Long horizontal runs from wells to GSHP

Discussion in 'Vertical and Horizontal Loops' started by Air Locked, Sep 28, 2018.

  1. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Absolutely, good insulation is recommended to protect the coil from the outside temps.

    However, you continue to be fixated on the idea that you will loose or diminish your delta T in the horizontal trench if the supply and return pipes are not insulated. You have a lot of Water going through the entering water line, which usually drops down to 32F, and is mostly surrounded by soil warmer than 32F. Trust me, the impact of a 5F colder pipe next to it is minimal. The surrounding soil will win. Your loop design and pressure drop is much more significant for overall system efficiency. You must pay attention to that one.

    1" pipes in boreholes raise a flag in general, so does 300ft one way for the header pipe, especially if it is only 1.25". The header pipe itself has over 30ft of pressure drop if you are flowing 15 gpm through it. Tell us about your exact configuration, tonnage, how many heat pumps, pumping strategy and antifreeze.

    A diagram would help.
  2. Air Locked

    Air Locked Member

    I'm not the geo designer, but what I know is this. The system in question is a single 5 ton [Sigh, how I wish HVAC people would move to KWH....] unit. There are 5-6 [I'm not exactly sure how many on each system.] paralleled wells. The manifold image:
    At 15 gmp, the Engineers Toolbox [.../pe-pipe-pressure-loss-d_619.html] seems to say 1.63 psi drop/100ft or 9.78 psi total. Is that excessive? The above-ground section will shorten the distance maybe 30 ft, so say 260 ft loop.

    On the horizontal length, waterpirate above said 3 ft of horizontal separation. If we can't accomplish that, and especially if we must stack vertically, what's the best practice?

    To be explict: my concern is about isolation from each other, not insulating from the surrounding earth.
    Last edited: Oct 9, 2018
  3. mtrentw

    mtrentw Active Member Industry Professional Forum Leader

    Please forgive my artwork. Not sure if I got this completely right, but what I understand you saying was installer option is on the left. Your proposed styrofoam add would be something like on the right. Not sure what your standard code specified frost depth might be in your neck of the woods. In most scenarios, a geo system can locally increase frost depth as some additional heat is removed locally (assuming it is running at margins below 32 degrees. Saturated soil can build up ice around geo pipes as well. If you add styrofoam (as on the right) you are potentially creating an insulated soil freezer above the styrofoam as heat has a much more difficult time getting to the upper pipe and area above. This is the same concept used when designing frost protected shallow foundations. It could potentially exacerbate any frost heave problems. Instead of a cold pipe connected to deep earth with heat flow opportunities to prevent ice buildup, the area may behave like a much shallower trench.

    As stated by others, the pipes (both inbound and outbound) will pick up some marginal heat, whether that overcomes slight losses from cross flow of heat in transit pipes is a theoretical gnat's ass problem for heavy modeling software and IMO isn;t worth any potential hassles or pitfalls.
  4. Air Locked

    Air Locked Member

    You draw better than I do. But yes, that's a good representation, except for dimensions.

    I was concerned about having a shallower trench and thus less spacing between inner and outer runs. I was curious about using the foam to compensate for that reduced distance.

    The customer was not keen on two trenches both because of the costs and the landscape.

    Things are looking up. They were going to use a 24" wide small excavator; I found a trencher with a 6" x 48" bar available semi-locally. The contractor rented it for a week but told me last night it's going so fast they may be done in 2.5 days, and that's getting down 4o"+. Plus, with its narrower width and less damage, they are making two trenches everywhere but part of the #2 manifold<->Well#13 run.

    I too would want larger pipe on the common legs but that's not my call, and the client is overseas for two weeks. (The best time to do all the ugly trenching.....)

    And the loop will have antifreeze good for -10C.

    Attached Files:

  5. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Here is my run with the drawing. Frost line in Seattle is likely less than 12", thus the ground by definition is warmer than 32F around the 28F pipe, likely 35F or more. Now if you take out some heat you chill down the ground a bit more, but it will be replenished by heat traveling horizontally and vertically through the ground. But in general the surrounding ground will transport more heat to the upper going out than it can "steal" from the warmer pipe coming from the loop field below. Plus the ground 2' lower will likely much warmer than the ground above. Installing an insulator (styrofoam) will likely be more counterproductive since it hinders the heat supply to the upper pipe from the ground below.

    Conceptionally, the ground will contribute marginal heat to both lines of the pipe, if you bring them closer together, you might reduce this even further, but you do not reduce the temperature in the "warmer" pipe much, since the surrounding warmer ground contributes much more heat than the "colder" pipe takes away.

    Attached Files:

  6. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Pressure drop with warmer water might be 9.78 psi total, but with colder water (32F) it will be much more, and will antifreeze it will be much, much more.

    Thus in 600 ft 1.25" pipe pressure drop with 50F water is 24.6 ft/hd (as engineering toolbox tells you, but with 32F water 26.5 ft/hd, with 15F protection with Methanol 31.3 ft/hd, and with 15F protection Glycol 38.1 ft/hd. Sorry for not being metric here, I share your pain. What antifreeze do you use?

    With Glycol, for the header line only, yep, that is excessive. Add to that the heat pump itself, the piping in the house and pressure drop of the loop field, and you end up with over 67ft/hd system pressure drop at 15 gpm. Lots of wasted pumping power. If you now have your 15 gpm spread out between (6) 1" pipes in the borehole, your Reynolds number is less than 1000, significantly reducing the heat exchange. Your water might not coming back with 5F delta T.

    I am trying to tell you that the pipe separation is your least issue, actually a non-issue. You said you are aware of the piping loss issues, when I brought it up.

    If this is my system, I would run 2" in the header trench, ensure that I use Methanol instead of Glycol (which you might already do), reduce the number of boreholes to 3, make them a bit deeper to put the same length of pipe in, and increase the pipe diameter in the hole to 1.25" or even better to 1.5".
    This would cut your entire pressure drop in the system from close to 70ft/hd down to less than 35 ft/hd, cut your pumping power in half for the life of the system, and give you a Reynolds number of 2800.

    But not a single time I would worry about pipe separation in the header trench, because both pipes would pick up heat from the ground.

    How deep are your wells, and what grout are you using? Any idea about grout and ground conductivity?
  7. Air Locked

    Air Locked Member

    The wells are all 100 ft. The explanation I got is the manifolds are limited to 1.5" (I was told WF didn't make larger manifolds....)

    So I ..encouraged... them to use 1.5 for the longer 5 ton unit. On the 8 ton/8 well system [200 ft away], are there 2"-1.5" HDPE reducers off the shelf? Don't know if there's an available socket-type fusion machine for 2" but would hope butt splicing would be possible.

    [A few years ago I watched as a crew spliced multiple 10" sewer line sections together almost like a fiber optic splicer...then burst-pulled them into WPA-era clay sewers; .]

    I know zip about fluid dynamics, and doubt I can do a Fourier transform on water....

    Doesn't methanol require bronze/stainless pumps? [If legal in Seattle..]
  8. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    My BS meter would be on full alert.

    Just because the heat pump manufacture (WF) does not make them larger, that should not mean we don't put in larger manifolds. Just build your own. The manifold gets the full flow of the system, in your case 15 gpm, the same as the supply and return lines, which is why I put in 2" as a standard, so I keep the pressure drop low in the pipes which see the whole system flow. Sure there are 2" to 1.5" reducers for socket fusion. Even if the WF manifold is only 1.5", the biggest pressure drop is in the 300ft one way supplying return line.
    Attached is a picture of (8) 2" manifolds, with (3) 1.5" lines coming from the wells, all socket fusion.

    Right now you have way to much pressure drop in the supply and return line, and not enough flow velocity in the wells, to get decent turbulent flow. This is all not gonna work well. You will pay much for pumping power, and you will not get much heat exchange due to the low Reynolds number. Any side restriction why you could not go lower than 100ft with the wells?

    You can use cast iron pumps with methanol. Are you using glycol? What is the size of the supply / return line in the 300ft trench.

    Attached Files:

  9. Air Locked

    Air Locked Member

    Reason beasides redrilling the wells being slightly costly? [And taking months to schedule the rig and crew?]

    The shorter lines would be 1.5" as well; unless we used the aformentioned 2->1.5 reducer.
  10. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Not sure if I can follow you here.

    What are the shorter lines, and where do you need the reducer?

    Plus you are not answering questions......but you are the boots (and eyes) on the ground.

    Nobody said re-drilling the wells. Going from Glycol to methanol would change the Reynolds number in the wells from 937 to 2058. Big difference for turbulent flow.

    Again, what is the size for the supply and return pipe in the 300ft trench....?

    A whole system layout would be helpful, we have to pull every piece of info out of you......
  11. Air Locked

    Air Locked Member

    I'm not on the ground {yet}, I'm on the other coast. [Tonight, however....]
    The wells are in place, the horizontal piping is not.
    They are 100 ft, with 1" pipe to their manifolds.
    The 5 ton/5 well system will have 600 ft of {now} 1.5" pipe.
    The 8 ton/8 well system will have ~300 ft of 1.5" pipe. Unless we can install a butt-splice reducer for 2"
    I don't yet know what the antifreeze will be, or what is legal in WA state. {Do you?}
  12. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    I don't think using Methanol in WA is an issue. A quick search showed no regulation in regards to the antifreeze used.
    Use Methanol in the pipes for lower pressure drop and higher Reynolds number. What is the minimum temperature your loop field is designed for. What heat pumps will be used? Yes, you can reduce from 2" to 1.5" for the manifold, the important part are the long supply and return lines for both the 8 and 5 ton system.

    Now, why wouldn't you combine both loop fields?
    Last edited: Oct 14, 2018
  13. Air Locked

    Air Locked Member

    So I talked to the designer/installer. It took some ..discussion... but we are going to 2" on the 8-ton/8 well/175 ft distance loop. The 2" pipe comes in 100 ft lengths, he'll use couplers; he has a socket fusion splicer, not a butt splicer. It will be Methanol.

    BTW, we're having to reexamine the route decision. Tunneling under the building had several dealbreakers; it would have two almost 90° bends & I could not see being able to pull replacement pipe though the casing. Plus the next stage is even worse. So we've cut the slab atop the previously mentioned buttresses/footer and will see how far that goes down. It's less than a foot wide/horizontally, so we can core-drill it if we can't get under.

    I don't recall why the 2 separate loops but that was discussed.
  14. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    We buy 2" pipe in 100,200 and 300ft length. Socket fusion is fine.
    Why not also 2" pipe on the longer run for the 300ft supply return lines?
    What heat pumps do you use. What model and how many?

    What is going to be your pumping solution/design? How do you route the water from the pumping/flow center to the heat pumps? Pipe size and length?

    I am asking those questions since the installer does not seems to be very experience with pressure drop, pumping power and larger system design.
  15. Air Locked

    Air Locked Member

    The first because that was what he could get quickly. The second because that line is already in place.
    We have time deadlines we must meet; getting the land backfilled before SWMBO returns.

    There are 2 WF units; that's all I know to date. The well loops and the output loops will be in the new mechanical room for the 8, and adjacent to the 5 under the house.

    There are large commercial installations around here, and resi ones; but there are only two installers doing residential. And from what I was told, the few drillers first work for their chosen installers [vs. a 18 month lead time.] The installer on the job was the one recommended by a majority of customers.
    Last edited: Oct 14, 2018
  16. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Sounds a bit like "design by inventory".

    Sound like a 8 ton water-water.

    Pretty steep pressure drop. If you you assume Methanol as antifreeze and you have, with 24 gpm nominally, 22.9 ft/hd pressure drop, 1.7 ft/hd for the boreholes (now with Methanol your loop Reynolds # is around 2800) and 21.2 ft/hd for the 175 ft 1.5" pipe each way. Add to that about 17ft/hd for the heat pumps, and about 9.5 ft/hd for about 50ft supply and return line inside the house. Now you are at 50ft/hd system pressure drop. How do you plan to pump it?

    What make this job different from the typical residential job is the 175ft and 300ft supply and return line, which sees the entire flow of the whole system, and the spread of the flow trough (5) and (8) 1" lines in the wells.

    Not understanding the impact on turbulent flow and pressure drop, especially with glycol, and the fact that you had to have a "discussion" (and that the 175ft line is already in) makes me wonder about the expertise of the installer, no matter how many customers recommend them. Sorry to be that frank.... I have seen too many bad geo jobs, and kind of have a feeling what their lack of knowledge is, and why geo systems lack performance. Next on the list is the pumping solution.....

    The next question is, assuming both are heat pumps for radiant (at least the 8 ton seems to be), what is the radiant system like, and how do you control the supply temperature.
  17. Air Locked

    Air Locked Member

    The 2"/100ft was on the spot. I pushed for 2" and he pulled out his phone and started looking at suppliers.
    (Then he looked for a 2" socket for his fusion splicer.)

    I'll have to ask for more details on the pumping.

    The 5 ton is to an air handler. The 8 ton is for radiant floors and a swim-spa. (I think, I'll check on that.)
  18. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    If he is not used to 2" pipe even for residential application, he does not understand pressure drop and pumping power.

    Is the 5 ton a split, or a water-water with hydronic air handler?
    Is the radiant throughout the house? Swim-spa all year long?

Share This Page