Finding fluid flow rate 1

[davechem]

Here is the problem i'm trying to solve.
I'm designing a water line to a tank from a service water line. The water line is 3" carbon steel under a pressure of 100psig. There is no back pressure from the tank. The line will be 1.5", contain 2 ball valves, a check valve, run approximately 25', and will drop down 20'. I need to know the flow rate or a way to obtain the flow rate.

Also what if you consider that one of the ball valves is 75% open or 50% open etc. Can anyone reference something that could help me determine my flow rate?

Thanks,
Dave

 

[jsummerfield]

You can rent a clamp on transit-time a.k.a. time of flight ultrasonic flowmeter from Controltron or another maker.

John

 

[davechem]

I really need a way to do this through calculations and not through a field test.

 

[Montemayor]

davechem:

You should never employ ball valves as throttling devices. If you need to throttle, use a globe (or "needle") valve type design. Ball valves are designed to serve as block valves - on or off. Using them for throttling will result in seat damage.

If you are "designing a water line to a tank from a service water line", and you have the pipe diameter, the pressure drop, the pipe length, the fluid viscosity and density you should have no problem in calculating the design water flow rate. As a Chem Eng, this should be a piece of cake. All it involves is Darcy's equation with a friction factor relationship such as the Colebrook-White or the Moody Chart. All of this is explained in great detail in Crane's Tech Paper #410 or Cameron's Hydraulic Handbook.

I hope this helps you out.

 

[TD2K]

I estimate about 200 gpm to 250 gpm to use up the available 100 psi pressure. I'm assuming 1.5" sch 80 pipe and that the 3" supply line can deliver the 200 gpm at the 100 psi pressure. I added in a entrance and exit loss and 75 L/D for your valves and fittings (ballpark).

Art's given you some great references if you don't already have them.

For the case(s) with the valve partially closed, you'd have to get a Cv curve from the vendor to factor that in. I don't have a feeling for that off the top of my head (for a ball valve 75% open my guess is it's not going to reduce the flow much).

 

[quark]

I agree with TD2K. I am getting around 280 gpm. I didn't consider the entry exit losses and also 20' of available static head. So they will cancel out(atleast for this lazy soul). You should definitely worry about the high velocity around 14.5 m/s.

Regards,

 

[katmar]

There is information that you have not given that is important, especially as the line is relatively short. It is always so easy for someone like myself (and TD2K and quark) who have the software at there fingertips to simply plug in a few numbers and get an answer. And indeed, the numbers calculated by TD2K and quark are probably close to the correct answer. But we do not know why you need this number and how accurate you need to be.

With such a short line the relative pressure drop in the pipe fittings will be important. You need to estimate these with sufficient accuracy for your requirements using the methods referenced by Art Montemayor.

As quark has pointed out, the velocity is high for a water line. More than 10% of the available pressure drop is consumed in accelerating the water to this velocity. Depending on how accurate you want to be, this may be important to consider. Plus, as Art said, you will damage your ball valves.

You did not give the schedule of the pipe. This is particularly important with small bore piping as the change in diameter is relatively large. A Sched 40 pipe will give about 20% more flow than a Sched 80 pipe under these circumstances.

It would be worth your while doing the calculation described by Art to discover for yourself which is the most pertinent information, and how sensitive your answer is to the various input data.

 

[davechem]

The pipe is schedule 40, and the ball valve will not be used as a throttling device so i'm not really sure why I asked in the first place. Which would mean over 250gpm by what you've said. That figure seems to be higher than expected. The purpose of the line is to dilute Acid polar solvent.

Can anyone provide online references. I will check my texts at home, but i'm away on business for a few more days.

I would like the data to be within ~20gpm of the actual. I may end up using a flow meter but I have had poor experiences with flow meters especially if I'm using a spare one.

 

[davechem]

sorry make that schedule 80 seamless steel

 

[mbeychok]

Quark:

Perhaps, you made a typo. Using 3 inch schedule 80 pipe (with an inside diameter of 2.90 inches), a flow rate of 280 gallons per minute would be a velocity of 13.6 ft/s rather than 14.5 m/s.

Milton Beychok
(Contact me at www.air-dispersion.com)

 

[sailoday28]

Another thing to consider in the calc for flow is the affect the flow will have on the upstream 100 psi pressure.

 

[quark]

Mbeychok,

The line is of 40NB and the main header is 80NB. If it is Sch.80 then the velocity is around 15.4m/s.

Here is how I did it.

First of all, I neglected initial velocity of water(this is one caveat and I feel even TD2K omitted it). No positive static head.

Just got the equivalent lengths for a schedule 40 pipe with bends and valves and plugged in the Darcy pressure drop equation for a pressure drop of 100PSI(another caveat).

Davechem,

You can start from here,

http://me.queensu.ca/courses/mech451/losses.htm

but this is not a straight forward calculation. We are just trying to give you a ballpark figure.

You should also check what Sailoday says. The pressure may not be constant at 100psi once the flow starts in the 40NB pipe and etc.

Regards,

 

[mbeychok]

Quark:

Your earlier response said that you got 280 gpm. That converts to 0.623 ft³/second.

A 3 inch schedule 80 pipe has an inside diameter of 2.9 inches. That converts to a cross-sectional area of 0.0459 ft².

Finally, 0.623 ft³/second divided by 0.0459 ft² equals a linear velocity of 13.6 ft/second.

This isn't about your use of the Darcy equation. It is simply the conversion of your 280 gpm in a 3 inch schedule 80 pipe to a linear velocity of 13.6 ft/second. To have a linear velocity of 15 m/second, the flow rate would have to be very much higher than 280 gpm ... in fact, about 1014 gpm. Are you sure that your answer was in meters per second?

Milton Beychok
(Contact me at www.air-dispersion.com)

 

[Montemayor]

Dave:

My calcs also agree with TD2k & quark's on the estimated flow rate. However, this is an estimate and - as Katmar correctly points out - subject to the accuracy of the method of calculating the result. Katmar astutely reminds all of us to be wary of the fitting loss contribution in such a short run. I would expect the fittings (& valves) to dominate the pressure loss and, as such, would use the 2-K method of pressure drop calculation outlined in the Crane TP #410.

My tables and experience indicates to me that you won't get a calculated answer within the 20 gpm margin you're looking for. I wouldn't expect to get below 15% accuracy on the calculation of the pressure drop for this type of application. I wish I could be more accurate, but past experience indicates otherwise.

I also come up with 13.6 ft/sec average water velocity within the 3" sch80 pipe when handling 280 gpm. I suspect quark merely got the usual units' Leprechaun doing strange things to his calculations as we also experience from time-to-time. My expected design water velocity would have been in the 5-6 ft/sec range. 14.5 mt/sec is astronomical. I consider even 13.6 ft/sec as too excessive under a sustained operation, and the source of possible pipe errosion.

 

[quark]

Mbeychok,

Yeah, I am quite sure.

The main source line is 3" and the branch which feeds the tank is 1.5"(see the initial post of OP)

The line will be 1.5", contain 2 ball valves, a check valve, run approximately 25', and will drop down 20'. I need to know the flow rate or a way to obtain the flow rate.

A 1.5" pipe will have 1/4th cross sectional area to that of a 3" line and the velocity will be 13.6*4 = 54.4ft/sec(as per your calculation)

Regards,

 

[mbeychok]

Quark:

Please accept my apology. You are right and the velocity in the 1.5" pipe would be about 54 ft/sec. I should have read your response more closely. As you and Art both stated, that is much too high a flow for a practical design.

But this does raise a point. The 3" main service line must be serving other sources of water use as well as this new use of 280 gpm. If the 280 gpm by itself amounts to a velocity of about 14 ft/sec in the 3" main line, then it indicates that the main service line was not intended nor designed for supplying a new user of that large a flow. I expect that Davechem will have to throttle the flow into his 1.5" line to a much lower rate than 280 gpm.


Milton Beychok
(Contact me at www.air-dispersion.com)

 

[quark]

Mbeychok,

There is absolutely no need for apologies.

You are right that there are good number of problems in this situation. Care should be taken while designing the system and predicting the flowrate within 10% may be difficult(and just impossible to me)

Regards,