Turbulent flow

Discussion in 'General Discussions' started by docjenser, Feb 10, 2011.

  1. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    This is getting scary, I thought about that, too. Which means I start to think like JOE:eek: Yep, more load on the pump, we will see. I will dig out my amp meter tomorrow. Stay tune...
     
  2. AMI Contracting

    AMI Contracting A nice Van Morrison song Industry Professional Forum Leader

    You'll be OK I promise:D
     
  3. Turbulence can be created under low flow rates: see AMASOND - Geothermal Energy and their concentric, corrugated pipes (out of the auto industry). I just returned from Eastern Europe and these are becoming commonplace and definitely extract more btu's per ft (m) than our smooth hdpe.

    I've also keep hearing of the "contact time needed to absorb heat" argument which is something that maybe should be discussed here...

    -Adam
     
  4. Looby

    Looby Member Forum Leader

    I see multicolor brochures with lots o' warm 'n fuzzy techno-babble,
    but not a performance specification to be found. My first impression
    is that it looks like a spendy way to install dozens of underground
    mechanical joints.
     
  5. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    I guess this is why a decrease in flow increases the delta T...:cool:
     
  6. I agree that they are violating the K.I.S.S. rule with all those joints...They do have some limited emperical data (not EWT/LWT or flow rates) to back up their analytic claims and even one install in Denver at http://www.amasond.com/uploaded-docs/Referenz en 10-10 Web.pdf Everything is in kW but conversion isn't too bad and the COPs in the high 4's--if true look pretty good (approaching an open loop hooked to a flowing artesian well)--which have to be coming from lower pumping costs.

    Now I wonder how much of the ground loop's "efficiency increase" is due to the concentric design where the LW is separated from EW and how much is due to the corrugation is questionable. Corrugation adds contact area which is an obvious reason for the heat transfer increase but that means more friction and pumping costs. Now my feeling is that any method of forcing the fluid flow out along the outer wall (why we rely so much on high Reynolds numbers) is going to give the most effect so I surmise the concentric design is most important.

    Now to the OP's question, the above isn't important unless he had some minerals build up inside on his smooth wall pipe, so I come back to the question of contact time since it isn't in any of our ground loop sizing equations but it's in all of my groundwater flow equations. Can time be much of a factor Looby?

    -Adam
     
  7. AMI Contracting

    AMI Contracting A nice Van Morrison song Industry Professional Forum Leader

    "Can time be much of a factor Looby?"
    I mentioned recently in another thread my tests and observations of loops over extended operations and noted that EWT is not necessarily reflective of the actual ground temperature though obviously impacted by it. Testing included a switch from heat to cool during which significant climbs were noticed though it would be unreasonable to assume we had significantly elevated the deep earth temp.
    It can take more than an hour for a system to stabilize (EWT level off) and few techs wish to spend that long during a maintenance test.
    So not only can time be a factor, but few take the time to evaluate it's impact.
    J
     
  8. Looby

    Looby Member Forum Leader

    Check the manufacturers' spec sheets. Decreasing water flow through
    the coax always reduces both HE and HC.

    The same is true for air coils: decreased cfm = reduced BTUH. Always.
    This has nothing to do with geo or heat pumps -- it's just basic physics:

    ................... Temperature is not heat.

    Color me highly skeptical. Notice that their 2-ton, 2400 sqft (7.3kw, 225 m^2)
    Colorado example claims an overall(?) COP of 4.4 while using an unspecified
    WaterFurnace heat pump. Please point me to any WF product that's capable
    (even theoretically) of a closed loop 4.4 COP -- after throwing in the circ pump
    and air handler of your choice.

    And I wonder how much of the "efficiency increase" is due to marketing fantasy.

    Specifications trump marketing hype. Show me the numbers!

    The concentric design can also be viewed as a highly efficient counterflow
    heat exchanger ...unfortunately, in this case, it's working against you. Also
    notice that the coax arrangement cuts the length of pipe in ground contact
    by half. For example, the AMASOND Colorado example uses 4 x 37m bores
    for a total of 485' of coax HX in the ground. However, a "conventional" design
    would have twice the length of pipe (970') in contact with the soil -- and it
    would seldom/never need 4 x 121' vertical bores to support a 2-ton unit.

    Yes. Increased contact time = reduced BTUH.

    ...temperature is not heat!

    Looby
     
    Last edited: Feb 20, 2011
  9. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Please check what I said!
    Whenever you slow down the flow, you increase the delta T.
    HE=delta t(F) x flow (gpm) x500 (for water) at 3 gpm in a 0.75 HDPE pipe (SDR-11).
    In the real world, you obviously cannot go to zero, and I do not question that if you get to low flow that you do not pay penalties in heat extraction, or that HE with extremely high flow cannot further increased. It is a saturated curve on both ends. But there is a wide spectrum where the curve is almost linear.
    I also believe that there is a difference wether you run an 800 ft slinky/ton or a 300ft straight pipe in a borehole/ton. I would assume that the 300 ft straight pipe is much more sensitive to turbulent flow given that it must extract much more heat/ft. I do realize that temperature is not heat, that is what this whole discussion here is all about.
     
  10. Looby

    Looby Member Forum Leader

    No, it is not. This has absolutely nothing to do with geo or heat pumps,
    (or Reynolds numbers, or turbulence, for that matter). It is the very nature
    of heat exchangers that (ignoring turbulence) the BTUH vs. gpm curve
    saturates only when flow becomes high enough to make (EWT - LWT)
    vanishingly small compared to the temperature difference between the
    brine and the external heat source/sink.

    Just examine the HE vs. gpm specs for any heat pump coax. There's a
    measurable slope to the curve, thus demonstrating that delta-T vs. gpm
    is non-linear. And the departure from linearity must be even greater in
    the loop piping -- where the brine-to-soil temperature difference is much
    smaller than the brine-to-refrigerant difference in the coax.

    Throwing Reynolds numbers into the pot only makes things less linear
    at the low-flow end of the curve -- such that BTUH vs. gpm falls off even
    more steeply as flow approaches zero.
     
  11. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Quite a surprise, at least to me. 1 pump running, 2nd one disconnected: one pump draws 0.79 amps at 232 volts.

    connected the second pump, now the 1st one draws 0.91 amps, second one now does 1.01 amps at 232 volts

    I am not a pump guy per se, but disconnecting one pump cut the power consumption in less than half. Kind of the opposite that what I would have expected. Anybody care to verify this with a 2 pump flowcenter?
     
  12. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    At the low end of the curve, with low flow, the water will have to much "contact time", heated up too much, and for the rest of the pipe lenght will not exchange much heat because it is already too warm, so heat extraction is limited (saturated). At the other end, HE is limited (saturated as well) where the water does not have enough contact time (e.g. the pipe is too short for the flow), and just pushing more water through does not increase the heat extraction.
     
  13. Looby

    Looby Member Forum Leader

    OK, call it "contact time" if you like, but the real problem is that
    reduced gpm allows the brine temperature to get closer to the
    source/sink temperature -- thus lowering the T-gradient across
    the pipe wall, and reducing heat flux proportionately.

    That is just plain wrong. There's no such thing as "not enough
    contact time." As contact time approaches zero, heat transfer
    rate must increase. There's no maxima in the BTUH vs. gpm
    curve; it asymptotically approaches a horizontal line as flow
    approaches infinity -- but it never peaks.

    The term "contact time" carries the implied assumption that
    the heat exchange process "cares about" how long it takes
    an individual water molecule to traverse the heat exchanger.

    ...It doesn't! (Because heat is not temperature.)

    Looby
     
  14. dgbair

    dgbair Just a hobby Forum Leader

    I have a switch/relay on my second pump.... with one pump running I'm seeing .71 amps and with two running I'm seeing 1.64 amps. All at 236.5 volts.

    I did not expect this neither... but could this be the same effect as those cycle stop values claim to have on well pumps???

    Update: I opened up a few more boxes.... seems like my second pump just draws more current then the first pump. Turning on/off the second pump doesn't seem to have any effect on the current draw of the first pump.
     
    Last edited: Feb 20, 2011
  15. dgbair

    dgbair Just a hobby Forum Leader

    Amasond...

    hmmm corrugated pipe, would definitely have a positive impact on the amount of turbulence you have in the pipe.

    my 2 cents....

    You really want the turbulence to mix the brine as it's flowing through the pipe. This will ensure the brine at the center of the pipe is at the same temperature as the brine running against the edges of the pipe. Without any turbulence, it would be like having many small pipes running inside the 3/4" pipe and the inner brine would need to transfer heat across a 'large' distance before reaching the outside pipe.

    So the question is do the slinky loops cause the brine to get mixed? Looking at the loops it's really made up of two parts... the loops and the 'straight' return runs. Even if the slinky part is mixing the brine sufficiently.... I'm pretty sure the 'straight' part of the run isn't doing a great job at transferring heat at a low Reynolds number.

    I think what we need is some clear slinky loops and some dye. :D Nothing like being able to see the brine being mixed as it travels down the pipe.

     
  16. Amasond was doing 120' per ton in Austria...

    The way too-complicated Kelix "thermocouples" are doing the same concentric pipe thing by providing laminar flow in the inner pipe with exterior "fins" and are able to get 12,000 BTU/Hr per 50' of borehole: Thermacouple System Operation

    Now with a 3.5" outside diameter pipe you also get a lot of surface contact with the ground per ft of borehole.
     
  17. engineer

    engineer Well-Known Member Industry Professional Forum Leader

    As to the pump issue, the more flow through a single pump, the more power it draws. This may seem counter-intuitive.

    To see an analogy, block the inlet or outlet of a shop-vac. The noise rises in pitch - motor is spinning faster since the impeller is unloaded - it no longer has passing air to "dig in to"

    Some pumps and blowers rated for high pressure must NOT be operated in open flow / free air or else they'll overload and overheat their motors.
     
  18. AMI Contracting

    AMI Contracting A nice Van Morrison song Industry Professional Forum Leader

    Interesting on the pump power (good point Curt). Next question Doc (this is probably answered by the welserver).......how but compressor amps? Is there an impact with the higher DT.
    j
     
  19. Looby

    Looby Member Forum Leader

  20. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    No, Amps, COP and HE did not change, at least not within the margin of error, which was 0.1F on the temperature side (delta T).
     
    Last edited: Feb 21, 2011

Share This Page