Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Fill time to compress an air tank

  1. Nov 16, 2013 #1
    Hello friends,

    I'm trying to calculate the fill time to compress an air tank. I currently have the compressor's flow rate[CFM] at various pressures[PSI].

    I honestly thought it was a trivial endeavor until I started gathering all the critical elements.

    So, there's obviously flow rate changing as we increase the pressure in the tank. There's going to be a choking case for given time. There's the temperature increase as we pressurize. How can we take all these cases and come up with actual time taken.

    It's a fun challenge, but I am loosing my hair trying to address all the changes that take place.

    It's easy to just run the system and calculate the time, but where's the fun in that? In fact, I would actually like to be able to predict how the time would change for various ambient temperatures.

    Any wisdom, pointers, directions are welcome.
  2. jcsd
  3. Nov 16, 2013 #2
    Actually, I don't think choking will occur. Someone please confirm this.

    As I have the flow rate vs pressure information from the compressor, this just tells me that when the air tank pressure is at ambient, the flow rate is just higher, and when the pressure increases in the tank, the flow rate decreases.

    So, there really should be no choking at any point, correct? I misunderstood the choking condition that exists within a compressor and applied it to this case.
  4. Nov 18, 2013 #3


    User Avatar
    Science Advisor

    Why make it all so difficult ?
    You can ignore temperature effects because cooling can be faster than filling.

    You know your target pressure in atmospheres. You know your tank volume.
    Multiply them and divide by the worst case cfm for a maximum time to fill.

    Alternatively, convert the psi to cfm pump efficiency into an equation and integrate the result.
  5. Nov 18, 2013 #4
    I doubt you can ignore the effects of temperature when you're looking at CFM's of 1.2. The density of air will be effected as result, and in turn effecting the pressure.

    When one has the psi to cfm chart, I don't know why use the worst case as the fill time, since that won't be accurate or reasonably acceptable.
  6. Nov 18, 2013 #5


    User Avatar
    Science Advisor

    This is an engineering forum. Without some pressure limiter that cuts out the motor when target pressure is reached, as well as a safety valve, the system will be unsafe.

    “when you're looking at CFM's of 1.2” Where in the OP was 1.2 cfm specified?
    Must I now assume the pressure vessel has a capacity between 1 pint and 1 cubic mile?

    You have not provided the psi to cfm chart. Maybe the chart is simply computed from a simple pump compression ratio. We have no idea how much information may available. The solution will be based on that unknown information.

    An air compressor usually has a cooling coil immediately between the pump and the reservoir. That is usually called an intercooler when installed after a turbo/super charger on motor vehicles. There is no point pushing hot air into a tank. It will take up to 50% more energy not to immediately cool the compressed air. The thermal properties of the pressure vessel will determine the cooling time, so do not insulate the tank.

    So long as you want to make it more difficult you will encounter obstacles to a solution. This forum is not a poker game where you can hide all your cards against your chest. It is time to put all your cards on the table to be seen.
  7. Nov 18, 2013 #6
    Well Baluncore, I apologize for not providing more info. Just ask, no worries. I appreciate your and others' time.

    So, to be clear: I have a compressor which operates at 1.30 CFM at 0 psi, and decreases with a polynomial relationship to 0.80 CFM at 300 psi (This relation is hard to put in an equation, as even polynomial orders of 7 and up didn't fit). Moving further, the compressor is rated at 300 psi, and is a two stage design with an integrated dryer. The operating temperature range listed for the compressor is between -40 to 80 deg C.

    I understand the certain compressors have intercooling, but can't be sure of this one, since none has been listed. Perhaps if you have more knowledge of a two stage compressor you can elaborate.

    The volume of tank in question is 1.5 gallons.

    Was I wrong initially to put value in temperature increase?
  8. Nov 18, 2013 #7
    And, I forgot to say that I do indeed have a pressure regulator capped at 300 psi.
  9. Nov 18, 2013 #8


    User Avatar
    Science Advisor

    Scuba diving tanks have a copper pipe after the compressor to radiate heat and so cool the compressed air. You should do the same. Then you may be able to ignore the temperature of the air.

    The operating temperature of the compressor should not be relevant. It is more to do with the rubber seals and lubrication.

    A change from 1.3 to 0.8 cfm @ 300 psi is almost no change at all. You could just use 1.0 cfm as a good estimate. When you say that even a 7th order polynomial does not fit I don't believe you. Any and every order polynomial will fit, with an r^2 or RMS error specified. Where is the data table? Please give the compressor make, model, and a link to the specifications on the manufacturers website.

    What is your target pressure?
    What type of gallons are you using US or 4.546 litre ones? What is 1.5 gall in cubic feet?
  10. Nov 18, 2013 #9
    I like the copper pipe idea. Is it safe to assume that the condensate formed due to temperature drop through the copper pipe would be negligible since the compressor has an integrated dryer? If not, how to keep the moisture out? Another desiccant type moisture collector?

    You're right about the the 7th order polynomial, my MATLAB code was not written well. Did a quick check on Excel, seems like lower orders do just as well.

    Compressor information here: http://tinyurl.com/nxyl6vz

    Target pressure is 300 psi, and I'm using US Gallons, so in cubic feet it's 0.2. There is a volume fill time for 1 gallon on the website as a reference. I currently have my fill time at 166 seconds for 300 psi. I'd appreciate if you can confirm.
  11. Nov 18, 2013 #10


    User Avatar
    Science Advisor

    A quick estimate...
    maximum (300 / 14.5) * (0.2 / 0.85) = 4.87 minutes = 292 seconds
    typical (300 / 14.5) * (0.2 / 1.00) = 4.15 minutes = 250 seconds
    minimum (300 / 14.5) * (0.2 / 1.27) = 3.26 minutes = 195 seconds

    Something is very wrong with that data table.
  12. Sep 28, 2015 #11
    Well, I'm no physicist, so I'm not nearly as smart as you fellows, but here's some homespun wisdom from down beneath the decks where we actually build and run real compressors for pay and fun. We get our hands greasy, we don't have fancy degrees, and we are mostly not even lowly engineering types, etc. :-) We do on occasion support the pumping of a hell of a lot of reasonably dry and pure compressed air, though.

    Two-stage compressors invariably have an intercooler between stage 1 and stage 2. (It's that finned pipe between the exhaust of 1 and the intake of 2). Typically it wraps around by the fan/flywheel on your reciprocating pump so that cooling air blows across those fins. Fins help a lot. :-) yes, it's copper. Now THATS your intercooler...boys and girls, this is not a turbocharger setup on your Subaru. "INTER" = between stages of compression. This helps stage one ram a more dense charge of air into stage two of your "two stage compressor"...got it?

    Well designed 2 stage compressors then ALSO have an after-cooler. That's a second heat exchanger (which drops the temperature of the full pressure air charge as close to ambient or below, as much as your wallet can afford). It does this by dumping the heat back to the ambient surroundings. This is pretty critical because the air leaving the exhaust at stage two is pretty hot - say about 130-150F at 100PSI if ambient was about 80f or so... The reason this is so critical is more than just pure efficiency, (though that is important.) The hidden reason is critical because aftercooling the compressed charge will condense (some of) the accumulated moisture out of that air. You really REALLY want to do this before your air goes to your accumulator. If you don't all that water will quickly fill up your accumulator, reduce total air volume, rust out your tank, pollute your downstream devices with moisture+oil+rust laden air and generally make your life miserable. This is bad. Shorty gets a call from engineering types if this water shows up above decks. Bad bad bad. :-/

    Now a decent after cooler can easily drop the temp from 135F to ambient, which will condense a fair amount of the water right out of that hot charge. Fancy refrigerated types can do even better. Smart physicist types can figure dew point and saturation and relative humidity and so forth to figure out how much moisture will actually condense. (You might have to ask a lowly engineer type to help you with the calculations and/or using real world experience, but we won't tell anybody that you asked :-)

    How much water? You smarter guys and gals will probably feel the need to figure that out, but it really doesn't matter how much since there will be literally gallons and gallons of it over time! That's bad. Just trust old Shorty here and plan on removing it. The first line of defense to do that is simple. Somewhere between the output of your stage two after cooler and your accumulator, you should have a moisture separator/trap. (I like to put them upstream of a one way tank check valve. It sounds like your existing compressor has a nominal trap, but plan on doing better by taking a few tips from Shorty)

    A fair amount of that moisture your after-cooler condensed will collect right there in the trap, if you build the system right. To do that, put a self venting trap UPSTREAM of your accumulator unloader valve. Lay out your after-cooler output pipes to run 100% downhill to the trap and 100% uphill from the trap to the accumulator. Shorty was taught in his little bit of schooling that water runs downhill. Laid out this way, all that water will condense in the after-cooler and drain right down to that separator. Of course, every time your pump rests the pressure in this part of the system will drop. When it does, the trap will automatically self-vent that moisture, saving you trip to the compressor ever few hours to accomplish the otherwise onerous task of remembering to drain your trap. So what you say? Well, if you don't drain it REGULARLY, The trap will quickly fill up and become useless, blowing that water right back into your accumulator. (This is bad. Water above decks, etc. shouting swearing ensigns calling on red phone, etc.)

    Shorty does want to point out that none of the above addresses downstream air purity of the functional air charge. That's a whole 'nother ball game which you must also address in your system design.

    p.s. Some nonsense in the original thread to ignore. 1.) Real he-man compressors don't have "rubber seals" in them. Rubber is non-existent in compressors and it seals in general. 2.) The intercooler is located between pump stages...got it? The heat exchanger between the pump and accumulator is called an AFTERcooler. 3.) You cannot "ignore the moisture", it is not negligible. 4.) Standards for air purity are well defined within the compressor industry - and breathable gas purity is among the highest and therefore most expensive standard to accomplish. Don't guess about this. Oil laden air is dangerous - even fatal - as a breathable gas. 5.) there is a direct relationship between required purity of air and dollars you must spend...and it goes in the direction you might expect.

    Shorty says "use the world wide inter web and Goggles to search out compressor design and engineering sources." There are lots of them with real heavy-duty Physicist quality info, calculators and stuff where they talk about bars and cycles and torrs and relative humidity and air grades and everything. Of course you physicist types will have to deign to speak with lowly engineering types and so forth to get the real scoop on how to get there, but relax. They don't bite.

    Cheers from the boiler room!

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook