Grinding & Classification Circuits

Grinding & Classification Circuits

  • To participate in the 911Metallurgist Forums, be sure to JOINLOGIN
  • Use Add New Topic to ask a New Question/Discussion about Grinding.
  • OR Select a Topic that Interests you.
  • Use Add Reply = to Reply/Participate in a Topic/Discussion (most frequent).
    Using Add Reply allows you to Attach Images or PDF files and provide a more complete input.
  • Use Add Comment = to comment on someone else’s Reply in an already active Topic/Discussion.

Pumps with troubles (6 replies and 5 comments)

P
Pabloski
7 years ago
Pabloski 7 years ago

Hello everyone,

After the SAG mill there is a SAG box, where the output of the SAG is transferred into. There are 4 pumps (2 operational and 2 in stand-by) that transfers the material to a cyclones battery and then to two ball mills.

When the material is fed into the SAG box, its distribution is irregular making that the pumps suck different particle sizes resulting in an irregular feed to both of the cyclones and to the ball mills. The objective is to have a mirror-type operation for each ball mill, as well as to decrease reactive maintenances for pumps.

The drawing of the SAG box along witht the pumps is attached.

Has anyone had a similar problem in their plant? or maybe some ideas to control the feed and the particles distribution in the pumps?

Thanks before hand!

sag box v
https://www.911metallurgist.com/wp-content/uploads/2017/03/SAG-box-v.jpg
A
AJNeale
7 years ago
AJNeale 7 years ago
1 like by David

Hi Pabloski - it's an age old problem, and in your case made worse by having the pumps on the side of the pumpbox, as opposed to the front (i.e. the side furthest away from the mill).  The segregation coming out of the SAG mill is due to the rotation of the mill, so the split changes if and when you change the rotation of the mill.  Do you ever reverse the direction of the SAG mill?

My guess is that the problem is made worse by the sloping floor, and you'll see higher segregation when pumps 2 & 3 are running, or when pumps 1 & 4 are running.  One option to consider is to move Pump #1 and Pump #2 to the front of the pumpbox.

Could you clarify what happens to the ball mill discharge, does it come back to this same pumpbox, or is there a second set of pumps and cyclones for ball mill discharge?  This issue is all about mixing and homogenization, so what you do with your ball mill discharge will have an impact on any proposed solutions.

Best Regards

Andrew

P
Pabloski
7 years ago

Hi Andrew,

The ball mill discharge makes a close loop with the hydrocyclones, doesn´t goes back to the same sump. SAG has its own sump, and then its transferred by pumps to 2 different sumps for each ball mil and hydrocyclones batteryl.

You are totally right about the segregation problem when different pumps operate, its a real issue here.

As the ball mill discharge is not mixed with the SAG discharge, I dont see it as an issue on a future solution. And as you said, its all about mixing and homogenization... my guess is that there should be only one big pump and dsitrubute the feed to the ball mill/hydrocyclones sump by means of valves, that maybe would help to homogenize.

Many thanks for your inputs Andrew.

Best regards,
Pablo.

D
DavidS
7 years ago
DavidS 7 years ago

Hi Pabloski,

I have had a similar experience where the SAG sump offtakes were at different heights.  The product pumped from each offtake was vastly different with regard to sizing and mineral composition.

We ended up running the two ball mills as separate units.  One had a single size charge added and discharged back into the SAG sump.  The other had a mixed charge added consisting of 50% smaller balls and had it's own sump.  Each had it's own cyclone nest as well with the same number and size cyclones but the one with a discrete sump obviously used fewer cyclones at any one time.

The products of each was also treated in it's own rougher circuit.

P
Pabloski
7 years ago

Thanks for your inputs DavidS!

So basically instead of changing your pumps configuration, it was decided to adapt the process to the issue.

Why you ended up running the ball mills as separate units? what was the decision maker there? did you had different but constant compositions for each feed that allow you to adapt your process to two parallel grinding processes?

Thanks!

Pablo.

b
Robert
7 years ago
Robert 7 years ago
1 like by David

I've searched out solutions to this problem and probably have records of several in files. The disappointment was that in most cases the operations after making steps forward were still not entirely pleased with performance and were still searching to further improve the split. Solutions included:

  • modified sumps - controlling positions of discharge lines to pumps, where material was dumped into sump, and design on the floor slopes.
  • discharge launders from SAG/Screen with control arrangement schemes (e.g., diverter arrangements or dart valve arrangements with control action based on ball mill load or related parameter)
  • pressurized pot arrangement (e.g., used at Escondida to split from SAG to 3 ball mills). 

Given particle sizes it's a difficult problem. Even when solution works for a time these really need good maintenance. As the load to different ball mill changes it's often possible to see major differences in work being done.

P
Pabloski
7 years ago

Hi Robert, thanks for taking the time to reply.

Any idea its a good idea right now. I could find anything on pressurized pot arrangements. By any chance do you have some information that you could provide me about it?

Many thanks in advance.

Cheers,
Pablo.

David
7 years ago
David 7 years ago

Pabloski, sadly your pump box is a "poor" design. Why are 2 pumps up high on the slope? Would have been better 2 on the sides and 2 on the front so all at the same elevation. I can certainly see sanding problems as well in the current configuration.

Maybe, try to local the discharge pipe of your slurry to maximize box turbulence/agitation. Not sure when in that box that would be, but maybe on the size of #3 & #4 to make sure the never sand-up and those are always 'live' for you to use.

Follow Andrew's lead.

Use the Social Share Bar on the Left. Tell everyone you can about https://www.911metallurgist.com/metallurgy/ It's FREE & GOOD.

P
Pabloski
7 years ago

We just to have a nice system before, it was a distribution box with 4 valves, so you just opened the valves the right amount and you could have a good particles distribution and a more constant feed flux, at least that´s what I´ve heard because I was not working here by that time. Now you have this pumps pumping everything they receive, it doesn´t seems like a way to control the process to me, but yeah, its a way to increase throughput to the plant.

I´m really not sure about moving the discharge pipe to the lower pumps, makes me think that #3 and #4 would have a lot troubles with big particles and they may become maintence dept. "favorites".. its just an assumption of course, it may work due to turbulence factors.

I think we need urgently a simulation of what the particles distribution looks like inside the box, but I´m not sure if that´s even possible.

Thanks for reply

C
czambrano
7 years ago
czambrano 7 years ago
1 like by David

One addition to David´s proposal to create more turbulence is to move the water addition to the pumpbox just on top of accumulation sites of the box. Another parameter to see is the level of slurry in the box which facilitates localized settling.

P
Pabloski
7 years ago

That would work, but the set point level of slurry is 60%, and probably accumulation sites are way down to change their position and disperse them into the slurry.
Pumps are difficult to move, I would go with a smarter feed into the box, install some distributor inside the box so to maintain similar particle sizes for each pump, a sort of riffle splitter inside.

Thanks for reply.

K
kam81
7 years ago
kam81 7 years ago

Using an intermediate chute or box to evenly distribute the feed into the pump box could help. The main problem however is the floor slope and suction locations.

If you don't mind losing some volume and suction head, and depending on your suction piping arrangement, you could raise the suction nozzle of pump #3/4 to match #1/2, and get creative filling in the floor.

A basic discrete phase CFD model could simulate your particle distribution with reasonable accuracy.

Please join and login to participate and leave a comment.