Horizontal Loop Field - Advice Requested - Southeastern MA

I'm running 4 700' horizontal loops in NE OH. (You can see my thread here https://www.geoexchange.org/forum/threads/starting-diy-horizontal-closed-loop-in-ne-ohio.7552/)
They are 5' deep. Deep earth temp here is about 53*. I have them hooked to a 3 T Q Energy HP. My entering water temp bottomed out at 40* (we do supplement with wood on the colder days.). I'm running 10 gpm. I would lean more towards the 2 gpm per loop and add more loops. Your pumping power goes down, your loops can be longer with less pumping penalty, and your potential heat stored in the loops when HP is off is higher. My unit can run for nearly 10 min before turbulent flow starts to matter due to the oversize field. Then the fluid takes longer to go thru the ground so has more time to pick up heat. All in all, the size of the field is far more important than turbulent flow (at least with 3/4” pipe).

 

I would agree with Eric. I would suggest looking at configurations using some combination of 600' or 700' loops. I believe the pipe comes in 600' rolls. The loops could be in individual 1' wide trenches, two loops in a 2' wide trench, or three loops in 3' wide trenches.

I have six 700' horizontal loops for a 4T WF7 in central Virginia. They are configured in two 350' x 3' trenches, each with three loops. My loops are probably oversized, by as much as 75-100%. The nominal maximum flow for the WF7 is 13.5 GPM, so the maximum flow would be 2.25 GPM per loop. It's a variable speed unit, so it runs on stage 1 or 2 out of 12 a lot of the time, so the flow per loop is under 1 GPM a lot of the time. Turbulent flow is not as critical when the loop is oversized, which would include when a staged heat pump is running on low.

 

Thanks for all the input, guys.

So as of right now, I'm leaning toward installing 10 loops. One option is the 36" Slinky at 11' C-C, which at 700' each which would fit nicely in the 140'x100' area I have sketched for the field. However, I'll discuss the excavation with the operator, and if it makes the most sense to just excavate the entire rectangular area, then perhaps I should then use pipe instead of slinky and go out and back with each pipe 150' twice to create ~600' loops, evenly spacing each pipe 2.5' apart, which would result in 10 600' loops in a 150'x100' area.

10 loops seems like a pretty good quantity to end up with reasonable flows within the range we expect to end up at for tons. At 7 tons, or 21 gpm, we're at 2.1 gpm/loop, and at 10 tons, or 30 gpm, we're at 3 gpm/loop.

So for all of this LoopLink experimentation I've been doing, I've been leaving my peak heating load at 80,000 Btu/hr. Should I be adjusting that up when I am entering the scenarios of bigger units and bigger system flows?

 

I don't think anyone is going to be able to answer your heating load question with more specificity given the very undetermined state of the house design. You could use CoolCalc (google it) to do a block load calculation on this fictional house to see how close the 80k BUT/hr is to something less fictional. I think if you model a two-story rectangular box that approximates your 4000 sq ft, oriented how you think the house will lay out on site, have walls and roof insulated values established, and throw a reasonable amount of windows on each elevation you'll get a pretty good approximation for heating and cooling loads (maybe +/- 10k BTU/hr accuracy). Outside design temps are based on location and inside design temps you can make what you want but 70 for winter heat and 75 for summer a/c are probably good values. Its all a guess so I don't think there is a wrong answer but probably a range of decent answers (much like your loop field).

 

I'm not actually asking if the 80,000 Btu/hr is correct or not, I know none of us can know that answer at this point.

I'm asking if, when I am experimenting with the different units and resulting system flows and EWTmin, should I also be adjusting that 80,000 Btu/hr accordingly? In the table that I posted with the different values I entered into LoopLink and the resulting EWTmin values, I left it at 80,000 for each. But I'm wondering if that was the correct approach, since I am varying in that table between 21, 27, 30 gpm system flows. As an example, for the 30 gpm system flow, since that kind of flow corresponds to something like 10 tons, should I be using something like 100,000 or 120,000 Btu/hr for that one instead of 80,000? Otherwise, is the EWTmin that I am getting higher than what it would actually be? Or is it sort of okay because the 80,000 is a peak load and the EWTmin is not really determined by the peak load but the average load?

 

80k BTU/hr is an instantaneous peak load that your heat pump equipment will have to handle. Its sole purpose in this exercise was to come up with an approximate equipment tonnage. The equipment tonnage dictates the system's total flow. Your loop field doesn't care what the instantaneous peak load is. Your loop field is influenced by how much heat you extract (or add) from the ground over a full season. A single day(s) heat extraction amounts to a small % of the overall year which is why this doesn't matter for the field in general. By using a range of system flow rates, you are in effect using a range of equipment tonnage which would suggest a range of instantaneous peak loads. So to answer your question, I don't think you need to play with the 80k number given the ranges you are considering in system tonnage and flow.

I guess my answer was more along the lines of getting a sense of if the 80k number is remotely accurate. My gut says it isn't wrong and is probably high, but some modeling with coolcalc would help solidify that.

 

There's a reason you cannot find a place to vary system flow. It's determined by the heat pump selection and is not normally something you vary. The thing you normally vary is the loop configuration. I would approach this differently. Pick a house load and pick a set of one or more heat pumps that satisfy that load, and that will determine the required total GPM.

Once that's done, set the desired EWT to 30/90 MIN/MAX in LoopLink and let LoopLink tell you how many feet of pipe you need for each candidate ground loop configuration. The amount of pipe required to meet 30/90 EWT will be determined by the ground loop configuration (number of pipes per trench, slinky vs straight pipe, pipe depth, trench separation, etc.) and your location. You can create multiple configurations (GHEX) to compare.

The required total GPM will not change. What will change is the GPM per loop, based on how many loops you need. As mentioned already above, you need to find a loop length and number of loops that keep the GPM for a loop within a reasonable range.

I would use a house load that's an even number of tons, like 72,000 96,000. Once you find a loop configuration that works for you, you can scale it to match your eventual house load. In other words, if you find you need twelve loops for 8 tons, then six loops should be about right for four tons, or fifteen loops for ten tons.

 

Here's an example of a result I get from LoopLink:

Deep earth temp 53*
Peak heating load 96,000 Btu/hr
Heat pump selected Waterfurnace Envision NDW100 (ELT-080/50) - Heating high capacity is 97,400 Btu/hr
System flow rate is 30 gpm

Even with the sort of "maximum" parameters that I've been testing for the 36" Slinky:
11' C-C spacing
10' depth

It still takes 10 loops of 685' to hit the 30* EWTmin

This is what I keep seeing: I have to go to the 10' depth, 700' loops, and 1.25 loops/ton in order to hit the 30* target.

Granted, there's a good chance that 8 tons is high, but if I scale back the number of loops proportional to the number of tons, this remains the same. So while I'd prefer not to have to go to the 10' depth, I may have to? Although I do think at this depth it might make sense to excavate the entire area and then I can evenly space pipe rather than slinky which should enhance the performance to some extent. Maybe that means I can get away with 9' depth instead of 10' but it's not necessarily a game-changer.

If I change depth to 7.5', the loop length goes up to 800'
If I change depth to 6', the loop length goes up to 900'

 

The NDW is a water-to-water heat pump. Is that really what you want?

Try it with two Waterfurnace, variable speed, 7 Series, 4 ton units.

Soil type has a major effect on required loop length. Are you using reasonable soil values for the location?

Look at a GHEX configuration of Standard, 4-pipe rectangular, 2' trench width, 6' deep, 10' spacing, 2 flow paths per trench, five trenches. See how that works.

 

I'm not 100% sure, but I had chosen water-to-water mostly because I would like to use radiant heat, at least in some zones, and I had read somewhere that water-to-water was more efficient. I don't necessarily have my heart set on radiant, though.

Okay, now I am able to reduce depth to 7' for 10 loops of 700'

I am using thermal conductivity of 0.72 and thermal diffusivity of 0.59, which was from the worst (I was trying to be conservative) of the two Sandy-Loam soil types in the estimator. The soil in the area is a "fine sandy loam" according to soil maps, and "sandy loam" in the test pit data.

Results:
Total Pipe Length 6,728 ft
Pipe Length per Trench 1,346 ft
Individual Trench Length 336 ft


I really appreciate the help, BTW.

 

The soil values seem reasonable. That's what I used for mine, but when the trenches were dug, the soil was more like clay damp clay, which is better. Seems like the 4-pipe with 5 or 6 trenches gets you into reasonable territory with 600' to 700' loops.

 

I know you had said that you did not want to clear more than necessary for the loop field, but I would suggest you consider that a little more. The reason had to do with the heat sink size. I have 4 700' loops in an area that is 80' x 350' with only a 3T HP. Now I know it is way oversized, but you are talking of putting double to triple the load on an area approximately half that. I would suggest experimenting with upsizing the loop area and seeing how that calculates. I'm sure slinkies work fine but common sense says that not having the pipes touching is technically better.

Of course, I don't know the lay of the land and other considerations, so just my 2 cents.

 

OK, I think we might overthink this too much.

You are talking about a 4000 sqf house which has to comply with the new energy code.

You like are looking at 50 KBTU/H for heating and not much more than 20-25 kbtu/h for cooling. You want to heat your pool with an air source heat pump, much less complex, much more efficient, much cheaper to operate (unless you plan to operate the pool year around).

Radiant is not really that much more efficient, but that can be a later discussion. So is the question how you make your domestic hot water.

The issue at hand right now is that you have no significantly higher expenses to install an oversized loop field. So go for it. 8 circuit, 10 circuits, 15 circuits. 500, 600, 800 ft per circuit. Go and knock yourself out. You will always benefit from a larger loopfield.
Use methanol, not glycol, for better pumping efficiency and a bit more turbulence. But at that size, you don't need turbulence, you have always an over performing loopfield. So if you are putting in more than 5000 feet of pipe, you have enough. You can worry about anything else later....

 

^^^ Thanks. So let's just say, as a hypothetical, I installed 15 circuits. But then it turned out my load is only 50 KBtu/hr, and I install a Waterfurnace 7-Series NV048. LoopLink shows a 13.5 gpm system flow rate for that unit. So if I still hooked up all 15 circuits, and each is only getting 0.9 gpm, that is still better than hooking up less circuits, just for the sake of achieving a higher gpm/loop, right? I mean, take it even further to the extreme, and use an unrealistic example of 30 loops, and 0.45 gpm/loop, flow won't be turbulent but the huge amount of surface area more than makes up for that, right?

The main reason I was leaning toward radiant was more about comfort than efficiency, but I'd love to hear your thoughts on this, or if there are references you would point me to that I could read, I'm very interested.

Same with the domestic hot water. I'd love to hear your thoughts or read more about the different approaches.

 

Yes, the larger loop will more than make up for the lack of turbulence. The water has more contact time since it remains longer in the pipe.

Radiant done well, on top of the subfloor, or embedded in concrete etc, works well. No carpets on top!

then you can run the supply temp between 85-90F. Heat pump runs efficient and you have the greatest comfort. It will cost you, since you are investing in two distribution systems. My house is radiant, I am a big fan of it.

Domestic hot water: Over the years we have migrated into making all the hot water via the geo system, with a big tank being conditioned by a small dedicated 2 ton water-water, or by a larger heat pump which switches over from floor heating to DHW.
The desuperheater were simply not efficient enough, most of the time there was not enough desuperheat.

See attached screenshots

 

^^^ Great info, thanks.

If I'm planning on having a propane on-site for generator and other small uses, does that come into play at all as far as decision-making on hot water or any other items?

When you go with the small dedicated 2 ton water-water unit for hot water, do you dedicate a few separate loops to that?

 

No, we have it all on the same loop field. I would not pu in a propane water heater, if I can have a dedicated geo solution. The infrastructure is already there. Why would I take a knee on the 1 yard line if I can score a touch down....?

 

Any suggestions on specific pipe type to purchase? Considering that the pipe needs to be purged and fusion fittings must be used (right?), it seems the general recommendation is that I just bury the loops and keep them capped and ready to be tied into the header later? I know a couple people in the gas industry that do fusion regularly, is it worth getting them involved to complete the header or is that something my geo installer will probably prefer to do themselves anyway since they are going to purge? In any case, since fusion technique seems to be at least somewhat brand-specific, I want to make sure I use the right type of pipe.

 

oilcreekplastics dot com or charterplastics dot com are two suppliers the installer used at my house for various pieces of the ground loops. You want high density polyethylene pipe (HDPE) PE-4710.

Not sure if you need DR-9 or if DR-11 will suffice (these are pipe sizes using dimension ratio sizing). I have DR-9 (thicker walls than DR-11) but I also have 400' vertical wells so the pressure gets pretty high at the bottom when you have a 400' vertical column of water. You might be able to get away with DR-11 for horizontal runs but see what others say who have done this. I would call and ask them what standard lengths they make so you aren't paying for something "special". They make straight pipe with u bend ends for vertical runs, guessing you don't need that for horizontal runs and can just use straight pipe bent around at the end of the trench. My understanding is that you site make slinky loops with straight pipe (if you are still considering that).

Heat fusion is the preferred technology. not aware of any brand specificity - the HDPE DR pipe comes with constant outside diameters and fittings with constant inside diameters so anyone's pipe can be fused to anyone's fittings. there are versions of PE pipe for water supply with constant inside diameters where you can't use fusion and have to use stab fittings. Maybe you are thinking of pex and their connections/brand specificity. I would leave the fusion welding to geo installers - they have irons to do this properly. Plus, at a certain point, you are asking the installer to trust the work of others which they may not want to do. I can't see them being too upset with the loop pre-install (take pictures/document it well)

 

SDR 11 pipe is fine, since you are likely looking at 3/4" pipe. Popular with DYI is to make the circuits a bit longer, let's say 700ft instead of 600ft, and run individual loop out to the loopfield. So you bring your pipes in individually, so you can purge with your flowcenter/looppump, since you can shut of all the other individual circuits.

You don't need a purgecart, and you do not need a header system outside. just some shutoff valves on the inside for each individual loop.