The Wall
Forum / Radiant Heating / Radiant Floor Heating System 404x40 shop
  • Post a Reply to this Thread

    Radiant Floor Heating System 404x40 shop (56 Posts)

  • CarGuy CarGuy @ 5:15 PM
    Contact this user

    Radiant Floor Heating System 40x40 shop

    Hello everyone, I am new to this forum but have been reading it for a while now.  The information is excellent, and I appreciate the professionalism here (not being a forum/message board guy due to the usual things that go along with them).

    A bit of background info first---

    I read here not to long ago, unfortunately it was after I purchased my radiant floor heating products, something to the effect that; "the company that you buy your components from is only as good as their after-sale service in helping you make it work properly"... like I said something to that effect.

    Gentlemen, (and Ladies), I am your next victim of the growing buyer beware world of DIY in-floor heating!

    Being mechanically inclined and able to do a lot in the build of own home I decided to take on our slab-on-grade hydronic in-floor heating system in a 40 x 40 shop.  After pretty careful research I decided to buy my components through an internet vendor, Pex Universe, with good pricing and what seemed to be a lot of PRE-SALE tech advice in what to buy etc.  And honestly, they and many other online outlets and system suppliers make a system like mine seem like a pretty simple task for the mechaincally inclined layperson.
    Then I found this forum.

    I noticed a number of people here who seemed to be be much like me were having problems, and the task which I was about to undertake seemed much more techinical and complex than I first thought.

    I approached the product vendor for a very basic request as follows "Can you supply a basic 6 branch installation diagram for the system/components we purchased from you to ensure proper and safe installation?"  Their instant first answer was "We do not design systems".  A bit of knee jerk, cover our behind answer to my pretty simple question I thouhgt, but the fellow did elaborate and agree to examine my purchase orders and say, "I will look at the list you have and try to give you some basic installation guidelines."  Great, so I though anyway.

    Several very proffessional and polite emails, phone calls and faxes later, simply asking for a rough guide as to an order of installation... even a photo would have helped... and 2 months later, not one reply!  Today we finally told their customer service rep how disappointed we are... and here I am.  

    I have a pretty good idea of how to put it together just want to make sure I have everything I need and have a pro verify the order of components so I do not make mistakes similar to others I have read here.

    I have the following:

    A 6 branch EverHot stainless steel manifold
    1/2 oxygen barrier pex going in the floor, 6 branches none more than 300'
    Taco 007 circulator pump
    Honeywell AM101-UT-1 Mixing Valve
    3/4" Threaded, Spirovent Jr Air Eliminator
    RX-15 Radiant Extrol Expansion Tank (2.0 Gal Volume)
    I know a lot of folks don't like this but it is what I have for heating the water --Takagi T-KJr2-IN-LP Tankless Water Heater and the Remote Temperature Controller for T-KJr (running on Propane)

    Thank you!


    This post was edited by an admin on April 29, 2011 10:32 PM.
  • Rob Blair Rob Blair @ 8:05 PM
    Contact this user

    First off.....

    Get rid of the mixing valve and the 007.  You will need a high head pump and can set the water temperature with the remote temperature controller.
    Send me a PM if you want to discuss more.

  • Chris Chris @ 8:23 PM
    Contact this user

    I Tend To Disagree

    I would rather see you pipe pri/sec or utilize a low loss header. Then the 007 could be used as a system pump for the radiant. I would also use the mix valve as a means of protection. You would need a properly sized circulator for the head of the tankless. The low loss header would be better piping option in my opinion. You can then have control of the flow rates on the radiant without having to worry about trying to maintain the min of 2gpm and max of 4gpm through the tankless. Personally I wouldn't use the 007 I'd use a Grundfoss Alpha or similar pump.

    You may want to consider finding another heat source..Hx only has a 1yr warranty.
    "The bitter taste of a poor installation remains much longer than the sweet taste of the lowest price."
  • NRT_Rob NRT_Rob @ 10:36 AM
    Contact this user

    probably no need

    for primary/secondary or the mixing valve if this is dedicated to a single zone slab system... as it should be.
  • CarGuy CarGuy @ 10:31 PM
    Contact this user

    Thanks for the info...

    As I said I am green, no wait, very green, at this.  This is my first dive into radiant heating.  I am capable of the BASIC assembly just was relying on the assistance of the vendor to guide me in their proper order of installation/connection of the components they sold me. 

    Sounds like the Taco 007 was a bad choice for a system pump?  Rob I will send you a PM to discuss the high head option.

    HVHEHCCA - So you are saying that I can use the Taco (although not your personal choice) but would need a second pump on the tankless?  I understand what the low loss header is but you said "it would be a better piping option" and I was wondering, a better piping option over what?  Sorry for my ignorance in this area, some of the terminology and design is quite new to me.

    Thanks again

    This post was edited by an admin on April 29, 2011 10:38 PM.
  • Ironman Ironman @ 11:05 PM
    Contact this user

    Some Other Things to Consider

    There will most always be some difference of opinion or preference in designing a radiant system. I would agree with Chris about using primary/secondary piping because of the high head loss through the Takagi tankless. (If you don't understand any of the technical lingo we're using, please let us know). And I believe any pro would agree that a tankless is a bad choice for a heat source. It doesn't meet code in many places. A mod/con boiler would be the proper appliance.

    With that said, let me ask some questions and make a few more suggestions:

    1. Have you done an accurate heat loss calculation for the structure? If not, I'm going to go on a "rough estimate". Please don't use this in place of the heat loss calc. It just gives me a reference point to explain what some of your requirements may be.

    Your building is 1600 sq. feet. Assuming 25 btu's pr. sq. ft., the load would be 40k btu's at design temp. Again, don't take this figure  as an accurate load calc.  You will have 1800 ft. of pipe as your emitter. 40,000/1800 = 22.22 btu's pr lineal foot of pipe. That's a little less pipe than I would want, but it's doable. You'll just need a little hotter water or more gpm. In a home we normally design for a 10 deg. delta T (temp drop) in the loop. This is to prevent cool spots in the floor. In a shop, you could have as high as a 20 deg. delta T. This means you only need to move 1/2 the gpm that you would vs. a 10 deg delta T. You'll still get the same amount of heat output, just not as evenly. Without going through all the math formulas, this means that you'll need about 4 gpm through the floor (total). That comes to just under 0.7 gpm, pr. loop. The Taco 007 and the 1/2 in. pex are sufficient for this, but not much more. The Grundfos Alpha would cost more but save some energy. The 007 will in no way be capable of also overcoming the head loss of the tankless. That is why you'll need to do some form of primary/secondary piping and have a separate circulator for the tankless.

    2. What type of insulation is going under the slab? You should have 1 -2 in. of blue board underneath and 2 in. min. around the perimeter. Over 40% of the heat loss is through the perimeter. Make sure you properly insulate.

    3. A mixing valve is a fixed temp. device and as such is not a good way to control water temp in a slab due to its high mass. There's something involved known as "the flywheel effect". It takes alot to get it going and alot to stop it once it is going. Simply using a mixing valve and a thermostat can cause the high mass of the slab to continue to give off heat after the stat is satisfied and thus over-heat the area. The proper approach is to control the water temp based on outdoor temp: the warmer it is outside, the lower the temp of the water in the slab. This is known as outdoor reset and can be accomplished in different ways. If you had a mod/con boiler, it would be capable of doing this for at least one zone. In lieu of that, there are other methods available such as variable speed injection controls or modulating valves with reset. By varying the water temp to match the load, The "flywheel effect" is eliminated.

    4. Don't attempt setback or using a setback thermostat for the same reason as stated above.

    5. You'll also need a back flow preventer / fill valve and a 30 psi relief valve. You may also need a Extrol # 30 expansion tank. The # 15 may not be enough. Not certain without checking sizing tables.

    6. I'm not impressed with the brand of manifold you have, but if it has flow setters and purge valves, it should be OK. Providing that it doesn't leak, of course. I just had that experience with one of these Chinese manifolds that a customer had gotten from the Internet

    I hope we've been of help and ask more questions if you need further explanation
    Bob Boan

    You can choose to do what you want, but you cannot choose the consequences.
  • Brad White Brad White @ 12:17 PM
    Contact this user

    What Ironman said

    I cannot add anything to what Bob wrote, but I did make a pass on a layout using LoopCAD (because it is fun to do). I have no idea about your actual heat losses, meaning no idea if radiant floor heating alone will do the job.

    That said, I forced the layout to provide 30 BTUH per SF and a slab temperature of not over 85 degrees F., which is a limit for most comfort applications. (It works out to 84F by the way.) The idea is, being at the upper range of what you would want, you can either heat the building with it and if too much, you can adjust the water temperature downward to suit. No greater output is possible without giving a feeling of over-heating to occupants, but you can always back off. Temperature adjustment is your friend here.

    As a caveat, this is a SAMPLE layout and is not a design for you to go run with it. There are sub-slab details such as footings, haunches, drains, ledge, etc. I cannot possibly know about. It is just an example of how I would heat such a generic open plan space with radiant tubing embedded in concrete.  A benchmark for where you may be and where you may want to go.

    I hope this helps!

    "If you do not know the answer, say, "I do not know the answer", and you will be right!"

    -Ernie White, my Dad
  • Gordy Gordy @ 10:10 PM
    Contact this user

    tubing layout

     Interesting tubing layout Brad. Does this program automatically provide a tubing layout by btu density required, and room shape?

  • Brad White Brad White @ 9:53 PM
    Contact this user

    The LoopCAD Program

    is by Avenir Software out of Calgary. I used to have their regular LoopCAD program but it was limited. Their 2010/2011 Pro version is much more robust.  It actually has a BIM engine in the program to do some building modeling including heat losses. So yes, you can define walls, windows, heights, the basic building.

    You can create your own building or import AutoCAD .dwg files as a basis. It is actually fun.

    You enter your data, floor structure or panel type, covering, optimum spacing, etc.

    Once you establish your circuit entry points the program draws the circuits aiming for equal circuit lengths, usually within 10%. It also selects the correct water temperature and a diagnostics tab walks you through any issues such as too hot a floor temperature or not enough RF heat which then needs supplemental heat. 

    If you find the water temperature too high, you can change your parameters, say narrower spacing, more circuits, higher or lower delta-T, etc. and the program re-draws and re-calculates the layout. It also furnishes a bill of materials and pricing (Watts is an add-on, there are others such as Uponor).

    Avenir offers a trial version if you want to test drive it. Check them out, Gordy!
    "If you do not know the answer, say, "I do not know the answer", and you will be right!"

    -Ernie White, my Dad
  • CarGuy CarGuy @ 10:47 PM
    Contact this user

    This is a huge help...

    Just want to thank you folks as already this is a huge help.  Unfortunately, I bought my stuff first... better late than not all, or after my mistakes were made though. Some of the technical terminology is new to me but I am educating myself as best as possible as I move along.

    IRONMAN/Bob - I will add info/answer the questions to th ebest of my ability:

    1.  I did have a bit of help with a rough heat-loss early on... done last year by a company whose products I did not buy after finding too much negative information about them.  It was based on a different design though.

    2.  I have 1.5" R7.2 EPS going under the slab... I needed a high PSI rating due to weight load on the floor and got exactly what I needed back before Christmas... then the ground froze a week later and I was done until now... (which has worked out to be a positive thing, given this situation). I will be putting in the floor insulation in next week or two (trying to do this in my spare time, after work).  As for perimeter, I went with the same insulation around the perimeter of the entire building, 2 feet down, including the door openings (this is actually done).  For the door openings, I cut the top edge on a 45 degree so the concrete would have a finished edge and set the insulation across the door openings in line with where the door seal will meet the floor.

    As a side note - I have R40 in my cieling and R22 in my walls.  Doors are all American made (not Chinese) R16 overhead doors.

    If I get a Grundfos will I need the secondary pump?

    3.  Yes, I do understand the flywheel effect. So should I eliminate the mixing valve and thermostat even though this is a shop, for personal use, and it will probably kept at a fairly constant 68 degrees?  My thought on the mixing valve and thermostat, (not knowing at the time that there was different/better way), was simply to keep the long time for the slab to heat and cool in mind when setting my temps. 

    4. I did not consider a setback thermostat for those reasons.

    5.  The backflow preventer and relief valve, yes I need to get those, thank you.  I do have a fill/drain valve off of the manifold.  I checked the chart on the expansion tank when I bought it and the 15 was good for this application (you know I was going to buy the 30... just in case... DRATS!)

    Brad - Thanks for taking the time on this.  My slab is very basic, nothing fancy other than the strength of the mix I am going with.  I just tamped the start of the base last week which is about 5 inches of a mix of black roak and sand stone mix over my native soil (clay).  I will be adding a few inches of clear stone next week to really level the floor off well and get a nice even thickness in the slab.  The actual concrete will be 5 inches thick.

    On the Takagi Heater - Understood completely, and I appreciate the honestly.  Unfortunately it is what I have, and I what I could afford, and I was obviously told otherwise by the internet preditors.

    Interesting tidbit--- so many people have said to me... oh a staright 40x40 in-floor system... that should be pretty easy to do, obviously they have not attempted this, or if they have they probably did not attempt to do it correctly.

    This post was edited by an admin on May 1, 2011 10:52 PM.
  • Ironman Ironman @ 11:27 PM
    Contact this user

    In Response...

    To your questions:

    You'll need 2 circulators, one for the primary (or boiler) loop, and one for the secondary loop. If you don't understand primary/secondary piping, look under the resources tab above or get Dan;s book on it. You may need a 3rd circ. if you choose variable speed injection to control the water temp.

    You'll still use a thermostat. If you use a good mixing control and get the reset fine tuned, the thermostat will act like a "high limit".

    The fill valve that I referred to is a pressure reducing valve (PRV) that maintains a minimum of 12 psi on the system. You can get it with the back flow as one assembly. Honeywell # FM911 is one. Can't recall the Watts # right now, but that is another one.

    We will have to know the head loss through the Takagi at maximum gpm to size the primary circulator. Look in the install manual or ask Takagi.

    My laptop crashed today, so I can't attach a piping diagram. I would suggest going to Tekmar's site and looking up their essay on injection mixing. It will also show some piping diagrams. Their 356 injection control would be my choice. Just did one today.

    You will also need a tubing un-coiller and one or two others to help you lay the pex as well as some means to fasten it before the concrete is poured. You might be ahead of the game if you could hire an experienced pro even for a day to give direction.
    Bob Boan

    You can choose to do what you want, but you cannot choose the consequences.
    This post was edited by an admin on May 2, 2011 11:30 PM.
  • NRT_Rob NRT_Rob @ 10:55 AM
    Contact this user

    I have to contradict

    this is a single zone slab with 6 loops and no real floor coverings. You can run pretty low flow rates and still not pay much of a penalty for high water temps. If you had even 3 GPM you'd be absolutely fine, and you could probably do it with 2 GPM and still keep your max temp under 100.

    this can be as simple as one pump out of the takagi and into your manifold, with just expansion tank/air separator and fill hardware. I am seeing the TkJr would have about 10 feet of head loss at 3 GPM. I don't use taco so I don't know the 007 curve, but as long as 2 GPM is an acceptable minimum for this unit then I think what you have would work, your loops would have a very low pressure drop at that point.

    No need for mixing... with an uncovered slab you are very unlikely to ever need more than 90-95 degree water here, if that. You might want a better thermostat, if you end up with any temp swings. one with PID logic like a tekmar 508 should handle it just fine.

    In short: don't use the mixing valve and everything else should be usable. but do NOT do domestic hot water from this heat source. Set the takagi to 90 or 100 degree water to start with, run your loops 12" o.c., check the head loss, dial your loops to 0.3 GPM each or so (depending on takagi flow minimums) and let it go.
  • Chris Chris @ 11:19 AM
    Contact this user

    Direct Piping

    I wouldn't direct pipe on this unit. You must maintain a minimum of 2gpm thru the unit. I would utilize a low loss header although you could primary/secondary if you wished. Your warranty just went to 1yr on the unit and I feel the LLH would be a better fit for a novice to maintain min flow rates. Size your unit pump based on the head loss in the unit. For my system pump I would use a Grundfoss Alpha or similar pump. 
    "The bitter taste of a poor installation remains much longer than the sweet taste of the lowest price."
    This post was edited by an admin on May 3, 2011 11:25 AM.
  • NRT_Rob NRT_Rob @ 11:35 AM
    Contact this user

    with a single zone system and flow meters

    minimum flow maintenance is a "dial it up" proposition. so why add a pump and LLH? if the 007 can't do it, upsize as necessary. LLH and a second pump is simply a waste of time and money in this case and all it would do is increase the risk of short cycling.
  • Chris Chris @ 11:46 AM
    Contact this user


    If it were you or I installing it I would be on the same page with you.  We are dealing with a green homeowner. He already picked a poision for the equipment and has no knowledge of gpm, flowrate, etc. Sure we can tell him but do you have confidence he can apply it? 

    As for short cycling. Its an on demand heater. Would you consider a faucet running for 45 seconds a short cycle? Isn't this units purpose in life is to short cycle..
    "The bitter taste of a poor installation remains much longer than the sweet taste of the lowest price."
  • NRT_Rob NRT_Rob @ 12:15 PM
    Contact this user

    for periodic demands, yes.

    for continuous heating demands, no. you'd short cycling 100x or a 1000x more in heating than you ever would washing hands.

    and I think adding LLH and pumps makes it more, not less, complicated. harder to screw up too. Reading a flow meter is not that tough.
  • Ironman Ironman @ 7:48 PM
    Contact this user

    Numbers Not Adding Up


    I highly respect your opinion and I certainly appreciate keeping a system as simple as possible. But the numbers aren't adding up.

    Without knowing CarGuy's location, and absent a heat loss, I took a rough figure of 25 btu's pr. sq. ft. That would give a loss of 40k btu's. Add another say 8k btu's for back loss = 48k btu's. If you could run 2 gpm, that would be a 48 deg. Delta T. With 100 deg. supply, that would mean a 52 deg return. How is 52 deg water going to heat the space to 65 deg? I know we could raise the supply temp., but we're getting pretty high for a slab, not to mention a very uneven floor with that un-acceptable Delta T.

    There is also (as Chris mentioned) the problem of the high head loss through the Takagi. At 4 gpm the head is 20 ft., and at 2gpm, it's 9 ft. Add the piping loss of about 12 ft., and we're over 20 ft. at even 2gpm.

    If you go with a 26-99, you could move enough gpm for a reasonable Delta T, but it's more than twice the cost of two 15-58's or 007's. And he already has one 007.

    I also don't understand your reasoning that primary/secondary will cause short cycling when the purpose of it is to allow complete heat transfer at different flow rates between two circuits. The Takagi modulates from 20k btu's to 140k. Would not doing what your proposing (2 gpm)  cause short cycling sooner than having 4 gpm?

    Please, don't mis-understand. I'm not attacking. And If you see something that I don't, I'd love to have you point it out to me. I'm just not able to see what you do from the info we have so far. :)
    Bob Boan

    You can choose to do what you want, but you cannot choose the consequences.
  • NRT_Rob NRT_Rob @ 8:09 PM
    Contact this user


    I've never seen a shop with an average heat load of 25 BTUs/sq ft, especially not "plus back loss", so I think your numbers are very high. Also he notes that his shop is very well insulated in a later post, so I think your heat load is probably at least twice as high as he's likely to see no matter where he is.

    Admittedly, I'm guessing and could be wrong, and you're definitely right to call me on it though, thanks for doing so. certainly a heat load should be done to settle that issue.

    a 15-58 can do 16 ft of head at 3 gpm, which was my guesstimation pump. I went off that but I just got off my lazy butt and found that the 007 is weaker, only has about 11 ft of head. so he'd get a couple of GPM if your frictional loss numbers are correct. again you're right to call me out, he'd never get 3 GPM here with this pump. *clears throat*

    but I do think 2 would do it. he probably isn't over 25kbtus, 2 gpm would give him a 25 degree temp drop, 100 to 75 for an average close to 90 which is just about enough to put out 25 BTUs/sq ft.

    primary/secondary could cause short cycling if primary flow is significantly higher than secondary flow, as it can reheat its own return water. with primary only this is impossible and you eliminate a potential issue... though one that can be rectified, if caught, by changing settings as well. but I don't see it as a benefit.

    if his pump WON'T do this, he wouldn't need a 26-99... a 15-58 would run away with this. but I don't think he'll need to upsize, so I'd gamble anyway, even if the step up was a 26-99.

    I was pretty glib and sloppy there though and I just had to justify myself with real numbers in response to your post, so thanks for poking. And attack all you like: I am very comfortable with being calling out on merit arguments. I've been wrong once, maybe even twice... since supper....
  • CarGuy CarGuy @ 8:24 PM
    Contact this user

    I am located in

    Thanks again for all of this info...

    I am located in Southern Ontario, Canada... Brantford is the city actually.  The climate where I am is actually not bad anymore for extreme cold.  Usually Jan and Feb are the cold months... but even they do not see extreme or extended cold periods on a regualr basis.  

    Our climate is similar to some central and northern States, actually less snow than places like Buffalo NY and areas of Michigan.  A lot like central/northern Ohio where we are.  80% of our customers are throughout the U.S. so we are always having the weather comparisons.

    Yes the shop is extremely well insulated, and in fact I am literally sealing every crack and seam I can find on both the exterior and interior before fall.

    This post was edited by an admin on May 4, 2011 8:27 PM.
  • Brad White Brad White @ 8:55 AM
    Contact this user

    I tend to take a Zen Judaism approach

    to radiant heat when it is an existing or "fixed variable" system as yours seems to be. I am not sure if you installed it yet or just have the parts and materials. 

    If the latter, there is room to make it ideal or close to it.

    If it is the former, that it is all installed and buried, you have one variable left and that is temperature.  This can be achieved by varying your primary supply water temperature and/or to a lesser extent, by varying your flow rates.

    Flow may be a limitation too, how many circuits at what diameter will dictate the up-side. Higher flow rates mean narrower temperature differences, but do not get too excited about this.

    There is a simplicity in the variables. Temperature is the "big dial". Flow is the "fine tuning dial".

    For example, say you now have say 130F water entering the tubing/slab and returning at 110F, (20 degree drop), your average water temperature will be 120F. Were you to double your flow (assuming the flow rates are within your tubing limits), your average water temperature would rise by only five degrees, to 125F. (Entering at 130F, leaving at 120F, naturally.)

    Not a huge jump considering you just quadrupled your pump pressure losses and increased your pump Watts by a factor of eight.

    So your best variable, if you are short on heating capacity is to increase your initial water temperature off your boiler. The downside is possible "striping" or perceived hot and cold lines at and between the tubing. Over time this will even out to a fair degree when the entire slab is up to temperature.

    If the slab temperature approaches 84 to 85 degrees and still does not heat the space (probably will heat it, but saying), you would supplement the capacity with other forms of radiation or bring it up the walls.  Just illustrating some options.
    "If you do not know the answer, say, "I do not know the answer", and you will be right!"

    -Ernie White, my Dad
  • CarGuy CarGuy @ 11:25 AM
    Contact this user

    Pump not doing the job?

    Well I have this system up and running, and I am definitely a believer in Hydronic Radiant heat! 

    The TACO 007 is only reaching 1.6gpm.  The specs on this pump are 1/25 hp, 0-20 GPM and 0-11 ft head range.  Why would this not reach say even 3gpm for this system?  Is it due to the amount of PEX in the floor (approx 1,700 ft)  and/or the restriction of the on-demand heater?

    The temp setting for the room is 59 degrees F and the system rarely comes on to maintain that right now.  The water heater is set at 99 degrees.  It is not cold outside now however, and it will be getting cooler soon so I want to change-up the pump...  should I go with the higher priced Grundfos 26-99 (which might be overkill)?  What specs should I be looking for in a pump to do the job here?

    Thanks once again,

    This post was edited by an admin on November 21, 2011 11:30 AM.
  • zacmobile zacmobile @ 1:31 PM
    Contact this user

    it's the tankless heater

    the tankless heater has a huge amount of pressure drop through it, that's why several people earlier on in the thread recommended either a bigger pump or primary/secondary piping, I really doubt the 26-99 would be too big.

    It could be a combination of that & the tubing though, what is your longest loop length?
    This post was edited by an admin on November 21, 2011 1:33 PM.
  • Chris Chris @ 1:35 PM
    Contact this user


    How many loops and what is your longest loop length? 1.6gpm running at the standard 10 degree delta-t for radiant means your delivering 8,000 btu's thru that radiant. What is the difference between the outgoing water temp and the returning water temp?

    Pump can't do the job. Anwser the above and we can help you out. I think Irons reply a few posts up explains what is going on and why I recommended you pipe pri/sec or use a LLH.
    Is this system using 100 percent h2o or antifreeze mix?
    "The bitter taste of a poor installation remains much longer than the sweet taste of the lowest price."
    This post was edited by an admin on November 21, 2011 1:42 PM.
  • CarGuy CarGuy @ 5:07 PM
    Contact this user

    Info ---

    6 loops, the longest I believe is 295ft.  All loops are pretty close to the same length.  I will check back in my notes to very the longest 100%.

    The system is 100% water, no antifreeze.


  • Gordy Gordy @ 6:39 PM
    Contact this user

    Did you adhere to

     The original design plan you PDFed in the beginning of the thread?  ther were 9 loops at 300' +/- Just wondering where this 1.6 gpm is playing in.

    If so at a 15*DT you are looking for just under 1 gpm per loop at just under 14' of head in the longest loop. Total flow needed for the 9 loops is 8.26 gpm. each loop averaging 5000 btus +/- . 

    You can not get that kind of flow rate through the Jr. If you are only trying to achieve 3 gpm at most through the jr. thats .33 gpm through each loop.

     If there are 6 loops spacing then changed from original plan. Assuming 1' spacing with 300' loops. At 3 GPM thats .5 GPM in each loop. Head is still 13.2'.  Nearly half of the original design flow rate. Was the heatloss wrong?
    This post was edited by an admin on November 21, 2011 6:57 PM.
  • Chris Chris @ 7:12 PM
    Contact this user

    Btu Delivery

    Based upon your 1.6gpm deliver your are given each loop approx .27gpm at less then 1ft of head. .The head in the tankless is beating you up.
    "The bitter taste of a poor installation remains much longer than the sweet taste of the lowest price."
  • CarGuy CarGuy @ 9:35 PM
    Contact this user

    6 loops--

    The designs in the earlier posts were given as examples after I had the 6 loop manifold bought and installed.

    I have a line on Grundfos 15-58 which has 19ft of head which I may give a shot.  This was one of the pumps recommended earlier on as well.

  • Chris Chris @ 9:44 PM
    Contact this user

    Don't Think So

    You have a line? Sounds like E-Bay and you'd be wasting your money like you have wasted the great advice and all the recommendations here. Why ask us if you have all the solutions?
    "The bitter taste of a poor installation remains much longer than the sweet taste of the lowest price."
  • CarGuy CarGuy @ 9:56 PM
    Contact this user

    I have taken

    No not ebay, don't touch that stuff.  I have taken most every bit of advice given here, that is within my means to do.  In fact, off this board I was in communication with one the posters most every step I made.  He was reasonable, understood what I already had, and gave advice as how to best utilize it.  He said to give the 007 a shot, so I did, he clearly added that the worst that could happen is that it would not do the job, and it didn't.

    So you are saying that the 15-58 will not do the job either?  I would certainlly go with the 26--99 if I have to.

    I wouldn't ask for advice I did not truly value the answers.  I take all the advice, and do my own due dilligence as well, and determine a course of action from there.

    This post was edited by an admin on November 21, 2011 10:01 PM.
  • Chris Chris @ 10:08 PM
    Contact this user

    Not Picking

    On you. Read Irons post back in May. He gave you the anwsers.
    "The bitter taste of a poor installation remains much longer than the sweet taste of the lowest price."
  • zacmobile zacmobile @ 10:34 AM
    Contact this user


    like Chris says, don't bother with the 15-58, it isn't much different from the 007. get the 26-99 I can pretty much guarantee it won't be too big if you're only pumping 1.6 gpm with a 007. but make sure you get the 3 speed so you can trim the speed to what you want.
  • CarGuy CarGuy @ 12:22 PM
    Contact this user

    Sounds good

    I will look for a 26-99 3 speed in my neck of the woods.  Seems a lot harder to get the equipment here.

    Thanks again,  I will update once I get it in place.

  • cattledog cattledog @ 3:48 PM
    Contact this user

    pumps in series

    Your lowest cost option may be to put another 007 (or the 15-58) in series with the pump you have. See Ironman's comments of May 4.
  • CarGuy CarGuy @ 9:24 PM
    Contact this user

    Yes, great post I read over and over for sure.  Had I been able to grasp then, what I now have, I would have went that route (either two pumps or just dropped the $250 for the 26-99.) 

    I believe I have a use for the 007, so by the sounds of it my best option (and most basic for me to do) would be to change out the 007 with the 26-99.

  • Ironman Ironman @ 9:32 PM
    Contact this user

    Driving a Spike with a Tack Hammer

    I would still recommend that you use pri/sec. piping with two smaller circs. You may find that the 007 is sufficient for the radiant loop. I don't have pump performance charts open in front of me so that may need to be checked out.

    We don't know all the details or sizes of your piping, particularly between the heater and your manifold. Some pics. would be helpful.

    But to simply keep getting a bigger circ. is like getting a bigger hammer each time. You may eventually get a big enough one to drive the spike, but you may also cause un-necessary damage as well. The Problem is the high head loss of the Takagi; it should be isolated on its own circ. A 15-58 on high speed is the equivalent of an 008. That may be enough for the Takagi alone and would cost a lot less than a 26-99. You need to check the charts for the Takagi and the circs and stop guessing. Simply using one over-sized circ is asking for trouble.

    "Do it right, do it once".
    Bob Boan

    You can choose to do what you want, but you cannot choose the consequences.
  • CarGuy CarGuy @ 4:55 PM
    Contact this user

    Got the new tack hammer installed..

    and you're right still not doing the job.  (I had ordered it right away) Most I can muster from the Grundfos 26-99 is 3.2 - 3.4 GPM... and that is on Med speed, switching to high does not change the flow. 

    I tried to fill/purge the system as best I could after loosing the water between the pump/line shut off valves and suspect it may still not be full.  Would this affect my GPM?

    I will take some photos.  I guess what still sort of burns me a bit is being lead down the garden path by the component seller pex universe and not provided ANY follow-up service as promised by them (I am almost over that).  Then I found this board and after reading and being in contact with one of the posters I eventually decided there was a lot more to this than the product seller had lead me to believe, so I contacted a few local Pro's to help. 

    Let me get some photos posted over the next day or so.

    Thanks again,

    This post was edited by an admin on December 8, 2011 10:21 PM.
  • NRT_Rob NRT_Rob @ 3:09 PM
    Contact this user

    by what measure

    is that "not doing it"?

    are you maintaining temp? If the unit short cycling?

    If not, what are your water temps?
  • CarGuy CarGuy @ 3:04 PM
    Contact this user


    Here is a photo of the system with the Grundfos in place.

    Any comments are appreciated.

  • CarGuy CarGuy @ 3:18 PM
    Contact this user

    Good question

    Hi Rob, good question, "By what measure", since I have no real experience here I guess I honestly do not know the answer to that.  It is not short cycling.  I have the room temp set at 57 degrees and the system comes on every few hours in these 40 degree temps.  It only runs for 10-20 minutes to bring it back up to temp. 

    Seems to be using more propane than I personally expected, but then again I have no "actual" info to draw from, only 3rd party propane useage talk.  I just now knocked my room temp down to 55 as I am not using the shop much right now.

    The heat "feeling" is amazing, far better than imagined.

    When you ask what are my water temps, do you mean on the "in" and "out" sides while the system is at room temp and not pumping?  I have a temp gun I can check with once I know this info.

    This post was edited by an admin on December 8, 2011 3:22 PM.
  • NRT_Rob NRT_Rob @ 3:23 PM
    Contact this user

    then I would say

    as I said before, you probably could have stuck with a smaller pump. if you are maintaining temp and are not short cycling, your flow portion of this equation is satisfied.

    propane usage will be based on your heat load and something like 80% efficiency on the takagi I believe. I don't think you have a real heat load so any number would be a guess.
  • zacmobile zacmobile @ 4:53 PM
    Contact this user

    what temperature?

    what do you have the setpoint on the unit controller set at? it sounds like it may be a little higher than needed based on your description of it's operation (being off for so long & running in relatively short bursts) though the unit may just be plain oversized.

    nice install by the way, I was envisioning a real rats nest. better looking than most "trades people" installations I've seen.
  • CarGuy CarGuy @ 10:33 PM
    Contact this user

    Temp setting for water heater

    The temp setting for the water heater is 102 now.  I was at 100 and that worked well, I was just experimenting a bit.

    I may be a little off on my 10-20 minute run time estimate.  I have been working out there organizing and not really paid attention to the time the system is pumping but I will tomorrow.

    Thanks on the system.  I cannot take all the credit.  I did the pex layout, pex install, the manifold install and plumbing and a few other things myself (with some advice from one of the Pros on this site), then I called in a local Pro to plumb the remainder of it.

    Here is a photo of my Pex install... would you believe that there are NO pex uncoilers available for rent (or for sale) within 1/2 hour of where I live... and some rental places had never even heard of it!!  So I improvised and used the Pex carton with a few other handy items around the garage, modifying it into and crude "uncoiler" .  I did it myself to save any family or friends that wanted to help the brunt of an occassional bout of frustration!

    This post was edited by an admin on December 8, 2011 10:37 PM.
  • furnacefigher15 furnacefigher15 @ 11:56 AM
    Contact this user

    Looks like a clean install

    The install looks clean, but a couple things.

    #1 - I don't see a relief valve in the photo (may be angle of pic). If there is not one, get one in pronto.

    #2 - The heater will cycle no matter what you do b/c its output capacity is more than the load and flow you are giving it, and I would leave it that way, unless once its colder, you don't get the heat you want. As long as the off cycle is lengthy.

    3/4 inch pipe is good for 40,000 btu at 4 gpm and you are moving fairly close to that at 3.4 you will move 34000 btu

    The tankless is likely a 3 stage burner (also has internal bypass to temper to setting) with a peak output near 100,000 btu and lowest setting just a hair above 40,000 btu

    If you find you need more heat, you will need more flow (bigger pump, or a 2nd in series), but there is a good chance everything will be fine as is.

    Good luck
  • CarGuy CarGuy @ 4:11 PM
    Contact this user

    Pressure relief valve?

    Thanks Furnacefighter15.  Do you mean a pressure relief valve? There is a system air vent/bleed on the manifold.

  • Chris Chris @ 4:35 PM
    Contact this user

    Relief Valve

    I also don't see one. Need to install one and make sure its a 30lb boiler relief.. Your delta-t, which should be 10 degrees for radiant slab is only going to allow you 15,000 btus delivery.

    3gpm x 10 x 500 = 15,000 btus

    Why did you pick this unit that can only make 6.6gpm when your design called for more then 8gpm? I also noticed the design shows 40% gylcol mix thus why the high head of 14' in the radiant. The Takagi manual clearly states that the pump must not deliver no more then 4gpm through the tankless. At 4gpm head is 20'. At 3 gpm head looks to be on their chart 12'.

    Your getting pretty much all the unit can do.
    "The bitter taste of a poor installation remains much longer than the sweet taste of the lowest price."
    This post was edited by an admin on December 11, 2011 4:59 PM.
  • NRT_Rob NRT_Rob @ 8:32 AM
    Contact this user

    10 deg dt??

    this is a shop. you can push 30 degree dt easily. the heater is built for 60+ degree DT.
  • furnacefigher15 furnacefigher15 @ 10:09 PM
    Contact this user

    Yes pressure relief valve

    The air vents are fine. The pressure relief valve is to allow the water an escape in case something goes wrong with the compression tank, or burner, or a fire etc....

    If you don't have one sized at 30 pounds blow pressure and at least 510,000 btu capacity, shut off the boiler NOW, because you have a possible bomb on your hands!!!
  • CarGuy CarGuy @ 10:16 PM
    Contact this user

    Will get the pressure relief valve

    I will get the prressure relief valve in place, thanks!  Chris, are you reffering to the 26-99 pump max'ing at 6.6 gpm?  I didn't have a deifinate design, I honestly did not realize/understand that I would need 8 gpm fo this (to get a specific outcome)... the examples Ironman posted were for my layout of the pex at the time. 

    I already had all the components and wanted to try to make them work if possible.  I followed Rob's advice and used it all except the mixing valve and it seems to be working exactly as he predicted it would.  Comments on the Taco 007 not doing the job jive with everyone, so I chose to up the pump to the 26-99. 

    Some more info on performance... I was in the shop today for 4 hours with the room temp thermostat set at 55 F.  Outside right around freezing all day.  When I got there the room temp was 55.7 degrees and the system did not come on once during that time.

  • Chris Chris @ 10:27 PM
    Contact this user

    Brads Post

    I based it on Brads post. I did also look up the unit. The manual specifically states unit can only make 6.6gpm. It slso goes on to say pump should be sized to move no more then 4gpm through the hx. At 4gpm the head in the tankless is 20'. Your getting all you can out of the 26-99 and just about all the tankless can give you. As long as you happy with the temps your good to go.
    "The bitter taste of a poor installation remains much longer than the sweet taste of the lowest price."
  • CarGuy CarGuy @ 10:45 PM
    Contact this user

    Very happy...

    Thanks for the extra info/details Chris.  Yes I am very happy with the temps and the performance. 

    Also, I just checked, and the Isolation Valve Kit that I bought to go with the Takagi have the pressure relief valves built in.

  • woodworker88 woodworker88 @ 12:07 AM
    Contact this user

    To much temp drop

    I also had a company that sold me a system but was no help when it didnt work right.  Here is a little background on the system and layout.  I have 2 inch foam on the perimeter and the floor, I ran 7/8 pex cause thats what they told me to use(bad idea), I have 3 - 400 foot loops on a 24 by 40 floor.   I know you are supposed to to not run water more than 120 in a concrete floor but I turned it up to 140 cause I was only getting 80 degrees on my return with 1.6 gpm so I did some modifications to the piping and now am getting 90 on the return.  Oh yeah forgot to mention I have a 15-58 pump and a Tagaki Jr once again cause that is what they told me to get.  I would like to drop my outgoing temp to 120 and have a temp drop of 15 or so, Do you think I need a bigger pump or i seen someone talking about a pump for the heater and one for the tubes, how is that done?
  • NRT_Rob NRT_Rob @ 9:55 AM
    Contact this user

    I don't deny

    that your setup is suboptimal. but why do you care if you have a 15 deg dt? in a shop, even 30 is fine.

    140 is the max temp for concrete.
  • woodworker88 woodworker88 @ 11:13 PM
    Contact this user


    I guess I don't care what my drop is but with 140 degree water going out and all I get coming back in is 85 or 90, I just want my return water to be higher so it helps heat the floor better.  And this is my house not shop
  • NRT_Rob NRT_Rob @ 8:58 AM
    Contact this user

    ah, gotcha.

    you are probably going to want to go primary/secondary then.
  • woodworker88 woodworker88 @ 5:12 PM
    Contact this user

    primary secondary help

    Any help with how that gets piped in to the system
  • Tim Tim @ 10:34 AM
    Contact this user

    Invaluable Books

    I sure got a bunch of help from this site when redesigning my system, but I also think
    Dan's books are worth their weight in gold
    Click "SHOP" in the title bar,

    It sure helps to understand how to set up correctly, and why the forces work like they do
Post a Reply to this Thread