Horizontal Loop Field - Advice Requested - Southeastern MA

I am purchasing a 5+ acre lot in Southeastern Massachusetts and plan on building a ~4,000sf house with a closed-loop system. Now I know that the only way to properly size a system is to determine the heat loss for the home and test the soil, but this is a very unique situation in that I have the opportunity to install a loop at a hugely reduced cost if I do it right now. When I say hugely reduced, I mean almost to the point that the cost is irrelevant, the excavation and installation will be free and the materials will be deeply discounted to the point that I will happily oversize the field significantly even if the expected benefit is expected to be small, as long as it doesn't come with major drawbacks and gives me a better chance of the system being suitable for the final design. So all in all, the loop cost is almost irrelevant, but only if I do it now.

The house design is still totally up in the air and definitely won't be settled on and/or finalized in time to actually know the heat loss for it, but at the same time, that also means that I can adjust the design to bring it in line with some targeted heat loss that the system was designed around, if necessary. I can also probably get the soil tested relatively quickly if that would be helpful for this process. I should also mention that we plan on installing an in-ground pool, and if it were possible to heat it using the geothermal system, again I wouldn't mind upsizing the field to account for this. FWIW, I also plan on installing a solar PV system, also with size to be determined based on what will make the most sense after seeing how everything else shakes out.

So all that being said, I would love to get some advice from some of the resident experts here on my approach for this unique situation. I experimented a bit with LoopLink RLC and came up with an preliminary guess of a potential suitable design of 3/4" 36" Slinky, 10' trench depth, 10' C-C Spacing, 15 trenches of 500' of pipe at 100' trench length, with 2" headers. I am totally flexible on changing any and all aspects of this design. Again, with excavation being free, I don't mind going deeper than usual if it is helpful at all, and the field size is also extremely flexible so it can be longer, wider, whatever works best. As far as I understand it, doing multiple trenches of shorter length in parallel allows use of a smaller circulator pump, which is why I chose the 15 trenches of 500' pipe length for this initial guess, but maybe there is a different approach that would be better for minimizing pumping cost?

I would appreciate any and all advice I can get from the resident experts here on my best path forward. Thanks!

Edit: FWIW, here's the soil data from perc testing in case it is helpful at all...

 

15 loops of 3/4" x 500' seems like more than necessary for a new construction 4000 sq ft home, assuming it is built to current energy codes or better. That is 7,500' of 3/4" pipe in the ground. I would roughly estimate 5 to 7 tons of HVAC (one unit or two?) with one 3/4" x 600' loop per ton, so say 7 loops of 3/4" x 600' (once you start going above 600' of 3/4" pipe per loop, your pumping losses start going up). There are a lot of unknowns regarding your future pool heating need, but just a guess of 4 tons W-W unit would add another 4 loops, so I'm guessing closer to 11 loops total of 3/4" x 600' = 6600' of 3/4" pipe in the ground. So, not far from what you proposed (~90% of what you proposed). Some would argue that using a geothermal unit for pool heating doesn't make a lot of sense economically due to the time of year when you are heating (assuming you aren't trying to heat it year round). Ambient air temperature is likely warmer than ground temperature when you are trying to heat the pool, so there is little to no efficiency gain from geothermal, plus an air source heat pump for heating pool water is less complicated and less costly than a water source unit.

 

I appreciate the response.

As for one unit or two, I assume one would be better, so it would seem my best bet would be to design the home to accommodate a single unit design, unless there are other considerations I'm not thinking of?

Understood on the pool heating. That makes sense. I guess I was just trying to take advantage of this opportunity to build the loop field for dirt cheap to take care of that as well, but I won't do that if it makes little sense to.

I know I am putting the cart before the horse in a lot of ways here, and I appreciate that you are doing your best to work around that and give me good advice with a lot of unknowns.

When you refer to 600' loops, we are talking slinky with trench length of around 120', right? That would be perfectly fine.

I guess my whole approach with this is that this entire field is going to cost me very little whether it's 8 loops of 500' at 6' deep, or 15 loops of 600' at 10' deep. The cost difference between those two extremes would be something like $750 all-in, so if option two would gain me any kind of significant advantage, even a relatively small one, over the life of the system, it would seem to be worth doing. So I know this situation is quite a bit different from a normal cost/benefit analysis, and with the uncertainty of the home design and load, I feel like I want to err on the side of oversizing the system. But that being said, I don't want to shoot myself in the foot and end up needing a much larger circulation pump than necessary, or two pumps, or something like that which would cost me more every day over the life of the system. So I'm trying to strike the proper balance between oversizing the system to an extent that would probably be unnecessary/foolish with normal field installation costs, but also not setting myself up for unnecessary future cost because of it.

So my question really is, knowing what you know about this situation, and if you were in my shoes, where the cost of the field is essentially irrelevant but the cost of running the system is fully relevant, how would you size the field?

 

Without a house design it is hard to properly estimate heating and cooling loads, but since you don't necessarily care that a bad assumption might turn into a huge loop, just make some conservative assumptions to come up with a load. Since you are in a heating dominated climate, I would probably assume something like 20 BTU/hr/sq ft for heating (so 80,0000 BUT/hr) and probably something like 10 BTU/hr/sq ft for cooling (so 40,000 BTU/hr). Good construction with current energy code would likely produce smaller loads, but since there is no cost premium on the loop for you, I would just go with those unless you have better info. With those loads and size of the house, I think it is unlikely you could use only 1 heat pump and even if you did, you may not like the heat/cooling distribution. Regardless, how do these numbers compare with what you put into LoopLinkRLC?

Because horizontal slinkys generally use 3/4 pipe, friction losses can get big on longer individual loops (say versus a vertical bore with larger diameter 1.5" or 1.25" pipe). Friction losses are proportional to the flow velocity squared. But you need higher velocity in the slinky loops to get low level turbulent flow in order to properly transfer heat in the ground. Its a double edge sword. To optimize pumping, you break up the overall flow into multiple parallel loops (like you and arkie6 are considering) to reduce the velocity to a range where you balance maximizing heat transfer while minimizing friction losses. The individual loops can connect back to a singular junction point were a header pipe (that collects all the slinky loops together) can run into the house to the heat pump(s).

Since you aren't worried about oversizing the ground loop, I would probably assume you need 2 units - say one 5 ton + one 4 ton to get the highest possible flow. Typical flow is 2.5 to 3 gpm per ton but why not assume 4 gpm/ton (again, higher to get bigger ground loop). So with 9 tons of heating/cooling, that would be 36 gpm total flow (in the header). If you had 12 loops, that would mean 3 gpm per loop (I believe that is within the sweet spot for 3/4 HDPE minimizing friction but maintaining turbulent flow). So using worst case numbers, 12 loops. In reality you probably will need less closer to what Arkie6 is proposing. But if cost doesn't matter, install 12 (or 15 or whatever you want), bring them into a header box in the ground set away from where you think home will be. Then when you get house more finalized, connect however many you need and leave others capped in the box for future use (or never use).

I'm one of the people that discourages geo for pools unless you plan on swimming year round, have an indoor pool, or swim well before and after the typical memorial day to labor day timeframe. air source is just too easy for the typical pool heating timeframe and air will almost assuredly be warmer than your ground temp for the typical late May to mid September swim season. If you were in Florida and attempting to cool the pool, geo might make more sense for that.

 

Why are you proposing slinky loops rather than straight out-and-back (race track) loops? Slinkies are usually done due to space constraints. Usually, the amount of earth exposed to the loop is the limiting constraint for heat transfer, not the feet of pipe in the ground. The amount of pipe matters, but the amount of earth in contact with that pipe usually matters more (unless you have extremely high heat transfer in your soil). With a 3' radius slinky having 600' of pipe we have an earth contact patch of 3' x 120' = 360 sq. ft. With a 600' race track loop, you could dig with a 2' bucket (~50% faster digging than with a 3' bucket) and would place the pipes on either side of the trench, but would have to go out 300'. The earth contact patch for this loop would be 2' x 300' = 600 sq. ft. And you don't have to build the slinkies.

Another consideration is the amount of flow down each loop. Assuming a 6 ton unit that would need ~3 gpm per ton or 18 gpm of water flow. Having 15 separate loops means you only have slightly over 1 gpm per loop which is less than recommended for turbulent or near-turbulent flow in 3/4" pipe. From a per loop heat transfer perspective, it would be better to have closer to 3 gpm per loop. If you did put in that 15 x 500' loop field, I would consider splitting the field up into say 10 + 5 with the 10 loops on one header back to the house for the main heat pump and the 5 loops on a separate header for the future pool heater if you decide to go down that path. Or maybe three separate groups of 5 on their own header such that you could valve them in and out of the system as necessary as future loads are added. If you went with 5+5+5 loops, you might be able to drop back to 1.5" header pipe depending on the distance to the loop field.

Is your home single story or multi-level? You can zone a single unit to serve multiple floors, but generally you get better comfort with a unit serving each floor. In the winter, the downstairs heat pump would be doing the bulk of the work as the warm air rises to the upper floor while in the summer the upstairs unit would be doing the bulk of the work as the cooler air sinks to cool the lower floor. This would likely provide better dehumidification in the summer as the one smaller unit upstairs would be operating for longer intervals than one large heat pump covering both floors and these long run times are needed to pull the moisture out of the air. Plus, having two separate units gives you some redundancy in the event one fails. Plus, if you ever plan to have a back-up electrical generator, you could size it smaller to accommodate starting one medium size heat pump vs. sizing it larger to start a large heat pump.

 

arkie6, not sure if you are addressing this to me or OP. I was only attempting to calculate the maximum feasible number of horizontal loops given the OPs fortunate situation where he can have individual loops installed seemingly for nothing now but pay handsomely for it later. Weather they are slinkies or out and back straight loops wasn't really my focus - I'll let you guys have that discussion. Since he doesn't even have a house design, load is really a guess. If there is no financial penalty for installing the loops, I would vote to install more, bring them to a header box, and leave them capped in the box until the house is planned/developed. You don't actually have to connect every loop that was put in the ground to the header. How many you connect depends on the house load and attempting to split flow to your (and my) 3gpm per loop target.

I'll let the OP describe this potential house better, but my reading of his original post indicated he didn't have any idea other than approximate size. If its two story, I would vote for two heat pumps, 1 serving each floor. I think you could be hardpressed to heat a 4000 sq ft house in MA on one heat pump anyway. Without details its hard to know which is why I rounded all of my math up to come up with a massive field that the OP wouldn't be disappointed in at a later date when it would cost $$$ to install.

 

Thanks for the responses, guys.

In LoopLink, I had used an assumption of 60,000 Btu/Hr peak heating and 50,000 Btu/Hr peak cooling, but then selected a Waterfurnace
5-Series NHW084 (ELT-080/50), which seemed to be way oversized for that load, but again, I wasn't sure and was just erring way on the side of oversizing everything for this initial design guess while admittedly not really being clear on the downsides of doing so. I then played around with 36" 3/4" Slinky fields, with 10' C-C spacing and trench depths of 8' or 10', different numbers of trenches and noting the lengths LoopLink was suggesting. I then switched to Fixed Length mode and used a length of 100' because it corresponded to pipe lengths of 500', which seemed like reasonable lengths to reduce friction losses, again based only on my limited knowledge. FWIW, there is also a 100' x 150' area that I sketched out in the backyard, out of the way of the house, pool, patio, etc, which works well with my proposed lot layout and seemed to be a suitable size for these oversized slinky fields I was playing with.

I could certainly do race-track loops, it's just that I thought with that 100' x 150' footprint that seemed to work really well for my proposed lot design, a slinky field would make a lot of sense, and it seemed that I was already over-sizing the field so there wasn't much to be gained by doing something else. We have excavators of all sizes, so I was envisioning getting a huge excavator in there for easy digging of the 8'-10' deep trenches, and the guys I would be building the field with put pipe in the ground all day every day, so I thought building the slinkies would be easy enough for us, but I am definitely totally open to other suggestions, so I'm glad we're talking about these details. I could expand that footprint to some extent but doing so significantly would probably require me to clear more trees than I'd like to.

The home will be 2 stories (plus a walkout basement not included in the 4000sf), and I'm not sure if this is relevant, but ~800sf or so of the 4000sf I mentioned earlier will be a dedicated theater which will be completely sound-proofed, meaning it will be totally air-tight other than the HVAC. I also do plan on having a backup generator, so it seems that everything is pointing toward having two units.

I also have taken into consideration what you guys have said about heating the pool, and I agree, I will likely be heating it during the normal seasonal timeframe rather than year-round, and so it probably makes more sense to go air source for that.

Any advice on trench depth? Obviously it's more work to dig deeper, but I figured that would improve my chances of being in moist soils, and in LoopLink, going from 8' to 10' seemed to improve the field performance significantly. This decision could change I suppose based on field conditions, such as if we hit the water table or if there's ledge or something, but I'd still be interested in hearing your thoughts on this specifically based on your experiences.

Thanks again for the great responses!

 

its anyone's guess on the house load. my gut says your heating might be okay possibly low for this oversized loop field exercise. Cooling is probably fine unless you anticipate lots of windows facing west. You will definitely be heating dominated so either your heating is too low or your cooling too high. adjust as you see fit. I would anticipate two heat pump units.

On trenches, all other considerations being equal, deeper is obviously better because soil temperature becomes more stable with depth. But for actual construction, it gets quite difficult to dig a 3' wide trench (or 2' or 4') that is 10' deep without the need for bracing/soil stability to prevent cave-in. Certainly wouldn't put a person in a trench like that. If you want it 10' deep, you probably have to clear out a whole area to that depth, run piping on ground, and then backfill whole area rather than have individual trenches. You can search this forum and I think you will find more installs are like 4 or 5 feet deep. Of course that comes with greater soil temp change so you need more pipe in the ground to heat/cool the same load. run some scenarios in looplinkrlc to see how it affects the amount of pipe you need. I'll let others respond on reasonable limits for the excavation based on actual experience but it could become pretty extensive pretty quick if you want it all 10' deep.

 

In case it is helpful, I also have access to a trench box/shield, which may make sense to use to do one trench at a time, place the slinky, and then backfill before moving on to the next trench. Just some food for thought, possibly more trouble than it's worth.

 

I would not expect a deeper trench to improve the performance by a significant amount. It can reduce the amount of pipe needed, but would require moving significantly more dirt. I doubt using 10' deep trenches makes sense. My 4T WF7 system uses 700' of trenches that are 3' wide and 5' deep. According to LoopLink, going to 10' deep would reduce the trench length required to 460'. However, the 10' trench would require moving 30% more cubic feet of dirt, which doesn't seem like an advantage.

700 x 5 x 3 = 10,500 cu-ft
460 x 10 x 3 = 13,800 cu-ft

 

I was replying to the OP. I had typed up my reply the previous night before your post, but forgot to hit the "Post Reply" button, so it didn't get posted until the next morning.

 

Okay, I did some more experimenting with LoopLink, and based on an "ideal" 3gpm/ton and 3gpm/loop, and thus 1 loop/ton, it seems each loop must be very deep, or EWTmin ends up very low. I realize I could be missing something or making a bad assumption, but even if I go with 11' C-C spacing (it shows no benefit beyond this point) and 10' depth (which as discussed is deeper than ideal for excavation difficulty), and 600' loops, achieving 3gpm/loop results in EWTmin of ~25-27 F, and going down to 2gpm/loop only brings that up to ~32-33 F. What am I missing, or what is my bad assumption?

 

you are in a cold climate and using this for heating, so the loop temp is going to get low. I have vertical loops and got down to around 35 this winter. Typical winter design temp for heat pumps is 30 deg entering water temp (EWT) into heat pumps which means the leaving water temp (LWT) from the heat pump is on the order of 5 degrees (and possibly more) lower than the EWT (heat pump extracts heat from the loop to heat your house). So it wouldn't be uncommon for a loop to be at 25 degrees at the coldest point of the winter. This is achieved by using antifreeze in the loop fluid instead of straight water. Typically ethanol or methanol is used (PG can be used but is much thicker increasing pump wattage). Some states (like mine in NJ) require use of ethanol (rather than methanol) due to concern regarding the potential for a leak to put the fluid in the ground. Methanol is preferred from a design/construction standpoint because it is the least viscous at these temps providing less pumping power needed. Not sure if MA requires ethanol or if methanol is allowed. I can't remember the looplink inputs, but if you have 30 deg F EWT for design, then I would expect a minimum loop temp hovering in the mid 20s. Fluid won't freeze due to antifreeze mixture, but you will freeze the water in the soil around the loop. I don't know if there are design considerations for this aspect or not - ie locate away from stuff so if ground heaves it doesn't damage anything, or if this dictates minimum spacing, etc.

I was just looking at the LoopLink output from when I did this and you should have to enter an EWT min and EWT max. I would use 30 for heating (EWT min) and 90 for cooling (EWT max). You can make these anything you want, and these numbers are widely used to balance loop field size, diminishing returns from an ever bigger loop field, and the range of heat pump ability.
If you make the EWT min higher, its going to make your loop field larger (same if you make the EWT max lower). I don't know how much larger, but at a certain point there are diminishing returns. I'm not sure what the earth temp is where you are at, but just using some made up numbers, if 6' down its 50 deg in dead of winter, then you can't have your EWT min at 50 because your loop field would have to be like the size of your county. Maybe it could be 40 but it might take up your whole property. Anyway, 30 is widely used so I would't worry about that and heat pumps are rated to extract heat from 30 EWT so it isn't like that temperature will negate the system ability.

 

I did know the loop would be fine at <32F with the antifreeze, I was just trying to avoid heaving if possible, even though with my proposed field location it wouldn't be a major problem, it would be nice to avoid it, especially if it came with more efficient performance from the unit(s).

After experimenting with these LoopLink designs, I'm guessing that, say, 7 loops at 3gpm is more effective than 10 loops at 2.1gpm, otherwise we wouldn't too much about getting close to that 3gpm mark, right? Or does LoopLink already account for turbulent/laminar flow, and so this is reflected in its calculations? If so, why would 10 loops at 2.1gpm not be preferred? Again, I'm sure I'm missing something, not trying to suggest I know more than anyone here, just want to make sure I fully understand the balancing act on field design and flow rate

 

To avoid using antifreeze in your loops, you would need to assume an Entering Water Temperature (EWT) of well above 40F (closer to 45F) because you are looking at around 5F drop through the heat exchanger, plus the refrigerant side of the heat exchanger has to be colder than the fluid to extract heat from it. This isn't likely to be feasible in a northern US heating dominated climate.

 

for design conditions (dead of winter, ewt at 30, outside temp at ASHRAE 1%, interior temp at say 70, etc.) you want the loop operating at turbulent flow so that mixing of the fluid within the pipe occurs. Otherwise, at laminar flow, you tend to get flow through the middle of the pipe cross-section while the outer fluid against the pipe wall is somewhat stagnant. Heat transfer occurs at the pipe wall so the turbulent flow guarantees mixing of the fluid which guarantees heat transfer. Laminar vs turbulent flow is dictated by Reynolds number. in geo systems, it is commonly believed that Reynolds numbers above 2500 will get you enough mixing to guarantee heat transfer. In fluid mechanics, Re of 2500 is probably more likely in the transition range between laminar and turbulent but that can apparently be ignored because enough mixing occurs to allow for proper heat transfer. So you want flow (gpm) high enough to get Re>2500 but low enough that it doesn't create excessive friction between the fluid and pipe wall (ie develop head loss that your circulating pump has to overcome). Friction is proportional to fluid velocity squared so higher velocity has a penalty to system efficiency just like lower velocity has a penalty. When I quoted 3gpm in each loop the other day, that flow rate in 3/4" HDPE DR11 pipe with water/methanol mix at 30 deg gives you an Re above 2500 while also keeping friction at around 2 to 2.5 ft of head loss/100' of pipe (middle of appropriate range). I wouldn't go below 2gpm (for fear of not getting enough mixing) or above 4gpm (friction losses start to pile up and you need more pumping). Remember, you can dictate how much flow goes in each loop because the total flow needed for the system gets divided up between the loops. So if you need 6 tons of heating (likely low end for 4000 sq ft house) your total flow would be like 18 gpm (2.5 to 3 gpm/ton is what heat pumps want), so you would want to have 6 loops to stay near 3gpm per loop. If you need 9 tons of heating (very leaky not well insulated 4000 sq ft house) your total flow would be like 27 gpm so you would want 9 loops to stay near 3gpm per loop. Since cost to add loops isn't critical for you, put in 10 loops (or 12 or 15 or whatever) and call it good. Then when design is all done, connect how ever many loops you need to get flow in each loop near 3 gpm target.

 

Oh, I'm not trying to avoid using antifreeze in the loops. I wouldn't take that chance, knowing there's some possibility almost no matter what I do of temps dropping below freezing.

 

I appreciate your detailed explanation, and that sort of planning ahead with different loop quantity tie-in options for loads of 6-9 tons is exactly what I was trying to do earlier today when experimenting with LoopLink.

I played around with different heat pump choices to change System flow rate and then checked what the different loop options would result in for EWTmin:



I tried to stick with the recommended 3gpm/loop, although I did have to veer from that on the 21gpm 600' loop option, or EWTmin was too low and it was giving me an error.

I know I am listing some 700' and 800' options, which may be too long and start to cause friction problems, but that's what it took in some cases to hit the 3gpm target and keep EWTmin reasonable. It seems I am teetering on the edge of feasibility for a slinky field design if the system flow rate is 21gpm, unless going longer than 600' per loop is still okay.

 

i don't have horizontal loops so I'm not familiar with their design - others are going to have to help out there. arkie6 in post 5 and SShaw in post 10 give good advice.

In looplink, you typically can select a specific heat pump for each floor/zone. I'm guessing that is what you did. Only thing I would point out is that sometimes, the heat pump manufacturer will specify max flow much greater than what is needed to actually operate the heat pump. For instance, waterfurnace does this in their literature and it may translate into LoopLink (not sure). So max flow rates for WF equipment are in the range of 4 to 4.5 gpm/ton. If you ran 3gpm/ton through that heat pump you would get like 96% of the capacity that running 4.5 gpm/ton gets you. Its a strategy WF uses (and I'm sure others too) to squeeze every bit of heating/cooling capacity out of their unit for specification charts. I don't recommend that level flow because pumping power goes way up with increased flow and resulting friction all to squeeze every last BTU out of the heat exchanger. Flow of 2.5 to 3gpm/ton is totally fine for the heat pumps to operate. I point this out only because the "System gpm" table in your chart would seem to translate to 7 to 10 tons of equipment (assuming 3gpm/ton). there maybe some way to adjust flow per unit in looplink I can't remember but my when I first looked at your chart, the system flows looked pretty high assuming a system between 6 and 9 tons

 

The only reason I used those particular system flows in my chart is those were just some I was able to get LoopLink to settle on after experimenting with different units, that were close (7, 9, 10 tons) to our target range of 6-9 tons. I couldn't find anywhere to adjust system flow in LoopLink, so I was selecting different units and quantity of 1 or 2 and then going to see what it resulted in as far as system flow.

It's important to note that this was all with 80,000 heat load, which I probably should have been adjusting along with this, which may mean my resulting EWTmin values are flawed.