Click here to Find a Contractor in your area.
Welcome! Here are some tips for using this forum.
Need to contact us? Visit

How to bleed in-floor loops?

jchjch Posts: 29Member
Converting the electric-heated part of my house over to hydronic. Looking at installing underfloor heat plates with 1/2" Al-PEX 8" OC fed from a manifold with actuators. Have worked my way through the entire Siggy book and have done heat loss calcs for all rooms.

One of the smallest loops would have a tube length (including leader) of 186 ft, flow rate 0.6gpm at 26F, supply temp 117F for 10F delta-T. Heat output 2,000 btu/h, matching heat calc.

A bathroom would be even smaller: 100 ft of tube in-floor, 0.2gpm (0.32 fps), 1,000 btu/h.

Lochinvar WHN-055 with Grundfos Alpha circulator directly feeding a manifold. Whole-house heat loss of 22,000 btu/h.

At 0.6gpm, this 1/2" PEX loop would have a water velocity of only 1fps, making me think that it would be prone to air locking (not in the range of 2-4 fps).

Given that the loops would be above the level of the boiler and manifold, how do people incorporate an air-bleed mechanism for such an under-floor loop?? The water velocity is too low to entrain air bubbles and bring them back to a central Air Separator and the floor is in the way of putting a bleed valve at the high point.



  • ZmanZman Posts: 4,212Member

    Your 1/2" tubes should not have trouble as long as you pipe the boiler room correctly. Be sure you are pumping away from your expansion tank and use a microbubble separator.

    For the initial purge it helps if you can isolate and purge each zone separately using a purge pump or the high pressure from the fill valve.

    "If you can't explain it simply, you don't understand it well enough"
    Albert Einstein
  • jchjch Posts: 29Member
    edited January 2014
    What about at season start-up?

    Makes sense that if you blast water into each circuit one at a time during initial fill that you should be able to avoid big air bubbles.

    My question was more thinking about what happens after a long summer sitting idle. During that time, wouldn't air bubbles gather at the high points of the system? If so, then how do you then get them out once heating season starts again if the water in those in-floor circuits is only flowing at 0.3 fps??
  • GordyGordy Posts: 7,989Member

    Should be between 2-4fps some of the reason is to keep air entrained . At .3 FPS it would take 17 min. For water to travel a 300' loop.

    To answer your question once bleed it should not be necessary every season with proper air removal devices in place.
  • GordyGordy Posts: 7,989Member
    edited January 2014

    Your post I would increase your flow rate a bit it's not going to hurt anything.
  • jchjch Posts: 29Member
    How would I increase the flow rate... the small single-zone bathroom (1,000 btu/h)? It's only 40 square feet and would have in-floor loops. If I increase the flow rate to 2fps in that room then my delta T would be only a couple of degrees.

    When hooked up to a boiler that has a minimum firing rate of 10,000 btu/h, wouldn't the boiler quickly overheat? Or short-cycle like crazy?
  • GordyGordy Posts: 7,989Member
    1000 btu zone

    I certainly would not have a room of 1000 btus on its own zone
  • delta Tdelta T Posts: 616Member
    I agree with Gordy

    That is too small to be on one zone. As far as air goes, as others said, once you bleed it at start up, and if you have a good air eliminator installed, after the system has run for a while, it will be free of any air that could cause an air lock. This can be easy or hard depending on how the boiler is piped. After startup There may well be some very small bubbles stuck somewhere but these too will eventually either oxidize some system component, or will eventually be removed by your air eliminator. Sitting and not being used during the summer will not let more air in (unless you have a leak somewhere in which case fresh water and fresh air are constantly being added)
  • GordyGordy Posts: 7,989Member
    edited January 2014
    Delta T worries

    I should add don't get to wrapped up in the delta T. It's a number to use for design, and can be used to trouble shoot. I'm getting the impression you think it has to be 10 or it some how won't work. Rarely in real world once the system is started and equalized will deltas be what was designed for unless you force it through your flow rate.

    We really know very little about your whole system. My advice is really based on if the circulators curve allows you to up the flow in a low flow loop. You have flow control on your manifolds I assume.

    A narrower delta will just give a more even temp across the floor. Keep your velocity in that 2-4 FPS which helps carry air back to the separator.

    I hope you used o2 barrier pex. If not you will have continuing air problems amount others.
  • jchjch Posts: 29Member
    Still... the design stage.

    The proposed 1,000 btu/h zone will be a small bathroom (60 sq ft) with a tile floor. I'm trying to avoid using electric in-floor heat for it and am reluctant to combine it with another room because of its unique heating requirements: floor kept at 80F+, year-round use, etc.

    How do people usually handle such a situation??
  • GordyGordy Posts: 7,989Member
    edited January 2014

    In the summer? Up the cooling.

    a neutral floor in summer is still nice 75-77. Personally I have tiled baths, and never thought it to be uncomfortable in the summer.

    I think for such a small load, and wanting the floor warm in non heating season dare I say electric radiant makes sense. Or due both then for the bulk of the heating season you would be on a cheaper fuel to heat the space. Don't know your utility rates but seems silly to fire any boiler to heat such a small load.
  • jchjch Posts: 29Member
    Summer cooling... is not an issue. It rarely gets above 75F so unheated tile floors always feel cold.

    What if I used a buffer tank to make DHW and fed the manifold off that? i.e. Use the tank as a giant hydraulic separator that's kept at 120F?
  • REKBDRREKBDR Posts: 22Member
    I have virtually the same issue as to how to successfully bleed a 168 square foot Warmboard floor where all the controls will be below the level of the tubing. In my case I am tapping into a conventional high temp system, so I will have a mixing valve and essentially a nearly closed loop where most of the water in the tubing never gets out (so as to pass through the air separator down near the boiler). I have to count on the separator(s) in my low temperature loop to purge the air during operation.

    The problem is the discrepancy between the recommended 2 fps (equals 1.2 gpm for 1/2” pex al pex) and the lower flow rate that otherwise might be found in such a set up (Warmboard has suggested .5 gpm). I could do a flow of 1.2 gpm since the flow meters in some manifolds go as high as 2.0 gpm, but that would give me a design Delta T of only about 5 degrees. Is that feasible?
  • ZmanZman Posts: 4,212Member
    Almost every system I have seen that has 1/2" tubing runs between .5 and 1 gpm flow rate. Air lock problems are rare after initial purge. I think this is partly due to the smooth flow path that pex allows. In general I like the 2-4 fps rule. I don't think it need apply to radiant pex loops.
    "If you can't explain it simply, you don't understand it well enough"
    Albert Einstein
Sign In or Register to comment.


It looks like you're new here. If you want to get involved, click one of these buttons!