Mining Magazine Congress,

Min
Oct. 8-9, 2009, Niagara-on-the-Lake, Canada

HPGR – Accepted Technology in Mining

High Pressure Grinding Rolls have been used in grinding cement clinker and associated raw
materials since 1985, and had found a niche market in diamond ores and in pellet feed
preparation. However their entry into hard rock mining was not secured until recently, with
the successful commissioning and operation of the rolls at Cerro Verde in Peru.

Since then, the growth has been steady and persistent. HPGRs are now accepted as a vi-
able alternative to SAG mills. This has prompted an endless number of trade-off studies
comparing the two systems. With energy prices going up, and the rolls requiring less energy
for grinding than SAG mills, decisions are coming down on the side of HPGRs particularly in
the case of hard ores. This is not only applicable to large scale projects but also to smaller
operations, as will be shown in this paper.

Cerro Verde’s four HPGR Polycoms 24/17

Cerro Verde is a copper-molybdenum concentrator. The circuit consists of four (4) HPGRs,
rolls 2.4 m diameter x 1.65 m wide, working in closed-circuit with eight (8) 3 x 7.3 m wet
screens. Each is powered by 2 x 2500 kW variable speed drive units. They were commis-
sioned at the end of 2006. By mid-2007 the plant reached its design capacity of 108,000 t/d,
and a few months later was operating above design capacity; this continues to be the case
two years on. Each unit handles between 2600 and 2900 t/h. Feed to the circuit is nominally
<50 mm from secondary closed-circuit crushing. The product of the wet screens is < 5 mm,
P80 between 2.5 and 3.0 mm. Grinding is done in four 12 MW ball mills.

Earlier feasibility studies had identified that two (2) 40’ SAG mills, each 22 MW, and four (4)
ball mills, each 12 MW, as well as three (3) 750 kW pebble crushers, would be required.
Four (4) 750 kW secondary crushers are used in the HPGR circuit.
Z:\Websites\miningcongress.com\MM Congress Presentations\Final MMC presentations for web\Polysius - Rene Klymowsky\Polysius, Rene
Klymowsky.doc 1 / 7

Polysius AG
Graf-Galen-Straße 17, Chairman of Supervisory Board: Dipl.-oec. Hans-Peter Breker
D-59269 Beckum-Neubeckum, Germany Executive Board: Dr.-Ing. Detlev Rose, Chairman
Phone: +49 2525 99-0 Dr. Friedrich-Wilhelm Dierkes,
Telefax: +49 2525 99-2100 Dr.-Ing. Detlev Kupper, Dipl.-Ing. Norbert Patzelt
E-Mail: polysius@thyssenkrupp.com Trade Register: Amtsgericht Münster HRB-Nr. 7717
Internet: www.polysius.com Certified to ISO 9001 Registered Office: Beckum-Neubeckum
Page 2

The savings are not only in power, but also in the elimination of expensive grinding media.
Long term safety risks associated with SAG mill liner replacement were also eliminated. (1)

SABC circuits are sensitive to variations in size distribution and hardness of the ore. HPGR
circuits are far less sensitive to these variations. There are other issues associated with
SABC circuits which most trade-off studies fail to mention. These include the following:
- segregation on coarse ore stockpiles which results in SAG mills always receiving different
size distributions;
- pebble crushing circuits operating with wide fluctuations in feed and in contact with tramp
metal / grinding media add to the complexity of the circuits. Thus one of the benefits – sim-
plicity of the SAB Circuit – is gone.
- SAG mill screens that are at the limits of their capacities when handling the output of the
larger SAG mills.

Flexibility to adjust to harder ores is lower for SAG mill circuits than it is for HPGR circuits,
and the upside potential is limited, often by constraints in primary crushing. A further advan-
tage is that HPGRs can be designed with a significant catch-up capacity without affecting the
specific energy consumption, wear life or product size distribution.

Finally SAG mills have longer lead times, risks with larger castings, and far longer installation
times. The ramp-up time to full production is shorter for HPGRs as indicated above.

Wear has been the major impediment to acceptance of HPGRs into hard rock mining. Wear
to the sides of the rolls and welding of the edges has virtually been eliminated. To protect the
roll surfaces from wear, each roll is embedded with tungsten carbide wear resistant studs.
Wear lives have been steadily improving. It is anticipated that the newest set of studs will
have a wear life of close to 6000h.

HPGR feed chute, load cells, feed control gates, and diverter by-pass chute.

Z:\Websites\miningcongress.com\MM Congress Presentations\Final MMC presentations for web\Polysius - Rene Klymowsky\Polysius, Rene
Klymowsky.doc 2 / 7
Page 3

The units operate with one fixed position roll and a floating roll to which grinding pressure is
applied. The operating pressure is 120-130 bars and average operating gap is 60-70 mm.

The HPGR feed chute system is installed with metal detection and a diverter bypass chute to
avoid tramp metal entering the HPGR. The feed chute is on load cells to help maintain a
constant level for choke feeding the rolls. There are also feed control gates that help position
the feed in the rolls and prevent differential power draw.

Main comments from the operators have been that the production capacity can be main-
tained even with fewer (3) HPGRs, and that swings in throughput are avoided.

Whereas Cerro Verde was a milestone in the application of HPGRs to hard rock mining, it
was quickly followed by another “first” in copper-gold.(2) PT Freeport Indonesia installed two
(2) HPGRs, rolls 2.0 m diameter x 1.5 m wide, each powered by 2 x 1800 kW motors, at their
C1 and C2 concentrators in 2007. These employ conventional three-stage crushing.
Throughput is expected to be up to 80,000 t/d. The primary purpose was to reduce the flota-
tion feed size and to increase copper and gold production by 4-5%. The units were set up for
quaternary crushing duty, taking a feed that was already 80% < 7.5 mm and reducing this to
a P80 size of 3-4 mm for ball mill grinding.

PT Freeport Indonesia’s HPGRs with feed and by-pass chutes

The main drivers were a lower capital cost than a comparable ball mill option, the small foot-
print of the HPGRs, and the fact that the HPGRs could be commissioned 6 months ahead of
the ball mills. The latter is becoming increasingly important in today’s volatile markets.

The upside potential of the HPGRs also fit their strategic plan to be able to handle future ca-
pacity increases.

Z:\Websites\miningcongress.com\MM Congress Presentations\Final MMC presentations for web\Polysius - Rene Klymowsky\Polysius, Rene
Klymowsky.doc 3 / 7
Page 4

One feature of the HPGR circuit is that they are able to recycle part of the product to reduce
the fineness even more. Despite the fine product produced by the HPGRs, there have been
no detrimental effects in flotation due to overgrinding.

Too many fines can lead to flow problems in fine ore bins particularly at higher moisture lev-
els. Preparations were made to overcome these, but the problems did not occur.

Wear on these Freeport units has been quite low. Over 16 millions tons have been proc-
essed on each unit in the period to end of August 2009. Most recent measurements indicate
a projected wear life of > 20,000 h. This appears largely to be due to the fact that the feed is
much finer than the operating gap, thus minimizing ore-to-metal contact.

A third large HPGR was commissioned in the 1st Q 2008 for a very hard Platreef platinum ore
in South Africa. Bond Work Index 23-27 kWh/t. The rolls are 2.2 m in diam. x 1.6 m wide,
and is located at the Anglo Platinum Mogalakwena North Concentrator. Feed < 65 mm.
Product < 8 mm, 50% < 1 mm. Flowsheet includes dry screening. The technology was se-
lected over AG/SAG milling after a 7 month on-site trial had demonstrated 96% availability.(3)

The company has had more than 10 years

experience in operating AG mills. From the
outset variation in the ore due to rock char-
acteristics and differences in mine frag-
mentation resulted in operational instabili-
ties and varying monthly production levels.
In-circuit-crushing was installed to remedy
the situation. The ore was so hard that at
times the crushers could not keep up and
the stock of uncrushed ore was building. A
pilot HPGR was leased, and run for 7
months treating in excess of 185,000 ton-
nes of Platreef ore. At the end of the test-
ing period, 1000-ton batches of very abra-
sive UG2 and Merensky ore were run on
the machine. The tests were successful
demonstrating low wear, high reliability,
and producing size distributions that were
much finer than those obtained from con-
ventional crushing of the material. Most
importantly, the operating and maintenance
HPGR Polycom 22/16 at Mogalakwena plant
staff overcame their initial scepticism and
became supportive of HPGR technology.

The HPGR operates at a rate of 2100-2400 t/h. Circulating load is between 90 and 120 %.(4)
The design capacity of the whole circuit was 900 t/h, with the HPGR specified at 1200 t/h,
giving a catch-up rate of 300 t/h. The HPGR is not fitted with VSD drives, but the feed con-
veyor is fitted with these drives to maintain a constant level in the feed hopper. Predicted
wear life of the rolls is > 20,000 h.

Z:\Websites\miningcongress.com\MM Congress Presentations\Final MMC presentations for web\Polysius - Rene Klymowsky\Polysius, Rene
Klymowsky.doc 4 / 7
Page 5

Following the successful pilot plant runs on UG2 and Merensky ores, Northam Platinum in-
stalled a similar size unit in its UG2 concentrator in South Africa. (UG2 is a high chromium
bearing ore.) (5)

The plant was commissioned in June of

2008. The rolls are 0.95 m in diam. x 0.65
m wide; installed power is 2 x 200 kW;
throughput 160-200 t/h; feed < 32 mm;
product 75% < 1 mm.

The objectives were to increase produc-

tion, increase product fineness and reduce
chrome content in the concentrate.
(Chrome interferes with the recovery of
PGMs in smelting).

It achieved all of these aims with a 20-30%

reduction in grinding energy and a 4%
increase in recovery resulting in a payback
in 90 days.
Northam UG2 Plant Polycom 09/06-0

Newmont is currently commissioning four (4) large HPGR units at Boddington Gold Mines in
Australia. The roll are 2.4 m in diameter x 1.65 m wide. Each unit is powered by 2 x 2800 kW
variable speed drive units. The plant will have a capacity of 35 Mtpa.

The decision for Boddington to go ahead with HPGRs was in large part due to the successful
demonstration of HPGR technology at the Newmont Lone Tree Plant in 2003.(6)

The ore was characterized as being very hard (JK Tech index A x b = 27) with little variability
within the deposit. (7) Criteria often given for selecting HPGRs over SAG mills are an A x b
index < 40, consistent hardness, and a large deposit. Boddington is certainly a large deposit.
A trade-off study showed that while capital costs were a little higher for a HPGR circuit, 7%,
vs a SABC circuit, the operating costs and risks were significantly lower.(8)

Boddington will have a capacity 108,000 t/d, and the HPGRs will be working in closed-circuit
with wet screening. However the layout is different from Cerro Verde. The crushers and
HPGRs are all in a line using a forward crushing scheme. The Cerro Verde layout is more
compact and uses conventional screening before secondary crushing.

The newest HPGR coming to hard rock mining will be at Peñasquito in Mexico. The unit, with
rolls 2.4 m diameter x 1.65 m wide, will be installed in a SABC circuit to process crushed
pebbles. It will be the first application of HPGRs in pebble crushing in the non-ferrous miner-
als industry. (A small unit was installed in pebble crushing in an iron ore plant in the US in
1996, and proved its worth.)(9)

Z:\Websites\miningcongress.com\MM Congress Presentations\Final MMC presentations for web\Polysius - Rene Klymowsky\Polysius, Rene
Klymowsky.doc 5 / 7
Page 6

Peñasquito Circuit
HPGR 
Feed 2417

Primary  AG or  Pebble  +6 mm Wet 

Crusher SAG Mill Crusher Screen
O/F

Cyclone Trommel  +12 mm ‐6 mm

Product U/F Screen

‐12 mm
Ball Mill
Cyclone 
Feed 
Sump
Peñasquito circuit flowsheet.

The unit will have an installed power of 6000 kW, and is expected to increase the capacity of
two SABC circuits by 30% (from 100,000 tpd to 130,000 tpd) for an outlay of 4.8 kWh/t.(10)

The ore has been classed as hard, Bond WI 13.5, but low in abrasiveness. The pebbles rep-
resent the harder portion of the ore. The larger diameter rolls were selected to minimize the
risk of damage from tramp metal / grinding balls coming from the SAG mill.

The HPGR product will be wet screened on 6 mm; the undersize will be fed to the ball mill
cyclone feed sump, the oversize will be returned to the pebble crushers.

Conclusions

Examples have been given of different applications of HPGRs in hard rock mining:
HPGRs used in tertiary crushing duty, in quaternary duty, with recycle, with wet screening
and with dry screening, as alternatives to SAG mills and complimentary to SAG mills in peb-
ble crushing duty.

It has been shown that even on a smaller scale, installing HPGRs in existing operations can
increase production and reduce operating costs within a relatively short payback period.

An important feature of HPGRs is that they can maintain production even as the rolls wear.

The sheer proliferation of these machines in such a short time is testimony to the world-wide
acceptance of HPGR technology.

Z:\Websites\miningcongress.com\MM Congress Presentations\Final MMC presentations for web\Polysius - Rene Klymowsky\Polysius, Rene
Klymowsky.doc 6 / 7
Page 7

References

1) HPGR implementation at Cerro Verde, by J.L. Vanderbeek, T.B. Linde, W.S.Brack, and
J.O. Marsden, SAG 2006, IV-45 to IV-61.

2) A pre-production review of PT Freeport Indonesia’s High Pressure Grinding Roll Project,

by Mike Mular and John Mosher, SAG 2006, IV-62 to IV-79.

3) Developement of a process flowsheet for the new Anglo Platinum concentrator, incorpo-
rating HPGR technology, by Chris Rule, SAG 2006.

4) HPGR – Revolution in Platinum, by Rule C.M., Minnaar D.M., Sauermann G.M., Third
International Platinum Conference ‘Platinum in Transformation, SAIMM, 2008.

5) Cracking the Chrome Conundrum, Danie Minnar, Modern Mining, Dec. 2008, pp. 20-23.

6) HPGR Demonstration at Newmont’s Lone Tree Mine, I.B. Klymowsky, T.C. Logan, CMP
2005, Ottawa, CANADA,January 18-20, 2005.

7) Boddington Expansion Project, Brendan Parker and Gerald Veillette, Randol Conference
Perth 21-24 August 2005.

8) Case study – Investigation of HPGR suitability for two gold/copper prospects, by J.

Seidel, T.C. Logan, K.M. LeVier, G. Veillette, SAG 2006.

9) Application Of High Pressure Grinding Rolls In An Autogenous-Pebble Milling Circuit,

E.C. Dowling, Paul A. Korpi, R.E. McIvor, D.J. Rose, SAG 2001, III-194 to III-201.

10) Squeezing an extra 30% of a typical SABC circuit for 4.8 kWh/tonne, by Steve Dixon,
Barry Olson and Ed Wipf, SME 2010.

Z:\Websites\miningcongress.com\MM Congress Presentations\Final MMC presentations for web\Polysius - Rene Klymowsky\Polysius, Rene
Klymowsky.doc 7 / 7