Bulk Engineering, BULKtalk

BULKtalk: Bulk material flow and associated problems

Steve Davis summarises some of the problems and realities he has found when it comes to the design of chutes, bins and hoppers.

Steve Davis summarises some of the problems and realities he has found when it comes to the design of chutes, bins and hoppers.  

Recently, I have become involved with evaluating the designs and installations of chutes, hoppers and bins where the functionality is not asexpected. 

Why am I asked? Because performance is reducing productivity through low flow or blockages, or the cost of maintaining flow through repairs and cleaning is excessive, and to get a different perspective. 

A typical flow sheet is and hourly steady state representation of the desired annual throughput. Some organisations apply an annual downtime factor, e.g. 65 per cent utilisation, and this boosts the steady state flows to 154 per cent of the annualised flows. It does not reflect that most materials handling systems are not steady state in operation, even though they operate at the required average rate. 

Examples are apron feeders, where flow varies as the pans are discharged and cyclic fluctuation of plus or minus 25 per cent or more of average is possible with some materials, filter discharges are typically cyclic at no flow to perhaps 400 per cent average flow as the cake is discharged. Many processes take time to stabilise or fluctuate and bulk properties may be different from the “normal”, even if from a process perspective the product is the same. Stockpiles often must be refilled at a significantly higher rate than average when drawn to empty. If equipment is not sized correctly for these flow variations, then poor materials handling result.

Related stories:

Avoid gross errors of understanding. I reviewed a flowsheet that included a 50,000-tonne bin with a single small outlet to load a train. The bin was sized on the highest potential compressed bulk density, and when full would have stopped the entire process. The bin would have been monstrous, and the outlet almost certainly would not have functioned. The bin was changed to a ground stockpile with overflow capacity from push out. Another flowsheet had all conveyors designed at 1/8760 of the annual throughput, i.e. 100 per cent utilisation. Another plant was constructed with rock box chutes for ore that was wet and sticky.

Define the material. The simplest mistake can be a problem. For example, a large conveyor system designed with a higher bulk density than actual, resulting in overfilling and spillage from surging at design flows. 

We have a choice of excellent bulk testing laboratories, and the cost of testing is minor compared to the losses from a single error over life of operation. I recommend a basic understanding of the test outcomes and how they benefit design is used to discuss expectations with the laboratory. All laboratories will offer a “standard” test option list and can provide design guidance. 

Testing requires a suitable sample, not a pile scraped from the toe of a stockpile. Test results are pointless unless they are used correctly as the basis for design. 

Don’t assume that testing of a pit core sample will be representative of the ore as it is mined, crushed, screened, crushed again and processed further. 

Don’t test against wear material that is irrelevant. Having results for carbon steel is pointless if ceramic or other wear plate is to be used. 

Be realistic with the amount of moisture likely to be added for dust suppression or in processing. A fine filter cake at the end of a process is likely to handle very differently to the ore at the beginning. If you have an existing facility with the same material, use this as correlation, but for example, hard cap iron from one mine is not the same a below water table iron from another. Ore in the wet season may handle differently from the dry season. 

If a standard laboratory test suite mentions the bulk material to be (very) difficult to handle, take note, understand why and whether anything can be changed to improve flow. A bulk material was transported from China to Australia in containers. Testing showed it was probable we would have a solid block on arrival, so mechanical handling was not a sensible option.

Having understood the material properties and the rate of flow, the next step is to develop a design. The quick, easy and cheap design comes from taking an existing design and stretching it to suit. 

The first problem with this approach is that the design office rarely gets site feedback so doesn’t know whether the original works. The second problem is that this design could be inappropriate for a different bulk material, and the flow properties are often ignored. The third problem is that space constraint often forces the shape, whereas the shape should force the space required. The arrangements of chutes and hoppers is defined in the layout before understanding whether they will function well.

Growing in favour, Discrete Element Modelling (DEM) has been with us for some time. I believe the main reason for use to be the generation of attractive coloured animations. 

The problems I have seen with DEM start by assuming that being able to computer model a system to handle bulk materials comes with knowledge of bulk flow and manufacturing methods. 

Does anyone see a benefit from creating a model that does not have specific bulk material flow input data, including size distribution, shape variation and clarity on interparticle and interwall reactions? 

I have seen DEM models of ore flow where all particles were spherical and the same size and were proposed for a complex chute design for a problem material. I have seen DEM models where the flow is good for the 60 second in the animation, but with indication of potential material build up. Fine if the same 60 seconds is repeated over and over but given the variability of ore properties over short and long term, I take any such indication as potential failure. 

Animations often only show the flow condition and ignore the possibility of start from full, such as a blocked chute or bin that has been full for a few days. 

The worst observation, in my opinion, is that the models do not replicate the manufacturing process, with smooth 3D walls that could only be made by esoteric methods if at all. The internal walls of most chutes and bins and similar are never completely smooth, often have many small ledges and valleys that entrap fines and initiate build up and resist flow. I have not seen DEM consider a worn surface. 

Our common system appears to be one where the DEM design is passed to a designer who then replicates the DEM “on paper” for fabrication. I find that many designers and engineers do not understand the finer points of fabrication, and the result is something that looks like the DEM at a glance, might be fabricable and misses details of installation and maintenance, and where accessories such as access doors, belt cleaners, dust sprays and air canons are literally stuck where they fit, and not where they should be. 

Fabricators will do their best to replicate the design drawings, but they are under pressure and take short cuts. I have yet to see any fabrication detail go back into the DEM, with final shape and dimensions as a check of flow. I have tried but there never seems to be time or budget.

On site, we find subtle differences in dimensions, and to save time the contractor makes it fit, without understanding consequences. During commissioning, the designed in adjustment capability such as in chutes is rarely altered, and the chute operates in the position shown on the drawing, which could be good, but many aren’t. 

I have seen chutes where the head was mounted 0.2 m low allowing some material to miss altogether. I have seen others where the design hides the upper chute adjustment inside an inaccessible head box. Feeder outlets are still designed without flow relief.

Finally, the issues with materials handling components rarely get back to the designers, and so lessons learned are not passed on, good or bad. 

To make matters worse, it seems common for operations to make their own quick fixes when system don’t functions as expected, and we see fundamental changes in liner materials. 

Doubling the liner thickness at a hopper outlet reduces the outlet size and provides an all-round ledge on the hopper wall. Welding in random wear blocks (chocky blocks especially) rarely does anything positive except to go for another piece of low hanging fruit. Again, the operator generally does not chase the design organisation except for significant problems. 

I have seen well-meaning purchases of cheaper components, mainly liners, which result in reduced operating life. Strangely, the obvious solution of going back to the original or something better doesn’t always happen.

Finally, the upgrade comes along, and well-intended changes are made. Conveyors are easy to speed up, but chutes are unchanged, the chute has a problem that is fixed without engineering input, and the chute gets worse. I have seen chutes become a scheduled maintenance item because it is easier than engaging specialist guidance and fixing it well. Hoppers, bins, and stockpile are left unchanged during an expansion and are insufficient to provide expanded surge causing outages. 

Is the problem that we are unwilling to commit the time, money and resources to engineer a good solution? Are we bound by other benefits of using a one-stop shop, but where in depth knowledge is lost in being a generalist? 

The knowledge and capability to get things right is certainly available, and like many aspects of good bulk materials design, the smaller specialised organisations are available and often have proven results from implementing a solution and feeding the results back for continuous improvement. I have, however, seen a few examples of projects paying for good advice or product and then ignoring the advice or installing the product differently to recommendations – strange but true. 

Send this to a friend