6 replies to this topic

[iry56]

Hi
Please i really need help urgently on this issue,i just got a project work in school which is basically to design a CO2 pipeline.I am using hand calculations based on the basic gas equations but i'm really getting confused

Firstly i am confused on whether darcy weisblach pressure drop equation can apply to the flow of supercritical fluids since co2 is normally transported in this condition.I am also confused on the velocity constant to use inorder to determine my erosional velocity.

Secondly i am stuck on how to determine the viscosity of CO2, the initial model i used gave me an extremely high reynold's number to the order of 10^11 for a 10" pipe at a pressure of 2205psi and 32degrees centigrade.

I am not really into piping design but would be grateful of any help i can get from this forum.Pls anyone who has previous knowledge on co2 pipeline should pls help me out

Thanks

[NGL Licensor]

Hi,

At 32 C and 2205 psia CO2 is in supercritical liquid phase. I see referring to gas equations.

Can you re-visit your conditions and make sure about the actual phase conditions

Regards,
A.King

[Art Montemayor]

Irv56:

Well at least you are aware that you are dealing with a Supercritical Fluid (SCF). That is a positive start. Now to answer your specific questions:

1. Yes, the Darcy-Weisbach equation can be used to calculate the pressure drop. However, stick to the limitations inherent to that relationship. There is nothing wrong with “hand” calculations. I would recommend you use a spreadsheet; it is much more accurate and efficient. Set up the equations using an algorithm that you organize.
2. Why are you worried about a so-called “velocity constant”? Are you trying to apply API 14E with respect to expected erosion effects? I normally would not expect any erosion in this application, but you can account for it if you wish. Refer to API 14E for the erosion constant.
3. You can easily obtain (for FREE!) all the thermodynamic values you ever desired (including viscosity) for CO2 by going to http://webbook.nist.gov/chemistry/

The reason CO2 is transported as a SCF is that it has very low viscosity values in that state.

[iry56]

I appreciate all your contributions and am really grateful
Please Mr Art i just wanted to ask about the restrictions i would consider while utilizing the Darcy Weisbach formula.My supervisor thinks the formula only applies to liquids and can't be used for gases.
Secondly is the general gas equation suitable for designing for CO2 in its suoercrtical state?
Thanks in ernest

[vicini]

supercriticle fluids can be simulated in pipelines with the Hazen Williams and general friction factor calcs. I use an excel spread sheet for supercritical ethylene. What you have to watch is density as it changes due to pressure drop. I usually break the line up into 1 to 5 mile segments and recalculate density. You can use NIST webook to get the density. convert your flows to bbl/hr to make it easy. you just solve the drop for each mile and then have excel carry the outlet pressure as the inlet to the next segment and replicate the formulas. I use a double itteration for reynolds number to calc friction factor change the density every line as needed.

CO2 is sold by the cubic foot in the oil industry. As a liquid or dry ice it is sold by the pound and under cap and tax it will be taxed by the tonne.

Attached Files

•  liquid_pipeline_PD.xls   39KB   95 downloads

[Art Montemayor]

Irv56:

While it is strictly true that the Darcy-Weisbach equation is valid for fully developed, steady state and incompressible flow, it can still be used for compressible flow within certain restraints. To identify the limits under which the Darcy-Weisbach equation can be used, refer to Crane’s Tech Paper #410. But the issue here is Supercritical Flow (SCF) – not gas flow, and I maintain that you can use the Darcy-Weisbach equation if you constrain your calculations in a manner recommended by vicini in the previous post. What he/she is recommending is exactly the manner in whichI would lay out the algorithm for the calculation of a pipeline for SCF flow.

I would not use the general gas equation for designing a CO2 supercritical pipeline. Unlike vicini, I don’t use the Hazen-Williams equation – even for common water flow. I have found it to be relatively very inaccurate as compared to Darcy-Weisbach. The ease of using the Hazen-Williams equation has been equaled in the introduction of such relationships as the Churchill, Swamee, and Serghides' Equations for resolving the friction factor directly. For an example, refer to my attached workbook and obtain a sample of what I mean by an algorithm setup. Even though this spread sheet is set up for estimating a pipe size using API 14E, you can also set one up for any other relationship.

By the way, you failed to answer my question regarding use of API 14E. Please note that I always try to answer every question asked on the Forum. It would be not only courteous, but also more efficient, if everyone answered the questions asked or stated that no answer would be forthcoming.

(As an aside, vicini’s workbook contains some nostalgia for me while I sit here in Madrid, Spain. Even though he uses abbreviations, I know what they mean from my personal hometown experience. Words like Mount Belvieu, Farm to Market Road 565, Cedar Bayou, San Jacinto, La Porte, Clear Creek, Friendswood, and Alvin are all embedded in my memory and remind me of my youth and growing up in Galveston, Texas and the Gulf Coast.)  Art__s_Line_Sizing___Line_List.xls   48KB   87 downloads

[iry56]

Thanks Vicini and Mr Art.I now have an idea on how to create a spread sheet for the design of the pipeline.
I have decided to use the API 14E recommendation for my erosional velocity calculation based on the constant value of 100.
I have also taken note of the restrictions that apply to the usage of the Darcy Weisbach formula and would apply it in my calculation

If you are to suggest Mr Art which equation do you feel if more accurate for calculating supercritical CO2 flow in pipes?I am looking into the Hazen's formula and would like to compare other methods.

Thanks