Geothermal "horizontal" loop circulation

Discussion in 'HVAC Heating & Cooling' started by scott_r, Jul 30, 2008.

1. scott_rCursed Dilletente

Joined:
Dec 17, 2007
Hello all,

I've posted here before, but always with Q's about wells. This is a little different - an HVAC issue. I'm getting ready to replace my 35yo oil furnace before the cold weather arrives - it's inefficient, and oil is killing me. I'm ready to stop enriching Arab Princes (and a few American Oil Princes as well!).

I'm going to go for a Geothermal System (Ground Source Heat Pump) - I've barely got enough well water to serve my domestic water needs, much less "waste" running through a heat pump. So, I'm going to go with a closed loop system - 2000ft of PE buried about 6ft. deep in the yard - this is achieved with 4 500ft. loops, connected back to a breakout manifold, which is in turn connected to a "flow center" (two pumps in series).

The system is sold with a "QT flow center" which has two Grundfros 2699 pumps to circulate the coolant (water + prop. glycol solution) through the loop. My concern is that I live on the side of a "mountain" (really, a ridge) and I have a very steep yard. I'm concerned that these pumps won't be sufficiently strong to circulate the coolant through the loop well (and will burn up pretty quickly). The vendor assures me that based on the loop length, one pump by itself would "almost" be enough. However, I think this is calculated based only on the resistance of pumping through the pipe, and assumes the loop field is level.

So, I have two questions:

1) what is the formula for calculating the friction 'head' of a linear pipe run (this is 3/4" PE pipe)?

2) If the loop is on a hill, regardless of where the inlet/outlet is on the hill, does the energy gained from gravity on the downhill portion of the loop mostly (some loss to friction) cancel out the energy required for the uphill portion?

I assume there is some internal resistance/friction factor for the pipe which is responsible for the 'head' in a horizontal loop, which would cost me some of the downhill energy.

2. Southern ManDIY Hillbilly

Joined:
Jun 27, 2008
Location:
North Carolina
1. The system manufacturer may have tabular values of friction loss based on the flow rate required, the fluid mixture used and the pipe recomended. It would also have the equivalent length of pipe for each fitting. Since the velocities in the pipe are going to be low, these numbers are probably very low also, and its normally not an issue with the type of pumps used.

2. If its a closed loop you don't have to worry about where the pipe goes vertically. Think about how a siphon works.

The only exception is that you want the pump low in the system, so you don't end up creating a vacuum that could vaporize the water at a low point in the pump. The highest a pump can practically "lift" water is about 10-15', and is reduced as the temperature of the fluid is raised.

4. jadnashuaRetired Defense Industry Engineer xxx

Joined:
Sep 2, 2004
Occupation:
Retired Systems engineer for defense industry.
Location:
New England
Once you get the air out of the system, you've essentially got water falling downhill helping the pump to push it back up the other side. If it can overcome the friction of the pipes and fittings at the desired flow rates, it will work as stated. Until you get the air pockets out...it may be tough pumping it.

5. Bill ArdenComputer Programmer

Joined:
Sep 30, 2006
Occupation:
computer programmer
Location:
MN, USA
Yes, but you want the air remover at the highest point in the system.

Edit: Is the Grundfros 2699 Bronze or cast iron?
I would also suggest using bronze or brass pumps with all non steel/iron fittings.

Last edited: Jul 31, 2008
6. scott_rCursed Dilletente

Joined:
Dec 17, 2007
So, this is a tricky part - I am more or less stuck with putting the loop field uphill from the house or downhill - both the pump and the bleeder will be located in the house. However, it seems like putting the pump at the bottom might be best - easiest to drain, and I'll always be pumping up.

It is possible there will be air trapped at the very top of the system; this is a coiled loop system, so that may be unavoidable.

I hadn't even thought to check this, but it's a very good point; I have a terrible time with iron (galvanized steel) parts of the plumbing rusting. I expect the oxygen in this system to be exhausted as it is in a hydronic system, but I'd still rather not have iron stuff rusting out. The Grundfos appears to be Bronze.

Well, I'm going ahead and springing for this system. I'll post pics and share the outcome!

7. Bill ArdenComputer Programmer

Joined:
Sep 30, 2006
Occupation:
computer programmer
Location:
MN, USA
You might end up having to run the return loop up near the ceiling in the basement and place the bleeder as high as you can before running the return back down.

The key will be keeping the pressure high enough so that the water does not turn to steam at some point.

You will also need enough flow to force the bubbles to the bleeder.

Joined:
Oct 20, 2005
Location:
New Hampshire
If you maintain pressure in the loop so that it is above the boiling point at the highest point in the loop you will not have any problem.

You can do that with a bladder tank or expansion tank.

Someone (presumably the seller of the system) should have done an analysis to determine the economic optimum for pipe size, length of each loop, and characteristics of the pump.

You need the following information:
1. A curve that shows head versus GPM for the pump. I assume that each pump operates two of the loops.

2. A curve or table that shows head loss for the pipe. http://www.pacificpumpandpower.com/docs/Friction_Loss_Table.pdf
Values in the table at the link are feet of head loss per 100 ft of pipe.

On the graph of head versus GPM for the pump, plot the head loss versus GPM for the pipes and heat exchanger.

Example:
Calculate the head loss in your system for the GPM range at which you expect to operate the pump. Here is an example for the 10 GPM point, which I take to be 5 GPM through each of two 500 ft loops.

At 5 GPM the head loss is 7.9 ft per 100 ft = 39.5 ft for 500 ft of pipe.

Add head loss for fittings (same link). Assume 100 ft of pipe for fittings FOR ONE RUN, 7.9 ft of loss in fittings.

Since two pumps (I'm assuming they are in parallel; not in series) are pumping through the heat exchanger/evaporator, determine the head loss through the evaporator. Your supplier should give you a chart of that head loss. Let's say two pumps in parallel at 10 GPM per pump puts 20 GPM through the heat exchanger so you need the head loss at 20 GPM. Let's say it is 20 ft of head loss through the heat exchanger.

Add the head loss through any control valves or other valves or equipment in the loop. I'm going to use 15 ft of head loss through controls and other equipment.

Pipe friction + fittings + evaporator + controls and other equipment = 39.5 + 7.9 + 20 +15 = 82.4 ft.

Plot that point on the pump graph at the 10 GPM point. If the calculated point is above the head line of the graph then your next point should be calculated at a lower GPM. Try lower flows until your head loss graph (the curve connecting several head loss vs. GPM points) crosses the curve of the pump graph. That is your operating point.

If I have made wrong assumptions (if your pumps are in series rather than in parallel) then you will have to adapt the approach to your system.

If you are buying a system from a supplier then he should have done (HE MUST DO) this calculation for the system. You should ask him for it. If he doesn't know what I am talking about, and if he can't show you the curves and how he got them, then he shouldn't be peddling systems.

If he is selling systems that have been "designed" by others then he should get that data from them.

I would not buy such a system from someone who could not demonstrate to me that the system is economically optimized, and that the length of pipe in the ground, the pumping power, and the compressor power are the right combination. He should be able to tell you what the temperature of the ground will be, what the temperature of the return water will be, and the temperature of the air delivered to your house.

He should be able to tell you (guarantee, for a specified condition) the kW necessary to run the pumps and compressor.

He should also give you some gaurantee on coefficient of performance, which is the ratio of BTUs out to your house to the BTUs of electricity necessary to run it. It should be "Energy Star" rated.

9. scott_rCursed Dilletente

Joined:
Dec 17, 2007
Well, that would be a neat way of ensuring the highest point in the loop is in an easy place to bleed, but my house sits on a hill - so bringing the pipes in above the freeze line or below won't affect the "high spot" problem for bleeding, but will affect the freezing pipes in winter!

Oh, boy, so this loop doesn't operate anywhere near that kind of temperature! Summertime loop temps are around 78F and winter temps are around 38F.

The flow center (two pumps in series) is set up for >3GPM flow, and is purportedly sized significantly larger than what's needed. What I was originally looking for is a way to calculate how large a pump I need for a given length of pipe in order to maintain good flow.

I've given some thought to that - the tubing is rated for 160PSI so

Bob, wow! That is exactly what I was looking for. Yes, the pumps are in series, not parallel, and yes, they did a design analysis - in fact, the pumps are part of an application specific package - I'm not sure if it's really kosher to post links to manufacturers, but:

http://www.bdmfginc.com/html/qt___ea_models.html

I've got the 2 pump QT flow center coming. It's purpose designed for this job, and sized to handle a range of loops - mine is at the small end of that range, so the pumps should create more than enough flow. I really believe my main issue is this question of adequate bleeding. However, given your pointers, I know where to go look to work out the solution for myself. Thanks very much!

10. Southern ManDIY Hillbilly

Joined:
Jun 27, 2008
Location:
North Carolina
Water will turn to vapor (steam) at low temperature when it is under a vacuum.

11. Bill ArdenComputer Programmer

Joined:
Sep 30, 2006
Occupation:
computer programmer
Location:
MN, USA
You can run the pipe into the house below the frost line as long as the flow rate threw the pipe is faster than the bubble rising rate for the diameter of the pipe.

The water will turn to steam at room temperature if the high point is more than 32 feet above the bleed point. This height decreases with temperature and becomes zero at the boiling point.

I think that things will work out just fine as long as you use a good size bleeder that will let bubbles settle out.

PS: The really great thing about ground loops is that you can always add more if needed.

12. scott_rCursed Dilletente

Joined:
Dec 17, 2007
Makes sense - this is 3/4" PE - I'll check the manufacturer's literature to find out what the bubble rising rate is.

Sure, good old PV=nRT

I hadn't really thought about this, but certainly, I can see where the fluid moving downhill would pull enough vacuum to create vapor (cavitation?) at the top. I just figured the pressure differential (the water is being pushed up the loop as well) wouldn't be that great.

So, now, I wonder: what difference will it really make if there's a vapor bubble trapped at the top of the loop? The pipe is rated for 160psi; is the simple answer to pressurize the heck out of the system?

I'll do a few calculations based on the water volume (weight) and see what the "pressure" is at the top of the loop.

I think that things will work out just fine as long as you use a good size bleeder that will let bubbles settle out.

PS: The really great thing about ground loops is that you can always add more if needed.[/QUOTE]

Joined:
Oct 20, 2005
Location:
New Hampshire
The problem with cavitation has nothing to do with PV=nRT. It has to do with the vapor pressure of water at temperature. If the absolute pressure at a point becomes lower than the vapor pressure of the water at that point it will evaporate (boil). You need a substantial pressure margin to keep from getting cavitation at fittings and pumps because the dynamics of the flow reduces the pressure. With pumps the required pressure is called the "Net Positive Suction Head Required", usually abbreviated simply NPSH required. It is rarely a problem in household applications.

That is also why you can't recover all of the energy from water being returned to a well in a closed GSHP system if the water level in the well is more than about 30 feet below the highest point in the system.

All of that is completely unrelated to the usual problem of air bubbles in pipes. If air gets trapped in a pipe it will collect at high spots. If it is on the suction side of a pump it can make it impossible to draw water through the pipe because of the way it affects the static pressure in the pipe. If the suction pipe has a high spot you need to provide means to remove the air. A big centrifugal pump with a "self primig" configuration will usually handle a modest amount of air if the highest point in the suction line is the inlet of the pump. Most small pumps have problems starting.

Air on the discharge side is usually entrained and discharged with the water.