Project Progress Report: ReInvestigation of rice kernel cracking and its control in the field and during post-harvest processes in the Mekong Delta of Vietnam - MS4 " docx

Ministry of Agriculture & Rural Development

Project Progress Report



026/VIE05
Investigation of rice kernel cracking and its control in
the field and during post-harvest processes in the
Mekong Delta of Vietnam


MS4: BASE LINE INFORMATION





Dr Vinh Truong, Nong Lam University

Associate Professor Bhesh Bhandari and Professor Shu Fukai
The University of Queensland






May 2007

1




BASE LINE INFORMATION

In order to determine the actual post harvest losses mainly due to rice cracking, the
baseline data was collected systematically based on the farmers practice and also by
experimentations. There is a chain of activities during harvesting and post-harvest
processing of rice. Each of these factors will contribute to the losses. Some of these
factors can be dependent to each other. The main factors which were considered in this
study during the collection of data were:

• Harvesting time- before and after grain maturity
• Harvesting methods- manual, reaper, combined harvester
• Drying methods- sun drying and mechanical drying
• Milling losses- small, medium, large rice milling plants and milling procedure.

The behaviour of the rice grain to cracking will depend on the variety and the season.
Considering this factor the data were collected for 6 different varieties in 4 different
locations of Mekong River Delta (MRD). The information presented here in this report is
only for rainy season rice crop (harvesting season June/July). Collection of more data will
continue for dry season crop (harvesting season February/March).

The survey data from farmers was collected during the winter-spring season for the same
varieties. The experimental data for the winter-spring season will be presented in the
forthcoming report.

1. Effect of harvesting time on the rice cracking and head rice yield

Incorrect harvesting time is one of the major factors that cause the losses due to cracking.
Cracking can develop in the field as a result of changes in grain moisture or moisture
cycles after the rice matures due to hot sunny days followed by humid nights. The
cracking behaviour of the rice in the field is expected to depend on the season due to the
different patterns of temperature fluctuation during day and night, degree and strength of
sunshine and frequency of rain. During the rainy season, the rice grain can develop cracks
during the late maturity stage due to rewetting. The objective of this experiment was to
determine the effect of harvesting time on kernel cracking of some rice varieties in rainy
season in the MRD.

Harvesting time field experiments were conducted in the three locations on four most
cultivated rice varieties in those areas (Table 1).

Table 1: Base line data collection to determine the losses due to current harvesting
practices (harvesting time and methods)


Locations Rice variety Harvesting
period
1 Tan Thoi 1 cooperative, Can OM 2718, OM1490 30
th
May-13
th
June

2
Tho province

2 Tan Phat A cooperative,
Kien Giang province
An Giang 24 (AG24) 22
nd
-30
th
July
3 Seed centre, An Giang
province
Jasmine June/July
According to local survey results, the one or two most cultivated rice varieties OM 2718
and OM 1490, An Giang 24 and Jasmine were chosen in Can Tho, Kien Giang and An
Giang provinces, respectively. Using a randomised block design, the rice was harvested 6
days prior and 6 days post-maturity stages in 2 days intervals for OM 2718 and OM 1490
(Can Tho) and 1 day interval for An Giang 24 and Jasmine varieties. The percentage of
rice grains with cracks and head rice recovery in a laboratory milling system were
measured for both brown and white rices.

The full details including the experimental design will be included in next six-monthly
report.

1.1 Harvesting time and rice cracking
Some selected data on the amount of cracked grains as influenced by the early or late
harvesting from the day of maturity are presented in Figures 1 and 2. The maturity day is
taken as an estimate from farmers experience and available data from the extension office.
This value was 90 days for OM 1490, OM 2718 and An Giang 24 and 98 days for
Jasmine rice varieties. The head rice recovery was analysed for both brown (after
dehulling) and white (after whitening) rices.

1.1.1 Cracking in brown rice and head rice recovery

Experiments conducted on four common rice varieties in three different locations
indicated that the rice cracking is obviously influenced by both the variety and time of
harvesting. Harvesting the rice a few days prior to maturity will not have much impact on
rice cracking, but delayed harvesting will result in significant rice cracking (up to 24% of
total brown rice) depending on the variety. Interestingly, early harvesting has shown
lesser proportion of grain cracks and higher head rice recoveries. This indicates how
important it is to harvest the rice in time. Any over-drying in the field (or in the plant) can
result in increased number of cracked grains and reduced head rice recovery.

Our results indicate that there is a varietal difference on rice cracking. It should be noted
that the maturity or optimum harvesting time was an estimate which was almost the same
day for all varieties used in this investigation. If varieties were harvested about the same
time, then we could conclude that 1. varieties differ considerably in the cracking (hence
intervention opportunity of growing low cracking varieties such as AG24 for farmers and
developing such varieties for rice breeders), 2. harvesting optimum harvest time had
rather small cracking problem but delay of 6 days can cause major problem (and hence
intervention opportunity), for which economic analysis can be made 3. varieties differ in
their response to time of harvesting hence time of harvesting is more critical for some
varieties than others, and hence opportunity for intervention (recommendation would be
to ensure quick harvesting for particular varieties).



3
0.80
3.20
9.60
4.80
10.80
15.20

23.60
0
4
8
12
16
20
24
-6 -4 -2 0 +2 +4 +6
Harvesting time (days from maturity)
Grain cracks (%)
0.40 0.40
1.20
2.80
10.80
4.00
5.20
0
2
4
6
8
10
12
-6 -4 -2 0 +2 +4 +6
Harvesting time (days from maturity)
Grain cracks (%)

Variety: 1490 Variety: 2718

0.13
0.67
1.6
0.4
0.53
1.34 1.33
1.73
5.47
0
2
4
6
8
10
-4 -3 -2 -1 0 +1 +2 +3 +4
Harvesting time (days from maturity)
Grain cracks (%)
4
3.92
5.18
8.66
6
5.14
7.6
0
2
4
6
8
10

-3 -2 -1 0 +1 +2 +3
Harvesting time (days from maturity)
Grain cracks (%)

Variety: An Giang 24 Variety: Jasmine

Figure 1: Proportion of cracked brown rice grains as influenced by harvesting time, 4-6
days earlier (-6 days) and 4-6 days later (+ 6 days) than the predicted day of maturity.

1.1.2 Cracking in white rice
The cracking in the whole white rice kernels were also measured for the same rice variety
samples which were used to determine the brown rice cracking. It is important to know
the level of cracking in the white rice because this will also be important when rice
breakage occurs during post-milling conditions. There is a possibility of split of the grains
with severe cracks during storage, particularly if there is moisture and temperature
variations or stresses. This is the area which will need more investigation (although may
not fall under the scope of this current project).

The cracked grains were more in the case of white rice than those in brown rice samples.
This is because the proportion of cracked grains is calculated based on the whole white
rice kernels, excluding the broken rice. The weak and fissured brown rice would normally
break during the whitening process. The cracking in the white rice kernels can be
developed due to the shear during the whitening process. Some of the brown rice kernels
with minor fissures or cracks may not break during the whitening process. The varietal
difference on the cracking and head rice yield is obvious in Figure 2.



4
1.60

4.40 4.40 4.40
5.20
4.80
11.60
0
2
4
6
8
10
12
-6 -4 -2 0 +2 +4 +6
Harvesting time (days from maturity)
Grain cracks (%)
0.40
1.60
2.80
2.40
3.20
4.40
6.00
0
2
4
6
8
10
12
-6 -4 -2 0 +2 +4 +6
Harvesting time (days from maturity)

Grain cracks (%)

variety 1490 variety 2718
0.27
0.67
1.47
0.4
0.53
1.07
1.27
1.73
4.13
0
2
4
6
8
10
-4 -3 -2 -1 0 +1 +2 +3 +4
Harvesting time (days from maturity)
Grain cracks (%)
7.4
8
7.8
11.8
8.4
9.4
12.2
0
2

4
6
8
10
12
14
-3 -2 -1 0 +1 +2 +3
Harvesting time (days from maturity)
Grain cracks (%)


Variety: An Giang 24 (AG 24) Variety: Jasmine

Figure 2: Proportion of cracked white rice grains as influenced by harvesting time, 4-6
days earlier (-6 days) and 4-6 days later (+ 6 days) than the predicted day of maturity.

1.2 Harvesting time and head rice recovery
The head rice recoveries as a function of harvesting time for four varieties of rice are
presented in Figure 3. The results indicated that the head rice recovery follows the same
opposite trend to rice grain cracking. This obviously means that the presence of cracks in
the grain influenced the head rice recovery. The head rice recovery was less at late
harvesting period. A delay of 4-6 days reduced the head rice recovery by 7-13.4%.

5

45.41
51.47
43.54
43.91
38.76

36.83
40.72
0
5
10
15
20
25
30
35
40
45
50
55
-6 -4 -2 0 +2 +4 +6
Harvesting time (days from maturity)
Head Rice Recovery (%
)
51.06
52.3
50.73
47.99
42.23
36.51
34.53
0
5
10
15
20

25
30
35
40
45
50
55
-6 -4 -2 0 +2 +4 +6
Harvesting time (days from maturity)
Head Rice Recovery (%
)

Variety: OM 2718 Variety: OM1490

42.35
41.75
40.76
42.51
43.5
46.33
46.99
42.72
35.9
0
5
10
15
20
25
30

35
40
45
50
55
-4 -3 -2 -1 0 +1 +2 +3 +4
Harvesting time (days from maturity)
Head Rice Recovery (%
)
41.59
54.65
51.82
55.36
54.59
48.15
49.46
0
5
10
15
20
25
30
35
40
45
50
55
60
-3 -2 -1 0 +1 +2 +3

Harvesting time (days from maturity)
Head Rice Recovery (%
)

Vaariety: An Giang (24) Variety: Jasmine

Figure 3. Influence of harvesting time on the head rice recovery of white rice

The overall results as influenced by harvesting time are presented in Table 2. It should be
noted that the head rice recovery was determined by a laboratory milling system. Thus,
the head rice recovery will also be a function of milling efficiency. Therefore, the head
rice recovery data presented in Table 2 is in relative term. In this case, the recovery on the
harvesting at maturity (0 day) was considered as 100%. In addition, due to the limited
number of experiments undertaken (as feasible), the values are presented in the range. The
varietal factor has been incorporated within this range.

Table 2: Effect of harvesting time before and after maturity (4-6 days prior and 4-6 days
later than the expected day of maturity) on the proportion of cracked grains and head rice
recovery. Head rice recovery is relative to the recovery on maturity day.

Proportion of cracked grain % Head rice recovery relative %
Rice variety
Before maturity After maturity Before maturity After maturity
OM1490 0.8-9.6 10.8-23.6 106-109 72-88
OM2718 0.4-1.2 2.8-10.8 104-117 84-93
An Giang (24) 0.5-1.6 1.3-5.1 93-97 83-108
Jasmine 4-4.5 6-7.7 75-99 87-99

6
Table 3: Effect of harvesting time before and after maturity (4-6 days prior and 4-6 days
later than the expected day of maturity) on the proportion of cracked white rice kernels.

Proportion of cracked grain %
Rice variety
Before maturity After maturity
OM1490 5.2-6.1 7.2-11.6
OM2718 0.4-2.8 3.2-6.0
An Giang (24) 0.3-1.5 0.5-4.2
Jasmine 0.4-2.8 3.2-6.0


1.3 Evaluation of farmers’ practice

The farmers harvesting practice data were collected in TanPhat and TanThoi cooperatives
during winter-spring harvesting season (2007) for the same varieties that were used in the
experimental study. Twenty to thirty farmers who had a good idea about rice harvesting
and loss information participated in the survey (Annex 1). A further data collection during
wet season will continue in next wet season.

1.3.1 Value losses due to late harvesting:
Based on the farmers harvesting time, the potential loss incurred by farmers was
calculated from the experimental value as determined in section 1.

In order to calculate the losses of head rice yield (HRY), the regression models (Table 3)
were used for each variety based on the experimental data of spring-summer crop 2006/7.
The losses of HRY were taken by the subtraction between HRY at the actual harvesting

date and HRY at maturity date. The maturity date of 90-92 was used for OM2718,
OM1490 and AG24 varieties, and that of 98 was used for Jasmine variety as indicated in
Table 5. The loss due to late harvesting by farmers was calculated by a regression fit of
the data presented in Table 4. The value of the broken rice was considered as 50% of the
whole kernel as an initial estimate. A further realistic analysis will be necessary by
considering various scenarios of utilisation of broken rice (such as mixing with bulk rice,
sold as broken rice, sold as animal feed with our without mixing with husk or exchanged
for the milling service).

Table 4: Losses of head rice yield due to late harvesting experimentally estimated for
OM1490, 2718, and Jasmine varieties (using the data from Figure 3).

Growing
time
(day)
OM1490

OM2718


Jasmine

92 0.00% 0.00% 0.00%
94 5.77% 5.65% 0.77%
96 11.49% 7.08% 7.21%

7
98 13.47% 7.08%
5.90%
Equation

HRY loss =
0.0119Growing time –
1.0671
R2=0.961
HRY loss= -
0.0703exp(92-Growing
time)+0.0699)
R2=0.994
HRY loss =
0.02414Growing time -
2.367
R2=0.743


AG24: No loss model can be estimated due to the inconsistency of HRY with harvesting
time (Figure 3). It was considered that no losses at all for this variety. The actual
harvesting survey data from farmers also indicate that the farmers harvest the rice at early
stage (Tan Phat A cooperative Kien Giang province, see appendix excel file).
For calculation of losses due to late harvesting based on the above models (as shown in
the appendix), the actual harvesting time was considered to be early (late) if it was less
than (more than) the maturity date collected from extension resources as shown in the
below Table:

Table 5: Definition of late or early harvesting for different varieties

Variety Early harvesting
if less than (day)
Late harvesting
if more than

(day)
Experimental
maturity day
Maturity day
mentioned by
extension service
OM1490
OM2718
Jasmine
AG24
90
90
96
85
92
92
98
90
92
92
98
90
87-92
90-95
95-105
85-90

As based on the above definition, farmers normally harvest 1-3 days after the maturity
date due to the lack of labour. Based on our data, 80%, 90%, 55% and 50% of crop of
OM1490 and OM2718, Jasmine and AG24, respectively are harvested late.

The following Table 6 provides the equivalent loss data for four different varieties due to
the late harvesting practice of farmers.

Table 6: Losses of HRY in equivalent to losses of paddy (kg/100kg yield) for different
varieties in the MRD due to current practices of late harvesting of the farmers
Variety OM1490 OM2718 Jasmine AG24
Losses (%) 4.16 2.9 1.17 0.06

On average, losses due to late harvesting are around 2.1%. These data are obtained based
on the harvesting practice of the farmers in the dry season (Winter-Spring crop, Feb/Mar)
when the weather is favourable for harvesting, only the limiting factor being the labour
shortage. The losses due to late harvesting are not very high because most of the farmers
have the knowledge of maturity day. However, in the wet season (Summer-Autumn crop,
July/August) the weather is changeable with rain and storm. Thus, in addition to losses
due to late harvesting (and field/sun drying) more losses are expected from uncontrollable
bad weather.

8


Conclusion and project intervention methods:
The following conclusion can be drawn from the above information:
• The harvesting time is one of the important factors to control the rice cracking and
eventual head rice recovery.
• Varieties differ in their grain cracking and those with low level of cracking such as
AG24 is recommended for cultivation (after confirming this results in the next
season). When head rice recovery is also considered, OM1490 is better.
• There is a clear trend that a few days early harvesting (before maturity) is better
than the late harvesting. The farmers’ survey results indicate that they harvest the
rice mostly late due to labour shortage during the peak harvesting period. The

intervention opportunity of early harvesting to reduce grain cracking and increase
head rice recovery should be conveyed to the farmers and extension agency. This
extent of this harvesting time effect is also dependent on the variety.
• This information gathered will be very useful for the farmers and will be made
available through training. This will highlight the importance of rapid harvesting
of the crop. Farmers are encouraged to make their crop management in such ways
that they can harvest earlier (eg, organize labour for harvesting etc).
• This is expected to impact to the farmers decision to determine the harvesting
time. In some rice varieties this is expected to reduce the losses substantially, since
one of the varieties tested in this investigation had a proportion of cracked rice as
high as 24%.


2. Effect of harvesting methods on the rice cracking , head rice yield and losses

2.1 Effect of harvesting methods on the rice cracking and head rice yield
Harvesting method used can influence the extent of rice cracking in the field. The
harvesting can be done by hand or machine. As a current practice, harvesting by hand is
widely used. In relation to harvesting method, the cracking of rice is related more to the
time duration necessary to harvest than the methods itself. A fast harvesting during wet
season and harvesting at correct time during dry season is necessary to avoid rewetting or
over-drying of the grain while in the panicle. Unfortunately, due to the lack of the labours
during the harvesting period farmers are not always able to harvest the crop in time which
results in losses.

This work gathered the actual data to determine the effect of harvesting methods on kernel
cracking of some rice varieties in Summer-Spring season (June/July) in Can Tho and
Long An provinces. The following harvesting methods were used:

1. Hand (+ mechanical threshing)

2. Reaper (+mechanical threshing)
3. Combined harvester (harvesting and threshing combined)

The data were collected from our own experiments as well as from selected farmers field
after their traditional harvests.


9
Comparison of harvesting methods by experimentation on the head rice recovery was
undertaken in cooperatives in Can Tho and Long An provinces (Table 7). As additional
information, cracking behaviour of the grain due to threshing was also investigated in
those two cooperatives. In each cooperative, the experiments were undertaken in two
fields, where the popular rice variety was grown. The following results were obtained
(Table 7) for each harvesting method applied.

Table 7: Effect of harvesting methods on the head rice yield
Location Harvesting method Notation
Average head
rice recovery
(%)
Hand and heaped immediately HH 41, 50
Hand and dried in the sun (one
day)
HD 37, 47
Reaper and heaped
immediately
RH 49*
TanThoi
co-operative,
CanTho

(OM 2718, OM1459)
Reaper and dried in the sun
(one day)
RD 52*
Hand and heaped immediately HH 45, 60
Hand and dried on the sun
(one day)
HD 43, 62
GoGon
co-operative, LongAn
(Bu Tin, VN 95-20)
Combined harvester CH 36**,60
*Only one replication due to rain
**Low value due to rain during harvesting

There was a large variation in cracking losses. It might be due to varietal difference and
other uncontrollable factors. It was raining a lot during the experimental period. To
achieve an accurate result, a large number of experiments should be undertaken to reduce
the variability in field condition. This was not feasible due to lack of time and restriction
on the resource. Therefore, this result should be taken as indicative only. The experiments
will be repeated in the dry season (Feb/March). The results have indicated that the head
rice yield by reaper is better or as good as harvested by hand. The main advantage of
using the mechanical harvesting is to shorten the harvesting time in order to avoid the
effect of rain or weather change difference on the head rice yield. We have previously
shown that the late harvesting than at maturity will make the grain more sensitive to
cracking. Therefore, any delay or longer harvesting time can cause more losses, as is the
case when the harvesting by hand is practiced.


2.2 Effect of harvesting methods on the threshing losses

The above harvesting methods also affected the losses of grain during subsequent
threshing step. The threshing losses are the fractions of paddy kernels mixed with the
impurities removed by the thresher. These losses are shown in Table 8.

10

Table 8: Effect of harvesting methods on the threshing losses
Location Harvesting method Notation
Losses (%)
Hand and heaped immediately HH 1.4
Hand and dried in the sun (one
day)
HD 1.2
Reaper and heaped immediately RH 1.1
TanThoi
co-operative,
CanTho
(OM 2718, OM1459)
Reaper and dried in the sun (one
day)
RD 0.8
Average 1.1

In general, leaving the rice one day in the sun after cutting prior to threshing (treatments
HD and RD) reduced threshing losses by 0.2%. This reduction can be explained by lower
moisture content of rice after sun drying. The average threshing losses by reaper and hand
were 1.0% and 1.3%, respectively. The overall threshing loss was 1.1%.

Conclusion and project intervention methods:
• The above information indicates that a quick harvesting by reaper method is

beneficial to improve the head rice recovery, but this needs confirmation.
• Since, a rapid harvesting is not possible for small farmers due to their limited
financial capacity; operation of harvesters through farmers cooperative is the best
alternative. This project implements this concept by providing mechanical
harvesters to two cooperatives.
• The cooperatives equipped with harvesters will be used for the demonstration
purpose. Data will be presented to the farmers through training. The farmers will
be trained through extension workers. It is expected that the dissemination of the
information to the local farmers within the catchments area of the cooperative will
be spontaneous due to the engagement of the cooperatives in the activity.
• Head rice recovery varied greatly within a system (eg 45-60% in hand and heaped
immediately in GoGon Cooperative), and identification of the source of the
variation would help increase the head rice recovery.


3. Other harvesting factors contributing to the losses

There are other factors which can contribute to the post-harvest losses. These possible
factors are:
1. Threshing method- hand or machine
2. Shattering of grain due to harvesting method applied


3.1 Effect of threshing method on grain cracking and head rice recovery

The threshing method applied can cause the cracking in the rice kernels and eventually
reduce the head rice recovery. The data were collected in two provinces at the same time
when experiments were conducted as described in the previous section 1. The results are
presented in Table 9. These results indicated that the grain cracking is not significantly

11
affected by the method of threshing. However, some reduction of head rice recovery was
observed in the case of rice threshed by machine.

Table 9: Effect of threshing method on rice cracking and head rice recovery

Grain cracking (%)
Brown rice White rice
Head rice recovery
(%)
Rice varieties
Hand Machine Hand Machine Hand Machine
OM2718/
OM 1490
4.1 3.9 3.0 1.8 49.9 46.7
An Giang 24
0.9 2.4 1.5 0.7 45.6 44.0


Losses due to grain shattering during harvest

Although this type of losses is not related to rice cracking, it can be important if a
mechanised method is used to minimise the rice cracking. The losses due to shattering
during harvest were analysed by two methods:

a. Collection of shattered grains in the selected farmers’ field in two provinces (Tan
Phat and Tan Thoi)- harvested by hand: A total of eleven farmers were selected.
Experiments were implemented after farmers harvested and threshed the crop. The
results are presented in Table 10. The losses due to shattering in the rice field was
2.9 % in average, but was as high as 5%. This loss for Kien Giang was 2.5 ± 1.9

%, while for Can Tho it was 3.3 ± 1.2%. These values are comparable or higher
than those obtained by PRA method for Summer-Autumn crop (2.4% for Kien
Giang and 2.3% for Can Tho, as reported by DANIDA project, 2003). There is
also a clear trend that the shattered grain is increased when the grain moisture is
lower, suggesting that late harvesting also causes more shattering of grain.

b. Collection of shattered grain from farmers field- harvested by hand and
mechanical methods: These experiments were conducted in two provinces, Cantho
and Long An. The data were collected during the experimentation involving the
effect of harvesting method as described in the previous section. The results are
presented in Table 11. Our experimental results show that cutting by reaper gave
lower shattering losses than cutting by hand (approximately one third). The results
also show that cutting by combined harvester gave lower shattering loss than
cutting by hand.

12


Table 10: Losses due to shattering in the field during hand harvesting (Tan Phat and Tan
Thoi cooperatives) by farmers
Farmers’
number
Rice
varieties
Initial
moisture
content
(pre-
harvest)
Grain

moisture
during
threshing
Shattered
grains (g)
per 25m
2
Productivity
(kg/1000m
2
)
% Loss
1 OM 2517 23.4 28.9 445 550 2.6
2 20.4 26.5 320 450 2.2
3 24.5 29.3 182 380 1.6
4 23.8 27.9 290 440 2.2
5 22.9 28.4 220 480 1.5
6
AG 24

18.9 19.4 538 400 5.2
7 22.5 24.8 285 350 3.6
8
OM 1490

20.3 23.7 298 300 4.6
9 22.8 19.4 262 380 2.3
10 27.6 22.1 305 400 2.4
11
OM 2718

18.9 19.4 318 380 3.4
Average 2.9±0.9
Note: AG 24 is the variety taken at Kien Giang province and the rest are at Can Tho. The
average loss (%) of Kien Giang is 2.5 ± 1.9 and that of Can Tho is 3.3 ± 1.2 (confident
level 95%).

Table 11: Influence of harvesting method on the grain losses due to shattering
Method of harvesting Shattering losses (%)
Hand 1.2-3.0
Reaper 0.7
Combined harvester 1.3-1.5


Estimation of harvesting losses
Harvesting losses consist of shattering and threshing losses. Therefore, a combination of
data in Tables 8 and 11 provide a picture of average harvesting losses. Table 12 shows
that harvesting losses can be as high as 4.4%. A threshing loss of 1.0% of combined
harvester was estimated by the manufacturer. On an average, mechanical harvesting
reduced harvesting losses.

13

Table 12: Effect of harvesting methods on the harvesting losses
Harvesting method
Shattering
losses (%)
Threshing
losses (%)
Harvesting
losses (%)
Hand and heaped immediately 1.4 2.6-4.4
Hand
Hand and dried in the sun (one
day)
1.2-3.0
1.2 2.4-4.2
Reaper and heaped
immediately
1.1 1.8
Reaper
Reaper and dried in the sun
(one day)
0.7
0.8 1.5
Combined
harvester
1.3-1.5 1.0 2.3-2.5


Conclusion and project intervention methods:

• Shattering loss due to harvesting method and also due to time of harvesting
(particularly late harvesting) is an important factor to consider to reduce the grain
losses during harvesting.
• Machine threshing is beneficial in terms of quickly harvesting the crop; however it
can affect the head rice recovery due to some increase in rice cracking. Therefore,
the farmers and thresher operators should be aware of this effect in order to make
them more cautious on smooth operation.
• As a general trend above results demonstrate that the mechanical harvesting can
reduce the losses by more than half of the losses incurred during traditional
harvesting method.
• The information will be communicated to the farmers through training sessions


c. Effect of drying method on rice cracking and head rice recovery

4.1 Experimental work

In order to determine the effect of drying on the rice cracking and head rice yield, the
experiments were undertaken using three types of drying systems:

1. Laboratory scale drying- hot air drying and sun drying
2. 1-ton mechanical dryer
3. 8-ton mechanical dryer

Due to the difficulty of standardising the laboratory milling system, no analysis on the
milling recovery was undertaken in the case of 1 and 8-ton flat bed driers (an example of
data is presented in Table 13). There were 8 experiments undertaken with two types of
variables (air reversal- no air reversal, two-stage drying temperature 43
o
C and 50

0
C
followed by 40
o
C). Due to the variability of moisture of the paddy samples and
differences in air flow rate the drying time was not consistent. At this point, these
experiments were limited to the drying operation optimisation, characterisation of dryer
and determination of cost of drying. Further improvement of the dryer and experiments
will be undertaken during the dry season harvest (Feb/March) which will include the head

14
rice recovery due to the current sun drying practices and using the mechanical dryer. Due
to the problem of variability of moisture in various rice batches, the drying parameters
will also be determined by simulation of drying models and some experiments will be run
to verify the models. The current information obtained on the dryer operation optimisation
and the estimate of the cost will be made available to farmers through training materials
targeted to farmers and extension workers.

Table 13: Part of the drying data collected during drying of rice using 1-ton reversible flat
bed-dryer
Drying conditions
50+43
o
C
With
Air
Reversal
50+43
o
C

Without
Air
Reversal
43
o
C
Without
Air
Reversal
43
o
C
With
Air
Reversal
43
o
C
With
Air
Reversal
43
o
C
Without
Air
Reversal
50+43
o
C

with Air
Reversal
50+43
o
C
Without
Air
Reversal
Paddy MC Before drying
(%)
24.7 18.8 24.4 21.1 27.7 27.4 26.0 21.8
Paddy MC After drying
(%)
13.4 12.8 12.6 13.7 13.5 13.4 12.5 12.6
Final Differential MC =
Top-Bottom
2.1 1.7 1.9 2.2 1.8 2.2
Coal consumption (kg) 22 10 18 9 20 8 17 -
Coal consumption (kg / hr) 1.93 2.50 2.25 0.84 1.74 1.23 1.55 -
Drying time (hr) 11.42 4.00 8.00 10.67 11.50 6.50 11.00 7.20


The laboratory experiments were undertaken to determine the effect of drying methods on
the rice cracking and head rice recovery. These methods included:
• Low temperature air drying at 35
0
C using lab flat bed drier
• Sun drying on asphalt road
• Sun drying on plastic sheet

Samples of rice with 24% moisture were dried down to 14% moisture in three different
thickness of bed (1, 2, 3cm). The results are presented in Table 10.

As estimation based on the local price system when compared between broken rice and
head rice, 1 kg loss of head rice recovery (converted into broken rice) resulted in value
loss of rice which is equivalent to 0.35 kg paddy. As head rice recovery is calculated
based on mass of paddy, then a reduction of 1% head rice recovery is equivalent to a loss
of 0.36%. Table 14 shows that sun drying on asphalt road reduced 2% of head rice
recovery or contributed a loss of 0.72%.

In general, the results indicated that there was no influence of thickness on the head rice
yield, whereas the head rice yield was higher in the case of air drying as expected. The
surface temperature of rice reached as high as 51
o
C in the case of sun drying. Due to the
relatively stable unchanged condition during the sun drying in the experiment, the head
rice yield was not affected much. In field situation it will be very different due the
variability of moisture of paddy, weather condition and unevenness of the spread sample.
Farmers sometime do not complete sun drying in a day but 2-3 days. In fact, sun drying
losses include losses due to spillage, birds and rodents, and drying condition. The results

15
in this work will be compared with the field data which will be collected during the next
rice harvesting period.

Table 14: Effect of drying method on the head rice recovery
Maximum rice
surface temperature
(1,2,3 cm)
Drying time

(1,2,3 cm
thickness)
Head rice recovery
(%)
Air drying 35
o
C 33, 34, 35
o
C 7, 9, 11 h 58.0
Sun drying asphalt
(amb. temp (31-38
o
C)
50, 51, 51
o
C 3, 3.3, 4 h 56.0
Sun drying
Plastic (ambient temp.
35-44
o
C)
44, 44, 45
o
C 4.5, 5, 6 h 55.8


4.2 Evaluation of farmers’ practice

4.2.1 Value losses due to field drying and sun drying
According to report of Cuong (2003) (presented at the seminar on post-harvest handling

in Can Tho dated 28/4/2003), field drying increased grain cracking by 10% resulting in
reduction of the head rice recovery by 8% during milling. This value is similar to the
experimental value of 9.6% reported by Thanh (1998) for IR64 variety. This loss of HRY
applied to full time field drying (3-4 days continuously). However, currently most of the
farmers apply a half time field drying (1-2 days) followed by half time sun drying (1-2
days). When full time sun drying is applied in the winter-spring crop, the following data
can be used for different varieties (Binh 2007):
Table 15: HRY (%) of different drying methods for some varieties of winter-spring crop
Variety Shade drying
(control)
Mechanical
drying (flat-bed
dryer)
Sun drying
(plastic, Ts =
42-48
o
C)
Difference between
mechanical and sun
drying (SD%)
OM1490
OM2517
Jasmine
Sticky rice
49.12
a
48.67
a
50.36

a
50.56
a
49.62
a
49.05
a
50.67
a
51.31
a
47.45
b
45.25
b
44.64
b
48.23
b
2.17
3.80
6.03
3.08
Note: the values with the same supper script letter in a row are not significantly different (P > 0.05). Ts is the surface
temperature of the grain layer.
The results presented in Table 15 suggests that flat-bed drying is as good as control
sample (in shade drying, SD) whereas, sun drying on plastic net as current practice of the
farmers reduced the HRY significantly (SD = 2.17 – 6.03%). For the OM2718 and AG24
varieties, average losses of HRY 3.77% can be used. In order to calculate the HRY losses
of the farmer who applied both field drying (FD) and sun drying (SD), the following

equation was introduced:

Losses of HRY = FD% f + SD% s


16
Where, f and m are the days of field and sun drying; FD = 8% is the losses of HRY due
to field drying (as reported by Cuong, 2003) and SD% is the losses of HRY due to sun
drying. The fractions f and s are assumed to be the constants in the following Table 16
depending on the actual drying situation:

Table 16: Loss factors f and s for combination of field and sun drying currently practiced
by farmers.
Actual drying method f s note
Full time sun drying
Full time field drying
Half time field drying and sun drying
0
1
0.5
1
0
0.5
3-4 days
3-4 days
1-2 days field drying and 1-2
days sun drying

Table 17: Losses of HRY in equivalent to losses of paddy (kg/100kg yield) for different
varieties in the MRD due to current practices of field and sun drying of the farmers

Variety OM1490 OM2718 Jasmine AG24
Losses (%)
7.32 8.97 10.45 7.87


On average, losses due to field and sun drying are around 8.7%. This loss is for winter-
spring season. The losses in the wet season (Summer-Autumn crop, July/August) are
expected to be much higher due to changeable weather of rain and storm. Data collection
from farmers will continue this year in wet season.

References:
Cuong, Doan Phu (2003). Results and experiences on pre and post harvest problems from 1998-2003. In: Workshop on current status
and developing trend for post harvest activities of the Mekong River Delta (April 28, 2003, in Vietnamese).
Thanh, Le van (1998). Study the effect of field drying, sun drying and mechanical drying on the milling rice quality of IR64 variety .
Study Report under DANIDA project, in Vietnamese.
Binh, Ngo (2007). Study the effect of sun drying and mechanical drying on the milling rice quality of different varieties. Study Report
under CARD 026/05VIE project.

17


5. Effect of milling on the head rice recovery

The head rice recovery will not only depend on the initial rice quality (existing cracks or
weaker grain), but also on the efficiency of the milling operation. The currently available
data are based on the surveys with the millers. In this work, actual milling loss data were
collected in two provinces, Tien Giang and Kien Giang.

There exist three systems of rice mills in both provinces:
1. The traditional white rice milling system: A complete rice milling plant without

polishing (accounts for 91%).
2. The brown rice milling system (accounts for 3%).
3. The final rice whitening/polishing plant (account for around 6% of total rice
mills).

In this work, the data collection was done on three milling types:
1. Small- <1 ton/hr
2. Medium- 1-4 ton/h
3. Large- >4 ton/h

The data were also collected in three types of hulling mills:
1. Hulling by stone disc huller
2. Hulling by rubber roll huller
3. Hulling by both stone disc huller and rubber roll huller
The data were collected by two methods:
1. Surveying with the millers (all three size rice mills)- in both provinces
2. Real data collection from the mills (four small and two medium size rice mills)- in
only Kien Giang province (the same data set will be collected in other province
this season).

The results are presented in Tables 18 and 19. The real data and data collected by survey
were quite coherent. Both data suggested that the head rice recovery in small scale mills is
the lowest and can be as low as 33%. Large rice mill had the highest of 55% head rice
recovery. In the actual ideal condition the head rice recovery and total rice recovery
should be around 59% and 69%, respectively (as rice is comprised of around 10% bran
and 20% husk). In literatures, the head rice recovery and total rice recovery have been
achieved as high as 60% and 70%. Therefore, there is still scope of improving the head
rice recovery even in large scale mills, let alone a poor performer small scale rice mills.

The importance of improving the quality of rice can be substantial. As far example, in

Kien Giang province, out of 715 rice mills, 67.6% are small, 28.1% medium scale and
4.3% large scale mills. Similarly in Tien Giang province there are more than 900 small
household mills. Simple facilities, product mainly supplied for local demand, not for a
commercial production, are the main causes leading to low rice recovery in a small scale
factory. By proper awareness, training of operators and maintance of mills the head rice
recovery can be substantially improved.


18
In Tien Giang province, the surveying data has also found that in the area where the
paddy is milled at high moisture content, 16-17% even 18%, has lower head rice yield
than the area where the moisture of the paddy is at 14-15% moisture.

Table 18: Head rice recovery data collected by surveying the millers in Kien Giang and
Tien Giang provinces
Scale
Grain moisture
(%)
Average head rice
recovery (%)
Broken rice
(%)
Small 16 47-48 18-22
Medium 16 50-52 17-18
Large 16 52-55 16-17


Table 19: Actual data collected from the mills in Kien Giang province
Scale Name of agencies
Capacity

(tons/hour)
Paddy
moisture
(%)
Milled rice
moisture
(%)
Head rice
recovery %
Duong Nguyet 0.8 15.5 15.8 48
Huynh Thanh Dung 0.4 15.2 15.1 42
Luu Van Thieu 0.7 14.2 13.7 50
Small
Coffee dehusker
(mobile)
0.2 15.3 15.2 33
Tan Phat 1.2 16 16.5 52
Medium
Luu Thanh Xem 1.5 15.9 15.9 49

In Kien Giang province, the survey results also suggested that the rice mills using rubber
roll huller had a better head rice recovery than those using stone disc huller or coffee grain
huller (Table 20).

Table 20: Head rice recovery (%) in mills having three different types of dehulling
systems in Kien Giang Province
Rice mill size Stone Disc
huller
Rubber
roll huller

Combined
(Stone+rubber)
Coffee grain
huller
Small
47 51 49 43
Medium
50 54 53 -
Large
- - 55 -


19

Conclusion and project intervention methods:
- Size of mill is an important factor that determines the losses. The small mills which
are used by small farmers showed a low head rice recovery. Medium and large scale
plants had a high recovery, but still it was far from ideal. The maximum head rice
recovery in large plants is still around 55%, a well below the ideal level (60%). This
means that the milling is another important factor to improve the head rice yield.
- It is necessary that the millers and farmers are aware of the efficiency of mills in order
to augment the value of their produce.
- The medium scale rice mills should be promoted even in village level to improve the
head rice recovery.
- The results obtained in this survey will be conveyed to the farmers, millers and mill
operators through training and workshops.

20

GENERAL CONCLUSION

• Any step of harvest or post-harvest contributes to a loss. This loss is the
summation of handling losses and value losses. For instance, improper harvesting
resulted in high shattering (handling losses) and low head rice recovery (value
losses). Similarly, sun drying in bad weather condition resulted in high spillage
(handling losses) and high colourful kernels and low head rice recovery (value
losses). The value losses can be converted into handling losses and inversely. As
calculated in section 4 on drying as an example, 1% reduction of head rice
recovery is equivalent to 0.36% loss of paddy. This concept is used to determine
the total losses of all chains in harvest/post-harvest process. The proportion of
small holder farmers in MRD possessing less than 0.5 ha is 14.1% (ranging from
8-30% depending on the province) who occupy around 4% of rice field. The
proportion of small holder farmers in MRD possessing less than 0.5 ha is 14.1%
(ranging from 8-30% depending on the province) who occupy around 4% of rice
field. The proportion of small holder farmers possessing 0.5 to 1 ha is 56% who
occupy about 32% of rice land. If 36% of rice land is considered to be the
percentage occupied by the small holder farmers in MRD then the contribution of
this group is a saving of 0.13 million tons of rice losses per season. This translates
to the increased cash income of the small holder farmers by more than 7.2%. If we
consider that the farmers recover 50% of the losses from the broken rice, a small
farmer holding about 1 ha of land with 5t/ha paddy yield can save up to US$25-35
per season. The small holder farmers who are part of the cooperatives will receive
the full benefit by applying the recommended techniques.
• The overall base line picture of the existing post harvesting practices and
introduced new harvesting methods were collected in specified locations in three
cooperatives during wet season. The harvesting time was found to be very
important in terms of losses occurring due to cracking. As a trend a late harvesting
was causing more rice cracking and consequently less head rice recovery. It can be
seen from the results that any intervention applied such as a correct harvesting
time practices, a farmers can reduce the losses significantly. Further confirmation
of this result will be undertaken in this dry season harvesting period (Feb/March).

• Similarly the losses due to rice breakage in various size scales of rice mills in two
provinces were collected. Small mills were having a significantly lower head rice
recovery. In fact, this is not good news for small farmers since these mills are
normally located in their area.
• More data will be collected on the rice cracking as influenced by drying methods
practiced by the farmers (such as sun drying) and also due to the introduction of
mechanical dryers.
• As an initial very rough estimate the following Tables (Table 21 and 22) depicts
the head rice recovery losses at different critical points of rice harvesting to
milling chain of activities. These values will be revised after the collection of
information during the dry harvesting season. These values should be taken very
cautiously.






21




Table 21: Estimated head rice recovery losses at various critical points

Critical loss points Estimated loss on head rice
recovery (%)
Harvesting time 3-28 (relative value)
Harvesting method 6-17
Shattering 1-3 (handling losses)

Drying losses To be undertaken
Milling (mainly small and
medium size mills)
6-13









Table 22: Conversion of head rice recovery losses into equivalent losses for estimation of
total losses at current situation and after intervention for improvement.
Equivalent Losses (%)
Critical loss
points
Handl
ing
losses
(%)
Value losses
in terms of
head rice
recovery (%)
Current situation
experimental
(0.36 x value loss)
Farmers

practice*
After
improvement
(ideal)
Harvesting
time
7-13.4 2.5-4.8 0.1-4.2 To be identified
Harvesting
method
5-15 1.8-5.4 To be
determined
To be identified
Threshing 0.8-1.4 0.8-1.4 To be identified
Shattering
1-3 1-3 1-3 To be identified
Drying losses
To be undertaken 7.3-10.5
Milling
(mainly
small and
medium size
mills)
8-12 2.9-4.3 To be identified
Total 9-18.9
*winter crop only (2007)
• In economic terms, a range of 9-18.9% losses (excludes drying losses) of the crop
can be estimated. At least 7.2% (= 9–1.8) of losses can be saved by improving the

22
harvesting practices, e.g. about 0.36 million tons of rice losses per season saved

out of 5 million tons of rice production in MRD. The dollar value of this volume
of rice is very significant. As such the average economic losses suffered by the
small farmers will be much higher than the average estimate. Their losses will be
towards the higher range due to their inability to apply efficient harvesting, drying
and milling technologies. With very limited capabilities of these farmers, the
improved technologies can be initiated to be applied through cooperatives. This is
the intervention approach which is being undertaken in this project.
• It should be pointed out here that due to the difficulty of sampling, analytical
infrastructure and availability of only one harvesting period during the duration of
the project, the baseline information available to date on drying is still limited.
After the data collection in the dry season, a better estimate of the base
information will be generated.

• As specified in the initial project proposal, the data generated will be used for
training the farmers, extension workers and millers. The training and information
is expected to improve post-harvest practices at various levels of post-harvest
chain.

• Economics of application of new harvesting methods, optimum drying conditions,
possibility of employing compact drying systems, impact due to increased
awareness as provided through training, technology adoption and spill over to
other areas will be assessed towards the end of the project (in final 6 months)
through actual observation and surveying the farmers in the cooperatives and their
vicinity areas.


23