1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Compressed air aftercooler questions

  1. Jan 31, 2010 #1
    I am the head electrician at Panorama, a ski-resort in SE BC. It is my job to identify & take action on energy improvement measures… This usually doesn’t involve consulting physicists, but now I have something beyond my experience.

    Our snow-making plant uses 3 x 800 HP centrifugal air-compressors to feed the mountain the compressed air we use to make snow. The system was originally designed in the mid 1980s, and the air-system has an aftercooler located immediately next to the compressor building. From there is approx total 12 miles of piping that runs up the mountain to an high-point elevation approx 3500’ above the compressor building. The old aftercooler is an air-air type, where the discharge air runs through a cooling tower with a double-walled pipe. The inside pipe contains the moisture laden discharge air, and the outside pipe is simply a conduit for blowing the ambient outside air (typically -10 to -25 DegC). Moisture that condenses inside the discharge air inner pipe, runs down to a header with a drain. The maximum discharge rate is 10,000 cfm at 130 psi.

    The commercial compressed-air aftercoolers of this scale that are presently on the market are both expensive & energy hungry.

    Here is an overall general primer on snowmaking systems
    http://peakstoprairies.org/p2bande/skigreen/Ch%2011%20Snowmaking.pdf" [Broken]

    Moisture in the discharge air is bad for a number of reasons.
    - Water vapour inhibits the super-cooling effect of the air at the snow-gun nozzle as it emerges & expands. The super-cooling effect is very desirable for snow-making.
    - Water vapour condenses in significant quantities inside the 12 miles of underground mountain piping, and gives us operational problems
    - Pushing the unwanted water vapour up the mountain pipe-plant obviously is consuming energy, and is considered a loss.

    Colder discharge air is better because of the higher density, it takes less of it (by volume) to achieve the same atomizing effect at the snow-gun nozzles

    I want to make a significant improvement in the operation of our existing aftercooler, which shouldn’t be too hard, since the existing air-air system is very ineffective. My plan is to simply remove the blower from the outside air jacket (a sealed pipe, sch. 40), and convert to a circulating chilling-water outer jacket. I have a big cold creek right next to the plant (also a pumphouse), so lots of very cold water (+1 DegC) is available. The drained off chilling water (now warmer) will simply be discharged back to the creek.

    Questions :
    1) What ratio of aftercooler performance improvement can I expect by changing from a air-air heat exchanger, to an air-water heat exchanger. Or, what method would I best use to approach the problem?
    2) By removing additional water vapour, I can now use my 2400 HP of compressor power to push more air (now dryer) up the mountain. How can I calculate “how much more”?
    3) The cooler air will be more dense & effective at generating water atomization at the snow-gun nozzles. How can I estimate how much?

    This job has the prospect of significant energy efficiency gains (perhaps 100-300 MWh annually). I need to answer the above questions before I can approach Sr. management. Only then can I hire an engineer for the job. At that point I’ll begin looking for a good air engineer (that isn’t a salesman)

    MThornton
    "www.panoramaresort.com"[/URL]
     
    Last edited by a moderator: May 4, 2017
  2. jcsd
  3. Feb 2, 2010 #2

    Lok

    User Avatar

    So you wish to change the -10 -25'C cool air for a 1'C cool water.

    Best water cooling systems use brine that freezes only at -15'C. Closed systems don't need much of water replenishment, and can be cooled of in a creek.

    What is the temperature of the compressed air after it passes through the aftercooler?
    Depending on the amount of cool air that is blown through the outside pipe, -10'C of air might be better that 1'C of water.

    What is the surface of the aftercooler? If improving that might make a better difference.

    Are there enviromental effects to the heated water you will supply back into the pond? You might evaporate the creek :P.

    Dimensions are very important (usually necessary for any estimate), but usually a water cooler is smaller so it will fit, but water is heavy so you have to take into account the extra weight ( might even be around an extra tonne). Air pipes are not engineered to carry that much weight (no need no bother).

    Water is corrosive, even at 1'C, even more so if you take it from a creek. Again air systems do not bother so much about corrosion. (From creek acidity and dust abrasion).
    A redesign is possible and might even be cheap, but it has to be very well thought.

    These are mostly the bad things that have to be taken into account, as you already know the benefits.
     
  4. Feb 2, 2010 #3
    Thanks for your reply
    The typical discharge air from the compressors (before the aftercooler) is 10-15 DegC
    The existing aftercooler arrangement drops the discharge air temperature only a couple of degreesC , and we would like it colder, but the primary concern is that it isn't very effective at removing the suspended H2O, which I understand remains as small droplets of condensed liquid water, suspended in the high-velocity air-stream.

    The existing aftercooler structure is made from sched-40 steel pipe. I may be able to take a photo of the structure tomorrow. If we went this route we would hire a mechanical engineer, but at this point I am only preparing feasibility information for a energy conservation proposal. Once the project is funded, then I have money for an engineer, who will make sure the structure is suitable.

    Present air volume is 10,000 cfm, with design capacity at 12,500, so it's not that small. The present aftercooler structure is ~12m tall, 15m long. I'll mark up the photo with actual dimensions.

    Regarding the warmed discharge water from the after-cooler, we would likely pipe this to our pumping intake gate, where it would be used to manage ice at the intake gallery. The creek is a large, (actually somewhat larger in flow than the still small Columbia River before they converge 20km downstream).

    M
     
  5. Feb 3, 2010 #4

    Lok

    User Avatar

    If the temp of the compressed air is only 10-15'C, water will be better, also look into ways of maximizing the heat exchange surface, a bit of piping and aluminum radiators will do.

    The 12m high installation will present some pumping problems as at this height for water a vacuum can appear, or take a chunk of energy to get that high if the pump is on the ground level. One theoretically simple way (practitioners will laugh) to escape this is to lay the installation almost flat to the ground (if at all possible) as water has to drip at an angle.

    The released water won't be that hot so glad to hear that the creek will be fine.
    And do not forget about corrosion, a good paint job can save millions.
     
  6. Feb 3, 2010 #5
    Thanks again for the replies
    Whenever we are running these air compressors we are also running high-volume water pumps up to 1750 psi (it's a mountain). So no problem getting water to the top of a 12m aftercooler.
    We can feed the aftercooler outer jacket via a PRV, with an outlet ball-valve for flow-control. Probably minimal flow requirements (I have lots of free water +1C)

    I am intrigued by use of brine to -15, but the corrosion issue is significant, and a spill would be toxic to the fish. If we stick with straight water, the corrosion will simply typical for what we are already dealing with, and it allows an open-loop design.
     
  7. Feb 7, 2010 #6
    Is there any documentation from the original installation? Whoever designed the air to air aftercooler might have also considered the nearby creek and made a decision against using a water-cooled cooler. Or maybe such considerations appear in the original bids, and someone picked the bid with the air cooler... What might have been the right decision in 1985 or whenever might not look so smart today (different electric rates, etc).

    If you don't have any of that old paperwork, maybe the original contractor / vendor does. It may be worth looking for, if nothing else you may gain some insight into why it was built the way it was.

    On another note, it seems to me like the long uphill pipe is a 'natural' aftercooler. Is the compressed air at the top of the hill cooled to ambient temp? If so, the advantages you list for drier air at the end won't really come into play (though your assessment of pumping that water up the hill seems like a very reasonable thought).
     
  8. Feb 7, 2010 #7
    Attached is a photo taken of the after-cooler in discussion. I've added labels

    Yes there are prints (paper only). The system was designed by an engineering firm that is no longer around.

    Over the past week I've come to believe that this unit was designed more for stripping out water droplets suspended in the high-velocity air-stream. There are only 2 cooler/stripper tubes installed because the original plant had only 2 x 3350 cfm compressors, but the final plan was for 4. We have 3 in place now, and we will never add a 4th.

    When the 3rd compressor was added (1994), cooler tube #3 was not added as the belief (then) was that this was a waste of money. This was based solely on the subjective judgment of the final discharge air temperature, and did not account for efficiency gains due to stripping out the suspended water.

    However, instrumentation was (& still is) incomplete, as there is only one temperature sensor (visible just above the insulation at final discharge). I'll add another thermowell & sensor to the air-in pipe next time we have a welder on-site.

    Q1 : If I add in the 2 missing cooler/stripper tubes, will the internal pressure remain constant & just the velocity decrease? Or, will the pressure drop & the velocity remain constant. Or both? Apologies in advance for my basic deficit in thermodynamics understanding.
    Q2: I ask because obviously I would prefer the velocity to decrease as much as possible without the suspended water droplets vaporizing. If the H2O remains liquid perhaps it has a better chance of falling out in the lower velocity air-stream. Am I right?

    We want as little liquid water up the mountain pipes for a number of reasons. A big one is that it collects in low-points when we shut-down & then gushes about in great mass when we start back up. The vertical relief of this system is 3500' 12 mi long, with a number of dips. Water/air-locks can form & are time consuming to vent & purge.

    This coming week I write up the feasibility study, and submit it. Based on concept approval, we'll get funding for an engineering study. But in the meantime, I need to know enough about this thing so I don't come across completely ignorant. (I'm the electrician)

    Thanks again to Lok & gmax137 for their responses... very helpful.

    M
     
  9. Feb 8, 2010 #8
    Looks like you forgot to make the attachment?

    I am familiar with much (much) smaller capacity compressed air systems. More like 500 scfm. In those you typically see an aftercooler followed by a cyclone separator, and then a receiver. The receiver dampens variation in demand, and may also allow for cyclic control of the compressors (loading & unloading). In you case, I imagine the system just runs at full blast or not at all. But, another important function of the receiver is to allow remaining liquid drops to fall out (due to the low flow velocity, as you suggested). Maybe you can consider adding a large tank with drain traps (or, is that what your aftercooler really is?).

    Do the dips (low spots) in the piping going up the mountain have drain traps? This would be a way to remove the liquid to prevent accumulation. Also, does the condensate freeze up there? Seems like that could be a big problem, and cause flow blockage/increased pressure losses.

    It is an interesting design problem.
     
  10. Feb 10, 2010 #9
    here it is again
     

    Attached Files:

  11. Feb 10, 2010 #10

    Lok

    User Avatar

    Looks old but functional enough.

    Adding the #3 and #4 pipes would double the surface and cut the flow in half (which means air takes twice the time in the cooler, very good). Plus that adding a better blower (this one seems of the centrifugal type) would increase performance and cut power costs.

    It is a tall thing, but it was designed to hold 4 pipes. which means the supports can easily carry the water jacket by weight with 2 pipes.(no ideea if the pipes wil hold, so get an engineer on site).

    Btw. with water there will be some freezing problems which implies emptying every time the compressor and cooler shuts off (it is quite a bother to empty the pump and pipes).
    The creek pipes might get ice plugs over night, as i suppose there are shutdown times. So an relatively simple temporary heating mechanism, like an exhaust pipe diverged from the engines, has to be installed.
     
  12. Feb 10, 2010 #11
    I think Lok is right on target. It might be worthwhile to take some temperature measurements (compressed air temperature in & out) to see how close you're getting to ambient. Try to do that with one and then two compressors running.
     
  13. Feb 10, 2010 #12

    Lok

    User Avatar

    One ideea is to run the pipe of the compressed air through the creek, that implies no water pumping mechanism.

    The condensed water can be evacuated from a small tank situated lower than the pipe through a small outlet pipe with a faucet (you get the idea) as the pressure will force it outside anyway. The compressed air pipe inside the creek sholud have an incline according to the speed of air flow, so that the water dropets don't get carried away by the air flow. A zigzag pipe might be a solution.
     
  14. Mar 7, 2010 #13
    An air to air intercooler is just as good as an air to water intercooler.

    Making the comparison to a supercharged, or turbocharged car...
    The air coming out of the compressor is at a high pressure and the temperature is higher.
    The air to water has an advantage in a car because you have the additional thermal mass of the water, and the car is accelerating and decelerating so that thermal mass is helpful.
    But a ski hill does runs at steady state, so air to air is ideal.

    You would probably want to have an initial cooler that brought the temperature down to ~0 C and trap water, then further cooling once the majority of the water is removed.
    Maybe an air to water system here would be wise.

    The air flow for snow making is pretty high. The water trap should ideally be pretty large in size so that the flow is slowed and the water can be condensed and not just blown up the pipe.

    In any case I would doubt that you would be seeing significant improvements in energy savings.

    I'll look at the photos now...
     
  15. Mar 7, 2010 #14
    What sort of blower is in this system?
    What sort of current and Watts pr HP does it draw?
    I would suspect that it is not a whole lot compared to the 3x800 HP compressors.


     
    Last edited by a moderator: Apr 24, 2017
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Compressed air aftercooler questions
  1. Compressing Air. (Replies: 2)

  2. Compressed air can (Replies: 12)

Loading...