Slurry pumping in a vertical pipe 1

[firefighter26m]

I am pumping an iron ore slurry approximately 60' through a vertical pipe. After calculating the limiting settling velocity I have found the actual velocity to be roughly 1 ft/sec slower than the settling velocity, I am trying to avoid dropping pipe size due to a drastic rise in velocity.

My question is, since the particles have a very limited possiblity of settling out due to the upward movement of the water, is the limiting settling velocity a major concern in vertical pipes? I have looked thrugh Warman Slurry Pumping Handbook as well as the Slurry Systems Handbook, however neither cover the issue, they are strictly horizontal piping. Would it be better to simply reduce the pipe size, to speed the actual velocity up above the settling velocity, until the last couple feet at which point I could enlarge it again to slow down the fluid flow.

Any advice is appreciated.

 

[BigInch]

I'm not claiming to know anything about slurry pumping, but your question does raise one red flag when you say... "drastic rise in velocity". Therefore, due to the effects normally accompanying a drastic increase in velocity, I would say, that it is probably best to avoid that, but I would also say it would depend on what that meant in quantitive terms.

In an effort to get better advice from regular slurry pumpers, why not specifically state the exact pipe size (ID), the settling velocity, and the fluid velocity you have now.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[Artisi]

Try searching google or

http://www.ask.com

- type in settling velocities in vertical pipes - or similar description - there is a wealth of information there (I think)

I came up with the following - might be worth a try.

http://internet.processmanual.com/ipmcontent/SLRY_2.htm

 

[BigInch]

Artisi, at first glance that looks like a very good reference. It has inspired me to read up on this aspect of pipelining.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[firefighter26m]

Thank-you for the link, I am looking into a licence for the site, definately some useful info there.

The specs for the system I have are as follows:
-pipe length (actual) is 60' all of which is vertical discharging into a bottom fed distributor
-Actual slurry velocity is 4.25ft/sec
-Settling velocity is 5.86 ft/sec
-Concentration by volume is 5.94%
-Maximum d50 size is 210 micron
-Flow rate is 93.57 GPM
-Pipe size is 3"

My thinking is that, although the slurry velocity is less than the settling velocity, the vertical flow and low quantity of solids, will keep the particles in suspension, allowing me to use velocities slightly lower than the settling velocity. I don't have a ton of experience in this area so am hoping others might be able to offer a suggestion. If I downsize to a 2" pipe, my velocities jump to 9.56 ft/sec actual and 4.79ft/sec settling.

 

[davefitz]

You might be able to increase the effective velocity , and improve entrainment, by isnatlling a spiral wound ribbon turbulator inside the pipe.

 

[dcasto]

Try 2 3/8 tubing

 

[moltenmetal]

firefighter: depending on your material, you actually CAN get 2.5" NPS pipe. Or you can go up in schedule on the 3"- that'd give you more erosion allowance as well. Or if erosion IS an issue, you could go with rubber-lined 3"- that would give you a bit smaller diameter and higher velocity.

I'm sorry, I have difficulty wrapping my head around what physically happens to a "rocks in water"-type, fast-settling slurry in a vertical pipe, flowing at a velocity less than the settling velocity for the largest diameter particles. Does it function as a particle size classifier, ultimately leading to plugging? Or do the larger particles merely get entrained in slugs out the pipe? Or does something else happen? If you find out from credible sources, I'd appreciate a description!

 

[BigInch]

I would think it would look like a coffee percolator, but with rocks shooting up then snowing down. I would also quess that there are several flow regimes depending on velocity, the particle size, effective density of the mix the density:viscosity of the liquid only which will also vary with the slope of the pipe. Then again, I'm not a credible source.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[firefighter26m]

I do plan to use rubber lined for the entire system as there is an erosion issue, I will check to see the availability of the 2.5" however. My fear is that it will act like a particle classifier should the particles become too large. I was more curious that since you can not truly get a 'sand dune' effect in a vertical pipe, would the settling velocity drop...however thinking now about the classifier effect, it might actually increase the minimum settling velocity. I think I will use alternate size pipe to stay on the safe side though....that or I need to come up with a way to gain 60' elevation through horizontal pipe to which I know the calculations for...lol

 

[katmar]

FireFighter, are you sure of your settling velocity? A settling velocity of 5+ ft/s sounds like 1/2" steel ball bearings and this does not square with your d50 size of 210 micron (i.e. less than 1/100"). It would make more sense if the units for your settling velocity were inch/s rather than ft/s, in which case your 3" pipe is no problem. How did you get to your settling velocity - was this a calculated or measured value?

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[firefighter26m]

It was calculated using the Durand's formula for settling velocities of particles of widely graded sizing (from Warman's Slurry Pumping Handbook). The formula is:
Settling Velocity = FL * square root of (2gD((SGs-SGL)/SGL))
where FL is obtained from Durand's Modified Limiting Settling Velocity Parameter, g is gravitational constant of 32.2, D is pipe diameter in feet, SGs is specific gravity of the solids, and SGL is Specific grvaity of the carrying liquid.

In our case SGs = 3.27 and SGL is water and therefore = 1

From what I see in the book and experience with the iron ore and waste material, I suspect the settling velocity is pretty close to accurate (of course in saying that, it wouldn't be the first time I was completely wrong either). I have since increased the rubber lining thickness slightly in the piping and have made the velocities better balance out so that I am above limiting settling velocities in all aspects of the pipng. I am still curious about the vertical flow though if anyone has any information, advice, or comments. Thank-you for all the advice and comments so far though, it has definately been a help.

 

[katmar]

What is the maximum particle size you are allowing for? With the SG's that you have now provided, I estimate that the particle size for a sphere with a settling velocity of 5.8 ft/s is almost 2 inches diameter! What units does your equation specify for the Settling Velocity? If I take your 5.86 as inch/s and not ft/s I calculate a particle size of 750 micron, which would make more sense with your d50 of 210 micron.

Take a few of your ore particles to the lab and watch them settle in a tall measuring cylinder. There is nothing to beat seeing it with your own eyes. This will give you an appreciation of just how fast 5.86 ft/s really is.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[firefighter26m]

Based on the units in the equation...gravity is in ft/sec squared and pipe diameter is in feet...giving you feet squared per second squared...take the square root and you are left with feet per second.

Out of curiosity, how are you calculating the sphere size? I am thinking that you are actually calculating bouyancy effect rather than slurry settling velocity, two differant effects. In the case of terminal velocity of spherical particles (where you calculate the rate of fall in a vertical chanber with no liquid velocity) the determination of a 2" diameter sphere would be accurate for a velocity of 5+ft/sec. However for limiting settling velocity, the speed of the slurry is being calculted to keep the spheres in suspension, essentially how fast the carrying liquid is moving over the suspended particles in order to create a bouyancy effect. As the particles are moving, as well as the carrying liquid, the speeds are relatively high in order to get a velocity differential between the two that is high enough to prevent the particles from settling (similar to an aircraft wing travelling through wind that is blowing in the same direction as the aircraft is flying)

 

[firefighter26m]

Katmar...I have been thinking more and more about your post and think you may have solved it for me. In the horizontal piping, the limiting settling velocity is greatly increased due to the requirement for the velocity differential. In the vertical plane however, the settling velocity would be equal to the terminal sinking velocity of the largest particles since they can not settle out as they do in the horizontal plane. Therefore if the upward flow of water matchs the terminal sinking velocity, the particles will stay in the same position, not gaining or losing height, and therefore, any upward velocity greater than the terminal sinking velocity will carry the particles upwards. Since the limiting settling velocity on the horizontal plane will always be greater than the terminal sinking velocity (vertical plane), the velocity used in the horizontal sections will always be enough to carry the particles upwards assuming the piping is the same diameter, etc.

Thank-you for all your help, this has been greatly appreciated.

 

[katmar]

Your dimensional analysis is correct, but what about the FL factor? I do not have a reference to the Durand Equation here so I cannot comment further - it just seemed to me that the calculated velocity was way too high. I will not be back in my office until the end of the week, and then I can check on various references I have there.

Yes, I am calculating the terminal velocity. In my opinion it is the same thing because all you are interested in is the relative velocity between the particle and the fluid. How the fluid is moving is irrelevant IMO. The only difference between terminal velocity calculations and settling velocity in a pipe is that the terminal velocity calc looks at a single particle, while in a pipe system you are interested in hindered settling. In hindered settling there are so many particles that as they settle they cause a significant upflow of fluid which hinders the particles. Hindered settling rates are therefore lower than calculated terminal velocities and I would expect my TV calculations to be conservative.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[HEC]

firefighter
It seems you have your answers, I have just noticed your post. One comment in an earlier post in relation to the low solids content to keep the slurry in suspension. This seems to be correct however ther are some exceptions. A noteable exception was with a kaolin clay that remained in suspension above 25% solids. Below 20% solids we had trouble with the clay settling out. This limit was established with settling tests at different percent solids. The application was to solve an erosion problem, we had been dosing small amounts of clay from a recirculation pipe at high velocity to keep the clay in suspension. By keeping the solids high we were able to reduce the velocity and the erosion problem. Relevence here, check the behaviour of the slurry across a range of solids content, it may behave counter to your assumptions.

Mark Hutton

 

[firefighter26m]

Thanks Mark, I will definately run some tests once the entire system is up and running.

The main pumping system pumps the slurry at 40% solids by weight, then we seperate out the iron (approx 48%) and recirculate 25% of the waste material through a separate line, so the recirculation is a relatively low concentration of solids. It'll be interesting to see how the two differant concentrations react.

 

[katmar]

FireFighter, I am back in my office and have been able to get hold of my references. As you said in your first post, the correlations you have used are strictly for horizontal piping. They will therefore predict much higher velocities than are necessary for vertical piping because in a horizontal pipe the liquid flows at right angles to the direction of settling. It seems that the general suggestion for vertical piping is to use a velocity of at least twice the free settling velocity.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[firefighter26m]

Thank you Katmar...with the horizontal velocities being as high as they are, they will more than be adequate then to suspend the particles. Thanks again for your help.