Cement Process Engineering Vade-Mecum: 4. Milling
Cement Process Engineering Vade-Mecum: 4. Milling
Cement Process Engineering Vade-Mecum: 4. Milling
VADE-MECUM
4. MILLING
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
Table of Contents
Index - i
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
1. Mill
1.1 Ball Mill General
a) Mill design
General L/D ratio
• Raw mills: 1.5 < L/D < 3.2
• Finish / cement mills: 2.8 < L/D < 3.2
Length of first Compartments relative to total mill length
• Raw mills: First compartment length equals 35 – 45% of total mill effective length.
• Cement mill: First compartment length equals 30 – 35% of total mill effective length.
• When L/D>1.5, classifying liners might be used.
• The lower the L/D, the higher the circulating load needs to be (see below).
b) Percent loading of mill
2π
αr 2 − r sin α (h − r )
0.9
h/d
• % volume load = 360 0.8
πr 2
where: 0.7
- r is the radius
0.6
- h is the free height
h−r
- α = arccos 0.5 0 10 20 30 40 50%
r % volume load
- α in degrees π = 3.14
Rules of thumb
• % vol. Load = 111.87 – 123.98 (h/d), 25 – 50%: error max 0.6%.
• It is estimated that material increases the actual ball filling ratio by about 2%.
• Another method (quick but not as accurate) consists in counting the number of visible shell liner plates (n) and
to divide by the total number of shell liner plates per circumference (N): Angle α = n x 360 / N.
Values of angle h/d ratio in relation to the ball load (% filling degree)
Ball load (%) h/d n/N Ball load (%) h/d n/N
20 .7459 .667 31 .6516
21 .737 32 .6434 .590
22 .7281 .653 33 .6352
23 .7193 34 .627 .580
24 .7106 .639 35 .6189
25 .702 36 .6109 .569
26 .6926 .625 37 .6028
27 .685 38 .5948 .558
28 .6765 .611 39 .5868
29 .6682 40 .5789 .549
30 .6598 .601 41 .5709
42 .563 .539
4.1
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
4.2
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
B = 24 d 80
(Other formula exist that result in value differences
of ± 5%)
- B = ball dimension (mm)
- d 80 is the sieve with 80% passing
10
.1 1 10 100
Clinker Size d80
Rowland Formula
d 80 ρ . Wi
• B = 25.4 . d 80 is the sieve with 80% passing
K 100 . Ψ . 3.281 Du
4.3
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
Quick calculation:
• Ball diameter (mm) = 3 250 P (P = weight in g)
785 2
• Specific surface of balls of diameter = m / mt (d = diameter in mm)
d
4.4
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
Wear rates:
Ball diameter mm Wear GT ball Wear / Ball Wear ∂ diam
g/h.T g/100h mm/100h
100
90 12.9 38.6 0.4
80 14.4 30.4 0.4
70 16.7 23.6 0.4
60 19.3 17.2 0.4
50 23.1 11.9 0.4
40 28.9 7.6 0.4
30 38 4.22 0.4
25 46.5 2.98 0.4
20 58.5 1.92 0.4
17 68.2 1.38 0.4
Bulk density for ball load
(coarse to medium ball size distribution):
• In first compartment; 4.3 –4.5 metric tonnes per tonne of balls (3.0 to 2.0 inch balls would be fine).
• In second compartment; 4.5 – 4.65 metric tonnes per tonne of balls (2.0 to 0.75 inch balls would be considered
medium to fine).
• In single compartment: 4.5 – 4.55 metric tonnes per tonne of balls.
c) Others internals
Partitions
• Total slot area: 10 to 20 cm2/tph production:
Slot Size Central Part Discharge Part
FM 7 mm ± 1 mm 9 mm ± 1 mm Max: ½ min ball size
RM 10 mm ± 1 mm 12 mm ± 1 mm
Liners
• Liners must be changed when 60% of their effective lifting height has worn away:
- -8 to –10 % production
- reference points to measure lifting height are the lowest point on the liner to the highest release point
(contact points between grinding ball and liner plate)
• American Lorrain pattern: diameter (ft)*2=# bolt holes/row, 18.8” center to center.
• DIN pattern: diameter (m)*10== # bolt holes/row, 31.4 cm center to center.
• Classifying liners if L/D>1.5 and volume load<35%.
• Without classifying liners, keep a maximum of 3-4 ball sizes.
4.5
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
1
The recommended volume loading for minimum kWh/t is based on an acceptable compromise with production. For minimum kWh/t
the volume loading can be as low as 22%.
4.6
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
4.7
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
4.8
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
A limited amount of 5/8” balls should theoretically be added but the designer decided to use ¾” as the smallest ball size.
e) Fineness in Finish Mills:
In the first compartment before intermediate diaphragm
• 95% passing of 2.365 mm (2360 µ or 8 mesh) for the material leaving the first compartment 33% of energy.
• Particle size distribution recommended on other sieves:
- 86 – 92 % passing 1.0 mm (1000 µm or 18 mesh)
- 80 – 90 % passing 0.6 mm (595 µm or 30 mesh)
- 75 – 85 % passing 0.5 mm (500 µm or 35 mesh)
In the second compartment before discharge diaphragm
• 95% passing 0.5 mm (500 µm or 35 mesh)
• 70- 80 % passing 0.2 mm (212 µm or 70 mesh)
4.9
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
1.27
rpm π
• P = L *
* J*K j* * Fr2.379 * d * C
Vcr 4
Simplified formula
RPM 100 Fr
• P =T * * *Kj * * 9.5
V cr 75 1.366
Kj Function of Volume Load
Volume load Kj
40% 0.9
30% 1
20% 1.1
Rules of Thumb
• One metric ton of balls increases the mill power draw by 10kW.
• Usually, 8 to 12 kWh/t is absorbed in the first compartment for clinker grinding.
b) Grinding Laws
General Law: Charles
• dW = cx −n dx
- If W = Comminution work, x = Size of
particles (initial, final)
Value of n
Energy Law Value of n: Applies well over range of:
Rittinger 2 10 – 1000 µm
Kick 1
Bond 1.5
4.10
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
4.11
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
2. Separator
2.1 Circulating Load
a) Junction with Three Streams
A • A, R, F are the feed, reject and fine of the separator
- ai , ri , f i are the cumulated % passing at a defined sieve(i).
- da, dr, df are the % retained corresponding to the sieve interval dx.
- A= R+F
- A da = Rdr + Fdf
R F
R df − da F dr − da
With: da = ai + 1 − ai , = , =
A df − dr A dr − df
b) Drawing c) CL calculation
• Plot ( f i − a ) vs ( f i − ri ) • Using the least square line calculations,
If the mill circuit is steady, the graph has to be a with α = 0
straight line:
( f − a) = α + β( f − r ) d) Quick CL calculation
- α should be close to 0 • With one set of results of sieving:
R R f −a
- β is the most probable value of =
A F a−r
R β
- The circulating load is defined as: =
F 1− β
∑ ( f − a )( f − r )
n
i i i i
R i =0
=
∑( f − r )
A n
2
i i
i =0
4.12
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
b) Imperfection
• I = d75 − d 25
where:
2 * d 50 - d25 is the size of the particle which has 25 % chance of going to rejects
- d50 is the size of the particle which has 50 % chance of going to rejects
- d75 is the size of the particle which has 75 % chance of going to rejects
I= 0.4 -0.5 for a high efficiency separator
0.6 - 0.7 for a Sturtevant
0.45 - 0.6 for a Raymond separator
Imperfection vs Circulation Load
0.44
Imperfection
0.42
0.40
0.38
0.36
0 100 200 300 400
Circ. load (%)
c) Acuity Limit
• AL is the abscissa of the intersection of the two Tromp curve lines.
• It’s the size at which selection is initiated
Rule of thumb
• Cement mill = Acuity limit: 20-30 µm, Raw mill = Acuity limit: 30-60 µm
d) Bypass
Definition:
• By-pass is the ordinate of the intersection of the two Tromp curve lines.
• The bypass is the lowest percentage of feed that will go to the separator rejects.
Bypass vs. feed rate – Sturtevant Bypass vs. feed rate O’Sepa/Sturtevant
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
70
Sturtevant
0 0 0 0 0 0 0 000 0 0 0 0 0 0 00 0 0 0 0 0 0 0 00 0 0 0 0 0 0 000 0 0 0 0 0 0 0
60
80
Bypass (%)
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
50
Bypass (%)
0 0 0 0 0 0 0 000 0 0 0 0 0 0 00 0 0 0 0 0 0 0 00 0 0 0 0 0 0 000 0 0 0 0 0 0 0
60 40
0 0 0 0 0 0 0 000 0 0 0 0 0 0 00 0 0 0 0 0 0 0 00 0 0 0 0 0 0 000 0 0 0 0 0 0 0
40 20
O-Sepa
10
20 0
1.0 1.5 2.0 2.5 3.0 3.5
0 Qf/Qa (kg feed/m3 separator sweep)
50 100 150 200 250 300
Feedrate to Separator (t/h)
4.13
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
00 00 00
QF/Qa vs. bypass
40
0 0 0 0 0 0 0 0 0 00 0 0 0 0 0 0 0 0 00 0 0 0 0 0 0 0 0 000 0 0 0 0 0 0 0 0
• If Qf is the separator feed rate (kg/h) and Qa the
00 00 00 00 00 00 00 00 00 0 00 00 00 00 00 00 00 00 0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
30
Bypass (%)
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
separator ventilation (m3/h),
• Qf/Qa is an important ratio for the separator 20
efficiency.
Qf
10 00 00 00
0 0
− f1
Qa
• Bypass = 1 + e
0
- f1: coefficient for the separator 0 1 2 3 4
Qf/Qa (kg/m3)
4.14
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
4.15
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
4.16
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
b) Heat balance
Mill heat generation
• Kcal/h = kWh (power measured)*factor*860
- Factor = 0.75 for vertical mill and 0.9 for ball mill
Furnace wall losses
• About 5kcal/kg fuel
4.17
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
4. Grinding Aid
Type of Products
• Surface active agents tend to saturate the free valence and inhibit the pack-set. Typical surface-active agents
are:
- ligno-sulphonates
- polyoils
- amines
- organic acids
• Polar compounds (water, ammonia) are known to have some action on such bonds through their polar moment.
However, their practical use as surface agents is limited by their other impacts on the cement properties.
• Other agents, particularly coal dust, have been used in the past.
• Commercial products available as grinding aids are essentially (60-800 g/t ck):
- Triethanolamine
- Polypropyleneglycols and polyethylene
• HEA2, DDA& and other products cause a definite reduction of pack-set but do not prevent agglomeration or
lump-formation problems that are caused by:
- Alkalis ( K 2 SO4 )
- Moisture
The effect of grinding aid on milling process:
- Enhances the flowability and prevents agglomeration
- Prevents coating on liners and grinding media- Decreases the "Blaine: Passing 325" ratio
- Is lLowers effect on coarser product (below 320 m2/kg)
- Reduces contraction
Example Bath
• HEA2 on the feeding belt, Range: 0.1-0.2 kg/t of kk
• Specific gravity: 1.195 kg/l, % active agent: 70%
0.15
• Price: 1.31 $/l, Cost: * 1.31 = 0.234 $/t kk
0.7 * 1.195
• At kWh/t cte, 200g/t of glycol: +80 SSB
• Production increase and pack set decrease. Mixed with water (3/1) and injected in #1 comp
HEA2 (Grace) HEA2/rm (Grace)
0.025% weight per weight of kk 0.078% weight per weight of kk
2.02$/kg 1.48$/kg
6% production increase for a higher setting time (20 min)
reduce cracks
4.18
Rev. 2002
CEMENT PROCESS ENGINEERING SECTION 4 – MILLING
VADE-MECUM
5. Sieve
Sieve Screen Micron Iso alter Screen Micron Iso alter
#400 37 38 #14 1400
#325 44 45 #12 1700
#270 53 53 #10 2000 2000
#230 63 63 #8 2360
#200 74 75 #7 2800
#170 88 90 #6 3350
#140 105 106 #5 4000
#120 125 125 #4 4750
#100 149 150 #3.5 5600
#80 177 180 1/4" 6350 6300
#70 210 212 5/16" 8000 8000
#60 250 250 3/8" 9510 9500
#50 297 300 7/16" 11200 11200
#45 354 355 1/2" 12700 12500
#40 420 425 5/8" 16000 16000
#35 500 500 3/4" 19000 19000
#30 595 600 7/8" 22600 22400
#25 707 710 1" 25400 25000
#20 850 1"1/4 32000 31500
#18 1000 1000 1"1/2 38100 38100
#16 1180 2" 50800 50000
4.19
Rev. 2002