Looking for gas pipe flow experiments with published results

[fpnut]

Hi All,
I've just made a simple spreadsheet for estimating pipe diameters. Since my application is for gas, lower pressures (<30 bar), short lengths of pipe (<30 metres), relatively low pressure drop, I've opted for a simple Darcy-weisbach equation, coupled with Swamee-Jain friction factor estimation.

Now I'm looking to test to see if it's correct. Does anyone know a resource for experiments with published results for this type of calculation?

Any pointers would be appreciated, thanks.

 

[goutam_freelance]

Darcy Weisbach equation is being used for about 100 years. It is well established now. For comparison purposes you can verify with formulae and charts available in Crane Manual 410M.

Engineers, think what we have done to the environment !

https://www.linkedin.com/in/goutam-das-59743b30/

 

[katmar]

Program verification is a very important step. I do not know of any actual experimental results that have been published that you could use to verify your spreadsheet. For verifying my own software I used calculations published in widely distributed texts because those calculations would have been seen and reworked by many engineers and if the calculations contained any mistakes they should have been brought to light.

You can find references to these calculations and the actual numbers at

https://www.katmarsoftware.com/examples/aioflo-example-index.htm

See examples 8, 9, 10 and 12 for calculations involving gases.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[fpnut]

Hi Goutam, thanks, I'll have a look at Crane.

Hi Katmar, thanks, I'll do some verification based on your links.

 

[LittleInch]

Not sure what application this is for but you must be looking at very small pressure drops to the point of insignificance, unless you're running a jet engine.

At those sorts of lengths there will be many factors other than the pipe length which will give you pressure drop including swirl, gas composition, small imperfections etc. Different equations work better with different velocities, pressure and pressure drops but you haven't given us much or any real information for anyone to give you their input.

I don't know what the number is but I doubt any of the established equations will work on only 30m of pipe, but I guess this depends on the diameters (which we don't know)

My guess is that unless it is longer than about 30-50D you won't get established flow to the point where the pressure drop becomes established.

Why don't you just measure it? Then you can write your own paper on it.

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

 

[1503-44]

If its 1/2 to-12" pipe, 30m is going to be > 50D and if its straight, so much the better. What's the problem? Surely there is ample cause to investigate the frictional pressure drop contribution to the total pressure loss within a 30m or even shorter length, especially if he is talking about the smaller diameters and large flow rates.

 

[katmar]

LittleInch - not everyone is designing cross country pipelines. In a typical chemical or petrochem plant there will be many pipes that are less than 30 m in length transporting gases and vapors, and they all need to be sized. Of course the gas composition is important in determining the density and viscosity, but I have yet to see a commercial pipe sizing application that requires the calculation of swirl or the accommodation of small imperfections.

In real life we are deciding between (say) an 8" pipe and a 10" pipe where the pressure drop in the 8" line will be around 3x that in the 10" line. It's not like we are determining pipe bores to micron tolerances. Do you have access to the Crane TP 410 manual? It describes the sort of piping applications relevant to this discussion.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[LittleInch]

True. My point was that not really knowing what sort of DP the OP is actually looking at or whether they are limited by something else like velocity it wasn't clear whether such small additions to the diff pressure would be significant or not.

fpnut just says "relatively low pressure drop". That could be single mbar or 2 bar. Ditto velocity.

We don't know if its straight or has bends, tees, tappings, insertion probes etc etc. All will make the pressure drop calculation less accurate than long straight lengths of pipe.

Hell, we don't even know what diameter range we're talking about here, or level of accuracy fpnut is looking for even - 1% or less will be a lot harder to achieve than 5 - 10%.

so the answer to the question "Does anyone know a resource for experiments with published results for this type of calculation?" is no, mainly because real life tends to get in the way....

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

 

[1503-44]

Even on a good day, if you can get within 10%, that's excellent annd he's just talking friction anyway, so nothing wrong with using his proposed equations for that. I think he just wants to compare his answers against something else; real data if possible.

fpnut, if you post a couple of problems, someone here just might check them against their spreadsheet.

 

[fpnut]

Thanks all for the discussion. Well my application is relatively simple and not complex/critical like those in O&G requirements. And it's more of scratching a personal itch rather than doing it for production work.

I recently started work in a company mainly doing gas distribution piping/manifold for smaller workshops, e.g. compressed air, nitrogen, and other welding gases. Pipe sizes rarely exceed 30m total length (including fittings converted to equal length), and rarely go above 2". Think of "the gas cylinders are just at the back of the wall here, I need you to install piping to the workspace across the room there".
Pressure drop is perhaps negligible - or at least I can't really see the difference at the source and at the endpoint, since I'm using standard gauges with 2.5% accuracy.

When I joined, my predecessor left a table of flow rates vs pipe lengths vs pipe sizes, but no calculations whatsoever. I assume it was at least sufficient as the company had never received any complaints about the pressure/flow not being enough. But I am personally bugged by the uncertainty and looked into at least doing some validation on my part, even if management is probably still going to go ahead with the tried and true table.

 

[LittleInch]

fpnut,

Fair enough - I can see why you want to do this to understand what the background is.

For small sizes you can get huge differences by going up or down one or two sizes and clearly those table work on the basis of a single pressure or did he do different tables for different pressures?

If you want more experimental stuff why not try your local university or training college to see if they do experiments or would be interested in "real life" issues affecting gas flow and calculations. Sometimes the lecturers really like introducing this as it makes it "real" for students and they have a lot more time and equipment than you do.

Big companies that deal with the same stuff day in day out often produce similar tables, but with a bit more background. sometime they publish it so try a gas distribution company or similar to see if they will share their data and compare it.

Having pipe that is bigger than you need isn't always a bad thing as the customer gets what he wants - gas at a flow and pressure that isn't being affected by the piping, but if you're paying for it then it might be different.

Hope we've helped a little and please let us know how you get on.

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

 

[1503-44]

Can you post the tables? We can easily compare those against others and see if they contain commonly accepted values.

 

[fpnut]

1503-44,
I initially thought about it, but decided against it for privacy reasons.


Littleinch,
He did do a table for different pressures, but only for a few pressures, since that's all we are dealing with at the moment.

I plan to gather data using existing field installations, perhaps that would be representative of real life data.

Not sure about the big gas companies, but I'll give it a shot.

Agreed, I also always err on the side of caution and recommend a bigger size than necessary, citing reasons for future-proofing (which is not unreasonable), but it comes back to the customer's decision in the end.

You've all been helpful by just responding and sharing thoughts. The next step is to get the Crane TP410. I don't live in the US, but I have friends who do and can send me that book.

Well the next best progression for the company is to start upgrading our resources and do bigger projects - at which point I'll recommend management to invest in proper software. That said I feel it's still necessary to understand the formulas and calculations and not just blindly plug numbers into software.

 

[1503-44]

For now, maybe you can check your results against air flow charts like this one.

https://www.engineersedge.com/pipe_flow_capacity.htm

 

[LittleInch]

If you haven't already found it I find this site very useful.

https://www.engineeringtoolbox.com/amp/gas-air-systems-t_22.html

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