system of cane sugar factory control - 3rd ed

7/27/2019 System of Cane Sugar Factory Control - 3rd Ed

1/87

SYSTEM OF

CANE SUGAR FACTORYCONTROL

Gardens Point.

A22810994B

System of cane sugar

factory control

THIRD EDITIONEd it ed by J. L. CL AYT ON

P U B L I S H E D F O R T H E

I N T E R N A T I O N A L S O C I E T Y O F

SUGAR CANE TECHNOLOGISTS

b y

Q.S.S.C.T.

1971


2/87

A22810994B

Wholly set up and printed in Australia byWATSON, FERGUSON AND COMPANY

Brisbane, Q.1971


3/87

PREFACE TO THE THIRD EDITION

The f i rs t ed i t i on of th e book "S yst em of Cane Sugar Factory C o n t r o l "al t hou gh p ubl i she d by a C o m mi t t e e of th e I .S.S.C.T. was, in fact, w r i t t e n by D r .

F. W. Ze rb an , and i t inc or po ra te d no t onl y his sty le b ut also cer tai n of his

personal opin ions.

W h e n I revised th e boo k in 1955 I t o o k pains to al te r i t no m or e th an neces

sary th ou gh I did n ot subscrib e to some of th e claims and pol ic ies th er ei n.

Af te r th e 12th Congr ess I set o ut to per fo rm a fu r th er revi s ion bu t I was so

dissat isf ied w i t h th e results tha t I abandon ed th at l ine of ac t i on . I deci ded tha t th e

only sat is factory course was to rewr i te the book ent i re ly .

To th e 13th Congress I t oo k the maj or po r t io n of th e new te xt in pr oo ffo rm and d i s t r ibu te d i t to a panel of techno logis ts re pres ent i ng a wi de range of

sugar cou nt r i es . I in vi te d comm en ts fr o m th em al l , and I her e express my tha nks

to those who were good enough to assist me with their advice. I have done my

best to use i t co ns tr uct ive ly .

J . L . C L A Y T O N .

1971


4/87


5/87

PREFACE TO THE SECOND EDITION

When the Committee on Uniformity in Reporting Factory Data met duringthe Eighth Congress in Brit ish West Indies in 1953, one of its resolutions wasthat the booklet "System of Cane Sugar Factory Control of the I.S.S.C.T." be"brought up to date and reprinted".

On the same occasion the resignation of Dr. F. W. Zerban as Chairmanof the Committee was accepted with regret, and the writer was appointed tothat office, with Mr. J. L. Clayton as Secretary. The major portion of Dr. Zerban'sexcellent composition in the original booklet remains unchanged in the new

edition, the preparation of which is due to Mr. Clayton.

The question of bringing the booklet "up to date" has occasioned considerable thought. Certain revisions had been approved by the Committee asa body but others arose for consideration and it was a question whether theyshould be included without formal approval. The policy adopted was to selectthose items which needed revision but were not of a highly contentious natureand to submit proposals to the regional sub-committees. The response wasgratifying, the proposals being either approved or commented upon in constructive manner. Our thanks are due to the members of the sub-committeewho assisted.

In the current edition two appendices have been included. The first presentsdata on the units employed in various countries and deals extensively withEnglish-Metric conversions. The second deals with the Tables used in sugartechnology. Though there were several requests for the publication of standardTables it was decided not to print these, for reasons explained in Appendix II .

Thanks are due to the Queensland Society of Sugar Cane Technologists,

which body has undertaken the financial responsibility for the printing of this

booklet.

It is hoped that the current edition will prove a worthy successor to thefirst and that its publication wil l serve to further the good wo rk of the Committee.

NORMAN J. KING,

Chairman.


6/87


7/87

PREFACE TO THE FIRST EDITION

HISTORY AND AIMS OF THE COMMITTEE

The Special Committee on Uniformity in Reporting Factory Data was created

at the Second Conference of the Society, held at Havana in 1927, upon mot ion

made by M. A. del Valle, who pointed out that the confusion of terms used in

sugar factory reports, and the multiplicity of control methods employed made

it impossible fairly to judge the results obtained and to make mutual comparisons.

He urged a study of this question and the establishment of uniform methods by

common consent.

The Committee members appointed at the beginning were M. A. del Valle(Puerto Rico), P. C. Tarleton (Cuba), and F. W. Zerban (U.S.A.), chairman. At

the Third Conference (Soerabaja, Java, 1929) E. C. von Pritzelwitz van der Horst

(Java) was appointed, vice P. C. Tarleton ; he resigned prior to the Fifth Conference

(Brisbane, Queensland, 1935), and no new appointment was made to replace him.

The chairman was empowered to co-opt further members. In order to make

the Committee truly international in scope, the co-operation of prominent

sugar technologists in all the important sugar cane countries was solicited, those

connected with official institutions or technical organizations being chosen

wherever possible. The response was most gratifying, and the following menhave served on the Commit tee at various times, many of them throughout the

entire period during which it has been functioning:

Argentine: W. E. Cross;

Australia: Norman Bennett, W. R. Harman, A. Jarratt, G. S. Moore;

British and French West Indies: Walter Scott, J. G. Davies;

Cuba: A.J. Keller, W. B. Saladin, H. D. Lanier, A. P. Fowler, Jose Santos;

Hawaii: W. R. McAllep, W. L. McCleery, S. S. Peck;

India: Noel Deerr, J. H. Haldane, K. C. Banerji;

Japan: Migaku Ishida, S. Kusakado;

Java: P. Honig, C. Sijlmans, Ph. van Harreveld;Louisiana: C. E. Coates, A. G. Keller;

Mauritius: Louis Baissac;

Mexico: T. H. Murphy;

Natal: H. H. Dodds, G. C. Dymond;

Peru: Gerardo Klinge;

Philippines: Herbert Walker, E. T. Westly;

Puerto Rico: M. A. del Valle, Jaime Annexy, E. M. Copp;

Santo Domingo: Rafael Cuevas Sanchez.

In some of these countries sub-committees were formed to advise themembers of the Committee.


8/87

To all these men the chairman again expresses his sincere gratitude because

without their willing, sympathetic, and efficient help it would have been im

possible to accomplish the large and difficult task allotted to the Committee.

Since the members of the Committee are scattered all over the world it

was necessary to undertake the work largely by means of questionnaires and

correspondence. The replies to the various questionnaires were analysed,

arranged, and summarized by the chairman, who then prepared comprehensivereports for each of the meetings of the Society. These reports were discussed

at the Conferences by the Committee members present or their proxies, and

all amendments agreed upon were entered in the reports. The revised reports

were submitted to the Society as a whole, were adopted by it, and published in

the Proceedings. The general terms and definitions, and the system of milling

control were thus disposed of at the Third Conference in 1929. The cont rol

of the boiling house, and the methods of weighing, measuring, sampling, and

analysis were reported at the Fourth Conference in 1932, but since the Commit tee

was poorly represented at that meeting it was decided to resubmit the report

to all the members of the Committee by correspondence. The final reportwas presented at the Fifth Conference in 1935, and was adopted. But the

Committee was retained in office in order to keep the methods up to date and

to revise them from time to time, in keeping with progress in sugar technology.

Accordingly, a report was presented at the Sixth Conference in 1938, but only

few changes in the methods were made at that time.

In the present book the reports of the Committee have been rearranged and

systematized so that they may be more readily consulted. In some instances the

original text has been somewhat condensed, in others amplified, always keeping

in mind the intent of the Committee's wo rk . The purpose of this treatise is

not to serve as a complete manual of control methods, but rather to explainthe principles which guided the Committee in arriving at its decisions, and to

emphasize essential points. Those well known methods the details of which

may be readily found in any of the generally used handbooks of factory control

are not described at length, but simply referred to by name. But the methods

recommended by the Committee which are not so widely known, are given in

full in Chapter VII. In other words, the book is not meant to replace existing

manuals, but to supplement them and to be used as a guide for those countries

or associations that desire to bring their control methods in line with those

recommended by the Committee, and thus to accomplish the purpose for which

the latter was created. It is hoped that the publication of this international systemof factory control may be found helpful toward that end.

The Committee accepted for its guidance the following principles enunciated

by S. S. Peck: "Your committee should strive for three main objectives, namely,

accuracy, clar ity, and simpl ici ty; and of these three I consider the last as important

as the first two. In s triv ing for greater accuracy, formulas have become so complex

that they are practically useless. If the committee stress simplicity of statement

which will not conflict with accuracy and clarity, they may be able to do some

persuading to an agreement on terms." It was also postulated that wherever

direct determinations can be accurately and simply made, they should be preferred

to indirect determinations or calculations, and further, that practical considera

tions should be favoured against theoretical speculations.


9/87

CONTENTS

PAGE

CHAPTER I

IN TR OD UC TI ON AN D SOME PREMISES . . . . . . . . . . . . 11

CHAPTER II

PRIN CIPL ES OF MI LL IN G CO NT RO L . . . . . . . . . . . . 13

CHAPTER III

DETERMINATIONS AND CALCULATIONS FOR MILLING CONTROL . . . . . . . . . . . . 18

CHAPTER IV

CO NT RO L OF T H E BO ILI NG HO US E . . . . . . . . . . . . 29

CHAPTER V

ME TH OD S OF WE IG HI NG AND ME AS URI NG . . . . . . . . . . . . 39

CHAPTER VI

ME TH OD S OF SA MP LI NG . . . . . . . . . . . . 47

CHAPTER VI1

ME TH OD S OF AN ALY SI S . . . . . . . . . . . . 55

CHAPTER VIII

DE FI NI TI ON S AND INT ERP RET ATI ONS . . . . . . . . . . . . 75

APPENDIX 1

CO NV ERS IO N OF U N I T S . . . . . . . . . . . . 81


10/87


11/87

CHAPTER I

Introduction and Some Premises

The manufacture of cane sugar from sugar cane is a distinctive industrial process in that it involves no element of synthesis.Sucrose comes into the factory in the cane, and, subject to somephysical losses and some destruction, emerges in the product,crystal sugar. The process is essentially a combination of separationand concentration.

The materials other than sucrose in the cane are collectivelyknown as impurities, and may be classified as dissolved and insoluble respectively. The first step is the separation of sucrose and

the impurities in solution from the insoluble impurities, togethercalled fibre. This is the function of the milling plant, and the processis commonly called extraction. The second step is the treatmentof the extracted juice for the removal of some insoluble and somedissolved impurities, and this is known as clarification. A considerable proportion of the water present is then removed in the processcalled evaporation. The further stages constitute the separationof impurities by crystallization of the sucrose. The processing of

juice to crystal sugar takes place in the boiling house, and theseparation of sucrose is called recovery. The overall separation of

sucrose from cane is also known as recovery.The proprietor of a sugar factory is naturally interested in

knowing how much of the sucrose in the cane he purchases ispresent in the sugar he sells, and he also wishes to know how goodor how bad is the achievement. This demands a series of measurements, analyses and calculations which constitute the systemknown as chemical control.

The main purposes of chemical control are threefold

1. To ensure that the various unit operations in the process of

manufacture are conducted at the highest efficiency. This is themost important function of chemical controlto provide"live" data for the immediate guidance of the plant operators.As this book does not pretend to be a manual of instructionfor operators, it touches lightly on this function of chemicalcontrol.

2. To provide a quantitative account of materials and their components entering the process, in transit, in stock, and leavingthe process. From the basic data convenient measures of per

formancesuch as sucrose extraction or overall recoverymay be derived, but most of such records are dispassionatelyfactual.

11


12/87

3. To assess the merits of performances achieved. Sugar technologists realized long ago that it is not reasonable to judge sugarfactory results against the standard represented by perfection.Other things being equal, an extraction of 96 is better than anextraction of 94; but in practice other things are commonly

not equal, and the 94 may represent a more meritorious performance than the 96.The system commonly adopted is to compare the actual result

with an abitrary standard resulta standard result which tempersprefection by recognition of the prevailing circumstances. Since theformulation of a standard involves speculation, none of the arbitrarystandards is above reproach, and different lines of speculativereasoning can lead to different standards for the same operation.This book discusses many of these standards and attempts to selectthe best for common usage.

Not all the "figures of merit" involve a comparison against astandard. Some are purely factual, and express merit only by virtueof an accompanying assumption; for instance, lost absolute juiceper cent fibre is a statement of fact but it is regarded as a figure ofmerit by thoseand there arc manywho accept as a generaltruth that lost absolute juice is an index of milling efficiency.

Sucrose and Pol.

Although the material of primary importance in the sugarfactory is sucrose and the accounting for sucrose should constitute

the main material balance, this is not generally so. The determinationof sucrose as such is laborious and more prone to error than themeasure of apparent sucrose by direct polarization, known as pol.

There is no doubt whatever that pol is used more generallythan sucrose as the basis of chemical control, and, until there is somesignificant change, pol must be the common basis. It is all verywell to point to the superior merits of sucrose, but, if the sugarindustry of a country will not adopt sucrose in its own interests, itis hardly likely to do so for others.

In plain fact, the normal relationships between pol and sucrose

are gratifyingly stable and most of the time chemical control on apol basis is entirely satisfactory. If suspicious results arc recorded atany time, the pol-sucrose relationshipscan be checked for abnormality.

For every pol and every derivative from pol there is a sucroseequivalent. This should be kept in mind, because, to save wearisomerepetition, this book deals primarily in pol. Those who prefer thealternative, sucrose, as a basis of control, are welcome to adhere toit, but should qualify reports accordingly.

There is one important exception. The concentration of opticallyactive impurities in final molasses is so high that the pol and apparent purity of final molasses are practically meaningless. Thosefigures should be used, as required, for purpose of calculation, butonly purities based on sucrose have any absolute significance.

12


13/87

CHAPTER II

Principles of Milling Control

The che mical con tro l over the oper ati on of milling involves accou ntin g for four mat eria ls can e, ba gasse, mixe d juice an d wat erand their components. Before procedures may be discussed somefeatures of the materi als are due for cons ider atio n.

Cane.

Cane may be taken to comprise fibre, being the aggregate of allcomponents in the solid phase, juice, being the aggregate of allcomponents in the liquid phase, and, possibly, hygroscopic water,being water physically adsorbed by some of the fibre.

At this stage it is pertinent to mention that terms like "solid"or "insoluble" or "undissolved" have to be accepted with somereservations. In many ways Nature does not deal in clear distinctions,an d the technolog ist wh o studies the structu re of cane in mi nu tedetail will find components to which classification as solid orliquid cannot be applied with certainty. For ordinary purposesfibre is a solid or insoluble component, but in finer degree fibre hasto be distinguished as a "non-liquid" component, and even thisdistinction is not absolute.

Th e co mp on en t called juice is really a hete roge ny of jui ces the rich juic e of the p ith cells, the po or er juice s of the r ind a nd th einte rnod es, an d the watery con ten t of the fibro-vas cula r bund les .

In earlier years there seems to have been some doubt aboutthe existence of the thi rd co mp on en t "hygr osc opi c wa te r" . It cannow be stated categorically that hygroscopic water exists and thatwhe n cane fibr e ha s been stee ped in dilute jui ce or water, th e hyg ros copic water which it adsorbs is somewhat variable in quantity, butis o f the o rd er of 25 pe r cent of the weig ht of f ibre.

Hygroscopic water was referred to as a "possible" componentof can e bec ause , alt ho ug h its existence in baga sse is beyo nd ques tio n,its existence in cane is not proven, and there is substantial evidenceth at if it is at all pa rt of can e it is pres ent in a very small pr op or ti on .

Undiluted Juice.Probably in order to avoid the complexityof allowing for the variations within the true juice of cane, thetechnologi sts of the Jav a sugar ind ustr y ad op te d a concept th at the

juice left in cane afte r dry crushing had the same Brix as the juiceexpressed by dry crushin g, i.e., pr im ar y juic e. T he who le juiceof th e cane, co mp ut ed on this basis, was terme d undilut ed juice .

In genera l the Brix of the resi dual juice is lowe r th an th at of thepr im ar y juic e, so if the re sid ual juic e is credited with a high er Brix

13


14/87

than it possesses, its quantity is less than the true juice. This leaves adeficiency in the materials balance, and the term "undeterminedwat er" was applied to the closur e. Can e was pr es um ed to consi st offibre, undilut ed juice an d und ete rmi ned water. It can be rea son edthat this undetermined water must consist of water which is part

of the true juice but not part of the undiluted juice, together withany hygroscopic water.

A wea kne ss of this con cep t for prac tic al pu rpo se s is th at theBrix, and ther efore the qua nti ty, o f undi lute d juic e dep end s on theBrix of pr im ar y juic e whi ch is subject to extern al influences, specifically, the crushing conditions.

Absolute Juice.To pro vid e a mo re st ably based q ua nti ty tha nundiluted juice the concept of absolute juice was adopted. Theassumption is the ultimate in simplicitythat cane consists entirelyof fibre and absolute juice. If there is any hygroscopic water, it isrega rded as par t of the a bsol ute juice.

Abs olu te juice wa s not presume d to exist as such, an d it certainly does not in bagasses, but its existence in cane may be closer toreality than is generally imagined.

It is well known that the milling factorthat is, the factorconverting the Brix of first expressed juice to the Brix of absolute

juice, is of the order of 0.975. This factor may be regarded as thepr od uc t of tw o su bsi dia ry fac tor s, one to conv er t the Brix of firstexpressed juice to the Brix of tru e juic e, and one to conv ert th equantity of true juice to the quantity of absolute juice (that is, toallow for hygrosco pic wa ter ). If the first of these tw o factors wereunity, that is, if the factor 0.975 were solely to correct for hygroscopic water, the correction would represent 17.5 per cent hygroscopic water at 12.5 fibre in cane; but it is invariably found thatthe Brix of the true juice is below that of the first expressed juice,and therefore the hygroscopic water allowed for is less than statedabove. Actually the factor to convert Brix of first expressed juiceto Brix of tr ue jui ce is of th e same o rd er as the overall factor, 0.975,and therefore the second factor is approximately unity.

It is not proposed to pursue this subject exhaustively butevidence from practical milling results, from press tests, and fromalt ern ati ve me th od s of de te rm in at io n of fibre in cane all leads to theone conclusion that, for practical purposes, there is no hygroscopicwater in cane as cane. There is a suggestion that the adsorption ofhygroscopic water begins when cells are disrupted and proceeds ata quite moderate rate. This can explain why cane pieces crushed in apress yield juice of declining Brix ; but when the pressing is interrupted a nd r esum ed later , the Brix of the jui ce jum p s to a new level an d

declines again. Wh ate ver be the expla nat ion, experime ntal resultssuggest that hygroscopic water should not be allowed for in originalcane, but must be allowed for in bagasses and disintegrator slurries.

14


15/87

Bagasse.

Bagasse her e me an s final bagasse, the end pro du ct of the millingtrain. It comprises fibre, juice and hygroscopic water. The fibre isalm ost the whole of th at originally present in the cane from whichthe bagasse was derived. Losses are negligible, but a proportionap pro xi ma ti ng 0.5 per cent on cane passes out of the milling tra m

with the mixed juic e. In strict ac cou nti ng this quan ti ty has to b eallowed for, but for general purposes it is assumed that the fibre incane all becomes fibre in bagasse.

Th e juice is a mixtu re rang ing from virtually water to origin aljuice still enclosed in a few inaccessible cells. The fibre containshygroscopic water in quantity usually assumed to be 25 per cent ofthe weight of fibre.

Since a final bagasse co nt ai ns some 50 per cent of fibre whichin turn hol ds some 25 per cent of hygroscop ic water , the quan ti ty of

this last item is sub sta nti al 12. 5 per cen t of the bagasse. Th isma ke s an apprecia ble difference between the true residual jui ce,37.5 per cent , and th e abso lu te resid ual jui ce, 50 per cent. If th ebagas se cont ain s 4 per cent Brix, then the conc ent rat ion of theabsolute residual juice is 8 Brix, but the concentration of the trueresid ual juic e is abo ut 10.7 Brix.

There is rarely occasion to consider the average compositionof the res idual juic e. Resi dual juice is co mm on ly reg arde d as amixture of some standard juice and water; the most popular standard juice was undiluted juice, first expressed juice has been usedby ma ny , bu t the rec omm end ed choice is absolut e juice. If the Brixof the abso lu te jui ce of the ca ne was 20, the n the bagas se referred topreviously may be said to comprise 50 per cent fibre, 20 per centabsolute juice and 30 per cent water, the water being made up of12.5 per cent hygroscopic water and 17.5 per cent imbibition water.Th e pro po rt io n of abs olut e juice was derived on a Brix basis, a ndthis is the standard practice; but it can be derived on a pol basis, andsuch a procedure is inherent in the Reduced Extraction formula ofNoel Deerr.

The standard method of analysis of bagasse at present involves the dete rmi nat ion of dry substa nce an d pol. T he Brix is generally derived from the pol usin g the pu ri ty of last expres sed juiceor last mill juice. For generations it has been acknowledged thatnei the r of these pur iti es is even close to the pur it y whi ch is reallyinv olve dt he purity of the resid ual juice in baga sse.

In earlier years the direct determination of Brix in bagasse wasknown to be possible but was considered to be too exacting forroutine use. Nowadays the high speed wet disintegrator provides aready means of releasing the Brix into an extract, and a precision

refrac tomet er serves to det erm ine the Brix of the extract. Dire ctanalysis for Brix is still not recommended in relation to every sample,

15


16/87

bu t it is pract icab le to analyse a sufficient numb er of ext rac ts for polan d Brix to mai nt ain an ade qu at e mea sur e of the prevailing pu rit y ofresidual juice in b agasse .

Mixed Juice.

Mix ed juice is the main liquid pro du ct of the mill ing train .

It inco rpora tes all the extrac ted juic e of the cane to ge the r with themajor pa rt of th e imbi bit ion water. It also cont ain s a solid co mpo ne nt mad e up of soil, particl es of fibre, an d othe r mi nor items .Cane of a high standard of cleanliness yields about 0.3 per centinsoluble mat te r in mixed juice , an d when the cane is dirty thisfigure may rise well over one per cent.

Thi s insolu ble ma tt er is technic ally fibre , no t jui ce, an d it sho uldbe tak en into acc ount accord ingly. W he n the mixed juic e is weighed,a sample should be analysed for insoluble matter, and the grossweight of mixed juice sh oul d be app or ti on ed between fibre in mixed

juice and clean mixed juice. N o t only is this correct as regardsaccounting for materials, but also it relates the analysis of themixe d juice to th e mat eri al actual ly analy sed.

Th e po l of mixed juice will nor mal ly be det erm ine d by the drylead me th od . T he pol thus meas ure d is the pol of the liquid pha se,the clean mixed juice. The Brix should be determined on filteredmixed juice, because, if the juice is not filtered, the inflationaryeffect of th e susp end ed ma tt er is inter pre ted as extra dissolvedsolids which do not really exist.

Water.The water referred to in this context is the net quantity of

water whi ch is ad de d in th e milling proce ss. Mo st of it is, of cou rse ,applied as imbibition, and a little may come in through hoses orsteam lines. On the other hand, a substantial quantity is lost byevaporation, particularly when hot milling is practised.

The Mass Balance.

Acco rdi ng to the previ ous edition of this bo o k "t he fund amen talequ ati on for the weights of the pro du cts entering an d leaving themill states tha t cane plus wate r equ als mixed jui ce plu s bag as se ".This is a da ng er ou s over-simpli ficat ion, for it fails to specify thatthe ter m "w a te r " has to mea n the net balan ce of ad de d water.

More precisely, the fundamental equation is:Ca ne + wate r ad de d = juic e + bagasse + wa ter lost.

In earlier days it was not practical to weigh bagasse, and thequantity thereof was calculated by subtracting the weight of mixed

juice from the combined weights of cane and added water. Theresult is not truly the weight of bagasse but the combined weights

of baga sse an d wat er lost. Th e loss of wat er is mai nly by ev apor at io nfrom the milling train. Extensive tests on a milling train workingund er hot condi tion s have disclosed a loss of water by evap or at io n

16


17/87

repr esen ting 3 per cen t of the weight of in co mi ng mat eri al. Un de rthes e con dit ion s, the weight of baga sse calcul ated by difference willbe inflated by 18 per cent. Th e examp le may be extr eme , bu t in anycase the error due to evaporation is serious enough to discourage theuse of the simple ma ss bal anc e to de ter min e the weight of baga sse.Furthermore this method is cumbersome, and the ban on the

use of un me te re d wate r at the mills ha s a high nuisanc e valu e.

No wa da ys the weighing of baga sse can be carr ied out as aroutine operation and the statement that "the best way to determinethe weight of bagasse is to weigh it" is no longer facetious.

Th e weights of mixe d juic e and bagas se, indi vidua lly a nddirectly determined, represent a powerful combination. The weightof ad de d wate r can be dispens ed with, a nd even the weight of can eis unnecessary if one key componentfibre or Brix or pol per centcanecan be determined.

When it is impossible or inconvenient to determine the weightof bagasse the best proc ed ure is to weigh the cane a nd dete rmin eits fibre content. The weight of fibre in cane, less the weight of fibrein mixed juice, is the weight of fibre in bagasse. This leads to theweight of baga sse and its co mp on en ts . Thi s system was originallyaccepted by the International Society with some misgivings as tosampling cane for fibre. It is sufficient to state here that reliablesampling can be achieved and the scheme works well in practice.

A very important contribution to factory controlan innov ati on since the previous edition of this bo ok h as been the

dev elo pme nt of the direct analysis of can e using the wet dis inte grat or.This operation has now been established as practical and reliablesubject to the normal demands on diligence and maintenance. Itssuccess de pe nd s up on the pro vis ion of reliable sample s of ca nebut this requirement can generally be met.

If th e weight of ca ne is kno wn , a nd th e analysis of th at can e isdet erm ine d, mos t of the purpo ses of the mo re complic ated systemsare achieved. Fibre in cane, corrected approximately for fibre in mixed

juice, gives fibre in bagasse, hence the weight of bagasse and it s compo ne nt s. Pol in can e less pol in bagas se gives pol in mixe d jui ce, hen ce

the weight o f mix ed jui ce an d its co mp on en ts . I f th e weight o f ca neis unk no wn , then the weight of mixe d juice sho uld be dete rmine dand the above procedure applied with the necessary modifications.

Th e previ ous edi tion s of this bo o k listed five oth er bases ofcon trol . Ea ch of th em inc orp ora tes an arbi trar y factor or a questio nable assumption, and none was formally approved for use. There isno point in re-stating them here.

It remains only to add that, in some factories, cane, water,bagas se an d mixe d juic e are not the only material s entering orleaving the milling train. Any other material involved, such asdecant fluid from mud treatment, must be accounted for as toquantity and composition and taken into the materials balances.

17


18/87

CHAPTER Ell

Determinations and Calculations for

Milling ControlIn Cha pt er II th e general ba sis of milling contro l was discussed,

and several schemes were outlined broadly. It is now desirable toclassify these schemes, express them in more detail, and go on toderive all the items of a full co ntr ol pr og ra mm e.

For this purpose a system of symbols has been devised, asfollows. The materials and components are represented by lettersCan e C, c Brix B, bBagasse (A mp as ) A, a Pol P, pNe t Ad de d Wat er I, i Fib re F,fMixe d Juice (actual ) M , m Wat er W,wMixe d Juice (clean) J, j Pu ri ty Q

Excepting purity, the capital letters are used to refer to quantities, and the reference material is identified by a lower case subscriptletter. Hence Ac is the amount of bagasse obtained from the originalweigh t of cane , and Fa is the weight of fibre in that bagasse. Lowercase letters are used to express one item as a pr op or ti on of an ot he r,and, for simplicity, a unit basis has been adopted. Hence p c is pol

per unit cane and fa is fibre per unit bagasse. Two subscripts areoccasionally necessary as in pfa, pol per unit fibre in bagasse.

It follows , for exa mp le , t ha t

A x pa = Paand Fc fc = C

Preliminary Data. When cane is analysed the items determined directly are water, wc, Brix, bc, and pol pc. Fibre is determinedby difference, fc = 1 wc bc

Bagasse is analysed directly for water, , and pol, pa. Brix isdet erm in ed from pol and th e pu rit y of the residua l jui ce, he re expressed as Qa, (again on a uni t basis ). Henc e ba pa Qa. Fibreis then determined by difference, fa = 1 wa ba.

Mi xe d juic e is ana lyse d for fibre, / , , and it follows th at J=M Fm or jm = 1 fm. The clean mixed juice is analysed forBrix, bj, and pol, pj. It is assumed that, in all cases, ba, pa, fa, wa,

fm, pj, and bj are known.

Basic Control Schemes.

Class IWhen the weight of cane is known.Scheme ACane weighed and analysed.

18


19/87

Scheme DCane, Mixed Juice and Added Water weighed.

This is the familiar "mass balance" method. The derivationof results usually starts with the assumption that the actualwe igh t of wat er a dd ed equal s the n et wei ght J. This is adubious assumption but, in the absence of knowledge regar din g inc ide nta l gains or losses of wa ter , it ha s to be m ad e.

19


20/87

20


21/87

This listing of schemes may not be exhaustive but it shouldcover all the cases likely to be encountered in practice and amenableto absolute calculation. Conspicuous by its absence is the casewhere nothing is known of the caneneither its weight nor any of itscomponents quantitatively. In such a case the weight and composi

tion of the cane cannot be determined by any absolute method; someempirical factor or arbitrary assumption must be invoked.

The procedures above outlined lead to full knowledge of theweight and composition of each of the materials involved, and it ispossible to draw up mass balances for Brix, pol, fibre, etc., asdesired. Certain transitions from one material to another havebeen based on pol where Brix might have been used instead. Thereason is that the main goal is considered to be a pol balance; theconsequence is that the Brix may not, and probably will not, balanceexactly but the error should be tolerable. If, for a particular purpose,

it is desired that the Brix balance be exact, then let Brix be usedinstead of pol for the transitions.

Having pursued the subject so far, the reader should not needto be instructed in the derivation of such obvious quantities asmixed juice per cent cane or pol in bagasse per cent fibre; howevera few terms are worthy of some explanation.

Absolute juice, as explained in Chapter II is that part of thecane which is not fibre. Ifbc is the Brix per unit cane, then the Brixper unit absolute juice is bc (1 fc); likewise the pol is pc

(1 fc). These are used to find quantities of absolute juice in othermaterials, usually on a Brix basis, but optionally on a pol basis. Thequantity of absolute juice relative to fibre in bagasse is an importantcriterion of milling performance.

For reporting purposes the net added water is called Imbibition,which explains the symbol I. Part of the imbibition water addedappears in the mixed juice, and this part is known as Dilution;the rest emerges as Imbibition Water in Bagasse. For the purposes ofdetermining the division of the imbibition it is assumed that the

original juices extracted into the mixed juice were the same as thoseleft in the bagasse; both are treated as absolute juice, and calculationhas traditionally been on a Brix basis. However, the milling performance figures which will be recommended for reporting are on apol basis, and such a basis might well be adopted here.

Since the absolute juice in the mixed juice is regarded as beingidentical with the absolute juice of the cane, it follows that theextraction of absolute juice equals the pol extraction. The quantityof clean mixed juice is known from the mass balance; the quantityof absolute juice therein is the quantity of absolute juice in cane

multiplied by the pol extraction (unit basis). The remainder of themixed juice is the dilution, which can then be expressed relative tocane or absolute juice in cane.

21


22/87

Th at qua ntit y of abs olu te juice no t accou nte d for in the mixedjuice is in the bagasse, and the dif ference between this absolutejuice and the absolute residual juice in the bagasse is counted asimbi bitio n wate r in baga sse a figure of doubt ful accurac y. T heas su mp ti on tha t the origi nal juice left in the bagas se has the sa me

pol as absolute juice in the cane may be expected to be appreciablyin error.

Figures Used for Judging Milling Results.

Extraction (sucrose, pol, Brix, juice) is purely a quantitativest ate me nt of fact. If all canes were of one com po sit io n then ex trac tionwo ul d also be a figure of me rit b ut cane is not uni for m, an d tothe extent that it departs from uniformity so extraction becomesdeficient as an index o f the per for ma nce ac hiev ed.

Recognizing this, technologists have sought criteria whichwoul d take acco unt of variati ons in the com pos iti on of the cane andallow for them in assessing milling results. The three significantvari able s in cane (the only thre e th at can be consi dere d) are fibr e,Brix and pol. The various criteria which have been proposed forindicating milling efficiency differ fundamentally in the assumptionsmade regarding the effects of these variables.

It is a feature of all the crite ria that , eithe r in th e first ins tan ce orentirely, they regard milling efficiency as independent of fibre inca ne ; in oth er word s, a t a co ns ta nt ord er of meri t, the loss of po l or

Brix or juice in milling is expected to vary directly with fibre incane. Efficiency is judged either by expressing the loss relative tofibre , or by "r ed uc in g " the loss to wh at it wo ul d ha ve been at astandard fibre in cane.

One may argue that when the fibre in cane rises, a mill grindingat a constant cane rate is operating at a higher fibre rate and maybe expe cted to incu r highe r losses o f po l pe r un it of fib re. Th eargument is sound enough, but it encroaches into the field ofperf orma nce per unit of equ ipm ent , which is be yond the presen tconsiderations. Even so, it is well to keep in mind that milling

performance criteria envisage a constant fibre rate rather than aconstant cane rate.

Whereas in respect of fibre in cane the various criteria arevirtually at one, this is not so as regards pol in cane. Here there aretw o fundamen tal prop osi tio ns one , th at the ratio o f pol to f ibrein bagas se is in de pe nd en t of po l in ca ne : the other, tha t the rati o isdirectly pro po rti on al to pol in cane, or pol in juice.

It is con ven ien t at this poi nt to discuss vario us criteria of millingper formance.

1. Extraction: The word alone normally signifies pol extraction;it is possible to calculate also the extraction of sucrose or Brix orab so lu te jui ce, in whi ch case th e te rm shou ld be suita bly qualified.

22


23/87

The for mul a for calc ulat ion of pol ex trac tion (he re expres sedon a unit basis) is:

but, when a mass balance is taken out, pol extraction is normally

calculated as Pj Pc.Th e last expres sion in the form ula is the com pl em en t of ex

trac tion , that is, the pro po rt io n of pol lost in bagasse. It ma y berendered into the form

rat io of pol to fibre in bagas se

rat io of pol to fibre in cane

Mill engineers in general main ta in th at the respon se of this term ,and therefore the respo nse of extra ction , to th e rati o of pol tofibre in ca ne is a false index of milling pe rfor ma nc e. Ex tra ct ion is astate ment o f fact and an inevitable pr od uc t of qua nti ty accou nti ng,but it is universally acknowledged to be a poor figure of merit.

2. Milling Loss: Milling loss is the r at io of po l to fibre inbagasse, usually expressed as a percentage. It is a simple expressionof the contention that, regardless of the pol and fibre in cane, millingperformance is best when the pol lost per unit of fibre is least. Ithas the advantage of simplicity and the disadvantage that as performance improves it decreases.

3. Whole Reduced Extraction: In a paper presented before theInternational Society in 1962, B. L. Mittal introduced the termWhole Reduced Extraction which he defined as:

W.R.E. (unit basis) = 1 p-"

c

JcThi s express ion pr ob ab ly ha d an eye to the availabi lity of the

data, pol in bagasse per cent cane and fibre per cent cane. Its significance is more readily appreciated when it is converted to the form:

W . R . E . - 1 Pa

This shows that W.R.E. is the complement of Milling Loss.Th ro ug h the reversal of th e sign of the va riab le, th e result rises,with improving performance, towards a limit of 1 (100 per cent).Like milling loss, it igno res the co mpo si tio n of the cane .

4. Extraction Ratio: Extraction ratio is normally defined asth e per cen tag e rat io of (100 ext ract ion) to fibre pe r cent ca ne.Mathematically, on a unit basis:

E.R. _ JL=liJc

23


24/87

This is more readily understood when rendered into the form:

24


25/87

25


26/87

26


27/87

ma ny years indicates tha t they mu st have some elemen t of reali sm.It is suggested tha t the tru th , if the re be any , lies betwe en the two .Th e absolut e juice theo ry depend s up on the ass ump tio n that, if theave rag e juice of Ca ne A ha s twice the Brix or pol of th e jui ce of CaneB, th en the same rati o app lie s to the juic es of the last few cells inbagasse. Tests suggest that there is a trend towards uniformity

(at a level well ab ov e ze ro , by the way) . If th e end resul t were uni formity, then W.R.E. would be the best choice.

In a factory report pol extraction will be either recorded orreadil y avail able by reference to the loss of po l in ba gasse. R. E.(Mit ta l) is easily cal cu lat ed from th e loss of pol in bag asse or fro mW.R.E. It is recommended that the milling performance figures forreporting be Whole Reduced Extraction and Reduced Extraction(Deerr) .

Imbibition

Milling perf orm anc e is so responsive to the pr op or tio n of wat erused in the process that a milling result cannot be properly assessedwithout an accompanying expression of imbibition.

When the actual weight of imbibition water can be measured orded uce d, the intensity of imb ibi tio n is best expressed as part s ofimb ibi tio n wate r per 100 pa rt s of fibre in cane, co mm on ly calledimbibition per cent fibre.

It is sometimes convenient, but somewhat less satisfactory torelate the added water to cane, rather than fibre in cane, hence theterms added water resp. imbibition per cent cane.

Lack of data may make it necessary to relate the added waterto th e weight of a juic e, such as ab sol ute juic e, und ilut ed juice or firstexpressed juice. The calculation requires only the Brixes of theoriginal juice a nd the d ilute d juice (mixed juice) but t he result hassignificant limitation. Technically the "original" juice is the extracted

juice as it would be if undiluted, but this juice is not available, nor isits Bri x; howe ver , any on e of th e th ree juic es specified ab ov e willserve. The second po in t is tha t the mixe d juice does n ot conta in all

the im bi bi tio n wa ter , so me of whic h passes ou t of the milling train inthe bagasse. In an earlier example a final bagasse was found to contain 17.5 per cent imbibition water. This bagasse contained 50 per centfibre so that the imbibition per cent fibre was 35. This is a substantialqu an ti ty to ignor e in a tot al of the o rde r of 200, b ut at least th ediscre pancy is fairly con sta nt. A third poin t is th at the mix ed juic emu st be the un ad ul te ra te d pro du ct of the milling tra in, for if it con tains filtrates or any other additives its Brix no longer reflects thedilution due to imbibition.

Despite its limitations, dilution per cent "undiluted" juice is auseful criterion, certainly worth reporting if imbibition per cent fibreor cane is not available.

27


28/87

Milling Plant Performance.

It is reiterated here that the performance figures discussedrepresent attempts to compensate for variations in the cane not theequipment. The only feature related to plant is the inherent assumption of a constant fibre rate. As most of the mills of a trainare affected more by fibre rate than cane rate, the basis is reasonable,

and accords well with plant capacity formulae which are usuallybased on fibre rates.

28


29/87

CHAPTER IV

Control of the Boiling House

Mixed Juice.Though mixed juice is the raw material of the boiling house,

it is better regarded as an end product of the milling process, sincemost of the interest in mixed juice is related to milling. In theboiling house, the mixed juice is about to be supplemented byfiltrates, limed, perhaps treated with phosphoric acid or sulphurdioxide or special additives, then boiled and settled. In these processes and operations the components of the mixed juice are so thoroughly re-shuffled that the original composition hardly matters.

The weight of mixed juice is known from the milling records,and the juice should be analysed for suspended matter, and poland Brix of filtered mixed juice. The densimetric Brix of mixed juiceas weighed is a false figure that gives rise to an erroneous purity anda purity rise on clarification that is mostly spurious. Previouseditions recommended the determination of reducing sugars andash in mixed juice but these would not seem to provide informationof any use.

Clarification.

In the clarification process pH at various points is kept under

control or observation but the only figure of record is the pH ofclarified juice. The phosphate and starch contents of mixed juicemay be of local significance from time to time, and clarified juice maybe analysed for sulphur, calcium, phosphorus, starch, turbidity andsuspended matter, but the results are not for publication.

It has been customary to report the amount of lime used forclarification, nominally as available calcium oxide per 1,000 parts ofcane. This quantity is responsive to so many features of material,process and conditions that it has only local significance.

The clarification process yields primary mud which nowadaysis usually supplemented by bagacillo and additives and passed tothe rotary vacuum filters. Other types of filter may be used, and thealternative process of extracting sugar by decantation survives to alimited extent, but in any case there is a filter cake or mud leavingthe factory, and filtrates or decant fluids returning to process.

The waste product, which is referred to as filter cake, contains aquantum of sucrose lost from the process, and therefore the weightand sugar content of the filter cake are essential components of thechemical control. Dry substance and bagacillo content may be of

interest in relation to filter performance, and the purity of theliquid component serves to check on deterioration, but only the

29


30/87

qua nti tie s of cake and pol in cake ar e of exter nal interest. T he re isa techni cal poi nt th at s om e of th e loss of po l in filter cak e, bei ngassociated with its bagacillo content, has already been accounted foras pol in bagasse. It is possible to make allowance for this, but nottoo precisely, and , as the err or is of the ord er of 10 pe r cen t of th epol in cake, it is usually ignored.

Clarified Juice.

F r o m a chemi cal poi nt o f view, clarified jui ce is the r aw mat eri alof the boiling process, and, whe re mixed juice is not weighed,some times the analysis o f clarified jui ce is used as th e analysis ofmixed juice for milling control. Its Brix is important in judging theperf orman ce of the evapo rat ors , and, as ment ion ed earlier, it maybe analysed and tested extensively, but only its pH and purity arecommonly reported. Its weight is usually determined from theweight of mix ed ju ice , with allow anc e for th e pol lost in filter cak e.

Syrup.

Syrup is a comparatively unimportant intermediate product.Its Brix is im po rt an t in reference to the per fo rma nc e of th e eva por ato rs an d the provision of suitab le ma ter ia l for the p a n s ; its pur it yis of inter est as the star tin g level for th e suga r boi ling proce ss, a ndthese two items, Brix and purity are normally reported. It is notnor mal ly weighed, a nd its weight is rarely of con cer n.

Pan Products.

Var ious grades o f massecu ites an d mola sses are invo lved inthe sugar boiling system, and their purities and Brixes are undercon sta nt o bserv ation or con tro l in the interests of the process. Afull repo rt will norm all y include the average puri ty o f each grad eof massecuite an d molasses. Th e puri ty of m ag ma s hou ld also bereported.

Sugars.

Several grades of ship ment sug ar ma y be tur ne d out, a nd eachshould be accounted for separately as to weight and analysis. Thest an da rd da ta for a suga r are po l (corr ected to 20 C) an d wat er.Optional extras are reducing sugars, ash and other organic matter.Sugars may also be tested for grain size, starch content, filterability,colour, etc., but such data are not usually published.

Suga rs retur ne d to proces s in the factory are of in ter nal interes tonly.

Final Molasses.

As final molasses contains one of the major losses of sucrose

in process, its wr

eight and analysis are important. The weight shouldbe determined directly, and if this is not possible, a densimetric basis

30


31/87

must be adopted. Pol should be determined for the pol balance,and apparent purity for the general record. As the apparent purityof final molasses may be very deceptive, true purity should also bedetermined and reported, together with reducing sugars, ash andsucrose, so that exhaustion formulae may be applied.

Recoveries and Performances.Introduction. - T h a t pro po rt io n or percent age of the pol in

cane which passes in to the mixed juice is referred to as the ext ract ion .Th at pro po rt io n of the pol i n mixed juice which passes into the sugarmanufactured is referred to as boiling house recovery. The productof the two represents the prop or ti on of pol in cane "r eco ve red " aspol in sugar, and is known as overall recovery. Like extraction,these two recoveries are purely quantitative statements and do notnecessar ily cons ti tu te measur es of efficiency. Let the puri ty of th emixed juice decline, the boiling house recovery will normally do the

same when the standard of performance in terms of merit remainsunchanged.

In the case of extr acti on, th ere was only one vari abl e to considerthe cane. The counterpart in respect of recovery is the mixed

juice; but there is another variable to consider alsothe qualityof the sugar pro du ce d. Po ur som e of the fina l mola sses over thesugar before the latter is weighed and analysed and the recovery willrise, but the pe rfo rma nc e of the factory has certainly not imp rov ed.Hen ce, in att emp ting to assess the merit s of a recover y figure, on ehas to take account of both the mixed juice and the sugar.

Actual recoveries, standard recoveries and various performancefigures to be discussed are almost invariably expressed as percentages,but it simplifies mathematical expressions and derivations greatlyif the term s are referred to the basis of un it y ra th er th an 100. Th eformer basis has been adopted.

All the terms can be, and ideally should be, based on trueanalysessucrose and dry substancebut pol and Brix are acceptedas the working standards and the terms, unqualified, are takento be based on these apparent measures.

Actual Recoveries:

Boiling House Recovery.As previously stated, Boiling HouseReco ver y is pol in sugar as a pr op or ti on of po l in mixed juic e.

Overall Recovery.As previously stated, Overall Recovery ispol in suga r as a pr op or ti on of po l in cane.

Standard Recoveries.

The S-J-M Formula.When a raw material containing sucrose

and impurities is processed into a final product, rich in sucrosean d a waste pr od uc t con tai nin g mo st of the impurities, all co mponents being accounted for, there is a mathematical relationship

31


32/87

between the materials and their components. The most familiarexpres sion of this is the s-j-m form ula o f Noel D eerr , whic h picturesa no tional juic e being proce ssed into su gar and molasses only. Ifth e puri ties o f the materia ls respectively be j, s and m, the n therecover y, r, i.e., th at pro po rt io n of the sucrose in the juice which iscontained in the sugar is determined by the formula:

j (s m)

This is mathematically true when s, j and m are literally truean d the re is no loss of any mater ial . W hen ap pa re nt values are usedfor s, j an d m, the fo rmula is no long er cor rec t in gene ral. T he largedispa rity between the true an d appar en t purities of final molassesmig ht suggest th at the use of ap pa re nt purit y wou ld yield ab su rdresults, bu t the re is a significant mea su re of comp ensa ti on . Thedifference bet ween th e two puri tie s for mix ed jui ce is appr eci ab ly less

th an for mola sses, b ut the form ula is mu ch more sensitive to j th anto m. A difference of 3 unit s of pu ri ty in jui ce is equ ivalent to ab out7 units of pur ity in mola sses, a nd these figures are not to o remo tefrom the real differences between true and apparent purities in thetwo cases.

Given values of s, j, m and r from the records, and adoptingthe s-j-m formula, it is possible to compare the actual recovery withthe idea l figure. Th e forme r result will be the lower beca use of th ekn ow n and unk no wn losses of sucrose other th an in molasses. Th erat io of the tw o recoveries wou ld be a figure of meri t, ta king accoun t

of the pur ity of mixed juic e, s how ing up losses othe r th an in molasses ,but accepting the purities of sugar and molasses at their actualvalues.

The Winter Formula.Before 1900, Winter and Carp indepe nden tly con clud ed that the yield of comm erci al sugar to beexpected from a mixed juice could be predicted by deducting fromthe pol in the mixed juice 40 parts for every 100 parts of impurities.This is expressible in the form:

r = 1.4- 0.4

Jwher e r was , originally, the recovery of comme rci al suga r perunit of pol in mixed juice of purity j (unit basis).

When the s-j-m formula was devised it was soon recognizedthat the s-j-m formula would yield the same values of r when s wasfixed at 1 (100 per cent) an d m at 0.2857 (28.57 per cen t). T hefor mul a of Win ter an d Car p, generally called the Win ter formu lano wa da ys , is generally regard ed as a special case of th e s-j-m fo rm ul a,but this is mathematical rather than historical.

Regar dles s of its origin, the Win ter for mul a is co mm on ly usedto provide a standard recovery for the boiling house, designatedBasic Boiling Ho us e Recovery . It recognizes th e pu rit y of th e

32


33/87

mixed juice, it allows no losses of pol except in molasses, it providesfor a molasses of 28.57 purity, and it takes no account of the purityof the sugar made.

Equivalent Standard Granulated Sugar.It has been mentionedthat the quality of the sugar is a factor in recovery. Not all of thesucrose in a commercial sugar is truly "recovered" because, in the

ultimate removal of the remaining impurities, some sucrose willinevitably be lost. Hence the sucrose content of any commercialsugar has to be discounted for potential loss if recovery is to betruly assessed.

Many standards of comparison of sugars, e.g., raw value, nettitre, titrage, and many standard sugars, e.g., 96 degree, StandardMuscovadohave been or are used in practice, but most of themhave a local and usually a commercial significance.

Noel Deerr suggested that the sugar itself was as good a subject

for the assessment of a process recovery as any other factory product.He proposed the use of the Winter formula, and to the materialexpected to be recovered he gave the name Equivalent StandardGranulated, usually abbreviated to E.S.G. The Winter recovery,being a recovery of pol, is applied to the pol of the sugar to yieldthe E.S.G. factor. Hence:

pol of sugar x Winter recovery = E.S.G. factortons sugar x E.S.G. factor = tons E.S.G.

The actual recoveries referred to earlier are of pol in actual

sugar; the Basic Boiling House Recovery is of pol in pure sugar. Forcomparison purposes either the actual recoveries might be adjustedto pure sugar, or the basic recovery to actual sugar.

Conventionally the first choice is adopted; comparisons aremade in terms of pure sugar, that is, E.S.G.

Criteria of Performance.

Boiling House Efficiency.The ratio of Actual B.H.R. toBasic B.H.R. is frequently worked out (as a percentage) and reported

as Boiling House Efficiency. As a criterion it compensates for thepurity of the mixed juice and it responds to losses, but it takes noaccount of the quality of the sugar produced, and it postulates astandard purity of 28.57 for final molasses.

Boiling House Performance.The neglect of sugar quality inB.H.E. can be rectified by expressing the actual recovery as E.S.G.instead of pol. The result can then be matched against Basic B.H.R.which, being a recovery of pure pol, may legitimately be entitledBasic B.H.R.E.S.G. Thus B.H.P. is the ratio (usually as a percentage)of Actual B.H.R.E.S.G. to Basic B.H.R.E.S.G.

By previous specifications, the term B.H.P. was to be associatedonly with sucrose and gravity purities. No name was specified for

33


34/87

34


35/87

35


36/87

36


37/87

of boili ng ho us e per fo rma nce would ap pe ar to be the ratio ofR. B. H. R. E. S. G. to Basic B. H.R. (s 100, j 85, m expect edpurity). For the pol basis of reporting, the expected purity wouldneed to be adjusted from true pur ity to app are nt puri ty accor dingto the prevailing local difference.

Overall Performance.- Wit h the perf or man ce of the milling

sta tio n expressed in one figure, and the per fo rma nce of the boili ngho use expressed in anot he r, it was natur al to think of co mbin in gthe two into one overall performance figure for the factory. Thiswas done originally by combining the two Noel Deerr criteria,R.E. and R.B.H.R.E.S.G. as Reduced Overall Recovery E.S.G.

In view of the preference for G u nd u Ra o 's recovery formulaReduced Overall Recovery E.S.G. may now be defined as the percentage product of R.E. (Deerr) and R.B.H.R.E.S.G. (GunduRa o) . It is no t as significant a term as mig ht be expected a t firstthought, because the milling department and the boiling house

arc so distinct tha t the expression o f their per for mances toge therin a single result gives little satisfaction.

In the consideration of overall performance there is a strongtendency to go back to a cane basis. The Overall Recovery E.S.G.multiplied by pol per unit cane gives the Yield of E.S.G. and theReduced Overall Recovery E.S.G. treated similarly gives ReducedYield of E.S.G. These arc not performance criteria at all, but,as actual or corrected yields of pure sucrose they are commonlyma tc he d against expected yields and so efficiency is ju dg ed . Tech

nically the venture is bold because the meagre data available inrespect of the cane are inadequate for the prediction, within reasonable limits, of a sta nd ar d yield; economicall y the com pa ri so n is wellan d truly justified when the price of the cane pu rcha sed is basedupon the projected yield of sucrose. Various formulae have beendevised an d are in regu lar use for the derivat ion of a stand ar d yieldof sucros e from cane ac cor din g to the comp osi tio n o f the cane,but these are mainly commercial formulae and are not recognizedfor technical evaluation of perfo rmance s.

Recapitulation.In the considerat ion of ter ms to be ad op te d

to express the performance of a sugar factory, for the purposes ofInternational comparisons, it is necessary to subordinate precisionin detail to the wide range of con dit ions to be catered for. V ari ouscriteria discussed in previous editions have been ignored here, notso mu c h on the gro un ds of lack of mer it as on the cont en ti on thatthey belong in a different fieldthe co nti nu ing study of the perfor man ce of on e factory. This applies part icul arly to the ma nyformulae based on detailed accounting for impurities.

Given as da ta the pu rit y of mixed juice, t he pur ity of suga rand the actual boiling house recovery, it does not seem possibleto derive any bette r criterion of th e wor k of the boiling house tha nR.B.H.R.E.S.G. (Gundu Rao). The most obvious avenue for im-

37


38/87

pr oveme nt lies in using the com pos iti on of the molass es to nomina tea target purity for that material, and using that purity to providea valu e for m in the formula, i nstead of the tr ad it iona l an d indis criminate figure of 28.57.

There is probably no need to depart from the Winter formulafor the pur po se of det erm ining E.S.G. The effect of a chan ge of

molasses purity is very slight, the relationships between impuritiesin the sugar are not necessarily the same as in the final molasses,and one can find some virtue in a common formula for all sugars.

38


39/87

CHAPTER V

Methods of Weighing and Measuring

Chemical control necessarily involves the determination ofweights of material, either directly or inferentially. The variationsin layout and procedure within factories and the range of equipmentavailable make it necessary to restrict the discussion of weighingand measuring to general principles of operation and types ratherthan specific brands or equipment.

Weights and Measures.

Whilst the universal adoption of an international decimalsystem of weights and measures is the ultimate goal, the fact remains

that the local systems are strongly entrenched and most of themare not likely to be supplanted in the foreseeable future.

Fortunately the great majority of data in a record of chemicalcontrol are relative within themselves and thus independent ofunits. International trade and communications have fostered theadoption of one or other of the major systems of units in preferenceto minor local systems in many countries, and the cane sugarworld may be said to be divided between the British and the metricsystems, with minor local variations. The adoption of either oneof these would appear to be the most that one can ask for at this

juncture. The tendency to express parts of a major unit in decimalsis developing and is to be encouraged, for this is a positive steptowards ultimate uniformity.

Weight of Cane.

Cane is almost invariably weighed on a platform weighbridgedesigned to accommodate the vehicle by which a load of cane iscarried.

As a weighing machine used for trade purposes, the cane

weigher often comes under the jurisdiction of the Authority controlling weights and measures within the Country, and, if so, calibration and certification are performed by that Authority. If not,the procedures for testing, which arc too lengthy to be set out here,are readily available from any recognized testing Authority. Theoperation of testing and calibrating is usually performed annually.

Certain routine checks are called for. The practice of testing thezero regularly should be observed; this is readily achieved bymaking it the first task of each weighbridge operator coming on

duty. He should also check the tare counterweights for identity andposition. There should be a mobile test weight on the premises,weighing about the same as the average commercial load. Correct

39


40/87

rec ord ing of th e weight of the test load sho uld be verified at leastweekly, and for preference, daily.

Cane normally loses weight from the time it is harvested untilit is crushed, and therefore it is desirable to minimise the storageof weighed cane. Ap ar t from the no rma l loss by eva porat ion, the remay be a significant gain if rain falls on the stored cane; the effect

in either case is to create an uncertain difference between the caneas weighed and the cane as milled and analysed.

It goes wi tho ut saying th at the accuracy o f the net weigh tthe weight of ca ne -- is no better tha n the accurac y of the tar e of thetransport vehicle. Circumstances vary so widely that only a generalword of caution is app rop ria te.

Weight of Field Trash.

Th e total a mo unt of fiel d tra sh canno t be weighed directlyan d must be estimated by d etermin ing the pro po rt io n of fieldtr ash in represe ntat ive samples of cane as received. Sa mpli ng isno t easy because the distri but ion of tras h is far from un if or m; thisapplies particularly to soil. Undoubtedly the best sample unit is awhole car of can e, bu t the stri ppi ng of such a qu an ti ty is a formid able task. It is more common practice to resort to partial unloading,aiming as far as possible to leave a section undisturbed. A convenient residual weight is from 100 to 200 lb, 50 to 100 kg. Thesam ple is str ipp ed and the original weight acco un ted for as fieldtrash and clean cane. In order to gain an acceptable average forthe whole supply this operation should be carried out not less thantwice every shift.

Th e determi nati on of field trash is no t an absolut e meas urefor there is no clear line of separation between clean cane andtr ash , par ticu lar ly at the top of the stalk. A te am instructed instandard procedures and arbitrary working rules can turn outconsi stent result s, but all repo rted meas ure s of field tra sh haveto be accepted with reservations.

Weight of Mixed Juice.

As the factory control is largely based on mixed juice, particular care must be taken to ascertain its weight correctly. It isnormal to weigh the mixed juice as expressed, filtrates, lime oradditi ves being in tro duced after the weighing . If any of these mus tbe add ed before the weighing, the weight o f ad de d mat eri al mustbe determined and allowed for.

Va riou s types of scale are ava ilable for the weighing of mixedjuice (or other liquids). They are al l batch weighers, mostly of thebeam balance type, but the hydrostatic pressure principle is alsoused.

The original juice weighers were essentially steelyard scalesan d were manu all y oper ate d. A ta nk of juice was isolat ed, weighe d,

40


41/87

emp tie d an d tar ed , and the cycle rep eat ed. Ma n y of thes e hav e beenautomated; the filling is halted at a predetermined gross weight,and the emptying is stopped at a set minimum weight or the tankis allowed to drain to an expected constant tare.

An ot he r po pu la r scale uses the principle of the unst abl e benta rm bal ance . Wh en the scale ta nk fill s to a critical weight, th e be am

tips . Thi s actio n cuts off th e inp ut of juic e and o pen s the outl et.When the weight reaches a critical minimum the beam tips backto the original position and the control valves revert to the fillingcondit ion.

The tw o types of scale ju st m en tioned a tt em pt to fill to a constant weight, empty to a constant weight and so discharge a constantweight each cycle. Their main limitation is that, at the momentwhen the inlet and discharge streams, respectively, are interrupted,som e juice is in trans it and , to the extent that this ma y vary , so ma ythe actual weight disc harg ed. Because of this limi tatio n, win d an dtemperature effects and mechanical defects, constant weight scalesdo not discharge a weight that can be accepted as constant over longperiods.

An independent tank scale is to be regarded as a necessarycompanion to constant weight scales. The tank scale should be solocated that one or more batches from the process scales may bediverted into it for check weighing. Thus the actual discharge pertip is determined. This should be checked every shift.

An ot he r ba tch weigher of later years is s oun der in princip le.

It is an automatic steelyard type which fills the tank approximatelyto a selected weight and then sets the counter weight to balance thescale. This is the zero position. The tank is discharged down appro xim ate ly to a selected weight an d the co un te r weight the n move sto the new po sit ion of bal an ce. T he integral of the mo ve me nt of th eco un te r weight is a tru e me as ur e of the weight discha rged .

Thi s is a par tic ula rly goo d type of juic e weigher, so goo d th at ,when it is operating properly it needs no check scale; but like allmachines it can develop leaks and mechanical faults. The per

for ma nce as to weighing a nd the int egr ati on of tota ls can be che cke dreadil y with the aid of built-in ch eck weights . L ea ky valves an dfroth overflows crea te inco rrec t results, bu t faults of this na tu reshould be observed and corrected promptly in any case. The provision of a check scale in this instance may be classified as desirablebut not essential.

Th e principle of weighing by pn eu ma ti c balanc ing of thehy dr os ta ti c pre ssure at the b ase of a tan k of jui ce is accep ted asso und an d reliable eno ugh . Successful appli cati on of the theo rydemands the following:

1. Th e juice tan k mu st be of uni form cross sectional area ,to p to bot to m. Th is is a ma tt er of con str ucti on.

41


42/87

2. The cross section must remain uniform on a time basis.This is mostly a mat te r of cleanliness .

3. Th e juic e mu st be of un ifo rm density with in eac h ba tc h.Errors due to this cause are generally random and thereforenegligible in the total of ma ny bat che s.

Juice weighers, other than those which integrate totals, haveto be provided with batch counters. Mechanical counters arestrang ely unre liab le, a nd the prov isio n of tw o does not help, forwhic h of two different results is cor re ct ? Pr ob ab ly the best co un te ris a pre ssure differential rec order re sp on di ng to the level of jui cein the weigh tank because it pro duc es do cu me nt ar y evidence of itsfunctioning. A mechanical counter can serve as a stand-by.

Th e dete rm in at ion o f the weight o f mixed juice from its vol um eand density has been recognised in the past, with some misgivings.It is doubtful whether the practice survives, and if it does, it should

be abandoned.

Weight of Clarified Juice:

If the mixed juice cannot be weighed the clarified juice may beweighed instead and the notes on weighing mixed juice will apply.Oth erwi se clarified juic e is not n or ma ll y weighed , its weight bei ngdete rmin ed from the weight and the pr op or ti on of pol that it contains.

Weight of Bagasse.

Bagasse has now been successfully weighed using severaltypes of equipment. One of the most convenient devices is thecontinuous belt type of weigher, using beam balance or load celldetection of the weight. The bagasse is transported on a rubberbelt. Because of the low loa din g per foot, the su spen ded section of theconveyor has to be relatively long, and the relatively high proportio n of the weigh t of the belt in the tot al mak es it necess ary to ke epthe tare correction under close supervision.

A convenient device for calibration is a wheeled trolley of

known weight placed on the belt over the suspended section, andtethered by a string parallel to the belt. In operation the weigherrapid ly acquires a ma ntl e of bagacillo , an d the calibr ation at theend of th e week will differ from tha t at th e beg inn ing wh en th eunit was clean. It is necessary to judge by observation how long thebagacillo takes to build up to the angle of repose, and proportionthe calibration factors accordingly.

A du mp in g ty pe of ba tc h weigher has been use d successfully.Naturally it needs generous doors, but they should not be snapacti ng, bec ause th e sud den plu nge of a great wad of bag asse in toa bin creates a back rush of air that picks up clouds of bagacillowhich can create a serious nuisance.

42


43/87

A rad iat ion typ e of bagasse weigher has been used in Ha wai ifor som e years . It ope rat es by virt ue of the fact th at the abso rp ti onof gamma radiation by bagasse is reasonably constant per unit weightover the ra nge of comp osi tio ns en coun tered in practice. Hence theinteg ral of the abs or pt ion by a str eam of bagasse is pr op or ti on al tothe weight pass ing the source of rad iat ion. App are nt ly the accu racyof the system is acc ept abl e.

Weight of Imbibition Water.

Th e weight of imbibitio n wat er has traditio nally been rega rdedas an item for precise determination, despite the rather obviousweakness of the mas s balan ce me th od of deriving the weight ofbagasse.

It is time to face up to the fact that such a proportion of theimbibition water will be lost by evaporation that it is not worthdetermining the initial quantity with great accuracy.

Scales,where they exist, will continue to provide a reliableweight, but it is not worth installing scales for the purpose. A goodwater meter will provide acceptable results, subject to the qualification that any water meter must be checked regularly.

W'eight of Filter Cake.

When filter cake is removed promptly from a factory withoutany significant change in composition, it may be convenient toweigh it in loading hoppers or transport vehicles. When this is notpracticable, or when the composition is promptly altered, e.g. byre-pulping with water, it is necessary to resort to sampling methods.

As filtration takes place usually on a well defined area of flator gently curved surface the general principle is to weigh samplesfrom a convenient fixed area and expand the average weight to theto ta l effective area for th e pe riod of time involved. In the case offilt er presses the uni t for weighing ma y be the con ten ts of on eframe or a section iso lated by a cu tt ing frame. In the case of ro ta ryfilters, the cake from one frame may be transferred to a metal tray

an d weighed; th e nu mb er of frames on the filt er and the nu mb er ofrevolutions for the period together provide the amplification factor.

Unfortunately there are other situations not so easily dealt with.There are rotary filters not filtering on frames, and there are othertypes of pressure or va cu um filter in which the " a r ea " of cake isvariable. W

7here no convenient or reliable unit of area or plant is

available, it is necessary to assess the total as well as possible bywhatever means can be devised. Re-pulped muds may be weighedan d analysed as such , for th e real inter est is more in th e loss of po lth an in the weight of filt er cak e. Th e de ma nd on accu racy is nothigh, but it is still much better to measure the mud loss than to estimate it.

43


44/87

Weight of Final Molasses.

Subject to allowances for slow filling and emptying, liquidweig hing scales in general a re sui tab le for th e wei ghi ng of finalmolasses; those which depend on a constant tare draining must beexcepted.

In some cases the final molasses is shipped promptly in tankcars and, by weighing the quantity so shipped, a reliable record ofproduction may be provided.

When the final molasses is stored in bulk tanks at the factoryit is usually possible to arrange for it to be passed through a weigher,or, failing tha t, thr ou gh measur ing ta nk s of kn own capac ity. Asatisfactory mea su re of aver age density may be ob tai ned by regula rlyweighin g samp les of kn ow n vol ume . Fo r furth er details refer toWeights of materials in Process.-

Final molasses is commonly diluted to a standard density forvarious reasons. It matters little: whether the quantity is measuredbefore or after.dilution, but it goes without saying that, for chemicalcontrol purposes, the t material measured and the material analysedmu st be the same.

Weight of Sugar.

In the past, the commercial sugar manufactured was almostinvariably stored and transported in sacks. Where this practiceconti nues , an d shipmen t is pro mp t, the best mea sur e of the weightof sugar is provided by weighing the loads of sugar leaving thefactory.

If to o mu c h ha s to be held i n store , the weight of suga r m ad ein a period has to be calculated from the number of sacks filled andthe average net weight per sack. In such a case the procedure is:

(1) by automatic, or controlled manual filling, to keep theweight of sug ar per sack as nearly uni for m as poss ible, a nd

(2) to check weigh a sufficient num ber of sacks on a pl at fo rm

scale reserved for the purpose. The number of sacks to bechecked dep end s upo n the sack-to-sac k vari ati on, an d is amatter for local determination.

In all cases where sacks are used , th e weight of th e sacks mus tbe deducted from the gross weight. Data for the correction are bestprovided by weighing a batch of 50 to 100 sacks from currentstocks.

Most commercial sugar nowadays is stored and transportedin bulk. Between the sugar conditioning equipment and the storageor loading facilities room can be made for a sugar weigher. Batchtypes and continuous types of scale are both in use and have bothbeen found satisfactory.

44


45/87

Because of the high pol c ont ent of com mer cia l sugar, its weightis of mo re significance in th e pol ba lan ce th an t ha t of any of theother materials. However, the general tendency for storage to beminimized at factories and concentrated at communal warehouses,where first class weighing facilities are installed, means that anauthentic check on factory weights is usually available at shortdelay . Wh en no such check is for thc om ing the weighi ng of com mercial sugar at the factory must be treated with the importancethat it merits.

The Weights ofMaterials in Process.

Wh en a po l or Brix bal anc e is to be ta ke n ou t for a pe rio d dur in gth e cur ren cy of a sea son, it is necessary to ta ke ac cou nt of thequa nt iti es of pol or Brix in stoc ks at the begin nin g an d the end ofthe period. Subsider juice, m ud , syrup an d a ran ge of massecu itesand molasses may be involved.

Th e mat eria ls are seldom, if ever, weighed, th e weigh t of ea chbeing usually calculated from the volume and the nominal oractually measured density.

Mas sec uite s an d mola sse s of the lower gra de are liable to carryocc lude d gases and a layer of foam, a nd pr op er estim ati ons of vo lum eand density are not made easily. The pneumercator, which measuresthe static pressure at the base of a tank, is not affected by bubblesin the liquid, but, except in special cases, the pneumercator can beused only in tan ks of uniform ho riz ont al sectional area.

When it is necessary to measure volume and density in a foamymaterial the following expedients may be adopted. To find thedepth of liquid, take a tube large enough for the depth gauge topass through it, and fit, at one end, a plug which may be dislodgedfrom inside. Immerse the sealed end until it is below the foamlayer and dislodge the plug. The liquid rises to the equivalent trueliquid level and this may be measured with the depth gauge.

Fo r the meas ure me nt of density there sh ould be available avessel of op ti on al sha pe, norm al ly coni cal, with a sh ort cy lindricalneck. A cap aci ty of ab ou t 250 ml is suitable . By pre vio us tests

with water filling the vessel to a flat free surface at various temperatures, the temperaturevolume relationships of the vessel shouldbe determined. The vessel is then filled with the test liquid at thetem per atur e of the bulk, the temper atur e being determin ed andrec ord ed. Th e weight of th e co nte nt s and the volu me of th e vesselat the recorded temperature yield the density at that temperature.

If the ma ter ial is heavily aera ted , its den sity will var y fromtop to bottom in a non-linear gradient. The density at middle depthis not average. The best policy is then to use a long tube to findthe free level, deeply immersed so that it fills from near the bottom.Then take the density sample from the tube or from near the bottomof the tank.

45


46/87

The der iva tio n of dens ity from the Brix of a dilu ted sol utionof th e pr od uc t is not reliable an d will be seriously in e rro r wh enthe product is aerated.

Miscellaneous Materials.

In factories of unusual layout, or for special purposes it may benecessa ry to kn ow the weight of mat eria ls not no rma lly involvedin the chemical control. Only general rules can be laid down. Theacc ura cy soug ht ha s to be related to the im po rt an ce of the qu an ti tymeasured, and, in general, weighing directly is more reliable thandetermining weight inferentially.

Condenser Water.

An at te mp t is ma de by s ome t o acc oun t s epara tely for suga rlost into the condenser water by entrainment from the last effetan d the pa ns . Whilst the meth od s of analysis of conde nser waterfor sugar are ad eq ua te for the dete ction of sugar an d an indi cati on ofconcentration, they are not to be depended upon for much accuracy.If the pr op or ti on of sugar in the wate r is dete rmi ned with satisfaction, it becomes a question how much water is to be allowed for.The quantity can be calculated from evaporation and temperaturedata with great satisfaction to the mathematician and no accuracyto speak of, and it i s bet ter to ask the engin eer ho w mu ch injectionwate r he is pump in g. Th e loss of sugar by en tra inm en t mu st becon sta ntl y wat che d an d kept to a mi ni mu m, but it is no t a figure

for separate recording in the pol balance.

46


47/87

CHAPTER VI

Methods of Sampling

It is a truism worth repeating that no analytical result can beany more reliable than the sample from which it was derived. Itshould be the constant care of the laboratory staff to ensure thatthe samples presented for analysis were taken properly.

When a process is continuous the sampling of a materialinvolved should ideally be continuous at a rate proportional to therate of flow of the material. Practical considerations often enforcea departure from this principle because the rate of flow of thesample stream would necessarily be too low for reliability andregulation. Flow splitting is unreliable and limited in scope when

the liquid to be sampled contains suspended solids. In such a caseit is better in practice to establish a reliable continuous flow ofthe liquid as a sub-sample, and then to regulate the quantity takenfor the actual sample by diverting the stream into the samplereceiver at intervals in accordance with a regular time cycle.

When the material to be sampled does not lend itself to flowsplitting at all, sampling must necessarily be by way of a series of"grab'' or "snap" samples. Batch operations naturally call forbatch samples.

In so far as sampling can be made automatic, it should be,but every automatic sampler requires regular inspection, maintenance and cleaning. Manual sampling is as reliable as the personnelwho carry it out. It is used extensively in practice and is quiteacceptable so long as the sampling personnel are reminded regularlythat their work matters.

There is a natural tendency to regard the process materials ofthe sugar factory as very variable in composition and the truth ofthis is easily demonstrated. It is therefore important to appreciatethat the variations are mostly frequent, random and of limited

amplitude, and that, on a broader scale, the process materials arecharacterised by a high degree of uniformity of composition.

The proof of this is to be found in the incredibly small samplingratios which are found to yield acceptable average results for thefactory. A mere 30 ten-pound samples have been found to givereasonable representation of 30,000 tons of cane. That is a sampleratio of 1 in 200,000. In Queensland, a minimum of six determina-tionsof fibre in cane, each starting from 12stalks of cane, can betakento provide the average fibre in cane for a week's supply of 20,000 to30,000 tons of cane. The sampling ratio is even less than 1 in 200,000.

These instances are to be regarded only as illustrations of theinherent uniformity of cane on a large scale.

47


48/87

Cane.

Cane has been sampled for many years for the determinationof field tr as h and fibre con te nt . Field tr ash ha s been deal t with inCh ap te r V. Fo r the de ter min ati on of fibre in cane it is des irab leth at the can e be samp led in the pr ep ar ed sta te, an d if this ca nn otbe do ne , " c o r e " sampl es of the cane are the next prefere nce. The

last resort is to stalk samples.An old establish ed pr oc ed ur e of sam pli ng cane, by stal ks, for

fibre determination is as follows. At least twice, preferably threetimes, a shift for thr ee shifts select at ra n do m the parc el of ca neto

system of cane sugar factory control - 3rd ed

Documents