Welcome! Here are the website rules, as well as some tips for using this forum.
Need to contact us? Visit https://heatinghelp.com/contact-us/.
Click here to Find a Contractor in your area.

rust in water using tankless hw heater in radiant system

joeinmo
joeinmo Member Posts: 22
I installed a closed radiant floor heating system using a takagi hw heater. It works pretty well when I get all the air out. But there is a tiny filter on the heater that often clogs reducing gpm flow from 4.5 gpm down to 2.2. I have 6 zones and 132 gallons of water in total. I know the takagi is not made for radiant heating, but that's what I've got at the present. Any suggestions on how to eliminate the rust and rust particles in the water in order to keep the tiny filter on the takagi from clogging? Thanks. Joe

Comments

  • Larry Weingarten
    Larry Weingarten Member Posts: 3,273
    Hi, How about adding a wye strainer in line ahead of the heater?

    Yours, Larry
  • joeinmo
    joeinmo Member Posts: 22
    Hi Larry. I already have a wye strainer, but it strains out the larger debris, not the fine rust particles clogging the Takagi very fine filter. I assume these fine particles are coming from the ferrous metal in the Grundfos pumps, but I don't know. Last year, I removed the tiny filter all together, but the rust in the water caused the propellor on the flow sensor on the Takagi to get stuck and when the propellor doesn't spin, the heater will not fire up. I had to replace it. I have no air in the lines, so I can't understand where this rust is coming from. All pipes are copper except the feed and return lines to the manifolds which are 7/8" pex. There is 1/2" pex in my floors and the manifolds are copper. Thanks. Joe
  • Ironman
    Ironman Member Posts: 7,367
    Did you use O2 barrier pex?
    Bob Boan
    You can choose to do what you want, but you cannot choose the consequences.
  • Jamie Hall
    Jamie Hall Member Posts: 23,162
    You may also need a magnetic particle trap.
    Br. Jamie, osb
    Building superintendent/caretaker, 7200 sq. ft. historic house museum with dependencies in New England
    Ironman
  • hot_rod
    hot_rod Member Posts: 22,022
    Steel expansion tank?
    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream
    Ironman
  • GroundUp
    GroundUp Member Posts: 1,889
    132 gallons? Are you sure? This sounds like a good candidate for something like a Rusco spin down filter
    Ironman
  • joeinmo
    joeinmo Member Posts: 22
    Thanks all for your replies. I installed this system myself and am not a professional radiant heat guy. But this is my second system, one in my shop and now this in the house, so I get the basics. One thing I learned is that for a large system, you should probably go with a boiler. Having said that, first, I do not have O2 barrier pex, unfortunately. I wish I did, but this is where I'm at. Yes, my expansion tank (Watts) is, in fact, steel. I never considered that hot_rod Bob, and so that could be the source of the problem. But I still need to fix it. On the capacity, yes, I'm sure I have right around 132 gallons in total. Last night it got down to 8 degrees outdoors and the heater has run constantly. It can't keep up: inlet water at 85 degrees and outlet around 118 and room temperature around 66. A magnetic particle trap sounds intriguing. My primary loop lines are 1 1/2 inch, and the webstone magnetic particle trap has 1" lines. I don't think reducing down is a good idea. I'm not familiar with a Rusco spin down filter either. Can you tell me how it works, GroundUp? After running all night, my gpm flow on the Takagi has gone down from 4.6 to 3.1, so I know the filter is getting clogged again. Thanks. Joe
  • joeinmo
    joeinmo Member Posts: 22

    Here's a picture of the system. Joe
  • Alan (California Radiant) Forbes
    Alan (California Radiant) Forbes Member Posts: 3,979
    edited December 2020
    Oxygen entry through the non-barrier tubing will be an on-going problem. If you want to keep the on-demand, install a heat exchanger to separate the oxygenated water from the Takagi, but oxygen will continue to attack the cast iron pumps. You can also change all the pumps and x-tank to non-ferrous.

    The elephant in the room is the non-barrier PEX. If money grew on trees, I would change the boiler, pumps and x-tank to non-ferrous.

    BTW, what model Grundfos pump is on the primary loop at what speed setting?
    8.33 lbs./gal. x 60 min./hr. x 20°ΔT = 10,000 BTU's/hour

    Two btu per sq ft for degree difference for a slab
    Ironman
  • duffy_4
    duffy_4 Member Posts: 79
    Few questions ,do your oh system have a fill valve. At 132 gallons is your expansion tank size correctly. Are you pumping towards your expansion tank? The Adey magnaclean units are excellent at capturing ferrous materials,I would think your total flow rate should be more than 4.5 gpm also with clean filter.
  • Ironman
    Ironman Member Posts: 7,367
    I'm pretty sure you're piping arrangement is all messed up. Can you explain the path of water travel please? I don't think your using your primary loop correctly. That is if I'm seeing this right. 

    The piping looks correct. It's p/s moose antler. The non-barrier tubing with the use of ferrous components is the problem. A 1" pipe will carry 8 gpm at 4 fps. So, a 1" dirt filter will not restrict flow. There are also treatment products that will help reduce the effects of O2 in the system.
    Bob Boan
    You can choose to do what you want, but you cannot choose the consequences.
  • Ironman
    Ironman Member Posts: 7,367
    edited December 2020
    It’s not a matter of efficiency. The tankless has an extremely high resistance to flow. The purpose of p/s is to allow different flow rates to mix between the closely spaced Tees without one flow rate effecting  the other.

    In this case, the tankless needs an extremely large circulator to overcome its resistance to flow while the house loops can use a smaller sized circulator.
    Bob Boan
    You can choose to do what you want, but you cannot choose the consequences.
  • Ironman
    Ironman Member Posts: 7,367

    The expansion tank could be better located to the bottom of the air separator so the boiler circulator isn’t pumping towards it.
    Bob Boan
    You can choose to do what you want, but you cannot choose the consequences.
  • joeinmo
    joeinmo Member Posts: 22
    Youngplumber, the primary pump is on the (left) return side of the closely spaced Ts. That's where the wye strainer is. The temperature gauge is on the outgoing water.

    Alan, I think long-term I will end up putting in a boiler; short term, I may just keep cleaning the tiny filter. But are you saying that if I do put in a boiler, I still have to put in pumps and x-tank with nonferrous metal? And, another question, would stainless steel pumps do the trick even though they are steel? BTW, the primary pump you see is a Grundfos UPS26-99FC.


  • Canucker
    Canucker Member Posts: 722
    I guess what I'm saying is. The common piping between the primary tees should not exist and the supply and return headers should be the loop. That's how see them, but hey @Ironman your the man, not I. 
    If he did that the boiler probably wouldn't fire until multiple zones were calling and the flow through the boiler would probably be too low unless they were all calling at the same time. 
    You can have it good, fast or cheap. Pick two
  • Ironman
    Ironman Member Posts: 7,367
    If I understand you correctly, I think you’re referring to p/s series. That’s not the correct method if you need the same SWT at each secondary branch since the return from each branch lowers the SWT to the next branch. 
    It looks like this:





    Bob Boan
    You can choose to do what you want, but you cannot choose the consequences.
  • Ironman
    Ironman Member Posts: 7,367
    edited December 2020
    This is what the OP has and it’s the correct method for his application:


    Bob Boan
    You can choose to do what you want, but you cannot choose the consequences.
    Canucker
  • Ironman
    Ironman Member Posts: 7,367
    What you described above is simple direct return piping. That won’t work with the scenario here which the OP posted because of the high head of the tankless and then the resistance of the house loops being added to that.

    I would highly suggest that you get Dan’s book “Primary Secondary Piping” from the book store on sight since the concept appears to be foreign to you.
    Bob Boan
    You can choose to do what you want, but you cannot choose the consequences.
  • Ironman
    Ironman Member Posts: 7,367
    edited December 2020
    That’s p/s with a crossover bridge. It’s used when balancing each branch is critical. It’s the most complex and expensive form of p/s and usually not necessary.

    I think you’re missing the point that the piping method that is chosen should be the one that’s most appropriate for the system it’s being incorporated in. Since the heat exchanger of the tankless has such high head, the proper piping method is the one which adds the least PIPING resistance to that circuit and its circulator. That’s the one that’s been used here.

    There’s nothing magical about the diagram you posted that would necessarily make a system that incorporated it any more efficient than another form of p/s. In fact, in this case it would actually be the wrong choice because of the additional resistance the additional components would add to the primary circulator.
    Bob Boan
    You can choose to do what you want, but you cannot choose the consequences.
    Canucker
  • Ironman
    Ironman Member Posts: 7,367
    I'm not cross, just trying to get you to see there's more than one way.
    Bob Boan
    You can choose to do what you want, but you cannot choose the consequences.
  • joeinmo
    joeinmo Member Posts: 22
    That was an interesting discussion on p/s loops, and with respect, I am not concerned about the p/s piping. It works just fine. The problem is the rust and the clogged tiny filter.

    I cleaned the tiny filter on the tankless and, just to make sure, put a magnet to the debris. It was, in fact, ferrous metal flakes. From what I've been reading from all posts, the source of the rust could be the expansion tank or pumps. Regardless, I'll always have O2 in the water because of the non-barrier pex. So, that's where I'm at. After cleaning the filter, I'm running at 4.8 gpm, which is close to capacity on this heater. Outgoing water is 117 and retuning at 108. So, for now, everything is fine. But it only got down to 25 degrees last night. But the system performed as it should. This issue of rust will be ongoing.

    On the issue of a fill valve; I have to fill manually if I need water.

    On relocating the ex-tank to beneath the air separator, at present I don't have the room since there's only about 8"-10" between the pipes of the secondary loop. And I am not certain if the ex-tank is sized correctly. I will look into that.

    I am also focused on a magnetic particle trap. Where would I place that? I assume it would go in the primary loop and I assume it's preferable to set it vertically. As I said, it comes in 1" and not 1 1/2", so I would have to reduce down.

    Finally, long term, I may need to put in a boiler. If I do, that's when I can consider pumps and ex-tank without ferrous metal. We'll see. For now, I may just be stuck with cleaning the filter every week.



  • PC7060
    PC7060 Member Posts: 1,155
    edited December 2020
    I’d put it right after the steel expansion tank since that appears to be the source.  Alternatively, right in front of the boiler. 
    Caleffi DirtMag is a good product and can be flushed with live system. You can get in a variety of sizes and connections. 
    I placed mine right at the return since that’s where the majority of the particle originate in my old system.   
    I’d also get a stainless tank and pumps in ASAP.  Better to get the contaminate situation under control sooner than later  
  • Ironman
    Ironman Member Posts: 7,367
    You don't have to mount the expansion tank directly below the air separator, just CONNECT it there. That way, the boiler pump will be "pumping away" from the Point Of No Pressure Change.
    Bob Boan
    You can choose to do what you want, but you cannot choose the consequences.
  • joeinmo
    joeinmo Member Posts: 22
    Thanks, Ironman Bob, I thought about that possibility, but I wasn't sure if I could have that much separation between the tank and the connection. So I appreciate you confirming that's what I should do. I think I'm going to buy an expansion tank that has a plastic lining to protect from corrosion. Amtrol RX has a 10.3 gallon one that I can mount on the floor and run copper tubing up to the air separator. My current 4.5 Watts is steel and probably too small, so I'll size up.

    I'll hold off on buying that expensive Caleffi DirtMag that PC7060 recommended above, although I think that would help too. It may become necessary in the future, so I appreciate that recommendation. As has been pointed out above, my non-barrier pex will allow my pumps to start rusting sooner or later, so I will eventually have to put in stainless steel pumps. But for now, I'll replace the expansion tank and see how far that takes me. Finally, I may some day put in a boiler, but only when the tankless fails.
  • PC7060
    PC7060 Member Posts: 1,155
    edited December 2020
    Certainly understand financial concerns in these times but get the DirtMag when you’ve got the $. It will prolong the life of your boiler. 

    Be sure to  flush your system a few times after you get new tank. 
  • joeinmo
    joeinmo Member Posts: 22
    Thank you all very much. This has been an incredibly helpful website and I've learned a lot. I may well go ahead and spring for the DirtMag since PC7060 makes a very good argument. I can place it in the primary loop where I will soon remove the expansion tank. Then, down the road, I will have to get the stainless steel pumps and I will no longer have to worry about the oxygen in the water.
    PC7060