Pressure Vessel

[Subystud]

Hello -

The pressure profile for a Process Tank was assessed based on field observations, shown in the attachment. The tank recirculates a thick media with added chemicals during a sealed, reaction phase. A gas evolves and collects in the headspace of the tank.

The pressures shown are a steady state balance of the head and the pump discharge pressure - when PI1 increased, PI2 increased accordingly (maxed 27,52 psig respectively)

The pump is rotary lobe, rated for water at 260 gpm at 3 bar (for 10hp). The max seen on this pump is 150 gpm for this media - viscosity of pumped media reduces max pump output.

Question:
1. Is this assessment correct based on a sealed, recirculated tank regarding Suction Head due to liquid (from tank bottom) and Filling Head in pipe (inlets to tank top) negating, and that the pump only has to overcome the headspace gas pressure (which rises due to reaction), and pressure drop due to fittings/pipe.
And that the pump only adds about 5 psi to the fluid - which is typical of positive displacement pumps to only add enough energy to overcome pressure presented before it.

2. Is this pump appropriately sized? A 15hp model will pump at 65 psig.

Thanks -

 

[LittleInch]

Is this assessment correct, I don't think so. Any recirculating system needs to ignore head and head pressure and simply look at the piping and fixture losses. Everything else balances out. The only reason you need to calculate pressure on the pump is for pressure rating of the pump and it's seals.

I actually think you have some serious problems with cavitation and vapour in the pump. Your liquid looks like it has a vapour pressure equal to the head pressure. Hence the NPSH at your pump inlet is only 20ft minus losses in the inlet pipe. You don't say what the inlet piping length and number of bends are, but it should be a short as possible. For a lobe pump to loose 40% of it's capacity is a big loss. How are you calculating flow? What is pump inlet pressure?

The pump is actually probably running too fast, but as said above all it is doing is overcoming discharge pipework friction resistance, so your 10hp pump is more than sufficient. What you need to look at is the vapour issue.

Some of the figures don't make sense though. If you have pt1 of 20psi and pt2 of 45 psi with 9 psi vertical head to overcome, then the piping losses in the discharge are 45 - 9- 20 = 16psi This makes me wonder what the inlet piping friction losses are.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[chamjam09]

I agree with LittleInch regarding the cavitation concerns. The vapor pressure of your liquid shall be equal to the headspace pressure as you will be in equilibrium with the liquid and vapor phases. However that being said you do have 20ft worth of static head and I wouldnt expect the NPSHr of the pump to be that high.

I'm perhaps reading your original response wrongly but the pump is basically adding enough energy to account for losses in the suction and losses in the discharge. As the fluid enters the pump the pressure will be equal to the vessel pressure (20psig) + static height (9psig)- losses (???psig). The discharge pressure on the other hand is equal to the destination pressure, i.e. the vessel pressure (20 psig) + static to overcome (9psig) + losses (???psig). The amount of energy the pump adds is the difference between P2 (discharge pressure) and P1 (suction pressure). Obviously we can cancel out the vessel pressure and static, therefore the pumps differential (energy added) is simply equal to the losses in the suction piping + losses in the discharge piping. On your attachment you have "actual discharge pressure" as 27.5psig. That can't be right, if the liquid needs to overcome 9 psig of static and the vessel pressure (20psig) to get into the vessel and the losses it obviously can't be that low. It will be 29 psig + losses.

I'm not I fully understand the query as well. Are you asking for explanations of why the pump is only pumping 150gpm instead of the rated 260gpm? I had considered the first point regarding cavitation leading to reduction in flow but we would need further info on NPSHr of pump to quantify. Is the pump operating at the correct speed?

The 10HP motor would be sufficient for the desired flow and head requirements.

(first time poster so go easy on me folks)

 

[LittleInch]

Do you have a variable speed motor? If so then it could all be Ok - but your logic is defiently a bit twisted.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Subystud]

Thanks for the input.

Cavitation: there is no cavitation present. The media is thickened WWTP sludge.

VFD: yes, the pump operates with a VFD.

This is not a VLE system. The headspace gas is CO2 and NO from the proceeding reactions - lots of vigorous chemistry here. Plenty of CO2 remains in solution, but the Nitric Oxide all collects in the headspace due to its non-polar nature. There is no equilibrium at play.

I see now the errors in my calculation. Obviously the pump discharge pressure must account for the Fill Pipe rise.

Thanks -

 

[LittleInch]

If the Co2 remians in solution, then the liquid will contain gas which is subjected to a lower pressure than the headspace pressure will release this CO2.

Pump discharge pressure accounts for pipe friction losses plus head rise. Head rise is balanced by the inlet pressure in terms of pump differential pressure / head / power requirement.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way