CASE STUDY

Mining Operation Conveyor Chute Optimization Benefits

from DEM Investigation Insight
By Dewald Tintinger, Technical Manager and Alwin Nienaber,
Technical Directos WEBA South Africa (Pty) Ltd

Discrete element method (DEM), a numerical technique, An engineering firm was contracted to review the flow
is often used to gain better understanding of particle functionality of the existing chutes with this scope of work:
flow dynamics. Mining equipment engineers leveraged
• Review and assess test work performed and identify
this practice in the initial stage of transfer chute design
any further test work required
optimization at a gold mine in northern Mexico. They
were mindful, however, that many DEM users apply the • Review current transfer chute designs
wrong input, which can result in misinterpreted results. • Assess feasibility and modification requirements for
Engineering firm WEBA South Africa conducted a series conveying three material subsets
of tests and calibrations on a subset of materials to
▫▫ Blend of waste rock and filtered tailings
determine the best direction for conveyor chute redesign.
▫▫ Blend of waste rock and filtered tailings with
The referred mine has an existing overland conveying minimum waste rock
system with a tripper conveyor and radial spreader. The ▫▫ Filtered tailings without waste rock
mining equipment currently handles waste rock (nominal
400 mm) at a rate of approximately 5,000 t/h. The facility • Recommend design modifications and cost at a
feasibility level for modified transfers to permit
planned to introduce filtered tailings onto the conveying
handling of a range of products from waste rock
system at a rate of approximately 1,200 T/h, generally only, waste rock plus filtered tailings, and filtered
along with 5,000 T/h of waste rock introduced upstream tailings only
of the filtered tailings. The filtered tailings would also be
conveyed when waste rock was unavailable.
 CASE STUDY
THE CASE FOR USING DEM The test work was performed on four material sets:

DEM is a numerical technique that models the interaction • Sedimentary dry tailings
between individual particles and boundaries to predict • Sedimentary filter cake
bulk–solids behavior. This tool can easily model moving
• Breccia dry tailings
boundaries in an overall effort to glean insight about
particle flow dynamics. In most applications, the • Breccia filter cake
knowledge is then applied to design more-efficient
equipment, thus improving process efficiency and product MATERIAL INTERNAL FRICTION ANGLE
quality as well as reducing design and operating costs. Sedimentary Dry Tailings 67°
Sedimentary Filter Cake 47°
DEM software has been proven to accurately predict bulk- Breccia Dry Tailings 59°

material flow patterns, bulk-material flow rates, impact Breccia Filter Cake 66°

forces on particles and boundary surface, wear patterns Table 1. Internal friction, continuous flow
on boundary surfaces, velocity profiles and dead zones, Performed by Jenike & Johanson
and particle distribution in segregation and blending.
Testing was conducted at the materials’ as-received
As with all forms of technology, DEM users must lay a moistures of 18.5%, 17%, 18.5% and 17%, respectively.
solid foundation at the outset to glean insight that can Samples were sieved to 1/4 inch for testing.
be trusted. Some investigators unknowingly use the
wrong input, which can lead to inaccurate results. Many MATERIAL BULK DENSITY RANGE (PCF)
Sedimentary Dry Tailings 84 - 132
engineers believe that this concept is robbing DEM of
Sedimentary Filter Cake 60 - 122
its credibility.
Breccia Dry Tailings 94 - 135

In addition to identifying the correct input, researchers Breccia Filter Cake 77 - 132

must ensure that material tests are done correctly. DEM Table 2. Summary: compressibility test results
runs its simulations based on material properties — bulk Performed by Jenike & Johanson

density, packing ratio, inter-particle friction coefficient,

Internal friction, compressibility and wall friction test
particle-to-friction coefficient, adhesion, cohesion, and
(Tables 1 and 2) were conducted, as relevant to the
particle-size distribution. Laboratory tests must always
original application: bin and hopper designs.
be performed using material screen between 0 mm to
-10 mm (typically -4 mm); they should not be assumed Further material behavioral tests, such as angle of repose
to be accurate if only a fraction of the material conveyed and critical chute angles, were performed; their results
consists of fine particles. When particle size is too fine, were considered during calibration and simulation.
the calculation is overwhelming, and the simulation will
not converge. To compensate, researchers must adjust Since the study mainly addressed dead-box chutes,
material properties so that big elements/lumps behave material calibration was based on internal friction angles
like fines. obtained during testing. Wall-friction angles measured on
Tivar® and mild CS HR plates were not applicable, since
MATERIAL TEST WORK AND CALIBRATION these materials flows onto themselves through the chutes,
not on a liner surface.
The mine’s combined bulk material was to pass through
four conveyor transfer points:

• CV-73 to CV-75A — inline chute transfer

• CV-75A to CV-76 — 90° transfer
• CV-76 to CV-77A — tripper conveyor to spreader

• CV-77A to CV-77B — transfer to rotating spreader

www.rocky-dem.com | © ESSS - All rights reserved.

 CASE STUDY
SEDIMENTARY DRY TAILINGS

After evaluating internal friction angles, the WEBA South Africa team decided
to use the worst-case scenario of sedimentary dry tailings measured at 67°.
Consequentially, the study did not consider sedimentary filter cake, breccia
dry tailings and breccia filter cake in the simulation.

Since the material test report does not include dry angle of repose, the assumed
angle is 38° (Figure 1). The first drop test that considered 0% moisture was
sedimentary dry tailings. Rolling resistance for 20-mm spheres was adjusted
until that angle was achieved.

Figure 1. Sedimentary dry tailings drop test

After calibrating the dry angle of repose, friction and material adhesion were
adjusted in an attempt to replicate the behavior of sedimentary dry tailings
including moisture. Figure 2 shows how the withdrawal angle of 67° matched
the internal angle of friction.

Figure 2. Sedimentary dry tailings including

moisture drop test

Once 20-mm spheres replicated fine-material behavior, some bigger particles

were added to represent the clumped material illustrated in the material test
report. Although 20-mm spheres create the same effect, bigger lumps were
added to increase simulation solve time. The material was then calibrated to
replicate the behavior of sedimentary dry tailings including moisture.

Figure 3 shows that the withdrawal angles are now close-to-vertical.

Figure 3. Sedimentary dry tailings with larger

particles drop test

www.rocky-dem.com | © ESSS - All rights reserved.

 CASE STUDY
MATERIAL BLEND CALIBRATION

SEDIMENTARY DRY TAILINGS BLEND WITH ROM CALIBRATION

Since there is no material test available on run-of-mine (ROM) particles,

material behavior was calibrated with particle size distribution (PSD) together
with previous information gathered on similar material.

Figure 4. Sedimentary dry tailings blend with Figure 4 illustrates a drop test for ROM material, a result of ROM only. The
ROM drop test
ROM was then blended with sedimentary tailings calibrated earlier in this
report and simulated through the inline chute.

CV-73 TO CV75A — SEDIMENTARY DRY TAILINGS

This simulation considered conveyor transfer points CV-73 to CV75A using

sedimentary dry tailings only without any blending. In terms of sticky material,
the chute handled the calibrated sedimentary dry tailings quite well. Simulation
showed no signs of chute blockage.

Considering the low tonnage of 1,480T/h, there appears to be available area

between the first dead-box and head pulley. The assumption is that when
material builds up close-to-vertical, there will be enough space when running
low tonnage.

Figure 5. Rocky DEM simulation of inline chute

transfer CV-73 to CV75A with sedimentary
dry tailings

PARAMETERS
Withdrawal angle 67°
Material build-up over time
PSD, 20 100-mm particles
Bulk density 1.6 T/m³
Tonnage 1,480 T/h

Figure 6. DEM simulation showing available area in first dead-box

www.rocky-dem.com | © ESSS - All rights reserved.

 CASE STUDY
It is not possible to simulate material that builds up during long-term operation.
Most chute blockages that handle sticky material, such as filtered tailings,
occur over time. Wet material settles in and dries quickly, then the next wet
material is ready to settle down. Over some time, this behavior can narrow
down and close available area, causing chute blockage.

In addition, DEM software cannot simulate the rhino-horn effect, in which

material builds back toward the trajectory. Because this effect happens slowly
over time, it is extremely difficult to simulate — as engineering simulation
software has little capability to handle longer time frames.

Material build-up over time

Air Cannons

Figure 7. DEM simulation showing available area and material flow behavior between first and second
dead-boxes

Although there seems to be ample area available between the two dead-boxes
(Figure 7), DEM software does not include tools to simulate scraper fines
and material splatter that builds up over time. The area circled in red appears
to be an open section exposed to no material; in reality, it is a low-pressure
zone that accumulates such splatter and fines. Blockages occur generally in
this region if flow is restricted so much that the main material stream cannot
pass through. One solution is to install air cannons programed to occasionally
blast open the zone.

CV-73 TO CV75A – MATERIAL BLEND

The simulation considered conveyor transfer points CV-73 to CV75A with a

material blend: sedimentary dry tailings with ROM material.
Figure 8. Rocky DEM simulation of inline
chute transfer CV-73 to CV75A with
material blend

PARAMETERS
Withdrawal angle 67°
PSD, 20 100-mm particles
Bulk density 1.6 T/m³
Tonnage 1,480 T/h
ROM
PSD, 20 100-mm particles
Bulk density 1.6 T/m³
Tonnage 5,000 T/h
Figure 9. DEM simulation showing relationship between head pulley and first dead-box

www.rocky-dem.com | © ESSS - All rights reserved.

 CASE STUDY
With the feeding conveyor traveling at 5.8 m/s, the impact angle into the
dead-box is approximately 85°. Such steep impact angles generally present
low resultant velocities, causing material to build up to angles close-to-vertical
before flowing down the chute. The trajectory angle from the first dead-box
is a good indicator of the material stream flow angle.

Figure 9 shows that the chute has ample area available after impact to handle
both ROM at 5,000 T/h and tailings. No blockages occurred during this
simulation.

Figure 10. DEM simulation showing relationship between first and second dead-box

Figure 11. Rocky DEM simulation of CV-75A

to CV-76 with 90°-chute transfer using Simulation results of the first and second dead-box area (Figure 10) are similar
sedimentary dry tailings
to those in Figure 9. This inline chute is big enough to handle high tonnage,
PARAMETERS even with sticky material building up close-to--vertical.
Withdrawal angle 67°
At first glance, the 3.5-meter drop height between the first and second
PSD, 20 100-mm particles
Bulk density 1.6 T/m³
dead-boxes raised concerns related to high velocities reached after free-fall.
Tonnage 1,480 T/h Generally, high wear and tear is an obvious factor in such cases. Continuum
calculations on material flow revealed that free-fall-induced velocity is
required to ensure that the resultant velocity after impact onto the conveyor
is approximately the same as belt speed.

CV-75A TO CV76 – SEDIMENTARY DRY TAILINGS

The simulation considered conveyor transfer points CV-75A to CV76 with a

90° chute using sedimentary dry tailings only without any bleeding.

In terms of sticky material, the chute handled calibrated sedimentary dry

tailings quite well, with no blockages. With the low tonnage,1,480 T/h, there
is sufficient area available between the first dead-box and head pulley (Figure
12). It can be assumed that when material builds up close-to-vertical, there is
enough space when running low tonnage.

Figure 12. DEM simulation of CV-75A

to CV-76 90° chute with sedimentary
dry tailings

www.rocky-dem.com | © ESSS - All rights reserved.

 CASE STUDY
However, there was one subject of concern: Figure13 shows the area available
between the first dead-box and material build-up from the bottom dead-box.
Taking into account minimal clearance, chute blockages are expected when
material build-up is any steeper than anticipated.

It appears that this dead-box was installed in an attempt to eliminate direct

belt impact. To alleviate any blockages, the dead-box must be shortened or
completely removed.

Figure 14. Rocky DEM simulation of CV-75A

to CV76 with 90°-chute transfer blending
sedimentary dry tailings with ROM material

PARAMETERS
Withdrawal angle 67°
PSD, 20 100-mm particles
Bulk density 1.6 T/m³
Tonnage 1,480 T/h
ROM
PSD, 20 100-mm particles
Bulk density 1.6 T/m³
Tonnage 5,000 T/h

Figure 13. DEM simulation showing chute blockages where material build-up is steeper than expected

CV-75A TO CV76 – MATERIAL BLEND

The DEM simulation considered conveyor transfer points CV-75A to CV76

with blended sedimentary dry tailings with ROM material. Figure 15 shows the
feeding conveyor traveling at 5.8 m/s, and the impact angle into the dead-box
is approximately 85°. Steep impact angles generally present low resultant
velocities, causing material to build up to close-to-vertical angles before flowing
down the chute. The trajectory angle from the first dead-box is a good indication
of material stream flow angle.

Figure 15. DEM simulation showing relationship between head pulley and first dead-box

www.rocky-dem.com | © ESSS - All rights reserved.

 CASE STUDY
The chute has ample area available after impact to handle both ROM at 5,000
T/h and tailings. No blockages occurred during blended material simulation. The
sedimentary dry tailings simulation, described earlier, shows chute blockages
where material build-up is steeper than expected (Figure 13). That area of
concern poses an even-greater threat when transferring blended material.
Considering the higher tonnage, the blended materials simulation shows that
clearance between material build-up and dead-box (Figure 16) pushes the
limit in terms of available area.

If material build-up becomes steeper or tonnage increases, blockages could

occur. Furthermore, if any foreign object passes through the chute, it could
get stuck between material build-up and dead-box. This would further reduce
any available area, increasing the probability of blockages.

This section should be closely monitored, as it will be the first to choke up. One
remedy is to shorten or completely remove the dead-box on the right side.

Dead-box

Figure 17. Rocky DEM simulation of CV-76

to CV77A tripper conveyor to spreader
with sedimentary dry tailings

PARAMETERS
Withdrawal angle 67°
PSD, 20 100-mm particles
Bulk density 1.6 T/m³
Tonnage 1,480 T/h

Figure 16. DEM simulation showing minimal clearance between material build-up and dead-box

CV-76 TO CV77A – SEDIMENTARY DRY TAILINGS

The simulation considered CV-76 to CV77A using sedimentary dry tailings

only with no blending, performed with the boom conveyor in a 90° position.

In terms of sticky material, the chute handled the calibrated sedimentary dry
tailings quite well, and no blockages occurred.

Considering the low tonnage, 1,480 T/h, there is sufficient area available
between the first dead-box and head pulley (Figure 18). When material builds
up vertical, enough space will be available for running such low tonnage.

Figure 18. DEM simulation of CV-76 to

CV-77A 90° chute with sedimentary
dry tailings

www.rocky-dem.com | © ESSS - All rights reserved.

 CASE STUDY
Figure 19 shows the available area and material flow behavior between the
first dead-box and bottom section. Although it appears that there is ample
area available between the two dead-boxes, DEM software does not include
tools to simulate scraper fines and material splatter that builds up over time.

Material build-up over time

Air Cannons

Figure 19. DEM simulation showing available area and material flow behavior between two dead-boxes

The area in red appears to be an open section exposed to no material; in

reality, it is a low-pressure zone that accumulates splatter and fines. Blockages
occur generally in this region if they restrict so much that the main material
stream cannot pass through. One solution is to install air cannons programed
to occasionally blast open the zone.

CV-76 TO CV77A – MATERIAL BLEND

The simulation considered conveyor transfer points CV-76 to CV77A blending

sedimentary dry tailings with the ROM material. Figure 21 shows the feeding
conveyor traveling at 5.8m/s, with the impact angle into the dead-box
approximately 85°. Such steep impact angles generally present low resultant

Figure 20. Rocky DEM simulation of CV-76 to

CV77A tripper conveyor to spreader blending
sedimentary dry tailings with ROM material

PARAMETERS
Withdrawal angle 67°
PSD, 20 100-mm particles Material build-up over time
Bulk density 1.6 T/m³
Tonnage 1,480 T/h
ROM
Air Cannons
PSD, 20 100-mm particles
Bulk density 1.6 T/m³
Tonnage 5,000 T/h Figure 21. DEM simulation showing relationship between the head pulley and first dead-box

www.rocky-dem.com | © ESSS - All rights reserved.

 CASE STUDY

Such steep impact angles generally present low resultant velocities, causing material to build up
to angles close-to-vertical before flowing down the chute. The trajectory angle from the first dead-
box is a good indicator of material-stream flow angle.

The chute has ample area available after impact to handle both ROM at 5,000 T/h and tailings.
No blockages occurred during blended material simulation. The same is true of material build-up
over time.

Figure 22. DEM simulation showing the chute’s worst-case impact angles

Considering the various angular positions between the feed and receiving conveyor, Figure 22
shows the absolute worst case in terms of impact angles within the chute.

Despite the material stream’s sudden directional change between the head and discharge chute,
the design proved to handle the tonnage quite well. There was no indication of blockage during
this simulation.

CV-77A TO CV77B – SEDIMENTARY DRY TAILINGS

The simulation considered conveyor transfer points CV-77A to CV77B using only sedimentary dry
tailings without any blending, with the boom conveyor in a 100° position.

In terms of sticky material, the chute handled calibrated sedimentary dry tailings quite well, and
no blockages occurred.

www.rocky-dem.com | © ESSS - All rights reserved.

 CASE STUDY

Figure 23. Rocky DEM simulation of CV-

77A to CV77B transfer to rotating spreader
with sedimentary dry tailings

PARAMETERS
Withdrawal angle 67°
PSD, 20 100-mm particles
Bulk density 1.6 T/m³
Tonnage 1,480 T/h Figure 24. DEM simulation with sedimentary dry tailings, no blending

Considering the low tonnage, 1,480T/h, there is sufficient area available

between the first dead-box and head pulley. Even when the material builds
up vertical, there will be enough space while running low tonnage.

Figure 25 shows area available and material flow behavior between the first
dead-box and bottom section. Although there seems to be ample available
area between the two dead-boxes, DEM software does not include tools to
simulate scraper fines and material splatter that builds up over time.

Material build-up over time

Air Cannons

Figure 25. DEM simulation showing relationship between first dead box and bottom

The area circled in red appears to be an open section exposed to no material;

in reality, it is a low-pressure zone that accumulates such splatter and fines.
Blockages occur generally in this region if they restrict so much that the main
material stream cannot pass through. One solution is to install air cannons
programed to occasionally blast open the zone.

www.rocky-dem.com | © ESSS - All rights reserved.

 CASE STUDY
CV-77A TO CV77B – MATERIAL BLEND

The simulation considered conveyor transfer points blending sedimentary dry

tailings with ROM material.

Figure 28 shows the relationship between the head pulley and first dead-box.
With the feeding conveyor traveling at 5.8m/s, the impact angle into the
dead-box is approximately 85°.

Figure 26. Rocky DEM simulation of CV-77A

to CV77B transfer to rotating spreader using
sedimentary dry tailings with ROM materials

PARAMETERS
Withdrawal angle 67°
Material build-up over time
PSD, 20 100-mm particles
Bulk density 1.6 T/m³
Tonnage 1,480 T/h
Air Cannons
ROM
PSD, 20 100-mm particles Figure 28. DEM simulation showing relationship between head pulley and first dead-box
Bulk density 1.6 T/m³
These steep impact angles generally present low resultant velocities, causing
Tonnage 5,000 T/h
material to build-up to angles close-to-vertical before flowing down the chute.
The angle of the trajectory from the first dead-box is a good indication of the
material stream’s flow angle.

It is evident that the chute has ample available area after impact to handle both
ROM at 5,000 T/h and tailings. No blockages occurred during the simulation
of the blended material. The same is true of material build-up over time.

Figure 27. DEM simulation using sedimentary

dry tailings with ROM material

SUMMARY

Detailed discrete element modeling studies of Figure 29. DEM simulation showing relationship between head pulley and first dead-box
conveyor transfer chutes set the stage for future
equipment design optimization at a gold mine in
Considering the various angular positions between feed and receiving conveyor,
northern Mexico. The engineering team contracted
to perform the assessment used best-practice this simulation using a blend of tailings and ROM materials was the absolute
techniques to accurately predict behavior of bulk
materials processed during mining operations
worst case in terms of impact angles in the chute (Figure 29).
— behaviors that cannot be observed nor
incorporated into prototype testing. Researchers Regardless of the sudden change of direction in the material stream between
worked diligently to ensure that they used proper
inputs, thus ruling out imprecise solutions. The the head and discharge chute, the chute proved to handle the tonnage quite
information they gleaned from the studies will be well. There was no indication of chute blockage in this simulation.
applied in the project’s next phase.

www.rocky-dem.com | © ESSS - All rights reserved.