Designing a pneumatic system for consistent pressure/flow

Question

Fellow engineers, I need your help in improving my design for a pneumatic system for maintaining a constant supply of at least 90 PSI and 20 CFM for a particle analyzer with stringent air requirements. Its expected to draw air every 2 seconds. Here's the breakdown of the components used in the setup (in this order). Picture attached.

A: Bulk water separator. This should help with removing moisture from the air because our plant air is heavily moisture-laden at times.

B: Particulate filter: 0.01 micron air filter to ensure clean air, free from particles.

C: Desiccant air dryer: Heat-less desiccant dryer for dry air. -40 C pressure dew point to ensure that there is no further condensation beyond this point and the air stays dry till it makes its way all the way to the analyzer. The dryer is rated for an inlet flow of 45 SCFM. I up-sized it because this is a heat-less dryer so a certain percentage of the air going in to the dryer will be used to regenerate the desiccants as well. The dryer has an in-built filter that will prevent any of the desiccants from leaving the dryer.

D: Globe valve: This is to contain the flow rate coming out of the dryer. This was suggested to me by another engineer. Since we are filling up a large tank, the air coming out of the dryer could potentially exceed the rated flow rate of the dryer which could damage the desiccants.

E & F: Check valve and booster pump in parallel. The booster pump is meant to charge the receiver tank to a pressure of 95 - 100 PSI. The check valve is there to reduce the charge time by allowing air to pass through up to the inlet pressure. This could help take some of the burden off of the booster pump so that the pump only cycles when the tank pressure needs to be raised to setpoint.

G: Air receiver tank: This will help create a reservoir of air that will ensure that the analyzer never has to wait for air. The tank is rated for 200 PSI but I won't be charging it more than 100 PSI. Still debating if I should go with a smaller tank because 80 gallons is an awful lot of volume for the booster to fill.

H: Filter: To get rid of any of the remaining particles and oil vapors left in the system.

I: Pressure regulator: To maintain the desired pressure for the equipment

J: Equipment - Particle analyzer.

Any suggestions on improving this design?

Answer 1

I see a few possible issues:

• do the amounts of air add up? you need 20 CFM @ 90 psi, that's about 120 CFM @ 15 psi (close to normal conditions) but the dryer is rated for 45 CFM

• In all applications I've seen (that I talked through with sales engineers for the air supply) the dryer was installed downstream of the compressor

• The globe valve seems suspect, wherre should the excess airflow you fear com from? Size the compressor so that airflow does not exceed the rated wirflow of the dryer

• I don't understand the check valve paralell to the compressor. When will outside pressure ever be significantly higher than pressure in tank?

• One compressor? Did you test the actual runtime it will experience, to supply the air? Is the compressor designent for that? This depends on actual air demand and size of tank.

• You did not tell us how long the ai demand is every two seconds. Just make sure you size your tank accordingly

• 95 psi tank pressure for a process that needs 90 seems low. Check your components from back to tank for actual pressure loss

• A lot will depend on the qulitiy of the pressure regulator. Get a chart with the pressure loss at the flow you expect and talk your application through with an application engineer. Often, pressure regulators have a minimum delta p of about 15 psi.

Alternativly, you can get a sales engineer (many compressor manufacturers will deliver complete stations) to design the system for you, and learn as much as possilbe from them.

Answer 2

100 psi in the buffer tank doesn't give you a lot of headroom. The closer that is to your 90psi output pressure, the larger the tank will have to be. You'll have to calculate how low the pressure will drop during the output burst. Make sure the minimum pressure is higher than output + regulator pressure drop.

Also just at a glance, placing the pump in the middle of all this seems a little suspect. I would be worried about debris from the pump, although to be fair they would probably be caught by the expansion tank and oil filter. Why not place it at the beginning?

It appears that the check valve will not do anything after start up. And you want to limit the inrush airflow anyway right? In steady state the buffer tank will be fluctuating between 100 and 90 something PSI, so why include an extra part that can fail?