Evaluation of a Prototype Fish Cleaning Machine With Proposals for aCommercial Processing Line

J. M. MENDELSOHN and J. G. CALLAN

Background

Heading and gutting small fish byhand is both time consuming andlabor intensive, making it an expen­sive operation. In cases where specialhandling is absolutely necessary orwhere the products can command ahigh price, processing fish by handmay be economically feasible; butwhere the fish product has to be com­petitive in price with other high pro­tein foods, or the value of the fisheryproduct is relatively low, machineprocessing must be used. Not onlymust the economic aspect be consid­ered but where the need to produce aheaded and fully cleaned fish forfurther processing is important, amachine must clean the fish com­pletely in order to be justified. This isespecially evident where the fish mustbe headed and completely freed ofviscera and belly lining, particularly

J. M. Mendelsohn and J. G. Callan are with theGloucester Laboratory, Northeast FisheriesCenter, National Marine Fisheries Service,NOAA, Emerson Avenue, Gloucester, MA01930.

ABSTRACT-A prototype heading andcleaning machine for small whiting wasevaluated under commercial conditions.The major finding was that the machine issound in theory and principle but needssome of its components redesigned forhigh speed production. Throughput wasfound to be about 2,250 fish/hour (lessthan was expected) of which about 95 per-

38

when the lining is pigmented as in thecase of whiting, Merlucciusbilinearis, prior to processing in ameat/bone separator. As a solution toheading and fully cleaning whiting, aLaPine Model 22 smelt processingmachine 1 was extensively modified(Mendelsohn et aI., 1977).

Preliminary Results

The modified machine (Fig. I) waslaboratory tested for processing vari­ous sizes of whiting 8-16 inches(20.3-40.6 cm) in length in test runsinvolving 500 pounds (225 kg) to5,000 pounds (2,250 kg) of fish. Fishoutside of this size range were culledand discarded because the processingeffectiveness of the machine was notacceptable for these. Since themachine was designed to handle fishthat varied in length by no more than4 inches (10.2 cm), we adjusted themachine to handle fish in the range of10-14 inches (25.4-35.6 cm) because

'Reference to trade names or commercial firmsdoes not imply endorsement by the NationalMarine Fisheries Service, NOAA.

cent were cleaned well enough for furtherprocessing or to be frozen directly as areverse-butterfly or pan-ready product.

From these tests, we believe that newmachines could be built to handle fishfrom 6 to l8 inches (l5.2-45.7 em) atproduction rates up to l2,000 fish/hour.For increased production, three machinesshould be used together in a system com-

this is the center of the 8- to 16-inch(20.3- to 40.6-cm) range.

The results of the laboratory testsindicated that the prototype machinehad a capacity of about 3,600 fish/hour and that it could effectively headand clean the fish and remove most ofthe peritoneum (black belly lining) ofwhiting. Subsequent trial runs beforeindustry further demonstrated themachine's potential, and only anevaluation of the machine's perfor­mance under actual commercial condi­tions was necessary to encourage itsassimilation by industry. Coordinatedby the authors, the machine wastested by a Boston processor undercontract to the New England FisheriesSteering Committee (NEFSC)2.

Processing in Commercial Plant

Gloucester-landed whiting werebrought to the Boston commercial fishprocessor for heading and cleaning bythe modified LaPine machine. The

2NEFSC is a nonprofit educational associationorganized to further the interest and welfare of allthose engaged in the domestic New Englandfishery (Anonymous, 1977).

prising auxiliary equipment. Eachmachine would process a specific sizerange offish with an expected total outputfrom the three machines of about 36,000fish/hour. The fish would be scaled,graded, oriented, conveyed, headed, andcleaned automatically with maximum re­covery of edible fish flesh and withminimum waste.

Marine Fisheries Review

Figure I. - Prototype fish heading and cleaning machine.

size range of the fish varied from wellbelow 6 inches (15.2 cm) to wellabove 18 inches (45.7 cm). Althoughthe machine was laboratory tested fora size range of 8-16 inches (20.3-40.6cm), an attempt was made to head andclean whiting between 6 and 18 in­ches (15.2 and 45.7 cm) to determineif the enlarged size range could be ac­ceptably accommodated.

Those fish below 6 inches (15.2cm) and above 18 inches (45.7 cm) inlength were culled and discarded.Again, as in the laboratory experi­ments, we adjusted the machine tohandle the middle of the range 10-14inches (25.4-35.6 cm). Once set, thecontrols were not changed during theruns.

This sequence was followed in atypical run using either penned 3 (5-6day old) or day-boat (1-2 day old)whiting in the heading and guttingmachine in the commercial fish pro­cessing plant:

I) Boxed iced whiting, landed inGloucester, Mass., in the morning,were trucked to the plant in the after­noon and held overnight in a cooler at35"F (rC).

2) The next morning, the fish werescaled in a commercial rotary scaler.

3) The scaled fish were manuallyplaced on the wooden cleated con­veyor leading to the heading and gut­ting machine.

4) Every fish was headed and gut­ted by the machine.

5) Each fish was inspected as it leftthe machine. Uncleaned fish werehand-cleaned or discarded.

6) If the fish were to be sold aspan-ready (reverse-butterflied) fish,they were frozen in 10 pound boxes.

7) If the fish were to be deboned,they were transferred to a conveyorleading to the Bibun Model 18 meat/bone separator. The minced fish fleshwas collected and made into 18.5­pound minced blocks and plate fro-

3Penned whiting refers to fish from boats thathave been out fishing 4 to 5 days. These whitingare 5 to 6 days old prior to processing. Day-boatwhiting refers to fish from boats out fishing foronly I day. These fish are I to 2 days old prior toprocessing.

January 1980

zen. Frozen fish blocks are used as theraw material in the manufacture offish sticks, portions, and other spe­cialty items (Ryan, 1978).

Results and Discussion

Operation UnderCommercial Conditions

Operation of the machine undercommercial conditions indicated thatwhile the theory and the principlesemployed were sound and effective,the theoretical capacity of 3,600 fish/hour could not be met. The maximumprocessing rate reached was about2,700 fish/hour with four peopleoperating the machine-two peoplefeeding the machine, one person keep­ing the machine running properly, andone person inspecting the fish. Evenat this rate, the machine had to bestopped for about 4-5 minutes every25 minutes for a complete wash-downto remove viscera collecting at thecleaning wheels which would other­wise impede the smooth flow of fishthrough the machine.

With a throughput of 45 fish/minutefor 25 minutes operating time and al­most 5 minutes cleanup, the machine

would head and gut about 2,250 fish/hour. With the weight of the whitingaveraging about one-half pound, themachine throughput was about 1,125pounds/hour. Even if the machinethroughput could be doubled by ex­perienced operators, it would fall farshort of the production rate desired bythe larger processors.

Still, the economics of using themachine compare favorably withmanual gutting and heading of whit­ing. The speed of an experienced per­son that hand processes whiting isabout 30 fish in 9 minutes or about200 fish (100 pounds)/hour. Even if aperson could continue to handle 100pounds/hour throughout the day, theprototype machine works over II timesfaster, with no loss in speed as mightoccur with a human as the day wearson. Thus while the machine might fallshort of commercial processor expecta­tions, the fact that one machine plusfour workers can replace II skilledcutters is evidence that it still representsan economic advantage. Also, thequality of the fish remains constant inmachine processing, whereas the qual­ity of the manually processed whiting

39

is variable and generally poorer, espe­cially as the workday wears on.

Yield Recovery

Starting with whole whiting, the re­covery of headless and fully cleanedreverse-butterfly whiting (pan-ready)was determined. Results from twoseparate runs showed that from a mix­ture of sizes of whole whiting (6-18inches or 15.2-45.7 cm), yields of46.0 percent and 46.3 percent wereobtained. These yields were approxi­mately 5 percent lower than thatfound under laboratory conditions.This can be explained somewhat be­cause only whiting between 8 and 16inches (20.3 and 40.6 cm) were usedin the laboratory tests. With fish oflarge size ranges, the yield is apprecia­bly decreased for the machine was setfor maximum recovery in the 10- tol4-inch (25.4- to 35-cm) range.

Even if whiting are to be sold as aheadless and dressed item, the 46 per­cent yield from the prototype machineas compared with a 50 percent yieldfrom the conventional headless anddressed operation should be no deter­rent, especially as the quality of theformer is better than that of the latter.

The improved appearance of thefish with all of the viscera removedand the potential shelf life extensionof the product from the prototypemachine should be able to commanda higher price than that of convention­al headless whiting. Most whitingprocessed in the conventional waycontain a small amount of viscera.Therefore, they cannot be labelled asheadless and dressed fish but must belabelled as headless whiting.

However, in a plant under U.S.Department of Commerce inspection,the fish from the prototype machinecould carry the U.S. Grade A stampbecause they would easily conform to

the headless, dressed whiting stan­dard. Where the whiting are to be putthrough a meat/bone separator or areto be further processed, the completeremoval of viscera and black belly lin­ing is essential.

Cleaning Efficiency

The number of fully cleaned whit­ing, requiring no followup hand clean-

40

ing, obtained from the machine wastaken as a measure of its effectivenessunder commercial conditions. Ap­proximately 30 percent of the whitingwere found to fit this category. About65 percent had only a very small pieceof black belly lining on them andabout 5 percent had both pieces ofviscera and black belly lining. An in­spector at the end of the cleaningmachine should be able to remove bitsof black belly lining with a plasticbrush or by handpicking so that about95 percent of the whiting would becom pletely clean. The remaining 5percent could be recycled or hand­cleaned at the end of the run.

Although most of the black bellylining was removed by the two clean­ing wheels and serrated roller at theend of the machine, it was observedthat the 6-inch (l5.2-cm) diametercleaning wheels would only cleansmall fish whose belly flaps did notexceed 2.5 inches (6.35 em) in size.The shaft on the wheel restrained thebelly flap from being in full contactwith the cleaning wheel. Since mostof the whiting in the commercial trialwere quite large, they tended to havebelly flaps longer than those thatcould be cleaned by the 6-inch (15.2­em) wheel presently on the machine.This is believed to account for thesmall amount of black belly liningremaining on the large whiting.

Problems Encountered

In addition to the time lost forwashing the machine during the run,several other problems were encoun­tered. One of the major downtimeproblems was the binding of the shaftsin certain sections of the machine.These shafts did not have sealed bear­ings, and fish juice which seeped intothem tended to increase resistanceforces until the shafts eventuallystopped turning. Prior to each run, theywere freed and oiled; but during therun, the fish juice would cause someshafts to stick and/or to stop turning.

Several other minor problems wereencountered during the commercialruns. One of these was the turning ofthe fish head on the cleated conveyorprior to its being removed. Since onlythe body of the fish is supported by

the conveyor, the head resting againsta stationary flat plastic guide is re­tarded slightly under the frictionalforce as it is dragged along to theheading blade. This causes the part ofthe fish that meets the rotating head­ing blade to be cut at an angle. If thehead were also conveyed on its ownsystem, the cut could be made per­pendicular to the axis of the fish formaximum yield.

Until recently, most conventionalheading machines had this inefficientaspect, but the trend is now leaningtoward the multiple metal pocket con­veyor where the fish are secured inposition while the head is being re­moved and other cuts are being made.This operation, which takes place atthe operating speed of the machine,measures the head of the fish as itcomes down the conveyor and clampsit in position to sever it at the correctangle for maximum yield. We haveconcluded that a higher and significantefficiency is possible and should be at­tained.

Another problem was the wetting ofthe electronic circuit boards, espe­cially during clean-up operations. Thecircuitry had been mounted in water­proof housings and all connectionshad been made waterproof with rub­ber gaskets and other seals. However,during the moving of the machinefrom Gloucester to East Boston,Mass., the cables for power, water,and air had to be disconnected and re­connected at their destination. Sub­sequently, some of the connectionsleaked and, as a result, the main elec­tronic circuit board broke down due tothe water and had to be replaced. As atemporary measure, clear plastic cov­ers were used to protect the electronicgear which will be redesigned so thatthe controls and other susceptibleequipment would be completely pro­tected.

During operation of the machine, alarge amount of fish waste collectedunder it. This is acceptable in smalloperations; but in a large continuousoperation, gurry conveyors should beinstalled. Most large fish processorsalready have these conveyors avail­able; therefore, no design work isneeded, only installation.

Marine Fisheries Review

Table 1.-Annual cost 01 processing whiting in theprototype machine vs. a manual operation'.

1Based on the need for 11 cutters for manual operation forthe same production of fish cleaned using the machineand 4 people.

commercial whiting processor topurchase a machine similar to the pro­totype machine described in this arti­cle.

After careful consideration of thecost and complexity of the modifica­tions needed to increase the speed andefficiency of the prototype headingand cleaning machine, we recommendbuilding a completely new secondgeneration heading and cleaning unitrather than modifying the existing pro­totype machine. The second genera­tion model would have improved fea­tures to: I) Increase throughput offish; 2) lower number of rejects re­quiring further handling; 3) reducedown-time (for any reason); and 4)maximize effectiveness of operatingcomponents in the production line.

The new line of machines would bebuilt around the successful engineer­ing principles found in the prototypemachine but would have a muchlarger throughput. In the smallerplants, one machine would besufficient to handle their productionwhile in the larger plants, as many asthree machines would be installed,depending upon the amount of fish tobe processed.

The proposed high capacity fishprocessing line is shown in Figure 2.Each unit would have a throughput ofabout 200 fish/minute; and if all threem'lchines were operating at full capac-

Products From Whiting Machine

Although headed and cleaned whit­ing can be used as the starting mate­rial for a wide variety of fishery prod­ucts, especially extrudable minceditems (Mendelsohn, 1974a), only twotypes of whiting products-a frozenreverse-butterfly or pan-ready itemand a frozen minced block-wereprepared. The variety of product typeswas limited by the equipment that wasavailable in the participating plant.

The pan-ready product needs nofurther processing and can be sold asa small consumer or large institutionalpack. At the processing plant, theclean whiting were packed in layers(head to tail), separated by plasticsheets, in 10-pound waxed boxes andfrozen in a plate freezer. After freez­ing, the boxes were placed into a mas­ter carton and the carton kept at-10°F (-23°C). National MarineFisheries Service personnel and oneindustry member examined the pan­ready product made from very freshwhiting and judged it highly accept­able. The pan-ready product madefrom 5-6 day old whiting was judgedas somewhat less acceptable due to itsslightly "fishy" odor.

The frozen minced whiting blockswere made by putting the cleaned,headless whiting through the meat/bone separator. The minced productwas then put into 18.5-pound waxedcartons and frozen in a plate freezer.After freezing, four blocks wereplaced inside a plastic bag and into amaster carton. The cartons were thenplaced into a freezer held at -1O~(-23°C).

To determine the acceptance ofthese blocks, they were cut into fishsticks and battered and breaded. Theresults of an organoleptic evaluationof the sticks made from fresh whitingby laboratory personnel showed thatthey were highly acceptable. They re­ceived an overall score of 7.3 (good tovery good).4 Sticks made from thepenned whiting (5-6 days old) using

4 Based on a scale of 9 to I where: 9 = excellent, 8=verygood,7=good,6 =fair,5 = borderline,4 = slightly poor, 3 = poor, 2 = very poor, andI = inedible (Mendelsohn, 1974b).

the same processing conditions wererated 6.4 (fair to good).

The largest quality differences be­tween the penned whiting and day-boatwhiting were recorded in their appear­ance and flavor. In the sticks madefrom the older fish, more black spotsappeared because of their softer skinwhich tended to squeeze through theholes in the meat/bone separator drumand be collected with the edible por­tion.

Recommendations andConclusions

The result of our tests indicated thatthe prototype whiting processingmachine performed largely as ex­pected. However, the commercial op­eration uncovered the need to furthermodify it and to add accessoryequipment in order to attain itsmaximum production potential.

Even though the throughput of theprototype machine is too low to meetthe demand in the larger processingplants, there is still a sizeableeconomic advantage in using amachine of this type. Based on cur­rent labor practices in the Gloucesterarea, where fish processors pay about$7.70/hour (including fringe benefits)for general help and $8.65/hour forcutters and semiskilled workers, wehave calculated the annual cost forprocessing a similar volume of fishusing the prototype fish heading andcleaning machine and compared itwith the cost of heading and cleaningthe fish manually. We assumed thatthe cost of building the prototypemachine to be $40,000. This estimateis based on our experience with otherfish processing machines. Interestrates were calculated on 12 percentper annum, and the machine wasamortized over a period of 5 years.Utility costs were calculated from ac­tual commercial water and electricitybills. The total costs are itemized andtabulated in Table I.

As shown in Table I, using the pro­totype machine with its limited capac­ity could save a fish processor over$115,000 on a yearly basis. Coupledwith a more consistently cleaned prod­uct of higher qual ity, we concludethat it would be worthwhile for a

Ilem

Prototype machineMachine (amortized over 5 yr)

Principal and interestMaintenanceLabor

3 General help @$16.016/yr1 Semiskilled help

ElectricityWater

Total

Manual operationLabor

11 Semiskilled help (cutters)@,' $17,992/yr

Waler

Total

Difference

COSI

$ 10,6905,000

48,04817,992

903373

$ 83.006

$197,912186

$198,098

$115.092

January 1980 41

WHOLE FISH

TATION

OR

\ING and CLEANINGHINE FOR SMALLFISH (601-1001

)

~HOPPERSCALINGMACHINE

~LONGITUDINAL

LONGITUDINALADJUSTABLE IORIENTOR

I GRADER ORIENTOR

+!LONGITUDINAL

I I CONVEYONVEYOR

) ORIENTOR

CONVEYOR-HEAD

NING MACRGE

') SLITTING BLADE

CLEANING WHEELS \/ ;)\ .... HEADING S°35 ::0

C

HEADING and CLEAMACHINE FOR LA

FISH (14"-18'

HEADING and CLEANING MACHINE FORMEDIUM FISH (10"-14") OR ADJUSTABLE

FOR RANGES OF FISH BETWEEN 6"and 18"

Figure 2. -Proposed high capacity processing line to head and clean 6-18 inch(15.2-45.7 COl) fish.

ity, the line would be able to handle600 fish/minute varying within a sizerange from 6 to 18 inches (15.2-45.7em) in length.

As shown in Figure 2, the highcapacity processing line would con­tain the following pieces of equip­ment:

I) An in-line rotary scalingmachine to scale all the fish whilethey are still in the round. In the con­ventional processing of whiting, thescaling operation is done by a rotaryscaler in a later part of the processingsequence. We are suggesting to do itas a first step when the fish are in the

round to eliminate immediately theproblems caused by the scales.

2) An adjustable size graderwhich will separate the fish into threesize ranges, 6-10 inches (15.2-25.4em), 10-14 inches (25.4-35.6 em),and 14-18 inches (35.6-45.7 em).

3) A reservoir to hold each size of

42 Marine Fisheries Review

fish until they are ready to be con­veyed into a machine that: a) Orientsthe fish longitudinally, and b) orientsthe fish vertically.

4) A machine to position the fishto be loaded automatically four at atime onto the infeed conveyor.

5) A conveyor to move the fishinto an automatic device which mea­sures the length of the head and setsthe cut-off knife to remove the headwith a minimum loss of fish flesh.

6) An offloading device thattransfers the fish to the processingconveyor where the fish is held in posi­tion between the carrier belts of thecleaning machine.

7) Semiautomatic air pressurecontrols to change the pressure ad­justments to regulate: a) The depth ofthe rotary slitting blade used to openthe fish's belly; b) the lateral pressureon the belts that hold the inner portionof the fish against the cleaningwheels; c) the vertical downwardpressure on the hold-down rolls whichkeep the fish tight over the cleaningwheels that remove the viscera alongthe backbone. The size of the fishpassing through the machine deter­mines the pressure that each pressurecontrol should apply to do the clean­ing job without losing excessive fishflesh. Sorting the fish into three sizecategories and running one size cate­gory at a time makes it possible to ad­just one set of regulators at a time tohandle the one size category of fish toits best advantage. In a small plantwith only one cleaning machine, eachset of regulators would be preset andcontrolled by a selector switch tohandle a single size range of fish. Thesize of the range would be no greaterthan 4 inches (10.2 cm) in length. Bypressing the preselector button, thecontrol system that had been in com­mand for one size category would beshut off, and the controls for the sizecategory to be run would be cut in.The time for switching controls willbe in the order of 5 seconds. In the

January 1980

larger processing plants where threeheading and cleaning machines areneeded to handle their throughput, thecontrols on each machine would bepreset for one size range of fish -alsonot to exceed 4 inches (10.2 cm) inthe variability of length. Althougheach machine would have the capabil­ity of manually switching from onesize range to another, it would be un­necessary when using the threemachines. This system increases theeffectiveness of the machines bymatching the size range of the fish tothe pressure range of the controls.One range of pressure settings doesnot work for a full size range of fish.

8) Cleaning wheels of increaseddiameter (from 6 inches (15.2 cm) to12 inches (30.4 cm)) in addition to athird cleaning wheel. This changeshould greatly improve the quality ofcleaning the belly cavity, especially inregard to the black belly lining in thelarger fish. With a l2-inch (30A-cm)cleaning wheel, the carrier bel t can be7 inches (17.8 cm) wide to hold theentire belly flap against the side of thewheel with optimum pressure to re­move all of the black peritoneum.With three 12-inch (30.4-cm) cleaningwheels, the extension conveyor at theend of the present machine equippedwith a conical skinning wheel to cleanup the tips of the belly flaps will beunnecessary. This system should giveadequate internal cleaning so that thecarrier belts can release the fish di­rectly onto the inspection belt whichthen drops the cleaned fish onto aconveyor for further processing.

9) Carriers with high belt speedsand a transfer system for conveying12,000 fish/hour.

10) A large and efficient waste dis­posal system. In the prototypemachine, the gurry disposal systemwas overtaxed necessitating shut­downs to clean the gurry ducts andareas where buildups occurred. A so­lution to this problem is to install oneor more high pressure fan-jets at each

station where fish parts and gurry ac­cumulate. The hoses to the fan-jetswould be secured to a rack connectedto an air or hydraulic cylinder whichwould operate on time intervals of 15minutes. At the preset time cycle, thehoses would make one traverse andthen return to a starting position. Fishparts which had accumulated duringthe previous minutes would bewashed down the gurry ducts to con­veyors for quick removal, thus mak­ing it a smooth continuous operationand eliminating the need to stop themachine for cleaning.

A system of machines as describedabove is not only commercially feasi­ble but is necessary for processinglarge quantities of fish between 6 and18 inches (15.2 and 45.7 cm) inlength to keep production costs com­petitive with costs in other food indus­tries. Also, since more and more em­phasis is being placed on quality, theproposed processing line would pro­duce consistently high quality seafoodproducts.

Acknowledgments

This work was supported in part bythe New England Fisheries SteeringCommittee under contract with Chan­nel Fish Co., East Boston, Mass. Theauthors also gratefully acknowlege thehelp given by the personnel of Chan­nel Fish Co. during the commercialtesting of the prototype machine.

Literature CitedAnonymous. 1977. New England Fisheries De­

velopment Program Report of Progress 1977.New England Fisheries Steering Committee,New Bedford, Mass., 20 p.

Mendelsohn, 1. M. 1974a. Minced fish in a newform. Mar. Fish. Rev. 36(8):34-36.

___. 1974b. A study: Expanded process­ing techniques, production costs and marketsurvey of underutiIi zed fish species. U.S. Dep.Commer., Econ. Dev. Admin., Tech. Assist.Proj. 01-6-09131-2, 51 p.

___, T. J. Connors, and J. G. Callan.1977. A machine for heading and evisceratingsmall fish. Mar. Fish. Rev. 39(2):11-18.

Ryan, J. J. 1978. Preparation offish fillet blocks.Mar. Fish. Rev. 40(1):5-12.

43