Laboratory Testing & General Mineral Processing Engineering

Laboratory Testing & General Mineral Processing Engineering

  • To participate in the 911Metallurgist Forums, be sure to JOINLOGIN
  • Use Add New Topic to ask a New Question/Discussion about Mineral Processing or Laboratory Work.
  • 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.

Computing and Mineral Processing (31 replies and 4 comments)

Victor Bergman
8 years ago
Victor Bergman 8 years ago

Do you think mineral processor engineers should have great computer programming skills?  Mineral processors work with huge amounts of data. Often this is done in Excel spreadsheets. Therefore in order to automate much of what they do, I would have thought basic computer programming skills would have been of value.

But are mineral processors expected to have this skill, if not why? If so, do they?

If they could write simple computer programming skills (such as VBA or Matlab), would it be valued? Or would they be considered not focusing on their job. 

What about statistics for metallurgists?

The underlying issue I am trying to come to grips with is what is the clear role of a mineral processing engineer.

If the mineral processing engineer is not responsible for detailed analysis, then who is? I recognise that a mineral processing engineer has different roles, and some of those roles are not analytical. Yet I want to understand the Company 'culture'.

A group of young metallurgical engineers was recently asked what they considered there role. Their Answer was Carry buckets.

Yet when I discussed that their role should be that of analysing the process with a view to improving profit, they all agreed that is what they would like to do.

So the question is at various levels; and people may answer however they see fit.

But an additional comment: I worked at a research Centre, and once made a general comment that I get surprised when a PhD student is able to write a thesis, yet not do any computer programming. This particularly upset one PhD student.

So I am not specifically asking whether ALL mineral processing professionals should know how to computer program but whether there is expectation that within (say any group of 5) at least one would know how to; and does so frequently.

And recall this particular group is 'Innovation' so I am definitely considering the question from an innovation perspective - rather than a maintenance perspective.

I would be interested in your thoughts on this.

Maya Rothman
8 years ago
Maya Rothman 8 years ago

Due to nature of the industry we are,computer programing skills, is of great value. for the better run of our processing plants. we need auto-process control because, always mineral processing plants faces process dynamics which therefore require the use of computer programming particularly mat lab to determine philosophy and tuning type.

Mechanical basics has nothing to do with chemistry of processes, sometimes process dynamism need to be controlled, how can you control it? manually or automatically. manual control is not as efficient as auto- control. for that matter, computer programming skill have to be given to mineral processing Engineers to make their lives easier.

Helena Russell
8 years ago
Helena Russell 8 years ago

Yes they need programming skills. Unfortunately many people self learn these skills, and are never exposed to concepts of documentation or naming conventions for their variables. De-bugging spreadsheets that have evolved with no functional spec and various recorded macros as generations of incumbent mets try to make their job easier is not an easy task.

Teaching the concepts of well written code and documentation is needed. Metallurgists should be more than capable sorting out the details of the coding once they have a written a clear spec.

(unknown)
8 years ago
(unknown) 8 years ago

If you are a process engineer in a plant, I think you must focus on mechanical and Industrial view of your work rather than a computer programming view. But by stepping in higher levels of this job ( e.g. A designer) it will be a valuable skill to have innovative ideas in that stage.
By VBA and matlab maybe you could simulate any Ideas you come up with but it will depend on what benifications it will contains. So The benefits from your programming is the priority of any company or your self-employed company. In conculsion, I think it 's not necessary to be a good matlab or VBA programmer for mineral processing engineers as it 's so important to know mechanical basics for thems as they or they not work in process.

(unknown)
8 years ago
(unknown) 8 years ago

I know the necessity of chemistry knowledge but what would you do when your equipments go wrong in thier working positions? If you are a manager of a working shift of a concentration plant, does the VBA and matlab programming help you to manage the process?
Yes, I agree with you to operate the process by automatic control but if this automatic control helps you to react upon the wrong equipments? you can just control the electrical view of your equipments not surely i 'll say the mechanical points.

JohnnyD
8 years ago
JohnnyD 8 years ago

Great question asking about the key competencies required of mineral processing engineers. I have observed these are often ill-defined & occasionally driven by particular interest as opposed to operational needs.

Some key roles for mineral process engineers include the analysis and understanding of operational performance, problem solving, process stabilization, process optimization. Some understanding of programming would be useful where development of new approaches are required and in the understanding of limitations of existing software. Further understanding of the maths would be great as basis for understanding the limitations of such analysis and modeling / simulation. Few engineers are occupied at the extremes of this work, however, most need to perform some level of computer-based analysis and improving ability here would improve competency and job-performance.

Maya Rothman
8 years ago
Maya Rothman 8 years ago

Hi Tony, 

Mineral Process Engineers sometimes if not always perform metallurgical works in metal production industries. I agree with you that of mechanical knowledge of the equipments however this is mostly for mechanical engineers. consider this practical example, you are a process engineer in a gold processing plant, at the CIL circuit you are more often facing tank leaching tank overflow. is this overflow results from equipment defects or something goes wrong in the process? here the solution is not surely the maintenance of the equipments rather than process analysis of the dynamic nature as a result control strategies of such situation can be suggested and implemented. for mineral processing engineers/metallurgist, optimization of process operating parameters, process stabilization as well as process data analysis are their key roles which i think need computer programming.

(unknown)
8 years ago
(unknown) 8 years ago

A Mineral Processing Engineer in a plant that has a leach tank overflowing will most likely not need to write a computer program to help analyze the data.

Depending on the career path that a particular engineer wants to follow will determine the need for computer programming. Many of todays students do not want to actually work in a plant -- they want to sit in the office and analyze the data -- they should be programmers. Of course, this is only my opinion.

Computer applications of simulation can be used as a complementary tool development that often lead to accurate solutions in less time and with much less resource consumption. These computational tools are not used in order to replace traditional, but have been shown to be a useful adjunct in the technological development and design engineering. The simulation tool can help has solved some of these problems, with low cost, high reliability and usually in less time. Otherwise these tools can aid the process engineer to understand what is happening and what are the problematic points in the process. I can mention some of the simulators:

  • Matlab Simulink
  • Simulator ModSim
  • Solidworks Simulator
  • FlowWorks Simulator. 

Victor seemed to be liking VBA and Matlab when there are a number of software tools that can achieve the same result. Personally, I enjoy analysing data and graphing different variables BUT when trouble shooting you need first to see and inspect what is happening on the "Ground". So many managers come up with findings from their desks, with fancy analysis, without seeing what is happening at the plant AND as a result are way off the mark and waste everyone's time.

Helena Russell
8 years ago
Helena Russell 8 years ago

On the topic of maintenance, a good process engineer (equipped with a data historian and some VBA skills, to use in excel) can generate reports that greatly assist in keeping an eye on (say) flotation agitator current draw and make sure that the wear parts are being changed. This is just one example of condition monitoring using standard instrumentation (most motors report their current to some DCS/PLC).

I agree that these maintenance issues are often large value drivers. I think you've got more chance of teaching the mets to do the analysis than the reliability engineers (although Mr Weibull did a fair job 😉 ).

(unknown)
8 years ago
(unknown) 8 years ago

It depends on your role as a mineral processor. If you are in production, then you are flat out making sure 'production targets' are achieved, reviewing production data and preparing reports. No real need for computing skills has pointed out. If you are in research, then perhaps, depending what area you are working in. The more information collected, then the more likely it is that you will use some form of computational analysis. While the data may be collated in Excel, the analysis may not occur in Excel. If you are design engineer or in consulting then it is probably unlikely.

So where does this idea come from that mineral processors handle 'huge amounts' of data?

The guys that handle mega amounts of data are the geologists.

Victor Bergman
8 years ago
Victor Bergman 8 years ago

The issue of the amount of data really depends on each plant's strategy. If data-handling techniques were available then obviously the amount of data could increase. Obviously there isn't much point in obtaining data if one does know what to do with it.

I guess the comment mineral processors don't have much data but geologists do indicates that information from the geologists is not utilized by the mineral processors, hence the 'geometallurgical' approach to mineral processing is not valued.

Yet even if one uses mineralogical data, one could say use the summary average mineral compositions, or go deeper and actually use the detailed mineralogical data. At least for the MLA this was stored in a database.

So it would therefore appear that a large proportion of metallurgists are focused on trouble-shooting rather than incremental improvement; and therefore don't have data-handling issues.

I don't think it is fair comment to say I am promoting VBA and Matlab. I could have listed a whole set of languages such as:

  • C++
  • C
  • Python
  • R,
  • Java

but I wasn't trying to focus on the details of the best language. If you consider alternative languages are preferred by all means say so,

However the context of my opening discussion was that many Mineral Processors use Excel for data-handling.

If this incorrect, I am more than happy to be enlightened.

The question was largely what it was. One thing I certainly wanted to know was whether the average mineral processor (forgive the stereotype) would even know what VBA was. I have recently received feedback that many mineral processors do not know what VBA is, so I wondered whether they ( a large majority) actually knew how to write any computer program at all. Hence my question. I am leaning to the answer 'no'. However there exists a few computing experts who (like me) have been blissfully unaware how little computer programming skills are valued.

Some have discussed Matlab; with natural lead to Simulink. Personally I never got anywhere with Simulink, and ended up going down the path of using Visio for flow-sheeting with integration being dealt with via VBA.

I have heard that one could achieve the same objective (for free ) using Python. I don't know.

The comment that a Manager wastes his time (and everyone else's) by doing fancy analysis is concerning. Surely if a Manager is incompetent if it is the fault of the recruitment process; not the software environment.

So what is concerning, and yet not surprising, is that anyone trying to develop method of analysis to improve profit is deemed out of scope by a large number of his colleagues.

This is very sad. Yet I thought that might be the case which is partly why I opened up this discussion. I am definitely trying to identify those Companies with a successful progressive culture.

(unknown)
8 years ago
(unknown) 8 years ago

The geological data is drill hole data, assays and the like and used for geological resource modeling. Not so relevant to metallurgists unless ore types have been identified and some metallurgical testwork has been done. Metallurgists generate most of the geometallurgical data however it is not 'huge' and does not require computational methods to handle it.

With regard to your other comments, I would suggest that you spend some time on site and see what is involved in managing a plant. Data is all and well and a wonderful space to hide but when the ball mill has tripped and the lime ring main has blocked and pyrite is going everywhere.

Collecting and analyzing data is an onerous task, requiring equipment, techniques, people, resources....not forgetting the effort and resources to check, calibrate and maintain any data measuring equipment unless the analysis of the data allows the flotation recoveries to suddenly improve by say at least 3 or more percent on a statistically demonstrable basis, then it is simply not going to happen.

Make the case for the improvement and management may listen, but by and large, it is the bare basics.

Exceptionally few sites have MLA facilities on site, so that is a research project conducted off-site, and not in the volume or depth that you are used to from your JK research days.

Victor Bergman
8 years ago

I agree with many of your points. The JK scenario was awful. One works in an office and every half year or so, I get to give a 15 min. presentation; and then the industry guys go in a private room and report back their feedback - in no means a constructive manner. I far prefer my current scenario - although I agree I need to get out more.

JohnnyD
8 years ago
JohnnyD 8 years ago

It's interesting to see much of latter discussion here being mineral processing engineers solely as firefighters. No wonder many leave the industry! In fact, it's particularly interesting seeing where these engineers reside at many concentrators and discussing the nature of their roles.

The present discussion of 'big data' and the formless nature of that team may have lead people to think only large (petabyte data sets are being discussed and of potential value).
I've seen there's definitely space for more than that base level of firefighting and much less than hugely demanding analytical campaigns 'mining' the data set. The programming demand doesn't need to go to absurd length, most of the analysis can be done using simple, easily available tools; and there's value awaiting attention.

Victor Bergman
8 years ago

Absolutely, and it is a discussion that has to occur.

Victor Bergman
8 years ago

This is a critical issue. Some young mineral processing engineers see the blatant inefficiencies and presume it is a deliberate Corporate strategy; so they don't even bother recommending improvements.

Victor Bergman
8 years ago
Victor Bergman 8 years ago

Yes, the mining industry is unattractive. I advised many competent PhD students and in the end thought it was pointless trying to bring these people into the mining industry when they could do far better elsewhere (particularly when there IP is taken away from them). Many of them left anyway. 

I gave a course on Simulation in South Africa last year; and one of the bright young Metallurgists decided to go work somewhere else; as she realized what she wanted to achieve for Company (financial optimization) was only going to happen (if at all) after about 20 years.

I would suggest even a 20 year target was unrealistic.

Point about 3% improvement. You basically turned the issue onto me and indicated I must indicate how to improve efficiency by 3%. Yes, 3% is a realistic target, and no I don't yet have case studies to prove my case. (early start up looking for case studies). However I was rather hoping the response was going to be, some Companies believe 3% targets are realistic and therefore have an analysis group dedicated to incremental improvement.

Obviously it is a better strategy for me to try and identify Companies where a 3% improvement is considered realistic and just want to discuss options on how this can be achieved.

A comment suggest that he may be an exception (and there may well be a number of others), but it would almost seem a conclusion that for most Companies a 3% improvement is not considered realistic.

Therefore the hyperbolic extension is:

  1. Companies don't think 3% improvements can be made.
  2. Therefore don't both collecting much data (just enough for firefighting)
  3. This means we don't have to worry about data collation or analysis.
  4. Computing programming skills are not required.

All is not lost however, as once positive case studies emerge one would have reasonable hope to change the culture. It is just that this is not where the Companies are at the moment.

Helena Russell
8 years ago

Yes, and the last two of these can be done much more quickly and effectively with some programming skills.

Victor Bergman
8 years ago
Victor Bergman 8 years ago

As you point out reports can be automated. 

There was a case where a metallurgist claimed he spent most of his time preparing reports; and I mentioned that it could be automated for which there were two benefits:

  • Save the time of the metallurgist
  • Produce the reports fast so that profit opportunities could be realised.

This doesn't mean that the metallurgist/mineral processing engineer is taken away from the data interpretation; rather most of the routine data handling can be done automatically

Yes, and this can be done with Matlab and VBA; and no I am not specifically advocating one language over another.

The question here is whether metallurgists/mineral processing engineers actually know that most of the datahandling can be dealt with automatically.

Rather like the parable of the axeman.

'Why don't you sharpen your axe?"

The axeman responds: "Can't you see how many trees I have to cut down. I don't have time to stop and sharpen an axe."

Now the point here is that yes, it takes a long time to learn programming so it may turn out that the time involved (for a mineral processing engineer) to learn may not be practical. But they should at least be aware that it can be done; and consider outsourcing the programming to a contractor. A straightforward cost/benefit analysis can identify whether it is viable.

Oh yes, what was the response of the metallurgist? He agreed that efficient datahandling and reporting could save the Company millions of dollars, but did not think the Company would be interested.

Bob Mathias
8 years ago
Bob Mathias 8 years ago

Seems to me that the trend continues to have one person do everything. A metallurgist or engineer is hired for that purpose, not to develop programs for data handling, that's what IT specialists are for (and faster at it). Get an axe sharpening specialist to sharpen your axes while you are busy chopping down trees at the same time. Of course programming knowledge is a benefit, but it can't be required. there would also be a real danger of one metallurgist developing his or her own software to deal with data and then when they leave; the data and program is lost. There needs to be continuity and one information management system that can be mined by many not altered by many.

I completely agree, Mineral Engineers with a computer knowledge back ground in VBA would be great to develop customized spread sheets for their own requirement. My self is learning a lot on Microsoft Excel every day in my working profile. This may not be THE case with people involved in plant operation as they have limited time to spare the time with computers.

(unknown)
8 years ago
(unknown) 8 years ago

My career goes back to when programmable calculators were the state of the art and an Apple II was a super hot machine. In those days we had to program in Basic or Fortran to generate the simplest reports. Programming largely ended for this type of thing when Visi Calc (remember that one?), the first commercially viable spreadsheet, first came out. Lotus followed with functions like standard deviation, linear regression, and graphing capabilities, and then Excel took over.

I guess using Excel is programming in a way, but with a much higher level language.

I think computer programming skills are a little bit of a stretch for a metallurgist in an operating plant today. Some metallurgists are tasked with doing DCS or PLC programming changes, but that is something for an individual company to decide. Those systems are so customized and constantly changing that teaching them in school would be futile. Most engineers get some rudimentary programming in college.

That being said, being able to fully and competently use data crunching programs like Excel to point out correlations and trends, build cost models to evaluate projects, and generate reports is vital. This goes hand in hand with the plant metallurgist's primary task of being the process expert when it comes to plant operations and relating the plant to the mine and the downstream customer's needs.

Actual programming skills are more appropriate for academics and research type people, maybe supplier people, along with dedicated on site programmers. They have the time to learn their particular system.

Maya Rothman
8 years ago
Maya Rothman 8 years ago

I don't think its quite true that metallurgists see the inefficiencies and assume that it's corporate Policy, but rather that unlike mining engineering graduates who are in people management from day one the metallurgist becomes a problem resolution expert. Thus the better their skill as solving problems the more they are tasked with that job, and the more they are sidetracked from becoming superintendents and managers.

I graduated learning on a slide rule and developed spreadsheets in APL, DOS and told the Programming guys what the PLC had to do in terms of reaction times and resets, so when Lotus and Excel came along life got easier, not so much for spreadsheets but for those complex regression analyses to prove if it was residence time or grind that was affecting recovery. As to the comment about using an axe sharpener now that really is a recent innovation. Most IT staff I've met know less about process, than process staff know about IT!

But to admit too much knowledge of any topic, is a surefire way to get landed with that job for life. With experience I may not be as skilled as the authors of Metsym or others on writing a mass balance software, but I can use the data to analyse bottlenecks and even look at by-product markets

Victor Bergman
8 years ago
Victor Bergman 8 years ago

We are actually in agreement but you have expressed it differently to me. The issue of learning computer programming is a 'death sentence' in many professions (not just mineral processing). There was a particularly relevant discussion by a financial analyst who basically concluded that if he learnt computer programming his salary would effectively be halved (in the long term).

This is why this discussion is important.

I have tried to avoid explicitly advocating a position (as I opened the question with sincere questions). I think it would be a poor conclusion that a mineral processing engineer decided not to learn computer programming over fear his career would be undermined. Therefore I think the middle path is they indeed gain exposure to computer programming with a view to understanding what is achievable.

Then, they should be able to work with software developers with a view to improving processes.

However I would like to convey something very important. In theory mineral processors are 'processors'. Not only should they understand the way minerals are processed, but how data is processed.

Therefore I think that it does make sense that mineral processors are aware of how data is 'processed' within the organisation with a view to making decisions. I did construct a course on VBA for Process Engineers:

http://processhub.com.au/

Yet I state in the course that the purpose is not to turn mineral processing engineers into software developers. The purpose is to expose mineral processors to some of the easy-to-access capabilities.

The decision as to whether they should write the code, or someone else should, is independent.

Bear in mind that my opening remark was: should mineral processors be able to write a computer program? Now there appears to be a large number of mineral processing engineers who aren't involved in any analysis at all. So let us confine the focus to those who are involved in analysis. I would indeed suggest they should indeed be able to write a computer program.

However I don't think they should ALL be 'software developers'; yet there are exceptions(such as Tom) who clearly contribute to their Company by being able to provide advanced analytical skills.

Yet the competent mineral processing engineer who can also write high level computer programs is an exception, and not mainstream.

I would tend to think if a graduate completed an engineering degree, and had never been taught computer programming, that this is wrong; and this is partly why I tried to get the course going as I think there was a hole that needed to be filled.

I think there is a serious issue with the profession when engineers appears to take on only a maintenance role. This means that the actual issue of overall improvement of an operation must be taken on by another profession. Yet let us hope it isn't the accountants. Or is it already too late?

Helena Russell
8 years ago
Helena Russell 8 years ago

If a site wants some improvements they can contact Outotec, their equipment OEM or some local consultancy. Our process engineers have programming skills and platforms to apply them in (eg HSC-Sim, some custom CCD simulations).

The mining companies concentrate on finding and extracting ore, they should not be expected to have the same level of processing expertise as the OEMs. Nevertheless, site personnel should use programming in addition to Excel for ad hoc analysis, routine reporting, and production management.

Maya Rothman
8 years ago
Maya Rothman 8 years ago

The one question I have is mining in all its forms (including mineral processing) more complex than it was 20, 50 or 100 years ago? Yes, by assigning axe sharpeners with the sharpining the axe user can chop more trees, but only large operations can afford to hire task specific employees such as axe sharpeners! Within the time line I've laid out above the Mine "Engineer" designed the mine layout, ventilation, company housing and surface roads. One of the reasons why mining is getting more expensive, other than lower grades & a lower value of money, is that more people are needed to do the same amount of work, this despite faster drills, higher blast productivity and better instrument control.

For a process engineer to be able to understand the modern Tools, which incorporate computers, is mandatory, and they need to know which have value. GIGO very certainly applies in IT, but buying software is not unlike buying reagents. Abdicating responsibility of any metallurgical task is suicide, just as deadly as losing focus and concentrating on the Tools rather than the objective.

But it is amazing just much control purchasing agents have, those who know the cost of everything and value of nothing. If engineers were permitted to deal directly with engineers such as Outotec then costs would be reduced rather than increased.

(unknown)
8 years ago
(unknown) 8 years ago

I have to echo the comment that buying "packages" for something like metallurgical accounting without validating them thoroughly is very poor practice and likely to get a metallurgist in trouble. By the same token, developing your own Excel worksheet without validating it is just as risky.

Automated reports are the epitome of garbage in = garbage out. One bad assay result from night shift (strange things happen on night shift!) or one of Thomas's magic online analyzers and the report is just that, garbage. Anything you put out has to be checked for reasonableness and accuracy, even if the machine generated it automatically the same way every time.

For this reason an autmated report may not save as much time as you think.

Just run some known audit samples through your lab sometime and see what you get back.

By the way, if the chain saw gets dull when I'm cutting down trees for firewood I sit down and take the time to file it. So does every logger I know, though some may just put on a spare chain.

Victor Bergman
8 years ago
Victor Bergman 8 years ago

Every decision should be based on a cost/benefit analysis (whether well-developed or back of an envelope). Over and over again, I hear reasons for inefficiency more to do with cost-cutting than being profit driven. I guess the point I (and others) are making is that knowing the basics of computer programming (particularly what you can do with automation) is simply adding to the options on ways to increase efficiency (and therefore profit).

I mentioned 'other profession' when a Company does not encourage its Mineral Processing Engineers to focus on the overall operation (i.e. analysis methods).

One can tangent off this to indicate the advantages of outsourcing to Service Groups (specifically OutoTech).

My comment was hyperbole meant to imply that if mineral processing engineers aren't focusing on the analysis of the operation, then it would be left to the Accountants. Why I specifically made this point is because that is exactly what happens (over and over again).

You may be aware I do contracting. And yes I talk to very large Companies. It is generally very easy to convince an engineer that a 3% profit increase is realistic. Why? Because they already know that. So I simply use the 'pitch' that I can provide a quantitative demonstration that a 3% increase is realistic.

The problem is not the engineers. The problem is:

  • They don't actually collect the relevant data (i.e. regular audits)
  • They don't have the time to get involved.
  • They need to convince the financial controller. i.e. the Accountant.

I am aiming to try and resolve problem no. 3.

That situation No.3 exists is almost incomprehensible. However when you realize that the mineral processing engineer is prepared to 'surrender' their position to the accountants it all makes sense.

Therefore I see the issue as either:

  • A continuous conflict
  • The mineral processing engineers have already surrendered their roles.

Now I would suggest (and more than happy to be shot down in flames) that the more a processing engineer can talk the language of the accountant the more chance they have of winning their argument.

In particular I was first shocked when I joined JKMRC and saw the simulation software, which was supposedly was being used for plant optimization. And then I would ask (rather naively) why doesn't the software show profit.

The answers might make sense to others, but they made no sense to me, and still don't. Hence I quickly formed the opinion that mineral processing engineers are technical rather than financial. This view has been enforced when all too often I see mineral processing engineers who are actually interested in profit improvement told that is not part of their job.

So part of the motivation for both discussion and my own world view is that we need to both change and empower the mineral processing engineer (whether directly trained as a mineral processing engineer or alternative background) to move into true financial optimization, and to develop this expertise to such an extent that they can convince the accountants of the required improvements.

I definitely see understanding the basics of computer programming as part of the skill-set.

Maya Rothman
8 years ago
Maya Rothman 8 years ago

Of the Chief Executive Officer, Chief Accountant and Chief Metallurgist only one is not surprised when the company fails! In some companies secrecy exists not to protect confidential information but because no one knows how to communicate.

(unknown)
8 years ago
(unknown) 8 years ago

Very interesting discussion and many valid comments were made.

I would like to jump-in about what has been implied - that mineral processing engineers are not encouraged/welcome in the financial analysis of the production data for uncovering opportunities.

Here the Chief Metallurgist has to content also with the Chief Geologist (the "discoverer" of resources) and the Chief Mine Engineer (the "enabler" for extracting the value).

However, traditionally, the Chief Metallurgist reports to the Mill Manager/Superintendent while the Chief Geologist, Chief Mine Engineer, and Chief Account report directly to the General Manager. It would be interesting to go back in time and identify if this has been so from the beginning, or if this has developed over time in an effort to better manage mine-mill sites.

Nevertheless, and going back to the original questions, some advanced computer skills - whether macro-programming in Excel, process simulation software, and/or statistical analysis software are always beneficial to have - whether for the current position or in preparation for the next position.

Victor Bergman
8 years ago
Victor Bergman 8 years ago

What a number of people have pointed out is that the ability to understand computer programming opens up opportunity to do analysis. And if one can do analysis then one can both communicate (analytically) with others.

The contrary view is that there is no point doing analysis at all. I possibly should have made this the discussion :should mineral processors be analytical? - but I think all mineral processors would argue that they do do some analysis (otherwise what else do they do?). Running around a plant making sure equipment works does not require a 4 year course in mineral processing.

However I would conclude from this discussion that most mineral processors should indeed have some level of understanding of computer programming. The fact that many do not is indicative that they are disempowered from carrying out their job function.

This then raises the question of why they are disempowered; and unfortunately the only conclusion is that Companies have inbuilt failure mechanisms. The most common failure mechanism is a deliberate lack of effective communication (don't talk to me about improving plant performance I am too busy). This is largely what is being said. If some comments were isolated they could be ranked with crackpots. Unfortunately I am hearing the same theme over and over again - which therefore leads to the inescapable conclusion that the viewpoint is valid. The people who make these comments are often highly-experienced mineral processing plant managers who found that technical innovation is highly frowned upon. I hear war-story after war-story of yuppie managers getting into lead positions and failure becoming status quo.

One of the most interesting discussions I had was an ex-plant manager who was referring to his boss, and the boss explained how he learnt to rise up the organisation by deliberate failure!

To take Peter's comments further, if a plant does fail, what do Companies do to identify the reasons for failure? What I hear is 'round up the normal victims (those who are innovative) and shoot them' - not the actual culprits.

So from my viewpoint if a Company wants to improve profit, the professional staff need analytical skills - and fundamental computer programming skills are fundamental to this.

Bob Mathias
8 years ago
Bob Mathias 8 years ago

Programming is the act of instructing a machine to perform a repetitive sequence of tasks. Quite useful if you would otherwise be doing the sequence 10,000 times yourself, but this has nothing to do with analysis.

Analysis is the process of deducing cause from effect - seeing a phenomenon and being able to unravel root causes (or to deduce the missing information that you will need to unravel said root causes), which in turn empowers you to achieve a more desirable effect.

You contend that you need programming in order to have analytical skills, but the only value that the programmer adds is that which he does to take the burden of mundane number crunching off the person who actually knows how to think and generate value. In that sense the programmer needs his four year degree much less than the person who walks the plant.

Victor Bergman
8 years ago
Victor Bergman 8 years ago

Thanks for the comments. There are many tasks that do not require sever number crunching for which a computer program is useful. When I give my courses in simulation, I generally ask for a show of hands of any one who uses Excel. Generally all hands go up. I then ask how much time do you spend using Excel with variations of between 30-100%.

I then ask how much of the time is spent doing copy and paste. the answer varies - but still some spend most of their time cutting pasting and reformatting data and writing routine Word reports.

Most of this can be automated. This is an example where if one knew what could be done, they could at least delegate the role to a computer programmer and improve efficiency. On the analytical side, in my course I give a simple problem to mineral processors.

Suppose (for a particular size-class) you have the feed density distribution, and the average densities of the heavies and sinks, estimate the Ep (efficiency) and Rho50 (cutpoint) of the cyclone.

I have not yet had one participant say they have been taught (in an Engineering course) how to solve this problem - most would have been taught they need to measure the full density distributions of either the heavies or lights.

After explaining the problem most mineral processors can solve the problem (with some guidance) using Excel Solver. They cannot write (nor what they be expected to) a numerical algorithm to solve the problem.

Problem 2) Similarly in my course I then extend the problem to if you have two known feeds going through the same DM cyclone (and presume that the cyclone processes the particles in the same manner) then from the flow rates only, estimate the Ep and Rho50.

One can find papers that indicate Problem 1 is known by one or two researchers, Problem 2 and its solution is new (and is part of a much larger patent I filed a month ago).

Similarly their are many ways we can increase our analytical understanding of plants by having a core set of skills beyond what was taught (or should have been taught ) at University. I can many provide other numerous examples.

I think the quote of me that I am saying that 'You contend that you need programming in order to have analytical skills' is not a true reflection of my viewpoint.

For example, a chess-master is highly analytical, and does not necessarily require a computer program. I think the issue is the level of depth required - in particular the level of complexity of the problem. My viewpoint here is that in many instances a mineral processing plant is 'complex'. This doesn't mean that basic quick fix analysis is not of value - because it is of great value.

But some plants may have over 100 streams. Therefore the problem of identifying which units are under-performing and best strategies for plant improvement becomes 'quantitatively analytical'. That is we have to look at the data and analyse it.

Once again, the issue is that many mineral processors do not see their role as requiring complex data analysis (i.e. do not even know how to estimate the efficiency of a dense medium cyclone as given in Problem 1). I would have this was a core skill; yet I don't want to get into a 'yes-no' argument.

I recognize than many mineral processing engineers do not see such analysis (identifying the efficiency of units )as useful - yet there are also many other would who think this is fundamental.

Bob Mathias
8 years ago
Bob Mathias 8 years ago

I had the job many years ago of trying to teach stats to engineers, and quickly came to the realization that the biggest problem that people had was framing the problem. If you set it all up for them in such a way that it was a mechanical operation, most people got it. If you took away the context and just gave them a bunch of data (including some completely redundant data) then everyone just froze, and maybe two people out of 30 had the intuition to frame the problem and select the useful data. Therein lies the problem - to solve any real world problem you need to know which data to use, in a universe of noise and dead ends.

Now here's the thing I find amusing - as you say, some plants may have over 100 streams, but the first thing I would say would be, "Ah, I see your problem! Too many streams!" We encountered a couple of examples of that in the last year as part of our plant support business, and simplification was a magic bullet in both cases.

Victor Bergman
8 years ago
Victor Bergman 8 years ago

And I am not specifically advocating complex circuits - just the recognition that mineral processing is complex. I think you and I agree that if you do have over 100 streams you need some smart analytical tools to deal with the analysis.

Solver is dreadful - although has the illusion of being great. in the course I also give simple problems that Solver collapses on, just so that participants don't rely on Solver.

The issue of maths engineering is well-known; and I don't think there is a solution. Albeit recognition of the problem is a start. Engineers (compared to mathematicians) will take maths and need a practical problem to understand it; given the practical problem they can use the maths well. However if they are given another problem which requires the same maths they have difficulty contextualising the maths to the new problem.

With respect to stats I start my course by saying" isn't stats boring". Everyone agrees. I then teach probability theory (applied to mineral processing) and explain how it is absolutely fascinating. Eventually I reintroduce stats, and they are far more interested. Unfortunately stats has a bad image because it is largely taught form the perspective of data quality (confidence intervals) rather than data analysis.

Maya Rothman
8 years ago
Maya Rothman 8 years ago

Personally I love Solver when it comes to performing tasks which require a lot of iteration, but I agree it does have its limitations (particularly with non-linear problems). This is where I find graphing the problem via Excel/Matlab useful.

But at the end of the day, all that the simulations and computers do is provide tools that makes the data easier to understand. It's the people that interpret the results and make the recommendations.

However, I do agree that you have to understand the limits of the problem and set up your boundaries logically. Yes Kate, I agree, graphing the outputs makes sense since you can see if the results are logical or trending as expected.

Statistics are also invaluable when solving problems. Programming has its advantages for repetitive tasks.

I would say all these skills have value and one skill is not necessary better than the other. As long as you can do the job at hand - as the saying goes, "there is more than one way to skin a cat".

JohnnyD
8 years ago
JohnnyD 8 years ago

In my experience, programming skills are becoming more and more important to mineral processing and although not all processors need to have these skills, some should. Although plants vary in the quality and volume of data available, the functionality and data quality that can made available from the advanced historians such as PI can greatly enhance the capability of the mineral processor to make meaningful analyses of plant performance, and is becoming more and more routine.

I've also found that the functionality available in excel has come forward in leaps and bounds in recent years. I recently wrote a metallurgical accounting spreadsheet that imports data, cleans data, dynamically sorts data, aggregates the data from various sources, then performs auto-checking and produces preliminary reports for final checking by the metallurgist - functions that were unheard of only a few years ago. Unfortunately the complexity starts to become excessive and a more specialized program is probably required.

I lament the lack of a suitably powerful, diverse and flexible metallurgical accounting tool that can also be understood and programmed by mineral processing engineers

In addition, as the data available becomes more abundant, so too do the demands on the reporting metallurgist. Some of the more advanced functions such as metal balancing by error minimization using statistical models and 'reconciliation engines' really start to go beyond the realistic functionality of spreadsheets and the programming capabilities of the typical processing engineer yet they are becoming more and more common.

I would also differentiate two functions - the metal accounting function, and the plant performance analysis &/or testwork analysis function. For the former, advanced excel skills inclusive of VBA would be the minimum requirement and specialized programming skills optional in my view. This function need not be done by a plant metallurgist but could probably be done by an IT specialist with some mineral processing and mass balancing understanding (unfortunately very rare indeed). For the latter, spreadsheets can be used for most of the requirements in conjunctions with some specialist statistical and processing software, but this also needs to be coupled with a comprehensive understanding of statistics as applied to mineral processing in addition to the normal fundamental understanding of the process under study.

Short answer: yes, programming skills are required for at least some processing engineers, they will be in increasing demand in future and they need to be coupled with skills in statistics in addition to the regular comprehensive understanding of the process itself.

Please join and login to participate and leave a comment.