tromp curve dynamic seperator.pdf

Modernisation of Mill Separators Presented by Dr Joe Khor

Cemtech Asia 2014 - Kuala Lumpur JK/17.06.2014

1. Introduction

2. Characterisation of separator performance

3. Replacement of 1st generation separators in a German

cement plant (Case Study 1)

4. Modification of a 3rd generation separator in a Malaysian

cement plant (Case Study 2)

5. Investment consideration

6. Conclusion / Discussion

Contents


1. Introduction

Separators in grinding

o Cement manufacture is energy intensive: consuming

typically > 3,000 MJ of fuel/t-clinker and > 90 kWh of

electricity/t-cement;

o Approximately two-third of the electricity consumed is used

for the grinding of raw materials, fuels and cement;

o Grinding is inherently an inefficient process -especially

when mills are operating in open circuit- but closed circuit

mills may also not be as efficient as they should be when

equipped with separators of the older design.


1. Introduction

Grinding without separator: open circuit mill

o In open circuit, material

leaving the mill must comply to

the finished product fineness,

entailing over-grinding &

wasted energy, at times also

overheating & coating of the

mill, which can adversely

affect the mill performance;

o The PSD of the product is

wider & for cement, the higher

coarse fraction can results in

lower strength. Coarse Fine

% r

eta

ined

PSD - Open circuit


1. Introduction

Grinding with separator closed circuit mill

o In closed circuit, the finished

product is separated externally

& the material leaving the mill

can be ground coarser: the mill

residence time is lower, over-

grinding reduced, throughput

higher & specific energy

consumption lower;

o The PSD of the product is

narrower & for cement, the

strength generally higher.

Conventional

Separator

High

Efficiency

Separator %

reta

ined

PSD - Closed circuit

Coarse Fine


Principle of dynamic air separation

o Air separator relies on the balance of

opposing drag (FD) & centrifugal (FC)

forces imparted by the air stream on

the particles as they spiral down the

casing by gravity (FG), to cause the

particles to be either moved into the

rotating cage & discharged as fines,

or continued to spiral down &

discharged as rejects for return to

the mill

FC

FRf

FD

(FG)

1. Introduction

Separator Feed

Separator

Rejects

FRc

Separator

Fines

Separation is imperfect due to uneven air/material distribution, turbulence, particle shape, obstruction of the descending particles, agglomeration, etc


Characterisation of separator performance

o Efficiency -or selectivity- of a separator is a function of the

particle size & is generally characterised by the percentage

of a given particle size in the feed discharged in the rejects;

o Plotting the size selectivity [T(x)] against the particle size [x]

produces a Tromp Curve, which is used to define the separator performance in terms of :

Bypass, Cut size, Sharpness of cut or separation, Imperfection, Agglomeration / mechanical state, of the separator.

2. Separator performance


Tromp Curve

o Tromp curve of perfect

separation -such as by sieving-

is characterized by a step

function from 0 to 100 %, or

perfect sharpness of separation

o Tromp curve of air separator is

imperfect and shows the particle

size selectivity [T(x)] with a

rightward slope indicating the

sharpness of separation & a

minimum value called BYPASS



Construction of a Tromp curve

The 1st step is to calculate the average recovery of the

separator fines & reject, vf & vg :

Where:

Vf = fraction of fines as a function of particle size [wt-%]

Vg = fraction of reject as a function of particle size [wt-%]

Qa(x) = fractional amount of feed passing size x [wt-%]

Qf(x) = fractional amount of fines passing size x [wt-%]

Qg(x) = fractional amount of reject passing size x [wt-%]

%100)()(

)()(

xQxQ

xQxQv

gf

ga

f %100)()(

)()(

xQxQ

xQxQv

gf

af

g



Particle size selectivity [T(x)] is calculated for each

particle size analysis from the laboratory as follows:

Where::

Qa(x) = amount of feed passing size x [wt-%];

Qg(x) = amount of rejects passing size x [wt-%];

Vg = separator rejects as a function of particle size;

)(

)()(

xQ

xQvxT

a

g

g

Calculating the size selectivity of separator



The circulating load [u] of a separator is calculated from

the particles size analysis as follows:

Where:

Vf = separator fines as a function of particle size;

Qa(x) = fractional amount of separator feed of size x [wt-%];

Qf(x) = fractional amount of separator fines of size x [wt-%];

Qg(x) = fractional amount of separator reject of size x [wt-%];

)()(

)()(1

xQxQ

xQxQ

vu

ga

gf

f

Calculating the circulating load



Bypass

o Bypass is the most important

value: the lower the bypass,

the higher the efficiency

o A high bypass is an indication

of over-grinding & energy

wasting

o Bypass of the latest generation

high efficiency separator

should be < 10%



Cut size

o Cut size is defined as that

particle size (x50) of which

half the particles are collected

as fines & half in the rejects

o If the sharpness of separation

is poor, >50% of the finest

particle sizes may end up in

the rejects so that no definite

cut size exists.

o Cut size of the latest

generation high efficiency

separator should be < 15 m



Imperfection

o The imperfection is given by:

o Imperfection of the latest


separators should be < 0.35

50

2575

2

)(

x

xxI



Sharpness of separation

o Sharpness of separation is

defined by Eder as follows:

o Sharpness of the latest


separator should be > 0.50

75

25

x

xx



Comparison of separator performance

Generation 1st 2nd 3rd Latest

CPB

Bypass [%] 30 - 60 10 - 35 8 - 20 2 - 10

Min. Cut size [m] > 20 15 - 20 < 15 < 15

Imperfection [-] > 0.50 0.35 - 0.50 < 0.4 < 0.35

Sharpness of cut - < 0.5 > 0.45 > 0.5

Max Blaine [cm/g] 3,800 4,500 5,500 6,000

The latest CPB G4 separator is develped based on extensive CFD modelling & pilot plant trial;

It has an extremely low by-pass of 2 - 10% depending on the product fineness, and can improve mill output / energy consumption by as

much as 20 25%.



Replacement of G1 separators in Phoenix Cement*,

Beckum, Germany (2010)

3. Case Study 1


Project outline

Existing plant data (CM 1)

Ball mill 3.8 m x 12 m L equipped with Horizontal Impact Crusher & 2 x 1st generation Heyd separators (installed 1969)

Cement type : CEM I 32.5 R & 42.5 R, CEM II/A-LL 32.5 R

Fineness : 3,800 - 4,200 cm2/g (according to Blaine)

Objectives / Constraints Improve cement early strength Produce high strength cement (> 5,000 cm2/g) Reduce production cost Adaption of restricted plant space Minimal production interruption from changing-over

3. Case Study 1


CPB study & findings

Plant inspection, including axial & circuit sampling to determine the existing mill & separator performance;

Zeisel test to verify the clinker grindability;

Confirmed low performance of Heyd separator due to: - low bypass / separation efficiency,

- low sharpness of separation,

- insufficient range of fineness setting.

3. Case Study 1


Replace the two1st generation Heyd separators with a single QDK 29-NZ high efficiency separator with the

following specifications:

Rated air flow : 143,350 Am3/h Operating temperature : 120 C Installed motor rating : 200 kW Max. feed rate : 258 tph Max. fine product : 115 tph

CPB recommendations

Adapt the ball charge to the new operating conditions & higher product fineness required for the new type

of cement to be produced.

3. Case Study 1


Project implementation - adaptation to restricted

plant layout / space & minimise stoppage time

2 existing Heyd

separators 1 new QDK

separator

Existing

Ball Mill

Erection & commissioning (6 weeks)

Swith-over

(3d)

2 new

Cyclones

3. Case Study 1


Production rate & fineness of CEM II/A-LL 32.5R

after the modernisation

3. Case Study 1


Production rate of CEM II/A-LL 32.5 R of equivalent

strengths from the Heyd and QDK29-N separator

Sample

Separator

Blaine

fineness

[cm/g]

+63 m

residue

[%]

2-day

early strength

[N/mm]

28-day

final strength

[N/mm]

Production

Rate

[%]

CEM II/A-LL 32,5R Heyd 4,100 6.5 8 24 48 100

CEM II/A-LL 32,5R QDK29-NZ 4,144 < 1 31 59 110

CEM II/A-LL 32,5R QDK29-NZ 3,844 1 27 54 112

CEM II/A-LL 32,5R QDK29-NZ 3,450 2 4.5 25 49 120

3. Case Study 1

Cement produced with the new QDK separator at lower fineness has lower residue and the equivalent early and late strength


Comparison of the 63-m residue of CEM II/A-LL 32.5R

from the old Heyd & new QDK29-N separator

3. Case Study 1

Old

New


Tromp curves of the new QDK29-N separator for

various cement types

o For CEM I 32.5 R (3,044

cm2/g) -corresponded to

the new production

capacity (light blue)- the

bypass is < 5%

o For CEM I 52.5 R (5,005

cm2/g) -unable to produce

previously (olive green)-

the bypass is < 10%

3. Case Study 1


Results Before After Difference

Cement type CEM II A-CC 32.5R

Fineness acc Blaine cm2/g 4,100 3,450

Residue 63 m % 6.5 8.0 2 4.5

Output t/h 68 84 + 23.5%

Power consumption kWh/t 37 30.4 - 17.8%

Cement quality 2 D N/mm2 24 25 + 1

Cement quality 28 D N/mm2 48 49 + 1

3. Case Study 1

Overall benefits of the separator replacement

Specific energy savings of approx 10% for CEM I 42.5 R

Production of CEM I 52.2 R for new market

* Refer ZKG 12, 2011 for more details


Modification of a competitor-modified CPB G3

separator in Holcim Malaysia, Pasir Gudang (2014)

4. Case Study 2


Project outline

Reasons / objectives of the modification The original CPB separator installed in 1997 was modified

by a competitor in 2008 to attempt to improve performance

Instead, performance worsened due to excessive vibration & cement residue also increased

CPB was requested to propose a solution at minimal investment & production interruption

4. Case Study 2

Existing plant data Ball mill 4.4 m x 12.75 m L, equipped with roller press Competitor-modified CPB separator QDK 36-N Cement types: HTS, HDC, HEW, HRF, HQC. Fineness: 3,800 - 4,200 cm2/g (according to Blaine) R 45 m: < 3.5%


CPB study & findings

Plant inspection -including axial & circuit sampling- to determine the existing mill & separator performance;

Confirmed high vibration due to insufficient stiffness of the modified rotating cage;

Competitor-modified air Inlet guide vanes unsatisfactory;

Gap of the modified labyrinth seal is excessive, allowing coarse particles entrainment in the fines;

Ball charge should be optimised, mill ventilation improved, separator motor & sealing air fan upgraded to further

improve the performance out of the project scope

4. Case Study 2


Differences in CPB & competitor design of rotating cage

The blades spacing were wider -170 mm cf 40 mm in CPB

design- bolted -not welded- & resulted in decreased cage

stiffness, even with the stiffener ring

4. Case Study 2

Stiffener ring

170 mm

Bolted


Overview of the CPB & competitor rotating cage design

4. Case Study 2

CPB-design Competitor-design


4. Case Study 2

Differences in CPB & competitor design of air guide vanes

CPB-design Competitor-design


4. Case Study 2

135 mm

25 mm

The gap between the inlet spiral

housing & modified rotating cage

was 135 mm, compared with 90

mm in CPB design

The labyrinth seal gap between the

fines exhaust duct & modified

rotating cage was 25 mm,

compared with 3 mm in the CPB

confined air seal

Differences in CPB & competitor design of inlet & air seal


There was no deflection plate at

start of the inlet spiral housing

4. Case Study 2

Position of deflection plate

in CPB QDK separator

Differences in CPB & competitor design of spiral air inlet


CPB recommended re-modifications

4. Case Study 2

Feed inlet chutes

Distributor plate

Air guide vanes

Rotating cage

Confined air seal

* Opportunity taken to also renew the cartridge bearings in the shut-down

1

2

3

4

5


Confined air seal & cartridge bearing of CPB separator

Arrangement of the

confined air seal in

CPB separator

Cartridge bearing

4. Case Study 2


Installation of the CPB air guide vanes, confined air

seal & rotating gage

4. Case Study 2


Installation of the CPB feed distributor plate

4. Case Study 2


Dismantling & assembly of the CPB cartridge bearing

after refurbishing

4. Case Study 2

Renewal of cartridge

bearings during the

shut-down


Final assembly of the CPB re-modified QDK separator

4. Case Study 2

Modification took less than 10 days from start to finish


Original

CPB

Latest CPB

Competitor

Tromp curve before & after the modification

o CPB original (1997):

Bypass > 15%

o Competitor-modified (2013):

Bypass > 34%

o CPB latest (2014):

Bypass < 3%

4. Case Study 2


Results Units Before After Difference

Cement type HRF

Blaine fineness cm2/g 3,892 3,674 -218

+45 m residue % 2.81 2.40 -0.41

Output t/h 113 122 9

Power consumption kWh/t 47.7 45.6 -2.1

Cement quality 7 D N/mm2 40.0 40.0 0.0

Cement quality 28 D N/mm2 51.1 51.9 0.8

4. Case Study 2

Summary of the improvement results*

For HQC, output increased by 10 t/h & power reduced by 4 kWh/t ROI estimated to be < 4 months, without considering higher sales

* Mill has not been fully optimised by CPB; data provided by Holcim Malaysia


ROI of a separator modification project in Germany*

CEM II/A-LL 32.5 R Basis: 7,200 h/yr running

Total investment 210,000

Electricity saving (38.8 34.1) kWh/t x 127 t/h x 7,200 h/yr = 4,297,680 kWh/yr

Electricity cost saving at

av. cost of 0.08/kWh 0.08 /kWh x 4,297,680 kWh/yr = 343,814 /yr

Return on investment 7.3 month (on electricity saving)

Production increase (127 111) t/h x 7,200 h = 115,200 t/yr

Extra sales at 10 /t profit 115,200 t/yr x 10 /t = 1,152,000 /yr

Total benefits 1,495,814 /yr (on electricity & sales)

Return on investment 1.7 months (electricity + extra sales)

5. Investment consideration

* Based on data of Heidelberger Cement Schelklingen Works


1.Separators play an important role in mill performance:

affecting the output, power consumption & cement quality;

2.Poor performance of separators are often overlooked, due

to either their locations out of sight, out of mind- or failure to audit & benchmark the performance;

3.The latest CPB separators are highly efficient -capable of

achieving bypass of < 5% for most cement types &

improve mill performance by as much as 25%, as well as

ability to produce higher Blaine / quality cement;

4.Higher grinding efficiency & cement quality means higher

margin & market shares -ROI is typically < 6 months- as

well as a lower carbon foot-print for the environment.

6. Conclusion


Thank you

for your

attention, any

questions or

comments?

6. Discussion

For more info, please visit www.christianpfeiffer.net

tromp curve dynamic seperator.pdf

Documents