Way to calculate flow rate fed by drain? 3

[Zonkytonkman]

Hello, we have just designed/installed a new system to recover hot water. Previously we generated (through a cooling process) large quantities of hot water that had to be dumped to drain for lack of enough space to store all of it, only to have to heat similar quantities of water mere hours later for another process. We devised a system where the overflow on the existing hot water tank would be redirected to fill a new hot water tank. The over flow line on the existing tank is just a straight 3" sch 40 line coming from the bottom of the tank straight up. From the overflow line (when this pipe starts to fill) to the top of the tank there is only 6" of height. On its initial run the old tank (which should have filled the new tank as it overflowed) overflowed from its roof, leaving me with quite a mess and a water recovery system that is useless. I walked in on the middle of this project and the engineer that handed it off to me assured that he calculated the flow velocities and they checked out fine.
Anyway, i'm going back throught the calculations now and I realized that I don't know how to account for the line filling as a drain, I'm wondering if the vortex (I'm assuming it will act similarly to a drain on a sink) cause a slowdown in the flow? Will the pipes flow be turbulent, and will it fill the entire cross section of the pipe?
I havn't done much work with piping before, so i'd appreciate any help you could offer. The liquid is plain water, approximately 60-80 degrees celsius.

 

[quark]

Vortexing is definitely not a problem here. Where is the receiving tank located and what is the elevational difference between the two tanks? What is the flowrate into the first tank? How long the 3" pipe is?

Your main problem may be the head required to create flow of water (6" maximum). As a rough estimate, water will not flow if your transfer pipe lenght is more than 11 meters if water flows at a velocity of 1m/s.

Regards,


Eng-Tips.com : Solving your problems before you get them.

 

[PEDARRIN2]

Also, something to consider is whether the overflow tank is open to atmosphere. If not, you could be creating an air cushion that will prevent water from flowing into the tank.

There is a formula for flow out of a container through an apurtance.

It is

Q=C(d)*A*(2gH)^.5

Q is flow rate in cubic meters per second
C(d) is .6014
A is area in meters
g is 9.81
H is height of the tank

If the flow rate out is less than the flow rate in, the first tank will overflow.

 

[PEDARRIN2]

I realized I need to make a correction.

H is 6"

The formula is for an apurtance at the bottom of a tank. It would be modified for this case since the "bottom" is where the pipe opens into the tank.

I apologize for that oversight.

 

[Montemayor]

Zonky:

You give a very difficult description of what is happening but I believe I visualize what you are experiencing. You say "The over flow line on the existing tank is just a straight 3" sch 40 line coming from the bottom of the tank straight up"; of course, if you are overflowing from one tank to another, you must mean that you have gravity, vertical flow. Is that correct? I also believe you mean to say that the 3" overflow line vertical and extends down through the roof of the new, excess capacity tank and down to its bottom section. Sort of like a "dip pipe". Is this also correct? If so, a very long discussion was had on thread 378-81608 on an analagous Diesel flow problem.

I strongly recommend you read: "Designing Piping for Gravity Flow"; P.D. Hill; Chemical Engineering; Sept 05, 1983. This is a classical paper on the subject and has been used by no less than Bechtel Engineering to showcase the problem and give examples of how to deal with it. You can also go to:

http:/

http://www.freecalc.com/selffram.htm

and use this service to calculate the correct size of the overflow pipe for gravity flow. I would also refer you to the NORSOK standards on Process Design (P-CR-001 Rev. 1, December 1994) which should be available on the InterNet. There, you will also find that the subject is dealt with just as in the above references. Specifically, NORSOK recommends as follows:
"Vertical gravity lines (such as liquids from sea water returns and produced water discharge) shall be designed such that the Froude number is less than 0.3 to avoid air entrainment.

Froude Number = V^2/Dg

where,
V = Velocity assuming full pipe (m/s)
D = pipe inner diameter (m)
g = gravity constant (m/ s2)

Gravity drains are very important items on Off-shore platforms and NORSOK makes sure of dealing with this subject. I have a copy of their document in a Word for Windows file and can send it to you if you're interested. Many people have a lot of trouble understanding the phenomena of free-fall, gravity flow in a drain and that's why I give all these references. Some people still don't believe me when I recommend self-venting design.

I have always used this principle on vertical, gravity flow lines and never had any problems. I always try to design for self-venting flow. I believe it is impractical and naive to expect that a vertical overflow line will flow 100% liquid-full across it's cross-section 100% of the time. This just doesn't occur as the references and other authors will concur with me. Your hot water being close to its saturation point doesn't help matters any.

Only two things can be happening if your overflow doesn't allow the excess water to exit the first tank and fall into the next reservoir: 1) The overflow line is not venting; or 2) the overflow line is plugged up. I'm betting the line is not self-venting.

I hope this helps you out.


Art Montemayor
Spring, TX

 

[quark]

Mr Montemayor!

From the OP's description, the arrangement seems to be different to what you said. The pipe is not a dip pipe but a pipe taken from a nozzle connected at the bottom of the tank and taken above the free surface of the liquid. He might have mistaken by considering static head equal to tank height.

PS: Btw, could you please let me know whether you received my mail dated 31/12/2003?

Regards,


Eng-Tips.com : Solving your problems before you get them.

 

[MortenA]

The NORSOK (Norwegian petroleum standards) are avaiable at the internet (for free).

The general process standard has someting on this subject:

http://www.standard.no/pronorm-2/data/f/0/01/41/1_10704_0/P-001r3.pdf

Check page 12/13

Best regards

Morten

 

[stevolando]

Montemayor,

I am currently working on a gravity fed piping system that requires diesel to flow from a surface tank to an underground tank located 2km below the surface. I am interested in looking at the "Designing Piping for Gravity Flow" by P.D. Hill that you mentioned. I am unable to find this document and I am wondering if you could tell me where I could find a copy, or possibly if you could send me a copy.
Thanks in advance

Stevolando
SMOrlando@mcintoshengineering.com

 

[Zonkytonkman]

Hello, thanks for all the help. In order to clarify my system, you can think of the original hot water tank (tank 1) as a bucket with a pipe coming straight up through the middle, but cut off before the top of the bucket so that as the bucket fills to the top of the pipe it simply falls down into that pipe(drain) which feeds into the wall of the recovery tank(tank 2) near the top. The drain line runs 5 feet down, 90s into a horizontal run of approx 40-60 feet(not strictly horizontal, very slight slope) into tank 2. Drain line is 3 inch copper sch40.

 

[PEDARRIN2]

You might have something similar to a plumbing drain. Plumbing piping typically has a very slight slope (1/16 to 1/4" vertical drop per 1 foot horizontal run.)

Is the bottom tank open to the atmosphere or vented?

If there is no way for air to move in the system, you could get air locked and that would keep flow from going to the lower tank.

 

[Montemayor]

quark:

Thanks for your useful explanation and insight as well as for your note of 31/12/2003. I had my beautiful granddaughters for the Holidays and they took over my computer in my study, trying to send emails. They accidently erased your message before I could reply. I'm grateful for your message and hope that you enjoy the same friendship and camaraderie as I obviously have.

I believe that I now have an accurate picture; there is a gravity overflow from one tank to another and, as such, is subject to the Froude relationship which is basically the ratio of inertia forces to gravity forces - independent of the Reynolds number.

MortenA:
You've helped me out again by not only advising of the specific Norsok website, but also noting that there have been recent revisions to their excellent Process Design Guidelines. However, the portion on gravity draining is still the same writeup: limit the size to less than a Froude number of 0.3. P.D. Hills' article is not the only specific reference in concert with this recommendation on gravity drains; Larry Simpson (Union Carbide Guru) wrote a similar article in Chem Engr., June 17, 1968, and also suggested basing pipe outlet diameters on a limiting Froude number of 0.3.

The Norsok process design guidelines are an excellent indication of proper and safe design, in my opinion. While some may criticize that they are too conservative, the good thing about the Norsok is that it places a high value on safety and stable process conditions. The free accessability to these design guidelines is a commendable resource for all design engineers and everyone should download these very helpful engineering jewels.

stevolando:
I have a Xerox copy of the original Bechtel Process Design copy that was furnished for their Design Handbook. A friend made a copy for me some years ago because I couldn't find a copy of the original article. Bechtel E&C thought so much of this writeup on the subject that they featured it in their design manual. I'm sure if you have access to an engineering or technical library with back issues of Chemical Engineering, you should be able to get a copy made for you. I am presently transposing this information to an electronic spreadsheet on fluid flow, but it is tediously manual and will take time. As soon as I finish, I can email you a copy. I presume you have visited the thread I previously mentioned on Diesel draining in a mine. Hills' article, in my opinion covers the practical answer and alternative ways very well.

Zonky:
Your configuration is hydraulically the same as if you had welded a 4" half-pipe to the exterior of your primary tank and allowed the overflow to exit through that conduit on down to the stand-by tank. You haven't addressed my questions nor given a design flowrate for your overflow stream, so I can't comment on what you have. But I can state that if you follow the design criteria spelled out in Hills' article as well as in the other references, you will not have any problems. Liquid flowing vertically down does so as an annular film. In such cases, low superficial velocities are necessary to avoid gas being sucked down with the liquid. I agree with Quark: vortexing is not a problem nor a factor in your application.

I hope this has given you some help.


Art Montemayor
Spring, TX

 

[unclesyd]

Here is website that was posted as source for technical reprints of all sorts. Sorry I didn’t note the posters name.

http://lindahall.org/

 

[katmar]

Zonkytonkman,

By my calcs, if there is no air lock occuring, you should get a flowrate of about 150 gpm from the one tank to the other. Check the rate at which the level rises in the first tank (before it starts draining to the second tank) and this will tell you the rate at which the water is entering the tank. If this is more than 150 gpm your pipe is too small and the only answer is to fit a larger one.

If the rate of filling the first tank is less than 150 gpm then there is some reason why the pipe is not carrying its full capacity. If this is so, I would guess it is either because of vortexing or an air lock.

With a maximum head of 6" over the entrance to the drain pipe I would not expect a vortex, but have a look and see if you have one. If there is, it is easily stopped by welding a 12" x 2" piece of 1/8" plate across the diameter of the entrance of the pipe. Its hard to describe, but one of the 12" edges should be tacked across the diameter of the pipe so that approx 4.5" sticks out horizontally on each side. This will offer minimum restriction into the pipe and will stop any swirl.

I believe the more likely problem is an air lock, especially as the "horizontal" section is sloped down to the second tank. As the water starts flowing it will try to flush the air out of the horizontal pipe, but because of the slope the bubbles will want to move in the opposite direction. If possible, extend the vertical section down a bit so that the slope is actually UP towards the second tank. This will make it easier to flush out the bubbles as the water flow and the bouyancy of the bubbles will be acting in the same direction. This would be a problem if automatic draining of the pipe is necessary. Do not adjust the slope by raising the discharge point as this will lower your driving force.

If there is no vortexing and you can flush the bubbles out of the horizontal section, then self-venting flow is not an issue and you can run the pipe flooded. This will give you a much higher flow than with a self-venting design. If you decide to go with self-venting (which is a safe but conservative answer) remember that the self-venting criteria are different in the vertical and horizontal sections. The references given by Art will give you the details.

regards
Harvey

 

[Zonkytonkman]

Hello monte mayor, I would really appreciate that document on drains that you mentioned. But I have no way to email you. You can try me at joecoady@molson.com.nospamthanks
Thanks for your help.
I made a mistake in my second description in that the pipe is cut off shorter than the bucket, not the other way around. For simplicity imagine a 12" high bucket with a open pipe coming straight vertically through the bottom and up 6", so that 6" of head can develop on top of it.

Both tanks have atmospheric vents.

And yes, this is a gravity fed system, no pumps on the drain, but water is pumped into tank 1 to fill it, but only at a rate of about 80-90 GPM

 

[stevolando]

Montemayor,

I would be very greatful if you could send me a copy of that document. Just wondering how long it will be. I'm not in a rush, I am just not sure how tedious it is.
Thanks again,

SMOrlando@mcintoshengineering.com

 

[stevolando]

Montemayor,

I'm sorry but I have not visited the Diesel draining in a mine thread that you posted earlier. I cannot find this thread, can you send me the link?
thanks

Stevolando

 

[PEDARRIN2]

Katmar,

I would like to know how you arrived at your result of 150 gpm.

The reason is that this situation is very similar to what happens in a roof drain, i.e. a layer of water on a roof drains into a vertical pipe which offsets horizontally fairly quickly.

Using that model, and using some of the formulas for vertical and horizontal piping in storm drainage piping flowing full, I got a result of about 23.83 gpm. This is with a slope of 1/8" per 1 foot. If the slope is 1/16" per 1 foot, then the flow rate is 16.86 gpm.

If the flow in is 80-90 gpm, then it would be expected to overflow the first tank.

 

[Montemayor]

stevolando:

I apologize for not citing the referenced Diesel thread number. It is the following:

thread378-81608

I hope it gives you some insight as to what some answers and problems can be related to the subject.


Art Montemayor
Spring, TX

 

[katmar]

Pedarrin2,

The difference between my calculation and yours is that you are neglecting the head developed by the static head in the 5 ft vertical leg. If you are assuming self-venting flow in the vertical section then you cannot rely on having this head. However, I have assumed that the vertical leg is flooded, i.e. it is full of water and it is driving the water through the horizontal section.

Using the relationship from the Hills reference above with a flow of 150 gpm and a pipe diameter of 3", the vertical section will be flooded if the level in the tank is 4" above the drain pipe. This is rather tight, as the maximum available is only 6", but it should work.

I therefore took a pressure drop of 5 ft of water over a length of 80 ft of 3" pipe and got a flowrate of 150 gpm.

regards
Harvey

 

[Montemayor]

Zonkytonkman:

Just want to let you know I sent you a copy of the article you requested to the email you furnished and it was returned "undeliverable". Sorry.

It seems that at 90 gpm in a self-venting drain you need a pipe that is approximately 6". I can easily visualize your primary tank backfilling. At least your atmospheric vent has the capacity to handle that overflow; otherwise, you'd bust your tank roof for sure.

Art Montemayor
Spring, TX