Spec'ing Check Valves for Chemical Injection - Pumps higher in elevation than injection point 2

[ajz]

Hello,

First time poster -- hopefully I'm posting in the right place and following forum rules. Please let me know otherwise.

I was tasked with spec'ing check valves for a boiler chemical injection system. My conceptual understanding of the fluid properties is a bit rusty, so I'd appreciate any insight.

* Metering pumps (~0.005 GPM) are on the second floor of the building, elevation 139 ft
* The chemical injection point is 30 ft below the pumps.
* The pipe at which it discharges into draws upwards of 120 GPM from a deaerator, and it's pumps are located in the basement, ~46 ft below the metering pumps.
* A single 50 psi cracking pressure check valve is installed immediately after the metering pumps on the second floor.

My task is to install two check valves after the metering pumps, and properly spec them to prevent siphoning of chemicals.

I have attached a sketch. Boiler chemical pump is the blue line, and the boiler feed water is the green pipe. The deaerator feed pumps are in orange. The labeled pressures were recorded off of actual pressure gauges. Flows were approximated. Pumps are in parallel.

My rational for spec'ing the check valves was to sum all of the suction forces after the metering pumps to the lowest point. From pumps to basement is a free-fall drop of 46 feet, equivalent to 20 PSI water column. I do not know how to estimate the effect of suction from the pumps in the basement on this system--I'm not sure if it is negligible or not.

My hunch is that check valves with a cracking pressure of 25 PSI should be satisfactory for this application, rather than the 50 PSI cracking pressure check valve that was previously installed.

Can anyone offer any input on this scenario? How could I best choose two check valves to install in series to ensure that boiler chemicals on the 2nd floor of a building are not siphoned throughout the system (minding the effects of hydrostatics/dynamics in this particular case)?

Thank you

 

[LittleInch]

Very good first past. A diagram, an explanation and good data.

All you need to think about is a worst case of absolute vacuum at the top of your pipe. Allowing for the riser section of say 10 ft the pressure on the d's side of your check valves will be 4 psia.If your head on the u/s side is atmospheric pressure only then so long as your valve has a cracking pressure greater than 15 psi you will never get flow without the pumps running.

At normal pressure the tie in point looks to be about 10 psi so pressure at the top of your loop is about 13 psia. So with the head and the cracking pressure of 15 psi you will need about 20 psi from your pump.

Try to just use feet head. Makes it easier to work out.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[ajz]

I knew there was a trick to it! Thank you so much for jogging my memory! I naively forgot to consider the absolute vacuum limit, which in retrospect, should have been painfully obvious...

15 psi cracking pressure makes perfect sense now.

Greatly appreciate it!

 

[LittleInch]

No problem. Glad it made sense.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[ajz]

Pardon me for reviving this question... I did some more thinking about this scenario and was looking for a concept check.

Based on my new understanding, my original diagram can be simplified substantially, as such:

Performing a rough energy balance on this scenario, assuming an absolute vacuum were possible, the maximum necessary rating for a check valve would follow the following relationship.

Cracking occurs when the pressures working against the check valve over come the forces working with the check valve...

Pressure_"left side" + Pressure_cracking = Pressure_"right side"

Pressure_cracking = Pressure_atmosphere + Pressure_ρgh - Pressure_"left side"

Using some real world numbers,
Pressure_"left side" = 0 psi ("Pressure_suction per the diagram")
Pressure_atmosphere = 14.7 psi (@ sea level)
Pressure_ρgh = assume 3 feet of water = 1.3 psi

Therefore, the maximum necessary rating for the check valve would be 16 psi. (In reality, it's less.)

So challenging this understanding, we could say...
*If the tank were substantially taller, say as tall as a skyscraper, than the hydrostatic pressure due to the weight of the column of water would be forcing against the valve, and thus a greater check rating required.
*If atmospheric pressure were lower, say at a very high altitude, than the force of the weight of air would be lower and thus a lower rating required.
*It is abundantly obvious that perfect vacuum would never be achieved, especially if water were performing the work which would end up eventually boiling, so there would be some vapor pressure term in there eventually. Also, my terminology for "Pressure_suction" is obviously bad. Thus, the pressures on the left side will always be quite remarkably greater than 0 psi, and highly improbably less than 14.7 PSI. Thus, it could be said 16 PSI cracking would be VERY generous in reality.

I think I got it now! (or do I?)