Geothermal units locking out etc! Design troubleshoot?

Hi. New to the forum and am trying to troubleshoot our geo system. Any thoughts or suggestions would be appreciated! (Sorry I may not have all the specs needed (like soil conditions etc) to be able to figure out the answer)

New construction home. 2 floors and unfinished basement, which does not have HVAC. Approximate sq ft of finished space is 5,400.

HVAC consists of horizontal closed loop geothermal. 2 geo units.
Climate Master
*Tranquility27 split unit (air handler in attic) for upstairs -- 3 tons (12,000W aux electric heat)
*Tranquility30 for main floor -- 4 tons (15,000W aux electric heat)
Total ground loop is 4,740’
Consists of (8) 1 ¼" loops
We have 4 trenches (supposedly) 8’ deep and 4’ wide with (2) loops per trench
Header pipe is 500’ and 1 1/4” diameter. (250’ from house to loop connection and 250’ from loop connection back to house)
We cannot confirm how the loops are connected together (ie series or parallel, or if there is reverse return flow etc)

Brief synopsis:
We have never been able to heat our house adequately in the colder portion of the winters without using space heaters and running fireplaces (or noticing that we are running a lot of time on the aux heat). Can only get temp to 68 or so with geo HVAC alone. We have upsized interior pipes to/from the geo units to the loop field (from 1 1/4” to 1 1/2” where the outdoor underground header pipes remain 1 1/4”), added an in line booster pump, replaced both a geo unit variable speed pump as well as failed compressor. Have had to fix numerous interior pipe leaks and also flush out a significant amount of air that was in the loops. Most recently, systems have locked out completely this summer (which necessitated the leak repair, air flushing, and compressor replacement).

Basically, when both geo units are running, they seem to be starving each other and locking out or operating just below lockout threshold (and thus not heating or cooling us to desired temp). When one unit alone (whether the 3 ton or 4 ton system) is running, we are getting upwards of 8.5-9GPM flow. When both are running, max flow is no better than 11GPM total (each unit only being able to get up to 5.4GPM). Also, the heat pumps (despite having a variable speed pump) seem to always be running at 90% capacity even now with more mild temperature.

Would digging up the 500’ header 1 1/4” pipe and upsizing it to 1 1/2” (or larger) help achieve the flow we need (based on ClimateMaster recs of 2.5-3GPM per ton in a closed horizontal loop system) or could the horizontal loop field itself or how it is all connected together be the culprit? Should I be ignoring GPM and looking at some other variable (such as delta T, which I understand is dependent on GPM anyhow) to know that our system is functioning appropriately?

Thanks in advance for any suggestions/thoughts!

 

where are you located (for purpose of determining heat and cooling load)? I'm less familiar with horizontal ground loops but I would have figured a much larger header for that distance (2" or more) and the individual loops should be plumbed in parallel and likely should be smaller (3/4"). 1.25" is way too small for a header and certainly a header of that length. Plus 1.25 pipe is really big for the individual ground loops - are you sure about all of these sizes? I would think you need to confirm the parallel configuration too as that could be a major problem if they are plumbed in series. Your flows seem to be very low for the tonnage of units and likely the units are being starved for flow which suggests the problem is all in the ground loop. I don't have the specs for the tranquility units but I would have expected something more like 3 to 4 gpm/ton for each unit so 12 to 16 gpm flow for the 4ton unit at max heating/cooling, 9 to 12 gpm flow for the 3 ton at max heating/cooling, and a total of 21 to 28 gpm for both units running a max heating/cooling.

 

Thanks for the reply.

Located in south central Pennsylvania.

I am sure of the horizontal loop and header pipe diameters and length. We have some photos and documentation that confirm it. Just don’t know how it is all piped together in the field.

Am I correct in mainly looking at GPM when evaluating if our 2 units are running efficiently? Or does delta T, pump speed, pump watts, or some combination of it all give a more accurate picture?

 

There's a lot we don't know about your system, such as whether you are using pumps internal to the GSHPs or an external flow center.

However, we can say for sure that 500' of 1.25" header is a big problem and a very poor design choice. At 21 GPM the pressure drop of 500' of 1.25" header piping would be around 46 ft-hd. That's probably around 70% of the pressure drop for the entire ground loop. The fact you get very little flow when the second GSHP is active is consistent with the pumps being unable to handle the pressure drop of the loop field.

I would use 2" pipe for the header. That would reduce the pressure drop for the header down to about 7 ft-hd, which would vastly reduce the pumping power required.

 

Thanks for the info. Each geo unit has it’s own variable speed pump contained within it. Then we have an external grundfos flow center that is connected in line right where the incoming header pipe enters our basement. Our big concern as you allude to is that the pipe diameters are not appropriate for the length of piping we have for a 7 ton system.

 

three flow centers? Would it be correct to assume that you have the units plumbed in parallel relative to the ground loop header piping/external flow center? Do you have check valves in the header piping to stop each internal pump as well as the external pump from pushing backwards through a each unit? I think you need some diagrams here for us to help but it sounds like there maybe two problems - your header piping is woefully undersized and maybe (depending on the diagram) pumps pushing backwards or at least putting backpressure on the internal unit pumps.

For reference, I have a two unit system (a 3ton and a 4ton) in central NJ (so same climate) both are plumbed in parallel relative to the header piping and I have 2" header piping that goes ~100ft between flow centers and splitting up into individual parallel (vertical) loops and I remember from my design calcs that I would have had way too much head loss at anything below 2" header piping. I probably don't have more than 40ft of head loss in my whole system and based on SShaw's numbers you have more than that in just the header. You want high flow (which allows for better heat transfer) at only two points - in your unit heat exchanger and in your ground loops. The higher flow causes higher head loss (exponentially) but you need the flow at these points for the heat exchangers to do their job. You want low flow (and therefore low head loss) in all of the other components that just transport the water between the units and loop piping. So generally you want bigger pipe in the header and smaller pipe in the loops and you want each loop plumbed in parallel so that flow is split up between them.

 

Yep. 3 flow centers. 1 built into each geo unit and 1 external on incoming header pipe as per pictures. Units are plumbed in parallel in the basement. No check valves as far as I can tell - just a few valves to just turn on/off and a couple valves to possibly bleed the system or purge any air.

 

PVC piping? I know it can be used for the internal portion of the header pipes, but it is frowned upon. HDPE is preferred. Is the outside header and loop piping all PVC too or does it switch to HDPE at some point? I don't know I have heard of the outside piping ever being anything other than HDPE. Its a little hard to tell with all the pipe insulation but I suspect you are correct that there aren't any check valves in the header piping (or the desuperheater piping - another problem).

For parallel piped units like that, you need a check valve on the ground loop piping - I prefer one on the output side of each unit. This is so that water can't go the way it wants to and can only flow in one direction. Water will flow the least restrictive way which means the one with less friction which will be the shortest path. Flowing out to your ground loop is where all the friction is while flowing backwards through your other unit likely has less friction (even though there is a big pressure drop in the unit heat exchanger). Draw a picture for yourself and act as a water droplet - with #1 running, water will leave that unit through the "out" side and can choose to either flow out to the ground loop or backwards through #2 (non running) to the "out" side of the unit because there is no check valve. Water will continue to go backwards through t#2 unit and leave through the "in" side, run to the "tee" in the ground loop piping and then can flow to the "in" side of #1 unit. Going in a circle with non of that water going out to the ground loop. With both units running, there is less chance of this happening, but one is likely pushing backwards on the output side of the other unit not allowing enough flow compounding the huge head loss of the undersize header piping. Get check valves in the ground loop piping and it will most likely solve some of the issues and may even allow the system to run (albeit way inefficiently due to the undersized header).

Seperately, your desuperheater piping probably has the same issue if both units are plumbed into the same buffer tank. The units have a small pump that pushes water to the buffer tank and if you don't have a check valve (again, I like them on the output side of the unit) you get reverse flow too.

 

so I just re-read your first post and it is now making a little more sense - what we see on the inside (the insulated 1.5" PVC pipe+third grundfos pump) was added after the original installation (which was originally 1.25" pipe and no extra flow pump) but the outside header still is 1.25" pipe (HDPE or PVC?). If this were my system, I would do things in the following order:

1. confirm and/or install check valves on the ground loop piping so that all water flows in only one direction through each unit. This shouldn't amount to anything more than 1 check valve on each unit. I'm not sure what is available in PVC. My installer used a brass check valve that is installed at the output of each unit because they are available for well water systems in 1" and the unit connections were 1" (these were for WF units not sure about your units).

2. confirm and/or install check valves on the desuperheater piping for the same reasons noted in #1. My check valves are 3/4" copper sweat fittings on the copper piping.

I would think that these two changes should allow your system to function both to heat/cool the house and produce hot water. It will use a lot of electricity to pump water (using the three flow centers) but it should heat and cool the house as the units won't be starved of flow. How is the third flow center powered? Is it always on or does it come on when one of the units/flow centers come on? Keeping the third one off and getting it to turn on when one of the units turns on might be a decent way to save some electricity. Once everything works, then #3 should be planned for:

#3 redo the header piping from units (in the basement) all the way out to the loop field (in the ground) and get rid of the third grundfos pump. The piping you have is way too small and all of the efficiencies of a ground source heat pump are being eaten up by the power usage of pumping water against a very restrictive plumbing system. At a minimum, consider nothing smaller than 2" HDPE but do the flow calculations and get the head loss as low as practical which might mean something larger than 2" (haven't done the numbers so not sure but you have a long distance with alot of flow). Great source to do this is geo-flow and their online design calculators (its what I used - PM me and I'll send a link).

 

Thanks for the good suggestions. Didn’t realize you had replied a couple of times until now. Yes, the header and loop pipes are the black 1.25” HDPE. Just the interior basement pipes are PVC, which are now 1.5” (originally 1.25” and outside header/loops remains 1.25” - all 4,740’ of it). Basement pipe upsizing was 2nd measure taken. Then 3rd measure was to add the 3rd flow center grundfos booster pump. The grundfos comes on when either of the 2 geo units kicks on.

I will ask about making sure there are check valves on both units and both desuperheater tanks.

 

The check valves are a fitting that would glued into the PVC (or possibly threaded brass at the unit connections) and sweated into the desuperheater copper piping. They should be readily visible backing away the pipe insulation. If these yahoos spec'd 1.25" header piping for a system this size its unlikely they understood the importance of the check valves.

 

I had a thought - before doing anything, one test you could do to confirm that the check valves are the issue is to shut say #1 unit off, close the PVC ball valves on the loop piping for that unit (I can see them in the pictures above just before/after the PVC piping enters each unit), and run unit #2 to see how it operates. Maybe do that for a day or two and see what type of water flow and unit operating parameters that you get for #2. Then reverse it, turn unit #1 on and open ball valves on the loop piping for it, then turn unit #2 off and close its loop piping ball valves and see how #1 operates. If both work satisfactorily during this isolation, then you know the check valves are the issue because by closing the ball valves on the off unit, that essentially creates a system with check valves as water can only flow one way through the operating unit and out to the field.

 

Before you go crazy looking for check valves, take a look at the Product Manual for your packaged unit here-
https://files.climatemaster.com/rp1...ermal-heating-and-cooling-product-catalog.pdf
For your split unit here-
https://files.climatemaster.com/rp1...rmal-heating-and-cooling-product-catalog1.pdf
And compare the model numbers to Unit Model Key on page 10 for packaged unit and page 14 for split unit, to see if a check valve is factory installed on the units.

Also read pages 5-6 of the TE30 Product Manual which describes design applications and requirements for pump and pipe sizing, and calculating system head for closed loop. Obviously, your installer didn't. Excerpt follows-
"Typically residential loops consist of ¾” circuit piping and 1 ¼” supply
and return piping. If the available pump head is less than the
calculated pressure drop of all external piping and the loop, it is
recommended that larger pipe sizes be investigated such as 1” circuit
piping and/or 1 ½” or 2” supply and return piping. This will significantly
reduce system pressure drop with little change in Reynolds number.
If this causes the Reynolds number to fall to an unacceptable level, try
reducing the overall number of circuits. This will increase the flow rate
through each circuit, increasing the Reynolds number."

 

geoxne - good spot on the model number and whether or not there are check valves installed internally. OP, certainly do that before even my "test" in #12 above as it wouldn't be necessary if there are in fact check valves installed in each unit from the factory.

 

Thanks for everyone's help! Very informative and seems to confirm our suspicions that design from the start was problematic.

Our split 3 ton unit and 4 ton packaged unit do indeed both have internal check valves already factory installed per model numbers. Despite our continued insistence that something has to be wrong in the ground and that it is likely a resistance issue, we haven't made much progress until just now. Seems like the installers finally have involved an engineer or experienced designer to look into the specifics of our setup and are now willing to dig up our yard (as much as we hate to have to relandscape it all) to replace with a more appropriately sized header pipe and possibly redo the loop connections depending on what they find (no documentation exists on how exactly it is all piped together) -- as opposed to just trying to convince us to abandon and replace the split 3 ton geo unit with an air source heat pump (which I'm not sure would work anyhow given that I don't believe that we can achieve reasonable flow for one 4 ton GSHP unit alone to run efficiently with the current design).

Once they give us proposed plans for design remediation (I understand that they are going to plan for multiple scenarios, depending on how loops are piped together), we will ourselves have a third party review and assess for proper specs/design. We will look locally for this review, but does anyone here maybe have suggestions on anyone we could use, even if not in south central Pennsylvania? We just want to make sure it is truly done properly this time after all we've been through.

Will also look into having check valves between the 2 GSHP units and our 2 separate desuperheater buffer tanks (or is that not necessary if each GSHP unit is going to its own hot water system? ie we have 2 separate hot water heaters and the split unit is plumbed to the tank for our 2nd floor and the packaged unit is plumbed to our first floor water heater buffer tank) -- and also ensure that they account for pressure drop across the factory installed check valves going out via header to the ground loops, which I see is already calculated by ClimateMaster for you in their tables on their product catalog.

 

I took a look at your pressure drop. You should be somewhere in the 60 to 65 ft-hd range at 21 GPM. See attached.

Looking at the attached sheet you can see the largest components are the 3 Ton unit at 18.0 ft-hd and the 500 feet of header piping at 38.4 ft-hd. I had to guess at the inside pipe lengths and used 100 feet (50 to each unit and 50 back.) The worksheet also forces 1 1/4" piping, so I used that instead of 1 1/2". Neither of these factors affect the result significantly. The inside runs to/from the units are in parallel, so they don't add up either.

You haven't listed the pump models, so I assumed your heat pumps have the newer UPMXL pumps, which are a little less powerful than the older Magna 25-140 pumps. I also assume your external flow center has a single 26-99 pump. Based on those assumptions:
Your two internal pumps alone should be able to provide 17 GPM, which is more than the minimum recommended flow of 16 GPM.
Adding the external flow center should have gotten you to about 22 GPM.

Assuming the calculation is correct, and assuming nothing is misconfigured with your pumps or units, the results would seem to indicate a problem with flow in the ground loop. Maybe a blocked or pinched pipe or trapped air.

Note the flushing requirements at the bottom of the pressure drop worksheet: 74.4 GPM at 762 ft-hd. I don't think you can achieve that with any flush cart. For reference, the Waterfurnace flush cart has a maximum head of 120 ft. So, I wonder if your loop was able to be purged correctly, or could ever be purged correctly.

Changing the 500 feet of header piping to 2" would lower the head to 28 ft at 21 GPM. The flushing requirements are still too high though at 351 ft-hd.

I don't know what your best solution is, but if you had 2" header pipe and 3/4" loop circuit piping, you should be able to run with only the internal pumps at 21 GPM, and your flushing requirement would be only 28.8 GPM at 72 ft-hd, which is achievable. You would also have a better reynolds number for better heat exchange with the ground. As noted by gsmith22, 2" header and 3/4" parallel loop piping would be a more more typical design. This analysis shows the reasons for that--the pressure drop, reynolds number, and flushing requirements are all reasonable.

 

I'm guessing the calcs use parallel piped ground loops but I think that is one of the question marks too how all of that is connected. Regardless, if you can't flush parallel connected 1.25" ground loop piping, then that has to get dug up too. Don't let them keep any of the nonsense outside that got you into this mess. Maybe have an attorney watching the whole thing to keep things moving in the right direction!

 

Yes, the calculation assumes 8 loops in parallel, each 600 feet long. In series, such a system would have 428 ft-hd at 21 GPM. That would be a terrible design and seems unlikely. But, the calculation shows the OP's combination of three pumps should produce 10 GPM with that design, which isn't far from what he's getting, so who knows.

 

This is the same individual(s) that put in 1.25" header pipe for 500ft thinking that was good too Regardless, you have shown that the 1.25" loop piping needs to get ripped out too because it can't be flushed and I think that is the important part here. All the exterior piping needs to go!

 

Well, this is all a little disheartening. We are still awaiting redesign plans. I don’t think we want to have all the pipe dug up. The header would be reasonable but the loops themselves would be too much - as you can see they’d have to rip up a lot of landscaping, raised garden beds, and a playground. If only the header, they’d just have to deal with some irrigation lines and a few other pipes.

Loops are in 4 trenches that run basically the entire length of the yard. 2 loops per tench. They think that the loops are plumbed in series within each trench and then each trench is in parallel to each other with the header pipe that is in red running across the back of the house.

We will wait to see what is recommended but it’s looking more like we just need to abandon and perhaps replace with 2 air source heat pumps.