Plant & Equipment

Optimising a conglomerate processing plant with simulation software

Civilisations have always been dependent on the art of quarrying to provide materials required for civil infrastructures.

My research has revealed that sustainable quarrying has many factors that must be considered to be successful. The Metromix-owned open pit quarry in Teralba has been providing such material to the Hunter region since the early 1960s.

Metromix was established in 1985 as an equally split subsidiary venture of CSR and Pioneer, and at that time comprised of five concrete plants located in Sydney. Since its inception, Metromix has grown to become a medium-sized construction materials company that employs more than 160 workers across 12 different sites throughout New South Wales.{{image2-A:R-w:320}}

The Teralba Quarry is the largest of three surface extraction sites Metromix currently operates. These sites are Marrangaroo Quarry, which is capable of 220,000 tonnes per annum (tpa) of processed quartz products, Anna Bay, which extracts 30,000 tpa of silica sand, and Teralba, which is capable of producing 750,000 tpa of conglomerate products.

The Teralba Quarry’s long existence has been attributed to its adaptability to changes in the civil construction market and underlying governing regulations.

One of these changes was the implementation of a new development approval. This came in February 2013 and effectively reduced the possible crushing hours from a 24/7 operation to 67 hours per week. This in turn put more emphasis on the crushing plant’s productivity.

The optimisation of the Teralba Quarry’s processing plant was identified as a way to adapt to these new DA requirements. Optimisation objectives to ensure Metromix’s operations remained sustainable for the duration of its new DA were:

  1. To increase the annual throughput of the processing plant to one million tonnes.
  2. To achieve a productivity rating of ≥85 per cent.

Metromix’s area manager Bill Sanderson first put the need for the Teralba Quarry’s optimisation to me. I jumped at the chance to take on such a significant and important project.

Part A of this project required analysis of the Teralba Quarry’s processing plant, to identify any critical components and bottlenecks that required optimising.

The Teralba Quarry’s current plant process is divided into two sections called the dry and wet sides. The dry side of the processing plant (Figure 1) is used to make unwashed products such as sub-base and fill material and is made up of five different circuits, which are the primary crushing, reclaim, secondary crushing, 100 minus and sub-base circuits.

The primary crushing circuit comprises a boot hopper, pan feeder, single deck screen (Screen 1), jaw crusher and conveyor belt (CV1). It connects to the reclaim circuit via CV1 and the reclaim circuit is comprised of a reclaim hopper, belt feeder and conveyor belt (CV14). {{image3-A:L-w:320}}

The head of CV1 discharges material into a transfer chute that sits on top of a double-deck screen (Screen 2). Screen 2 marks the beginning of the secondary crushing circuit. The overs from Screen 2 are dropped into a bin (Surge Bin 2) and then transported to a cone crusher via two conveyors (CV2A and CV3). The crushed material is then transported back to Screen 2 via CV4 and CV5.

The 100 minus circuit takes material from the primary crushing circuit and stockpiles it via CV6. The final circuit is the sub-base circuit, which takes material from Surge Bin 1 and stockpiles it via CV7 and CV15.

The wet side of the processing plant (Figure 2) produces washed aggregate and sand products.To achieve this, the wet side consumes about 500m3 per hour of water, of which 80 per cent is returned to silt dams and recycled. The wet side consists of three different circuits, which are the washing, sizing and sand circuits.

The washing circuit begins at a banana screen (Screen 3), where minus 4mm material is washed out and pumped to a silt tank. The plus 4mm material is then conveyed to a wet Barmac vertical shaft impactor (VSI), where it is shaped.{{image4-A:R-w:320}}

The material is then transported via CV8A to a coarse screw washer (Aggscrew), where it is scrubbed and dewatered. The plus 4mm material is then transported via CV9 to the sizing circuit. The sizing circuit begins at a two-deck horizontal screen (Screen 4), which splits the material into three different pathways. The plus 12mm material is transported to a triple-decked screen (Screen 6) via CV10.

This material is then placed according to its size into either the oversize, 20mm or 14mm product bins. The plus 4mm material is transported to another triple-decked screen (Screen 5) via CV11. This material is then placed according to its size into either the 5mm, 7mm or 10mm product bins.

The sand circuit converts the minus 4mm material from the washing and sizing circuits into sand products. The minus 4mm material from these circuits is pumped to a fine material screw washer (sand screw) via a launder tank. The sand from the sand screw is then stockpiled via CV12.

The next step in my analysis of the current plant process was to calculate the throughputs required to meet the project objectives.

I assumed a 48-week working year to factor in public holidays, maintenance and breakdowns. I also assumed production time would be divided equally between making dry and wet products. This resulted in the dry side and wet side each having 1608 hours of annual production time. {{image5-A:C-w:500}}

I then factored in losses pertaining to the 85 per cent productivity target and a 10 per cent silt loss from the washing process of the wet side. This yielded optimised throughput targets of 400 tph for dry side products and 342 tph for wet side products.

Bottlenecks in the current plant process were then identified by comparing the maximum capacities of each fixed plant component with that of the optimised throughput targets. Any component with a capacity below the optimised target was identified as a bottleneck.

I then adjusted the optimised throughput targets of all fixed plant components located after Screen 3 in the wet side to 300 tph. This was to account for the amount of fine material removed by Screen 3. The bottlenecks identified using this method are shown in Table 1.

In this project I defined critical components as any component within the processing plant that causes the entire production process to stop when it fails. After analysing the current plant layout one critical component was identified as CV1, because material entering the system from either hopper must be conveyed by CV1 before moving through to the next stage of the plant. Therefore there would be no way to introduce feed into the system if CV1 failed.{{image6-A:C-w:500}}

Part B of this project involved optimising the components identified in Part A and providing solutions for the problem described.

The optimisation of the plant was undertaken using software called Aggflow, which is a flowchart simulation program that allows the user to model plant layouts and then simulate its production process. Aggflow was developed in Vancouver, Washington, in 1993 and has been designed specifically to suit the mining and aggregate industries.

Before starting simulations in Aggflow, I took numerous belt cuts from the processing plant for testing. The gradings of these belt cuts were then used in Aggflow to increase the authenticity of my simulation results.

The optimisation of the identified bottlenecks and critical component was achieved by inserting upgraded fixed plant components and modifying the flow stream of the modelled processing plant. I continued this process until I obtained enough data to compile several solutions that met the project objectives.

At this stage of the study I required a logical and concise method of recommending the most suitable solution from all the data I had acquired. I decided to use a recommendation matrix method (RMM).

This method uses criteria I considered important to the Teralba processing plant, which were then given weightings ranging from 1 to 6, with 1 the lowest weighting. {{image7-A:R-w:320}}

A ranking scale ranging from 1 to 10 was then applied to rank the solutions in terms of their suitability. The criteria weighting was then multiplied by the ranking to yield a score. The scores were summed to yield a total recommendation matrix score (TRMS) for each solution. The most suitable solution for this optimisation was selected as the solution with the highest TRMS.

I considered four different washer upgrade options in my analysis of the Aggscrew: a log washer, a bigger coarse screw washer, a blade mill and a rotary scrubber. The results of this RMM analysis are shown in Table 2. It identified Option 3 (blade mill) as the recommended solution. I believe the blade mill’s ability to produce a better washed product at the required throughput – while consuming similar levels of water and power when compared with the current Aggscrew – makes it the most suitable option.

I considered three options in my RMM analysis of the critical component.

Option 1 was to store surplus spare parts for CV1. This would reduce the downtime of a critical component breakdown rather than eliminate the critical component.

Option 2 was to add an additional hopper to the plant layout that bypasses CV1. This would successfully eliminate CV1 as a critical component but it would be expensive to implement.

Option 3 was to modify the existing conveyor (CV14) to discharge material directly onto Screen 2. This successfully eliminates CV1 as a critical component and requires fewer resources to implement than Option 2. Therefore I recommended Option 3 as the solution.

In my analysis of the VSI bottleneck, I considered three different VSI upgrade options from three different manufacturers.
These options were a CME Auspactor VS300RR, a Trio TV85-ROS-SD and a Metso Barmac B7150SE. After conducting an RMM analysis and considering the advantages and disadvantages of each option, I recommended Option 3 (Metso Barmac VSI).

I chose the Barmac because it has a similar design and power requirements to the existing VSI. The other two options required higher-powered electric motors to shape the material. {{image8-A:L-w:320}}

This higher power requirement was deemed unsuitable because the surplus power from the motors would not be utilised in shaping the material, due to the composition of the Teralba conglomerate.

This is due to the conglomerate having an adhering matrix of sand and clay, which therefore requires less power to shape
it when compared with other harder geological rocks.

The most suitable method for increasing the throughput on conveyors 10 and 11 was to increase the conveying speed. I determined the conveying speed must be increased from 1.26 metres per second (m/s) to 1.4m/s.

To achieve this I recommend the diameters of the driven pulleys of conveyors 10 and 11 be reduced to 256.5mm and 252mm respectively.

My analysis of the Screen 1 bottleneck indicated that replacing the screen with a bigger machine would not be practical, due to the existing physical constraints of the primary crushing area. The alternative solution was to increase the closed side setting of the jaw crusher to 75mm, so it could accept the increased unscreened overflow from Screen 1.

This would result in an increased amount of larger material entering the secondary crushing circuit. Fortunately, Aggflow simulations showed the secondary crushing circuit has enough surplus capacity to handle this increase.

In conclusion, the findings and recommendations from this project were made to help the Teralba processing plant overcome the throughput challenges imposed by a new DA. Currently, the blade mill recommendation is being implemented, with the purchase and delivery of a new GreyStone blade mill.

It is my hope that Metromix management will continue to consider the recommendations from this study to help preserve Metromix’s operations at Teralba.

Finally, I hope the Teralba Quarry continues to provide the Hunter Region with quality products while operating in an efficient, sustainable manner for another 30 years.

Muhammad Yunusa is an IQA member and a mechanical engineering honours graduate from the University of Newcastle. He previously worked at Metromix for more than two years, where he was involved in and contributed towards maintenance, training, environmental, quality and safety operations at the Teralba Quarry. He now works for Hanson Australia.

Leave a Reply

Send this to a friend