How long does it take to pressurize a tank? 2

[invent11]

In our plant we have a tank with capacity of 340000 lit. We intend to use instrument air and pressurize it to 20 kPa to look for leaks. A pressure regulating valve connects plant instrument air at 860 kPa through a 1/2" line to the tank and can provide 20 Kpa at 7.6 L/s. How long does it take to fill the tank? Apparently this method has been used in the past and we are told it took days to pressurize the tank. Appreciate If you could let me know the governing equations and your estimate for the time needed.

 

[zdas04]

Invent11,
Really? You can't just think about the problem and find an approach without a cookbook solution?

THINK ABOUT IT. When the tank is at 20 kPa (I'm assuming 20 kPag, because 20 kPaa would collapse the tank, see I'm thinking about it) you can determine how much mass of air is in the tank because PV really does equal mR_airT. So now you know how much mass is in the tank. You can convert that to cubic meters at standard condition, right? If you can't do that, then I would really recommend that you read Chapter 0 of my book, because if that concept is not automatic then you have no business working as an engineer. Now you have SCm, and you know the capacity of your compressor. DIVIDE THE TWO and get time. Easy, Breezy. If not, buy Chapter 0--I called it Chapter 0 because it is full of stuff that you need to know before you can call yourself an engineer in fluids.

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[invent11]

Main compressor supplies air to a # of users ours is only one connection to network out of many. what i know is the PRV and flow rate downstream of it.

 

[zdas04]

Seriously? You really can't read what I wrote and think to yourself "he didn't mean the whole compressor, just the flow rate to the tank" and divide the SCm of the tank at 20 kPag by 7.6 L/s to get the elapsed time? You really came back here and busted my chops about an irrelevancy? Think really hard before you do any more than say, "Thank you", or even "Thank you, I bought your book and see how to calculate SCm". Engineering is about problem solving, and you seem to be missing that point. By the way, never use "340000 lit" again, it is "340000 L", getting the nomenclature right is a big leap in the direction of getting the solution right.

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[MortenA]

invent11, i think the point David is trying to make - this is a task you should be able to complete if you are working in this field - and this is not a "free engineering service". If somebody had written the answer, would you just report it? If it was wrong and you didnt understand why - what would you then do? Remember to check unit as David point out, and then find a piece of paper and a pencil... At the end of the day you might even be better off just trying it in the field then

 

[invent11]

I was looking for an insight from someone who has previously done similar practice in field or can refer me to any related reference. I think the question is if the flow rate downstream of PRV remains constant or as the pressure builds up in the tank it is reduced. This impacts the time needed to pressurize the tank. MorenA i understand that we may try it in the field but i was looking for an estimate for the time needed to plan accordingly or whether we should seek other options. I have already had the calculation in hand as per zdas04 before posting this tread but our operators' past experience indicates that the time needed is much longer than what is calculated by this method. I thank you all who responded to this inquiry but i would like to hear from someone who had any experience in this regard or has another perspective.

 

[georgeverghese]

When you say fill rate is 7.6litres/sec, what reference basis does this volumetric rate refer to - at 20kpag and 20degC or is this at standard conditions (1.0atm, 15.6degC)?
My calcs tell me, if this fill rate references standard conditions, then fill time is 854secs or about 15mins. If it takes much longer than this, you've got one or more leaks at this tank, and in this case, it does sound like a big leak.
Fill time will be even less if this flow references 20kpag, approx 20degC.
Since the final fill press of 120kpa abs << 0.5x(860+101.3), then the filling flowrate is not affected by filling press throughout the pressurisation operation. Critical press ratio for air is approx 0.5.

 

[LittleInch]

You're asking a question we can't answer as we have no data about your regulating valve and how it works.

If it controls on pressure alone then it might give a lot more than 7.6 l/sec if the initial pressie is 0 kPa. You need to look at the valve characteristics to see what the max flow is.68000litres. If you are limited to 7.6 l/sec then time is about 140 mins.

If your volume is higher to start with it will be faster but a

At these pressures you can neglect compressibility and just work from basics.

If we assume start pressure is 100 kPa(a) and you want to get to 120 kPa(a), then you need to increase volume or mass odd air by the same amount. So 20% more volume is 68000 litres. If we tale the 7.6 l/sacred figure then this is about 150 minuTed.

George either you or I have am order of magnitude error.

Bit realistically 1/2" Air line to fill 340 cubic metres is like filling a swimming pool with a garden hose.

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

 

[LittleInch]

Last post a bit garbled and repeats itself but hope you get the gist of it.

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

 

[danw2]

>I think the question is if the flow rate downstream of PRV remains constant or as the pressure builds up in the tank it is reduced.

I suspect that the PRV in use is a proportional pressure (reducing) regulator which runs wide open at the beginning when the tank is at atmospheric pressure, then the flow rate backs off as the pressure builds up in the tank and the regulator's outlet pressure approaches setpoint. So the cited 7.6L/s is not a constant flow rate, it's Cv spec for a given size PRV at the 20kPa setpoint.

 

[georgeverghese]

Oops, got cross eyed with the number of zeros on the tank volume and I used 34000litres. With 340000litres, fill time is under 150minutes, so still a big leak somewhere in this tank.

@danw2 has a point, what is the PCV setpoint pressure? Flow would drop off as tank pressure approaches 20kpag, if PCV setpoint is 20kpag. On the other hand, this same PCV would go wide open for much of the pressurisation operation when tank press is less than say 17-18kpag or so, and flow would be >7.6litres/sec. This calc is based on a time averaged fill rate of 7.6litre/sec. Setting this PCV at 25kpag or so would enable a decent fill rate even as tank press approaches 20kpag, but I doubt if it is going to improve on fill time by much. If you want a constant fill rate, use a restriction orifice instead of a PCV.

 

[invent11]

Thank you all for taking time to respond to my question. Based on your comments i may conclude that if in the past took our plant's operators a long time to pressurize the tank that could be either as suggested due to a large leakage or caused by interruptions in air supply.

 

[itsmoked]

I want to know how putting air in the tank finds leaks?
You either need an ultrasonic gun.
Or fill the tank with a little bit of refrigerant and use a leak detector.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[KevinNZ]

Itsmoked

It common to check for leaks with low pressure air. Fill the object then go around the seams and welds with soapy water in a spray bottle.

 

[itsmoked]

Ah. Thanks Kevin. B I G leaks. I'm usually looking for much smaller ones.

Keith Cress
kcress -

http://www.flaminsystems.com