Expansion Tank for GSHP, is this necessary

[riscy]

I have been doing some research about the expansion tank sizing, antifreeze and so on. I experimented expansion equation from the ASHRAE and product data from various anti-freeze datasheets.

Vt = Vs((V2/V1)-1) - 3 ? ?t) / (1-(p1/p2))

ASHRAE demonstrated equation that account for both expansion coeff due to water *and* due to vessel tank (mostly steel). I found it is necessary to have expansion tank for HVAC and DHW within the building.

The expansion coeff for steel tank is 11.7mm/m.K

==================================================GSHP

I now looking into ground loop circuit where GSHP is connected. I need to work out how big the expansion tank needed for the project.

I knew brine expands 3-4 time faster than water (20% Ethylene Gycol + 80% water), this would lead to bigger expansion tank and thus an increase cost and installation space issue.

I worked out the expansion coeff of PE100 which is 0.150 mm/m.K (linear), this is many time faster than steel tank.

I don't know what is the linear expansion of the water, most of them are volumetric (~250e-6). Any suggestion?

Based on guideline document from website (non ASHRAE), popular equation occurs mostly for solar project:- Vt =Vv(B / (1-p1/p2), where B is expansion factor.

One source suggested B=0.13/K, but that apply to solar, not for GSHP.

I have acquired the volumetric coeff of the Ethylene Gycol from ASHRAE table (Fundamental Chap 22) based on 20% mix with 80% water.

-5 0.000 964 m/mK I used V=1/density conversion.
35 0.000 977 m/mk

With ASHRAE equation, I then calculated with PE100 pipe, and discovered that the PE100 expansion is actually faster than brine(!). Based on this, it seem to suggest that drop of pressure as you increase the water temperature (during cooling HVAC operation).

This calculation suggest I'm wasting the money with expansion tank as it does not do much. The PE100 is doing the works for you.

In addition, the water within pipe, tank, valve as part of HVAC system expands as well as expansion tank, this make sizing less critual.

On further thought, if you use plastic pipe within the DHW or heating network, it seem they expands faster than water and copper pipes, hence you may not need expansion tank.

Am i missing something, it too good to be true....

Comment welcome...

 

[riscy]

Correction
==========
Reviewing the calculation, the 20% brine actually expands (volumetric) two time faster than pure water at same temperature range, not 3-4 as quoted.

 

[GMcD]

Cut to the chase- you're overthinking this, keep it simple and ignore the expansion capabilities of the piping system whether it's plastic or steel. Put in an expansion tank on the geo side - it's not only good practice, but good insurance for the usual "what-if" scenarios. You can design the temperature differences down to the second decimal place, but in the real world, you have to design in an "envleope of performance" and system tolerances to account for not everything being installed perfectly.

In a geo-exchange system, the biggest variable is the soil conductivity, and therefore your geo loop temperatures, so whatever you assumed you'd be getting won't be the case once the system is operating.

 

[AbbyNormal]

I guess it depends on where the system is.

A static charge of about 30 psi tends to increase in winter as pipes shrink, then drops down to about 10 psi in summer when pipes expand.

You want the expansion tank to make sure pumps do not cavitate in the summer?



Take the "V" out of HVAC and you are left with a HAC(k) job.

 

[sterl]

If your in-ground grid is going to be operating at fluid temp of minus 5 C. occasionally and you are thinking the PEX is going to expand: Consider that some of it will at least short term be surrounded by frozen ground; and that its always going to have to push ground around in order to expand. If that ground is anything but really loose sand: the buried PEX isn't going to expand. Similarly for pressure swing. Ignoring that for the moment:

Unless this is a huge arrangement, or your pressure changes are huge: Size the tank draw down for the volume of glycol going through a temperature swing a little greater than the design, treating the system containment as rigid, and ignore thermal expansion of steel and so on....and use Specific Gravity, not density to chase it so your units are Vol to Vol. SO: -5 C: Sg =1.037; 35 C: Sg=1.023; Difference over 40-deg: .015 or 1.5%. Over 60 Deg: 2.25%. With a 10 Cu. Meter Charge: You will need draw down volume of 0.225 Cu. Meter. That 10 cubic meters is over 10,000 Kg of mixture!

 

[riscy]

I have not have real data on how the temperature varies during the winter so I took numbers from the textbooks.

AbbyNormal:- Just curious, have you made the measurement, could I have more details, have you fitted the expansion tank?

sterl:- The pipe will contracts when ground get colder and will be colder when taking heat out from the fluid during the heating operation, so it will also contracts, so I expects the pressure goes up because the pipe contracts faster than fluid. During summer, the pipe expands as the fluid is hotter, the water expands, but pressure drop.

I need to know how to balance the decision for sizing of the expansion tank......it save cost and maintenance cost as well. I have no intention to make it ultra cheap and ultra dirty. Also I have been very though about soil conductivity....

All I saying the expansion tank equation does not work with PE100 pipes because it expand 10 time faster than copper and steel pipe. It need to be factored in in order to design the tank correctly.

As for project, I have 500 Litre fluid for 2500m horz slinky pipe (10 x 20m trenches), this is for 20KW heating and cooling.

 

[AbbyNormal]

rsicy, I am just making comments on what I observed on quite a few residential systems installed. These were primarily horizontal loops, typically systems were started up in the fall, no expansion tanks



Take the "V" out of HVAC and you are left with a HAC(k) job.