Washington Hp Home DIY Geothermal

After doing some figuring on the pressure drop I'm not sure if a single Grundfos Alpha is big enough for my radiant either.

I will post another drawing with some pipe lengths, mabey I can get some expert opinions here.

But I think this might be better off with two pumps, or mabey one bigger one.

 

yep, you have a bit of a problem. Lack of balance between circuits.

To good part is that you can make the physics your friend, namely that depending on the flow, the circuits can self balance themselves.

You have to get the flow up so circuits with lesser pressure drop will see so much more flow, that their pressure drop will increase exponentially.

 

Rather than worry about circuit lengths I would install a factory manifold with balancing valves and flow meters. You have one problem manifold on each of the two circuits. Throttle the short ones down until all thre are equal in flow from and to the manifold.

From the manifolds you only need to count the head of one loop.

Did you get that what you drew is a ladder? Is it possible to re-pipe the ladder into a revers return system?

Mark

 

The manifolds all have balancing valves and flow meters. The (2) 3 circuit manifolds are getting an actuator and tstat for each circuit.

I think I could change it from a ladder to a reverse return, I would have to read up a bit I am only vaguely familiar with it.

All 3 manifolds are installed with the 1" pex ran to the mechanical room but nothing else is plumbed yet, just 6 pipes stubbed out of the mechanical room ceiling.

I was also able to get rid of the Wilo pump I had so now I can choose whichever pumps will work best.

 

If each run on the two three loop manifold is its own zone I see no flow or pumping issues except if it is noisy adding a differential bypass valve.

Have you started head calculations? The Burnham or Grundfos books will show you how to do it.

 

It kind of beats the purpose of keeping the pressure drop low for reduced pumping power when one introduces balancing valves on every circuit which themselves each have a significant pressure drop.
It is an old fashion thinking to rev up the pump(s) only to reduce the flow for each circuit via a flow restrictor.
And before advising someone to repipe to a reverse return configuration, one should understand the impact of such change.
How long are the headers and are you using antifreeze?

 

With calculating head loss I need to keep working on it. I came up with 15 feet but I'm not confident how accurate that is.

I will check out the Burnham and Grundfos books and see if I can get a more accurate idea.

The header piping from the buffer tank to where it tees off into 1" pex will be about 20' total. They are in the ceiling right above the buffer tank.

I was not planning to use antifreeze on the load side, but I will use some methanol on the source side.

 

HP:

You are getting this. It can not be very hard as it only took me 35 years to learn what I know.

Length of pipe, X head of size, then add fittings for a total of what the piping adds to the push the pump needs to do. Then we find out how many GPM the unit needs, Hand all that to a 5th grader and do what she tells you to do. Take her number to a pump manufacturer. Order a pump.

About I can not remember when, the Wilo guy, in the USA long enough to have business cards, told me that there are no undersized pumps in the USA. I have found him to have told me the truth.

Mark

 

If 1/2" pex needs 1.2 gpm minimum (2 fps) and there are 12 circuits total that means 14.4 gpm.

Or using the universal hydronics formula I get 36,000BTUH / 5000 = 7.2 gpm.

7.2 gpm would mean 0.6 gpm through each circuit of 1/2" pex. Too slow for the 2-4 fps range.

I'm feeling confused so I must be on the brink of learning something

 

Forget what you think 1/2" pex will do or take. FPS does not transfer heat.

Flow and GPM is what you care about. We will make it go through the piping as needed.

 

Ok so at 7.2 gpm and 15 feet of head a single Alpha pump isn't gonna do it.

I need to read up a bit and double check my math, make sure it's actually 15'.

 

So why don't you tell him what that number is a 5th grader would come up with, or you after 35 years.

 

No one says that 1/2" needs 2 fps of velocity, or 1.2 gpm minimum.

Your universal hydronic formula is based upon 10F delta T.
Also your heatpump at 32 F EWT will not make 36000 BTUs, it is more like 28-29000 BTUs.

Lets wait and see what number Mark comes up with....might be a while.

 

With 28,000 btu's and a 15F delta T I get 3.73 gpm.

At that flow rate the Alpha can do about 15' of head....

 

I don't need to give away a number. I am disabled/retired.

 

You don't want to have a 15F delta T with a geo system.
Again, geo is all about flow, keeping the delta T low. Fossil fuel is all about temperature (a flame as a minimum temp, sometimes even requiring mixing valves), not needing much flow. So imagine, coming from the old boiler world, applying the same principles to geo....Autsch.

Give me a moment here, we will get it straightened out.
Little hint, if all the zones are calling you should shoot for about 9 gpm of flow. The only key question is what goes into what zone zone.

 

Come on.... with your 35 years of experience you should be able to quickly do a 5th grader math!

 

15 degree delta T was more like me thinking out loud, I knew I didn't want it that high but I am seeing how the variables affect each other.

I had been thinking 9 gpm was about the right flow for delivery since that is the same rate the heat pump is producing it.

28,000 btu's at 9 gpm gives me a delta T of 6.22F.

Question on head calculating- On a manifold you only look at the longest loop. With the ladder piping is that treated like a manifold?

In other words am I just looking at the longest loop on the furthest manifold? Or longest loop on each manifold and add the three?

 

That is exactly the point. Otherwise you have the water short-circling through the buffer tank, and the heat pump will be cycling and not run full out.

You look at the loop with the higher pressured drop. All the others have less and don't matter. You might have to restrict flow depending on your flow rate, but you obviously don't want to do it on the circuit which has the highest pressure drop already. That is why balancing valves on each circuit is wrong.

BTW, a reverse return header only serves the function to add a certain flow restriction too to ensure that all the tabs are balanced.
But....here your entire radiant is unbalanced, so a reverse return header would not add anything, and adding a flow restrictor (balance valve) on each circuit would make things only worse as well.

 

Hands back the buffalo chip labeled flow.