Вы находитесь на странице: 1из 11


The Ball Mill

C E M E N T T E C H N O L O G Y N O T E S 2 0 0 4 43
4 . 1 I NT R ODUC T I ON
4 . 2 T E R MI NOL OGY
4 . 3 P R I NC I P A L P A R A ME T E R S
4 . 3 . 1 I NT R ODU C T I ON
4 . 3 . 2 DI A ME T E R
4 . 3 . 3 L E NGT H
4 . 3 . 4 V OL U ME
4 . 3 . 5 V OL U ME L OA DI NG
4 . 3 . 6 ME DI A C HA R GE
4 . 3 . 7 MI L L S P E E D
4 . 3 . 8 MI L L P OWE R
4 . 3 . 9 L I NI NG
4 . 3 . 1 0 DI A P HR A GM
4 . 3 . 1 1 ME DI A S I Z E S
4 . 3 . 1 2 ME DI A WE A R
4 . 3 . 1 3 P OWDE R L OA DI NG A ND R E S I DE NC E T I ME
4 . 3 . 1 4 MI L L A I R F L OW
4 . 3 . 1 5 MI L L T E MP E R AT U R E , C OOL I NG A ND WAT E R I NJ E C T I ON
contents chapter 5
C E M E N T T E C H N O L O G Y N O T E S 2 0 0 5 44
The ball mill, or tube mill, is essentially a lined steel tube, which
is partially filled with grinding media (usually balls). The tube
rotates, thereby causing a cascading action of the media, which
the feed material passes through. Comminution takes place within
the cascading, cataracting and attrition actions of the media.
The ball mill is used in conjunction with other equipment to
form a range of different mill circuits. These are discussed in
Section 5.
Figure 45. Ball Mill.
Inlet Trunnion
Step Lining Plates
First Chamber
Intermediate Diaphragm
Second Chamber
Classifying Lining Plates
Outlet Trunnion
Mill End Wall
Coarse Grinding Media
Fine Grinding Media
Outlet Diaphragm
contents chapter 4 chapter 5
C E M E N T T E C H N O L O G Y N O T E S 2 0 0 5 45
For simplicity, the ball mill can best be described by reference to
the main parts involved. (See Figure 45).
Mill Shell: The cylindrical tube
(usually rolled from steel plate)
Mill Ends: The circular castings bolted to the mill shell
Trunnions: The smaller cylinders at each end of the mill
which provide the inlet, outlet and main
bearing support
Liners: The metal plates, usually bolted, to the mill
shell and ends for wear protection
Mill Hood: The chamber between the mill exhaust
ducting and the mill end which is around the
outlet trunnion
Chamber: Compartment in the mill created from the
division of the mill by diaphragms
Diaphragm: The slotted plate which divides the mill tube
into chambers. Between chambers is the
intermediate whilst at the mill end is the
outlet diaphragm
Grinding Media: The steel balls (or cylpebs) which partially
fill the chamber and tumble with mill
rotation to achieve crushing and grinding
Mill Speed: Rotational speed (rpm) of the mill
Critical Speed: Mill speed at which particles just remain at
(rpm) mill liners as a result of centrifugal force
Volume Loading: The percentage of the mill volume (inside
(Vol%) liners) occupied by media
Powder Loading: A measure of the volume or weight of
material (not media) in the mill (See later)
Mill diameter: The diameter of the mill tube, either inside
(D) (m) the shell, or more usefully, inside the liner
Mill length: The length of the mill tube, either in total
(L) (m) (between mill ends) or a summation of the
effective chamber length
Chamber diameter: The diameter inside the lining in each
(D) (m) chamber
Chamber length: The effective length of each chamber (i.e.
(L) (m) between diaphragms, or mill end and
diaphragm, in chamber 1)
Mill Power: The installed power relates to the motor size
(kW) The gross drawn power relates to the actual
consumed power
The net power refers to either the gross
power less gearbox and motor losses, or an
estimate of power from the mill parameters
Specific Power The mill gross power divided by the product
Consumption: tonnes/hour
Slot Size: The average opening of the slots in the
(mm) diaphragms
Ball Size (mm): Ball diameter
Media density: The bulk density of the media in the mill
Media charge (t): Usually the weight of media in the mill
Mill airflow: The volume of air passing through the mill
/hr) (mill NOT filter)
Height above The distance from the liner to the media
charge: level
(HAC) (m)
Height above The distance from the liner to the material
powder: level
(HAP) (m)
Liner Step: Usually the depth of the protruding part of
(mm) the liner plate (in chamber 1)
Diaphragm The % diaphragm area that is open (i.e. slot
active area (%): area)
Diaphragm The % of the slot area blocked (usually by
blockages: nibs)
Media Grading: The size distribution of the balls
Mean Media The average ball size (usually on a weight
(mm): Size basis)
Clinker nibs: Usually the oversize clinker particles that
escape chamber 1 and build-up in chamber 2
(or the last chamber) or the diaphragm slots
Axial Samples: Material samples taken from inside the mill
(when stopped!)
Drum feeder: The large diameter drum, fitted externally to
the mill feed trunnion, with scoops and
lifters to pass material into the mill
Feed chute: A simple lined chute fitted into the trunnion
scrolls on the trunnion that assists material
into the mill
Coating: The adhesion of fine particles to the mill
internals, particularly to the balls and liners
contents chapter 4 chapter 5
C E M E N T T E C H N O L O G Y N O T E S 2 0 0 5 46
The preceding section attempted to list the most important
terminology involved in a ball mill. Below is additional
information concerning the principal parameters that are
involved in the evaluation of mill performance. Further details
are discussed in section 11 on mill testing.
In most cases the most practical diameter will be the effective
diameter inside the lining. However in many manuals providing
basic mill data, the diameter will be inside the shell, thus there is
a single diameter for a given mill. For the diameter inside the
liner there will be a diameter for each chamber depending on
the thickness of the lining (often 50-80mm).
4.3.3 LENGTH
Like the diameter, the length is most useful when it is the
effective length of each chamber.
4.3.4 VOLUME
The effective volume of each chamber can be calculated using
the effective chamber diameter and length.
The chamber volume loading can be determined from knowledge
of the chamber volume, tonnes of media and media density.
e.g. assuming a chamber volume of 77.14m
, a media weight of
101 tonnes and a media density of 4.3 tonnes/m
, the volume
loading would be :-
Bulk media volume = 101 = 23.49m
Volume Loading = 23.49 = 30.4%
However, more typically, the volume loading is measured directly
from an internal examination of the mill. (See Section 11).
The volume loading is calculated from the measured diameter
and the height above the charge. There are a number of
empirical formulae, which can be used very easily, although the
volume loading can be best determined from first principles.
Figure 46. Example of Mill Data.
Chamber: 1 Mill rpm: 14.6
Diameter: 4.45 HAC: 2.92
Length: 4.96 HAP: 3.17
Volume: 77.14 Media t/m
: 4.30
%Nc: 72.8
Vol Loading,% 30.4
Tonnes: 101.0
Powder, U 0.78
Net kW: 1435
Chamber %L: 38.4
Ball Piece No. Wt Tonnes
Size Wt. (%)
90 3015 5 25.4 25.6
80 2118 10 35.7 36.0
70 1419 10 23.9 24.1
60 893 10 15.0 15.2
0 0 0 0.0 0.0
0 0 0 0.0 0.0
0 0 0 0.0 0.0
Mean Size 77.1 mm 100 101.0
Chamber: 2 Mill rpm: 14.6
Diameter: 4.50 HAC: 3.05
Length: 7.95 HAP: 2.98
Volume: 126.44 Media t/m
: 4.50
%Nc: 73.2
Vol Loading,% 27.9
Tonnes: 158.5
Powder, U 1.16
Net kW: 2354
Ball Piece Average Pos 1 Pos 2 Pos 3
Size Wt. Wt Tonnes No. Wt No. Wt No. Wt
(%) (%) (%) (%)
60 893 2.2 3.5 1 6.7 0 0.0 0 0.0
50 517 4.2 6.6 2 7.8 1 4.7 0 0.0
40 265 9.3 14.7 8 15.9 5 12.0 0 0.0
30 112 6.4 10.1 12 10.1 8 8.1 1 1.0
25 65 17.9 28.4 50 24.3 40 23.5 10 6.0
20 33 30.8 48.8 90 22.4 80 24.0 150 46.0
17 20 29.2 46.3 85 13.0 150 27.7 250 47.0
100 158.5 100 100 100
Mean 24.7 30.0 25.0 19.0
Summary Feed T/hr 110.0
Total T/hr 330.0 T/hr/m
Chamber Volume Media Media Mill Mill kWh/t Powder Powder Residence
Loading Tonnes Mean Power Power U Tonnes Time
Size Net Gross Mins
1 30.4 101.0 77.1 1435 1538 14.0 0.78 11.7 2.1
2 27.9 158.5 24.7 2354 2522 22.9 1.16 24.5 4.5
Total 259.4 3789 4060 36.9 36.2 6.6
Average 28.8 45.1 1.01
Gross/Net Factor: 1.07
contents chapter 4 chapter 5
C E M E N T T E C H N O L O G Y N O T E S 2 0 0 5 47
According to FLS, the volume loading is determined from:-
Vol % = 50 - 125 x (h - D/
Where h is the height above the charge and D is the diameter.
For the data given in Figure 46, the volume loadings are
calculated to be 30.5% and 27.8%.
From trigonometry it is possible to use the height above the
charge to calculate the area of the segment less the area of the
triangle to determine the area of the media. (See TIS MS007).
From this calculation the volume loadings are correctly
determined to be 30.4% and 27.9%. (See Figure 46).
Volume loadings are typically in the range 25-35%, but can be
TIS MS007. Volume Loading.
Area of Segment = r
Area of Triangle =
* r cos * 2 r sin
= r
cos sin
Area below level = a r
- r
cos sin
= r
( - cos sin)
Area of circle = ! r
Volume occupied = r
( - cos sin) / ! r
= ( - cos sin) / !
Based on the measured volume loading the media tonnage is
determined using the media density. This is usually assumed to
be in the range 4.3 - 4.7 tonnes/m
The critical speed is where particles just centrifuge
(See TIS MS008).
The equations can be shortened to:-
Nc = 42.3
Where D is the effective diameter.
TIS MS008. Mill Speed.
Critical Speed when mv
/r = mg
/r = g or v
= rg
= D/2g or v = (D/2 * g)
Also = (2 ! N)/60
N = (60 ) / (2 !)
Also v = r
= v/r
= (60 v)/(2 ! r)
= (60 v) / ( ! D)
= {60 (D/2 * g)
} / ( ! D)
= 42.3 / D
m = mass
g = gravity
D = mill diameter
v = tangential velocity
= angular velocity
N = mill speed (rpm)
= mill critical speed (rpm)
Thus for the data shown in Figure 46 the critical speed is 20.00
rpm (using average of chamber 1 and 2 diameters) and thus the
actual speed of 14.60 rpm represents 73.0% of critical speed.
Typical mill speeds are in the range 65-80% although 75%
represents a typical average. Most mills have a fixed speed
although mills with variable speed drive do exist. This, in
theory, at least provides the option of obtaining the optimum
profile of cascading media as the liner wears or as the media
volume loading changes.
Since it is very easy to determine the mill speed (visually time 10
revolutions), it is recommended that this is done when
appropriate (relying on recorded data can sometimes lead to
The gross drawn power of a mill can normally be measured
from the kWh meter. Often, average values are based on the
kWh totaliser readings together with run hours. However in
most cases it is better to take a spot reading or readings directly
from the kWh meter. This involves timing the dial on the kWh
meter to determine the number of revolutions in 1 hour, and
then using the kWh/rev factor on the meter to calculate the kW.
(i.e. time 10 revolutions of the dial).
For example:
Meter factor = 3.210 kWh/rev
Time for 10 revs = 28.46 seconds
Therefore revs/hour = 1264.9
Therefore mill power = 4060 kW (revs x kWh)
hr revs
The net power of the mill can be estimated from the charge
weight, arm of torque and the angular velocity. Most plant
suppliers have their own formulae.
r cos = h - r
= cos
{(h - r)/r}

contents chapter 4 chapter 5

C E M E N T T E C H N O L O G Y N O T E S 2 0 0 5 48
One empirical formula which has been used is:-(See also TIS
kW = 0.2846 x D x (1.073 - VOL%) x W x N
where D = effective chamber diameter (m)
Vol% = chamber volume loading (%)
W = chamber media weight (tonnes)
N = mill speed (rpm)
The formula is used to calculate the net power of each chamber.
For the data in Figure 46, these are 1435 kW and 2354 kW.
The gross power is seen to be some 7% higher than this, i.e. the
gross kW/net kW ratio is 1.07, (4060/3789). This factor is
usually in the range 1.05 - 1.10. Values outside this range can
be indicative of measurement errors or operating and design
4.3.9 LINING
The liner plates are either bolted or sometimes boltless, relying
on fixing by key from one plate to another. Bolted plates can
lead to individual fixing problems and leakage at bolt holes.
Boltless fixing avoids this problem, although replacement due to
damage or wear can become a major job.
The liner plate profiles vary according to the chamber in which
they are used since the aim is to promote a different movement
of media in each chamber (given that mill speed is fixed for each
In chamber 1 the aim is to expand the charge to promote a
"cataracting" action for crushing of larger clinker pieces. Thus
the liner plates are stepped (See Figure 47), the step design
typically being 75-90mm. For efficient action, a minimum step
of 40mm is normally recommended. On average, the lining
thickness is typically around 75mm. Hence the difference
between inside shell diameter and effective diameter is usually
0.15m. (See Figure 46).
In subsequent chambers a "cascading" type of action is required
and smoother profile linings are used to promote a more
compact grinding. (See Figure 47).
In the second chamber of a 3 chamber mill a ripple lining is
often used with an effect someway between chamber 1 and 3.
Figure 47. Lining Plate Profiles.
In chamber 2 of 2 chamber mills it is often practised to have a
classifying lining to promote the classification of ball sizes, so
that the larger sizes remain near to the intermediate diaphragm
(See Figure 48) and the small sizes are displaced to the outlet
Smoother profiles tend to have a thickness of around 50mm
thus reducing the inside shell diameter by around 0.1m.
Figure 48. Classifying Liner Plate Profiles.
The diaphragm plates divide the mill into chambers. They can
be of single plate type but more often are of double plate design
(See Figure 49). The plates are slotted with radial or more
usually circumferential slots, which retain the media within the
chambers and allow passage of material and air (See Figure 50).
Figure 49. Diaphragm Double Plate Designs.
F L Smidth
Slegten, Adjustable
contents chapter 4 chapter 5
C E M E N T T E C H N O L O G Y N O T E S 2 0 0 5 49
Figure 50. Diaphragm Slot Arrangements.
The active or open area of the diaphragm as a result of the slots
is typically around 5-10% of the total diaphragm area. Slot sizes
(widths) are ideally around 6mm for intermediate and 8mm for
the outlet. The outlet slots should always be marginally larger
than those of the intermediate diaphragm to avoid the build-up
of nibs (e.g. large clinker pieces) in the last chamber and the
outlet diaphragm.
Double plate diaphragms have lifter plates between the plates
(See Figure 49) to promote material transport through the
diaphragm. Naturally the flow rate of material through the
diaphragm is only a function of the total feed rate (assuming a
steady-state) and so the lifter action effectively controls the
filling level in the preceding chamber.
In some cases the lifting action can be too powerful and thus 1st
chambers are often operating with a low level of material. As a
result of this the so-called variable (or adjustable) diaphragm
was developed (See Figure 49), to provide a degree of
adjustment of the material filling level.
Media is used in a range of sizes in each chamber to provide:-
- a range of desired impact energies
- an element of charge densification
- a degree of stability of media grading when allowing
for wear (i.e. equilibrium charge)
- the ability to deal with a range of feed material
Chamber 1
Here media is nearly always in the range 90 to 60mm. Sizes
above 90mm are rarely required and in any case risk breakage
of liners. Sizes below 60mm are often too small, and therefore
impact energy is too small, for sufficient preparation of clinker
in chamber 1. Mean size is typically in the range 75-78mm.
Chamber 2
Here media is often in the range 60-16mm depending on many
factors including overall mill design and manufacturers
When a wide range of sizes are deemed to be necessary then a
classifying liner is required to classify the media so that the
coarser sizes remain at the inlet end of the mill. It is often noted
that, particularly where there is no classifying liner, media
becomes naturally reverse classified, i.e. coarse sizes concentrate
at the mill outlet.
If nibs are a problem, these can often be classified to the mill
outlet with the small media thereby not being subjected to the
larger balls necessary for their breakage.
Mean media sizes are typically in the range 20-40mm
The influence of media sizes on mill performance is discussed
further in section 6.
The wear of media has significantly reduced in the last 10-20
years as a result of the developments in metallurgy. Typically,
wear rates are around 10-100g/tonne of cement ground.
Although wear is often quoted as g/t of cement it is more likely
that the majority of wear is a result of run hours or kWh.
Hence wear in terms of g/kWh is also used. An example of the
effect of media wear on annual media consumption is shown in
Figure 51.
Figure 51. Media Wear.
Circumferential Radial
Mill Power
Mill Production
Annual production
Run Hours
Media Wear
Chamber 1
Chamber 2
4060 kW
110 tonnes/hour
800,000 tonnes
50 g/tonne
101.0 tonnes
77.1 mm Mean size
158.5 tonnes
24.7 mm Mean Size
Annual Media Consumption = 40 tonnes
Media Wear = 1.35 g/kWh
Reduction in Mean Media Size
(assuming no top-up):
Chamber 1 Media Wear =
77.1 mm =
New weight =
New mean size =
Chamber 2 Media Wear =
24.7 mm =
New weight =
New mean size =
Reduction in Volume Loading:
Chamber 1 From
Chamber 2 From
(50 * 800000)
15.6 tonnes 15.4%
1899 g
1606 g
73.0 mm
24.4 tonnes 15.4%
62 g
52 g
23.3 mm
30.4 %
25.7 %
27.9 %
23.6 %
contents chapter 4 chapter 5
C E M E N T T E C H N O L O G Y N O T E S 2 0 0 5 50
For an annual production of 800,000 tonnes the media
consumption for 50g/t is 40 tonnes. In the example this equates
to 1.35g/kWh.
Mean media sizes (assuming no top-up) are seen to reduce to
73.0mm (from 77.1) and to 23.3mm (from 24.7) respectively.
Cement grinding in a ball mill essentially involves the transport
of material through a rotating cylinder. In the simplest case
where there is no mixing as particles pass through the mill the
type of flow is referred to as "plug flow" and all the particles
have the same residence time in the mill.
Conversely, in the case where the contents of the mill are well
mixed and uniform in composition, such that the exit from the
mill has the same composition as the contents of the mill, the
flow is referred to as "back-mix" of "fully-mixed" flow.
In both cases, for steady-state conditions, the residence time is
simply defined as the ratio of the mass of material in the mill to
the feed rate, i.e.
Residence Time = W/F
Where W = hold-up of material in the mill (tonnes)
F = feed rate to the mill (tonnes/hour)
In practice however, the flow of cement through mills is non-
ideal (plug flow and fully-mixed flow being "ideal") because of
partial back mixing as a result of the mill internals (e.g. media,
diaphragms). Hence in reality there is a distribution of residence
times for particles in the mill with an overall mean residence
time, i.e. some particles remain in the mill for only a short time,
whilst others remain for a very long period.
The residence time distribution (RTD) and mean residence time
(MRT) can be determined in practice using tracer techniques,
such as that proposed by Frank Mardulier and details of the test
method can be found in his published articles.
The Grace method of Frank Mardulier involved the use of
sodium fluorescein as a pulse tracer added to the feed of the
mill. The concentration of tracer in the mill product is then
monitored starting at the time of addition and at regular
intervals for up to 60 minutes (longer for open-circuit mills,
shorter for closed-circuit mills).
A typical tracer curve is shown in Figure 52 for an open and
closed circuit mill.
The trace from a closed-circuit mill is complicated by the
presence of tracer in the rejects (hence the cyclic nature of the
curve). It is possible to mathematically correct this curve to
produce the equivalent of the open-circuit (Austin has a
publication on this) but it is quite complex. The mean residence
time can then be determined. It should be noted that this rarely
coincides with the peak, as the distribution of residence times is
usually skewed (See Figure 52). In most practical tests the peak
of the curve was used to estimate the mean residence time,
although in general this will be marginally shorter than the
actual one.
Figure 52. Residence Time Distribution.
contents chapter 4 chapter 5
C E M E N T T E C H N O L O G Y N O T E S 2 0 0 5 51
From knowledge of the feed rate and the MRT, the hold-up can
be determined.
For example, from the data in Figures 46 and 51, we had a feed
rate of 110 tonnes/hour and say a circulating load (total feed
divided by fresh feed) of 300%. For a MRT of 6.8 minutes
determined by fluorescein tracer, we would calculate a hold-up
of 37.4 tonnes (See Figure 53).
Figure 53. Residence Time/Hold-Up/Void Filling Fluorescein
As a result of this, the steel/clinker ratio is 6.9. Typical ratios are
between 6 and 10, with the optimum around 8 (See section 6).
The determination of hold-up in this way, whilst accurate, is
somewhat involved and can be time consuming. Furthermore
the value of tonnes hold-up is not a very practical value unless
qualified by knowledge of other data such as feed rate,
circulating load and media charge.
A more useful value is the volume of material in the mill
compared to the volume of media. The steel/clinker ratio goes
someway to provide this, although can only be used to relate to
volume if densities are known and remain constant. This is
more or less true for the media but not so for the material (the
bulk density of clinker/cement can be anywhere between 1000
and 1700 kg/m
One means of relating volumes is to refer the volume of
material to the volume of voidage in the media charge, i.e. void
filling. This is a key parameter in the "Grace Factor" approach.
The media voidage will depend on the packing of the balls. For
dense packing the voidage can be as low as 40%, however when
allowing for end effects (i.e. the constraints of a chamber - sides
and ends) and a degree of charge expansion due to the material,
the value is typically in the range of 42-45%.
In our example (See Figure 53) the volume of voids for the
media in the mill is 25.86m
. For a bulk density of say 1400
for the material in the mill we have a material volume of
. Hence the percentage void filling is 103%. We shall see
later that this is on the high side for optimum grinding.
Another approach for assessing the powder loading is to directly
estimate during a mill internal inspection following a crash stop
(i.e. a mill stop with feed on). This involves an estimation of the
mean level of material relative to the media level and then the
combination of this with the height above charge measurement
to produce the effective height above powder (HAP) value.
Where the powder level is below the media level, the void filling
is simply the volume to the powder level divided by the volume
to the media level. In the example (Figures 46, 51, 53), in
chamber 1 we have HAC = 2.92, HAP = 3.17 (Powder was
250mm below media) and equivalent volume loadings of 30.4%
and 23.8%. Hence the percentage void filling is 78%.
Where the powder level is above the media level the void filling is:-
the volume to the powder level less the volume of
media steel (e.g. 0.57 * volume loading) divided by the
volume of voidage (i.e. 0.43 * volume loading). In the
example, in chamber 2 we have HAC = 3.05, HAP =
2.98 (powder was 70mm above media) and equivalent
volume loadings of 27.9% and 29.7%. Hence the
percentage void filling is 116% (not simply 106%).
See also TIS MS010.
From the void fillings it is of course then possible to estimate
the hold-up in tonnes using an assumed material density of say
1400 kg/m
. This is seen to be 36.2 tonnes from Figure 53.
Figure 53 (cont). Residence Time/Hold-Up/Void Filling Mill
Inspection Data.
Data (Fluorescein Tracer):
Feed Rate =
Circulating Load =
Mean Residence Time (tracer) =
Media, Chamber 1 =
Media, Chamber 2 =
Volume, Chamber 1 =
Volume, Chamber 2 =
Volume Loading, Chamber 1 =
Volume Loading, Chamber 2 =
110 Tonnes/hour
300 %
6.8 Minutes
101.0 Tonnes
158.5 Tonnes
77.14 m
126.44 m
30.4 %
27.9 %
MRT = Hold-Up/Feed rate
or Hold-Up = MRT * Feed Rate
Circulating Load = Total Feed rate/Fresh
Feed rate
Total Feed rate = (300/100 * 110) =
Hold-Up = (6.8/60 * 330) =
Steel/Clinker Ratio =
Media Volume, Chamber 1 =
Media Volume, Chamber 2 =
Void Volume, Chamber 1 =
Void Volume, Chamber 2 =
Void Volume, Total =
Cement Volume (assuming 1.4t/m
) =
Void Filling =
330 Tonnes/hour
37.4 Tonnes/hour
23.48 m
35.21 m
10.70 m
15.16 m
25.86 m
26.7 m
103 %
Data (Mill Internal Inspection):
Height above powder, Chamber 1 = 3.17 m
Height above powder, Chamber 2 = 2.98 m
Volume to powder level, Chamber 1 = 23.8 %
Volume to powder level, Chamber 2 = 29.7 %
Void Filling, Chamber 1 = 78 %
Void Filling, Chamber 2 = 116 %
Void Filling, Mill = 101 %
Hold-Up, Chamber 1 = 11.7 tonnes
Hold-Up, Chamber 2 = 24.5 tonnes
Hold-Up, Total = 36.2 tonnes
MRT = 6.6 Minutes
contents chapter 4 chapter 5
C E M E N T T E C H N O L O G Y N O T E S 2 0 0 5 52
Rarely will both methods (i.e. fluorescein tracing or mill
inspection) give the same results (or indeed be carried out at the
same time). The advantages and disadvantages are:-
Fluorescein Tracing:
Advantages: Theoretically more accurate for
tonnes hold-up
Provides additional RTD data
No need to stop the mill
Impressive technical service
Disadvantages: Involved procedure
Short-cuts will reduce accuracy (i.e.
no allowance for recycle and using
peak concentration only)
Mill operating tests only
Void filling relies on good estimate of
material bulk density
Mill Inspection:
Advantages: Direct measurement for void filling
Can be carried out in conjunction
with a wider mill inspection or audit
Provides individual chamber
Disadvantages: Hold-up requires good estimate of
material bulk density
Needs mill stop (can be advantageous if
in conjunction with a wider inspection)
Can be difficult to estimate (See
Section 11)
Often, in discussion with cement plants, it will be possible to
acquire an understanding of the typical or characteristic void
filling in a mill since they will undertake relatively frequent stops
on the mill (although the majority will be with the feed run out).
Knowledge like, "well below the media level" or "about the
same level" or "50mm" deep of material above the media" will
often be quoted. These could equate to "low void filling"(say
80%), "100% void filling" and "110% void filling" respectively.
Further information is provided in sections 6 and 11.
The mill airflow refers to the volume of air passing the mill.
However much available data will refer to the air passing the
filter and exhaust. In-leaks, for example at the mill end, are
important and thus sometimes only 50% (or even less) of the
exhaust air will have passed through the mill.
In the past, where the average mill size was somewhat smaller
than today, mill airflows were often targeted at around 2-3
volume changes per minute (where the volume is the free
volume above the charge). More often the air requirement,
which is principally for heat and moisture removal, will be
designed in terms of kg/kg of air to cement. A typical design
value of 0.4 kg/kg will often be used, although the majority of
mills will operate below this at 0.1 to 0.3 kg/kg.
The airflow can be measured by:
- pitot (at the mill filter exhaust). Then make
allowances for the in-leaks
- anemometer (e.g. vane or hot wire)
- gas tracing
One practical method is to determine the volume of air at the
filter exhaust (usually by pitot) and then carry out a
temperature balance. Details are provided in TIS MS011. This
- measure filter exhaust gas temperature and airflow
(correct to STP conditions)
- measure the temperature of the mill exit airflow
(mill exit cement temperature can be used)
- measure the ambient air temperature
- carry out a temperature balance to determine the
volume of air passing the mill
In the example in TIS MS011 for a filter airflow of 55,000
Nm3/hr at 74C, a mill exit temperature of 118C and an
ambient temperature of 15C, the mill airflow is estimated to be
only 31,500 Nm
The majority of the kWh/tonne appear as heat, which results in
a rise in temperature of the material leaving the mill.
High milling temperatures result in a deterioration of the
grinding efficiency as a result of an increased tendency for
agglomeration and coating. Also a greater proportion of the
added gypsum will be dehydrated to hemihydrate and/or soluble
anhydrite (See Section 1).
For these reasons milling temperature is controlled by one or
more of the following:-
- airflow through the mill
- external water sprays
- internal water injection
- cooling air to the separator
- direct cooling of the separator rejects
In practice, all mills have some cooling from airflow whilst
some have additional water injection for cooling. External water
sprays are only usually found on smaller older mills. Modern
separators (See Section 5) often have a "straight through" air
circuit which provides significant cooling to the mill returns.
Separators have also been cooled by water jackets on occasion.
For correct mill cooling it is necessary to determine a mill heat
balance which assesses the sources of heat and the removal of
heat. A summary of these is shown in Figure 54.
Water injection either to the first chamber or both chambers
generally offers the most controllable form of mill cooling. As
an overall guide water injection should be limited to around 3%
by mass of fresh feed to the mill - 1% in chamber 1 and 2% in
chamber 2.
contents chapter 4 chapter 5
C E M E N T T E C H N O L O G Y N O T E S 2 0 0 5 53
Figure 54. Mill Heat Balance.
Potential problems of internal water can occur, such as:-
- In chamber 1, localised cooling resulting in
hydration and blockages and coating of the
intermediate diaphragm
- In chamber 2, inadequate dispersion or targeting of
the water such that cement pre-hydration can occur
resulting in a reduction in strength levels.
The water should be dispersed into the media charge.
Dispersion is usually achieved by atomisation with air. However
a very fine atomised water spray can be directly taken by the
mill venting airflow, and therefore be less effective.
Second chamber injection at the intermediate diaphragm is the
most beneficial point of injection, although requires well
maintained equipment.
When using internal water any other sources of water should
also be taken into account, e.g. raw materials, air.
One method of assessing the maximum water input is to assess
the dew-point temperature of the mill exhaust. Ellerbrock (ZKG
1/88 pp1-12) has suggested a maximum dew point temperature
of 70C to avoid any deterioration of cement characteristics.
Mills typically operate with dew-point temperatures of 40C.
Details are provided in TIS MS012.
Mill Power
Other Feed
Water, injection
Water, air
Chemical Reaction
Water (evaporation)
Gypsum Dehydration
(1) Principal heat input
(2) Typical Water Injection 0 - 3%
Chamber 1: 0 - 1%
Chamber 2: 0 - 2%
Dew Point Temperature 70C maximum
40C typical
contents chapter 4 chapter 5

Похожие интересы