Minerals Engineering xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Minerals Engineering
journal homepage: www.elsevier.com/locate/mineng

Operational parameters affecting the vertical roller mill performance

Deniz Altun a,⇑, Hakan Benzer a, Namık Aydogan a, Carsten Gerold b
a
Hacettepe University, Mining Engineering Dept., 06800 Beytepe-Ankara, Turkey
b
Loesche GmbH, Germany

a r t i c l e i n f o a b s t r a c t

Article history: Vertical roller mills (VRM) have found applications mostly in cement grinding operations where they
Received 1 June 2016 were used in raw meal and finish grinding stages and in power plants for coal grinding. The mill combines
Revised 18 August 2016 crushing, grinding, classification and if necessary drying operations in one unit and enables to decrease
Accepted 19 August 2016
number of equipment in grinding circuits. Sustainable operations and proven benefits of the technology
Available online xxxx
in cement grinding applications attracted interests of the mineral industry. Within the scope of the study,
ore grinding performance of the VRM was investigated via a mobile grinding plant and relationships
Keywords:
between operational parameters were identified. In this context, gold ore was ground under different
Vertical roller mill
Dry grinding
process conditions. The samples collected after the test works and size distributions were determined.
Comminution As a result of the study, effects of operational parameters such as working pressure, and classifier rotor
Ore grinding speed on product mass flow and product size were investigated. In addition, correlations of specific
Modelling energy consumption of the grinding-classifying circuit with product rate and product size are presented.
The aim was to develop a VRM performance prediction model that is based on input data from charac-
terization combined with the knowledge of the interaction of the single grinding parameters.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction significant effect on this improvement of VRM grinding efficiency.

Comminution is highly energy consuming and the head opera-
tion of the entire processing plant where the target size required
for subsequent unit operations e.g., leaching or flotation is pro-
duced. Therefore, energy efficient grinding technologies have
become the main interest of the industry. Up to now, various types
of machines differing in grinding and operating mechanisms have
been developed to meet the requirement of the industry. With the
introduction of the high compression machines such as High Pres-
sure Grinding Rolls (HPGR), Vertical Roller Mills (VRM) and Horo-
mills, efficiency of the grinding circuits was improved in terms of
energy and downstream effects. Among these devices, vertical
roller mills have an important position in cement and mining
industry.
In literature, there are many studies performed in cement
plants on comparison of conventional grinding systems and verti-
cal roller mills in terms of energy consumption. It was noticed that
30% energy saving could be provided by VRM for cement grinding
(Schaefer, 2002). Similar conclusions have also been drawn by
Tamashige et al., 1991; Ito et al., 1997; Roy, 2002; Simmons
et al., 2005; Jorgensen, 2005. The classifier performance has a
⇑ Corresponding author.
E-mail address: deksi@hacettepe.edu.tr (D. Altun).
The classifier reject material creates coarser material bed on the
table. The coarser bed increases the grinding performance, because
finer material in bed cause cushion effect. An efficient classifier
leads to decrease in grinding energy consumption and vibrations
as well (Schonbach, 1988; Becker et al., 1993).
In recent years, pilot plant tests with the Loesche Ore Grinding
Plant OGP mobile were performed. The aim was to identify if the
benefits proven for VRM’s in the cement industry can be trans-
ferred to the mineral industry. Pilot plant test results with various
ores showed that low energy consumption feature of vertical roller
mills is applicable to the ore industry. Grinding test results for zinc
ore showed that it is possible to decrease the total grinding energy
consumption from 20.11 to 11.40 kW h/t by using vertical roller
mill instead of AG/SAG-ball mill circuit (van Drunick et al., 2010).
In another test performed in the Loesche test centre in Germany,
22.9% in airflow-mode and 34.4% energy savings in overflow mode
were obtained for copper slag grinding (Gerold et al., 2012a,
2012b). This situation was confirmed in an OGP test campaign
for chalcopyrite ore with 18% less grinding energy consumption
over to conventional ball milling circuit (Altun et al., 2015).
In addition to low energy consumption, an increased mineral
liberation achieved by the in bed grinding principle compared to
conventional tumbling mill grinding could be identified. Positive
effects on performance of downstream processes like flotation,

http://dx.doi.org/10.1016/j.mineng.2016.08.015
0892-6875/Ó 2016 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Altun, D., et al. Operational parameters affecting the vertical roller mill performance. Miner. Eng. (2016), http://dx.doi.
org/10.1016/j.mineng.2016.08.015
2 D. Altun et al. / Minerals Engineering xxx (2016) xxx–xxx
leaching, etc. could be proven. In a study, flotation recovery of cop-
per and nickel ore was increased from 74% to 84% and from 79% to
86%, respectively (Viljoen et al., 2001). Test studies showed that
high Zn recoveries could be obtained at efficient energy use (van
Drunick et al., 2010). In another study, performance of gold leach-
ing has been improved by VRM system (Erkan et al., 2012). Beside
all these the ability to crush, grind, classify and dry within a single
unit, low wear rates, less floor space, etc. are other advantages of
VRM system.
In this study, pilot scale tests were performed with a mobile
vertical-roller-mill grinding plant to investigate the relationships
between operational parameters. The identification and quantifica-
tion of the grinding parameter relationships was carried out to
implement these relationships into a computational vertical-
roller-mill performance model.
The model consists of grinding and separation sections. For the
modelling of the grinding operation, breakage rates of the particles
are correlated with design parameters and material characteristics.
Separation operation of vertical-roller-mill will be modelled by
application of current models. The model shall enable the
prediction of the throughput, the particle size distributions
around the vertical-roller-mill circuit, the specific energy
consumption and in a later stage the wear to be expected in
dependency of ore treated. The model combines performance data
from pilot and industrial plants with mineral characterization
methods. The mineral characterization is based on mineralogical
studies combined with results from mechanical rock fracture test
work like piston and die tests. The model accuracy is going to be
increased with further test and mineral characterization
campaigns to be carried out.
Within this aim, grinding tests were carried out at different
operating conditions. At the end of this study, the relations
between operational parameters were evaluated to create a first
basis for a computational performance evaluation VRM model.
2. Materials and methods
2.1. Description of the mobile plant
Within the scope of the study, grinding tests were carried out by
using an LM 4,5 installed in the OGP mobile constructed by
Loesche for ore grinding. The installation of the OGP in a gold pro-
cessing plant is illustrated in Picture 1 Design and operational
parameters of the plant are summarized in Table 1.
Loesche OGP mobile can be operated in two grinding
modes characterized by location of the classification systems as
Picture 1. Picture of the OGP mobile in gold processing plant.
Please cite this article in press as: Altun, D., et al. Operational parameters affecting the vertical roller mill performance. Miner. Eng. (2016), http://dx.doi.
org/10.1016/j.mineng.2016.08.015
Table 1
Technical specifications of the mobile VRM
plant.
Throughput rate (t/h) 0.5–3
No of rollers 4
Table diameter (mm) 450
Installed power (kW) 420
Mill power (kW) 37
Heater power (kW) 300
Separator air flow (m3/h) 1500–5000
air-swept and overflow mode. In this paper, studies performed in
air-swept mode are presented. In the air-swept mode, the
vertical-roller-mill and a high efficiency dynamic classifier are sui-
ted above each other as a single unit (Fig. 1). Material is fed to the
grinding chamber and transported by table rotation to the grinding
gap between the fixed grinding rollers and the mill table. Crushing
and grinding take place by compressive grinding. Ground material
is transported over the edge of the grinding table and lifted pneu-
matically to the dynamic classifier. After classifying, coarse mate-
rial is sent back to the grinding table and combined with the
fresh feed. Classifier fine product is collected as final product in a
following bag filter.
2.2. Grinding tests with the mobile plant
The grinding test studies were performed in air-swept mode.
During the tests, steady state conditions were verified through
the control room data. Then samples were collected around the cir-
cuit. Totally 16 grinding tests were carried out for gold ore, whose
Bond Work Index is 18.5 kW h/t. Table 2 presents the operating
conditions of the mill at steady state conditions.
3. Material characterization studies
In this part of the study, the particle size distributions of col-
lected samples were determined. To identify the effect of the final
product particle size respectively the degree of mineral liberation
in dependency of the product particle size a variety of different
product particle sizes has been produced. Initially, dry sieving
technique was applied from top size to 150 lm, then sub 150 lm
size was measured using laser scattering technique. These two size
distribution data were then combined to obtain a full size distribu-
tion of the collected samples. Particle size distributions of product
samples are given in Fig. 2.
In addition to particle size distributions, mineralogical analysis
with the focus on the degree of mineral liberation and determina-
tion of breakage characteristics are part of the vertical-roller-mill
performance model study. Mineralogical analysis is a good way
to evaluate the effect of vertical-roller-mill on downstream pro-
cesses. For this specific ore the degree of mineral liberation is still
to be determined. In previous works (van Drunick et al., 2010) an
increase in the degree of mineral liberation by the usage of the
VRM compared to conventional technologies could be proven.
The breakage behaviour of material is critical for an accurate model
structure. Therefore, it is planned to carry out compressed bed
breakage tests to determine the breakage distribution function
and the breakage characteristics of the gold ore.
4. Results and discussion
In order to evaluate the performance of a comminution process,
energy utilization and the necessary degree of mineral liberation
and hence the size reduction are the key features that need to be
considered for a given device. As to be expected the specific
 D. Altun et al. / Minerals Engineering xxx (2016) xxx–xxx 3

Fig. 1. Flowsheet of Loesche OGP mobile in air-swept mode.

Table 2
Test conditions.*

Throughput Working pressure** Classifier rotor Product size Specific energy consumption***
(kg/h) (kN/m2) speed (rpm) (P80) (lm) (VRM-grinding) (kW h/t)
T1 315 600 409 44.6 21.4
T2 393 600 320 61.8 18.2
T3 454 600 260 84.2 15.8
T4 137 600 930 18.5 32.2
T5 349 720 400 48.5 22.0
T6 469 720 310 64.0 17.2
T7 537 720 250 91.6 15.8
T8 188 720 750 25.9 30.1
T9 432 1000 400 44.5 22.6
T10 541 1000 310 62.2 19.1
T11 645 1000 250 90.0 16.0
T12 214 1000 750 22.2 30.9
T13 575 1000 310 69.0 17.4
T14 598 1000 310 68.4 16.9
T15 631 1000 320 61.6 17.8
T16 623 1000 320 62.3 17.3
*
Mill differential pressure and dam ring height are constant.
**
The kN/m2 value is the pressure of the grinding roller applied onto the grinding table. The value accounts for the projected area of the roller onto the grinding table. The
projected area consists out of the mean roller diameter multiplied with the roller width on the grinding table. The working pressure in the actual grinding gap is much higher.
***
Classifier and fan power are not included in specific energy consumption.

Additional to the grinding energy the classifier motor and the

Fig. 2. Particle size distributions of product samples.
grinding energy consumption is decreasing with increasing pro-
duct particle size distribution (psd), at increasing product size
the throughput is increasing (Fig. 3).
fan motor are consuming energy. The specific energy consumption
of the classifier and the fan are determined by a separate scale up
procedure and can’t be directly taken over from pilot plant. Hence
these values were not given for the pilot plant. In a future case
study, they will be given for a possible industrial scenario.
The main parameter related with the product rate and product
particle size is the classifier rotor speed. The cut size of classifica-
tion can be controlled by adjusting the speed of the classifier rotor
in relation to the volume flow and the grinding pressure. Variation
of product rate and product size with classifier rotor speed shown
in Fig. 4a and b, respectively.
The differences in the production rate marked in Fig. 4a with the
red circle are achieved through the variation of the table rpm and
hence more efficient crushing and grinding. Effect of table speed
on product rate is given in detail in Fig. 5. With increasing table
rpm at constant classifier rpm more ore is transported into the
grinding gap and hence more ore is ground in the same time. The
product particle size increases from P80 = 62 lm to P80 = 69 lm
which is another factor for the increase of production rate.

Please cite this article in press as: Altun, D., et al. Operational parameters affecting the vertical roller mill performance. Miner. Eng. (2016), http://dx.doi.
org/10.1016/j.mineng.2016.08.015
4 D. Altun et al. / Minerals Engineering xxx (2016) xxx–xxx

Fig. 3. Relationship between specific energy consumption (grinding) & (a) product rate and (b) product size (P80).

Fig. 4. Relationship between classifier rotor speed & (a) product rate and (b) product size (P80).

The VRM technology is characterized by the possibility to

Classifier rotor speed= 310 rpm and 1000kN/m² control and adjust the circulating loading of the internal grinding
working pressure classifying circuit. The control of the circulating load is achieved
by adjusting the pressure drop over the mill comprising out of
the gas flow and the solid loading inside the mill. Higher classi-
fier rotor speed reduces the product particle size and hence
increases the amount of internal grinding classifying circuits.
Adjusting a fixed internal circulating load, defined by the pres-
sure drop over the mill, the product rate decreases. Fig. 4a
clearly displays that an increase of working pressure increases
the product rate. Variation of product rate with working
pressure at 310 rpm and 400 rpm classifier rotor speed for tested
gold ore is given in Fig. 6a. Working pressure effect on product
rate was verified with different gold ore data (Fig. 6b) and sim-
Fig. 5. Relationship between table speed & product rate at constant classifier rotor ilar results were obtained. For the tested gold ore an increase of
speed.
the working pressure of 66% is resulting in about 37% increase of
the production rate. This ratio could be found for P80’s around

Fig. 6. Relationship between working pressure & product rate for gold ores.

Please cite this article in press as: Altun, D., et al. Operational parameters affecting the vertical roller mill performance. Miner. Eng. (2016), http://dx.doi.
org/10.1016/j.mineng.2016.08.015
 D. Altun et al. / Minerals Engineering xxx (2016) xxx–xxx
Fig. 7. Relationship between working pressure/classifier rotor speed & product rate.
63 lm as well as for P80’s around 45 lm. As a result of the high
working pressure, the grinding efficiency is increasing and more
fines are offered to the classifier and leaving the grinding-
classifying-circuit faster. Table product size decreases and reject
material which cannot be transported to the classifier decreases.
This result in higher product rates.
Classifier rotor speed and working pressure are effective param-
eters on product rate as indicated. When the working pressure is
normalized with the classifier rotor speed Fig. 7 is obtained. The
graph shows that there is a steadily upward trend between the
working pressure/classifier rotor speed and product rate. Previous
tests results showed a correlation between the working pressure
and the degree of mineral liberation (van Drunick et al., 2010;
Reichert, 2015). Up to an ore specific value the increase of the
working pressure results in a higher degree of mineral liberation.
Increasing the working pressure further results in unwanted over-
grinding of the valuable minerals. Therefore, the working pressure
has to be adjusted to find a compromise between throughput and
the degree of mineral liberation. Further mineralogical test work is
planned to quantify the effect of working pressure. With this aim,
grinding tests at different working pressures will be carried out for
different minerals and mineralogical analysis will be performed for
tests products. In this way, it will be possible to correlate the liber-
ation degree of mineral with working pressure.
5. Conclusion
Within the study, effect of operational parameters on product
size and product rate was investigated. For this purpose, pilot plant
Please cite this article in press as: Altun, D., et al. Operational parameters affecting the vertical roller mill performance. Miner. Eng. (2016), http://dx.doi.
org/10.1016/j.mineng.2016.08.015
5
grinding tests were carried out successfully at different operating
conditions. The results showed that specific grinding energy con-
sumption has a correlation with product size and product rate. It
is observed that classifier rotor speed has an inverse proportion
with product rate and product size. Besides, working pressure is
normalized with classifier rotor speed and effect on product rate
was presented.
The obtained relations between the single grinding parameters
will be combined with the results of the mineral characterization
to build a first basis for a computational performance evaluation
VRM model. The model accuracy is going to be increased by an
increase of the data base by additional pilot and laboratory cam-
paigns as well as parallel mineralogical characterization test.
References
Altun, D., Gerold, C., Benzer, H., Altun, O., Aydogan, N., 2015. Copper ore grinding in
a mobile vertical roller mill pilot plant. Int. J. Miner. Process. 136, 32–36.
Becker, J.F., Sweeney, J.W., Wentzel, W.C., 1993. The design and manufacture of
large vertical roller mills. Zement Kalk Gips No.2, 40–43.
van Drunick, W., Gerold, C., Palm, N., 2010. Implementation of an energy efficient
dry grinding technology into an Anglo American zinc beneficiation process. In:
XXV International Mineral Processing Congress (IMPC) Proceedings, Brisbane,
QLD, Australia, pp. 1333–1341.
Erkan, E., Umurhan, S., Sayıner, B., Cankurtaran, M., Benzer, A.H., Aydoğan, N.,
Demir, H.K., Langel, J., Gerold, C., 2012. Comparison of the vertical roller mill
and rod-ball mill circuit on the gold extraction. In: Proceedings of XIIIth
International Mineral Processing Symposium, Bodrum, Turkey.
Gerold, C., Schmitz, C., Stapelmann, M., Dardemann, F., 2012a. Recent installations
and developments of loesche vertical roller mills in the ore industry. In:
Comminution’12 Proceedings, Cape Town, South Africa, pp. 1–22.
Gerold, C., Schmitz, C., Stapelmann, M., Dardemann, F., 2012b. Recent
installations and developments of Loesche vertical roller mills in the ore
industry. In: XXVI International Mineral Processing Congress (IMPC), New
Delhi, India, pp. 18–29.
Ito, M., Sato, K., Naoi, Y., 1997. Productivity increase of the vertical roller mill for
cement grinding. Cement Industry Technical Conference, Conference Record,
vol. 39. IEEE, pp. 177–194.
Jorgensen, S.W., 2005. Cement grinding-a comparison between vertical roller mill
and ball mill. Cem. Int. 3 (2), 54–63.
Reichert, M., 2015. Research of iron ore grinding in a vertical-roller-mill. Miner. Eng.
73, 109–115.
Roy, G.R., 2002. Increasing cement grinding capacity with vertical roller mill
technology. Cement Industry Technical Conference. Conference Record, vol. 44.
IEEE, pp. 205–211.
Schaefer, H.U., 2002. Slag grinding: latest advances. World Cem. 9, 61–66.
Schonbach, B.H., 1988. High efficiency separators in roller mills. World Cem.
(November)
Simmons, M., Gorby, L., Terembula, J., 2005. Operational experience from the United
States’ first vertical mill for cement grinding. In: Cement Industry Technical
Conference. IEEE, pp. 241–249.
Tamashige, T., Obana, H., Hamaguchi, M., 1991. Operational results of OK series
roller mill. IEEE Trans. Ind. Appl. 27 (3), 416–424. 99.
Viljoen, R.M., Smit, J.T., Du Plessis, I., Ser, V., 2001. The development and application
of in bed compression breakage principles. Miner. Eng. 14 (5), 465–471.