The Homemade Heat Pump Manifesto

[superlen]

Steve,

I haven't checked the Reynolds number, but was assuming that it was not close to turbulent. I'll double check. FYI, the 15gpm was just a swag at maximum flow I might need to give me a max pumping cost to expect.

(Edit - using this calculator, I'm showing well into turbulent flow even for loops! That's how handy assumptions are. - in this case..wrong) http://www.gcisolutions.com/flow.html I would need to drop the loop gpm to less than .5gpm to get back into laminar. I'm thinking that calculating pumping costs for turbulent flow is more math to figure out.


No tight u-bends at all. It's out and back in 2' wide trenches.

The other thought I had was to scrap the 1.25" supply/return lines and instead just extend each 3/4" loop line the 400' (200' each way) needed to reach the house. This would give me an additional 2000' ft of pipe in the ground which would be awesome for the heat load, but cranks up the daily cost to $0.37/day (again at max 3gpm/loop which would NOT be that high with that much pipe in the dirt).

It's also 5 times the digging to go that route & as I mentioned above, my field/yard is starting to resemble a small city with the amount of underground widgits running in it. I would like to have all five loop lines popping up in my basement for experimenting, but I think I'll probably just run 1.25" supply/return as before.

Len

[AC_Hacker]

Originally Posted by stevehull

Len,

15 gpm is a lot of water to move through the pipe. Did you check Reynold's number to see if you are approaching turbulence? You will not in the parallel loops, but you may well be in the supply & return pipes.

Remember, those tables/calculators generally assume laminar flow and turbulence takes a lot more energy to push a fluid.

That said, if you remain in the laminar range, then I believe your numbers are spot on.

Love Kirkoff's law . . . . . .

SH,

Are you implying that Kirkoff's law pertains to turbulent flow?

Also, if the purpose of the loop field is to operate with minimum power, then laminar flow would be the way to go.

However if instead, the purpose of the loop field is to transfer heat from the ground to the thermal transfer medium, and ultimately the heat pump, then turbulent flow should be designed-in to the system.

From what you have written, you seem to be advising Len to design for laminar flow. Any engineer who works with thermal transfer knows the advantages of turbulent flow.

Here's a page that explains what's going on, in simple non-technical terms, and why turbulent flow should be used for thermal transfer.

-AC

[superlen]

Steve,

I'm still a little confused on the power calcs as it pertains to turbulent vs laminar. The equation I used (or tables supplied by mnf which match my numbers) give out head loss vs flow rate. So aren't those head loss numbers already taking into consideration laminar vs turbulent? If so, then the equations would be valid no mater what range I'm in, wouldn't they? After all, once I have head loss, then ft-lbs is ft-lbs is ft-lbs.

AC,

I agree that the turbulent flow will transfer more heat. I'm struggling with the thought that is stuck in my head that a lot of one's potential COP is wasted with pumping power. I've seen that sentiment posted before and I have always assumed that pumping cost was a major cost of the energy in a geo-thermal. Obviously not as large as compressor amps, but still quite high. Using my example above, it doesn't seem to be as much as I suspected, so I think that minimizing loop pumping power, while certainly a design goal, isn't a great worry. Particularly if you drop below turbulent and tank your thermal transfer.

From the online calculator I checked, it looks quite difficult to get laminar flow with the typical ground loop sizes and flow rates I see used. The flowrate would need to be so low, that hardly any heat would be moved.

Len

[stevehull]

AC - you make us laugh!

The take home point is that you cannot have turbulent flows in any of the distribution fields as Len has described. I am wondering about just his supply and return pipes. In fact, I doubt if there ever is anything but laminar in vertical loops as well.

But that is my perspective based on some years of education, field and laboratory work.

You may want to consider the inlet diameter ratio where pipe velocity decreases in a parallel pipe network. In general, at about ten diameters downstream, turbulent flow becomes laminar. This is an accepted rule of thumb (is still on the EIT PE test) and is certainly accepted in straight pipes. A "loopy" horizontal field is still going to look pretty much like a straight pipe from a hydraulic perspective as the ratio of the loop radius to pipe diameter is so large.

This wound mean that all our loop fields (vertical or horizontal are likely laminar and would be very poor at transferring BTUs (your opinion).

Oh my - another deletion from the thread!!

We can continue to have fun with this - and expand the number of replies to the manifesto - or we can have more salient discussions . . . .


Steve
aka SH

[AC_Hacker]

SH,

I would much rather have this discussion with someone more intelligent.

Things being what they are, this will have to do.

The above photograph is of the cover of the International Ground Source Heat Pump Manual, often referred to as "The Basic Manual." I doubt that you ever heard if it, and I am certain that you never read it, else you would be giving good advice, rather than the advice you do give.

The book is prepared and published by the Oklahoma State University Division of Engineering Technology, Stillwater, OK 74078. I think you have heard of that school.

The book is a compendium of information that has been distilled from theoretical studies and real world application that date back more than fifty years.

The book actually has the expert information that you claim to have.

For example, I would like you to look at the photograph below that is from page 71 of the manual.

Originally Posted by stevehull

AC - you make us laugh!

Who is the laughing "us" you are referring to?

-AC

[superlen]

Hold on Steve.. Don't drag me into this.

I think my horizontal loops ARE turbulent. According to the online calculator I used, my 3/4" 3gpm are highly turbulent. I'm all ears if there is evidence to say otherwise.

My loops are not slinky, btw. They are 150' out, one 12" radius 180 deg turn and 150' back in the same 2' wide trench.

Len

[AC_Hacker]

Len,

Here's a nifty graph regarding turbulence & thermal transfer:

...and here is the text to go with:

Heat Transfer Enhancement
Convective heat transfer coefficient increases drastically when the flow becomes turbulent, due to effective mixing of different fluid layers in the flow. This behaviour is shown in following figure.So it is a common practice among designers to covert laminar flows into turbulent by introducing suitable vortex generators in the flow.

Also another especially juicy bit:

...and the text to go with it:

Drag reduction
Coefficient of drag around a body reduces by a huge amount when flow changes from laminar to turbulent.This phenomenon is shown in following figure.This is the reason why golf ball has got lot of dimples on it.This irregularities on surface of the ball will help in transforming laminar flow into turbulent and reduces drag, with low drag ball can travel more distance.

The take away from this is that there is a small 'sweet spot' after turbulence is achieved, when there is a dramatic drop in fluid friction.

The big news is, at this sweet spot, there is both a 'drastic' increase in thermal transfer, and a 'huge amount' of fluid friction reduction.

Well worth working toward.

Fluid velocities beyond this sweet spot result in increasing fluid friction by the square of fluid velocity.

( above information from: What is Turbulence ? ~ Learn Engineering )

Best,

-AC

[superlen]

AC,

Thanks a ton for those two pics. I noted much earlier in this thread the discussions on the "turbulent sweet spot" & tabled it as I didn't quite wrap my brain about how one could get a little bit of the best of both worlds. I was planning on doing some more research on it, but was striking out in the google department.

It seems one could do a real world test with some of the new variable speed pumps by monitoring GPM & instantaneous power. Slowly bring the speed up and map the GPM vs Power. At that spot where the drag coefficient takes a dive, we should see a corresponding dip in the power required. It of course should be somewhere around the theoretical GPM to put us at that laminar/turbulent boundary. That may be a fairly wide range as I have seen Reynolds numbers anywhere from 2000-4000 as the boundary.

Given a good solid consistent EWT and measure the LWT and we could also plot the heat rejection at the same time & learn something about how sensitive heat transfer is around that boundary. I think that the increased transfer from the water to the tube due to the turbulence may be overshadowed by the extremely slow transfer from the plastic to/through the dirt. As you have mentioned & I like the analogy...molasses slow transfer into the dirt.


Len

[stevehull]

Len, - here is an easy way to tell if you have turbulent flow. Set up a rig with the flow rate and tube length you are using. Turn the water on. Put an inexpensive stethoscope on the pipe and listen - no noise laminar, noise (whooshing) then means turbulent. You only need 20-30 feet of pipe. We used this in engineering classes where we would look at clear plastic tubes with an air/water mix coming in (bubbles allow you to visualize turbulence). Fascinating how at about 10-20 diameters downstream, the fluid transitions from turbulent to laminar and the noise (via stethoscope) goes away.

But back to your point on energy use. I fully appreciate the importance of lowering a pump's electrical needs. Here I pay a lot for my pump and dump open loop.

Steve

[MEMPHIS91]

Originally Posted by stevehull

Here I pay a lot for my pump and dump open loop.

Steve

*OFF TOPIC- EXCUSE MY RUDENESS FOR ONE MOMENT PLEASE*
Do you have picture/details of this unit and set-up?