Dipl.-Ing. Th. Fahrland, Dr.-Ing. K.-H. Zysk, Loesche GmbH, …€¦ · · 2016-09-26Loesche...

CEMENTS GROUND IN THE VERTICAL ROLLER MILL FULFIL THE QUALITY REQUIREMENTS OF THE MARKET

Dipl.-Ing. Th. Fahrland, Dr.-Ing. K.-H. Zysk, Loesche GmbH, Duesseldorf, Germany

IN DER VERTIKAL-WÄLZMÜHLE GEMAHLENE ZEMENTE ERFÜLLEN DIE QUALITÄTSANFORDERUNGEN DES MARKTES

1304268_VRM_unrivalled_flexibility_02.indd 1 30.07.13 14:53

64 CEMENT INTERNATIONAL 4 2/2013 4 VOL. 11

ZUSAMMENFASSUNG

Cement production typically requires the grinding of three

separate types of material during the process: the raw ma-

terials and coal before the kiln, and the final cement product

once firing is complete and the clinker cooled. Looking back

on a century or more, ball mill systems were used for all

three grinding stages, but the development of more energy-

efficient vertical roller mills (VRMs) led to their replacement.

Initially, this focused on grinding coal and the cement raw

materials, with the adoption of vertical roller mills for cement

product grinding – with its finer grinding requirements – com-

ing more recently, in the late 1990s. The main reason for

the delay in uptake of VRM technology for cement grinding

was the concerns of producers that their product qualities

would not meet market requirements, specifically in three

key areas: water demand, strength development and set-

ting times. Over the past 15 years, however, it could be

demonstrated that these concerns are unfounded, and that

the quality of cement obtained from VRM grinding is as

good as, or in some cases better than, that produced in a

ball mill. In consequence, most of the world’s major cement

producers now use vertical roller mills for cement grinding

with no hesitation.3

Im Zementherstellungsprozess müssen gewöhnlich drei

unterschiedliche Materialien fein gemahlen werden: vor dem

Zementdrehofen die Zementrohmaterialien sowie die als

Brennstoff verwendete Kohle und nach dem Klinkerkühler

der gebrannte Klinker zur Herstellung des Endprodukts

Zement. Vor mehr als 100 Jahren wurden zur technischen

Realisierung dieser drei Mahlprozesse fast ausschließlich nur

Kugelmühlen eingesetzt, meist im Kreislauf mit einem Sich-

ter, bis diese Mühlen durch die Entwicklung der energieeffi-

zienteren Vertikal-Wälzmühlen immer mehr ver drängt wur-

den. Zuerst vollzog sich dieser Wandel bei der Kohlemahlung

und der Rohmehlerzeugung, bis in den späten 1990er Jahren

die Vertikal-Wälzmühle auch bei der Zementmahlung mit

ihren spezifisch hohen Anforderungen an die Mahlfeinheit

den Einzug hielt. Der Hauptgrund für die späte Akzeptanz

der Vertikal-Wälzmühle bei der Zementmahlung war die ver-

breitete Skepsis bei den Zementherstellern, dass die Qualität

des in der Vertikal-Wälzmühle hergestellten Finalprodukts

Zement den Marktanforderungen nicht entsprechen könnte,

speziell was die Anforderungen an den Wasserbedarf, die

Festigkeitsentwicklung und die Verarbeitungszeiten betrifft.

Über die letzten 15 Jahre konnte jedoch der Nachweis

erbracht werden, dass die bisherigen Bedenken unbegründet

waren und der in einer Vertikal-Wälzmühle erzeugte Zement

hinsichtlich seiner Qualität einem Kugelmühlenzement nicht

nachsteht, fallweise sogar besser sein kann. Im Ergebnis die-

ser Entwicklung entscheiden sich deshalb heute die meisten

Zementhersteller weltweit ohne jegliche Bedenken bei der

Zementmahlung für den Einsatz von Vertikal-Wälzmühlen.3

SUMMARY

4Dipl.-Ing. Th. Fahrland, Dr.-Ing. K.-H. Zysk, Loesche GmbH, Duesseldorf, Germany

(English text supplied by the author)

PROC

ESSI

NG

64-69_Fahrland.indd 64 02.04.13 16:061304268_VRM_unrivalled_flexibility_02.indd 2 30.07.13 14:53

CEMENT INTERNATIONAL 4 2/2013 4 VOL. 11 65

1 Introduction

There is no question that vertical roller mills like the Loesche Mill offer significant advantages over ball mills in terms of their energy efficiency. As noted in a current publication (1) the specific power consumption of a ball mill is higher than that of a vertical roller mill (VRM) carrying out the same oper-ations by a factor of between 1.5 and 2, depending on the degree of optimisation of the ball mill. � Fig. 1 illustrates this connexion, as well as showing the increasing energy bene-fit that can be obtained with a vertical roller mill as the spe-cific Blaine surface area rises.

Cements ground in the vertical roller mill fulfil the quality requirements of the marketIn der Vertikal-Wälzmühle gemahlene Zemente erfüllen die Qualitätsanforderungen des Marktes

PROC

ESSI

NG

matically some possible alternative flowsheets using VRMs and ball mills.

With the focus here being on ball mills and VRMs, Table 1 shows some comparative performance parameters for the two systems when used for grinding cement. It has to be remembered that there are major differences in the mech-anism of grinding between VRMs and ball mills, in terms of how grinding occurs, the residence time, the level of repeat grinding and recirculation factors, among others.

In a VRM, comminution occurs by pressure and shear forces that are introduced via the grinding rollers. In ball mills, com-minution is mainly done by impact, with the grinding balls being lifted up by the rotating shell, then dropped back onto the charge and other balls. There is some attrition as well.

There is also a major difference in terms of the average resi-dence time – the time the material particles remain in the mill system before they leave the classifier as product. Including both grinding in the mill body and a circulation factor, the res-idence time for a VRM is less than one minute, while parti-cles can remain within a ball mill system for 20 to 30 min-utes. Individual cement particles will be ground from one to three times in a VRM before being offered to the classifier, whereas repeat grinding in a ball mill is virtually uncounta-ble because of the grinding mechanism. Finally, while a ball mill will have a recirculation factor of 2 to 3, this increases between 6 to 20 for a VRM, depending on the pressure height, the grinding tools configuration, the grindability of the material and the required product fineness. The differences in all of these parameters are shown in � Table 1.

3 Outlining the issues

The first modern Loesche Mill for cement and slag grinding, a mill with the designation LM 46.2+2, was sold to Taiwan’s Lucky Cement Corp. in 1993 and commissioned in 1994, for grinding cement at its Pu Shin plant. While producers were initially concerned that the quality of the cement produced

Ball millsystem

LoescheVRM system

Pow

er c

onsu

mpt

ion

[kW

h/t] 100

9080706050403020

3 000 3 500 4 000

Fineness acc. to Blaine [cm2/g]

4 500 5 000

Figure 1: Specific power consumption of ball mill system v/s vertical roller mill system for OPC grinding

Ball millsystem

LoescheVRM systemPo

wer

con

sum

ptio

n [k

Wh/

t]

20

40

60

80

100

120

140

160

Fineness acc. to Blaine [cm2/g]

5 0004 0003 000 6 000

Figure 2: Specific power consumption of ball mill system v/s vertical roller mill system for slag grinding

Fundamentally, ball mills use proportionately more energy to produce a finer ground product than do VRMs, and while the energy consumption in a VRM obviously does increase with product fineness, it does so much less rapidly. Indeed, while the difference in energy efficiency is significant when a VRM is used for grinding OPC, the energy benefit is even greater in the case of blastfurnace slag which is hard to grind, as can be seen in � Fig. 2.

Having demonstrated the energy advantage of the VRM concept over ball milling, the remainder of this paper focuses on various aspects of cement quality, with the aim of putting to rest once and for all the outdated, disproven ideas that cements produced in a VRM differ markedly from those produced in traditional ball mill systems. The arti-cle looks first at the various grinding systems that are avail - able for cement producers today, then takes each of the pur-ported product quality issues in turn to show that these are neither justified nor valid.

2 Grinding system options

Today, cement producers have the option for using a range of different systems for cement grinding. A comprehensive list of all the available options would certainly include tradi-tional ball-mill systems, high-pressure grinding rolls in every kind of design types and their various combinations with ball mills and, of course, VRMs vertical roller mills. All of these systems treat the material to be ground differently, in that the actual grinding needed to achieve the desired character-istics varies from one to the other. � Fig. 3 illustrates sche-


66 CEMENT INTERNATIONAL 4 2/2013 4 VOL. 11

PROC

ESSI

NG

would not meet their clients’ specifications, results from the first installations showed that the cement qualities are indeed acceptable to the market. From the late 1990s, the majority of cement producers changed their preference towards the vertical roller mill system.

The producers’ concerns were centred on three specific areas: that cement produced in a VRM would have a higher water demand when mixed to a workable paste; that the setting times would differ drastically; and that the compres-sive strength would be lower when compared to the same cement produced in a ball mill system. The supposed reasons for these concerns were, respectively: a steeper particle size distribution, the different particle shape produced by a VRM, and a lower gypsum dehydration. In point of fact, operating experiences since then have shown conclusively that none of these concerns is justified, and that the cements produced by grinding in a VRM meet market requirements in all respects.

4 Particle size distribution

The particle size distribution of a cement is usually plotted on the well known RRSB-diagram. � Fig. 4 shows size distribution curves for cements produced by grinding in VRMs and in ball mills. The inclination of each curve, the slope ‘n’, is measured at the positioning parameter that repre-sents the particle diameter at which the residue, in terms of mass, is 36.8 %. A higher ‘n-’value produces a steeper curve, whereas the lower the slope, the more fine and over-ground particles are in the finished product at a constant Blaine value, measured in cm2/g.

Fig. 4 shows that in this case, ‘n’ for the VRM cement sys-tem is steeper than for a cement produced in a ball mill. This is caused by the higher proportion of fine (over-ground) ma - terial present in the ball mill cement, which in turn reflects the greater number of impacts and the inherent inefficiency of ball mill grinding.

In more general terms, a typical particle size distribution for a ball mill system in closed-circuit operation with a high- efficiency, third-generation classifier would be between 0.75 and 0.98. The equivalent would be between 0.82 and 1.05 in a Loesche vertical roller mill system. These ranges may differ when different types of laser sizers are used.

The area of concern in the past was that the steeper particle size distribution for VRM cements would lead, for example, to their having higher water demand and lower early strength development. This could, of course, generate problems, especially in precast concrete manufacturing with its sophis-ticated production processes regarding cycle and stripping times.

The question then has to be asked as to the reason for the different slopes in the particle size distribution diagrams. Firstly, and as explained above, the different grinding behav-iour of the vertical roller and ball mill systems means that particles remain in a ball mill system for about 20 to 30 min-utes before they leave the classifier as product. They are impacted repeatedly during that time, with the result that some particles are ground more than necessary. By con-trast, the limited amount of grinding for each particle in the VRM system before classification avoids any unnecessary over-grinding.

Clinker Gypsum Limestone

Product

Clinker Gypsum Limestone

Product

Ball mill system

Vertical roller mill system

Figure 3: Alternative flow sheets using vertical roller mills and ball mills

CharacteristicsBall mill

(closed circuit)Vertical roller mill

Comminution by Impact and attritionPressure and shear forces

Residence time [min] 20 to 30 < 1

Crushings before separation

∞ 1 to 3

Circulation factor 2 to 3 6 to 20

Wear rate [g/t] ~ 50 3 to 6

Table 1: Comparative performance parameters for the two systems when used for grinding cement

VRMSystem

Ball millsystem

Particle size [μm]Sum

of p

assi

ng Q

(x) [

mas

s %

]

(Slope n: 0.82 to 1.05)

(Slope n: 0.70 to 0.98)

Slope n = tan α

Position parameter

Figure 4: Different particle size distributions



PROC

ESSI

NG

If a similar product to one from an existing ball mill system is needed, however, the VRM can be adjusted to achieve this. As shown in � Fig. 5, adjustments can be made to the grind-ing pressure, the dam ring height, the mill airflow, the clas-sifier rotor speed and, for cements with a very high Blaine value, the table speed. � Fig. 6 shows the effects on the slope of the particle size distribution curve by adjusting each of these parameters.

A higher grinding pressure will result in more intense grind-ing with the development of more fine material, and hence a shallower particle size distribution. The higher the dam ring, the longer the material remains on the grinding table, and the more it is ground in one cycle. This again results in the generation of more fines, and a shallower particle size distribution. In addition, the airflow and the classifier set-tings can be adjusted in order to produce the desired prod-uct characteristics.

Testwork carried out in 2009 at the Vicat group’s Montalieu plant in France has proved conclusively that these adjustments will deliver the required results. The study involved grinding the same cement once in a closed-circuit ball mill system and once in a grinding plant with a Loesche mill LM 53.3+3. � Fig. 7 shows some results from the project, confirming that the positioning parameter, the slope of the particle size distribution and the product fineness of both cements are in exactly the same range. Furthermore, the water demand for both cements needed to create a cement paste with a stand-ard consistency are the same, reflecting the same slope ‘n’, positioning parameter and Blaine values. This testwork proved in a practical way that there is no difference in material properties due to the differing grinding mechanisms.

5 Particle shape considerations

Differences in particle shape have been another area in which vertical roller mill systems have been the subject of scrutiny. The suggestion was made that cement particles coming out of a ball mill are always much rounder then those coming out of a VRM, which are more shallow and elongated. This again, together with fewer fine particles, would result in increased water demand for the cement paste.

The roundness or circularity of the particles can be deter-mined by optical methods such as image analyses. Within those measurements, particle shapes are described with val-ues of between 1 and 0. A value of 1 indicates a particle that is perfectly spherical, and as the value approaches 0, it indi-cates an increasingly elongated, shallow polygon.

� Fig. 8 illustrates the particle size distribution of a cement with a product fineness of about 4 100 g/cm2 acc. to Blaine. This shows that the size of the smallest particle is about 0.1 µm and the size of the largest particle is about 52 µm for this particular cement. In general, while a wide variety of cements are ground to different finenesses, the largest par-ticles are usually between 45 and 55 µm.

In addition, 95 % of all the cement particles are below 45 µm in size, again depending on the final product fineness and the slope of the particle size distribution curve.

Testwork undertaken by the VDZ in its role as Germany’s Cement Research Institute compared the particle shapes produced in different types of milling systems. � Fig. 9 shows a plot of the shapes of cement particles from high-pressure grinding rolls, ball mills and vertical roller mills, with the par-ticle size being shown on the x-axis and the circularity on

the y-axis [2].

It is clear from this plot that there is no significant difference in particle shape between the cements produced in all three of these mill types, apart from at a particle size of around 58 µm, where the shapes begin to differ. How-ever, this particle size rarely occurs in most types of cement, and where it does, it only appears in such small proportions that there is no severe influence on the water demand, strength development or other issues.

Dam ring

Grinding pressure Dam ring height Mill air�ow Classi�er rotor speed Table speed for very high Blaine cements

Figure 5: The operational parameters

Grinding pressure

Mill air �ow

Height of dam ring

Classi�er speed

lower

higher

lower

higher

higher

lower

higher

lower

Slop

e n

<==

Shal

low

St

eep

==>

Figure 6: Simple adjustment options to achieve the desired product

Technical properties

Mill VRM Unit BM Separator LSKS O-SEPA Density g/cm3 3.164 3.152 Blaine cm2/g 4 258 4 095 Slope n - 0.93 0.92 Position parameter d' µm 11.7 12.6 Water demand (Standard consistency) % 28 28.5

Resi

due

[mas

s %

]

Particle size DK [µm]

Figure 7: A particle size distribution as needed can be easily produced within a vertical roller mill system


68 CEMENT INTERNATIONAL 2/2013 VOL. 11

PROC

ESSI

NG

6 Gypsum dehydration optimisation

Gypsum is mainly added to cement to act as setting regu-lator, with the precise amount and type of gypsum being chosen in relation to its solubility and the individual clinker. A producer will work out the correct proportion and type of sulphate to be used through product optimisation in order to fulfil optimal requirements.

The energy input into the mill and the hot gases will heat the cement, with the gypsum being partially dried and con-verted to hemihydrate, so-called bassanite or plaster. Mixed together with water, hemihydrate dissolves better than gyp-sum, so the dilution is more reactive in regulating the cement set. Partial dehydration is intended within the milling process, but if it exceeds or underruns a certain value, it can result in accelerated setting behaviour. This in turn can lead to prob-lems such as with the workability of the cement paste.

As the retention time in ball mills is 20 to 30 times greater than in vertical roller mills, the cement is exposed to the hot-gas atmosphere for much longer. In addition, ball mills use much more energy than vertical roller mills in order to grind the same amount of cement, with this additional energy heating the material even further. Because of this, the par-ticle temperatures at the exit from ball mills operated with-out water cooling are usually about 30 °C higher than those experienced with vertical roller mills.

Therefore, if the same cement recipe is ground in both mills, paste from cement ground in a vertical roller mill will exhibit different setting behaviour and strength development during

hardening. This is because the gypsum is dried more inten-sively in a ball mill system, resulting in the formation of more hemihydrate and hence higher sulphate solubility.

Of course, commissioning any new mill requires the system to be optimised, including adjusting the amounts of gypsum and the other additives needed to achieve the required set-ting and strengthening behaviour. The same applies when changing from a ball mill system to a vertical roller mill sys-tem, when the operator needs to increase slightly the amount of gypsum put into the cement, or substitutes the gypsum to increase the solubility by hemihydrate or anhydrite.

Other optimisation measures include increasing the mill exit temperature and/or reducing the moisture content of the mill gas flow, both of which will enhance the drying pro-cess of the added gypsum. As summarised in Fig. 10 these are standard process-optimisation procedures that can also be used to ensure that the cements produced in vertical roller mills have comparable setting times and compressive strength development to those from ball mills. Because it is partly true that cement produced in a vertical roller mill has lower gypsum dehydration, this can be adjusted through standard process optimisation.

7 Final remarks

Table 2 presents a summary of the results obtained from the comparative milling tests at the Montalieu plant in France. While key features include the very similar particle size distri-bution, positioning parameter and the fineness acc. to Blaine, the results show that the setting times for the two cements are the same. Setting begins after about 125 minutes and stops – in both cases – after 175 minutes. In addition, the compressive strength and the strength development meas-ured after 2, 7 and 28 days are basically exactly the same.

This proves that cements produced in a ball mill or in a ver-tical roller mill can have the same characteristics and qual-ities when required to do so by local markets, particularly in terms of the water demand of the cement paste, the set-ting times, the compressive strengths and strength devel-opments of mortars.

Today, out of the 260 Loesche mills sold around 185 are used worldwide for grinding cement products. On a regional basis, the largest proportion of the total is in eastern Asia, with sig-nificant numbers of the machines in Europe and the Amer-icas as well. It is also significant that around 60 % of these mills are used for grinding more than one grade of product, with around 40 % being used for more than three products. Indeed, the flexibility of this mill system is such that today

52 µm37 µm

5 %

Resi

due

[mas

s %

]


Fineness: about 4100 Blaine

Figure 8: Particle sizes in cements mostly below 55 µm and 95 % below 45 µm

Blaine-value B 3 000 cm2/g

VRMRoller pressBall mill

Particle size [μm] (VDZ 2007)

1.0

0.8

0.6

0.4

0.2

0.00.1 1 10 58 100

Circ

ular

ity

Figure 9: Similar particle shapes of cement produced on vertical roller mill systems

Increase gypsum content

Add an amount of hemihydrate

Increase mill exit temperature

Add an amount of natural anhydrite

Decrease humidity of mill gas �ow

resulting in

Same setting behaviour, i.e. setting time and compressive strength

Figure 10: Normal optimisation process by works quality department

64-69_Fahrland.indd 68 02.04.13 16:07Unbenannt-1.indd 1 30.04.2013 10:52:05 Uhr1304268_VRM_unrivalled_flexibility_02.indd 6 30.07.13 14:53

there are several mills used to grind five or six cement prod-ucts. Cement producers worldwide appreciate the higher energy efficiency of the vertical roller mill. In addition, the inherent flexibility of this mill system means that product specifications can be changed quickly and easily in contrast to a ball mill.

[1] Pohl, M.; Obry, C.; Zysk, K.-H.: Operating experience with a vertical roller mill for grinding blastfurnace slag and composite cements. CEMENT INTERNATIONAL 10 (2012) No. 2, p. 56–69.

[2] VDZ: Activity Report 2005–2007. Ed.: Verein Deutscher Zementwerke e.V., Forschungsinstitut der Zementindus-trie, Düsseldorf, 2009, p. 38.

Designation UnitVertical

roller millBall mill

Separator LSKS O-SEPA

Fineness acc. to Blaine cm2/g 4 258 4 095

Standard consistency % 28 28.5

Setting time, begin min 130 130

Setting time, end min 175 175

Compressive strength (w/c = 0.5)

2 d7 d

28 d

MPa 29.838.957.1

29.938.654.1

Specific energy consump-tion (measured at shaft)

kWh/t 28.6 39.7

Table 2: Same cements with the same characteristics

LITERATURE / LITERATUR


PROC

ESSI

NG










the y-axis [2].


Dam ring



Grinding pressure

Mill air �ow

Height of dam ring

Classi�er speed

lower

higher

lower

higher

higher

lower

higher

lower

Slop

e n

<==

Shal

low

St

eep

==>




Resi

due

[mas

s %

]



64-69_Fahrland.indd 67 02.04.13 16:06


PROC

ESSI

NG










the y-axis [2].


Dam ring



Grinding pressure

Mill air �ow

Height of dam ring

Classi�er speed

lower

higher

lower

higher

higher

lower

higher

lower

Slop

e n

<==

Shal

low

St

eep

==>




Resi

due

[mas

s %

]




Loesche – worldwide presenceLoesche is an export-oriented company run by the owner, which was established in 1906 in Berlin. Today the company is inter-nationally active with subsidiaries, representatives and agencies worldwide.Our engineers are constantly developing new ideas and individual concepts for grinding technologies and preparation processes for the benefit of our customers. Their competence is mainly due to our worldwide information management. This ensures that current knowledge and developments can also be used immediately for our own projects.The services of our subsidiaries and agencies are of key importance for analysis, processing and solving specific project problems for our customers.

Loesche GmbHHansaallee 24340549 DüsseldorfTel. +49 - 211 - 53 53 - 0Fax +49 - 211 - 53 53 - 500Email: [email protected]

Please visit our homepage at www.loesche.com for up-to-date information on our overseas companies.

1304268_VRM_unrivalled_flexibility_02.indd 8 30.07.13 14:53

Date post:	29-May-2018
Category:	Documents
Upload:	ngokien
View:	212 times
Download:	0 times

Dipl.-Ing. Th. Fahrland, Dr.-Ing. K.-H. Zysk, Loesche GmbH, …€¦ · · 2016-09-26Loesche...

Documents