Pressure drop for Slurries

[chogben]

Greetings,

I'm interested in some feedback on methods people have used to determine friction loss in slurry piping.

The applications have ore from -28 mm to +1.5 mm with a specific gravity of 2.5 and a dense medium for separation with a d50 of approximately 30 microns and a specific gravity of 7.

There are some people in the office that prefer the Hazen-Williams equation and those like myself that were trained with the D'Arcy equation.

What are the experiences of the forum?

Regards,

Chris

 

[SlideRuleEra]

Hazen & Williams is an empirical formula based on a fluid having a kinematic viscosity of 1.13 centistokes (water at 60 degrees F). It can be used for any liquid, but viscosity must be essentially the above value, or accuracy declines. I use H&W whenever practical, but it does have this limitation.

http://www.SlideRuleEra.net

 

[stanier]

Invest in Slurry transport Using Centrifugal Pumps by WIlson Addie and Clift. It will be the best couple hundred dollars you spend for this technology.

Hazen Williams is inappropriate for a slurry as you dscribed.

The performance of a slurry pipeline depends on
Size of particles
Size distribution of particles
Density of particles
Density distribution of aprticles (not all sizes have the same density)
Density of carrier fluid
% fines that may go to making up a fines slurry to rpevent deposition
Temperature
Velocity
Density hold up in pipe line (what goes in = what comes out but the bit in the middle will have a higher density due to hold up)
Roughness coefficient of pipe material
Diameter
Shape of particles (affects hindered settling characteristics)
On top of that you have to consider the starting and stopping of the pipeline. Its like a Jombo jet the hard bits are take off and landing, flying high is relatively easy.

 

[chogben]

Thanks for the advice!

I have read through some of "Slurry Transport Using Centrifugal Pumps" and concluded myself that H-W wasn't the best fit, but there are those who are wiser with a few more grey hairs that have experience with H-W in the office and prefer it to other techniques.

Cheers

 

[Artisi]

look for info on slurry pumping from Warman, GIW, and other manufacturers, they have spent a lot of time on the practicality of it all and understand the problems.

Naresuan University
Phitsanulok
Thailand

 

[sprintcar]

GIW offers a class once a year on slurry transport - it might be worth it for you. Here's the link

http://www.giwindustries.com/course.html

I went a few years ago - very technical but enjoyable too.

"If A equals success, then the formula is: A = X + Y + Z, X is work. Y is play. Z is keep your mouth shut."
-- by Albert Einstein

 

[stanier]

GIW also have a series of technical articles that they will compile and send to you if you demonstrate a legitimate interest.

There are also plenty of papers on theinternet. Authors I would recommend are Venton, Cowper, Boger, Wilson, Addie, Swift, Pullen, Jacobs.

Be wary of the mathematics in the papers. I recall a PhD student thesis (University of Melbourne Suspensions with a Yield Stress) where the student actually dissected all the published math. He found only three papers were correct!

 

[chogben]

Stainer,

thanks for the recommendations, and the caution about the math! I have a good starting point with all of the recommendations.

All I need to do now is convice others in the office that there is more out there than Hazen-Willams for slurry flow.

Regards,

 

[pumpmanagement]

Warman Pumps used to produce a slurry pumping handbook with all the information necessary.
I am afraid they are rather rare, but it would be worth trying them.
They are now part of the Weir Group, so you may find they are more reluctant to spend money on general technical support literature.
try the following

http://www.warman.com

http://www.weirwarman.com/co.uk

Brian Conway

http://www.pumpmanagement.com

 

[jici]

Metso can supply a pump sizing software called PumpDim (used to be the Svedala one). It is based on the Hazen-Williams method. But I prefer the GIW pump sizing software (called Slysel). It is more accurate and will handle differences in pressure loss between vertical, horizontal and sloped piping. The only drawback with the GIW software is that it is not too user-friendly.

For Slysel, go to

http://www.giwindustries.com/slysel.html

For PumpDim, you have to contact a Metso pump rep.

Both software can also do pressure loss calcs without sizing a pump.

 

[stanier]

Re Slysel, its better that it is not user friendly as only those who know the subject will use it. Programs that are too easy tempt others who dont know the subject to just put in numbers and then believe the answers.

SLurry modelling is very difficult and complex. I suggest to any new to the subject to follow the examples in a textbook and do the calculations manually first. I always cross check calculations from software manually before embarking on sensistivity analysis.

Undertaking a series of sensitivity analysis is essential as the number of parameters is so great. When a pipeline stops and starts these parameters change and you have to ensure the specified equipment can meet these requirements.

 

[bulkhandling]

Chris,

The EnviroTech's Slurry Pump Manual and Metso's software recommends H&W methods while Wilson (GIW) and other modern hydraulics use Darcy formula.

I assume that your application is in-plant service and the pipe length if normally with a few hundred feet. In this case the difference between the H&W and Darcy is not a major concern. The concern is how you select the C value for H&W or contingency for Darcy.

Although the slurry software Slysel or Pumpdim may give you some reasonable results for most slurry applications, you may need to take some special considerations for your application that has two types of solids in the slurry: dense medium and gravels.

When a D50 of solids gets down to 30 micron, I may need to consider the possibility of increased viscosity especially when the solid volume percentage is high. In your case, the solids SG is very high (7.0) and it is used as dense medium (dense medium separation? normally the very fine part <10 micron is screened out), I believe the viscosity is not a concern, that is, similar to water. So the job now is to size a right pipe to make sure no settling will happen. You may use both methods (Durand, Wilson) and make sure your velocity is higher than 1.5x of any of the methods.

The next step is to check the settling velocity of the gravels. The fluid SG is the mixture of water + dense medium. If your application is for dense medium separation, you may need consider the gravels as two parts: low specific gravity one and high specific gravity. You do not have problem with the low SG part for sure since it should be lower or close to the fluid SG. You have to check and make sure there will not be major settling of the heavier particles.

If there will be no settling and viscosity is not a problem, the friction loss in pipe will not be too far from that of water, with contingency you need to consider for wearing in pipe (rougher wall), impeller and casing.

 

[Artisi]

"bulkhandling (Mechanical)"

"with contingency you need to consider for wearing in pipe (rougher wall), impeller and casing."

I cannot agree with this part of your statement, from my experience, when handling solids, especially fine mnaterial, the surface on the pipe internals, impellers and casings are usually polished to a mirror finish. There maybe some local scoring in areas of the impeller and on the casing - but generally these are also polished to a high level.

Naresuan University
Phitsanulok
Thailand

 

[chogben]

bulkhandling,

do you have a link or contact for the Enviro-tech Slurry Manual? I have some of the GIW information that uses the D'Arcy friction factor and would be interested in seeing a presentation of slurry piping using the Hazen-Willams formula.

Regards

 

[sprintcar]

Why not contact your pump supplier and see what they can offer?

"If A equals success, then the formula is: A = X + Y + Z, X is work. Y is play. Z is keep your mouth shut."
-- by Albert Einstein

 

[bulkhandling]

Chogben,

you can ask local Weir rep for a copy of Enviro-tech Slurry Manual.
No matter Darcy or H&W, it's just a method to calculate the piping friction loss based on WATER if we are not talking about a viscous non-settling slurry. Like the older method Durand, the newer GIW, or the approx approach using "C" factor in H&W (often used by slurry engineers for in-plant pumping), they are all actually trying to find a correction factor over the water friction loss.

Artisi,

I think we are talking about normal carbon steel for piping. When you say the fine solids will polish the pipe internal surface, I normally even not get to that micro accuracy. No matter of fine or rough solids, you can hardly see a even wearing in a in-plant pipe. "Outer" side of down stream of a elbow normal has the worst wearing; Bottom side has worse wearing than top side (that's why you may see people rotate a field pipe to extend its life). You may now agree that the worn pipe is "rougher" than the new one?

If an open impeller is used for a slurry service, the vane tips worn out quickly - you get reduced impeller trim!! For closed impellers, the vanes get thinner and narrower- even if you do not think "local scoring" is a problem. If you have a fixed speed pump for a constant flowrate, you can find the liquid level in the pumpbox get higher and higher - an indication of reduced pump efficiency - and you eventually have to repalce the worn out impeller.

 

[Artisi]

bulkhandling

" Artisi (Mechanical) 4 Jan 06 3:23
"bulkhandling (Mechanical)"

"with contingency you need to consider for wearing in pipe (rougher wall), impeller and casing."

I cannot agree with this part of your statement, from my experience, when handling solids, especially fine mnaterial, the surface on the pipe internals, impellers and casings are usually polished to a mirror finish. There maybe some local scoring in areas of the impeller and on the casing - but generally these are also polished to a high level. "

Just to clarify my comments above, I was referring to the comment "rougher" in terms of pipeline, in my experience the pipe line internals will get smoother rather than "rougher".

As for the gradualy destruction of the impeller, wear plates, casing etc. I can only agree with your comments and would add the following - the finer the material being handled the greater the wear rate, that's why rubber lined pumps are usually the first choice for fines handling. Although good hydraulic design and the high chrome irons exceeding 700 Brinnel can now be used successfully in many fines pumping applications, likewise thick rubber lined pumps are proving successful in many area's that were once exclusively hard metal pump applications.


Naresuan University
Phitsanulok
Thailand

 

[katmar]

Chogben,

Sorry to jump in so late - been away for a while.

Years ago, while designing the piping for a fractional crystallization plant, I used a method published by Lin and the results I got were good (i.e. the plant worked!).

It is a reasonably simple and usable method.

The reference for Lin's method is
Chemical Engineering, May 17, 1982, pages 115-117. There were some errata published in the Nov 29, 1982 (page 5) edition, but they are minor and obvious errors in the example.

regards
Katmar

 

[bulkhandling]

Artisi,

Different types of slurries behave quite differently in piping. For a well-suspended fine slurry, the wearing in pipe is normally not a major concern for friction, or as you said, it even polishes the pipe surface. For heavier and larger solid slurries, it can be totally different. The velocity is higher because of greater critical velocity to suspend the solids. Sliding on the bottom of pipe, gouging, impact can all become problems. Erosion and abrasion can be very serious. The pipe wall surfaces do get rougher in many services. I did see sliding lines in a coal power plant discharge pipe (not remember how many years' use) where the slag particles have very sharp edges (the velocity might be not high enough since there was a sliding bed obviously). It's up to you to say it was fines handling or coarse one but the average solid size is below 1mm (do not remember exactly how much).

The behaviors on pump impellers are also quite different. You are right, we only use rubber-lined pumps for fines handling (<5~6mm as a rule of thumb) , but it's not because the fine slurry has greater wear rate (which is not necessarily true - see next paragraph). Rubber liner has a better wear-resistance than steel, but it's bad at to impact, gouging, edge-cutting, which are major problems in the coarse applications.

When we come to check if the fines or coarse solids have greater wear rate, let's go to visit some mining process plants where they bought all steel impellers for all slurry services since some major suppliers (GIW is one) only supply steel impellers. The grinding circuits handle coarse ore and flotation circuits and tailing pumps handle finer solids. The impellers in grinding pumps have normally two months to six months life and casing may double the impeller life. The minimum impeller life for flotation and tailing pumps are over a year as I know (many last for mine life). Although lower solids percentage in flotation can be a major factor, but many final tailings get close solid percentage to grinding and life is still much longer. There are also many other factors affecting the impeller life, so I do not want draw a conclusion here as what kind of slurry will have greater wear rate.

 

[bulkhandling]

It seems that there is nothing wrong with the performance of the pump and with the suction line.

I'm thinking the possibility of shear-thickening.

Shear-thickening happens on some fine slurries especially with higher solid concentration. When shear rate increases, liquid viscosity increases. The thickened fluid will increase the friction loss starting from the pump casing to all along the piping length. When the pump speeds-up (increased shear rate), it aggravated the shear-thickening and further increased the viscosity and friction loss in the pipe.

To check if the fluid viscosity is higher at the first dosing point should give some indication if the above assumption is true or not.