32 - Lab Automation

ICPS 2002

1 L a b A u t o m a t i o n & Q C

Laboratory Au A u t o m at atii o n & Q lit C t l 1

What is is lab lab aut autom oma ati tion on? ? 

Sampling



Sample mp le tra tr anspo ns port rt



Sample mp le pre pr eparat parat ion io n



Sample mp le Ana An alysi ly sis s



Control Actions



Storage to rage of data


2

What is is lab lab aut autom oma ati tion on? ? 

Sampling



Sample mp le tra tr anspo ns port rt



Sample mp le pre pr eparat parat ion io n



Sample mp le Ana An alysi ly sis s



Control Actions



Storage to rage of data


2

3 L Why Wh y a automate? au u t o m at e? a b yo u want want to have have a consistent consistent an and d  Do you A u m aybe yb e an im impr prove oved d QC QC, which whic h will wi ll also also t o give you an an improv imp rove ed product pro duct? ? m Dif iffere ferent nt ope op erators do d o thing thi ngs s diffe dif ferently rently.. Thi his s a result re sults s in shif s hifts ts in ana nalyt lytica icall level level from fr om one on e t i ope op erator to ano noth the er. Ha Have you se s een that? t hat? o n Do you know kn ow if your yo ur sampling sampling is is represe repre sent nta ati tive? ve? Con onti tino nous us sampl samplin ing g & the th eor ore eti tically cally the th e pr pre eferre ferred d nme nm etho thod. d. Con ontin tinou ous s Q samp sa mpli ling ng is i s easi easie er done d one by an an aut autom oma atic C samp sa mpler ler than than by ma m anual samp sampli ling. ng. 



3

Wh y a Why automate? au u t o m at e? 

Do you y ou want want to have have a consistent and an d m ay b e an im impr prove oved d QC QC, which whic h will wi ll also also give you an an improv imp rove ed product pro duct? ? 





Som ome e pr pre epara parati tion on//analyt nalytic ica al methods metho ds are no nott suited su ited for fo r some so me mate materi ria als. The The ri righ ghtt method metho d may ma y be b e di diff fficu icult lt in ma m anu nua al ope op eration ration,, but easy to aut utomate omate.. Woul ould d it be of va v alue to you yo u to inc i ncrea rease se th the e ana nalyt lytica icall sche s chedul dule e? So you wou would ld ga gain in more knowle know ledge dge of your yo ur proce pr ocess? ss? An A n autom automa atic syst sy ste em ma m ay be b e abl ble e to pr proc oce ess an inc i ncrea reased sed number nu mber of sa s amp mples les compa comp ared to an an ope op erator Is it i t easy easy fo forr you y ou to retrieve data fo forr evalu evalua ati tion on purpos pur pose es. Aut Automatic omatic systems helps helps you yo u store stor e and di distr strib ibut ute e valua valuabl ble e data data..


4

5 L Why automate? automate? a b  Do you want to tighten the quality A u control? Do you want to avoid producing t o low quality product, but to save fuel and m energy? a t You would need frequent and fast capture of i o analyses and production data. This is not easy n to do manually. & It may be obtained by moving the analytical instruments into the process thereby saving Q sampling and sample transport time, C 



preparation time and analysis time

5

6 L Why automate? automate? a b  Do you want to save construction and/or A u operational costs? t Tight quality control allows for optimisation of o m process lay-out, i.e. reduced blending a facilities and controlled use of t difficult/expensive raw materials i o n & Q C 

6

Why automate? automate? 

Lastly, you may also want to save labour cost. 

This may be done by exchanging the operator/lab tecnician with automated equipment from sampling to analysis


7

Main Purposes of Lab Automation 

Tight quality control during production  On-stream, Inline analysers etc  Automated Lab system



Quality requirements from customers on (end) product  Automated Lab systems


8

9 L a b A u t o m a t i o n & Q On-line Analysis: C

Laboratory Automation and On-line Analysis

or PGNAA

PGNAA

Central laboratory automation

XRF

FCaO/PSZ 9

10 L Automated Lab Systems a b  Based on very accurate analytical A u instruments, correct sample preparation t o methods, reliable sampling and sample m transport systems a t  Various suppliers (FLSA, Polysius, Pfaff, i o Herzog, Iteca …) n  Possible FLSA solutions will be discussed & in details, competitor solution mentioned! Q C

10

FLS Automation lab systems


FLS Automation lab systems QCX/RoboLab  integrates manual or semi-automatic analysis &

sample prep equipment  high flexibility in equipment selection high to very high capacity  built in manual back-up  modular design allows for easy integration of additional equipment

QCX/ AutoPrep AutoPrep  automated equipment units  well suited for small powder prep ‘only’

applications  small to medium capacity  built in manual back-up


12

Quarry & Piles

Raw meal Silo

Cement Silos

Raw mi ll

Clinker storage Kiln

Manual production lab

QCX Scalability: Small Small

Cement mill


Quarry & Piles

Raw meal Silo

Cement Silos

Raw mi ll

Clinker storage

Cement mill

Kiln

Automated production lab

QCX Scalability: Medium Medium size


ASIA Cement, Thailand


QCX Scalability: Large Large Quarry & Piles

Raw meal Silo

Cement Silos

Raw mi ll

Clinker storage

Cement mill

Kiln

AAS

Fully automated production lab

CST

Plant wide LIMS for ALL analysis & test data


QCX/RoboLab Mill

X-ray roo m

Sample prep room

Press

Mill

XRF

Composites

Sample transport Colour

XRD Dosing/ cleaning


Particle Size Fusion

17

QCX/RoboLab -- General features • Sample receipt

• Identification & registration • Sample splitting & dosing • Individual preparation recipes • Priority management • Alternative routing in failure situations • Integrated control and dynamic supervision of : - Robot - Prep eqp & Analysis instrms ( - Sample transport PLC )


18

QCX/RoboLab Configurations Configurations

AUT

LAB

AUT

LAB

AUT

LAB

AUT

LAB


20 L Recommendation a b  Allow your self time to study your needs A u carefully t o  Discuss your needs with qualified m suppliers, who knows the process and the a t problems most often seen i o n  If any doubt, have a pre-study made on & your materials to ensure the correct Q choice of equipment C



Balance your investments between sampling and lab equipment. Where do you gain the most?

20

21 L a b A u t o m a t i o n & Q On-line Analysis: C

Laboratory Automation and On-line Analysis

or PGNAA

PGNAA


XRF

FCaO/PSZ 21

Competing technologies ?

On-line analysis

On-line analysis


Main objective : quality control Applied on more and more sampling points

Central laboratory Main objectives : quality control and quality assurance manpower savings, correct sample preparation, ...


22

QCX & & On On -Line StockPile StockPile Control Control Process flow Analysis data

Limestone

Accounting Control loop (manual or auto)

Clay

Feed proportions

PGNAA

QCX/ BlendExpert OLA


QCX/Laboratory 23

QCX/BlendExpert & & Pile Cntrl Cntrl 

Circular stockpiles: accounting in accordance with circular coordinates of stacker and reclaimer, if coordinates are provided from the PLC. If not, accounting is simple ‘endless’ integration


24

QCX/BlendExpert & & Pile Cntrl Cntrl 

Longitudinal stockpiles: the accounting provides one total average composition of the stockpile section currently being stacked = simple batch integration.

After completion status changes to ‘reclaimed’ and later ‘historic’ pile


25

“ Raw mix on the stockpile” Sand

Process flow Analysis data

Pyrite

Limestone

Accounting Control loop (manual or auto)

PGNAA

QCX/ BlendExpert OLA


QCX/Laboratory Control scheme for pro ducing r aw mix on the pile

26





27 L “ Raw mix on the stockpile” a b Before: Manual stockpile control A 4 component mix cntrl before before mill u (Polab ) t o LSF std dev’s dev’s out out of mill = 4-5 m a Now : Two stockpiles (one circular, one t i longitudinal) equipped equipped with PGNAA o n analysers (ASYS (ASYS FSA) & Manual feed to pile control including additives Q from both piles to raw mills, C 100 % feed from

thus no mixing before raw mills LSF std dev’s dev’s out out mill = 1.7-1.9



Future:

Automation of additive feed 27

QCX & On On -Line Raw Raw Mix Control Control Raw m aterials

Raw mi ll Feeder set points

PGNAA

QCX/ BlendExpert

Fast OLA

QCX/ OnStream Fast

Process flow Sample

Av erage sample


Slow

QCX/Laboratory

Analysis data Control loop

Control can be based solely on on-line analysis or in a redundant set-up inco rporating lab XRF analys is

28

QCX/OnStream Compact and robust cabinet (IP65) with local control panel  Analyses all 4 main elements: Fe, Ca, Al, Si  Analysis time 2-5 minutes (EDXRF)  Lower investment required than for PGNAA-based alternatives 


Combines traditional and on-line techniques. Tighter quality control at lowest possible cost. 29

QCX/OnStream

alternative P ro ce ss f l o w

OnStream air sli de sampler OnStream scr ew sampl er

Dosing device

P r o c e s s f l o w

Analyzer cabinet Blower

Excess material return


QCX/OnStream cabinet 30

QCX/OnStream

P ro ce ss f l o w

OnStream OnStream scr ew sampl er

Dosing device


Analyzer cabinet Blower


31 L a b A air sli de sampler u t o m a Screw sampler t i o Air slide sampler n & Q C alternative

Dosing device 31

QCX/OnStream ::

alternative P ro ce ss f l o w

OnStream air sli de sampler OnStream scr ew sampl er

Dosing device


OSCA3000 analyzer Analyzer cabinet Blower



Calibration sample outlet 32

QCX/OnStream Control Control performance performance 130

LSF 120 110 100 90 80

Before QCX/OnStream

With QCX/OnStream

US Plant 1999

70

High frequency sampling and analysis Fast control response Reduced kiln feed variations with small (or poor) silos Reduced work load for sampli ng, preparation & analysis Robust installation


33

OnStream ttest results results Static

Dynamic

SiO2

Results 0.078

Results 0.111

Al2O

0.072

0.096

CaO

0.065

0.126

Fe2O3

0.008

0.022

LSF

0.57

0.72


Results achieved with 30 secs integration time for all elements. 34

Online Analytical Lab WDXRF ( 20 sec )

SiO2

0.05

Al 2O3

0.015

CaO

0.025

Fe2O3

0.004

ASYS

35 L Performance a b A On-line On-line u EDXRF PGNAA t o ( 100 sec ) ( 10 min ) m 0.10-0.20 0.15-0.30 a t i o 0.10-0.15 0.12-0.20 n & 0.10-0.20 0.25-0.40 ( 0.15-0.20 ) Q C

0.03-0.05

0.03-0.06

Precision ( = repeatability) data incl sample prep for Lab WDXRF & sample presentation for On-line EDXRF

1 2

5

9

36 L Precision a b if this is Lab WDXRF A u then this is t 20 1 o on-line EDXRF m 1 8 a t i 1 o 3 n & 1 Q 0 C 1 4

4 1

6

1 1

6 1

8

7

9 1

3

1 7

2

5

and this is on-line PGNAA 36

Error on ’Dynamic’ Precision C3S / LSF S.dev

6

4

: sampling : sample preparation

(10 min)

: analysis

4½

3

3

2

Traditional Lab WDXRF (20 sec)

1½

On-line PGNAA

1

On-line EDXRF OSCA 3000 (100 sec)

S Y A S


38 L Error on ’Dynamic’ Precision a b C S / LSF S.dev Including time for fillin g of flow cell, compacting, A 12 8 emptying etc, one OnStream cycle is 5 min…. u t so this would be the fair value to compare with o m 9 6 ----- On-line PGNAA ----- a t (5 min) i o n 6 4 & (10 min ) Q On-line EDXRF C 3

3

2

OSCA 3000 (5 min)

S Y A S

S Y A S

C3S/LSF S.dev

12 8

9

6

Manual/ off-line control + Lab-XRF

Based on hourly average samples taken at raw mill outlet and measured on Lab WDXRF ( = ‘tr ue’ value)

Lab-XRF Lab-XRF 6

4

manual sample prep (1 sample/hr)

automatic sample prep (RoboLab) (2 samples/hr)

3

On-line PGNAA

On-line EDXRF

2 : Advanced computerized raw mix blending control


Traditional Vs OnStream OnStream analysis analysis Traditional

OnStream



QCX/OnStream combines the high freq. of the OnStream analyser with the high precision and high accuracy of lab analyser



QCX/OnStream monitors the process dynamics; the lab analyser establish the accurate chemical levels



The ‘redundant’ structure is tolerant to errors in one of the instr instruments

Sampling error

Sampling error

Preparation error Analytical error

Analytical

& sample presentation

error


40

FLSA recommendations -EDXRF or PGNAA for raw mix control (Choice to some extent pending process lay-out details) EXDRF has the potential for lowest LSF std.dev. Out of mill -PGNAA for stock pile control where requirements to analysis precision are less stringent, and for raw mix control when long transport delays in the material department exists (e.g. Surge bins) QCX supports interfacing og all PGNAA models on the market

On-line EDXRF/PGNAA does not eliminate the need for the traditional lab analyser (WDXRF)


41

Future Trends 

On-line instrumentation for control of material properties (chemical analysis, fineness …) will increase to support JustIn-Time production:    

 

Reduce material storage's Recipe control Special products and qualities Fast change of production to reduce waste Utilise difficult raw materials Power savings


42

43 L Future Trends a b  Increased demand for documentation and A accountability from both customers (Total u t o Quality Control) and from authorities on m environmental impact: a t  System must handle large amount i o of data in a safe way n &  Laboratory procedures will be very stringent and automation will be Q C wide-spread in other areas of the

cement analysis (physical testing, fuel, …)

43

32 - Lab Automation

Recommend Documents