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.

Pilot Plants for Process Engineering (20 replies and 2 comments)

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

Often we hear and see discussions around pilot plants for process engineering around the mining industry and there are many questions / issues left unaddressed, e.g.,

  • Why - Most critically why would anyone run a pilot plant in 2015 given what's known about the processes we are using & the capabilities available to model and simulate based on small scale / lab test work? This is potentially a major investment. How does this relate with risk management / project financial analysis?

Is the story different for:
-mineral processing (comminution, flotation, etc.),
-hydrometallurgy,
-pyrometallurgy,
-etc.?

  • What comprises a meaningful pilot plant? Is it the entire circuit or is it acceptable to pilot only critical parts of the entire circuit where uncertainty / risk is greater?
  • When in the project lifecycle should pilot work begin / end?
  • Where is the best location to pilot - on-site or at a consulting house? What are concerns to consider around this?
  • Who should be running the pilot? What is best staffing plan? What are major considerations?

What are your thoughts?

Paul Morrow
8 years ago
Paul Morrow 8 years ago

As recently as two years ago I saw an operation that did use a pilot plant and found some operating considerations that changed their whole flowsheet. My take would be:

  • Why - Even if we do know how the equipment will perform on lab scale, this does not mean we know how it will operate under continuous conditions. Also lab tests are as close to ideal as you could ask for, and they use small size samples. Many an operation has found out unexpected things once full scale operations are commenced. The chemical based portions are the most likely to behave differently; flotation and hydrometallurgy in particular.
  • What - While it would be nice to pilot the full circuit, at least those that are most likely to cause problems should be (again flotation and hydrometallurgy).
  • When - After preliminary engineering, but before detail desing, in parrallel with basic engineering.
  • Where - As close to the site as possible, at the site, using site water and hopefully bulk samples. This is where a modular plant would come in handy.
  • Who - Probably start with a contract/consulting crew and tranistion to supervisors/senior operators as a training tool.

Maya Rothman
8 years ago
Maya Rothman 8 years ago

I believe that many vendors have in depth data from their customers these days - this data covers a very wide range of ore types and equipment in operation. Usually they can run modeling programs as a way for a general setting and then connect to like mine site circuit data for another indication as well. Unless it is a totally new concept of processing that involves several untried/proven steps in the process - then the modeling program needs to be built again. 

Remember, that we are only looking at reagent selection and known equipment. 
Operation people skills and location is also out of our scope to comment. 

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

There are certainly some applications where it's easier to see need for pilot plants:

  • new process, new equipment
  • need samples for customers or feed for evaluation of next process in line
  • difficult to process feeds
  • where lab testing suggests problems in achieving desired product quality

An issue with piloting is addressing variability related issue. While we may obtain a large sample and pilot around a feed, a multiplicity of ore types with potentially different performance can increase difficulty of executing meaningful pilot plants.

JohnnyD
8 years ago
JohnnyD 8 years ago

I have many fond memories of utilizing pilot plant studies to increase my mineral processing knowledge, easily observing the effects of changing variables and various feed grades especially in the flotation process, and transferring these observations to others back in the real operating facility. Thanks for bringing up the topic; I'm looking forward to reading the responses.

(unknown)
8 years ago

I remember operating those extended pilot plants with you in the distant past, requiring many truckloads of feed to be brought in.

Paul Morrow
8 years ago
Paul Morrow 8 years ago

Tony, 

An issue with piloting is addressing variability related issue. While we may obtain a large sample and pilot around a feed, a multiplicity of ore types with potentially different performance can increase difficulty of executing meaningful pilot plants.

This is when lab testing as the only basis can lead to big problems achieving desired plant performance, and when a pilot plant can save a lot of issues during start-up and commissioning.

(unknown)
8 years ago

That's the big question. I agree on all the knowledge and benefits can provided. However, when does risk management (mitigation) force us towards operating a pilot plant; and also provide a basis for the strong justification necessary?

Maya Rothman
8 years ago
Maya Rothman 8 years ago

A great topic - these are the times when I do get to learn new things. As with a lot of us, we tend to have tunnel vision based on our specialties which usually misses the bigger picture. These days that one big difference of SAG Vs. HPGR is a critical decision - I still end up with SAG (not only because we are ball designers/suppliers) when the ore body has a wide variation in characteristics within the target deposit. The next question can be screens or cyclones, with CIP, CIL or even leaching... 

These days I have great faith in you people to pick the plant from the Jorg drill results but if the ore body is so complex the answer may be two separate circuits to enable changes as the project goes forward? An each way gamble?

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

In addition to the comment from the experts above, who should start up, commissioning, ramp up to production target is a group of competent operator, usually retiree with short term contract plus several mechanic and electric/instrument modification squad. Some of the operator need to have sampling competency under the team leader direction. The team leader ideally someone who has a metallurgical background of the specific deposit. Some non skill helpers will be great to support sample handling, material movement, and general tasks. Should an assay and metallurgical lab not available, sample handling and labeling become critical.

Maya Rothman
8 years ago
Maya Rothman 8 years ago

About three years ago a customer discussed with me the availability of a small, transportable ore processing circuit. His intention was to move from small anomaly to small anomaly, treating deposits that were overlooked because of their smaller size (say < 1m ton of medium to high grade ore - Au). 

The discussion and subsequent research lead us to a reasonable set of economies - probably weighted towards good viability subject to the head grade. This was a basic crusher, small SAG and a ball mill in front of a small CIP. Ideas these days would now suggest a larger crushing circuit with only twin ball mills? Perhaps such a plan may also work as a pilot plant and then sold on for that purpose - at a cost offset? 

My comment earlier was a reaction of high cost pilot plant that would have a one hit use only.

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

Staples et al., Are SAG mills losing market confidence, SAG 2015, Vancouver, 2015, raises some interesting points for consideration:

  • interpretation of results from pilot plants – true for lab work as well
  • of samples used in pilot plants – true for lab work as well
  • of benchmarking as part of interpretation of results – broadly true whether or not piloting
  • analysis of results

This particularly as many of the problematic studies they cite included pilot plants. Some of their comments include:

“Autogenous (AG) and semi-autogenous (SAG) grinding mills are considered a mature technology, so why are we seeing a number of recent projects not achieving nameplate capacity and is SAG technology losing market confidence, particularly in financial circles? This paper presents a critical review of underperforming AG/SAG mills based on public domain information comparing design expectations with actual operational performance. Actions taken in operations to mitigate the issues and optimise mill performance are listed, including a high level discussion on capital and operating costs.

The current scepticism within financial circles of the ability of consulting and engineering firms to predict SAG mill throughput, as evinced by the Scotia Bank articles referred to above, should not be considered as a lack of sophistication of the available testing methods or the maturity of SAG milling technology. Rather it reflects issues in properly apply the available techniques for prudent SAG mill design when processing competent ores, including benchmarking of performance.

As stated in Lane (2010), in some cases poorly compiled JKSimMet models have been used to support the data provided by other sources. The issues relating to JKSimMet modelling of competent ores have been previously discussed by Bailey et al. (2009). There are several SAG mill specific design tests available to determine the amenability of an ore to SAG milling and the resultant SAG mill specific energy. Recent projects that have failed to achieve design throughput have typically failed to take sufficient heed of the indications of ore competency in the test work data or analysis of the data.

If A x b values are below, say, 40 or DWI values are above 6.5, due consideration should be given to the inherent competency of the rock and benchmark data relating ore competency to SAG mill specific throughput. The scale-up of pilot plant mill specific energy to large scale operation requires careful consideration of the pilot mill discharge function and associated breakage rates external to the mill in SABC mode. As a minimum, careful benchmarking of design outcomes against available data in the public domain is strongly recommended.”

Victor Bergman
8 years ago
Victor Bergman 8 years ago

Running a pilot plant is an expensive exercise; however, even conventional hydro-metallurgical processes require to run pilot plant for extensive periods of time to evaluate the effect of the recycling or buildup of impurities, evaluation of new reagents and additives as well as for the training purposes of the personnel. For a new processes the running of a pilot plant is fundamental to compile information for the CAPEX and OPEX of the commercial plant.

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

years ago in another role we worked with Hazen Research in Golden on three of four pilots. One was really large pilot for new process some folks had come up with for Trona. It failed in the pilot. I remember looking at the as a large expense to the client, however it has was the key point that the larger project was then put on hold. I would say some major cost and risk avoided was avoided. However the expense of the pilot plant was not small.

Helena Russell
8 years ago
Helena Russell 8 years ago

That SAG conference was indeed a great forum. The HPGR/SAG argument was again in full swing. I still get stuck on one reality though - a SAG is capable of reaching a final grind target HPGR cannot. Just more utility.

One comment on "who" that's worth considering, is that it's a good idea in my mind to have the technology vendors represented in the pilot plant planning and operation. Its one thing for Hazen, SGS, Corem, ALS etc to have the lab versions of my equipment, and it's even better if they've had training from us, but to have our direct hands-on participation helps ensure that 

1) poor results aren't a result of improper or non-ideal operation of the pilot scale unit operations and 

2) there is an opportunity to incorporate creative ideas that might arise from the vendor's experience with other projects or technologies.

Bob Mathias
8 years ago
Bob Mathias 8 years ago

Normally my off-the-cuff response would have been doing a flotation pilot plant is a critical step. After reading this and after thinking about it for a few days I’m not so sure.

Please see some thoughts below addressing the question for flotation (which is the area I spend most of my time in these days).

* Why? - There are probably four main reasons.

a) To better understand the process especially in terms of circulating loads. From a rougher perspective many of these circuits are run open circuit today (Chinese designs not withstanding). If it is an open circuit rougher a basic kinetic float should give good information for the size of the cells and to tailor the froth crowding down the bank. So really a pilot is only giving information of the interactions in the cleaners and recycles streams. It will definitely give some insight but unless the feed sample, water chemistry and grind size and liberation are identical it may not be the final picture. I firmly believe that it can demonstrate that a process is possible but how good the scale-up information will be is another matter…..I guess one could also argue that good lock cycle test would may also have the same result. We can get good ‘scale-up’ information for any given flotation bank from doing a batch test….the trick then becomes knowing how much material is feeding the bank as this will effect flow (>residence time) and tonnage pulled.

b) For looking at the potential impact of moving locations of various process streams (especially in Greenfield plants).

c) To create enough concentrate sample for further downstream testwork (though I don’t know how often this is attempted).

d) To train operators on decision making and impacts – though today one could argue that there are some really great ‘virtual training packages for this’

Saying this I think the benefit of piloting is proportional to the complexity of the circuit. For a simple two stage copper float it may be a waste of time and money but for a complex multistage cleaning process with multiple recycles and regrinds it may be of huge value.

* What? - As above I think it may be useful for getting a handle on cleaner circuits and circulating loads. But there will always be a question mark on chemistry, grind and sample representativeness.

* When? - Ideally toward the end of testing to prove the process up and potentially try out some circuit changes and prove up the float. For Greenfield plants potentially to play with configuration or try new reagents as a risk mitigation.

* Where? - For Greenfield ideally close to site on the same process slurry with the same grind as the existing plant. Brownfield in a lab environment but with the same process water if possible. The cost of sample logistics and pilot size will likely determine location.

* Who? - I'm pretty open on this one but I will say that I think if you are running a pilot to design a plant it is probably worth getting in some engineering company people and also technical guys from the flotation equipment supplier (as per Tim's comments) to look at the plant run. In particular its good to see the froth characteristics as this may impact float cell design and froth factor used. Obviously you’d want these people to be experienced float specialists. Seeing the bubbles and the numbers always gives a more complete picture.

I think the real key to a successful pilot run is getting a good feed that is actually representative of the 'average' ore (this seems to be the message from the grinding people also). In this case I mean ore that represents a ‘typical’ mill feed blend based on the mining plan. This is always a challenge as it takes a lot of ore to run a large pilot plant and its not always possible to get this, but if the ore isn’t going to be what the plant is fed then what is the point! As for variability testing I think a better way of doing this would be with a nice standard batch testing setup then a model relating batch performance to cleaner performance.

I’d also like to make a comment on the engineering design side. If we are indeed running a pilot plant to get a handle on the circulating loads in the circuit then this is fine but from my experience what tends to happen is that during the engineering phase safety factors are added. These generally occur in three places on tonnages, on residence time scale-up and on froth factor. I’ve also noticed the less experienced the engineer the bigger these factors become! It does add cost and is not necessarily the best way to design (especially when looking at froth crowding). If all we are doing is running the plant and then multiplying up the values we’ve found by a large safety factor when does this over-design then negate the test-work ? I understand (in most cases) why this is done but I wonder does it diminish the value of the pilot work all together? I’ve noticed that some reasonably large mining houses have even got to the point where they do minimum test-work but design enough fat into the circuit to cover most eventualities > this has risk but the return is less time and money taken for testing. Rather they focus on getting the plant built and producing, then tweak it. It’s an interesting model.

Bob Mathias
8 years ago
Bob Mathias 8 years ago

I think that a flotation pilot plant is often useful especially to:-

1. Confirm scale-up kinetics, although use of JK Tech will help.
2. Look at the effects of recycling water.
3. Produce a bulk concentrate for marketing purposes. I know of one tin project that had F contaminants that killed the project.
4. Raising funding.

I'd be genuinely interested to hear comments from anyone who has used the Eriez Mini Pilot Plant equipment, with respect to whether/how the ability to run a continuous flotation/regrind with flexible circuitry with a smaller amount of sample really improves the bang for your buck. There are always a lot of concerns about the cost of having a very large amount of sample, and I hear a lot of concern about the ability to scale up properly from locked cycle batch tests. These are the benefits we market this equipment to provide. Personally I wouldn't call myself an expert in that type of test work, so I'm open to feedback from those in the group who are.

Paul Morrow
8 years ago
Paul Morrow 8 years ago

I have run and participated in many pilot plants for engineering and minerals processing operations. Electrolytic manganese dioxide production, Chloride copper leaching and purification, Electric furnace operations, Gold leaching and recovery, solvent extraction, flotation and leaching, AG and SAG milling and Magnetic Separation, Copper Gold leaching and recovery. Most of these pilots, while costing some money, have been beneficial to the eventual process operations and some have been used later to test modifications to the various processes.
Size counts. Piloting at bench scale, and during daylight hours, does not always show up what will happen in a 24 h continuous process, and pumping, slurry handling and solids conveying processes can be different for the smaller operation.
In leaching processes, maintaining temperatures and flows at consistent levels and avoiding piping blockages(that happen in the cold of night shift) has to be balanced with transfer and retention times, as well as available surge capacity between stages of a complete process. There are Minimum mill sizes that you can use for piloting AG or SAG operations and the mill operations have to be appropriately calibrated. Dry screening or ore handling requires dust and fiber considerations, and with smaller material quantities,cyclones in a circuit might have to be substituted by screens. For magnetic separations I have used an Eriez pilot unit which we modified to allow use of a smaller section of the 0.5m wide drum, but we still used a high volume of sample.
Flotation pilots I have performed on bench scale,small plant and full scale plant sizes for reagent and ore changes and subject to success have carried through to the whole plant.
With Rougher flotation testing, the major component is collection and recovery of the desired ore minerals; It is only with the cleaners that the regrind size has to give maximum liberation without slimes and the float needs to recover pure ore minerals less the unwanted contaminants,; so they are two different pilot-plant situations.
In Metals, concentrate, general product recovery, there is always a purity requirement to be achieved, particularly if the products are to be used to sway prospective buyers in the market.

A Point often Overlooked. When Using or running a pilot plant, appearance can often be used in the marketing strategy. I have had Directors and managers ask "why should we spend money on a paint job, because the plant will only be used for a short period and then discarded?" The answer from me has been that in a green fields or even a brownfield situation, the company will announce that they are running a pilot plant. Then the directors will want to have their prospective investors,merchant bankers and magazine reporters come to have a look. If the plant looks like a useless collection of rusty gear, then they onlookers will not be impressed and will leave with the wrong impression to publish.If the pilot is dressed up like a real plant and newly neatly painted with appropriate related colors, They will walk away impressed whether the plant is operating properly or not. Thus the return from a properly set out, painted and kept plant will be its major part played in impressing investors and thus bringing in the necessary funding for the final plant, quite apart from the technical information gained from running of pilot operations.
I am happy to discuss more if anyone wants.I like setting up and running pilot plants, Sometimes they are more complex and have more interface challenges than the big plants.

Paul Morrow
8 years ago
Paul Morrow 8 years ago

First off pilot plants, if built on site, do not have to be discarded when the "real" plant is built. It still can be used for training and test purposes. Never knew a process engineer (including me) who did want to try something different (reagents, retention time, etc.) that cannot be done without disrupting operations.

Then there is the point, if anyone really thinks they can build a 1/4 to 1/2 billion dollar concentrator based on a few kilograms of samples run in an air-conditioned lab half a world away using clean water and a process modeling program, and will give a guarantee and warranty on operational performance, I have some land for sale.

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

I have seen a very large plant built on bench scale information, with a person advocating that a pilot plant was not necessary. The plant ran with difficulty; lots of big shutdown problems and the process fume destroyed the local environment; all of this would have been discovered and remedied if they had piloted the project, but pilot was eliminated as a "money saving" step which cost more in capital replacement and operational down time than the pilot plant would have. You can see I'm all for piloting an operation. 

Victor Bergman
8 years ago
Victor Bergman 8 years ago

The basic question which always asked before pilot plan studies is that 'what are the risk associated if directly commercialized' and pilot plant studies are mostly conducted to address the issues identified such as design issues, repeat-ability & reproducible, product quantity to establish market, more realistic cost benefit, material handling issues, Safety issues at larger scale which is difficult to model. 

It might not very critical for simple gravity separation where most of things are known or modeled but it can be very critical for a hydro and metallurgical process where many things unknown such as process safety as well as impact of variation in process conditions on product. For a totally new concept 7 to 10 years are enough but for a known subject 3 to 5 should be sufficient.

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

I think a pilot plant should be practiced (yes or no) based on the actual plant capacity that is to be installed. For small size plant say 50-100 TPH capacity it makes less sense to have a pilot scale plant because its fractional cost will be more...hence we can proceed with lab scale test with proper design margin and varying operating options so that the plant may be fine tuned during commissioning.

However, if we are thinking for a large investment with plant capacity of ~500 tph, then we should think to invest in pilot plant. The total cost of pilot plant in this case will be much less as compared to actual plant cost. Also the pilot scale results will give more idea in terms of equipment sizing.

The pilot plant should always be designed and run before engineering of a project so that the correct capacity/size of equipment may be selected.

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

Hear are details and links to some references around use of pilot plants for various cases:

Hydrometallurgy - challenges with complexity, new processes, stability, etc.
Molnar, R., Hydromet piloting: what the mineral processor can expect, Proceedings CMP, 2006, 607-625.

Differences between mineral processing and hydromet pilot plants:

  • chemical and physical environments,
  • reactions taking place,
  • reactions affect solid phases in particular,
  • - extensive recycling and long residence times,
  • need more than 10 days of run time for stability and experimentation
  • of pulp rheology, sampling requirements, operational training time, product disposal

Mezei, A. et al., Can complex hydrometallurgical pilot plants effectively reduce project risks? Part One: Key metallurgical and technical management issues, SGS, 2006.

http://www.sgs.com/~/media/Global/Documents/Technical%20Documents/SGS%20Technical%20Papers/SGS%20MIN%20TP2006%2004%20Hydrometallurgical%20Pilot%20Plants%20To%20Reduce%20Project%20Risk.pdf

Pilot plant metallurgical testing is commonly included in feasibility studies with the intent to limit the risks associated with the commercial implementation of mining projects. Still, metallurgical problems are among the main causes leading to disappointing project outcome. The most serious difficulties often arise in the case of projects involving difficult ores, generally requiring novel and/or complex flowsheets. The paper discusses pilot plant practices proven to be effective in producing metallurgical test data of sufficient reliability to effectively minimise metallurgical project risks. The causes and remedies of poor metallurgical performance at the pilot plant level are also discussed. Broader technical management issues covered include design, construction and operation of complex hydrometallurgical pilot plants. Particular emphasis is placed on pilot plant operations with respect to their preparation, management, design, operation and most importantly, overlooked metallurgical issues. Piloting experiences discussed include flow sheets treating poly-metallic feed samples containing nickel, cobalt, copper and gold as pay metals.

Accordingly, the main causes of pilot-plant failure include:

  • Insufficient time, sample and human resources - or succinctly - not enough money;
  • Work scope and expectations are poorly matched
  • Over-confidence in the proposed flowsheet - generally combined with lack of key metallurgical expertise, as well as a lack of bench scale testing.

The majority of the problems occur when new processes are tested, and this is generally in relation to non-conventional feed stocks, where processing is carried out hydrometallurgically. The role of comprehensive pre-pilot testing in these situations cannot be understated. This paper complements some of the ideas put forward in a recently released SGS Minerals Services paper [4], which provides an overview of issues relating to hydrometallurgical piloting.

Please join and login to participate and leave a comment.