta and the flow is whatever it is to achieve that delta...would I be better of to flow slightly higher temps/faster flow, and let the delta be what it may in order to run nice, even temps across my whole floor, and possibly heat a bit more of the pipe?

* What is the difference between a joist track system and using some extra coil stock a making a little jig?  Would it really help me get that much better heat extraction?

* Would pressurising the distribution side give better heat exchange?  I'm thinking instead of just heating up the bottom 1/3 of the tubing, I'd probably be heating more surface area of the piping and get a better heat exchange.

Thanks in advance...

heavy gauge plates are wildly better than no plates and very significantly better than lightweight plates. That said, if you only have a 10 BTU/sq ft load, what you have should do it with a 110 degree supply temp and 10 degree drop average. that's a pretty low load though and if goes up from there you need plates. with heavyweights you could cut 15 degrees off even that temperature and pick up a COP, but whether or not that makes any sense depends on how much you could save in energy there.

10 degree Dt is arbitrary. target that in high load/high water temperature requirement areas. use a higher number like 15 or more for lower load areas. don't go lower than that or you'll waste any efficiency you get in pump energy from upsizing your pumps.

I'm not sure what you mean about heating the "bottom 1/3rd" of your tubing?

Rob,

With no pressure in the system, I assume the water would simply lay at the bottom of the tubing rather than heat up the whole circumference of the tube?  I wasn't sure if putting pressure to it and running at a lower temp might be more efficient than no pressure and higher temps.  It's says it's flowing .3/gal/min...I have to think I'm barely touching the surface of the tubing at that as you can easily get several gallons/min out of 1/2 pex water lines (obviously higher pressures than you'd want with radiant).  I'm wondering if that added pressure might help force water against the sidewalls, and therefore better push the heat out of the tubing better.  Maybe it's my imagination, but I would think that put pressure to it would get more BTU's out.

You are right...at about 110 she gets warm, but I wonder if I will have any issues with a warping flow at those temps....not to mention my COP on the geo goes down bigtime at those temps!

Greg

110 is a low temp, but if you have a reset control, you only need to run those temps when it's cold out. you can run cooler the rest of the time. Or you can add plates.

if you only had part of the pipe filled you would not be able to pump water. pressurization influences how likely you are to have air problems, but not your water temperature requirements.

I would look at adding heat transfer plates and increasing the delivered water temp, even a small amount 5 -10 degrees if you can. The thin plates will work fine at about 1/4 the cost. you can get them in 2 foot sections, may work well with the current tube talon spacing. thick plates will work better but at what cost.
with 17 circuts you may have some with bad flow due to air in the lines.
If you can get the system up to about 15 PSI and put in an air scrubber that might also help eliminate the potential bad flow problem.
This is where I would start. If you are uncertain about the thin plates and the potential effectiveness try a test area. one zone, or a couple loops.
Dan

the cost differential between the two types of plates would likely be saved in additional energy savings with geothermal heating systems.

however, in this case if the load really is that low the answer is likely to simply be a reset control and a slightly higher maximum temp, as the cost of plates would be unlikely to offer a lot of payback at such low load/water temperature requirements at all.

more likely is that even if the average load is that low, that there is some area somewhere with a significantly higher load per square foot (corner bath, great room, etc). That area might benefit from heavy plates while the rest could stay plateless, perhaps... if zoned separately.

it's all speculation without a room by room load calc though.

I can't elaborate on the info others have provided, but I have some questions.

IF you use forced air heat to augment the radiant floor heat on cold days, how is the air heated? Heat exchange with the ground loops or heat strips?

Are your exterior ground loops sized for BOTH the forced air and radiant floor systems to operate simultaneously?

A geo distributor told me I'd need separate loops for a W-W and a W-A to operate at the same time. If I wanted radiant as the primary heat source and a heat strip in the W-A for backup, then I could get by with one loop. Dave

Two things I would not recommend.

1  Staple up radiant  floor with GSHP.

2  Forced air back up for radiant floor heating.

that's a wide recommendation geotek. at low output requirements, staple up radiant floor can work very well. in "typical" homes, I would echo your recommendation for plateless applications, but heavy plated apps can work very well, perhaps with a little help in a spot or two.

not sure what your #2 is based on.

Posted By geotek on 15 Feb 2010 03:09 PM
  ,,,I would not recommend.

2  Forced air back up for radiant floor heating.

Why? If one installs a geo water to air unit sized for A/C in a heating dominated area it seems like water to water radiant would be a great way to make up the difference for heat. And, in that situation, wouldn't it make sense to have the radiant as primary heat with the air unit providing the extra btus in the coldest weather?

I'm sorry, I did mean plate-less. In a northern climate it will probably have to be heavy plates unless well insulated.

As for forced air back up with radiant floor I think a small mod-con or similar product would be a better choice to avoid air drafts. One of the reasons why we use radiant floor.

Does that make sense or am I all wet here?

one thing I have seen with geo is that many geo installs are there "for the air" first. the hydronic portion is almost an afterthought.

since there are air units there in many cases, using them for backup makes some sense, but only when the hydronic cannot keep up.

I'd be flip and say it should all be water, but chilled water heating is a different skill set than air for sure.

All, thank you very much for your replies. I had forgotten to check the site lately, and see I have some answering to do.

- As far as the heat loads go, the 7-10 BTU's a foot are correct. It's a well insulated home that uses closed cell foam, and is above the code required R-value. They are measured, not from a manual J.

- I elected not to use joist trak as my geo is capable of making 125 degree water without a problem, and even at the decreased efficiency, it's still able to produce slightly more BTU's than my home requires on the coldest day. By bumping up the temperature and extracting more BTU's, the forced air would never have to run. Joist trak would've added close to $9000 to the cost of my home (about $2/ft) when I priced it out (mostly parts, as my labor isn't worth much!). I had figured that my best option would be to run the radiant slightly cooler, and make up any difference on the real cold days with the forced air. This would allow me to run the radiant more efficiently. Running the radiant hotter, or using teh forced air to supplement, come out to almost the exact same cost-wise. I did not put strip heat in as I have the whole 5-ton unit as a backup. Having teh forced air backup has been great.

- Regarding the size of the loop field, I actually didn't hire one sub-contractor as we have a discrepancy about the loop field. He insisted that if I was putting 10 tons of units in, I needed 2-5 ton loop fields in - one for each unit. I however, was of the feeling the 2nd unit (W-A) would run very, very little, and when it did run it would in 1st stage (approx 3 tons). I put in an 8 ton loop field, headered into a 2" PVC, and then splits off to my 2 pump packs. I had my installer put a solenoid on each one that opens when the unit/pump pack turns on, preventing water from running from one unit to the other and bypassing the loop field. It seems to be working great, and I suspect I'm getting slightly higher EWT's than if I had seperate loop fields. Oh yeah, it also saved me several grand on my system by having 2 less tons of loop field!

- If I had it to do all over again, I believe I would have used some sort of coil stock, or ran more tubing. I used a triple run on my vault's/tile bathrooms, and if I had it to do again, I might have done 4, and 3 runs per joist in my "normal" areas. That would probably have been the cheapest, and probably the best, option. I believe the Wirsbo Joist Trak system runs .050, whereas coil stock is .020 roughly. At 100 degrees, Joist Trak is showing about 15 BTU's/ft in their literature, whereas just 1/2" pex is saying about 8 BTU's/ft. Assuming the relationship is linear, coil stock should be around 10.5? Coil stock would've been perfect. By running 50% more tubing, I'd be at 12 BTU's/ft.

- I do have a reset control, and it seems to be working well. I've tweaked it a few times to get it just right (where the air hardly ever kicks in), but keeps the temps as low as possible. One nice thing about my system is that by having just enough capacity, my pumps are almost always on (one zone 20 hours/daily, the other 22-24), therefore my floors are a constant temp. This works well as my floors are brazilian cherry, which is to be a very unstable wood. My understanding though, is warping.cracking floors tends to have more to do with the humidity in your house, and less to do with temperature (until you get above 125 degrees or so). Anyone have any solid facts on this. My info comes from talking with the folks here: http://www.launstein.com/radiant-heat/test-results.html

- I did have to make some changes to my switching relay. Originally I had no priority, so when we got a cold snap, all the slabs, along with the 2 main house zones, would all be calling for heat at once. My unit can produce roughly 50K BTU's. My normal loads are 30-40K. When the slabs all kick in, even with the flow shut WAY down, I'm pulling 70K+. I took and put each of my 2 main house zones on their own switching relay and have each one as priority. I put the slabs all on the relay with the zone (bigger of the 2, it's 10 of my 17 loops) that calls for heat 20 hours a day. When that zones kicks off, the slabs have roughly 4 hours to capture a days worth of heat. It works great, as that is easily do-able, and it doesn't significantly drop my buffer tank temperature as the load/supply are almost equal at 50K BTU's (slabs and 1 house loop pull about 55K an hour). Now my unit runs almost non-stop, everywhere is getting heat, and instead of having my house drop temperature when the slabs call for heat, they have to wait until mid-day sun kicks in, and the house zones are satisfied.

- Once my open system was primed, for some reason it hasn't lost prime in quite awhile. I suspect I am going to run a loop up to the 2nd story, and put an expansion tank on that loop. If it works as I believe it will, the air will all rise to the tank/air eliminator.





-

more tubing doesn't help you that much in joist installs. you can't just add more pipe and get a linear benefit. it's a diminishing benefit that diminishes very rapidly.

also I do not believe the benefit is linear with plates as part of the equation is the quality of tubing contact.

running a geo system to 120-125 is leaving at least a full COP and probably two on the table that plates could have reached. it's not about meeting load, it's about meeting it efficiently. however if your heat load is no more than 10 BTUs/sq ft, you shouldn't need water that hot even if you don't have plates. I suspect if you need water that hot that you have some areas with loads over that average, which is likely. but if not, you should be running colder temps than that and plates would have been a waste of money.

If you actually need that water temp though, figure out what that COP increase would have done to your energy costs and figure the payback on the plates. If you need 8 tons of heat total, I wager the payback characteristics wouldn't have been that bad.

Rob,

You bring up some good points. One of the things a gentleman recommended was having more air space between the insulation.

I'm not running 125 degree temps...I'm typically around 110. I'm a COP of about 3.2ish, whereas with plates I might be at 3.7. I figured that the 15% or so in savings might save me $250 a year. It just didn't seem like a good investment to me (20-30+ yr payback at current electric rates), especially when I could run 100 degree temps that get me to that 3.7 COP, and then simply use the forced air for makeup heat at a similar, high, COP.

The one area I haven't got finished yet, but I'm thinking I really should have used plates, was under my bonus room. Unfortunately the garage is now drywalled, so I don't have the option. Being that it's southern exposure, I hope that will help some. I have 4 loops, or roughly 1000' of tubing under the floor, and the heat loss was 9000+, so I should be right in the ballpark.

The 8 tons they called for is actually being heated currently, with a little room to spare, with a single 5-ton W-W, when the priorities are set up properly. Prior to the priority being hooked up correctly, 5 tons was slightly undersized. That being said, I'm spending less than $1700/year for heating. Out of the space being heated, about 1/2 is slab, and 1/2 is staple up. The only thing I may have done differently would have been to have a seperate buffer tank for my slabs/joists, and set the slabs at a very low (maybe 80 degree) temps to gain the added efficiency. I'm currently sending the same temp water through my concrete as my joists as it all comes out of the same buffer tank, without any mixing valves. The slabs however, aren't consuming much heat at all (maybe 150K BTU/day vs. 750K or so for the rest of the house). I suspect the higher COP with a seperate tank would barely be worth the cost if installation, and adds one more compenent to an already complex system.

with heavy plates you could easily drop 15 degrees in water temp, and as the temps rise so does the differential between the two. I find it unlikely you'd only pick up half a COP with heavy plates installed, that's not what I've seen in engineering data on Geo units. but you are right that you can target them to the high load intensity zones to get that water temperature reduction benefit without installing them everywhere, which could take your initial payback calculation and radically alter it. I would even consider the cost of re-finishing under that bonus room. drywall isn't free, but it's not made of gold either and if you could drop significant water temps that would just extend the payback range a bit.

are you fully heated now for $1700/year? if so, I'd say you've got $500/year on the table. 3.2 isn't very good average efficiency on a geo unit.

Rob,

While you can probably make a case for them, and I can see where they would be helpful, I've gotta think there is a diminishing return on the # of ways you can be efficient. The spray foam guys want you to seal up a house super tight ($$$), the window guy wants to sell you argon ($), the well driller says do vertical wells and it's more efficient than horizontal trenching ($$$), the geo guys says go geo vs. air source ($$), etc. In the end, the more you do, the more the return diminishes. You've gotta pick and choose where to put your money as ultimately. The majority (probably 90%+) aren't going to do it unless it makes financial sense. If you isolate each part of all the above, they have the ability to have the potential for a decent payback, but the more you do, the more the paybacks become somewhat unrealistic. Capital investments can also be scary, considering how technology, and government subsidies, are ever changing and developing. I'm not sure if the full system cost (plates, tubing, etc) is covered in the new energy star systems, or if it's just the geo units. If they are putting the 30% up for radiant related items as part of the total geo system, I'd maybe have done it.

I was also relatively uneducated as to the plates when I made my decision and hadn't been aware of the potential for cost savings.

full systems are not covered as far as I know. all I know is given the numbers you are giving, plating the bonus room probably made good sense, along with any other higher intensity load zones.

there is a diminishing return there. it's just not at the plate level.

Gotcha...thanks for the info.  I might still put plates in...I'm getting tired of not being able to get enough BTU's without them.

I don't know why I was thinking that the sole purpose for flow meters/adjustments was for basically after the fact "adjustments" should the tubing/plating not be spot on to the load.  I don't think I could have been more wrong as, unless I am now mistaken, there is very little adjustment to be made after it's in (outside of bumping supply temps at risk of warping your floor).  Changing the flow seems as though it will still put out roughly the same BTU's, you'd just have a different delta.  I had in my head that one of the nice parts about radiant was you had all the control in the work after the fact, but seems like I was wrong (except for concrete slabs...changing flow should never really affect delta as the delta should always basically the same, irregardless of flow).  It seems for in joist applications, the design is much more critical than slab installs.  I guess I should've done my homework a bit more in the first place.  Hard to know everything the first time around...only wish I had a second chance now!

I no longer design sub-floor systems without use of heavy plates. They improve output and lower design water temperature, which will affect comfort. Heavy plates give the designer a margin of error unavailable when bare tube is suspended below or stapled to the sub-floor. Designing systems within certain margins makes for simple and efficient hydronic heating.

While flow is one of the lesser factors (temperature being the greater) it is still a critical factor. The speed and temperature of the heat transfer fluid determine available energy but the emitter and heat load determine output reflected in the delta T.

When working on (or converting) old gravity driven boiler systems delivering high delta Ts I have to smile when thinking about the various manufacturers dictating this delta T or that. Controlling room temperature involves a sensor, minimum flow and minimum water temperature for the present load.

That is; if it is relatively warm outside the flow and water temperature can be reduced and both can gradually be increased as the temperature outside drops and the load increases. As most pumps are fixed speed we must size them for design conditions and adjust water or depend on the thermostat to anticipate the output of our radiation. This can work very well but a proper heat load is necessary to ensure success.

I find flow meters on my radiant manifolds indispensable when balancing disparate heat emitters used on the same system.

Factory training, a few good books and 5 or 10 years experience and you're all set.

DIY design; always the first mistake.