a new system for continuous sterilization of...

145

A NEW SYSTEM FOR CONTINUOUS STERILIZATION OF OIL PALM FRESH FRUIT BUNCHES

A NEW SYSTEM FOR CONTINUOUSSTERILIZATION OF OIL PALM FRESH

FRUIT BUNCHESK SIVASOTHY*; ROHAYA MOHD HALIM* and YUSOF BASIRON*

* Malaysian Palm Oil Board,P. O. Box 10620,50720 Kuala Lumpur,Malaysia.E-mail: [email protected]

Keywords: bunch crushing, sterilization, continuous sterilization, continuous processing, mill automation.

Date received: 21 June 2004; Sent for revision: 14 August 2004; Received in final form: 31 May 2005; Accepted: 24 November 2005.

ABSTRACT

A system is proposed for continuous sterilization of oil palm fresh fruit bunches based on crushing bunches to

facilitate steam penetration into bunches and subsequently heating the bunches using live steam at atmospheric

pressure to an extent sufficient to facilitate stripping of the fruits. An attempt is made to evaluate this system

using a pilot-scale set-up.

INTRODUCTION

In the conventional milling process, fresh fruitbunches are loaded in cages and pushed intosterilizers where they are cooked for about 70 to 90min using steam at 40 psig. The process arrests oilquality deterioration due to enzymatic activity. It alsofacilitates the stripping of fruits from the bunchesand the extraction of oil and kernel (MonganaReport, 1955). Operations related to the batchsterilization process absorb much of the processlabour in a typical mill. Continuous sterilization haseluded the palm oil industry despite the manyattempts made over the past five decades. The needto use steam at high pressure with intermittentpressure releases to achieve good sterilizationcomplicates the problem of achieving continuousprocessing.

Exploratory studies (Sivasothy and Rohaya, 2000;Sivasothy et al., 2000) on continuous sterilization ofoil palm fresh fruit bunches, carried out using adouble-roll bunch crusher and a laboratory-scalebatch steam autoclave, showed that the extent of nutbreakage by the crusher was small and that therewas no significant deterioration in the oil quality dueto bruising of the fruit mesocarp if the bunches areheated immediately after crushing. The sterilizationtime and steam pressure required to achievecomplete stripping of fruits from the crushed

bunches (Figure 1) were significantly lower thanthose normally used for uncrushed bunches.

The main aim of this study was to examine a newconcept for continuous sterilization. The studyallowed us to gauge the extent to which the resultsachieved by heating bunches using steam atatmospheric pressure in a laboratory-scale batchautoclave during exploratory studies werereproducible in a pilot plant based on the proposedconcept. Since the concept is new, it was essential todemonstrate that it works on a pilot-scale beforescaling up to a commercial system.

The pilot plant study also provided insight intothe problems associated with integrating the newsterilization process with the rest of the millingprocess.

PROPOSED SYSTEM

The ability to sterilize bunches using steam at lowpressure and without the use of multiple-peak cycles,facilitates the development of a continuoussterilization process. The problem of continuouslytransferring bunches to and from the sterilizer cannow be more easily surmounted. Many differentmethods were explored, both for transferringbunches to and from the sterilizer, as well as forconveying the bunches through the sterilizer.

Ideally, the method used for conveying shouldensure continuous and uninterrupted flow of theproduct and ensure consistent product retention timewith little or no short-circuiting. There should alsobe little probability of product accumulating insidethe sterilizer as this is detrimental to the oil and

Journal of Oil Palm Research Vol. 17 December 2005, p. 145-151

146

JOURNAL OF OIL PALM RESEARCH 17 (DECEMBER 2005)

kernel quality. The bunches should also be conveyedwithout being subjected to very high forces as thecooking progresses as the application of pressuremay lead to oil being easily released from the cooked(softened) fruits and lost with the condensate andempty bunch stalks.

The system should also be designed to minimizesteam loss from the feed and discharge openings.This requires both openings to be made as small aspossible while allowing continuous anduninterrupted flow of the product through thesystem.

The system should preferably also be able to copewith fluctuating flow rates and permit changes inthe retention time to cope with varying bunchripeness.

Preferably, the system (both the valve and thecontinuous sterilizer) should be designed to enableheating the bunches using steam at, or slightly above,atmospheric pressure.

Figure 2 illustrates a proposal for a continuoussterilization system that sufficiently addresses theabove considerations.

As in the preliminary studies, disruption of theclosed-knit arrangement of the spikelets in the freshfruit bunches is achieved using a double-roll crusher.This crusher provides many advantages, includingminimal nut breakage, simple and compact design,low investment cost, low power consumption, lowmaintenance and operating cost, and ability tohandle all types of bunches. A single crusher shouldbe able to handle a typical palm oil mill’s entirethroughput.

As shown in Figure 2, the bulk of the heating isachieved inside an enclosed vessel called the

continuous sterilization chamber. The bunches areheated for about 60 min inside this vessel using livesteam while they are conveyed using a double-deckscraper conveyor or some other type of conveyor.By making use of both the forward and return pathsof the scraper conveyor for conveying the bunchesas shown in Figure 2, it is possible to reduce theoverall length of the chamber.

The bunches are pre-heated using steam thatbleeds from the continuous sterilization chamber.This facilitates heating the bunches immediately afterthey are crushed to above 60ºC to deactivate thelipolytic enzymes responsible for the formation offree fatty acids. Pre-heating the bunches alsofacilitates deaeration and minimizes the amount ofair entering the continuous sterilization chamber,thereby ensuring that the temperature in thecontinuous sterilization chamber is close to that ofsaturated steam. An additional pre-heating step maybe used if the delay between crushing and steamheating is enough to cause oil quality deterioration.In this case, the bunches are heated immediately afterthey are crushed using hot water at above 60ºC.

The bunches are further heated after they leavethe continuous sterilization chamber using steamthat bleeds from the discharge at the end of thecontinuous sterilization chamber.

The bunches enter and leave the continuoussterilization chamber through one or more rotaryvalves, flap valves or gate valves to minimize thesteam loss. The base of the feed conveyor can alsobe filled with water to seal against steam loss. Thiswater bath can be heated to above 60ºC to enablesimultaneous deactivation of the lipolytic enzymes.

Figure 1. Relationship between steaming pressure and streaming time to achieve complete stripping of bunches.

147


PILOT PLANT

The continuous sterilization pilot plant set-up(Figures 3 and 4) was specifically designed for heatingusing steam at atmospheric pressure. By making useof both the forward and return paths for conveying,it was possible to achieve a total conveyor length ofonly approximately 20 m.

The crushing was carried out using a double-rollbunch crusher with a fixed clearance of 15 mmbetween the rollers. The speeds of rotation of therollers were 22 and 32 rpm.

A steam-sealed inclined scraper conveyor wasused for conveying the bunches to the continuoussterilization chamber. A simple double-flap door wasused at the entrance of the continuous sterilizationchamber to minimize steam loss. The hinged flapswere kept in the normally closed position by counter-weights and opened by the impact of the bunches.This simple steam-locking mechanism was sufficientto accommodate the low throughput of the pilotplant, which was approximately 1.2 to 2.4 t hr-1. Thebunches were hand-fed to the crusher one at a time

at 30 to 60 s intervals. The steam that escaped throughthe inlet valve was used to pre-heat the bunches inthe feed conveyor.

The crushed bunches were pre-heated for 6 minin the feed conveyor and then a further 56 min inthe continuous sterilization chamber.

Steam was introduced at a number of points atboth decks. The rate at which steam was introducedwas manually adjusted to minimize loss from theentrance and discharge ends. It was observed thatthe temperature in the continuous sterilizationchamber could be quite easily maintained close to100ºC.

The continuous sterilization chamber wasinsulated to minimize heat loss and slightly tilted tofacilitate condensate discharge.

The bunches leaving the continuous sterilizationchamber were held in a hopper and periodicallydischarged by a hydraulically controlled slidingdoor.

The cooked bunches were stripped by a typicaldrum stripper in the mill where this study wascarried out.

Figure 2. Proposed system for continuous sterilization.

148


BUNCH STRIPPABILITY

The strippability of bunches achieved by heatingusing a laboratory-scale batch steam autoclave(Sivasothy et al., 2000) was comparable to thatachieved under continuous processing in the pilotplant under similar process conditions. Both the

Figure 4. Continuous sterilization pilot plant.

Figure 3. Fresh fruit bunch crusher.

autoclave and pilot plant allowed good contactbetween steam and bunches. The study providedconclusive evidence that crushed fruit bunchessterilized in a continuous manner using steam atatmospheric pressure for about 1 hr can be strippedusing a conventional drum stripper. Only a smallpercentage (approximately 8.3%) of bunches still had

149


unstripped fruits after passing through the stripper.Even the unstripped bunches had only very fewfruits left. There were none of the highly unstrippedbunches commonly observed with the batchsterilization process.

Examination of the unstripped bunches showedthat many of the fruits could be easily stripped byhand, indicating that the stripping efficiency of thedrum stripper was less than 100%. To maximize thestripping effectiveness, it is important that thebunches should have sufficient mass so that the forcegenerated by the impact after their fall in the stripperis sufficient to dislodge the fruits. This may be aproblem with small pieces of crushed bunches.

Another factor responsible for the unstrippedbunches could be poor crushing effectiveness. Largevariations in the effectiveness of opening up of thebunches to facilitate steam penetration duringsterilization were observed. Very small bunches werealso not effectively opened up by the crusher becausethe gap between the rollers was too large. Thisproblem can probably be rectified by fine-tuning thecrusher operation or by using more than onecrushing step (either in series or in parallel).

It was observed that there was a tendency forsmall pieces of stripped spikelets to pass throughthe bars in the stripper drum and end up with thefruits and subsequently pass through the digesterand screw press. It may be necessary to modify thestripper to prevent this.

ASSESSMENT OF COOKING EFFECTIVENESS

Sterilization is one of the key processes in the palmoil mill. It is believed that heating during batchsterilization results in: (i) deactivation of the oil-splitting enzymes to prevent an increase in free fattyacids (FFA); (ii) loosening of the fruits in the bunchto facilitate stripping; (iii) softening of the fruit pulpto facilitate further treatment of the fruit duringdigestion and pressing; (iv) heating and partialdehydration of the nuts to enable the nuts to becracked more readily; (v) coagulation of protein inthe oil-bearing cells to prevent formation of colloidalcomplexes to facilitate the separation/clarificationof the oil; and (vi) hydrolysis/decomposition ofmucilaginous material to also facilitate the oilclarification process.

Heating crushed bunches for about 60 min usingsteam at atmospheric pressure appears to beinsufficient to achieve all the above objectives ofbatch sterilization. Table 1 summarizes the results ofa comparative study of the cooking effectiveness ofbatch and continuous sterilization based on the easeof peeling off of the mesocarp from the nuts. Themesocarp was peeled off by a simple squeezingaction using our fingers without applying too muchpressure. Although the mesocarp was soft enough

to be quite easily peeled off by scraping, for thepurpose of this qualitative test, the peeling wasachieved without the use of scrapers.

It can be observed from Table 1 that the mesocarpcould not be peeled off from 2.13% of the fruits frombatch sterilization. These fruits were mainly from theinner layers of the bunch and had a very thinmesocarp. On the other hand, the mesocarp couldnot be peeled off from 24.10% of the fruits fromcontinuous sterilization. It can also be observed fromTable 2 that the peeled mesocarp constituted 45.47%of the initial weight of peeled fruits in the case ofbatch sterilization, whereas it was 37.68% in the caseof continuous sterilization. The above test clearlyshows that the fruits from continuous sterilizationwere not as well-cooked as the fruits from batchsterilization.

TABLE 1. ASSESSMENT OF COOKING EFFECTIVENESS

Source Hard Peeledfruits mesocarp(%) (%)

Batch sterilization(i.e. fruits from mill) 2.13 45.47Continuous sterilization 24.10 37.68

Exploratory studies were carried out using alaboratory-scale batch autoclave to assess the extentof further heat treatment that will be required to fullycook the fruits to a comparable standard as batchsterilization as gauged by the peeling test. A fewkilograms of fruits from continuous sterilizationwere cooked in an autoclave using steam atatmospheric pressure and at a pressure of 10 psig.

It can be observed from Tables 2 and 3 that about60 min of further heating using steam at atmosphericpressure were necessary, whereas 20 to 30 min weresufficient at 10 psig.

TABLE 2. EFFECT OF FURTHER HEATING ON FRUITSUSING STEAM AT ATMOSPHERIC PRESSURE

Heating time Hard fruits Peeled mesocarp(min) (%) (%)

0 24.10 37.6820 6.48 39.3640 2.48 40.8160 1.04 43.51

150


FURTHER PROCESSING

Although the fruits from continuous sterilizationwere not as well cooked as the fruits from the batchprocess, trials were still conducted on furtherprocessing using the conventional milling process.About two to three sterilizer cages of bunches werestripped using a drum stripper, the fruits heatedusing a conventional digester and the oil extractedusing a twin-screw press. It was possible to achieveabout 20 to 45 min of continuous processing by thescrew press. Samples of the oil and press cake werecollected from the same screw press when processingfruits from continuous sterilization as well as fruitsfrom batch sterilization.

The study showed that it was possible to extractoil from fruits by screw pressing in spite of thereduced cooking effectiveness of the continuoussterilization process. The moisture and oil contentof the pressed fibre (Table 4) were higher than withbatch sterilization. Further studies are needed togauge if the oil loss in fibre can be reduced bychanging the operating and design parameters of thescrew press.

The generally low FFA content of 2.1% in the oilfrom continuous sterilization (Table 5) shows thatthere is no significant increase in FFA due to bruisingby the crusher. This can be attributed to the shortdelay between the crushing and heating steps.

From Table 5, the FFA content in the oil from thecontinuous sterilization was slightly lower than thatin the oil from batch sterilization. This could be dueto the fact that the bunches used in our continuoussterilization were generally fresher than the bunchesprocessed by the mill. Typically, fresh fruit bunchesare stored in hoppers for at least a few hours beforethey are processed by the mill, whereas the bunchesprocessed by the pilot plant were mainly freshlybrought to the mill.

Table 5 shows that although the peroxide valuewas approximately the same for both processes, theDOBI value of the oil from continuous sterilizationwas higher by 0.5. This may have been due toreduced oxidation due to the sterilization beingcarried out at a lower temperature.

The carotene content of the oil from continuoussterilization (Table 5) was lower by about 50 ppm.The oil and press cake were lighter coloured in thecase of continuous sterilization. There is probably acorrelation between the cooking effectiveness andcarotene content of the oil. There may also be acorrelation between the pressure applied duringpressing and the carotene content of the oil. Studieson double pressing (Choo and Yap, 1991) showedthat the oil from the second press generally had ahigher carotene content than the oil from the firstpress. It is likely that lower pressure was appliedwhen pressing the fruits from continuoussterilization (reflected in the higher moisture and oilcontents of the fibre).

The iron content of oil from continuoussterilization was about 2 ppm higher than that frombatch sterilization. The higher iron content may bedue to increased contact with the metal surfacesduring the conveying. The pilot plant was notoperated daily and often left idle for several days.Rust was observed to have formed on the insidesurfaces of the conveyors and this could easily havebeen scraped off by the bunches. Further studies arerequired to establish if a comparable increase in ironcontent can be expected in a plant operatedcommercially.

TABLE 3. EFFECT OF FURTHER HEATING ON FRUITSUSING STEAM AT 10 psig

Heating time (min) Hard fruits (%)

0 20.6110 14.8520 4.3230 0

TABLE 4. ANALYSIS OF PRESS CAKE

Parameter Batch Continuoussterilization sterilization

Percentage fibre - 49.21Percentage brokennuts - 6.17Percentage wholenuts - 44.62F/N ratio - 1.00Percentage moisture 33.18 44.41in fibrePercentage oil infibre (wet basis) 4.07 4.65Percentage oil infibre (dry basis) 6.09 8.49

TABLE 5. EFFECT OF STERILIZATION PROCESS ONOIL QUALITY

Parameter Batch Continuoussterilization sterilization

FFA content (%) 2.68 2.10Peroxide value (meq) 0.30 0.29DOBI 2.77 3.22Carotene content (ppm) 598 546Iron content (ppm) 4.24 6.18

151


CONCLUSION

The pilot plant study has demonstrated the technicalviability of the proposed continuous sterilizationsystem and provided the information needed forscaling up to commercial-scale and integrating withthe rest of the palm oil milling process.

It has also demonstrated the possibility ofsterilizing crushed bunches in a continuous mannerusing steam at atmospheric pressure to an extentsufficient for most of the fruits to be quite easilystripped by a conventional drum stripper. The pilotplant study has also demonstrated that the bunchescan be heated using steam at atmospheric pressurewithout difficulty. Uniform temperature wasachieved throughout the continuous sterilizer inspite of the very low pressure. Steam loss fromcontinuous feeding of bunches to the continuoussterilization system was minimized by incorporatinga pre-heating section. The effect of bruising by bunchcrushing on FFA formation was also observed to beinsignificant, especially if the bunches were pre-heated immediately after crushing.

The study has also shown that there should beno major problems integrating the proposedcontinuous sterilization process with the rest of themilling process in a conventional mill. The focus inthis study has been on the processes thatimmediately follow sterilization, i.e. stripping,digestion and pressing. It appears that slight changesin the design of the stripper, digester and probablyeven the screw press may be needed. The effect ofthe new process on milling processes furtherdownstream, such as clarification and nut cracking,were, however, not properly evaluated.

ACKNOWLEDGEMENT

The authors wish to thank the Director-General ofMPOB for permission to publish this paper. Theauthors also wish to thank the management ofKumpulan Guthrie Berhad for their permission tocarry out this study in Rantau Palm Oil Mill.

REFERENCES

MONGANA REPORT (1955). Volume 1. PORIM,Bangi p. 62-80.

SIVASOTHY, K and ROHAYA, M H (2000).Crushing and sterilization of fresh fruit bunches: apromising approach for continuous sterilization.International Planters Conference. 17-20 May 2000.Kuala Lumpur.

SIVASOTHY, K; ROHAYA, M H and TAN, Y W(2000). Continuous sterilization of fresh fruitbunches. National Seminar on Palm Oil Milling,Refining Technology, Quality and Environment. 3-4July 2000, Genting Highlands.

CHOO, Y M and YAP, S C (1991). Valuablecomponents from palm pressed fibre. Seminar onRecent Developments in Palm Oil MillingTechnology and Environmental Management. 3-4July 2000, Genting Highlands.

a new system for continuous sterilization of...

Documents