Just curious what some of you more experienced designers consider the minimum distance between geo tubes in a horizontal system?  I am pretty much writing off the "slinky" deisgn due to all the additional friction loss and and possible reduced performance from the btu/sf density in the ground. 

I was running some calculations and since I want to target 8-10ft deep in the installation, trenching would simply take too long.  My other consideration is using our track hoe to dig the trench but at a width of approx 4ft and create a lateral U tube with spacing at 4ft apart.  I can separate each U tube assy by more than 4ft but just trying to ballpark the size of the excavation right now. 

I am trying to obtain a reasonable U radius to reduce head losses.  I am looking pretty hard at 1.25" HPDE for this installation but I have not completed thermal calcs to know for sure yet. 

There is no minimal distance. Pretty much all designs work to some extent, it's a question of overall cost effectiveness. Depth, pipe contact surface and volume of soil that heat is being extracted from are all factors, but I doubt that anyone is modeling it correctly without loop software.

Do you have some software like "loop link" or other? Writing off a slinky when you are talking about digging with a track hoe 4 feet wide makes me wonder if you have run the numbers. The exchange is all about feet of pipe in the ground to get the job done and the most cost effective way to bury all that pipe.
Eric

I am only working off of physical property data and built a few calcs in excel for now. We are able to model/design certain systems with our 3D CAD systems but they are not purpose built for it. Never the less, I think one of my driving factors is the head loss and velocity in the pipe. All those loops hurt flow on paper. I also found a few docs that indicated there might be some concern of changing EWTs with slinky designs due to the higher BTU/CF density and poor conductivity of soil. There was also some mention of tube pinching with soil compression at the tube cross points but I would have to see proof of that. In my pipe laying experience, we typically bury pipes in sand and one of the reasons is uniform pipe pressure.

Because I will not be contracting someone for the excavation, I am not as concerned with "cost effective" as much as "reasonable" and "correct". Common sense would indicate that long, straight runs would be beneficial for head losses. However, I am all ears if there are other thoughts and experiences with the slinky design. It certainly seems common these days.

We have done 15 or so slinky jobs and the all work great. You will dig less trench and get more pipe in the ground by doing slinkies. You will however not have fun building slinkies out 1.25" pipe. You really need a loop design program to model the different options.

I am not sure what lengths of pipe and/or flow rates you might run but I am really paying attention to flow rates and head losses right now. Because I will run a passive system at times of the year and cycle my geo field in with the radiant floor system, the combination could be a lot of loss. right now, I am working to stay at 10-15ft loss through the geo field.

I already started this from a thermodynamics POV and got to looking at fluid flow and ran into a wall. I am not sure what you guys use for conductivity of soil but I am trying to error on the side of caution until I get a sample or learn more there. The rest should be a fairly straight forward calculation.

Many of us here are guilty of "trying to reinvent the wheel". You certainly seem headed in that direction also. I urge you to purchase some simple software or use some share ware so you can quickly model the impacts of different designs. Geoexchange is a simple matter of pipe in the ground to carry the load. While head loss and feet of head and amount of pipe needed to carry the load are some what technical, do not make this excersise harder than it needs to be.
Sites on the web are full of happy stories of diy homeowners who borrowed a trencher from work, installed all the pipe they could afford or fit in thier yard at depths shallower than the soft ware indicated and they work well. Geothermal is one of those fields where thier are not very many hard and fast rules. If you do follow them the rest is diggers choice/ pun intended.
Eric

Your calculations are likely leaving out turbulant vs laminar flow. If you want the best heat exchange per foot turbulent flow is the ticket. Putting 1 1/4" utubes in the ground requires a greater supply of pumping power to get turbulent flow.
There are all sorts of systems that will deliver required btus but you are running equations that are not emphasizing the heat exchange side of loop design. This could make overall system less efficient while saving pumping power.

Distance between trenches of well designed loops does have minimums, but if you lay a few miles of 1 1/4" pipe it won't matter.

Since you are most interested in what is "reasonable and correct" why not listen to the pros here that have repeatedly suggested you may be on the wrong track.
j

Joe makes a good point about turbulent flow. You need to move the water fast enough to exchange heat. This will require some head pressure.

Maybe this has been said, but the difference in head loss between a slinky loop and straight pipe is negligable for the same length of pipe. For a given load, a slinky will require a bit more pipe, so as a practical matter there may be slightly more head loss in practice. I think in terms of pumping work you'd get more bang for your buck if you paid careful attention to pump sizing and efficiency. I went with straight runs for my diy loop mainly because backfilling a slinky loop properly seemed to be a bit harder based on people's experiences and I didn't want to be experimenting with making loops while the excavator was on the clock.

Posted By cnygeo on 10 Aug 2011 09:24 AM
1) For a given load, a slinky will require a bit more pipe, so as a practical matter there may be slightly more head loss in practice.

2) I think in terms of pumping work you'd get more bang for your buck if you paid careful attention to pump sizing and efficiency. 

1) Curiously the designers of CM's op cost calc software don't agree indicating shorter lengths for slinkies than multi pipe trenches in many of my calcs.

2) .....and $/kwh

Point is good design is specific not only to conductivity of soil but demand (load) and cost of energy. It certainly doesn't happen by accident.

Most importantly while loop design matters it is only one of the "fish" to fry and since most DIY folks tend to over compensate the bigger risk is on the load, duct and heat plant side of the equation.

j

Posted By joe.ami on 10 Aug 2011 09:52 AM

Posted By cnygeo on 10 Aug 2011 09:24 AM
1) For a given load, a slinky will require a bit more pipe, so as a practical matter there may be slightly more head loss in practice.

1) Curiously the designers of CM's op cost calc software don't agree indicating shorter lengths for slinkies than multi pipe trenches in many of my calcs.

Are we talking pipe length, or trench length? I can't think of any case where you would need less pipe in a denser configuration. Not doubting you, just curious. I guess if you're talking 4 or 6 pipes in a trench it would make sense. I was thinking of 2 pipes per trench 3 feet apart, which was what I happened to use.

talking pipe-

Obviously there is more work to go into this and I appreciate the info already presented. This thing will not be built for another 6-8mo. I am just working through some numbers to see which design type I might want to approach. Joe, you are correct that I did not account for turbulent flow. Those eddies will certainly be an advantage. Back to the Reynolds calcs I guess..

I am still trying to establish some ballpark head numbers. I realize I am getting knocked for considering 1.25" pipe but I do not see 30psi loss being anywhere close to ideal considering the power it will take to operate that.

Reynolds number over 12000. I need only 2000 for turbulent flow. Safe to say, turbulent flow will likely be a non-issue here but I will admit, I was not thinking about it..

Are you looking at a single long loop rather than parallel loops? 30psi and Re of 12K are very high comapred to typical loop numbers. What flow rate are you targeting?

I would have to have 3-5 parallel loops to be practical whether slinky or straight U tube style. I am targeting 2GPM/ton of cooling at a ballpark for now. 15tons total over 3, 5ton units. I will only instal 1 unit right now. the rest of the tubing will be used as a passive system linked with the radiant floor tubing. With 1.25" pipe, I can get more manageable 23ft loss. I very much doubt I will need this kind of flow for the passive radiant system but have to design it for a pair of 5T climatemasters. Again, these are only ballpark values right now because the flow rate will be determined by the EWT and deltaT across the exchanger among other things. Because I will stay well inside of turbulent flow, I see little point in going towards the smaller build size which could prove more problematic.

For the passive system, I should be able to run a 1/8-1/12HP circulator. As of now, it looks like I will need 2, 1/6HP circulators for the heat pumps. Obviously the pumps would be the last part of the puzzle but I am just crunching ballpark numbers here. I want everything to operate discretely if required which means 1, 2, or all 3 systems depending on demands.