Fish Feed Technologist
THA/89/003
Description of Duties
1.— Advise on the upgrading of the existing feed mill of NIFI.
2.— Advise on the economic and engineering aspects of feed mill construction with special reference to small-scale and research operations.
3.— Advise on the micro-ingredient distribution in feed materials and the related effects on the quality of mixtures.
4.— Advise on the improvement of nutritional value and reduction of toxic materials in the feed preparation process.
5.— Advise on the review the feeding regimes as well as the research programme related to fish feed technology.
6.— Advise on and review sinking pellet and floating pellet manufacturing by using locally available feed ingredients.
7.— Submit a technical report at the end of the assignment.
The consultant left the United States on September 2, and arrived in Bangkok on September 3. As this was a new consultancy on an unfamiliar project, discussions about the scope of the project and duties within the terms of reference (TOR) were held, first with the Imre Csavas, the Regional Aquaculture Officer in the FAO Regional Office, and then with Mr. Prasert Sitasit, the Government Project Manager in the Thailand Department of Fisheries. Mr. Presert said that most of the work to be performed within the terms of the consultancy would be done in association with Wimol Jantrarotai, PhD., of the research staff, and therefore the consultant's TOR should be discussed with him, to obtain the specific needs of the program.
The consultant and Dr. Wimol were aware that information relating to item 2 had been addressed most comprehensively by the book Feed and Feeding of Fish and Shrimp, by Dr. Michael New (UNDP/FAO publication ADCP/REP/87/26). This publication is a complete presentation of information needed by fish nutritionists and feed manufacturers and covers economic and engineering aspects of small-scale feed mill construction. It was decided that there was no need to cover this same material. Discussion within this topic would then revolve around issues related to the fulfillment of item 1 of the TOR.
Dr. Wimol also said that one of the previous feeds technology consultants, Dr. Nagai, had instructed them well on micro-ingredient distribution and mixing theory, assigned as item 3 of the present TOR. As there was no need to discuss the same material again, fulfilling the TOR of this item would best be served by passing on any practical information concerning mixing when, during the period of the consultancy, it became apparent that such information would help the staff to produce better feeds.
Dr. Halver arrived in Bangkok on Sept. 12, and Dr. Hardy initially arrived on Sept. 27. Dr. Halver focused on the continuation of training of Dr. Wimol in the use of the HPLC system for vitamin C analysis and later, after Dr. Hardy's arrival, supervised the work of Dr. Supis Thongrod in determining fatty acid profiles of a number of lipid sources for fish feeds. Dr. Hardy worked with Dr. Wimol and trained him to use the HPLC system for amino acid analysis.
The implication for the feed technology consultancy was that there were times when one or both of these researchers were occupied with the activities of Dr. Halver or Dr. Hardy during the period from Sept. 12 to Dec. 5. During these times it was necessary to perform some of the work with Somreuk Sinmuk, the technician in charge of the quality control/materials analysis laboratory, other NIFI staff, or after Oct. 15, with two student interns from Buraphar University, Penkhar Nasuwan and Siriporn Rajchapakdee. Therefore, during the four month period of the consultancy, the feed technology consultant was involved in advising and training activities with at least seven different individuals.
A number of feed formulations with potential were developed during the consultancy (Annexes N,O), but as the consultancy was outside the springtime catfish breeding season, a lack of small fish, and facilities to grow them in, did not permit extensive growth studies. Fiberglass tanks inside the research facility which were available to Dr. Wimol were used to conduct two growth experiments set up by the student interns. One study involved feeds based on warm water test diet (WWTD) and different preparations of soybean meal (Penkhar's Diets, Annex R), and the other study utilized diet containing WWTD and different amounts of broken rice and lipids as energy sources (Siriporn's Diets, Annex S). During the development of the feeds for Penkhar, Dr. Wimol had at first wanted to use equivalent amounts of the different soybean preparations in the different feeds. The soybean preparations to be used were regular soybean meal, soybean meal boiled 60 minutes, soybean meal treated with an acid solution, and soybean meal treated with an alkaline solution. Proximate analysis indicated that these preparations had different percentages of protein and fat. It was pointed out to Dr. Wimol that using equivalent amounts of these preparations would have resulted in feeds which were not isonitrogenous, and it would not have been possible to attribute the growth response on these feeds only to the soybean meal preparation method, but the different amounts of protein would have also been a factor leading to differences in growth rates. Dr. Wimol was shown how to do step-by-step calculations to develop isocaloric, isonitrogenous diets using these different materials.
Dr. Wimol was able to make three feeds on the extruder which were sent to another researcher in a collaborative study (Annex Q). These feeds were made without any supervision by the consultant, yet the final products were quite good and floated well. Dr. Wimol made one of these feeds twice, once putting it through the dryer, and once air drying it on the floor of the feed mill to compare the amount of shrinkage which occurs in the dryer feed compared with the air dried feed.
One day was spent in the Kasetsart University Library familiarizing Dr. Wimol's assistant, the quality control lab technician, with journals, periodicals, books, and reference sources relating to feed technology, food science and food analysis. Special emphasis was put on collecting information about alternative feed materials such as leucaena, cassava, winged beans, etc., that might be used as test ingredients in future research work.
Several books on the winged bean plant (Psophocarpus tetragonolobus) were brought from the United States by the consultant, and Dr. Wimol was able to copy some of this material. Dr. Wimol was encouraged to investigate the mature seeds of the winged bean, a plant indigenous to Thailand, as a replacement for soybeans, which are mostly imported. The winged bean, which has been shown in experimental production to produce the largest quantity of protein per hectare of any cultivated plant, and has extraordinary capability for nitrifying soils, has great potential in Thailand for widespread cultivation as a protein crop and as a fertilizer-free means of soil enrichment. Although not cultivated on a large scale basis yet, Dr. Wimol received a visitor from the Oilseed Producers Association of Thailand, and was told that they are interested in growing winged beans for vegetable oil, and were looking for marketing opportunities for winged bean seed meal. He was asked if he had any information about the nutrient characteristics of the meal. Dr. Wimol, in a cooperative effort with the Oilseed Producers, intends to perform proximate analysis and amino acid analysis of winged bean seed meal provided by them, and will initiate research into the suitability of using winged bean seed meal in fish feeds.
In addition to giving advice and assistance on most of the topics in the TOR, there a need for training in other areas related to feeds technology—
To help the staff to use the recently installed steam extrusion system to make better floating feeds, a better understanding of extrusion technology and the extrusion process was needed. Materials which were given to, and discussed with Dr. Wimol or Mr. Prasert concerning extrusion of feeds have been placed in Annex A.
Problems with the installed feed dryer created the need for a discussion of food engineering principles of dryer operations and the theory of drying.
Training staff members proper mixing practices in the production of semi-purified experimental diets.
To train Dr. Wimol and staff members methods for the determination of the adequacy of cooking of oilseed meals, how to inactivate and measure toxic materials in soybean meals (soybean trypsin inhibitor and lectins) and other similar oilseed meals.
To initiate training in procedures for conducting digestibility trials, and other techniques needed to determine quality and utilization efficiency of nutrients.
To train staff in conducting water stability trials on floating and sinking feeds.
UNDP/FAO had funded the purchase of an expensive Dura-Dry freeze dryer, which had not been used for 2 years because it did not seem to work right, and the instructions for its operation were quite complicated. In addition to the need to investigate the functional problem, training in the operation of the freeze-dryer was necessary.
To help the technician in the quality control/materials analysis laboratory to improve analytical methods, laboratory safety and working conditions.
The Government Project Manager, Mr. Prasert, asked Dr. Ron Hardy to organize a 4-day training class in fish nutrition and feeds technology for fisheries biologists, which was held from Nov. 25 - 28th at NIFI. Four lectures on the following topics were prepared and presented to the biologists at the request of Mr Prasert and Dr. Hardy—
In addition to these research and training activities, the consultant was able to participate in two regional workshops. The first was a workshop on “Aquaculture Feed Processing and Nutrition”, which was held September 19–21, 1991 in Bangkok. This workshop was sponsored by the American Soybean Association, to provide aquaculture feed producers in Thailand and other Southeast Asian countries with current information about selection of raw materials for feeds, nutrient requirements of fish and crustacean species cultivated in Asia, feed processing techniques, and the use of special feed additives such as carotenoid pigments and vitamins. Notes were taken from several lectures given by Charles Botting, an expert in the principles and practices of feeds extrusion. These notes were later shared with NIFI staff and used as the basis of discussion for the improvement of extruded floating feeds produced in the NIFI feed mill. The second workshop was a “Regional Seminar on Mud Crab Culture and Trade in the Bay of Bengal Region”, sponsored by the UNDP/FAO Bay of Bengal Programme and the Thailand Department of Fisheries, held in Surat Thani, November 5–8, 1991. The consultant attended as an observer, but was able to make a number of suggestions to participants regarding needed nutrition research for mud crabs and the problems of feeds and feeds manufacturing.
Mr. Chen Foo Yan, the director of the Network of Aquaculture Centers in Asia (NACA) asked for advice on the best means for initiating a regional fish nutrition project which would give small-scale fish and shrimp farmers the capability of making low-polluting, lowcost simple feeds using locally available ingredients. A meeting with Mr. Chen was held on November 26th. Based on input from Mr. Chen as to the aims and objectives of such a project, a statement outlining the needed scope and objectives and means of implementation of such a project was prepared and given to him by December 13th.
1.1 The feed mill staff now have the capability of making both dry compressed pellet sinking feeds using old CPM equipment, and floating or slowly-sinking extruded feeds using the recently installed Taiwanese Idah single-screw extruder. Delta Vet Co. Ltd, the company that installed the extruder, brought several people in, and spent a week training the two feed mill operators in extruder operation. A representative of Idah came and worked with the feed mill operators for one month to train them in feeds manufacturing. According to Dr. Wimol, the company representative wanted them to make a formulated feed using materials which were not readily available here, and after substitutions of ingredients were made, results were not very successful.
Grinding for both production lines is done with an old hammer mill, but it seems to be produce materials which are ground finely enough for most feeds which are produced here. There was a problem grinding some full-fat soybean meal, and large particles of soybean in the mash plugged the 2 mm extruder die during one run. A vibratory screener should have been able to remove the large particles, but on some occasions poorly ground materials were found mixed in with finely ground materials. The screener was not taken apart to see if there were perforations in some of the screens, but there is a good possibility that holes are present.
Ribbon Mixer—
There is a 4 mm gap between the outer edge of the ribbon and the mixer wall. Adding water to the mixer would aid aggregation of materials in the mixer, however since there is usually no water added, a layer of unmixed material remains constantly in the bottom of the mixer, even after emptying. A metal cover on the mixer for dust control makes it quite awkward to clean out the mixer completely, and the feed mill workers were not cleaning it out. This was creating a situation of carry-over from one batch to another, which is a poor practice in a mill trying to produce experimental feeds. The staff were told they should sweep out the mixer after it has been emptied, through the discharge port.
The mixer is emptied through a small 15 cm × 15 cm square discharge port which is opened by removing a sliding sheet metal gate. Feed then drops into a bucket elevator. The size and position of the gate makes it somewhat difficult to empty the mixer thoroughly.
Hot Air Tunnel Dryer—
This dryer is not at all suited for drying feeds. Annex B contains a copy of a fax which I sent to a food engineer at the University of Washington, which outlined the problem. Annex C contains his response.
Extruder— (manufactured by Idah Machinery, Taiwan)
This extruder was found to be fairly limited in it versatility and can produce only a narrow range of feed. The primary problem is limited capacity for conditioning the feed mash prior to extrusion. A secondary problem is very limited ability to monitor and control conditions in the extruder barrel. Upgrading of gauges and replacement of the existing butterfly valves with higher precision needle valves was recommended, as a way to have better control. The purchase of micrometer gauges for measuring wear on the extruder screw was suggested, as the operators currently have no way of checking the tolerances of the internal screw parts. Installation of a variable speed motor control to the conditioner and extruder would increase the usefulness of this machine.
Bucket Elevator—
There are several areas on the bucket elevator where poor welds allow feed mixes to blow out of cracks during operation of the lift. These gaps should be sealed with a sealer.
CPM Pellet Machines—
The CPM pellet machines have been well used by the Institute staff. However, the steam conditioning has been limited by a small capacity steam generator. Additional insulated piping to move steam from the new steam generator installed for operation of the extruder to the CPM machines was recommended early in November, but the work was not done before the end of this consultancy.
2.1 Dr. Wimol said that one of the previous consultants, Dr. Narai, had instructed them in on micro-ingredient distribution, and that there was no need to cover the same area twice. However, a number of practical lessons related to the use of a Hobart Mixer for preparation of test diets and premixes were passed on to the NIFI staff.
2.2 A broken segment on a mixing paddle for the Hobart mixer in the experimental feeds preparation room was found. This paddle had been broken for several years. This mixer-paddle combination was used not only for mixing feeds but also dry mineral and vitamin mixes. A demonstration of the poor mixing which this paddle obtained was shown by mixing together a white-colored caseingelatin test diet, and then adding 0.5% (w/w) chromic oxide to the center of the mix bowl. Within a few minutes, the center of the bowl was uniformly green, but a 1.5 cm band of white test diet, not containing chromic oxide, was evident. This showed that use of this paddle for preparing mineral and vitamin mixes was not a good idea, as it was not possible to get adequate uniform mixing with it. It was suggested that all premixes prepared in the previous few years had not been well-mixed, and replacement of the paddle was recommended.
2.3 Casein-gelatin test diets had been routinely prepared by mixing all ingredients together, and then adding ambient-temperature water to the mixture. Although producing a decent noodle when extruded through a die as a semi-moist mix, the water stability of feeds produced by this method are not as good as can be obtained by other means. The staff were advised that at the very least, using hot water instead of ambient-temperature water would allow some activation of gelatin and starch in the mixtures, and improved pellet water stability. Heating the gelatin (or starch as found in other test diets) in cold water to be added to the mixture, bringing it to a boil, and then adding this to the premixed remaining dry ingredients was shown to create a much more stable pellet, and was recommended as the best method for producing test diets.
2.4 Staff members were also shown how to judge when enough water had been added to a dry mix to extrude a noodle which would hold together, but not stick to other noodles simultaneously emerging from the die.
3.1. Most of the work in this portion of the assignment were performed using soybeans and soybean meals as the test material. Soybeans, when not properly prepared, present a number of difficulties when incorporated into fish feeds. Proper heating of soybeans and soybean meals can not only improve nutritional value, but also denatures and inactivates a number of components which are considered to be “anti-nutrients”— soybean trypsin inhibitor, lectins, and phytate. The methodology for measurement of degree of cooking and antinutrient levels in soybeans can be used with all beans and leguminous products, and many oilseed meals.
3.2 Dr. Wimol had been working with several preparations of treated and untreated soybean meals prior to the consultancy. He had started with a mechanically extracted soybean meal, and had made three other preparations: boiled for 60 minutes; treated with acid; treated with base. In addition there were a raw fullfat soybean meal prepared from whole beans in the laboratory, and a commercially prepared full-fat soybean meal obtained from a feeds commodity company. Since he had made these different preparations, he was interested in obtaining a complete set of data on the effect of these treatments.
3.3 Several assays for determining the degree of cooking of soybean meals were taught to the staff.
A. Protein Solubility Index
The method of Araba and Dale, based on the paper found in Annex F, was taught to Somreuk Sinmuk. Results of the analysis of Dr. Wimol's soybean meal preparations can be found in Annex G.
B. Cresol Dye Test
The method of Olomucki and Bornstein, based on the research paper placed in Annex H, was taught to Dr. Wimol and Somreuk Sinmuk, the QC technician. Dr. Wimol's samples of soybean meals, some of which he had prepared in different ways were analyzed. Results have been reported in Annex I. When full-fat soybean meal or winged bean meal is used in this test, lipids cloud the supernatant, and make spectrophotometric analysis inaccurate. For this reason, a simple room temperature procedure for extracting lipids from fullfat meals was taught as well.
C. Soybean Trypsin Inhibitor Assay
The method of Keshiun and Markakis, based on the research paper in Annex J, was taught to Somreuk Sinmuk and the two student interns. Dr. Wimol was able to observe the procedure on one day. results of the analysis of Dr. Wimol's soybean preparations have been reported in Annex K.
Determination of the digestibility of ingredients and feeds is a very important part of fish feeds development. It allows the researcher to assess the quality of feeds and feedstuffs, and in association with Kjeldahl nitrogen analysis, lipid analysis, amino acid analysis and calorimetry, the utilization of nutrients by fish can be determined. Two attempts were made with Dr. Wimol to determine digestibility of feeds using chromic oxide as a marker. Feeds used in both trials are shown in Annex L. In the first instance, a series of five feeds were made and fed to small (40 g) catfish in 200 L circular tanks with conical bottoms. The fish were weighed, put into the tanks, and then adapted to the feed. the shape of the tanks made it difficult to judge the gut passage time of the test diets. After several days, as the fish could not be sacrificed, collection of feces was done by stripping, but only a very small amount of feces, not enough for analysis, was recovered. It was recommended that he try using glass aquariums, where it would be possible to see the bottom of tanks and the presence of excreted feces. It would then be possible to determine the gut retention time of the fishes' morning meal. Secondly, he was advised to use larger fish which would make feces extraction easier. In the second trial, the recommendations from the first attempt were adopted. Ten (200 g) fish were placed in each of 6 200 L glass aquariums.
Training for the analysis for chromic oxide was done using the method of Stevenson and De Langen (Annex M). This method was taught because in addition to being a reliable and accurate assay, it is a safe method which avoids the use of perchloric acid. Perchloric acid work usually requires a specially designed fume hood to minimize the possibility of explosions in the laboratory.
Feeding regimes used during experimental trials to determine the effect of different dietary treatments was discussed with Dr. Wimol. the benefits of designing experiments in which paired feeding (or modified paired feeding) is conducted along with ad libitum feeding was emphasized as a way of estimating and controlling the effect of palatability of a feedstuff in growth experiments. This is particularly important when new feedstuffs are being used for the first time. To summarize, if fish eat different amounts of feed with increasing inclusion levels of a test ingredient, consumption may be affected by the concentration of the test ingredient in the feed. Ad libitum feeding will create a situation in which fish that do not like the taste of a feed will eat less than fish which do not object to the taste, and the poorly feeding fish will grow less as a result. In a paired feeding trial, all fish are fed the amount of feed which the poorest eaters on any given day have consumed. In a modified procedure, the amount of feed consumed by each treatment group is recorded, and the following day the fish are fed the amount that the poorest eaters consumed the day before. If this is done consistently, all fish will be fed the same amount of feed during the experiment, and differences in growth will be based on actual feed efficiency and not consumption rates.
Concern about the polluting effects of feeds has highlighted the need to carefully regulate the amount of calcium and phosphorus in feeds. In general, feeds are overly abundant in calcium from usage of large amounts of fish meal, and are low in phosphorus with respect to the appropriate Ca:P ratio needed to give the greatest benefit to the cultured animals. If calcium levels in feeds are reduced, phosphorus supplementation does not need to be as high, and the pollution caused by phosphorus will be minimized. A working area in one of the laboratories was set up, equipment was assembled, and reagents were mixed. The addition of the capability of analyzing feedstuffs and feeds for these two minerals should help the fish nutrition staff to control the balance of these nutrients when the need arises.
When the first few extruded feeds were manufactured, the extruder operating conditions were correlated with the density and water stability of the feeds, to show the effect when proper extrusion conditions had been achieved by the operators. The ability of some of these feeds to float were measured, and in most cases, more than 90% of the pellets were able remain floating after 5 minutes.
There are two freeze dryers available to the NIFI nutrition research group. One instrument is an old relatively portable unit with external mounting of freeze dry flasks over a small plastic drying chamber. This machine had been purchased by NACA. The second freeze dryer is a Dura-Dry freeze dryer with sophisticated temperature controls. In addition to a lower condenser/vacuum pump unit there is an upper tray heating/cooling unit which is integrated into the drying chamber of the freeze dryer. Neither freeze dryer was operational at the beginning of the consultancy. The small NACA purchased unit had not worked for some time, as the vacuum pump had a leaking oil seal. When the condenser unit was tested to see if it could be operated with another vacuum pump, it did not work. Inspection of the unit reveled that a family of mice had been nesting inside the unit, and a number of wires had been eaten through. This freeze dryer was moved to another laboratory from the feed preparation room, and replacement of the wires and repair of the vacuum seal were recommended.
The Dura-Dry had not worked properly from the time it was installed, and several other consultants had tried to make it operational but had not succeeded. The condenser chamber temperature would not fall below 0°C. even after an hour of operation, and therefore temperatures below - 40°C, needed for proper freeze drying, were not achievable. Refrigeration servicemen were brought in, and during a second visit, they recharged the refrigeration unit with more coolant. The difference was immediately apparent. Within 15 minutes of turning on the condenser cooling unit, temperatures of -50°C were obtained, and after 1 hour of operation, temperatures reached - 65°C.
The research staff were unclear about operating the Dura-Dry. It is a very good instrument and it has two control panels with a number of switches; its operation is not simple. The NIFI staff had a duplicated set of brief instructions, but no complete operations manuals for the Dura-Dry freeze dryer. The brief instructions did not explain in a way that was understandable to Ms. Nanthiya Unprasert, Dr. Wimol and others how to operate the freeze dryer. A simpler, more understandable, step-by-step set of instructions were written and posted by the freeze dryer. Correct operation of both the upper and lower units of the freeze dryer was demonstrated to all of the research staff in the nutrition group.
5.1 On two occasions, production of feeds with the working CPM machine was observed. Feeds were dried on the floor rather than being put through the tunnel dryer. Use of the dryer was recommended. The feeds appeared to be uniformly ground, well packed, and when dry, were hard and fairly resistant to breakage. The capability of the staff to vary formulations is limited because of the low output of the old steam generator, which reduces the ability to condition the pellet mash sufficiently. A request was made to Mr. Prasert to connect a steam pipe from the new steam generator to service the old CPM pellet machines. This work was not done before the end of the consultancy, and experimentation with sinking feeds for freshwater prawns that was proposed to Dr. Wimol and Mr. Prasert was not initiated as a result.
5.2 Dr. Supis was assigned the task of trying to develop a water stable macrobrachium prawn feed using locally available materials, and she asked for assistance. The first diet was made using uncooked broken rice, casein, fish meal, shrimp head meal, soybean meal, alpha-starch, squid liver oil, minerals, vitamins and CMC binder (Annex N). Dr. Supis was assisted in the mixing and extruding of the feed, as well as in the determination of its water stability. When water stability was shown to be poor, two modifications of the initial formula were made, one including glutinous rice flour instead of the broken rice, and a second, tapioca flour. These feeds were made using both uncooked and precooked rice flour and tapioca flour. To assess the effect of different drying methods on the quality of the feeds, 1/3 of the feed was dried in the freeze dryer, 1/3 was sun dried, and 1/3 was over dried.
5.3 Work with soft-dry feeds was suggested, as there are many fish which do not like to eat hard pellets. Such feeds do not require refrigeration, and are well suited to tropical climates. The formulation using full-fat soybean meal (Annex O, Diet 3) extruded well in a 4 mm size, and if preservatives and binders had been added, this might have made a good starting point for work with producing these types of feeds. This work was never started, however, because a recommendation to purchase a supply of feed preservatives commonly used in soft-dry feeds so that such work could be initiated, was never followed up on by the NIFI staff.
6.1 Most of the formulation and manufacturing activities conducted during the consultancy were for extruded floating fish feeds. The work started with a modification of an extruded practical diet for channel catfish (Ictalurus punctatus) published in Nutrient Requirements of Warmwater Fishes and Shellfishes. The formula was adjusted for variations in the nutrients between American and Thai feedstuffs (Annex O, Diet 1). The first time this feed was made for the purpose of seeing the staff operate the extruder, minerals, vitamins and oil were omitted. Subsequent to this first attempt, these additives were included in all formulations. Initially, a vitamin mix which had been prepared by Dr. Wimol was used. As there was a need for a considerable amount of vitamin mix, and not enough time to make it, Mr. Prasert was encouraged to contact a feed mill, and a 10 kg supply of mineral/vitamin mix was obtained. This formulation was successfully extruded, but relies heavily on corn and soybean meal, two ingredients that are grown in limited quantities in Thailand. Thai feed makers usually rely on imported supplies. Much of the soybean meal used is imported from China, and costs only 1–2 Baht less per kg than fish meal. Ground broken rice was substituted for the corn in the formulation and extruded (Annex O, Diet 2). Full-fat soybean meal was used instead of extracted soybean meal in a formulation (Annex O, Diet 3), but the limited conditioning capability of the extruder coupled with the high fat content of the feed created problems during extrusion (see 6.3 below). Soybean meal was removed from the soybean meal/broken rice formulation and replaced with peanut meal (Annex P, Diet 1). Some of the broken rice in this formulation was taken out and 10% cassava flour was utilized (Annex P, Diet 4). When Mr. Worawut Ungjitpaisarn, sales manager for Delta Vet Co. Ltd. visited NIFI, he was shown some of these warmwater fish feeds, and said that they were better than anything he had been able to produce on the extruder during the week that he and his staff were at NIFI setting up the extruder for operation.
6.2 An attempt was made to develop an experimental matrix in which different a single formulation was manufactured under different extruder operating conditions. However, the inability of the operators to control the operating conditions, previously mentioned in section 1.1, became a problem. It was impossible to accurately control water and steam input to the conditioner, and the moisture content of the feed coming out of the extruder varied enough that if a formulation was coming out well expanded, in a few minutes, without any changes being made by the operators, the moisture content would change, and the feed would no longer expand as well.
6.3 An attempt was made to develop an experimental matrix in which a number of isocaloric and isonitrogenous formulations varying in lipid and carbohydrate content were manufactured. Because of the inability of the operators to control the degree of conditioning or the length of conditioning time, it was not possible to properly condition the mash. In one instance, a formulation containing fullfat soybean meal was tried. This feed had a lipid content near the maximum recommended level for extruder production, but should not have been difficult to make. However, it was not properly conditioned, and the feed was coming out but not expanding. Then the die plugged from large pieces of soybean in the mash about 25% of the way though the run.. The die had to be removed, and a larger bore die was installed so that the feed could be removed before it set up in the extruder barrel and conditioning chamber.
7.1 With respect to employee health and safety, the main complaint of the staff in the laboratory was that they were subjected to frequent headaches and sore throats. A previous employee being paid by the UNDP/FAO project had quit because of this. This laboratory was retrofitted into a room which was never designed for this purpose. The laboratory was found to contain two muffle furnaces, two drying ovens, a refrigerator, a Soxtec lipid extraction unit, a Fibertec fiber analysis unit, and a Kjeldahl nitrogen analysis system consisting of a digestion unit and a distillation unit. Both units of the Kjeldahl system have acid traps/sodium hydroxide scrubbers which discharge into a drain pipe under a sink.
7.2 Only the Kjeldahl digestion unit and a hot plate for predigestion of samples with nitric acid were in a fume hood. The fume hood was not large enough to hold both the digestor unit and its associated scrubber unit. Inspection of the operation of the two scrubber units revealed that although they were being discharged under a sink, there was no vapor trap installed, and vapors were able to rcenter the laboratory environment through the sink drain hole. After consultation with the technician, a request was made in early November to have a trap installed under the sink. This work had not been done by the end of the consultancy. The short term solution was to plug the sink drain with a wet cloth during Kjeldahl and fiber determinations, which improved air quality significantly.
7.3 Because the Soxtec unit is operated on a benchtop, petroleum ether vapors are discharged into the laboratory environment whenever lipid extractions are performed. The staff usually turn on a fan and open a window when they are working with this apparatus. Inspection of this operation showed ether vapors rising into the laboratory even when the window was open and the fan was used. More careful placement of the fan helped. The person performing lipid extractions was advised to wear safety glasses and a face mask with a chemical air filter during this work.
7.4 The refrigerator and muffle furnaces were moved out of the laboratory into a foyer area with better ventilation. An old wooden fume hood was moved to the foyer area but installation had not been done by the end of December. The laboratory floors and some of the walls were thoroughly scrubbed, but it was not possible to clean a high, sloping glass ceiling. Purchase of two more fume hoods was requested on the supplies and equipment list submitted with the project extension report submitted by Dr. Halver.
8.1 As there was some question about some of the crude protein results obtained in the laboratory, Somreuk Sinmuk felt that he did not have enough ability to assess the reliability of their methods. Training was initiated in a procedure for the quantitative recovery of nitrogen from the Kjeldahl analysis apparatus. The manual for the apparatus contained a method for quantitative recovery using acetic anhydride, but we were not able to buy it, as it is a restricted chemical in Thailand. Therefore, ammonium sulfate was purchased, and a quantitative method was developed. The method was done twice by Somreuk under supervision, and he was advised to have his other staff try a quantitative recovery to assess whether they were performing the method reliably. Results of this nitrogen recovery effort are reported in Annex T.
In the Feed Mill
Check the screens in the hammer mill and in the vibratory shaker to see if there are perforations in the screens. Purchase and replace screens as needed.
Install an insulated steam pipe from the new steam generator to the old CPM pellet machines to increase the capability of this part of the mill operation and to improve the quality of feeds produced.
Refit the tunnel dryer so that it is functions efficiently.
Seal the air leaks in the bucket elevator with a silicone sealer or equivalent.
Purchase and install an air compressor and air storage system, so that the feed mill operators are able to properly clean machinery.
Purchase a set of tools, including micrometer gauges, so that the feed mill operators can operate and maintain their equipment effectively. The purchase of micrometer gauges will allow the operators to measure wear and tolerances of the extruder. A new bimetal thermometer was purchased and installed to replace one which did not work at all, but the operator did not have an accurate way to judge the depth to set it, which he could have done with a micrometer gauge. As a result the thermometer was set too deep, and the inner tip of the thermometer was bent the first time the extruder was run. The thermometer was not able to return to room temperature readings, and its accuracy was questioned. Another thermometer has been damaged in the same way, and because the freedom of the bimetal to move inside the thermometer base has been restricted due to bending of the tip, the thermometer only increases its temperature readings when the operator hits it with a stick.
Upgrade thermometer gauges and put steam gauges in a position where they can be read more easily. Install a more accurate flow control device on the water intake to the conditioner, as the present gauge is inadequate for accurate water metering. Replacement of the existing inadequate butterfly valves on the steam lines with higher precision needle valves is recommended, as a way to have better control over temperatures in the extruder barrel, and also to precisely regulate steam injected into the feed mash in the barrel. Installation of variable speed motor controls to the conditioner and extruder drive motors would increase the usefulness of this machine considerably.
Shorten the duct between the conditioner and extruder and install a motorized feeder assembly between the conditioner and the extruder. On many occasions conditioned feed mash became packed above the extruder screw, and had to be pushed down with a stick until it was caught by the screw. A feeder with tines mounted on counter-rotating rods would break up the feed and push it into the extruder.
There have been times when the extruder has been operated without running feed through it, for instance, during the many times that the feed mash became packed above the extruder mouth. Since the operators have no capability for reducing the speed of the extruder, this causes a great deal of wear on the extruder screw. If at all possible, the installation of a variable speed control on the extruder drive motor is highly recommended. This would allow the operators not only to reduce wear on the screw, it would also give the extruder more versatility in feed production.
The feed mill operators should be required to keep a log book of all work done on the extruder, times and duration of operation, problems during extrusion, etc. There is no way to determine the cause and rate of wear, the cause of problems without such a log. The operators should also keep track of formulations which have been made and the conditions under which they have been successfully made. Storage of a formulation or proper production conditions in the memory of an operator limits the ability of other staff members to make feed if the normal operator is away from NIFI.
In the Quality Control/Materials Analysis Laboratory
Install a second fume hood in the laboratory for the Soxtec lipids extraction unit. Install a fume hood in the foyer area for holding the muffle furnaces.
Install a trap in the sink drain to prevent vapors from Kjeldahl and Fibertec analyzers from entering the laboratory atmosphere.
Research in Feeds Technology
After increasing the conditioning capacity of the CPM compressed pellet machines by installing a steam pipe from the new steam generator, initiate a research project testing feeds for fresh water prawns.
One of the limitations at NIFI is the lack of ponds on site for feeds testing. Dr. Wimol needs to develop further ties with other DOF facilities and researchers so that he can test the ability of his feeds to grow fish.
A number of feeds were formulated and produced in the NIFI mill. When possible Dr. Wimol should remake these feeds at a time when they can be properly tested.
| Mr. Imre Csavas | Regional Aquaculture Officer FAO/RAPA |
| Mr. Sompong Hiranwat | Director of NIFI |
| Mr. Prasert Sitasit | Government Project Manager THA/89/003, NIFI |
| Mr. Chen Foo Yan | Coordinator Network of Aquaculture Centers in Asia |
| Mr. Charles Angell | Senior Aquaculturist BOBP/FAO Dhaka, Bangladesh |
| Dr. Dean Akiyama | Technical Director for Aquaculture American Soybean Association |
| Mr. Charles Botting | Feeds Extrusion Consultant C.A. Botting and Sons, Guildford, U.K. |
| Dr. Chin How Cheong | General Manager Ceylon Grain Elevators, Ltd. Colombo, Sri Lanka |
| Dr. George Pigott | Institute of Food Science and Technology University of Washington, Seattle, WA. USA |
| Mr. Worawut Ungjitpaisarn, B.Eng. | Sales Manager, Engineering Division Delta Vet Co. Ltd. Bangkok |
