Grinding & Classification Circuits

Grinding & Classification Circuits 2017-03-23T09:46:37+00:00
  • 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.

Using Abrasion Index for calculating Steel Balls and mill liners consumption (35 replies and 2 comments)

John Koenig
1 year ago
John Koenig 1 year ago

Are the old Allis-Chalmers Abrasion Index and calculations for Steel Balls and liners consumption still valid?  The Allis-Chalmers Standard Abrasion test has been used for several decades to calculate steel consumption. Current advances in alloys composition and on heat treatment may be producing steel balls and liners that would last longer than material in the past.  The equations in use today were derived from plant experience data, which was obtained excluding those places were Ni-hard material was used.

This equation is used to determine the steel balls consumption , for example.

Balls: lb/kwh-hr=0.35 (Ai-0.015)^0.33

I know of a practice that uses 50 percent of the theoretical consumption calculated by the Bond formula for ball mills; for SAG Mills 65 percent of the theoretical consumption is used considering the SAG mill as a ball mill.The rotating paddle used in the Bond Ai test is made of SAE4325 steel hardened to 500 Brinell and therefore not necessarily equal to the hardness and abrasion resistance of commercial grinding media.Grinding media producers may have in-house correlation factors of the Bond test steel and their own grinding steel wear rates but I have never seen any published data on this. Bond Ai's can vary widely on account of the difference in mineralogy between the particles tested where as media consumption operating plants vary typically only in a fairly narrow range.
Development of a more indicative wear rate test procedure would be in order.

Does anyone know of any paper documenting the practices you mention above? The interesting thing for me is that if I apply the factors mentioned above, everything start making sense to me for data that I have at the moment.
or do you know of any current projects or study where this has been used.

Bill Fraser
1 year ago
Bill Fraser 1 year ago

Moly-Cop have developed a new Abrasion correlation using Ai, F80 and pH, this new correlation has a 9% error in comparison with Bond which has 60% error. Our new model was published in several symposiums like Procemin 2013 and is going to be published in Copper 2013 and IMPC 2014.

In the paper we published by Moly-Cop, there is a comparison between real consumption and Bond predictions, and average error is about 60%. MoyCop published 46 real data and we have measured the Ai for each application.

Alan Carter
1 year ago
Alan Carter 1 year ago

1. In practice during 1978 to 2000---in a commercial plant of Pb-Zn beneficiation plant of 4000 tpd used hypersteel with low Cr. R&D in commercial plant proved that High Cr gave good results. 50%reduction in consumption, but double the cost. The main advantages of using high Cr is that it retains shape of ball till end. Only top size of 200mm is added for makeup.

1. Ball consumption----

a. When we had cyclone with ceramic/rubber apex consumption was high, due to more recirculation. But after we used Silicon carbide apex cost reduced drastically. Apex wear very less as compared to ceramic and rubber.

b. When feed size was 16mm ball consumption was high. When we used 12mm feed it reduced drastically.

c. Wt% solids in ball mill : At high wt% solids consumption was high. But at normal wt%solids it was low.

Similarly consumption depends on operating parameters which need to experienced by an operator, or do lab R&D tests.

Initial charge is mix of of all sizes.

John Koenig
1 year ago
John Koenig 1 year ago

I think that Molycop does a great job supporting their clients. I read the paper that you mention and I use the data to compare to the Ai that I obtained. I used the Original Bond equation and my estimated steel consumption was very high, almost unreal. Molycop's work has helped me to obtain data that is more reasonable 

The Ai tests are not very expensive. Any operation trying to improve their steel consumption should develop more work. Molycop has excellent tools, they are simple and powerful. Thanks a lot.

Bill Fraser
1 year ago
Bill Fraser 1 year ago

Regarding the model we have developed is only for using Moly-Cop forged balls for cast balls of low quality balls you have to correct due cast balls low quality balls usually has higher grinding consumption. In resume the model is dependent on the grinding media type.

Alan Carter
1 year ago
Alan Carter 1 year ago

Bill,

Moly-cop is new name of Armco automatrics.

  • Only proprietary grinding media in the world--who produce Exclusive chemical composition of media and hardness is guaranteed by armco.
  • I don't have much experience with Moly-cop. Please send chemical composition and hardness of both types of media.

Can you give few case studies in commercial application and their consumption, with ore hardness.

Bill Fraser
1 year ago
Bill Fraser 1 year ago

Alan, you are right Moly-cop is now the formerly old brand Armco Grinding Systems, in order to sent you grinding media specifications please sent me a PM.

JohnnyD
1 year ago
JohnnyD 1 year ago

This characteristic of ball media wear rate vs the bond index was investigate at part of the AMIRA P9L project 1996-1999.

It was found the systems dominated by corrosive wear mechanisms (ie: mild steel media) had wear rate in the 100-159g/kwh range and the systems that did not have corrosive wear mechanisms (Cr media or non-corrosive ores) and were dominated by abrasive and impact wear mechanisms had wear rate in the 50-100 g/kwh range.

“The objectives of this research project on media wear were:
(i) to define a total wear model incorporating abrasive, corrosive and impact wear mechanisms,
(ii) to develop three ore-metal-environment specific laboratory tests to determine model parameters,
(iii) to calibrate and verify the model with real system data.

All three objectives were attained and can be described as follows:
(i) A total media wear model was defined incorporating abrasive, corrosive and impact wear mechanisms. The total media wear model is of a decoupled nature which permits the identification of each wear mechanism individually. However, this decoupling assumes that experimentally such a decoupling is possible and will not greatly affect the resulting analysis.
(ii) Three ore-metal-environment specific laboratory tests were developed in order to identify the model parameters for this decoupled total media wear model.
(iii) In the case of impact, the laboratory set-up did not attain the desired impact energy levels. However, some experimental data was generated allowing the definition of a impact wear/impact energy relationship for different mill media. Further, it should be noted that this relationship does not reflect the expected relationship for impact wear as a function of impact energy.
(iv) In the case of abrasion, the laboratory set-up performed as expected in the range of applied abrasion forces. However, these forces were greater than those experienced in a batch mill and less than those experienced in a real mill. This undoubtedly contributed to the use of different correction factors (0.0235 for the batch mill and 0.2266 for the real mills).
(v) In the case of corrosion, the stainless steel batch mill test also performed as expected. However, determining the specific corrosion rate is dependent specifically on the assumption that decoupling works and on the abrasion wear test data determined independently.
(vi) Model calibration and verification, of course, is dependent on the data obtained from industry. However, as these tests were of an exploratory nature, it can be said that this was achieved with a measure of success. The total media wear model is sensitive to changes in media, ore and environment and illustrated total wear as a function of the proportions in impact, abrasion and corrosion wear. The jury is still out on this decoupled approach to predicting mill media wear.”

The papers that can be referenced on this work are shown below:

References
Blickensderfer, Tylczak (1989) Evaluation of commercial US grinding balls by laboratory impact and abrasion tests, Miner. Metall. Process 6:60-66.
Mishra, Rajamani (1994) Simulation of charge motion in ball mills, Part 2: numerical simulations, Int. J Min. Proc., 40:187-197.
Misra, Finnie (1980) A classification of three-body abrasive wear and design of a new tester, Wear 60:111-121.
Radziszewski P, (1997) Chapter 4, Ball Mill Media Wear and Prediction, JKMRC/AMIRA P9L Project report, December, 35-51.
Radziszewski, (1998a) Chapter 5, Preliminary Results from Media Wear Tests, JKMRC/AMIRA P9L Project report, June, 75-86.
Radziszewski, (1998b) Chapter 6, Predicting Steel Wear in Ball Mills, JKMRC/AMIRA P9L Project report, December.
Radziszewski, Morrell (1998) Fundamental Discrete Element Charge Motion Model Validation, Minerals Engineering, vol. 11, no. 12, 1161-1178.
Rajagopal, Iwasaki (1992) The properties and performance of cast iron grinding media, Min. Process. Extrac. Metall. Rev.,vol. 11, 75-106.
Scieszka, Dutkiewicz (1991) Testing abrasive wear in mineral comminution, Int. J. Min. Process., 32:81-110.
Yelloji, Rao, Nararajan (1991) Factors influencing ball wear and flotation with respect to ore grinding, Min. Process. Extrac. Metall. Rev., vol. 7, 137-173

Tony Verdeschi
1 year ago
Tony Verdeschi 1 year ago

Just published is some Newmont data in a paper at SAG 2011 on both SAG and ball mill steel consumption. The ball mill wear data reflected a 0.5 factor on the Bond equation consistent with Arts observation above. I also like the Molycop equation Levi has discussed for ball mills although it does seem to reflect slightly higher wear rates than what we see in circuits where a ball mill follows a SAG mill. I expect particle angularity of crushed vs SAG milled ball mill feed has an impact on the wear rate. Also we observe for SAG mills that ore competence and feed size distribution, being major influence on rock to steel ratio in the mill, are significant factors that influence SAG mill steel wear. We've observed the abrasion index to be a less useful means of predicting SAG mill grinding media wear. I have not found the Bond models for mill liner wear to be useful and it's normal to assume a certain number of liner change outs per year, and factor this up and down depending on generalized assessments of the wear environment (consider ore hardness, feed size, ball size/surface area, abrasion index, mineralogy and water quality). I can email a copy if the paper if you don't have access to it.

Victor Bergman
1 year ago
Victor Bergman 1 year ago

Agree with the wear factors but we also need to add the media type as another variable. Media can vary as much as the ore in their properties - grain size, chemistry, hardness, stability, some associated with size of the media...media breakage from impact or spawling is, in some cases, a much greater steel consumer than abrasion.

We, like the other media designers, do have a large data base - based on actual consumption data that is available on a multitude of circuit/ore variations that should be even more accurate that a drafted formula. I do fully understand that the thought of devising a formula is most interesting for many of our technical colleagues however, we have 7 SAG mill ball grades which are for targeting a range of environments - each of them will perform uniquely in a similar grinding environment.

There are only two cases that I can think of that will need this projection:

  • feasibility studies for projected costs.
  • to check if consumption is within "acceptable" levels in terms of "bang for your buck".

Great to see the level of interest and the level of expertise that this forum attracts.

John Koenig
1 year ago
John Koenig 1 year ago

Victor,

The industry needs to learn about the different grinding media out there. It cost money and takes time but when someone is expending between 10 to 20 or may be more millions of dollars per year in grinding media I think that it is reasonable to start doing some test to optimize grinding media and to optimize this parameter. This optimization may come a long way to also help optimize grinding efficiency as a team naturally will start looking at more parameters around the grinding circuit as they look at the grinding media optimization.


Tony,

thanks for your comment. I find it very interesting. Can I have the paper name please?
The logical conclusion, after analyzing all of these nice comments, is that the alloys composition has evolved into more robust liners and steel balls which makes the Bond equations obsolete.
Not only Molycop but Newmont found that factors are needed to gauge steel consumption when trying to come up with numbers for greenfield projects or current operations.
Good work guys ! 

Tony Verdeschi
1 year ago
Tony Verdeschi 1 year ago

I agree that media quality is an influence although in many cases differences in media type are very hard to quantify amongst the background noise and this is why it is typically not considered in these models. I don't believe anyone has published a model that relates media spec to wear performance, in addition to the process noise there's a lot of variables to consider when ranking media quality.

Helena Russell
1 year ago
Helena Russell 1 year ago

I totally agree with the comments above regarding the ball / rock ratio in SAG mills and its impact on the ball wear rate (particularly breakage by impact). This is precisely what I experienced in one of the plants I worked at. We were, however, able to reduce the ball wear rate (total and breakage component) by changing the composition and hardness profile of the alloy.

Moly cop is the best. Experience after operating ball mills for the last 25 years show that Cast Hyperstell balls are not only poor in quality but also reduces grade and recovery due to deformed balls causing inefficient breakage and size distribution of broken solids of minerals. When ball retains shape griding is good and effective in all respects. More profitable. But CAOEX is more.(Double the price) consumption is 50% reduced, in Ball mill round balls of 75mm dia size, for topup.

Quality forged balls will always out perform cast iron media - from what we see these basic cast iron balls are very rarely used these days. Our African office sees them occasionally but price was the attraction but there is usually no value.

John Koenig
1 year ago
John Koenig 1 year ago

I have used the Molycop database and obtained the following updated Bond Equation

Ball Mill Media (kg/kWh) = 0.0944*Ai^0.299

Aidan Giblet and J.Seidel published a paper on the 2011 SAG conference (MEASURING, PREDICTING AND MANAGING GRINDING MEDIA WEAR) and they obtained

Ball Mill Media (kg/kWh) = 0.0817x(Ai)0.498

the two equations are very similar.
two or three groups of people are finding updates to the Bond equation in different places using different data bases.

Is it time to officially declare the Bond equation as obsolete? if you agree with this idea, How do we do to reach consensus and publish a paper to finally update these equations.
We need supporters to put a database together.
We need to replicate the 60s Bond work and create a magnificent new data base, then create the paper and then contact universities so they can update what they teach.
Any ideas or leads?

Tony Verdeschi
1 year ago
Tony Verdeschi 1 year ago

When you refer to the MolyCop database do you mean the data in the paper Levi referenced and co-authored? I was interested that ball size had a stronger correlation to steel wear than Ai in that data set, and also that the steel wear rates in his data was higher than the published Newmont data on average. I think all of his points to a useful correlation for ball mills, but an Ai model wont work for SAG mills as feed size, competence and charge composition are large influences. Water chemistry, feed condition (particle shape - preparation effects) and mineralogy will at times be significant factors influencing wear in ball mills. To close he book on this requires more effort than i expect the ndustry is willing to support. But we have undertaken some work in this area at the University of Queensland that i would be happy to discuss further. It is a long and expensive road, and will require a significant collaboration to progress.

John Koenig
1 year ago
John Koenig 1 year ago

Yes, I meant the paper that Bill and others published. 

I obtained the steel consumption in g/t from Bill. He makes clear that kg/kWh is the parameter that must be analyzed, which I support. I made a quick correlation with what we see at the end of the day, the Ai vs the gr/tonne, which must be similar to what you guys did. In essence the gr/tonne is not a good indicator as the mill inefficiencies are included in the g/t, which includes mill loads, F80, P80, operation variability, operator’s knowledge about the relationship between circulating loads and efficiency, etc.
Molycop obtained a new empirical equation for ball mills
KdE= 1.36 [(Ai-0.05)/0.20)^0.166 * (F80/5000)^0.069 * (pH/10)^-0.243]
This equation takes into consideration more parameters so as rule of thumb it should provide more accurate correlations between observed and measured variables.
In any case and to obtain a quick flavor about how well the Bond equations are predicting steel wear the comparison I did is valid as it provides a first comparison for steel consumption. A big majority of people is still using the old Bond equations without asking themselves: do 950 g/t or 1,100 g/t of steel consumption in the ball mills make sense? When we apply our experience from the last 20 years we see right away that steel consumption in the range of 1,000 g/t in a ball mill is not correct, and that is a fact. But people needs a tool to forecast costs and they keep using the old Bond equation because they don't have the authority to say this is the correct equation we should use or even worse they don't even have the data that is needed to update these equations.
The mining industry must fund a project like this. It is in the best interest of everyone. Developing work like this pays itself very quickly. The small investment is already paid for if a mine can save 1 million dollars per year in steel balls. A mine expending 20 to 40 million is steel balls per year may reach gains of this nature.
Having a big budget for steel media consumption, based on assumptions of higher consumption, leads to complacence while the money is worn out in the mill, literally speaking.
I completely agree with your comments on the factors having an effect on steel media wear and we can add the ball quality and size. I also agree with your comment about the degree of support that the mining industry is willing to provide to support work like this. I wish that CEO’s would hear of these discussions were mineral processors from all over the world are telling the industry that there is room for more savings and for energy efficient initiatives.
Cost control can help business where there are inefficiencies but I also believe that cutting costs may lead to prevent improvements activities, which from my perspective is not the way of doing business. No one wants to expend the money because everyone is trying to get the lowest budget possible, so pennies are saved while dollars are transformed into noise, heat and waste.

David Kano
1 year ago
David Kano 1 year ago

For greenfield studies I can understand the need to come up with a reliable method to estimate ball wear. However there are other ways to select grinding media for an existing operation. Has anyone conducted marked ball wear trials to measure actual wear rates in an industrial ball mill? In one trial that lasted about 8 months, we trialed several different ball types from three different suppliers, including cast and forged balls, all at the same time. The trial was done in a ball mill. The results were conclusive.

John Koenig
1 year ago
John Koenig 1 year ago

That is an Excellent comment David because it opens up some more discussion doors 

Please let us know the procedure you guys used to mark the balls and details about the test, if possible.
I think that you are correct from the perspective that you describe. The approach you mention will give an overall average of all the effects over steel consumption. Did you guys determine the effect of those parameters usually discussed in the posts in this discussion? For example, once the steel balls provider were selected, can one continue optimizing the steel consumption by moving the grinding parameters? For example, pH, reagents (some people call them grinding aids), improvements in classification (for example changing screen sizes or apex and vortex), changes in slurry % solids, etc. I think that Molycop has addressed some of these issues with what they have done when they obtained the update to the Bond equation ( Benavente’s Empirical Correlation) and there is also a new methodology used by JKTech that is in the market now. The JK technology was developed I McGill and I think that this is becoming a better know technique now.

Victor Bergman
1 year ago
Victor Bergman 1 year ago

At SGI we have a different view of marked ball tests (MBT).

Our reasoning is:

  1. Suppliers can send several test types - some on the edge of stability but high on anti abrasion - how many survived? no one knows as the test controllers only need to find a single ball to grade its result - how many broke?
  2. The sample hole is only so deep - is it not the full life of the ball that is of primary interest and not the first 30%? What if the test media had a relatively soft center or was not heat treated through to the center leaving the remaining 70% with accelerated wear?
  3. The quantities required for MBT are usually only a few hundred balls - easily made in laboratory (perfect) conditions and so are not representative of the suppliers ability in typical production runs.

If we are going to be accurate in using MBT as a TRUE benchmark then random independent sampling of supplier stock material is THE ONLY method for sample supply for such a test. We have seen MBT results that "conclude" that a clearly lesser ball design has come out on top.

I am happy to share this information if anyone wants to take this further - however the 3 points above do not defy logic.

David Kano
1 year ago
David Kano 1 year ago

Details of the trial is long. In summary this is how we have done it: each grinding media supplier (3 off) nominated the type of balls they wanted to include in the trial, and arranged all of their balls drilled and placed numbered tags in each hole. Holes were then filled up with low melting point alloy. A total of 11 different ball type was trialed. All balls all had the same nominal diameter as what was being used as the top up size in the actual mill. But each ball type had different average weight. Additionally, each and every ball was individually cleaned, dried and weighed and their weigh recorded against their tag numbers. We had about 200 balls for each ball type, supplied to us in sealed 200 L drums. One of the drums contained control ball, that is the same ball as was being used in the mill. Representative number of balls were taken as samples from each drum for analysis (spectrophotometry, there may be better and faster ways to analyse balls these days) to confirm the control ball is the same as the one currently used and have records of all other ball types. Hardness profiles were determined for each ball trial ball type. Once all preparations were complete, all balls were added into the mill at the same time. We collected balls by going inside the mill in a pre-planned mill shutdown (actually we arranged the timing of the ball collections to coincide with regular maintenance shuts. I needed about 4-6 hours, which was given. 4-6 people entered the mill after all procedures were complete (i.e. isolation, permits, ventilation, fluoro lights to provide light inside the mill, buckets, hoses, water to clean balls, etc..) We were able to collect balls from full surface of the charge initially, then dug the contents up to about a metre. To give you an example for one supplier who supplied us with 1200 balls (in 6 types) we were able to collect about 350 balls. As for the drills on the balls, some had one drill hole, some two holes at certain angle, some had three holes at certain angle again. All this information was used in the final analysis. The collected balls were washed, dried at about 70 degree and weighed and their tags were removed my melting the low melting point alloy and recovering the metal tag inside the hole. This was we knew the precise weight of the original ball and the weight after, say 1-3 months. We also measured the ball charge inside the ball mill before we started the trial and when the trial ended. We tried to control the ball charge constant throughput the trial. All other relevant operating parameters were also recorded. However we were primarily interested in how the trial balls compare against each other. I do not want to go into further details here in short space (and time) available, but this is basically how the trial was conducted. Understanding the theoretical side of the trial and coming up with a reliable method to analyse the results was also important. Unfortunately even when we did the trial, there were not a lot of good published papers in the topic. In the end when we plotted ball diameter loss against grams of weight loss per surface area for each ball, we came up with a regression coefficient of about 0.998. Some of the balls had more than 20 mm depth, which is more than acceptable given that the nominal ball diameter was 65 mm (32 mm radius).

The results were very conclusive, so that the company decided to follow up with full plant trial with the winner ball, and the full plant trial also confirmed the results of the MBT. 

John Koenig
1 year ago
John Koenig 1 year ago

Thanks a lot for the test description. It is a very succinct but powerful description of a test for steel media wear. I like the fact that the description is not only theoretical but that also proivdes an insight about the hard work needed to do these tests.

Did you guys had the opportunity to determine the ball "survival size" for each one of the suppliers?

I'm not conducting or planing to do any ball tests in the near future but we may need to develop a test program and it is interesting that we can apply the knowledge already available rather than re-inventing the powder gun again.
I think that it is vital that we understand about the "survival rate" as there are indications that the balls size distribution in the mill may have an effect on metallurgical performance. 

Bill Fraser
1 year ago
Bill Fraser 1 year ago

Regarding the MBWT procedure, we have conducted many MBWT trials (more than 20 years ago), this has help to Moly-Cop to develop an improve products.

MBWT is a abrasion resistance trial an is very conclusive due you can evaluate different grinding media quality in the same mill environment, so you can avoid any disturbance in ore, an mill conditions.

MBWT has to be very well designed in number of balls to be added, time on MBWT (to assure that ball must wear 40% of ball at least), balls must be taken from industrial usage (not laboratory as mark mentioned), balls has to be drilled using special machine (like electroerosion) to avoid any disturbance in ball microestructure. After drilling, each ball has to be weighed and tagged with a code, and the sent to operation. The most difficult is to recover balls into the mill (is a very hard work), once you have recovered the balls, you have to perform a statistics analysis in order to suppor results.

Tony Verdeschi
1 year ago
Tony Verdeschi 1 year ago

Excellent input on the MBWT guys. We have performed many such tests in our ball mills and all of the issues raised here are valid. The test is not perfect but if well managed is certainly useful for ball mills. As you know we've also performed some recent MBWT's in SAG mills, and while the test is very fast due to he high wear rates, and limited to only one or two ball types due to multiple drill hole patterns promoting breakage, it is still a useful tool and the high wear rate at least makes for a fast test. Due to the uncertainties around real product quality, survivability and such we've attempted to develop a laboratory wear test procedure with Jeff Gates at UQ that i think has a lot of promise as a preliminary tool. On this scale we use 25mm balls of the same spec as the field grade product (variations in hardness profile with size are less of an issue with ball mill balls up to 3", so we can get away with this compromise) and run them in a lab ball mill with 2-3kg of ore and measure the wear loss. You can mix grades and test multiple balls at once to get a feel for comparative wear rates. Much less disruptive than a full scale MBWT, though wont predict any full scale breakage issues associated with poor quality control in manufacturing, so there is no substitute for working with a proven supplier, having clear quality metrics in the supply agreement and routine quality assurance testing. Ultimately there are many sources of noise in our processes and we cannot afford to spend time trying to perfect things like grinding media specifications. We need to narrow down to something that works, is cost effective and close enough to perfect to allow us to focus on the rest of the things that influence profit. I see the value in this endeavour i believe as clearly as anyone, though i also see how much our site people have to do and how little time and resources they have to do it. FYI Jeff published a paper on his lab test at MEI Comminution last year.

David Kano
1 year ago
David Kano 1 year ago

The MBWT, which I referred to was done a while ago when I was in production roles. I longer work for that company. The data belongs to them. Because I drove the program and involved heavily in the execution of the trial I thought I would mention. Another information useful to know, grinding media charge in the ball mill we used for trials, was replaced (on average) every three months. That means average grinding media residence time in the mill. As long as the MBWT lasts at least for that period (and samples collected at the end, and/or intermediate), then there is a level of confidence in the results. As for the soft center in the grinding media don't most (if not all) grinding media have soft center and they work harden inside mill? Once past the soft center wear rate increases rapidly. I think this is the point we are missing in in the data (difficult to collect). On the other hand once the grinding media becomes smaller and reaches the softer center, then their shape change become potato shape or hexagonal, or another shape, or coin like in the extreme. We have seen all these. This media occupies the grinding charge, draw power but their contribution to grinding efficiency is questionable. We have seen this in a number of mills, where we had to empty the contents of the mill. In one case we weighed about 2-3% of the mill charge occupied with such media. That was a SAG mill operating like a ball mill. The ratio is much larger in ball mills.

I would expect high breakage rates when doing MBWT in SAG mills, while it is not so much an issue for the ball mills. 

Victor Bergman
1 year ago
Victor Bergman 1 year ago

David,

Many grinding balls have a softer center. It was not so long ago that the technology only enables heat treatment to an effective depth of about 22mm. Now this is not so, we have through hard media up to and including 160mm. Shape will hold shape in SAG operations with even wear due to grain structure and heat treatment. Now we are doing more work on the smaller balls down to 80mm using the same methods. Many of the smaller balls are made by "roll forging" which has green bar feed with the same diameter as the finished ball and formed by rotating screws. As the tooling needs to be cooled (water) the top of the bar is partially quenched leaving uneven structure and hardness - the shape is further lost with media that are in the "kidney region" of the mill as this media is not free to roll but is "inched" which forms the cuboid rejects. Our media in the main is hammer forged from bars that have a diameter much less than the finished ball size which enables a very high degree of upsetting and eliminates the bar grain structure. The end result is no soft spots on. or in the finished product. The wear is even until the ball disappears.

MBWT in a SAG - I agree with you totally - the danger here is breakage - we offer <1% breakage in almost all SAG applications. Few, as far as I can remember, have done a MBWT in a SAG.

Grinding media has changed greatly over the past 10 years, however many of the operational guys have not seen the new materials and still believe that grinding media has a fixed performance - slowly the message is getting out.


Bill,

Totally agree with you how the procedure of MBWT should be done - on every point.
However, balls are sent by suppliers to the test site - there is room for "selection".
How many balls are collected to call the number "representative" and was there breakage, again no data here either.
Regardless if the samples are random or not, there is still room companies to supply lab made samples - independent selection from stockpiles you will agree is best?
Any test on any media must be a totally random selection from a large stockpile by an independent - as "random sampling" should always be in the end report.

Millions of dollars are being evaluated - let us both just ensure, as designers that the samples supplied by the makers have that "random" status attached?

Bill Fraser
1 year ago

You are right..MWBT must be random to be representative. what we recommend to customers is sampling balls before testing, and also get samples of drilled balls for comparative analysis..so balls samples from stock and balls added to the MBWT must have the same characteristics...if there is any special ball added you can detect easily.
In other way, for having a reliable test using statistical analysis you can detect is there high variance...an P value, so you can validate the test.
However, I think MBWT is the most reliable test to compare different kind of suppliers (we have tested some times 10 different grinding media) at the same mill operating conditions...and avoiding longer trials.

Maya Rothman
1 year ago
Maya Rothman 1 year ago

Has anyone done work on the influence of ball size (weight) on wear. I was involved in some work a long time ago that seemed to indicate (if memory serves me) a wear rate proportional to radius to the 2.14 power, or indicating that the wear on cast chrome steel balls was mostly a function of surface area, but also influenced by the ball weight.

Bill Fraser
1 year ago

the effect of the ball size has a stronger correlation with wear rate.., but specially with the exposed area (m2/m3), so if you have higher exposed area the wear rate will be higher. Regarding the comparison between HCr and forged balls, usually you will find that HCr balls has lower wear rate, but there is an effect that nobody notes...., HCr balls has lower density (Higher Cr content the lower density) than forged balls, so if you have lighter balls your ball charge will have lower weight and you will draft lower power, so if there is lower Kw draft by the mill you will be affecting the throughput or the product size, this is a point to take care when you are thinking to use HCr balls.

Alan Carter
1 year ago
Alan Carter 1 year ago

No. In general in any ball mill we add mixed balls as recommended by Vendor. Your question " Has any one done work on the influence of ball size(wt) on wear".

As far as my experience goes----I have no idea on such study. But the experience says, that in a ball mill once we start ball mill we don't see again for one year. We just top up with biggest ball having 4'' diameter as per calculations. But all balls gets ground till zero diameter. But when we opened ball mill after one year of operation we found that small balls below 2'' diameter got deformed when we used Cast Hyper steel balls having Mn. Consumption was high.

We could understand the problems. We changed ball composition to High chrome forged balls , we could not believe it performance. All balls were in perfect shape and size, with uniform wear. Wear rate was just half of the steel balls. Continued to use High Chrome balls.

John Koenig
1 year ago
John Koenig 1 year ago

Bill I find your comments very relevant. I've done some steel balls trial and studies too and the parameters indicated on Bill's comment can not be missed during steel balls tests.

Alan Carter
1 year ago
Alan Carter 1 year ago

Theoretically HCr ball wt may be less. But wt% reduced is negligible. Practically do not have much effect as compared to High recirculating load from cyclone under flow. Practically it is proved that HCr gave more throughput and Metallurgy. Improved grinding efficiency. Profit --balance sheet ---proved it is more profitable. Continued to use them for the last 15 years. No one need not have any doubt on use of HCr balls.

Maya Rothman
1 year ago
Maya Rothman 1 year ago

Bill, I am not sure that I understand your comment. We found ball wear was slightly higher than just a surface area phenomenon, indicating that total ball mass had an influence on wear as well as surface area. That is to say if wear was purely a function of the surface than it would be a function of ball radius squared. If it was purely mass it would be a function of ball radius cubed. Our measurements indicated wear was mostly, but not totally related to surface area, but also ball mass had an influencing factor as well. That is a larger ball will tend to wear slightly faster than a smaller ball when looking at total ball mass versus surface area. If I recall correctly it was a function of the ball radius to the 2.14 power. Has anyone else had this type of experience ?

Helena Russell
1 year ago
Helena Russell 1 year ago

This is an interesting discussion, and one I am sure many concentrator managers would be interested in. However, as a colleague pointed out to me, this discussion only concerns itself with one aspect of the economics of grinding media, that is the wear component. There are other aspects to be considered, particularly when dealing with base metal sulphide and gold ores. These are the impact of grinding environment on the surface chemistry of the minerals we are trying to separate, and how this impacts reagent consumption and metallurgical performance.

An excellent example of this is the Ernest Henry Mine conversion from forged steel to high chrome grinding media. At the time the conversion was effected the high chrome alloy recommended reduced the wear rate by nominally 29 percent, which incurred a cost of some $180,000 AUD. However, this cost was offset by a decrease in the operating pH of 1.3 pH units (equivalent to a saving of some $800,000 AUD) and an increase in copper recovery of at least 1.3 percent (worth $11,600,000 AUD). In effect, if the operation had only considered wear this change to the grinding circuit would not have gone ahead. But, as the manager took a more holistic approach he was able to reduce his operating costs by $620,000 AUD, and generate new revenue of at least $11,600,000 AUD, with a pay back period of 0.8 weeks.

We have several case studies which yield similar results from a variety of copper ores, as well as other base metal sulphide operations. We intent to present this work at the IMPC in Santiago next year, where we will compare the laboratory results obtained using the Magotteaux Mill with those obtained during plant trial.

I guess it is recognised that the cost of grinding media is a significant portion of the operating budget for most concentrators, but wear is only one factor that should be considered when selecting the right alloy for your process. By manipulating the pulp chemistry during grinding through the selection of the right alloy it is possible to reduce the consumption of other reagents (i.e. collector, pH regulator, activators like copper sulphate), and have a positive impact of concentrate grades and valuable metal recoveries.

John Koenig
1 year ago
John Koenig 1 year ago

Thanks for an excellent comment Helena. I completely agree. We need to give this a broader look. The example you provided must be imitated. It is good to think about the downstream or secondary effects that a change in operating conditions will cause. I've heard several times that we should "change one parameter at at time", when such thing will not to happen in plant operation. Your example is very good because provides solid actual confirmation of the effect of one change over the rest of the plant operation parameters. 

You have made us aware of more parameters that we need to look at when trying to assess grinding media. I think that someone mentioned the effect over metallurgical performance in this blog and you have just confirmed that not only met performance gets modified but also the other reagents consumption 

Victor Bergman
1 year ago
Victor Bergman 1 year ago

Agree that elena has made very valid points - focus on grinding media really does need detailed assessment for full effect evaluation.

I think that this discussion has been very clear in noting great differences in ball types and qualities. Also it is clear that ball metallurgical composition also can have a major effect in bottom line costs.

It has always been a major battle (Levi may agree here too) to explain to metallurgical and operational staff that there is a huge difference between the different steel media available today. "Balls are balls are balls" - clearly not.

One of the better customer procedures that I have seen is worth comment:

  1. Determine if improvements are possible, this should also be in consultation with media specialists as a balance to operational evaluation - media makers comment always invited.
  2. If confirmation of improvement is probable request the manufacturer what qualities does the proposed media have over existing media. Have random samples supplied for a full metallurgical laboratory analysis including volumetric hardness and micro imaging of cross sectional grain structure.
  3. If the quality of the product is confirmed and the logic of the product improvements are accepted then (a) for SAG a small - say 20 to 60 ton breakage or survival test is initiated. (b) Ball mill - MBWT with the existing supplier may be an option but a performance guarantee with a production trial is still better.
  4. For SAG last stage assessment - run a fully purged production test with comparison with a sister mill or carefully mapped ore variations throughout the trial. Ball mill media assessment covered in 3.(a).

All the best to the forum for this coming year - stay safe in 2016.

John Koenig
1 year ago
John Koenig 1 year ago

I would like to add that most of the time the operators assign little amount of resources to these endeavors. As you all know, it takes time to conclude if X or Y steel ball fits a n specific operation . There are site specif components that made make one type of media to be better than the other one. 

so the recommendation is to assign the proper amount of resources to these evaluations and to use the best practices in the market. I agree with Mark: Use the help of specialists for your evaluation. A plant has so many components that it is not possible to become specialist in every single aspects. While we are knowledgeable we are not specialist! 

Alan Carter
1 year ago
Alan Carter 1 year ago

My experience for the past 24 years in ball mills is that, the blow holes in the balls in CAST Hyper steel ball, contribute to wear and deformation of balls. But this wear rate is less in FORGED BALLS.

Normal friction wear due to rubbing, tumbling, fall of balls is mostly due to hardness of ore ground. Silica ores consume more balls, as compared to others.

To reduce consumption of balls particles as soon as ground to reduce size should come out of ball mill.

Uniform ball distribution (Mixed size balls) is a must during starting of ball mill after liner replacement.

Wt% solids in ball mill also play a major role.

Operating variables its effect on ball wear----

Case study:

In 40 tph ball consumption was high, metallurgy of concentrates was less, difficult to control grade and recovery. We did a technical auditing of the total circuit. To our surprise we found following errors in OPERATING PARAMETERS.

  1. Feed Size: Designed feed size of the ball mill was 12 mm. But to underground crusher problem we received 15mm and above.
  2. Cyclone apex life 15 days, causing high recirculation loads, reducing ball mill throughput.
  3. Cyclone having surging effect.
  4. Pulp density inside ball mill high due to damaged feed pipe of 2'' replaced by 1''.
  5. Cast iron steel balls got deformed in shape as it got ground to size of 2'' to 1''. Plate like balls. Causing inefficient coarse grinding.

ACTION TAKEN TO RECTIFY ERRORS IN OPERATING PARAMETERS.

  1. FEED SIZE--- Underground jaw crusher liners replaced with teethed liners. Secondary and tertiary crushing gaps reduced to get less than 12 mm product.
  2. CYCLONE APEX: Silicon carbide --hardness next to diamond was used--life 12 moths.
  3. SURGING OF CYCLONE: Sump was provided with level controller and VFD motor for all pumps.Both in control loop with DIDC controller.
  4. Water balancing made and pipes were changed, flow meters were erected to know water flow ratio.
  5. AAI Ahamedabad made a study on composition of balls and found best CHROME ratio with forged balls.

PROFIT: 30% improvement in feed capacity, and costing Rs/ton feed-- in all consumption rate was reduced. Power, Grinding media, Liner, etc. Metallurgy improved drastically with good grade , improved recovery, less down time etc.

John Koenig
1 year ago
John Koenig 1 year ago

I think you have made a point here: the grinding circuit needs to be under control before someone tries to assess steel balls consumption. "Under control" means that operators in each shift operate using the same criteria and that the maintenance is somewhat predictable. The input and output f the circuit must be the same. For example, what about if a new direction forces the circuit to produce a coarse grind ? obviously the steel balls test is going to be biased. Corollary: The planing stage must be thoroughly thought and everyone needs to be ready with plans B, C , etc in case the operations philosophy changes. 

All the aspects discussed here should be put on the discussion table and addressed by the testing protocol.

Please join and login to participate and leave a comment.