results and discussion - shodhgangashodhganga.inflibnet.ac.in/bitstream/10603/43269/14/14_chapter...
TRANSCRIPT
RESULTS AND DISCUSSION
CHAPTER IV
RESULTS AND DISCUSSION
This chapter deals with the results obtained on engineering aspects of copra
processing and the results obtained during evaluation of improved copra processing gadgets
developed during this study under the following headings.
® Copra Processing Practices Adopted and Constraints Experienced by Farmers
® Physical, Thermal and Mechanical Properties of Coconut
® Sorption Isotherms
® Development of Nut Splitting Device
® Development of De-Shelling Machine
® Thin-Layer Drying
® Deep-Bed Drying and
® Copra Dryer and Quality Characteristics of Copra and Oil
4.1. Copra Processing Practices Adopted and Constraints Experienced by Farmers
Copra processing practices adopted by farmers and constraints faced by farmers are
reported and discussed in the following sub sections.
4.1.1. Socio-personal characteristics of the farmers
The socio-personal characteristics of the coconut farmers under the study are
presented in Table 4.1.
From the Table 4.1, it is clear that a vast majority of the farmers (88.80 %) were
above 40 years old. Only 11.2 % of the farmers were less than 40 years. The observed
pattern of distribution of farmers according to age is in line with the general trend observed
in Kerala state where the younger generation keeps away from farming (State Planning
Board, Kerala, 2002). It was also observed that there were no illiterate farmers among the
respondents. More than 22.3 % of the farmers were having degree or higher educational
Table 4.1. Distribution of farmers according to their socio-personal characteristics
SI. Characteristics Category No. of farmers
No Frequency Per cet
1. Age <40 years 24 11.240-60 years 140 65.1>60 years 51 23.7
2. Education Illiterate 0 0Primary 47 21.9High school 82 38.1Pre degree 36 17.7Degree 45 20.9PG and above 3 1.4
3. Occupation Farming alone 77 35.8Farming +Agricultural labour 26 12.1Farming + Private job 13 6Farming + Government job 12 5.6Farming+ Business 87 40.5
4. Family size <5 89 41.55-10 120 55.9>10 6 2.8
5. Farm size <0.5 ha 39 18.10.5-1.0 ha 69 32.11.0-1.5 ha 67 31.2>1.5 ha 40 18.6
6. Farming < 10 years 29 13.5experience
10-25 years 84 39.1>25years 102 47.4
7. Annual income <10,000 8 3.710,000-14,000 147 68.4>14,000 60 27.9
qualification. The high rate of literacy in the state is reflected in the distribution pattern of
farmers according to their educational status.
The adoption of improved farming practices by the cultivators would be influenced by
the extent of their involvement in farming as a major source of income for their livelihood.
Among the respondents, only 35.80 % were totally depending on farming alone as their
source of livelihood. The remaining (64.20 %) farmers were engaged in other activities also
besides fanning for their income. This is also in line with the general trend observed in
Kerala state (State Planning Board, Kerala, 2002).
The distribution of farmers according to their family size showed that 41.50 % of the
farmers were having family size of below five members. Percentage of farmers having a
family size of 5-10 was 55.90. Only very few farmers were having larger family size of
more than ten. The distribution pattern of farmers according to their family size is in line
with the general trend in Kerala, where the population growth rate is on the declining phase,
there is an enhanced tendency towards nuclear families and the adherence of people to small
family norms (State Planning Board, Kerala, 2002).
The results indicated that majority of the fanners were having long years of
experience in farming. Only 13.5 % of the farmers were having less than 10 years of
farming experience. It is also observed that the distribution of farmers according to their
experience in farming followed a similar trend among adopters and non-adopters. This is
merely a reflection of the results observed in the case of distribution of farmers according to
their age, as illustrated earlier. Sunita (1998) reported that 52.5 % of the respondents had
high level of farming experience followed by medium 33.37 % and low 14.17 %.
The distribution of farmers according to their farm size indicated that majority of the
respondents (50.2 %) were having only less than one hectare of farmland. Fragmented farm
holdings is quite often cited as a salient feature of Kerala’s agriculture where the average
farm size is only 0.27 ha (State Planning Board, Kerala, 2002). It is a generally observed
trend that the extent of adoption of improved farming methods, including post harvest
processing technologies, was higher in the case of farmers having more acreage than small
and marginal farmers who were obviously handicapped by resource constraints. Similar
results were reported by Jnanadevan (1993) and Thamban and Venugopalan (2002).
The distribution of farmers based on their annual income showed that about two-
third (68. 4 %) of the farmers was having an annual income between Rs.10, 000 to
Rs.14, 000. Nearly one-third (27.9 %) of the fanners were having annual income of more
than Rs.14, 000. Only few (3.7 %). farmers were having their annual income less than
Rs.10, 000. As discussed in the case of farm size, the extent of adoption of improved farm
technologies influences annual income of farmers. Extent of adoption of improved farm
technologies tends to be higher by farmers with more income than their counterparts having
less income. Area production and productivity of coconut in different states, union territories
of India and in the world are given in Appendix - XII and XIII.
4.1.2. Dehusking of coconuts
The practices adopted by farmers for de-husking coconut are furnished in Table 4.2.
Table 4.2. Practices adopted by farmers for de-husking coconut.
Sl.No. Method No. o f farmers Per cent
1. Using crow bar 155 72.1
2. Using knife 6 2.8
3. Using de-husking tool ‘Keramithra’ 54 25.1
The results indicated that majority (72.1 %) of farmers were using the traditional
crow bar for dehusking coconut, especially when large number of coconuts are to be de-
husked for copra production. For dehusking nuts for household use knives were used by
about 2.8 % of fanners. The dehusking tool, ‘Keramithra’, developed by Kerala Agricultural
University has become very popular. About one-fourth of respondents adopted the same.
Dehusking of nuts using the traditional tool such as crow bar is a semi skilled work
and hence farmers mostly depend on hired labour to complete the job. The labour utilization
pattern for dehusking coconut is furnished in Table 4.3
Table 4.3. Labour utilization pattern for dehusking coconut
Sl.No. Source of labour No. o f Per centfarmers
1. Family labour 22 10.7
2. Hired labour 193 89.3
It can be seen from the results that a majority of farmers (89.3 %) employed hired
labour for dehusking coconuts. Only 10.7 % farmers utilized family labour.
4.1.3. Splitting of coconuts
It was observed that all the farmers contacted for this study resorted to knife for
splitting coconut irrespective of the purpose either for splitting small number of nuts for
household use or splitting large number of nuts for copra production. The de-husked nut
would be held in the left hand, which is split with the knife held in the right hand by
applying a small force. Plate 4.1 and 4.2 depicts the traditional method of nut splitting using
a traditional knife. The farmers opined that they adopt the traditional method of using knife
for splitting nuts mainly because there was no other improved tool available which was
superior to knife. Further, all the farmers perceived that the traditional method of using knife
for splitting coconut was less efficient, required some amount of skill and can cause injury if
not properly handled. The results indicated that there is an urgent need to develop a splitting
device which is simple, easy to fabricate, easy to operate, more efficient and economically
viable. The consumption pattern of coconut in India is given in Appendix-XIV.
4.1.4. De-shelling
The result of the study with respect to the mode of de-shelling nuts adopted by
farmers is presented in Table 4.4.
Plate 4.2. Splitting of nuts using traditional knife in a processing unit at Kasaragod
It is evident from Table 4.4. that a majority of farmers adopt the traditional
tool made of wood which is about 12-15 cm long and 5 to 8 mm thick, with flat and sharp
edge for de-shelling coconut for copra production. Plates 4.3 and 4.4 depict a farmer
de-shelling coconut using a traditional wooden knife.
Few farmers (3.7 %) resorted to metallic devices for shelling. Other methods such as
hitting two or three times on the ground with impact force to separate the shell and kernel
were also employed by some farmers (2.3 %). According to farmers all these methods,
required skill for operation and were less efficient. Further there are chances for injury, if
not properly practiced. All the farmers opined that there is an urgent need to develop an
improved de-shelling device. The sample size is representative in nature and large enough to
ensure statistical validity of the estimates made.
Table 4. 4. De-shelling methods adopted by farmers
SS.No. Method of de-shelling No. of farmers Per cent
1. Wooden knife 202 94.0
2. Using metallic device 8 3.7
3. Other methods 5 2.3
Total 215 100.0
4.1.5. Farmers perception about improved devices for splitting and de-shelling
From Table 4.5, it is clear that 84.7 % farmer wanted improved devices for splitting
and deshelling whereas 15.3 % of the farmers surveyed felt that there was no need for
improved devices. Majority of the farmers (92.4 %) felt that there was no need to develop a
de-husking machine as already Keramitra husking tool was very popular among farming
community.
Table 4.5. Farmers perception about improved splitting and de-shelling devices
Improved devices Sl.No. for splitting and de- Frequency
_____ shelling__________________________________________
1 Not required 33 15.3
2 Required 182 84.7
Total 215 100
4.1.6. Production of copra
It is a proven fact that the quality of copra produced is directly influenced by the
method of drying of coconuts. The methods adopted by farmers for production of copra are
furnished in Table 4.6.
Table 4. 6. Different coconut drying methods adopted by farmers
Sl.No. Coconut drying methods No. of farmers
Per cent
1 . Sun drying 121 67.2
2. Smoke drying 55 30.6
3. Indirect drying (Using small holder’s copra dryer) 4 2.2
Total 180 100
The results clearly indicated that a majority of coconut farmers (67.2 %) were
adopting sun drying (direct drying) of coconuts for the preparation of copra. Majority of the
farmers were adopting direct sun drying method for copra production perceived that the
quality of copra produced were inferior mainly due to fungal growth.
Nearly one-third (30.6 %) of the respondents were adopting the traditional smoke
drying method for copra making. In this method burning of coconut husks and shells were
carried out in shallow pits over which a grill of wooden platform was erected. Coconut cups
were arranged in layers over the grill and drying was carried out. Firing was done either
continuously or intermittently. Due to the non uniform heating and heavy smoke, copra
obtained was of low quality, brownish and with the smell of smoke. The copra cake obtained
after extraction of oil was also not of good quality for use as animal feed. The sample size is
representative in nature and large enough to ensure statistical validity of the estimates made.
The ‘copra atti’ the traditional smoking method for copra production in Kerala is shown in
Plate 4.5.
4.1.7. Constraints experienced by farmers adopting traditional methods in copra
drying
Details of different constraints experienced by fanners adopting traditional methods
in copra drying are given in Table 4.7. More than one constraint was reported by some
farmers and it was included more than once while calculating percentage.
Table 4. 7. Constraints experienced by farmers adopting traditional methods in copra drying
Sl.No. Nature of constraints No. of fanners
Per cent
1 Drying is not possible during rainy season
112 62.2
2 Inferior quality due to smoking of copra 75 41.7
3 Lack of space to spread the split nuts for sun drying
109 60.1
4 Inferior quality due to fungal growth on copra
105 58.3
5 Discolouration of copra 108 60.0
It is evident from Table 4.7 that drying coconuts during rainy season was an
important constraint perceived by majority of the farmers (62.2 %) adopting traditional
methods of copra drying such as sun drying and smoking method. Kerala State receives an
annual rainfall of more than 3000 min mostly from two season’s namely South west and
North East monsoon. Obviously drying of coconuts in open sun becomes difficult in rainy
days especially during the last week of May to August where the average monthly rainfall is
above 700 mm (section 4.7.9).
The next major constraint was space. Nearly 60.1 % of the respondents perceived
that lack of space to spread the split nuts for sun drying was a constraint experienced in
traditional method of copra drying. Open, shade free level ground surface was needed for
spreading the split coconuts for drying. About 12 m2 space is required for spreading 1000
split nuts. Only few farmers can afford to have permanent cement dying yard constructed for
copra dying. Traditional method of sun drying often resulted in fungal growth on copra due
to cloudy weather resulting in low temperature and high humidity. Fungal growth reduced
the quality of copra, thus fetching low market price. In the present study, 58.3 % of farmers
perceived this as an important constraint in copra production.
Direct smoking of nuts was another traditional method of drying followed by
farmers. This method of drying was used even during rainy season but the disadvantage was
that the copra gets smoked and its quality was affected. Inferior quality due to smoking of
copra was cited as a constraint by 41.7 % of the fanners in the present study (Table 4.7).
Discolouration of copra was yet another constraint perceived by more than half (60%)
farmers adopting traditional method of copra drying. In sun drying method, where the split
nuts are kept in open, lot of dust particles and other dirt materials from atmosphere fall on
the nuts thereby causing discoloration and in turn reducing the quality of copra. Discolored
copra fetches low price in the market.
It can fairly be concluded from the above results that fanners adopting traditional
methods of copra drying experience various constraints. Some of these constraints affect the
quality of copra produced and also reduce the market price of copra. Hence it is imperative
that appropriate copra drying machinery which is technically superior and economically
viable is the need of the hour to overcome the above constraints and to produce good quality
Small holder’s copra dryer is suitable for small holders who possess coconut
holdings of less than five hectares size and utilize 15000 nuts per annum for copra
production. It mainly consists of a drying chamber, plenum chamber burning cum heat
exchange unit and a chimney. It is simple but requires frequent fuel loading i.e. every
15 minutes. In spite of the obvious advantages over the traditional methods of copra drying,
farmers also perceive some constraints in the adoption of small holder’s copra dryer. The
constraints experienced by farmers while adopting small holder’s copra dryer is given in
Table 4.8.
Table 4.8. Constraints experienced by farmers adopting small holder’s copra dryer
105
4.1.8. Constraints experienced by farmers adopting small holder’s copra dryer
Sl.No. Nature of constraints No. of farmers
Per cent
1. Frequent fuel loading is an inconvenient and labour intensive activity
32 91.43
2. Non-uniform drying of nuts in different parts of the drying chamber
26 74.29
3. Drying time required is too lengthy 23 65.71
4. Maximum temperature generated is not sufficient for proper drying
20 57.14
5. Fungal growth on copra 21 60.00
6. Discoloration of copra 21 60.00
Majority of the selected coconut growers (91.43 %) adopting small holder’s copra
dryer, perceived that frequent fuel loading was an inconvenient and labour intensive activity.
For the effective functioning of the dryer the recommended practice should be to refill the
fuel tray once in every 2-3 h. Addition of fuel once in 15 minutes is inconvenient to the
farmers and involved lot of labour.
Three-fourth of the farmers (74.29 %) perceived that the nuts spread in different
parts of the drying chamber were not uniformly dried. This may be due to the fact that heat
was not uniformly distributed to all parts of the drying chamber. Unequal drying of copra
adversely affects the quality of copra and oil.
The drying time required for small holder’s copra dryer was 36 - 40 h. Though this
was much less than the time required for traditional drying methods, most of the fanners
(65.71 %) perceived that it was too lengthy. Their suggestion is to develop a dryer which
requires still lesser time for drying. Small holder’s copra dryer is shown in Plate 4.6 and 4.7.
According to 57.14 % of the farmers, the maximum temperature generated in the
small holder’s dryer was not sufficient for proper drying of nuts. They perceived that
slightly higher temperature could be tried for proper drying. This indicates the limitation of
the dryer in terms of its design. Hence, in the improved dryer design there should be
provision for generating higher temperature to achieve drying within the expected time.
Fungal growth on copra and discoloration of copra were the other constraints
experienced by farmers while adopting small holder’s copra dryer. About two-third (60 %)
of the farmers perceived these constraints. Improper drying of nuts due to the inherent
deficiencies in the design of small holder’s copra dryer resulted in fungal growth on copra
and discoloration of copra. Hence, it is imperative that a dryer having a proper design
without above constraints have to be developed to meet the requirements of coconut
farmers.
From the foregoing discussion it can be concluded that farmers adopting small
holder’s copra dryer experienced a number of constraints in copra drying mostly due to its
inherent deficiency in the design of the dryer. Hence, there is an urgent need to design and
fabricate an efficient copra dryer to enable the farmers to overcome the constraints
experienced.
4.1.9. Suggestions of farmers for developing an improved copra dryer
Based on the experiences of adopting various drying methods, both traditional
methods and small holder’s dryer, fanners gave suggestions for the design of an improved
dryer. The suggestions of the farmers are summarized in Table 4.9.
Easiness in operation was the most important attribute for the design of an improved
dryer as perceived by half (50.2 %) of the coconut fanners included in the study. Simplicity
of the innovation is an important factor influencing the extent of adoption of the innovation.
Hence, adequate attention has to be paid to this aspect while an improved copra dryer is
designed and fabricated.
Similarly, cost involved in the adoption of a technology also influences its rate of
adoption. That is the reason why 42.8 % of farmers suggested developing a dryer having low
cost.
Farmers also gave suggestions to develop a dryer: having a device to avoid frequent
fuel loading, providing uniform temperature distribution inside the drying chamber and
having a provision for enhanced temperature than that is available with the present design of
small holder’s dryer. It is worthwhile to note here that these suggestions were matching with
the constraints they experienced in the adoption of small holder’s copra dryer. Hence it is
important that these suggestions are taken care while designing the improved dryer.
Table 4.9. Suggestions of farmers for developing an improved copra dryer
SI.No. ItemNo. of
farmersPer cent
1. Develop a dryer with low cost 92 42.8
2. Develop a dryer having a device to avoid frequent fuel loading
36 16.7
3. Develop a dryer providing uniform temperature inside the bin
54 25.1
4. Develop a dryer having a provision for enhanced temperature than that is
38 17.7
5. Develop a dryer which does not use electrical energy
46 21.4
6. Develop a dryer which is easy to operate
108 50.2
More than one-fifth of the respondents (21.4 %) suggested to develop a dryer which
does not use electrical energy because cost of electrical energy for industrial use is higher
and if used the cost of drying will increase. It is an accepted fact that electrical energy is
costly and also the frequent power failures and voltage fluctuations create lot of problems to
the cultivators in Kerala state. Further, it would be ideal if the improved dryer utilizes the
locally available coconut shell and other agricultural wastes as fuel.
4.1.2. Large scale processing of copra
Three large scale processing units were studied and data was collected on method of
splitting, de-shelling and drying which were given in the following sub sections.
4.1.2.1. Kalpaka food products, Palayad industrial estate Palayad, Kannur district
Kerala
The processing capacity of the plant was 15,000 to 20,000 nuts / day. By eight
persons (skilled workers), within 3 hours 30 minutes to 3 hours 45 minutes, about 10,000
nuts were husked. The nuts were split, with a common heavy knife (with the sharp edge).
This was done in a continuous process. They were not looking for the larger eye positions to
split the nut. The average time taken to split one nut was 7 seconds (unpublished data).
Preliminary drying was carried out at about 100 to 120 °C for the initial two hours. Again
after 8 to 9 hours of drying, at about 90 to 100 °C, de-shelling was done. Just by visual
checking, they confirmed whether the cups were ready for shelling or not. De-shelling was
done with a specially cut wooden knife. It took 4 hours to shell 10,000 nuts by six people.
The total drying time was 24 h. The blowers used for drying were having a capacity of 5 hp
(2 nos., with 3 phase supply). To control the burning of firewood / coconut shell, a separate
small blower was used. By controlling the amount of firewood and the air blowing into the
furnace, temperature of drying air in to the drying chambers were controlled. After
de-shelling, drying air temperature was kept at 60 C. Copra was dried for 13 to 15 h after
de-shelling to produce good quality copra. The colour of copra produced was light brown in
colour. High pressure blowers were used to pump the hot air in to the drying chamber. The
1 0 3
depth of drying chamber was 75 cm. Mould growth was not observed due to very high
drying air temperature. Dryer used was indirect type manufactured by Premier Industries,
Ernaluilam, Kerala.
4.1.2.2. The Kottachery co-operative marketing society’s coconut processing unit
Ambalathara, Kanfaangad, Kasaragod district, Kerala
The processing capacity of the plant was 30,000 nuts per day. It was reported that
one kilogram of husked nut contains about 20 % water. Splitting method used was a
common heavy knife. They were not looking for the eye position. By one or two strikes, the
nut splits almost exactly into two halves. The average time taken to split one nut was two to
three seconds (unpublished data). De-shelling was done after partial drying. The capacity of
the primary dryer was about 15,000 nuts and the primary drying time was 12 h. (before
shell removal) and that of the secondary dryer was 15,000 nuts (after the shell removal, 8 h
of drying to reach 6 % moisture content). The method used for de-shelling was traditional
wooden knife. Average time taken was 4 to 6 seconds per cup (unpublished data). Moisture
content was determined by visual observation and they were not using any moisture meter.
In summer, sun drying in the courtyard was being followed and the dryer was not operated
due to high operational cost especially the cost of electricity. In rainy season only dryer was
used. Drying air temperature was 100 °C for the first ten hours and 80 °C for the rest of 10 h.
Total drying time was 12 h in the first stage (before the shell removal) and 8 h after the shell
removal. They were also using premier dryer. The fuel used in the dryer was coconut shell
and the approximate cost was Rs. 2.0 lakhs.
4.1.2.3. Karshaka bandhu coconut products, Kolathur, via Poinachi, Kanhangad,
Kasaragod district, Kerala
The capacity of the plant was 10,000 nuts / day. They used a wooden platform to
place the nut for splitting. The advantage of this method was that, with only one strike, the
nut was split into two halves. They used an ordinary heavy knife to split the nuts.
De-shelling was done after partial drying for about 12 h. De-shelling was done using a small
craw bar, to separate the shell and kernel. Almost all cups were separated with slight force,
after the initial 12 h of drying. For the secondary drying, it took nearly 12 to 18 h. The total
drying time was about 24 to 36 h. They were also using premier dryer. The fuel used in the
dryer was coconut shell.
A comparative study of three processing plants engaged in the production of copra is
presented in Table. 4.10
4.2. Physical, Thermal, Mechanical Properties of Coconut and Sorption Isotherms of
Copra
The results of physical, thermal, mechanical properties of coconut and sorption
isotherms of copra are presented and discussed in the following sections.
4.2.1. Physical properties
The results of physical properties such as size, shape, sphericity, bulk density, true
density, porosity, angle of repose, static and kinetic coefficients of friction were presented
and discussed in the following sections.
4.2.1.1. Constituents of coconut
The minimum, maximum and mean length of the fruit was 16.50, 23.50 and
20.14 cm, respectively. The minimum, maximum and mean width of the fruit was
13, 18 and 14.75 cm, respectively where as there was wide variation with respect to fruit
weight, the minimum, maximum and mean values were 535,1800 and 955 g, respectively.
Ratnambal et al. (1995) reported similar results in case of coconut of West Coast Tall
variety. The other characters are given in Table 4.11.
From the Table 4.11 it is clear that the coefficient of variation was below 10 % for
length, breadth and minimum perimeter of the nut. The coefficient of variation for other
characters was below 20 % except fruit weight. With coefficient of variation 20 %, the
sample size requirement estimates within 5 % relative error was 64. The sample size
requirement is given in Table 4.12. Meunier et al (1977) reported that sample size should be
Table 4.1 J. Descriptive statistics of fruit and nut characters of West Coast Tallvariety coconut (n=64)
Coconut Std. CV
characteristics Minimum Maximum Mean deviation (%)
Length of fruit (cm) 16.5 23.5 20.143 1.2534 6.22
Breadth of fruit (cm) 13 18 14.75 1.333 9.04
Weight of fruit (g) 535 1800 955 328.575 34.41
Thickness of husk
(cm) 2.3 3.8 3.271 0.3311 10.12
Weight of nut (g) 290 850 523.15 130.649 24.97
Husk, percentage 22 59 43.54 7.683 17.65
Thickness of kernel
(wet), cm 1.1 1.8 1.356 0.1568 11.56
Thickness of kernel
(dry), cm 0.8 1.3 1.066 0.1303 12.22
Minimum perimeter
(cm) 43 56 50.02 2.711 5.42
Weight of copra (g) 95 230 157.34 29.221 18.57
Weight of shell (g) 80 180 133.39 26.915 20.18
Table 4.12. Sample size requirement
5 % relative error 10 % relative error
CV=34%
CV= 20%
184.96
64
46.24
16
of 34-36 nuts for error to be less than 5 % and 18 nuts for error greater than 7 % in case of
coconut. Thus the sample size used in the study is adequate to estimate the various fruit
c haracteristics except fruit weight.
The ANOVA of fruit characteristics of WCT coconuts is presented in Table 4.13.
From the Table 4.13, it is clear that copra showed significant correlation with breadth of
fruit, fruit weight, nut weight and shell weight. Thickness of kernel does not show
significant correlation with copra weight whereas breadth of fruit was found to be correlated
with copra weight. Length of the fruit was correlated with the thickness of kernel. Round
nuts were found to have more copra.
4.2.1.2. Size
The geometrical minimum, maximum and mean diameter of fruit was (D) 14.07,
19.67 and 16.38 cm, respectively. The geometrical mean diameter values calculated by
using the equation proposed by Sreenarayanan et al. (1985) gives values very close to the
observed values (Table 4.11). Hence, the equation proposed by Sreenarayana et al. (1985)
can be used for calculating the geometrical mean diameter of West Coast Tall variety of
coconut.
4.2.1.3. Degree of sphericity
The minimum, maximum and mean value of sphericity of fruit calculated using the
equation 3.5 was 69.85, 97.0, and 81 %, respectively which indicates that West Coast Tall
variety of coconut is like a sphere as the values are in the range of 69.85 % and above. Thus
for all practical purposes the fruit of WCT coconuts can be considered as spherical. Similar
results were reported by Patil (1984 a) in case of WCT variety of coconut.
Ratnambal et. al. (1995) reported the shape of West Coat Tall variety as oval.
4.2.1.4. Mass of fifty split coconut
The mass of fifty split nuts (two halves) was in the range of 19.750 to 21.510 kg with
a mean value of 21.065 kg. Thus the mass of 1000 coconuts will be approximately
421.30 kg. These values will be useful in calculating the total load on the de-shelling
machine for design of de-shelling chamber and calculating the horse power of prime mover.
4.2.1.5. Bulk density
The bulk density of coconut kernel decreased from 464.23 to 411.674 kg / in3 when
the moisture content decreased from 94.93 to 6.15 % d.b. (Fig. 4.1). The decrease in bulk
density with decrease in moisture content indicated that the decrease in weight owing to
moisture loss in the sample was greater than the accompanying volumetric contraction of the
bulk. Similar trends have been reported for pistachio seed (Hsu et al, 1981), pumpkin seed
and kernel (Joshi et al., 1993), karingda seed and kernel (Suthar & Das, 1996) and coffee
parchments (Chandrasekar & Vishwanathan, 1999). The variation in bulk density (pb) was
found to be linear with the moisture content (M) and can be represented by the following
regression equation
pb = 0.5541 M+ 412.32 -—4.1
with a R2 value of 0.98
The bulk density of fruit, de-husked nut, split coconut cups, copra and husk varied
from 260.57 to 268.49, 462.40 to 496.86, 459.97 to 502.79; 402.96 to 417.77 and 109.98 to
11 3.86 with mean values of 263.57, 487.78, 497.97, 411.67 and 111.83 kg / m3, respectively
(Appendix - X). Similar results were reported by Senthil (1986) and Bosco (1997) in case of
East Coast Tall variety of coconut. The bulk density of split coconut with shell intact
decreased from 488.643 to 309.96 when the moisture content decreased from
94.93 to 6.15 % d.b. The variation in bulk density with moisture content is given in
Table 4.14 and Appendix - X.
1 1 7
Table 4.14 Bulk density of split coconut with shell intact
Average moisture content,
% (d.b.)
Bulk density,
kg / m3
6.15 309.96 (0.004)
11.60 322.876 (0.002)
18.48 339.869 (0.005)
26.26 361.565 (0.001)
35.04 380.593 (0.003)
43.47 400.626 (0.004)
54.32 426.196 (0.006)
68.35 453.602 (0.001)
81.15 472.292 (0.003)
94.93 488.643 (0.004)
Mean of 5 observations, Values in parentheses indicate standard deviation
4.2.1.6. True density
The true density of kernel decreased from 524.364 to 450.316 kg / m3 when the
moisture content decreased from 94.93 to 6.15 % d.b., respectively (Fig. 4.2). The decrease
in true density with decrease in moisture content indicated that the decrease in weight owing
to moisture loss in the kernel was greater than the volumetric contraction of the individual
kernel. Similar trends have been reported for pumpkin kernel (Joshi et al., 1993), Karingda
kernel (Suthar & Das, 1996) and coffee parchments (Chandrasekar & Vishwanathan, 1999).
The variation in true density (pt) was found to be linear with the moisture content (M) and
can be represented by the following regression equation
pt = 0.8696 M -i- 447.29 -—4.2
with a R2 value of 0.98
The experimentally observed data on bulk and true densities resulted in the following
linear equation
pb = 0.6372 pt +127.31 —4.3
with a R2 value of 0.99
4.2.1.7. Porosity
The porosity of coconut kernel calculated from relevant experimental data revealed
that the porosity decreased non-linearly from 11.29 to 8.58 % as the moisture content
decreased from 94.93 to 6.15 % d.b. It was reported that the porosity increased with the
decrease in moisture content in the case of soybean (Deshpande et al., 1993), pumpkin seed
(Joshi et al, 1993) karingda seed and kernel (Suthar & Das, 1996) neem nut
(Viswanathan et al, 1996), coffee parchment (Chandrasekar & Vishwanathan, 1999), melon
seed (Oje et al., 1999) but decreased in the case of guna seeds (Aviara et al., 1999).
4.2.1.8. Angle of repose
Experimentally determined values of angle of repose of coconut kernel are plotted
against moisture content as shown in Fig. 4.3. From this it observed that the angle of repose
increased from 30.24 to 34.67 as the coconut kernel moisture content decreased from
94.93 to 6.15 % d.b., respectively. This may be due to the fact that at lower moisture content
the coefficient of friction was more there by increasing the angle of the cone. Kaleemullah
(1992) reported similar type of results in the case of groundnut kernels. The relationship
between angle of repose (0) and moisture content (M) is non-linear and can be expressed as
6 = -1.0099 In (M) +38.148 -—4.4
with a R2 value of 0.99
4.2.1.9. Static and kinetic coefficients of friction
The effects of moisture content and surface nature of materials on the static and
kinetic coefficients of friction of coconut kernel are shown in Figs. 4.4 and 4.5. The static
coefficient of friction on the stainless steel surface (SS) varied from 0.436 to 0.284, on the
galvanized iron sheet (GI) from 0.561 to 0.343, on the mild steel sheet (MS) from 0.624 to
0.424 and on the bamboo plywood surface (BPW) from 0.668 to 0.448 while the kinetic
coefficient of friction on the stainless surface varied from 0.346 to 0.173, on the galvanized
1 2 0
iron sheet from 0.447 to 0.271, on the mild steel sheet from 0.519 to 0.358 and on the
bamboo ply wood surface from 0.558 to 0.377 for moisture contents between
94.93 and 6.15 % d.b., respectively (Appendix, XI).
The static coefficient of friction at any moisture content on any surface was higher
than the kinetic coefficient of friction. Similar type of result was quoted in the case of
soybeans, red kidney beans, unshelled peanuts (Chung, 1989) and groundnut kernels
(Kaleemullah, 1992). The maximum static and kinetic coefficient of friction were noted on
bamboo plywood (BPV) surface, followed by mild steel (MS), galvanized iron (GI) and
stainless steel (SS) surfaces. Chandrasekhar and Viswanathan (1999) reported that the
maximum coefficient of friction was in the case of mild steel surface, followed by
galvanized iron, aluminum and stainless steel surfaces, at all moisture contents. All the static
and kinetic coefficients of friction decreased linearly in the moisture range from
94.93 to 6.15 % d.b. Similar trends have been reported for soybeans, red kidney beans,
unshelled peanuts (Chung, 1989), groundnut kernels (Kaleemullah, 1992) and lentil seeds
(Carman, 1996). Since the static and kinetic coefficient of friction were more in case of mild
steel, hence this material was selected for fabricating the de-shelling machine so as to carry
ail the nuts to the top for maximum fall.
4.2.2. Thermal Properties
The moisture dependent thermal properties such as thermal conductivity thermal
diffusivity and specific heat were presented and discussed in the following sub sections.
4.2.2.I. Thermal conductivity
The line source transient heat flow method is the most commonly used method
which has been successfully applied to the measurement of thermal conductivity of various
biological grains (Kazarian & Hall, 1965; Wratten et al, 1969) but the same could not be
used for coconut because the half coconut cup has large air space and as such thermal
conductivity values obtained was those of air rather than coconut kernel. Hence indirect
method was used i.e. the equation proposed by Anderson (1950) and Sweat (1974).
The thermal conductivity of coconut kernel was estimated using the equation
proposed by Anderson and Sweat. The values obtained in the moisture range of
94.93 and 6.15 % d.b. was found to lie between 0.358050 to 0.280249 W / m. K. as per
Anderson’s equation and 0.150401 to 0.148286 W / m. K as per Sweat equation. The
variation of the thermal conductivity with moisture content is shown in Fig. 4.6 and Table
4.15. It was observed that the thermal conductivity, k (W/ m. K) increased linearly with an
increase in moisture content.
1 2 2
Table 4.15. Thermal conductivity of coconut kernel
Moisture Moisture Anderson Sweat
content,d.b.
% content, w.b. (decimal)
Thermalconductivity,
W/ m. K
Thermal conductivity,
W/ in. K6.15 0.058 0.280249 0.148286
11.60 0.104 0.296547 0.148513
18.48 0.156 0.314971 0.148769
26.26 0.208 0.333394 0.149025
34.04 0.254 0.349692 0.149252.
43.47 0.303 0.367053 0.149494
54.32 0.352 0.384414 0.149735
68.35 0.406 0.403546 0.150002
81.15 0.448 0.418426 0.150209
94.93 0.487 0.432244 0.150401
Mean 0.358054 0.149369
4.2.2.2. Specific heat
The specific heat of coconut kernel was determined by using Charm (1971) equation
in the moisture range of 94.93 to 6.15 % d.b. It was found to vary from 1425.246 to
3221.469 J/ kg K, respectively for the above said moisture range. Singh et al (1980) reported
that the specific heat of copra as 0.423 cal / g C in the moisture range of 8 to 12 % w.b. for
copra. The variation of the specific heat with moisture content is shown in Fig.4.7. The
specific heat value calculated by Charm’s equation was very close to the value reported by
Singh (1980) and hence, Charm’s equation could be considered to be quite accurate in
obtaining the specific heat values of copra at various moisture contents. Kaleemullah (2002)
also reported that the specific heat value calculated by Charm’s equation in case of chilies
was less by only 3.19 % and hence, the specific heat values calculated by using the equation
could be considered to be quite accurate.
4.2.2.3. Thermal diffusivity
The thermal diffusivity of coconut kernel was determined by using the values of
thermal conductivity, specific heat and bulk density of coconut kernel using the Eqn. 3.12.
1 2 3
Table 4.16. Thermal diffusivity of coconut kernel
Moisture content, % d.b.
Moisturecontent,
w.b.
Thermalconductivity,(k),W/ m. K
Specific heat (C„,),
J/ kg. K
Bulk Density (Pi.,), kg / ni3
Thermal diffusivity, (a)
m2/s6.15 0.058 0.280249 1425.246 411.674 4.78E-07
11.60 0.104 0.296547 1617.848 418.526 4.38E-07
18.48 0.156 0.314971 1835.572 422.327 4.06 E-07
26.26 0.208 0.333394 2053.296 429.369 3.78 E-07
34.04 0.254 0.349692 2245.898 434.268 3.59 E-07
43.47 0.303 0.367053 2451.061 438.314 3.42 E-07
54.32 0.352 0.384414 2656.224 442.216 3.27 E-07
68.35 0.406 0.403546 2882.322 448.626 3.12 E-07
81.15 0.448 0.418426 3058.176 456.821 3.0 E-07
94.93 0.487 0.432244 3221.469 464.23 2.89 E-07
From the Table 4.16 it is clear that the thermal diffusivity decreased linearly with
increase in moisture content. This phenomenon can be explained by the fact that the bulk
density of coconut kernel increased with the increase of moisture content. A relationship
between thermal diffusivity, a (m2 / s) and moisture content (M) can be expressed as
a = -9E-08 In (M) + 2E-07 -—4.5
with R2 value of 0.99
Verma and Suresh Prasad (2000) reported that the thermal diffusivity of maize
increased with moisture content up to 40 % d.b. and thereafter, decreased with further
increase in moisture content.
4.2.3. Mechanical properties
The results of mechanical properties such as surface area, shell thickness, nut splitting
forces and calorific value of shell and husk were presented and discussed in the following
sections.
4.2.3.1. Surface area
The results of surface area of kernel, weight surface area relationship, copra
weight - nut weight relationship and copra weight - shell weight relationship were given in
the following sections.
4.2.3.1.1. Surface area - kerne! fractionation
The drying rate is directly proportional to the surface area of the produce. By
fractionation of any produce, the total surface area of the produce will increase. The
percentage rate of increase of surface area of kernel by fractionation is given in Table 4. 17.
Table 4.17. Surface area - kernel fractionation
1 2 5
Fractionation. Total surface area
(m2)
Rate of increase
(%)
Two (halves) 0.04881 -
Four (quarters) 0.05444 11.53
Eight (one-eight) 0.06215 27.33
Sixteen (one-sixteen) 0.07205 48.53
Thirty two (one-thirty two) 0.09705 98.83
All values are mean of 5 observations
The relation ship between the surface area of the kernel and fractionation is shown in
Fig. 4.8.
The rate of increase was linear throughout the fractionation range studied. An
equation to predict percentage increase of surface area (SA) with respect to the initial
surface area and with respect to fractionation (F) is represented by the following regression
equation
SA = 0.0016 F + 0.0477 -—4.6
with a R2 value of 0.99
where,
SA= surface area, m2
Bosco (1997) also reported similar linear relation ship between surface area of the
kernel and fractionation in case triangular shaped kernel pieces of East Coast Tall variety of
coconut.
4.2.3.1.2. Weight surface area relationship of kernel! pieces
The surface area of any produce will increase as the weight of the produce of a
particular shape increases. The increase in surface area with respect to weight of a square
shaped kernel was linear and is shown in Fig. 4.9. Narayanan et al (1992) and Bosco (1997)
reported linear relation ship between surface area of the kernel and weight of kernel in case
of East Coast Tall variety of coconut.
The following regression equation was obtained with respect to weight - surface area
relationship of kernel pieces
SA=0.2677 W+ 0.0006 -—4.7
with Revalue of 0.99
where,
SA — surface area, m2
W= weight of kernel, kg
The minimum, maximum and mean surface area of whole nut calculated using the
equation proposed by McCabe el al. (1986) was 0.0621, 0.1218 and 0.0840 m2 respectively.
4.2.3.1.3. Nut weight-copra weight relationship
The split nut (two halves) weight and copra weight relation ship was explored and it
was found to be highly correlated (Fig. 4.10). An equation to predict the copra weight with
respect to nut weight is represented by the following equation
Y = 0.0004X2 0.0907X +134.26 -—4.8
with R2 value of 0.57
where,
Y= copra weight, kg
X= split nut weight, kg
4.2.3.1.4. Copra weight - shell weight relationship
The split nut (two halves) weight and shell weight relation ship was explored and was
found to be highly correlated (Fig. 4.11). An equation to predict the shell weight with
respect to nut weight is represented by the following equation
Z = 0.002X2-1.4453X + 371.42 —4.9
with R2 value of 0.77
where,
Y= shell weight, kg
Z= split nut weight, kg
4.2.3.2. Shell thickness
Arulraj (2002) reported that in the traditional method of splitting nuts using a knife,
it was observed that nut was held opposite to the larger eye (germinating eye) and impact
force was applied. To understand the basic principle, a study was conducted to determine the
shell thickness of the nut along the equatorial circumference. The average value of the shell
thickness was taken from 200 WCT nuts opposite to all the eyes. The thickness of the shell
varied between 2.9 and 7.3 mm. The descriptive statistics of the characters were shown in
Table 4.18.
.no
The relationship between measurements of the thickness of the shell opposite to the
three eyes was explored by fitting different curves such as linear, quadratic, logarithmic and
exponential (Figs. 4.12 and 4.13). In both the cases, the linear fit was found to be the best
based on the R2 values (Tables 4.19 and 4.20). The linear equations fitted for shell thickness
opposite to the larger eye (Y) to the thickness opposite to other eyes (X) are given below
Y = 1.3833 + 0.8721 X (least thickness opposite to other two eyes) —4.10
Y = 1.1433 + 0.8669 X (intermediate thickness observed opposite to other two
eyes) —4.11
Table 4.18. Descriptive statistics of shell thickness
Characters Minimum
(mm)
Maximum
(mm)
Mean
(mm)Std.
Deviation CV (%)
Thickness opposite to larger eye
3.50 7.30 5.35 0.82228 15.37
Least thickness observed opposite to an eye
2.90 6.40 4.55 0.83870 18.44
Thickness observed opposite to the other eye
3.10 6.60 4.85 0.86792 17.88
The measurements on the thickness of the shell opposite to the three eyes were found
to be highly correlated. The correlation between shell thickness opposite to larger eye and
least thickness observed opposite to an eye was 0.889 with the intermediate thickness
of 0.915. Correlation between the shell thicknesses opposite to eyes other than the larger eye
was 0.95.
Table 4.19. Summary statistics on curves fitted to reveal the functional relationship between shell thickness opposite to the larger eye and the least thickness opposite the other two eyes
Curve R-square d .f. Constant Regression coefficients
Linear 0.791 153 1.3833 0.8721 -
Quadratic 0.792 152 0.7653 1.1558 -0.0314
Logarithmic 0.788 153 -0.3468 3.8053 -
Exponential 0.784 153 2.4597 0.1681 -
132
From Table 4.18, it is clear that the thickness of the shell opposite to the larger eye
were maximum but the force required to split the coconut by hitting opposite to the larger
eye was much less compared to splitting the nut by hitting at any other place using the
traditional knife. The reason was that the area opposite to larger eye was the least fracture
zone where in by a small force the maximum thickness was split by the internal pressure of
water which in turn splits other lesser thickness zones. This technique should be popularized
among farming community especially where in women are engaged in splitting of nuts by
applying the impact force opposite to the larger eye.
Table 4.20. Summary statistics on curves fitted to reveal the functional relationship between shell thickness opposite to the larger eye and the intermediate thickness observed opposite to other two eyes
Curve R-square d.f. Constant Regression coefficients
Linear 0.837 153 1.1433 0.8669 -
Quadratic 0.837 152 1.2520 0.8199 0.0049
Logarithmic 0.B29 153 -0.9062 4.0035 -
Exponential 0.832 153 2.3458 0.1674 -
4.2.3.3. Splitting force
Different types of knifes with bevel angle from 15 to 35° were fabricated in a local
workshop using high speed steel and was fitted to the pendulum impact test apparatus to find
the splitting force required. From Fig. 4.14 it is clear that the splitting force required to split
othe nut was in the range of 0.155 to 0.456 N for the knife bevel angles 15 to 35 and the
position of the nut was the side opposite to the larger eye. Eye position has been explained
in detail in section 4.2.3.2. The splitting force was least for 25 bevel angle. To confirm the
results the knife angles were fitted on the newly developed splitting device and splitting tests
were performed to find the shape of nuts and the number of strokes required to split the nuts.
The results are presented in section 4.3.1.3
4.2J.4. Calorific value
The calorific value of coconut shell and husk of different maturity were depicted in
Figs. 4.15 and 4.16. The calorific values of shell of fully matured, matured and immature nut
were in the range of 4598.31 to 4724.61, 4108.68 to 4436.24 and 1856.23
to 2400.36 kcal / kg with mean values of 4640, 4290 and 2150 kcal / kg, respectively
(rounded off to nearest 10 kcal / kg). The calorific values of husk of fully mature, mature
and immature nut were in the range of 2200.18 to 2498.71, 2100.36 to 2203.68 and 1200.36
to 1462.86 kcal / kg with mean values of 2500, 2200 and 1460 kcal / kg, respectively.
4.2.4. Sorption Isotherms
The experimental adsorption and desorption isotherms of copra at 25, 35 and 45 °C,
temperatures and the average equilibrium moisture content at each ERH are given in
Table 4.21.
Figs. 4.17 and 4.18 show the adsorption and desorption isotherms of copra at
_ o25, 35 and 45 C. The reason for selecting the above mentioned temperatures was that in
Kerala the temperature range is 25 - 33 C and in the neighboring coconut growing states of
Tamil Nadu, Karnataka and Andhra Pradesh the temperature is up to 45 °C. All the isotherm
curves followed the same sigmoid shape. Similar experimental EMC values of copra were
reported by Rajashelaran, et al. (1961). The EMC values of copra reported by
Patil et al. (1982) were very low as compared to the present study because it was calculated
indirectly using the equation proposed by Brustillos and Banzon (1949).
4.2.4.I. Effect of temperature on EMC
The EMC values of copra decreased with increased surrounding air temperature in
both adsorption and desorption at constant ERH (Figs. 4.17 and 4.18). The reason may be
that as temperature increased the vapour pressure of the moisture within the copra increased
which in turn increased the transfer of moisture from copra to the surrounding air and hence
Table 4.21. Adsorption and desorption values of copra at different temperatures
Temp. °C RH, %Moisture content, d.b. *
Adsorption Desorption
25 ± 1 12.13 3.2 + 0.05 3.5+0.04
25+ 1 22.70 4.0± 0.07 4.4 ± 0.05
25+ 1 33.60 4.810.08 5.4± 0.12
25 ± 1 43.80 5.2+0.14 5.6± 0.14
25+ 1 64.30 6.2 + 0.12 7.2 ± 0.17
25 + 1 75.56 7.3.±0.17 8.5 ± 0.14
25+ 1 80.30 9.0 + 0.14 9.8 ± 0.17
25 ± 1 86.50 10.6± 0.12 11,2± 0.12
35+1.5 11.67 3.0± 0.06 3.2 ±0.06
35 ± 1.5 21.01 3.8 + 0.08 4.1 ±0.08
35 ± 1.5 32.47 4.6 + 0.12 5.1 ± 0.12
35 ± 1.5 43.48 5.2± 0.08 6.0± 0.06
35 ± 1.5 62.36 6.1 ±0.14 7.1 ± 0.1235 ± 1.5 75.49 7.2+ 0.1 I 7.9± 0.14
35 ± 1.5 79.80 8.4± 0.14 9.0+ 0.16
35 ± 1.5 85.86 9.5 ± 0.12 10.0 ±0.08
45 ± 1.5 11.53 2.8 ±0.02 2.9± 0.06
45 ± 1.5 18.38 3.6± 0.04 3.8 ±0.04
45+ 1.5 31.75 4.2 ±0.06 4.6 ±0.09
45 ± 1.5 43.18 4.8± 0.07 5.4± 0.1245 ± 1.5 60.50 6.0± 0.11 6.9± 0.1445+ 1.5 74.84 6.6± 0.14 7.6± 0.18
45 ± 1.5 79.35 7.9± 0.12 8.4 ±0.1245 ± 1.5 84.58 9.0 ± 0.09 9.4± 0.14
* Mean of three replications
recorded lower values. Similar decrease in EMC values with increase of air temperature was
reported in case of mushroom (Pandey and Aich 1999) and chilies (Wesley et al., 2000 and
Kaleemullah, 2002).
4.2.4.2. Effect of ERH on EMC
The EMC values of copra increased with the increased ERH in both adsorption and
desorption process at constant temperature (Figs. 4.17 and 4.18). This may be due to the fact
that biological materials are hygroscopic in nature and they absorb moisture. If ERH is
increased, copra absorbed moisture and hence recorded higher EMC. Similar type of results
was quoted by Rajashekaran et al. (1961).
4.2.4.3. Sorption hysteresis
The adsorption and desorption isotherm exhibited the phenomenon of hysteresis, in
which the EMC was higher at a particular ERH for desorption curve than for adsorption
(Figs. 4.19 and 4.20). Similar type of results was quoted by Rajashekaran et al. (1961). The
hysteresis (desorption moisture content minus adsorption moisture content) existed over the
entire ERH value. The hysteresis values increased gradually up to 70 % ERH and
afterwards, the values sharply decreased.
4.3. Nut Splitting Device
The design and fabrication of the nut splitting device and the results obtained for
splitting of coconut are discussed in the following sections.
4.3.1. Development of a manually operated nut splitting device
In order to overcome the problems faced by the farmers in splitting of coconuts and
to reduce the drudgery and need for semi-skilled labour a simple manually operated splitting
device was designed, fabricated and field tested. The design drawing of the splitting device
is given in Fig. 4.21.
4.3.1.1. Main frame
The main components of the splitting device designed and fabricated are given
below briefly. The materials required for fabrication with cost of each component is given in
Appendix-II
4.3.1.2. Nut holder
The nut holder’s diameter was fixed at the upper level of 9.5 cm based on the
average diameter for all the 200 observations. The average diameter of the nut was 9.47 cm
(Table 4.22). During tests conducted it was found that the nut holder was effective in
holding all sizes of nuts.
Table 4.22. Length and diameter of WCT coconuts (n = 200)
Character Length Diameter.
Mean 10.1095 9.4730
SD 0.6939 0.7751
4.3.1.3. Knife bevel angle
The ANOVA (Table 4.23) of observations on number of nuts broken in single stroke
and number of round-shaped split nuts showed significant difference among the knife bevel
angles studied. The Duncan’s Multiple Range Tests (Table 4.24) revealed that the bevel
angle of 25 was the best which was significantly different form others with regard to the
characteristics studied. Next best bevel angle was 30°.
Table 4.23. Summary ANOVA on number of nuts broken in single stroke and round-
shape
Source DF Mean Square
Single stroke Round shapeBetween Bevel angles 4 382.16** 194.06**
Error 20 5.72 14.70
**Significant at 1% level
Table 4.24. Duncan’s Multiple Range Tests on effect of knife angle on effort required and shape of nuts split in the splitting device
Knife bevel angle, degree
Single-stroke* Round-shape*
15 22.4a 32.4a
20 32.4” 33.8a
35 33.6b 34.0a
30 39.2C 39.8 b
25 46.0d 47.4c
*Treatments means with the same letters are not significantly different at DMRT1% level
4.3.1.4. Method of operation
Individual de-husked nuts were placed on the nut holder. Impact force was applied
through the knife attached to a handle. By a small impact force (0.155 N) the nut was split.
The split nut automatically rolled side ways where in it got collected in a basket. The nut
water drained into a collecting chamber from where it moved in to a bucket though a pipe.
The advantage of this device is that any unskilled person can operate with less strain and
chances for hand injury are almost eliminated. Thus, one of the constraints faced by farmers
has been solved by this new, simple, low cost device.
4.3.2. Performance evaluation of the nut splitting device
The splitting device fitted with 25° knife bevel angle was tested for its performance.
The average time taken to split one nut was seven seconds. The coconut water collected
could be used for preparing Nata-de-coco or vinegar as it was collected hygienically.
4.3.3. Comparison of splitting device with traditional methods of splitting
The splitting device developed was compared with the traditional method of
splitting. The relative advantages and disadvantages are given in Table 4.25
The output of splitting device was slightly more than manual method using
traditional knife, easy to operate and chances of injury are very less. Physical strength and
drudgery required is comparatively less as compared to the manual splitting. One of the
major constraint reported in splitting of nuts using a traditional knife was bending posture as
nuts are to be picked from the ground where as in the splitting device developed the posture
is straight and the nuts were taken from the basket kept near the operator. From the
Table 4.25 it is evident that the splitting device could remove the requirement of skilled
labour which was also one of the major constraints faced by the coconut farming community
in Kerala.
4.3.4. Comparative cost economics of splitting nuts
The cost of nut splitting device was Rs 2000.00 (Appendix-V). The cost involved to
split 1000 nuts in the newly developed splitting device was worked out using standard
procedures and found to be Rs. 5 / 1000 nuts if the labour is common for splitting and
drying, other wise the cost will be Rs. 36.25 / 1000 nuts. The cost of splitting using semi
skilled labour will be Rs. 31.25 /1000 nuts. The cost of splitting can be reduced if automatic
feeding device is provided. As the shape of the split nut through the splitting device was
round, copra will fetch higher price in the market.
4.4. De-Shelling Machine
In order to overcome the problems faced by the farmers and large scale processing
units in de-shelling, a simple machine was designed, fabricated and tested. The design of the
de-sheiling machine and the results obtained are presented below and discussed.
4.4.1. Development of de-shelling machine
The details of the main components of the de-shelling machine designed and
fabricated are given below. The design details and specifications are given in
Appendix -VIII.
4.4.1.2. De-shelling chamber
The dimensions of the cylindrical de-shelling chamber were calculated as 90 cm
diameter and 115 cm length to hold 200 split coconuts at an average moisture content
of 35 % d.b. based on bulk density and porosity already calculated and reported. Loading
and unloading of partially dried copra was done manually. The design drawing of the
de-shelling machine is depicted in Fig. 4.22.
4.4.1.3. Flights
Three flights having 29 x 22 cm2 cross section running throughout the length of the
drying chamber were fixed inside the de-shelling chamber at a distance separated by 120°.
The lip angle of the flight was fixed at 70° based on preliminary tests conducted, so that the
nuts can fall freely from the flight uniformly in the de-shelling chamber.
4.4.1.4. Motor
The horse power required to rotate the cylindrical de-shelling chamber at 10 RPM
was calculated as 3.0. To achieve the required low speed a reduction gear box with a set of
pulleys were fixed. A 3.0 phase 3 hp electric motor was coupled to rotate the de-shelling
chamber at the desired RPM.
4.4.2. Fabrication of major components
The materials required for fabrication with cost of each component is given in
Appendix-111
4.4.2.1. Main axle
The main axle was fabricated using 60 mm (<j> OD) mild steel shaft of 1.75 m length
which was provided with 2 bearings at the ends and one pulley to rotate the de-shelling
chamber freely.
4.4.2.2. Main support frame
The main support frame was fabricated using 75 x75 x8 mm mild steel angle to
withstand the load of the de-shelling chamber and copra. The height of the de-shelling
chamber was fixed at 110 cm for easy loading and unloading of partially dried copra from
the newly developed copra dryer (reported under section 4.7).
4.4.3. Method of operation
Partially dried copra was loaded in to the de-shelling chamber manually. The
capacity of the de-shelling chamber was 400 partially dried half cups. During the drying
tests conducted, the de- shelling machine was kept close to the dryer developed
(reported under section 4.7). When the average moisture content of the copra reached
35 % d.b. the door of the dryer and de-shelling chamber were opened and copra was loaded
directly from the dryer. The de-shelling chamber was allowed to rotate at a speed of 10 RPM
for different periods and the de-shelling efficiency was calculated.
4.4.4. Effect of moisture content on de-shelling efficiency
Copra was obtained from the drying tests conducted in the dryer designed and
developed. The de-shelling machine was tested for its performance evaluation with partially
dried copra having moisture content in the range of 66.4 to 25.7 % d.b. and it was found that
the optimum average moisture content for de-shelling was 35 % d.b. At 35 % (d.b.) moisture
content the deshelling efficiency was 82.5 %. Five replicate tests were conducted for each
moisture content and the average values were used for calculations. The increase in
de-shelling efficiency with respect to decrease in moisture content is given in Fig. 4.23.
The relationship existing between moisture content (M) and de-shelling efficiency
(De) is non linear and can be represented by a regression equation
D, =215.46exp0-0338A,/ —4.12
with a R2 value of 0.96
where
De~ de-shelling efficiency, %
M= moisture content of partially dried copra, % d.b.
The effect of number of rotations on the de-shelling efficiency were explored by
keeping the moisture content constant at 35 % d.b. Nuts (400 half cups) were loaded
manually and the machine was operated. It was observed that after 40 rotations practically
there was no further increase in the efficiency of de-shelling (Fig. 4.24). On careful
examination it was also observed that unless the coconuts are of the same maturity the
efficiency will not increase due to uneven moisture content in the partially dried copra.
At 30 rotations the efficiency of the machine was 85.64 %, at 40 rotations the efficiency was
and 92.16 % and this further increased to 93.5 % by rotating 50 and 60 times. Hence the
optimum number of rotations was fixed as 40. Thus the deshelling time based on the speed
of the reduction gear output was four minutes.
The increase in de-shelling efficiency with respect to number of rotations is given in
Fig. 4.24, The relationship between number of rotations (R) and de-shelling efficiency (Dc)
is non linear and represented by a regression equation
4.4.5. Effect of number of rotations 011 de-shelling efficiency
4.4.6. Comparative cost economics of de-shelling
The cost of de-shelling machine was worked out to be Rs. 27,100.00 (Appendix-IV).
The cost of de-shelling 1000 nuts was worked out using standard procedures and found to be
Rs. 53.00 / 1000 nuts, if labour available for copra drying is diverted for de-shelling other
wise the cost will be Rs. 62.00/ 1000 nuts (Appendix-VI). The cost involved in deshelling
using human labour was Rs. 36.00 / 1000 nuts but the time taken is more than four times as
compared to machine. Hence the use of de - shelling machine will reduce the drying time
drastically and improve the quality of copra.
4.5. Thin-Layer Drying
The drying characteristics of coconut, quality characteristics of copra and the
applicability of different thin-layer drying models were discussed in the following sections.
4.5.1. Effect of drying air temperature on drying time and moisture content of
coconut
The moisture content of coconut reduced exponentially as the drying time increased
(Fig. 4.25). The free moisture content available was more in the case of coconut dried at low
temperature (50 C) than the one dried at high temperature (100 °C) at the same drying time.
Reduction in moisture content of coconut at any point of time increased with increase in
temperature of the drying air from 50 to 100 C. Similar type of result was quoted by
Gurate et al. (1996) in the case of coconut in the temperature range of 40 to 100 C.
The drying data indicated that the coconut dried at higher temperature dried at faster rate as
compared to the coconut dried at lower temperatures (Fig. 4.25.)
It took 85, 63, 43, 29, 27 and 19 h to dry the coconut from its initial average moisture
content of 81.81% d.b. to the final moisture content of around 6.25 % d.b. at 50, 60, 70, 80,
90, and 100 C of hot air temperature, respectively. Gurate et al. (1996) reported that the
average drying time to reach the 7 % moisture content (w.b.) from the initial range of
45 - 47 % w.b. was 110, 79, 58, 46, 34, 21 and 18 h at drying temperatures of 40, 50, 60, 70,
80 ,90, and 100 C, respectively in thin layer drying experiment. Patil etal. (1982) reported a
drying time of 65, 70 and 80 h to dry coconut from the initial moisture content of
86 % d.b. to 6 % d.b. under thin layer drying under the sun on black painted Palmyra mat,
jute cloth and coconut floor respectively. He also reported a drying time of 50 h in solar
cabinet dryer.
4.5.2. Effect of drying air temperature on drying time and drying rate of coconut
The drying rate of coconut was 6.37, 8.94, 12.32, 15.15, 15.15, and 15.15 % d.b. /h
in the first hour and 0.88, 1.197, 1.75, 2.60, 2.79 and 4.09 % d.b. / h in the final stage of
drying at 50, 60, 70, 80, 90 and 100 C of hot air temperatures, respectively (Fig. 4.26.).
It may be due to the fact that the coconuts were having high moisture content in the order of
80 % d.b. and above in the initial period and was only 6.25 % d.b. in the final stages of
drying. From Fig. 4.26, it is clear that the constant rate period was absent and that for the
entire duration, the drying of coconut took place under the falling rate period at all the
drying air temperatures. Similar type of results was quoted by (Rachmat, 1999) where the
drying rate was 2 % w.b. / h at drying temperature of 50 to 60 C. The constant drying rate
period was absent due to variation in the moisture migration from inner to outer surface by
kernel which is a biological property of coconut.
The drying rate was more up to an average drying time of 14 h in the case of samples
dried at higher temperatures and the phenomena was quite opposite after that time. The
reason may be that at higher drying air temperatures, more moisture was removed in shorter
time and the free moisture available in coconut was less at latter stages. Less moisture was
lost in the initial drying periods of coconut dried at lower drying air temperature compared
to the one dried at higher temperatures. Hence, the free moisture available in coconut dried
at lower temperatures was more compared to the one dried at higher temperature which
resulted in longer time for drying with drying rate gradually decreasing. Also in the initial
stages of drying, the moisture content at the surface of the material which is in a condition of
free moisture was easy to evaporate while at the last stages of drying, the drying rate is
reduced because water availability is less and capillary forces from the inner to the surface
parts of the copra may be lower. This condition resulted in lesser water evaporation from the
surface. Similar results have been reported by Rajashekaran et al. (1961) and
Xiaoren (1989). They reported that in the initial stages of drying the rate of moisture
Fig. 4.25. Effect of drying air temperature on drying time and moisture content of coconut in thin layer drying
D r y i n g t i m e , I i
Fig. 4.26. Effect of drying air temperature on drying time and drying rate of coconut in thin layer drying
movement from the interior of the coconut meat to the surface was sufficiently high to
maintain the surface in a completely wetted condition. Moisture movement is known to be
dependent on diffusion and capillary flow.
4.5.3. Effect of drying air temperature on moisture content and drying rate of coconut
A plot between the drying rate and the average moisture content of coconut at
different drying air temperatures is shown in Fig. 4.27. The drying rate was more for
coconut dried at higher temperature than the one dried at lower temperature for the same
average moisture content of the coconut. The reason was that at higher temperature, the
relative humidity of the hot air was less as compared to the one at lower temperature.
Because of this, the difference in the partial vapor pressure between coconut and the
surrounding environment at higher temperature will be more as compared to the one at
lower temperature. Hence, the moisture transfer rate is more at higher drying air
temperature.
Like many of the agricultural products, coconut did not exhibit a constant rate drying
period. The entire drying took place in the falling rate period. At the beginning, when
moisture was high, drying rate was very high, and as moisture content approached to
equilibrium moisture content, drying rate was very low (Fig. 4.22). Similar results have been
reported by Gurate et al. (1996) and Rajashekaran et al. (1961) in case of coconut dried in
thin layer. Constant rate drying during early stages of drying was not observed as in other
studies (Cardenas, 1968). This may be due to the fact that the halves may have dried a little
during sample preparation (Lozada, 1978). This is in confirmation with other high moisture
content crops like onions (Mazza and Maguer, 1984), garlic (Madamba et al., 1996) and
lettuce and cauliflower leaves (Lopez 2000).
4.5.4. Effect of drying air temperature and drying time on moisture ratio of coconut
Curves of moisture ratio versus drying time for the different drying air temperatures
are shown in Fig. 4.28. The moisture ratio of coconut reduced exponentially as the drying
1 5 2
Average moisture content, % d.b./ h
Fig. 4.27. Effect of drying air temperature on average moisture content and drying rate of coconut in thin layer drying
Time, Ii
Fig. 4.28. Effect of drying air temperature on drying time and moisture ratio of coconut in thin layer drying
time increased. Reduction of moisture ratio from 0.922 to 0.070 occurred in 85, 63, 43, 29,
27 and 19 h at 50, 60, 70, 80, 90, and 100 C drying air temperatures, respectively. In these
curves, an increase of drying rate, exhibited by the slope of the curve, with increase in
temperature was observed. In general, moisture ratio reduced as the tissue dried out and this
attributed to the shrinkage of the cell structure and low concentration which may have
reduced the water diffusion coefficient as in the case of coconut (Rajashekaran et al, 1961)
scalded potato (Fish, 1958), onions (Mazza and Le Maguer, 1984) garlic
(Madamba et al, 1996) and) in lettuce and cauliflower leaves (Lopez et al, 2000).
4.5.5. Effect of drying characteristics of coconut in hot air oven
The drying time decreased exponentially with the increase in drying temperature.
Drying in hot air oven showed different drying characteristics. The influence of relative
humidity has drastic effect as the coconut is not exposed to the atmosphere as in the case of
thin layer drying. The free moisture content available was more in the case of coconut dried
at low temperature (50 C) than the one dried at high temperature (100 C) for the same
duration of drying (first 17 h of drying). However, reduction in moisture content of coconut
at any point of time was similar for the drying air temperature of 50 to 60 °C up to a period
of 17 h (Figs. 4.29 to 4.32). There after the loss of moisture increased in case of drying air
temperature range of 70 to 100 C. This may be due to the fact that in a hot air oven the
relative humidity is constant as drying takes place in a closed container and the kernel must
have dried at faster rate at higher temperature there by forming a hard outer layer which
restricted the passage of moisture for some time there by decreasing the drying rate.
It took 68, 51, 46, 40, 38 and 35 h to dry the coconut from its initial average moisture
content of above 80 % d.b. to the final average moisture content of around 6.25 % d.b. at 50,
60, 70, 80, 90, and 100 C of hot air temperature and relative humidity inside the hot air
oven of 66, 54, 48, 44, 36 and 32 %, respectively. The drying rate was higher for
temperatures 50 and 60o C up to 17 hours and later on it decreased where as in case of
drying air temperature of 70-100o C the drying rate was almost constant after 11 h of drying.
The drying rate was in the falling rate period up to 20 h and later on the entire drying
took place in the constant rate period. Similar results were reported by Patil et al. (1982) and
Hammonds et al. (1991). It was reported that two distinct periods of drying i.e. falling rate
period and constant rate period exists in thin layer drying of coconut. Similar results were
also reported by Drew, et al. (1993). The falling rate period also showed change in line
which may be due to change in mechanism of moisture migration from .capillary to
diffusion. Thus if coconut is dried under a constant relative humidity the drying
characteristics will be different. These times were also longer than those reported in
literature. Palmer, (1968) and Dumaluan et al. (1982) reported a drying time of 18 to 20 h
o .....
at 60 C. Sreenarayanan et al. (1989) reported a drying time of 20 h in a mechanical dryer to
reduce the moisture content from 50 % w.b. to 7 % w.b. at a drying air temperature of 65 °C.
4.5.6. Effect of air velocity on drying characteristics of coconut in thin layer drying at constant air velocity of 1.0 in / s
The drying characteristics of coconut at drying air velocity of 1 m / s and drying air
temperature of 50, 60, 70, 80, 90 and 100 °C are shown in Fig. 4.33. The moisture content of
coconut reduced exponentially as the drying time increased in all the cases. It took
69, 53, 48, 42, 39 and 35 h to reduce the moisture content from initial 81.81 to 6.25 % d.b. at
drying air temperatures of 50, 60, 70, 80, 90 and 100 °C, respectively. The drying rate of
coconut in the first hour was 6.37, 6.37, 15.15, 15.15 and 15.15 % d.b. / h and 1.17, 1.28,
1.83, 2.35, 3.01 and 3.64 in the last period at drying air temperatures of 50, 60, 70, 80, 90
and 100 °C, respectively (Figs. 4.34 and 4.35). The entire drying took place in the falling
rate period and there was no constant rate period of drying.
Reduction of moisture ratio from 0.9 to 0.06 occurred in 69, 53, 48, 42, 39 and 35 h
at drying air temperature of 50, 60, 70, 80, 90 and 100 C ( Fig. 4.36). At air velocity of
1.5 m / s it took 72, 55, 47, 43, 38 and 36 hours to dry coconut from initial moisture content
of 81.81 % d.b. to final moisture content of 6.25 % d.b. at drying air temperature of 50, 60,
70, 80, 90 and 100 C hot air temperature, respectively (Table 4.27). At 0.5 m / s air
velocity it took 85, 63, 43, 29, 27 and 19 h to dry the coconut from its initial average
moisture content of 81.81% d.b. to the final moisture content of around 6.25 % d.b. at
50, 60, 70, 80, 90, and 100 °C of hot air temperature, respectively (Fig. 4.25). From this it is
clear that air velocity has no effect on thin layer drying of coconut above 70 “C. This was
probably due to the fact that the heat bearing fluid escaped to the atmosphere at a faster rate
there by reducing the total contact time of heat bearing fluid with the coconut. This is also
clear from Table 4.26 (ANOVA) and Table 4.27 (DMRT) to substantiate the results. From
the above, it is clear that the moisture reduction, drying rate did not increase with increase in
air velocity in thin layer drying within the air velocity range studied.
Table 4.26. Summary ANOVA showing the mean sums of squares (MS) of drying time in four drying methods tested in six drying temperature in thin layer
Sources of df Drying Temperaturevariation
uoo
60 °C 70 °C 80 °C 90 °C 100 °cDryingmethod 3 299.7** 138.3** 23.3 208.3** 161.7** 334.5°**
Error 16 11.7 5 7.5 6 5.625 5.125
** Significant at I %
Adeyemo (1993) reported that the air How rate depends on the stage at which drying
occurs. It was reported that the effect of air flow rate was more effective at the constant rate
period compared to falling rate period since the internal resistance of moisture movement in
the falling rate period compares with the surface mass transfer resistance which makes the
air flow rate to have little effect on the drying rate. In a review of thin-layer drying,
Jayas et al. (1991) concluded that the air velocity has little effect on drying grains.
Shivhare et al. (1995) and Kaleemullah (2002) also concluded similar type of result in the
case of chilies dried in thin layer dryer.
Table 4.27. Duncan’s Multiple Range Test for grouping of drying methods at different
drying air temperature in thin layer
Drying methods Drying temperature50 °C 60 °C 70 °C 80 °C 90 °C 100 °C
Oven 68.0“ 51.0“ 46.0a 40.0a 38.0“ 35.0a
Air velocity (1 in / s)
69.0a 53.0ab 48.0a 42.0a 39.0a 35.0a
Air velocity (1.5 in / s) 72.0“ 55.0b 47.0a 43.0a 3 8.0a 36.0a
Thin layer (0.5 m / s)
85.0C 63.0° 43.0 a 29.0b 27.0 b 19.0c
4. 5.7. Effect of drying air temperature on quality characteristics of copra
The effect of drying air temperature from 30 to 120 °C on quality characteristics of
copra and oil content was explored. It was found that per cent oil content has no significant
difference among the treatments. The average oil content (%) varied between
60.3 to 63.9 % d.b. (Table 4.28). Statistical analysis of quality characteristics of coconut oil
(Table 4.28) at various drying air temperature revealed that there is no significant difference
in saponification number where as there was significant difference at 1% level for free fatty
acid, acid value and peroxide value. The mean saponification numbers were within the
ranges given by Kaufmann et al. (1956) and Banzon et al. (1982). Based on the FFA
content, the coconut oil extracted from different samples obtained during drying
experiments were of good quality (Varnakulasingam, 1974).
Rajashekaran et al. (1961) reported similar results stating that drying air temperature
has no significant difference at drying air temperature range of 55-70 °C on oil content.
Wuidart et al. (1978) reported that oil content was not the same at all places in the coconut
and is 61.8 % near the embryo, 69.3 % in the opposite pole of embryo and 70.7 % in other
parts of copra. Sreenarayanan el al. (1989) reported 65 % oil content in case East Coast Tall
variety of copra dried in a mechanical dryer.
Gurate el al. (1996) reported oil content of 60.08 to 64.95 % d.b. in case of Ivory
Coast local variety. Barrios et al. (1990) reported that in the traditional coconut oil
extraction process with expellers, the dried meat contained 53-60 % oil on dry basis and
varied with coconut variety and maturity.
1 8 1
Table 4.28. Duncan’s Multiple Range Test for grouping of drying air temperatures
based on quality characteristics
'■^Treatment means with the same letters are not significantly different at DMRT1% level
The analysis of variance of oil content of samples dried in thin layer drying is
presented in Table 4.29. From the Table 4.29 it is evident that there was no significant
difference in oil content with respect to drying air temperatures within the temperatures
range studied.
Table 4.29. Summary ANOVA sliowing the mean sums of squares (MS) of different
quality characteristics of oil
Sources of variation df Oil content Free fatty
%, d.b. acid, %
AcidValue,
mgKOH/g
Peroxide Value,
milli.eq. peroxide/ kg.
Saponificationnumber,
mg KOH/g
Drying T emperature
9 5.941 0.045** 0.009** 0.004** 5.680
Error 40 4.570 0.001 0.002 0.001 2.763
*K * Significant at 1%
From the ANOVA (Table 4.29) it can also be seen that the drying air temperature is
significantly different for FFA, PV, and AV but are well within the range prescribed. In case
of oil content and saponification values no significant difference among different drying air
temperatures was noticed. Hence it can be fairly concluded that drying air temperature up to
80 °C does not have any effect on oil content.
The summary DMRT of different coconut oil quality indicators is presented in
Table 4.30. From the Table 4.30 it is clear that taste of copra were not affected by drying air
temperature up to 80 °C. Smell was not affected up to 90 °C where as colour slightly dulled
at 80 °C and was brown at 90 °C, dark brown at 100 to 110 °C and black at 120 °C. Similar
results were reported by Guarte et al. (1996) for colour of copra in thin layer drying of
coconut. Copra dried at 30 to 40 °C was affected by fungal growth. The browning of copra
at high temperature is mainly attributed the Millard reaction. Since pale white to light brown
copra is traded commercially and there has been no report on loss of copra quality even at
light brown colour. From this it can be deduced that based on sensory qualities of colour,
smell and taste, copra can be dried up to 80 °C without loosing its quality. Thus sun drying
in Kerala has to be restricted to those months when the temperature is above 33 °C.
4.5.8 Effect of drying air temperature on fatty acid composition of coconut oil
The fatty acid composition (Table 4.31) did not vary significantly due to change in
copra drying temperature. Variations observed in fatty acid composition did not follow any
specific pattern with reference to copra drying temperatures. This indicates the genetic
variability in the sample composition itself but not due to the effect of temperature. Similar
results were reported by Guarte et al. (1996) in case of thin layer drying of coconut in the
temperature range of 20 - 100 °C. The fatty acid composition of coconut oil observed in this
experiment followed the typical fatty acid composition of coconut oil. According to Oo and
Stumpf 1970) and Naresh et al. (2000), coconut oil contains 48 % of lauric acid (Cl2:0)
followed by Myristic acid (Cl4:0) which was about 20 %. Gas chromatography on oil
samples indicated that coconut oil mainly contained the saturated fatty acids viz., C6:0
(Caproic), C8:0 (Caprylic), C10:0 (Capric), C12:0 (Lauric), C14:0 (Myristic), C16:0
(Palmitic), C l8:0 (Stearic) and C20:0 (Arachidic). It also contained the unsaturated fatty
acids viz., C 18: I (Oleic) and C18:2 (Linoleic) in smaller quantities (Table 4.31).
Table 4.32. Summary ANOVA showing the mean sum of squares (MS) of fatty acid composition of coconut oil extracted from copra dried at various temperatures
Sources of Fatty acid composition
variation C6:0 C8:0 00:0 C12:0 04:0 C16:0 08:0 08:1 08:2 €20:0
Temperature 40 to 110°C
0.003 0.106 0.142 1.53 0.468 0.286 0.032 1.400 0.056 0.000
Error 0.003 0.425 0.249 2.66 1.34 0.473 0.133 0.974 0.027 0.000
In the present study, all oil samples had typical fatty acid composition with above
mentioned fatty acids (Table 4.31). This indicates that the drying temperatures did not
influence the quality of oil, in terms of fatty acid composition. Summary ANOVA
(Table 4.32) showed the mean sum of squares (MS) of fatty acid composition of coconut oil
extracted from copra dried at various temperatures indicated no significant difference for
values of fatty acid compositions among different temperatures. Hence, it can be concluded
that drying air temperature has no effect on fatty acid composition of coconut oil in the
temperature range studied.
4.5.9. Optimization of drying air parameters
The drying air parameters of coconut was optimized based on the drying time, smell,
taste, colour, fungal growth, oil and free fatty acid content. The scores obtained at each
drying air temperature are presented in Table 4.33 based on scores given to quality attributes
(Table 3.3). Though the coconut dried at 100 °C gave a good score of 9 for drying time,
it gave a poor score of 3 for colour and 7 for smell, taste and colour FFA and oil content.
It gave a very poor score of 1 for drying time. The coconuts dried at 80 °C scored a total of
53 points, the highest score where as the coconuts dried at 90 and 100 °C scored 49 points.
Hence based on the above parameters 80 °C drying air temperature can be adjudged as
optimum. The coconut dried at 80 °C was light brown in colour which fetches equally good
price in the whole sale market.
160
Table 4.33. Scores obtained for dried copra (dried at different temperatures) based oil drying time, smell, taste, colour, oil content, and free fatty acid_________________________. ..
Drying airtemperature,
°C
Score TotalscoreDrying
TimeSmell Taste Colour Fungal 1
growth |Oil
contentFree fatty
acid50 1 9 9 9 9 7 7 5160 I 9 9 9 9 7 7 5170 I 9 9 9 9 7 7 5180 5 9 9 7 9 7 7 5390 5 9 7 5 9 7 7 49100 9 7 7 3 9 7 7 49
From Duncan’s Multiple Range Test for grouping quality characteristics of copra as
affected by drying air temperature in thin layer drying (Table 4.31), it can be seen that
drying air temperature of 80 °C is best succeeded by 70 °C. Hence the optimized
temperature was 80 °C.
4.5.10. Models
Reduction of moisture ratio during the drying process at different temperatures under
thin layer drying was characterized in terms of appropriate functions. Besides the standard
models viz., Lewis, Hustrulid and Flikke and Page, other models (both linear as well as non
linear) were also tried. The best fitted model was selected based on variability explained by
a model, the pattern of residuals, standard error of estimates (Es) and mean relative
percentage deviation (Em). After fitting models at different temperatures, a general model
that characterizes the moisture ratio at any temperature between 50 to 100 °C was also
proposed.
Estimated parameters of the fitted models of moisture ratio at 50 to 100 °C is shown
in Table 4.34 and the proportion of variation explained and the Es and Em statistics were
shown in Table 4.35. The pattern of residuals against predicted moisture ratio from
50 to 100 °C is shown in Figs. 4.37 to 4.48.
Irrespective of the drying air temperature, the standard models of moisture ratio
showed systematic bias especially at the initial period of drying. A close examination of the
data suggests that the rate of reduction of moisture at the beginning and end of drying
periods was different. Therefore it was decided to use models involving two terms. Based on
the lowest values of the statistics Es and Em (Table 4.36), and pattern of residuals
(Figs. 4.37 to 4.48) the best fitted model was selected for each temperature. Except for
50 °C, the best fitted model was Model-5, which is a sum of two Hustrulid & Flikke models.
Though the same model was good enough to describe the pattern of moisture reduction at
50 °C, model-6 was slightly better, which is a sum Hustrulid and Flikke model and a
modified Page’s model. The non-linear fit Page model did not converge for temperature
100 °C for the chosen ‘initial value’ used as starting value for the iteration process. The best
fitted model is given below
MR = A exp (-kx) + B exp (-lx)
The R2 of the best fitted models for temperature 50 to 100 °C varied from
0.996 to 0.999 indicating that the models explained the variation in moisture ratio
satisfactorily. The values of the Eltl and Es statistics were also less for the fitted model but
not lowest in some cases. Further the pattern of residuals of the fitted models was not
showing any systematic variation as well.
When temperature was included in the model, the best fitted model was one which
was found to have three components, the first term is a Hustrulid and Flikke model; the
second term is a product of modified Page’s model (Model- 6) and a linear function of
temperature, and the third term is quadratic for temperature multiplied by a Hustrulid and
Flikke model. The model is quadratic for temperature indicating temperature beyond certain
limit will not yield faster reduction. The R2 of the best fitted model was 0.98 and has the
lowest values of Em and Es statistics as well (Table 4.37). The fitted curve of this model
along with the original observations on moisture ratio at different temperatures is shown in
Figs. 4.49 and 4.50. The best fitted model is given below
MR = A exp (-lex) + B t exp (-1/x) + C t2 exp (-mx)
Table 4.35. Estimated values of mean relative percentage deviation (Em), standard error of estimates (Es), R2, Residual plot pattern of different models used to describe thin layer drying of coconut from 50 to 100 C
Temperature ModelNo.
R2 Em Es Residual plotpattern
1 0.980 -1.00 0.0421 Systematic2 0.841 15.12 0.0751 Systematic3 0.966 3.27 0.0347 Systematic
SO C 4 0.991 -2.56 0.0170 Systematic5 0.998 -1.12 0.0068 Random6 0.999 0.30 0.0049 Random1 0.996 0.05 0.0158 Random2 0.924 5.38 0.0637 Systematic
60 C 3 0.976 1.88 0.0355 Systematic4 0.994 -0.48 0.0174 Random5 0.996 -0.59 0.0149 Random6 0.992 -0.67 0.0201 Random1 0.991 -0.05 0.0278 Random2 0.921 7.65 0.0798 Systematic
70 °C 3 0.969 3.13 0.0498 Systematic4 0^992 1 0.25 0.0253 Random5 0.997 -0.46 0.0160 Random6 0.990 -0.65 0.0287 Random1 0.992 0.23 0.0309 Random2 0.886 24.63 0.1067 Systematic3 0.952 11.14 0.0702 Systematic
80 C 4 0.988 3.46 0.0348 Random5 0.999 -0.35 0.0092 Random6 0.995 -1.02 0.0220 Random1 0.995 0.22 0.0271 Random2 0.913 23.78 0.0993 Systematic3 0.958 12.14 0.0702 Systematic
90 “C 4 0.987 4.32 0.0395 Systematic5 0.999 -0.31 0.0103 Random6 0.997 -0.87 0.0201 Random1 0.999 0.05 0.0163 Random2 0.975 16.36 0.0690 Systematic3 0.983 11.41 0.0597 Systematic
100 C 4 - - - -5 0.999 -0.03 0.0115 Random6 0.999 -0.12 0.0113 Random |
Table 4.36. Estimated values of parameters of proposed models used to describe thin layer drying of coconut from 50 -100 °C drying air temperature
Table 4.37. Estimated values of mean relative percentage deviation (Em), standard error of estimates (Es), R2 and Residual plot pattern of proposed models used to describe thin layer drying of coconut from SO to 100 °C
The drying kinetics of coconut in different layers of deep bed and the suitability of
different models are discussed in the following sections.
4.6.1. Effect of layer position and drying time on moisture content of coconut
The moisture content of coconut reduced exponentially as the drying time increased
(Fig. 4.51). The free moisture content available was more in the case of coconut placed at
top layer than the one placed at bottom layer at any point of time during drying. Reduction
in moisture content of coconut at any time decreased as the bed layer moved from bottom to
top layer. It took 21, 26, 29 and 35 h to dry the coconut in bottom layer, 15 cm, 30 cm and
40 cm depth, respectively from its initial moisture content of around 94.32 % d.b. to the
final moisture content of around 6.25 % d.b. at 80 °C of drying air temperature.
It clearly shows that the coconuts kept at bottom layer dried quickly as compared to
the one kept at top layer (Fig. 4.51). The reason is that the hot air directly conies and
contacts the bottom layer first and removes more moisture from it. The moisture that is
removed from each layer is added to the drying air and thereby changing the psychometric
properties of air, which ultimately reduces the temperature of hot air and increases the
relative humidity. Drying took 6 h more to dry coconut layer present at 40 cm depth which
indicates that the layer depth has to be restricted to about 30 cm to achieve faster drying
time.
These drying times were longer than those reported in literature. Palmer (1968) and
Dumatuan et al (1982) reported a drying time of 18 to 20 h at drying temperature of 60 °C.
Sreenarayanan et al. (1989) reported a drying time of 20 h in a mechanical dryer to reduce
the moisture content from 50 to 7 % w.b. at a drying temperature of 65 °C.
Annamali et al (1989) reported an average drying time of 33-37 h in a 1000 nut capacity
natural convection dryer at an average drying temperature of 60-70 0 C. Anonymous (1984)
4.6. Deep Bed Drying
reported a drying time of 25-30 h in case of coconut dried in the I.R.H.O hot air copra dryer.
Some of these times were far below the ones in the present study.
Although such differences could not be explained it was assumed that they were due
to experimental techniques. The method of moisture content determination and the thickness
of meat could also account for difference between the past and present results.
Brennldorfer et al (1985) reported that the temperature, moisture content and the physical
dimensions of coconut meat were the most important factors affecting the rate of moisture
removal.
4.6.2. Effect of layer position and drying time on drying rate of coconut
The drying rate of coconut was 14.63, 14.19, 14.18 and 10.31 % d.b. / h in the first
hour and 4.206, 3.49, 3.03 and 2.51 % d.b. / h in the final stages of drying at bottom layer,
15 cm, 30 cm and 40 cm depth, respectively in deep bed drying (Fig. 4.52). It was due to the
fact that the coconuts have high moisture content in the order of 94.32 % d.b at the
beginning and was only 6.12 % d.b. in the final stages of drying. From Fig. 4.52 it is clear
that the constant rate period of drying was absent during the entire period of drying and the
drying of coconut took place under the falling rate period in all the layers of deep bed.
Sudaria et al. (1996) reported an average drying rate of 31.08 kg / h in case of coconuts
dried in improved farmer’s direct type dryer. Similar type of results was quoted by
(Rachmat,1999) where the drying rate was 2 % w.b. / h at a drying temperature of
50 to 60 °C in case of coconut. The absence of constant rate period of drying may be due to
non uniform release / migration of moisture through kernel tissues. In other crops similar
type of result was quoted in the case of red chilies dried in a waste fired drier
(Phirke et al., 1992). The constant rate drying period was absent due to quick removal of
moisture from the pericap (skin) of the chilies.
The drying rate was more up to an average drying time of 6.5 to 8.5 h in the case of
samples dried at bottom layers and the phenomena was quite opposite after that time. The
1 8 2reason was that at bottom layer, more moisture was lost in less time and the free moisture
available in coconut was less at later stages. Less moisture was lost during the initial periods
of drying of coconut at top layer as compared to the one at bottom layer. Hence, the free
moisture available in the coconut present in the top layer was more as compared to the one
dried at bottom layer, which resulted in the higher drying rate at later stages.
4.6.3. Effect of drying air temperature (80 °C) and average moisture content on
cl lying rate in different layers in deep bed drying of coconut.
A plot between the drying rate and the average moisture content of coconut in
different layers of deep bed is shown in Fig. 4.53. The drying rate was more for coconut
dried at bottom layer than the one dried at top layer for the same average moisture content of
the coconut. The reason was that at bottom layer, the relative humidity of the hot air was less
as compared to the one at top layer. The rate of drying decreased gradually as drying period
extended in all the layers. That is at the beginning, when moisture was high, drying rate was
very high, and as moisture content approached to equilibrium moisture content, drying rate
was very low (Fig. 4.53).
4.6.4. Effect of drying time on moisture ratio of coconut in different layers of deep
bed drying
Curves of moisture ratio versus drying time in different layers of coconut of deep
bed drying are shown in Fig. 4.54. The moisture ratio of coconut reduced exponentially as
the drying time increased. Reduction of moisture ratio from 0.86 to around 0.06 occurred in
21, 26, 29 and 35 h at bottom, 15 cm, 30 cm and 40 cm of deep bed, respectively.
During drying, as moisture content decreased, the difference in vapour pressure between the
inner surface of the kernel and out side drying air also decreases. Drying also causes
shrinkage of outer cells. These two factors slow down moisture diffusion. This may be the
reason for decrease in drying rate during extended periods of drying.
183 ‘
4.6.5. Optimization of drying bed depth
From the drying trials conducted at various depth (Fig. 4.51) it can be seen that copra
has to be left for 6 h more to dry the top layer at 40 cm depth. This will adversely affect the
quality of copra at the bottom layer and at 15 cm depth. Thus, to overcome over drying and
browning of copra, it is desirable to restrict the bed thickness to 30 cm and mixing copra
(bottom layer to top layer) after 21 h of initial drying is recommended.
4.6.6. Models
Models fitted for describing the moisture reduction in deep bed drying at various
depths (bottom, 15 cm, 30 cm and 40 cm) are described in this section. Standard models
viz., Lewis, Hustrulid and Flikke, and Page’s as well as modified ones were tried.
Based on variability explained by a model, the pattern of residuals, standard error of
estimates (Es) and mean relative percentage deviation (Em), the models were compared.
The models tried for moisture reduction at various layers of deep bed drying are
given in Table 4.38. The proportion of variation explained and the Es and Em statistics are
shown in Table 4.39. The pattern of residuals against predicted moisture ratio is shown in
Figs. 4.55 to 4.62. The R2 of the best fitted models ranged from 0.993 to 0.998 indicating
that the models explained the variation in moisture ratio satisfactorily. The values of the
Em and Es statistics were also less for the fitted model but not lowest in some cases. Further
the pattern of residuals of the fitted models did not shown any systematic variation as well.
Except for 40 cm depth, the best fitted model for describing moisture ratio was
Model-4, which is a linear combination of two Hustrulid & Flikke: models. The best fitted
model at 40 cm depth was Model-5, which is a sum of Hustrulid and Flikke model and a
modified Page s model. For 15 cm depth the Page’s model did not converge for the chosen
‘initial values’ of the parameters.
Table 4.38. Estimated values of parameters of different models used to describe deep bed drying of coconut from bottom iayer to 40 cm depth at 80 C drying air temperature
ModelNo.
Model Estimated parameter, k/liBottom layer 15 cm 30 cm 40 cm
1 Lewis : MR = exp(-kx) k = 0.162 k = 0.126 lc = 0.112 k =0 .0872 Hustrulid & Flikke:
MR = A exp(-kx)A = 0.964 k = 0.155
A = 0.857 k = 0.106
A = 0.871 k = 0.097
A = 0 .980 k = 0.085
3 Page: MR = exp(-kx“) k = 0.194 n = 0.907
- -
4 MR = A exp(-kx) + B exp(-lx) (Proposed Model)
A =0.148 k = 0.042 B =0.897 1 = 0.219
A =0.589 k =0.076 B =0.503 1=0 .437
A = 0.718 k = 0.081 B= 0.326 1 = 0 .484
A = 0.333 k = 0.085 B = 0 .646 1 = 0.085
5 MR = A exp(-kx) + B exp(-l/x) (Proposed Model)
A = 1.010 k = 0.202 B =0.065 1 = 0.781
A =0.754 k =0 .125 B =0.057 1 =-1.228
A = 0.797 k = 0.103 B = 0 .03 1 1 = -1.482
A =0.985 k = 0.090 B = 0 .015
1=2.121 I
Table 4.39. Estimated values of mean relative percentage deviation (E,„), standard error of estimates (Es), R2 and residua! plot pattern of different models used to describe drying of coconut at bottom layer to 40 cm depth in deep bed drying
Layerposition
Model No. R2 E„i Es Residual plot pattern
Bottom
1 0.987 0.1238 0,0256 Systematic2 0.988 0.1020 0.0250 Systematic3 0.992 0.0578 0.0211 Systematic4 0.998 -0.0034 0.0119 Random5 0.998 -0.0085 0.0120 Random
15 cm
1 0.940 0.1594 0.0511 Systematic2 0.968 0.0723 0.0383 Systematic3 <04 0.993 -0.0184 0.0189 Random5 0.985 -0.0222 0.0270 Random
30 cm
1 0.963 0.1211 0.0413 Systematic2 0.986 0.0444 0.0254 Random34 0.996 -0.0113 0.0133 Random5 0.993 -0.0126 0.0182 Random
40 cm
1 0.997 .0301 0.0136 Random2 0.997 .0188 0.0127 Random3 - - -
4 0.997 .0188 0.0131 Random5 0.998 .0011 0.0125 Random
In order to overcome the problems faced by the fanners (in the small holders and
conventional dryers as per the survey recommendations) a new type of dryer has been
developed. The performance of the dryer, the drying kinetics of coconut, quality
characteristics and the applicability of different drying models are discussed in the following
sections. The design drawing is given in Fig. 4.63.
4.7.1. Design and development of copra dryer
The detailed design calculations of major components are given in Appendix -IX.
Assumptions have been made wherever required. The capacity of the dryer was 1000 nuts.
The fuel used was coconut shell for evaluating the performance. The dryer developed was of
natural draft type, having two separate fuel chambers, which were arranged side by side.
The isometric view of the heating chamber is given in Fig. 4.64.
4.7.1.1. Description of the dryer
The dryer consists of a drying chamber, a burning chamber, a plenum chamber and
ventilation holes. This dryer was fabricated using locally available materials such as
asbestos sheet, galvanized iron sheet, mild steel angle and fire resistant plywood. Asbestos
cement sheet has been provided only at those places where the copra does not come into
direct contact. The contact areas (sides) of the drying chamber with copra are provided with
heat resistant bamboo plywood.
The shape of the burning chamber was designed to avoid the flame and flue gas
coming into direct contact with the copra. This was one of the constraints reported by the
farmers. As smoke did not come into contact the quality of copra obtained was good.
4.7. Copra Dryer
4.7.1.2. Air inlet chamber
This is the most important part of the dryer since fresh air enters into the dryer.
A 20 cm opening from the ground level was provided all around the dryer. Hinges were
provided to regulate the supply of fresh air in to the plenum chamber. The air entering from
the inlet passed to the plenum chamber through the passage provided and got heated up by
the heat radiated from the heating chamber. The hot air being lighter in weight raised and
passed through the drying chamber.
4.7.1.3. Heating cum heat exchanging chamber
Butterfly valves were provided in the exhaust pipes to control the entry of fresh air in
to the dryer thus controlling the inside temperature of the dryer. This in turn regulates the
entry of air for combustion and thus controls the rate of combustion of fuel so that the fuel
burns slowly maintaining the temperature.
Initially only one heating chamber was provided but during tests conducted it was
found that due to improper burning smoke was getting accumulated in the heating chamber
because of which the shells were not burning properly. Hence two fuel chambers, instead of
one was provided. Also it was found during the tests that after de-shelling the volume of
copra almost reduced to half the volume of drying chamber. Hence after de-shelling the
farmer can use only one heating chamber which in turn will reduce the quantity of fuel
required. Also if the farmer has lesser number of nuts than the capacity of the dryer he can
operate only one heating chamber there by reduce the cost of drying.
4.7.1.4. Plenum chamber
The empty space provided above the heating chamber is known as plenum chamber.
A door has been provided mainly for cleaning the top surface of burning chamber. During
loading and unloading lot of coconut pith and small pieces of fibre adhering to the shell fall
on the drying chamber. This has to be cleaned periodically to avoid its burning other wise it
will produce smoke and contaminate the copra quality. Also in case of sudden rise in drying
air temperature, the door can be opened so that fresh air enters there by bringing down the
temperature.
4.7.1.5. Drying chamber
The top portion of the dryer is known as drying chamber. The dimensions of the
drying chamber was 2.25 m (L), 1.5 m (B) and 0.30 m (H).The weld mesh for stacking
copra was made of 10 gauge, 25 x 25 mm size weld mesh. On one side of the drying
chamber a door was provided for easy loading and unloading the coconuts in to the dryer
and into the de-shelling chamber, the height of which is lower than the dryer. The sides of
the burning chamber was covered by 6 mm thick bamboo plywood sheets, and all other parts
of the dryer was covered with 4 mm thick asbestos cement sheet to withstand high
temperatures and to reduce the overall cost of the dryer.
4.7.1.6. Fuel preparation
Eighty half shells were required in each tray to make one row. The coconut shells
were interlocked and laid on the tray. The hollow end of the row of shells was ignited
(a little kerosene being added to help initial burning). When the shells begin to burn well and
without smoke, the tray was pushed inside the heating chamber. The shells burn uniformly
by the incoming air for which ventilation holes were provided on the door. The number of
holes required was standardized based on tests conducted. Additional ventilation door was
provided with provision to open and close in case of necessity and to retain the heat once the
shells were burnt completely.
4.7.1.7. Firing the dryer
Each fuel tray produced heat for 6 h with a temperature of about 80 - 82 °C.
Generally after about 6 h, when the temperature drops below 60 °C, the fuel trays were
removed from the dryer, cleaned and reloaded with fuel, refired and replaced in to the
respective burning chambers. About 4 loads of fuel were required to dry the copra to about
6.25 % d.b. moisture content. The heat generated by burning of the fuel heated the heating
chamber. The air above heating chambers got heated up and moved upwards through the
layers of fresh coconut kernel and the hot air laden with moisture escaped from the top of
the drying chamber in to the atmosphere, and fresh air entered through the ventilation holes
provided at the bottom (Fig. 4.63). This phenomenon was carried out with the help of
natural convection.
4.7 .1.8. Testing of copra dryer
Before conducting an experiment, the burner was loaded with fuel and charged for
30 minutes till the desired drying air temperature attained steady state. The dryer was tested
for production of copra during January - March 2003. After finishing the experiment, one
coconut was again dried in the experimental environmental condition till it recorded a
constant weight. This sample was used to determine the equilibrium moisture content of the
sample in the experimental environmental condition.
For making copra, the coconuts were broken into two halves and kept inverted for
4-5 minutes in order to drain the water. After the coconut water Avas completely drained,
the cups were stacked in the drying chamber, layer by layer in such a way that in the first
two layers, the kernels faced upwards (U) and in the subsequent layers, the kernels faced
downwards (f|). The cups in adjacent layers were stacked in a brick-laying-fashion,
one overlapping the other. Drying was carried out continuously for 24 h by firing the burner
four times at intervals of approximately 6 h.
4 . 7 . 2 . Effect of drying air temperature on drying time and moisture content of
coconut in copra dryer
The moisture content of coconuts reduced exponentially as the drying time increased
(Fig. 4-.65). The percentage moisture content available was more in the case of coconuts
p laced in the top layer as compared to bottom layer at the drying air temperature of 80 °C for
the same drying time. It took 22, 21, 26 and 25 h to dry coconut from the average initial
moisture content of 90.14, 88.34, 92.12 and 86.23 to 6.25 % d.b., respectively in the four
drying tests conducted. The average mean relative humidity of ambient air was 74.12, 72.56,
86.45 and 84.12 %, respectively in the four tests. The average time taken to reduce the
moisture content from 89.21 % d.b to 6.25 % d.b was 23.5 h where as in the small holders
dryer it took 36 to 40 h there by reducing the total drying time by 12.5 to 16.5 h.
Thus almost two batches of copra could be dried in the time taken by smalt holders dryer. In
case of sun drying it took 6-7 days where as in this dryer copra can be dried in one day.
Thus there is a saving of 6 days off time for the farmer during which he can attend to other
useful work. Also to dry 1000 nuts the total area required is only 3 m2 where as in case of
sun drying it required a minimum of 12 m2 open drying yard which is hardly available with
the farmers under Kerala conditions.
The times recorded with newly developed dryer for drying copra were different from
those reported in literature. Palmer (1968) and Dumaluan et al. (1982) reported a drying
time of 18 to 20 h at drying temperature of 60 °C. Sreenarayanan et al. (1989) reported a
drying time of 20 h in a mechanical dryer to reduce the moisture content from
50 to 7 % w.b. at a drying temperature of 65 °C. Patil (1983) reported a drying time of
36 - 40 h in small holder’s copra dryer. Annamali et al. (1989) reported average drying time
o f 3 3 - 3 7 h i n a l 000 nut capacity natural convection dryer at an average drying temperature
of 60 - 70 °C. Anonymous (1984) reported a drying time of 25-30 h in case of coconut dried
in the T.R.H.O hot air copra dryer.
Although reasons for different durations of copra drying could not be explained
correctly it is assumed that they were due to experimental techniques (Hall, 1970). The
method of moisture content determination and the thickness of meat could also account for
difference between the past and present results (Gurate et al, 1996). Brennldorfer et al
(1985) reported that the temperature, moisture content and the physical dimensions of
coconut meat were the most important factors affecting the rate of moisture removal. The
internal structure and composition of the coconut meat were also of importance. The age of
the coconuts also affects drying. Coconuts stored for 2-3 months after harvest drastically
reduces the drying time due to weight loss mainly because of decrease in water content.
Anonymous (1984) reported that the drying time depends on the thickness of the layer of
copra, its density, free circulation of hot air and fuel used. Singh et.al. (1999) reported a
drying time of 24 h at an average drying air temperature of 75 °C for drying coconut in a
kiln type dryer. Sudaria et al. (1996) reported an average drying time of 8.67 h in case of
coconuts dried in improved farmer’s direct type dryer.
4.7.3. Effect of drying air temperature on drying time and drying rate of coconut dried
in copra dryer
The drying rate of coconut was 11.74 % d.b / h in the first hour and 3.19 % d.b / h in
the final stage of drying at 80 °C drying air temperature (Fig. 4.66). It was due to the fact
that the coconuts were having high moisture content in the order of 89.21 % d.b. at the
beginning of drying and was only 6 to 6.25 % d.b. in the final stage of drying. From
Fig. 4.66 it is clear that the constant rate period of drying was absent for the entire duration
and the drying of coconut took place under the falling rate period at all the drying air
temperatures. Rajashekaran et al. (1961) reported that at constant air velocity the drying of
half coconuts in cup shape occurs entirely in the falling rate period.
The drying rate was more up to an average drying time of 11 h during which the
drying rate ranged from 11.74 to 6.15 % d.b. / h for drying air temperature of 80 °C and after
that the drying rate was 5.84 to 3.19 % d.b. / h. The reason may be that initially more
moisture was lost in less time due to the availability of free moisture and at later stages the
moisture available in coconut was less. Also, during drying, as moisture content decreased,
the difference in vapour pressure between the inner surface of the kernel and out side drying
air also decreases. Drying also causes shrinkage of outer cells. These two factors slow down
moisture diffusion. This may be the reason for decrease in drying rate during extended
periods of drying. Moisture movement is known to be dependent on diffusion and capillary
flow. From Fig. 4.66 it is clear that the constant rate period was totally absent during entire
duration and the drying of copra took place under the falling rate period. Similar type of
results was quoted by (Rachmat, 1999) where the drying rate was 2 % w.b. / h at drying air
temperature of 50 to 60 °C.
4.7.4. Effect of drying air temperature on moisture content and drying rate of coconut
dried in copra dryer
A plot between the drying rate and the average moisture content of coconut at drying
air temperatures of 80 °C is shown in Fig. 4.67. The drying rate was more for coconut dried
at higher temperature than the one dried at lower temperatures for the same average
moisture content of the coconut. The reason is that at higher temperature, the relative
humidity of the drying air was less as compared to the one at lower temperature. Because of
this, the difference in the partial vapor pressure between coconut and the surrounding higher
temperature drying air environment was more compared to the one at lower temperature
drying air environment. Hence, the moisture transfer rate was more with higher drying air
temperature.
Coconut did not exhibit a constant rate period of drying. The entire drying took place
in the falling rate period. At the beginning, when moisture was high, drying rate was very
high, and as moisture content approached to equilibrium moisture content, drying rate was
very low (Fig. 4.67). The absence of constant rate drying period may be due to non uniform
release / migration of moisture through kernel tissue. This is in confirmation with onions
(Mazza and Maguer, 1984), garlic (Madamba et al, 1996) and lettuce and cauliflower leaves
(Lopez et al., 2000).
4.7.5. Effect of drying air temperature on drying time and moisture ratio of coconut
dried in copra dryer
Curves of moisture ratio versus drying time for the different drying air temperatures
are shown in Fig. 4.68. The average moisture ratio of coconut reduced exponentially as the
1 3 9
drying time increased. Reduction of moisture ratio from 0.86 to around 0.06 occurred in
22, 21, 26 and 25 h at 80 °C drying air temperature, respectively. During drying, as moisture
content decreased, the difference in vapour pressure between the inner surface of the kernel
and out side drying air also decreases. Drying also causes shrinkage of outer cells. These
two factors slow down moisture diffusion. This may be the reason for decrease in drying
rate during extended periods of drying. Rajashekaran el al. (1961) reported moisture ratio
of 0.8 to 0.08 at the initial and final period in case of coconut dried at 70 °C for the initial
8 h and complete the drying at 60 °C. The total drying time was 13.5 to 14 h. This is in
agreement with the reports quoted by Mazza and Maguer (1984) in onions,
Madamba el al. (1996) in garlic and Lopez et al. (2000) in lettuce and cauliflower leaves.
4.7.6. Variation in drying air temperature in the drying bin of the copra dryer
The temperature profile in the drying chamber under no load conditions after
15 minutes of firing is given in Fig. 4.69. From the Fig. 4.69 it is clear that the temperature
in the drying chamber reached 80 °C at some places and at other places it was about 55 °C.
This was due to the fact that only one shell was burning in both the burning chamber at a
time and so at those places where the shell was burning (vertically above it) was heated up
faster. The temperature in the plenum chamber reached up to 82 °C within 30 minutes and
was more or less uniform with a variation of ±2 °C. Hence, the dryer was fired 30 minutes
before the actual loading of fresh split coconut halves in to the drying chamber. The
variation of temperature in the drying chamber in the bottom layer, 10 cm, 20 cm and 30 cm
depth from the bottom layer under full load conditions after 30 minutes of firing is shown in
Fig. 4.70. From Fig. 4.70 it is clear that the temperature of hot air in the bottom layer was in
the range of 82 to 76 °C. To avoid browning of copra the split nuts in the first two layers
were kept in such a way that the kernels faced upwards (U )• The variation in temperature at
10, 20 and 30 cm was in the range of 77 to 70 °C; 72 to 66 °C and 67 to 62 °C, respectively.
The average temperature in the drying chamber was 72 °C.
4.7.7. Performance of copra dryer
The performance data of the dryer is given in Table 4.40. From the Table 4.40 it can
be seen that at full load the thermal efficiency was in the range of 25.25 to 26.4 % which
indicates good performance of the dryer. Patil et al. (1984b) reported thermal efficiency of
42.65 % in case of electrically operated copra dryer. AnnamaJai et al. (1989) reported
thermal efficiency of 18.7 to 23.4 % in a natural convection copra dryer. Singh et al. (1999)
reported thermal efficiency of 31.25 % in case of copra dried in kiln type dryer. Rachmat
el al. (1999) reported thermal efficiency of 10.72 % in case of pit type copra dryer. Thus in
comparison it can be seen that the thermal efficiency is better as compared to other indirect
type of dryers available. It can be seen from the Table 4.40 that the thermal efficiency
reduced to 9.41 % at 50 % capacity indicating that the dryer should be used at full capacity
only. The quantity of fuel consumed was 86 kg at 50 % load where as it was in the range of
60 to 64 kg in case of full load indicating that the hot air escaped into the atmosphere at a
faster rate there by increased the quantity of fuel required. The heat utilization factor was in
the range of 0.17 to 0.19 indicating low heat utilization which is very common in indirect
type of dryers but the coefficient of performance was very high in the range of 0.80 to 0.86
indicating high efficiency of the dryer. Similar type of result was quoted by Chakraverty and
More (1983) and Kaleemullah (2002) in the case of drying of raw and parboiled paddy in a
baffle type grain dryer and chilies in a rotary type dryer, respectively.
4.7.8. Quality characteristics of copra dried in copra dryer
The oil content was in the range of 62.48 to 63.55 %. It indicated that there was no
loss of oil for the copra dried at drying air temperature of 80 °C (Table 4.41). The oil content
reported for West Coast Tall variety was in the range of 60 to 68 % (Thampan, 2003). The
average free fatty acid content was 0.0865. The average acid value, peroxide value and
saponification number were 0.265, 0.3375 and 253 respectively. Similar results were
reported by Singh et al. (1999) in case of copra dried in kiln type copra dryer. The colour of
2 0 6
copra was light brown as smoke did not come into contact with the kernel and the smell and
taste were typical indicating no loss of the natural aroma of copra.
4.7.9. Evaluation of weather parameters for suggesting effective production zones for
production of copra using newly developed copra dryer
Fig. 4.71 shows monthly variation of ambient relative humidity, rainfall, temperature
and sunshine hours based on monthly means from the year 1974 to 2002. The weather data
is representative for the whole state and similar conditions exist throughout the state. Hence,
the recommendations given will hold good for all those places where similar environmental
conditions exist. The higher value of relative humidity i.e. above 80 % exists throughout the
year in the forenoon session and the RH is above 80 % during the period June to August.
During the same period the rainfall is above 700 mm / month and bright sun shine hours are
about three hours only and the maximum temperature is in the range of 28.6 to 30 °C only.
Thus, during these three months (June to August) copra drying in open sun becomes
impossible. During this period the copra prices shoots up due to demand supply gap. The
copra dryer developed can effectively be utilized during the aforesaid period for getting
maximum profit. Similar results have been reported by Patil et al. (1982).
A bright sunshine of more than 8 hours per day is available during the period
December to middle of May. During the same period the maximum temperature is in the
range of 31.8 to 33.1 °C and RH in the afternoon was in the range of 50 to 60 %. Due to low
humidity and moderate temperature the farmer can adopt sun drying or combine both and
reduce drying time. From Table 4.21 it is clear that the EMC of copra under adsorption and
desorption at an average temperature of 25 °C and RH of 43.80 % is 5.2 to 5.8 % which
indicated that the copra can be stored safely during the above said period.
During the period September to November, the average rainfall is in the range of
218 to 239 mm again forcing the farmer to adopt artificial means of drying. Thus effectively
the farmer is left with only five and half months to adopt traditional method of sun drying.
4.7.10. Effect of weather parameters on storage of copra
Copra can be stored well except from last week of May to October because the RH
during this period is above 80 %. Thus storage can be safe during the month of October if
structure is well ventilated. But, even in well ventilated structures copra cannot be stored
during June to September as the relative humidity is above 85 %. From Table 4.21 it can be
seen that above 85 % RH the EMC was in the range of 10.6 to 11.2, 9.5 to 10 and 9.0 to
9.4 % d.b. at 25, 35 and 45 °C, respectively indicating favorable conditions for the growth of
bacteria and fungus. Lozada (1995) reported that moisture content above 8 % was favorable
for the growth of fungus. Thus the fanner using the copra dryer during the period June to
September should ensure proper demand before processing to avoid storage or should have
his own oil extraction mill otherwise he should not resort to mechanical drying of copra.
4.7.11. Effect of drying method on microbial characters of copra
The population of bacteria, fungi and actinomycetes in different copra treatments are
given in Table 4.42 and the average values in Table 4.43. The count of bacteria and fungi
were almost similar in the treatments Ml and M2 while the actinomycetes was high
(86.6 x 103 cfu / g copra) in the Ml treatment. However, the copra of M3 treatment (stored
for three months at RT) had maximum fungi (307.0 x 104 cfu / g copra) and actinomycetes
(1 84.0 x 103 cfu / g copra) population. The coconut dried in sun and stored for one month at
room temperature had 22.0 x 105 bacterial, 77.0 x 104 fungi and 11.0 x 103 actinomycetes
cfu / g copra. The S2 and S3 treatments had almost similar bacterial and actinomycetes
populations, however, a significant increase in fungal count was observed in S2
(117.0 x 104cfu / g copra) and S3 (173.0 x 104 cfu / g) when compared to SI treatment. In
the case of coconuts dried in copra dryer and stored at room temperature for 1 month, very
low bacterial (0.2 x 106 cfu / g) and nil fungal and actinomycetes growth were recorded. The
bacterial and fungal counts increased significantly in CD2 treatment where copra was stored
at RT for two months with population count of 36.0 x 106 cfu of bacteria and 23.0 x 104 cfu
209of fungi per gram of copra. In the CD3 treatment, the bacterial count decreased, fungi
increased slightly whereas actinomycetes population jumped to 17 x 103 cfu/ g of copra.
The microbial analysis clearly indicates that copra dried in smoke dryer and stored at
room temperature for three months is highly contaminated with fungi and actinomycetes
whereas the same treatment for one month and two months period had lower levels of
microbial population. The samples dried in the sun also carried some load of bacteria, fungi
and actinomycetes which were however significantly lower than the M3 treatment but
somewhat on par with M2 treatment. The copra dried in copra dryer and stored for 1 month
had minimum microbial growth which increased in CD2 and CD3 treatments. This shows
that drying of copra in copra dryer is superior to drying either in smoke dryer or sun drying,
from microbial quality point of view. The succession of micro flora is usually seen with
bacterial populations colonizing first, followed by fungal and then actinomycetes
populations.
Table 4.42. Microbial analysis of copra samples
Bacteria Fungi Actinomycetes
samplen x 106 cfu/g copra
RI R2 R3
n x 104
RI
cfu/g copra
R2 R3
n x 103 cfu/g copra
RI R2 R3
Ml 27 18 15 50 80 100 100 40 120
M2 13 7 9 30 150 80 30 10 10
M3 19 3 8 380 290 250 160 220 170
SI 31 20 14 60 120 50 22 00 10
S2 32 24 34 120 70 160 10 20 20
S3 11 21 24 240 120 160 10 20 30
CD1 0.2 0.4 0.1 00 00 00 00 00 00
CD2 53 24 31 20 20 30 00 8 10
CD3 12 16 14 25 31 23 10 10 12
Table 4. 43. Average microbial population of three replications
SampleBacteria Fungi Actinomycetes
n x IQ6 cfu/g copra n x 104 cfu/g copra ex 10 cfu/g copraAverage Average Average
Ml 20.0 77.0 87.0M2 10.0 87.0 17.0M3 10.0 307.0 183.0SI 22.0 77.0 11.0S2 30.0 117.0 18.0S3 19.0 173.0 20.0
CD1 0.23 0.0 0.0CD2 36.0 23.0 6.0CD3 14.0 26.0 17.0
4.7.12. Comparison of Small holder’s copra dryer (400 nuts) with the dryer developed
A comparative study of the copra dryer developed and the one being most commonly
used by the farmers is given in Table 4.44
Table 4. 44. Comparison of dryer developed with small holder’s copra dryer
*Source, Patil et ah (1983)
From the Table 4.44 it is clear that the dryer developed as compared to small holders
dryer has the main advantage of feeding fuel once in 6 h which was one of the major
constraint reported by fanners. The other advantage was that the drying air temperature is
maintained at 80 °C so that the effective drying air temperature of 72 °C inside the dryer was
achieved which reduced the drying time by 12.5 h. Though the cost of dryer developed was
double the cost of small holders the capacity is more than doubled. The cost of drying
is 6.05 / kg copra in the small holders dryer but the cost of drying is only 5.33 / kg of copra.
The thermal efficiency is also comparatively higher in the newly developed dryer.
4.7.13. Models
The best fitted model of moisture ratio of the dryer developed is a polynomial of
order three which explained variation completely. The estimates of coefficients are shown in
Table 4.45. The pattern of residual plot was seen random (Fig. 4.72) and the values of
Es and Em statistics were the lowest for the chosen model (Table 4.46). The observed and
fitted models were shown in Fig. 4.73. The selected model for describing the moisture ratio
is given below
4.7.14. Cost economics
The cost of the dryer was estimated to be Rs. 15,000.00 (Appendix.-7). The cost
involved to dry one kilogram of copra in the copra dryer was worked out and found to be
Rs. 5.33. The cost of drying one nut excluding the cost of nut works out to be Rs. 0.93. As
the quality of copra dried in the copra dryer is good it will fetch higher price in the domestic
as well as international market.
2 1 2
Table 4.45. Estimated values of parameters of different models used to describe copra drying at drying air temperature of 80 °C in copra dryer
Sl.No. Model Estimated parameter, k/h1 MR = b„+ biX + b2X2+ b3XJ
(proposed model)b„ =0.96956 ; bi = -0.10757; b2 = 0.00451; b3 = -0.00007
2 Lewis : MR = exp(-kx) k = 0.1243 Hustrulid & Flikke:
MR = A exp(-kx)A = 0.979; k = 0.122
4 Page: MR= exp(-kx") k = 0.137; n = 0.9575 MR = A exp(-kx) +B exp(-lx)
(proposed model)A - 0.996 ; k= 0.127 B = 0.0006; 1 =-0.159
6 MR = A exp(-kx) +B exp(-l/x) (proposed model)
A = -0.309; k = -1.000 B =-0.641; [ = -0.226
Table 4.46. Estimated values of mean relative percentage deviation (Em), standard error of estimates (Es), R2 and Residual plot pattern of different models used to describe the copra drying in copra dryer at drying air temperature of 80 C
Model No. R2 E,„ Es Residual plot pattern1 1.00 -0.0002 0.0025 Random2 0.997 0.0689 0.0138 Systematic3 0.997 0.0564 0.0130 Systematic4 0.997 0.0393 0.0121 Systematic5 0.999 -0.0007 0.0059 Random6 0.999 0.0018 0.0074 Random
