SIMULATION OF FOOD GRAIN STORAGE MANAGEMENT SYSTEM
IN BANGLADESH
BLK. Bala M.A. Satter
Departmrnt of Farm Power & Machinery
Bangladesh Agricultural University
Mymensingh, Bangladesh
and
Md. Golam Mohiuddin
Department of Industrial and Production Engineering
Bangladesh University of Engineering and Technology
Dhaka, Bangladesh.
ABSTRACT
A system dynamics model of food grain storage, govt.
procurement and release, and import in Bangladesh is presented.
The simulated results of the model for govt. procurement and
release, and import policies are also presented. Finally, the
policy implications of the model are discussed.
INTRODUCTION
Bangladesh is a food deficit country and the deficit is met up
by import and food aid. Major foodgrains are rice and wheat.
The supply of the foodgrains has a profound effect on the
movement of the price of the foodgrains over time. The price is
lowest at the harvest and then gradually increases with storage
time until the next harvest. Such fluctuations have serious
impacts on farmer’s income and production. Small farmers need
cash just after harvest and can not hold the grain for a long
time. Hence, they are likely to be more affected by the
seasonal fluctuations of the price. Government procurement of
the grain at a high price just after harvesting can help to
prevent the large amplitude of the fall in price. This
procurement immediately after harvest stabilizes the price,
encourages production, prevents smuggling across the border and
ensures adequate supplies of foodgrain (Islam 1980). Again the
supply of foodgrain through rationing system at a subsidised
price will stabilize price and provide consumer welfare etc.
Thus, an effective procurement and release of the foodgrains
can iron out the seasonal variations of price and provide
consumer welfare by correcting influence on income
distribution. Hence an analytical tool is of vital importance
for the foodgrains management system essentially consisting of
procurement and release, storage and import of grains.
75
76 System Dynamics ‘90
Islam (1980) discussed the foodgrain procurement, input
subsidy and the public food distribution system in Bangladesh.
The salient features of the policy package are the procurement
of foodgrain by the government at fixed prices during the
harvesting season, the provisons of subsidies for important
inputs like irrigation and fertilizers and the distribution of
foodgrains to consumers at a price lower than the market
prices. He provides an overview of the various aspects of the
policy package. Furthermore, he suggested for effective
procurement of cereal grains immediately after harvest at a
price sufficiently high level. He maintained that the time
series data does not show any relation between quantity
procured, level of gross production and market procurement
price relatives . The procurement policy has not been very much
successful in reducing seasonal fluctuation of price.
Alamgir and Berlage (1973) used time series data to analyze
the foodgrain demands in Bangladesh over the period 1950/51 to
1968/69 and to make projections for the future. They used their
projections to calculate import requirements and market price
of local rice that would occur if the productions were not
realised. They also maintained that it is desirable to build a
substantial stock of foodgrain both through local procurement
and import in order to avoid unusally large fluctuations in
rice price due to fluctuations in rice production.
Berlage (1973) developed a dynamic programming model to
design foodgrain import and storage policy in Bangladesh. The
model is essentially a multistage one in time period with each
period consisting of beginning. stock, import ordering,
production, local procurement and release from the system.
Gibson (1978) developed a general system simulation model for
foodgrain sector in Korea and used the techniques from various
disciplines such as system design, economics, operations
research, linear and non-linear systems and automatic feedback
control theory. The model itself is a non-linear dynamic system
model that has time varying parameters. The model was
developed for use as an on-line tool for government decisions
regarding price, stock, storage and trade of grain.
Shahabuddin (1987) developed an econometric model of the
public foodgrain distribution system in Bangladesh. He analysed
the impact of different production Scenario on market price,
ration offtake, internal procurement, import and total
consumption. He also explored the impact of ration price,
ration quota and share of rice in the total ration distribution
on market price of rice and food grain consumption. The model
does not consider the short term scenario and the seasonal
effects. The model developed by Shahabuddin is linear and does
not include time lag. Bala et al (1990) developed a system
dynamics model of food grain procurement, release and import in
Bangladesh. The model of Bala et al does not consider the
System Dynamics '90 77
scenario of different production levels, and policy
implications of local procurement and release rates on market
price and consumption.
The purpose of this paper is to present a system dynamics
model of food grain procurement, release and import system in
Bangladesh and to analyse the impacts of different production
scenario and to explore the policy implications of local
procurement and rationing on market price and consumption.
A BRIEF DESCRIPTION OF FOOD GRAIN STORAGE SYSTEM IN BANGLADESH
In Bangladesh foodgrains are stored by the farmers to meet
their own consumption and for seeds. The produce thus retained
is estimated to vary from 10 to 100%. The average retention is
about 70% of the produce (Mohiuddin 1988). Foodgrains are
handled either in containers or in bulk. The container in
Bangladesh is the jute bag whereas the bulk grains are stored
in silos and bins (dool, berh, gola etc.) 90% of the available
storages are in bulk.
There are about eight different types of storage systems at
farm level use in Bangladesh. Among these, dool, berh and gola
are the common types of storage systems. Dool is a closed
bottom bamboo structure normally cylindrical in shape and it is
kept on a platform. The capacity of a dool varies from 240 to
450 kg. The berh is essentially an oval shaped dool and it is
made of woven bamboo. Its top and bottom ends are open and are
generally placed above a wooden or bamboo platform. The
capacity varies from 700 to 1200 kg. Gola is made of bamboo
generally cylindrical or rectangular in shape. It may be indoor
or outdoor type. The roof of a outdoor gola is made of
corrogated tin or thatched. Its capacity varies from 4000 to
40,000 kg (Bala 1989).
Public foodgrain storage system includes local supply depots
‘LSD’ (at the rural and district urban areas), central storage
depots ‘CSD’ (at regional level) and bulk silos (at port and
national level). Local supply depots and central storage depots
are flat type of godowns where foodgrains are stored in jute
bags (Mohiuddin 1988).
The food deficit in Bangladesh is met up by large imports.
The average annual import of foodgrain in the last few years
was about 2.2 million tons (Mohiuddin 1989). To handle the
imported and internally procured food grain and to develop a
food security stock as against natural calamities, the public
distribution and storage system has been developed in
Bangladesh.
78 System Dynamics ’90
MODELLING OF THE SYSTEM
Food grain storage management system in Bangladesh is a complex
socio-economic system and the modelling of such a complex
system is a formidable challenge. System dynamics is the most
appropriate technique to handle such a complex system. To
simplify the presentation of the model, a verbal description is
followed by causal-loop diagram.
The rural storage stock is increased by production of grain
and depleted by the consumption and local procurement. The
government stock is increased by procurement and import and
depleted by release for sale. It is postulated that the amount
procured depends both on domestic production and the
procurement price relative to market price. The ration quantity
is assumed to depend on govt stock and ration price relative to
market price.The stock for sale is depleted by consumption.
Available food supply from rural stock and govt. stock would
determine the open market price and hence the per capita
consumption. The per capita food consumption influnces the
death rate. Procurement and release policies are of vital
importance for stabilization of price and to provide a minimum
subsistence amount of food for consumption.
Again, food grain production is seasonal and weather
conditions influnce the production. Hence a buffer stock is
needed to even out the fluctuations of price. However, there
should be a security of food grain against the natural
disasters. For reasons of national security and to assure
adequate buffer stocks for seasonal price stabilization, it is
necessary for the government to maintain reserve stock at some
minimum level. The govt. reserve stock should not dwindle below
minimum acceptable levels regardless of procurement and release
of foodgrains. Shortage of food grain is to be met up by
import. There will be some delay in ordering and receiving of
the food grains. The grains on order will be equal to the
amount ordered minus the amount delivered to govt. stock. The
order should be placed well ahead of the time i.e. when the
govt. stock is equal to the secuirity stock plus the amount
required for consumption during lead time. The lead time
requirement is computed from predicted three month’s
consumption minus expected production over this period. When it
is negative no import is needed. A computer model based on
system dynamics methodology of Forrester (1968) was developed
to explore the dynamic interrelationships among production,
procurement, import, ration offtake, market price and
consumption. The flow diagram of the foodgrain management
system is shown in Fig. 1. The model was programmed in
professional DYNAMO.
System Dynamics '90 73
+
AURAL STOCK
procureuent :
PROCUREMENT.
PRICE RELATIVE
STORAGE
CAPACITY
LEAD TIME
PRODUCTION s +
PROCUREMENT. i
RELEASE RATION >
+
+ + _
FOOD GRAIN
PRODUCTION
RATE PRICE RELATIVE
+ ca
GRAIN “STOCK CONSUMPTION
+
ON ORDER RATION PRICE
LEAD Time Food +7 BIRTH
. PER CAPITA
< + )
Vo LEAD TIME POPULATION
ef %
st. CONSUMPTION \
SECURITY
RESERVE
*NoeatH
FIG-1 CAUSAL LOOP DIAGRAM OF FOOD GRAIN PROCUREMENT RELEASE
. AND IMPORT SYSTEM IN BANGLADESH.
SIMULATED RESULTS AND DISCUSSIONS
Basic mode
The model was simulated for local procurement based on
production and procurement price relative to market price,
release rate based on govt. stock and ration price relative to
market price, and govt. inport order based on lead time
requirement plus security reserve for a simulation period of .35
months starting from January, 1989. The starting food supply is
1.275 million tons and the govt. stock is 0.725 million tons.
The security reserve is assumed to be 1.5 million tons.
En) System Dynamics ‘90
Simulated food supply, govt. stock and market price are shown
in Fig. 2 for a period of 35 months starting from January,
1989. Considerable fluctuations in the supply of food grains
are mainly due to the seasonal pattern of crop production and
these fluctuations are left even after internal procurement of
the grains. Govt. stock fluctuation is considerably low and the
dynamic behaviour of-the govt. stock gives a measure of the
storage capacity required. The simulated results in Fig. 2
gives an indication of the storage capacity of 3.75 million
tons. The simulated govt. stock increases with time because of
the food requirement for increased population and the
requirement of price stabilization to provide consumer welfare.
However, the price fluctuation does not die out with time but
the amplitudes of the fluctuation are considerably low.
FooD suPeLy
Govt. stock
Food suPPLY (0,7 £6) MY]
[#———-———covrr. stock (7 £6) wT
PRICE (0,7 £4) TAKA/TON
i 20 25
SIMULATED PERIOD (MONTH) 30 *
Fig. 2. Simulated food supply, govt. stock and
market price : basic mode.
Simulated population, food percapita ratio and the imported
foodgrain receiving rate are shown in Fig. 3. The fluctuation
of food percapita does not die out with time but the amplitudes
of the fluctuations are considerably low. This assures uniform
consumer welfare throughout simulated period. Again, the
foodgrains arriving at the ports are to be handled timely and
speedily to avoid any price instabilization and food crisis
Simulated results indicate that Bangladesh ports should have an
annual unloading capacity of about 2.4 million tons. Mohiuddin
System Dynamics '90 81
(1989) suggested that Bangladesh ports should have a minimum
annual unloading capacity of 3 million tons at both Chittagong
port and Mongla port. Thus, the simulated result agrees well
with the suggested unloading capacity required.
RECEIVING RATE
FOOD PER
CAPITA RATIO.
POPULATION (100663 107 £6) —————_}
FOOD PER CAPITA RATIO (0,7):
[oo neceivine Rate (0,7 £5)MT/MONTH
4 1 : 1
1 1. al 33
10 15, 20
SIMULATED PERIOD (MONTH)
Fig. 3. Simulated population, food per capita
ratio and imported foodgrain receiving
rate : basic mode
Policy planning mode
The model was also simulated to demonstrate the potentiality
of the model for policy implication for + 25% increase of food
production levels, procurement fraction, release fraction, and
both procurement and release fractions. The starting conditions
are assumed to be same for all the cases as in the basic mode.
Fig. 4 shows , simulated food supply, govt. stock and market
price for +25% increase in the levels of food grain production.
A reduction of 25% food grain production causes about 40%
increase of price from the basic run shown in Fig. 2 while an
increase of food production of 25% results about 20% reduction
in market price. Shahabuddin (1987) also reported similar
results. The low level of market price has resulted from poor
procurement and this may affect the farmers seriously. The high
level of market price is mainly due to the shortage of food
System Dynamics '90
grain. However, the patterns of the price fluctuations are
similar in both the cases. The food supply and govt. stock are
large for 25% increase of food production in comparison with
those for 25% reduction of food production. But, the govt stock
is larger for 25% increase of production than the food supply
for 25% decrease of food production.
Govt. sTOCK.—=}
FOOD SUPPLY.
GOVT. STOCK (0,8 £6) MT ——______
ice @, 8 £4) TAKA/TON————————4
oo surety 00,866) MT.
o cy 10 1S 20 25 30 Ec}
SIMULATED PERIOD (MONTH)
Fig. 4. Simulated food supply, govt. stock and market
price : policy planning mode, 25%
reduction of food production and ....... 25%
increase of food production level.
Fig. 5 shows simulated population, food per capita ratio,
and imported grain receiving rate. A reduction of 25% crop
production causes 25% reduction in consumption which results a
decrease in population while an increase of 25% food production
causes 25% increase in consumption which results an increase in
population. The imported grain is still higher in the years of
good harvest. This is mainly due to the poor level of internal
procurement that can not satisfy the minimum requirement for
food reserve.
System Dynamics '90 83
‘008 PER CAPITA RATIO (0,7)
CEIVING RATE (0,7 £5) MT/ MONTH.
—FOPULATION (100 £6, 107 £6)
A,RECEIVING Foo PER
JS RATERS A CAPITA RATIO
Fig. 5. Simulated population, food per capita ratio
and imported foodgrain receiving rate : policy
planning mode ; ———— 25% reduction of food
production level and .... 25% increase of food
production level.
Fig. 6 shows simulated food supply, govt. stock and market
price for +25% increase of procurement fraction. The price is
slightly higher for 25% increase of procurement fraction but
very close to each other. The govt stock is higher for 25%
increase of procurement fractions than those for 25% decrease
of procurement fraction. But the case is reverse for food
supply. The slightly higher price is mainly due to the
relatively higher reduction of rural stock in comparison with
the amount of grain released through rationing system.
System Dynamics °90
FOOD SUPPLY
GOVT. STOCK (0,767 ut
price (0, 7 £4) TAKAsTON ———————
(Foon suey (0,7 £6) MT
0 10 20
SIMULATED PERIOD (MONTH)
Fig. 6. Simulated food supply, govt. stock and market
price : policy planning mode ; 25%
reduction of procurement fraction and
25% increase of procurement fraction.
Fig. 7 shows simulated population, food per capita and
imported grain receiving rate for +25% increase of procurement
fraction. The per capita consumption is almost equal for both
cases. The imported grain is considerably lower for increased
procurement because of the fact that the food reserve is
maintained using a greater proportion of food from local
procurement. The population changes are also almost equal for
both cases because of the almost equal per capita food
consumption.
Simulated food supply, govt. stock and price, are shown in
Fig. 8. for +25% increase of release fraction. Food supply is
larger for 25% increase of relaesed fraction. But the case is
reverse for govt stock. Higher release causes higher depletion
of the govt. stock resulting high level of food supply. The
market price is slightly higher in case of reduced release
fraction. However, the difference between the prices for both
the cases are very small.
System Dynamics ‘90
Fig.
10766) ——
FOOO PER CAPITA RATIO (0,7)
RECEIVING RATE ,7E5) MT/MONTH
POPULATION (100 £6:
necevine /\
\
70 15 20.
SIMULATED PERIOD (MONTH)
Simulated population, food per capita ratio
and imported foodgrain receiving rate : policy
planning mode; 25% reduction of procurement
fraction and ...... 25% increase of procurement
fraction.
FooD suPPLY
FOOD SUPPLY (0,7 £6) MT
PRICE (0,7 £4) TAKA/TON
}———--—+—-— Govt. stock (0,7 £6) MT.
10 is 20 2s 30 35
SIMULATED PERIOD (MONTH)
Simulated food supply, govt. stock and market
price : policy planning mode ; 25%
reduction of release fraction and ....... 25%
increase of release fraction.
86 System Dynamics '90
Simulated population, food per capita and imported grain
receiving rate for +25% increase of release fraction are shown
in Fig. 9. Per capita food consumption is almost same for both
cases. But higher amount of imported grain is required for
increased release fraction to maintain the security reserve and
the food consumption level. The population is almost same for
both cases.
POPULATION: 4
1%]
|
ss
228). t
ges
ge8
ges
wiz :
Bee rs
fear ia
gis A i
bE? Po /\ roo | \
239 receiving | \ i \ oper | OY
gee fp. RATE ‘ 1 \eaeirat \
3 K H 1 Naaor
2 NS i ONMRPP OS
[Ti in WAY
if Le tkX\ >) a
\ j ae St :
Sy
| A A : : _|
7 30 35
Fig. 9. Simulated population, food per capita ratio
and imported foodgrain receiving rate : policy
planning mode ; 25% reduction of release
fraction and ....... 25% increase of release
fraction.
Simulated food supply, govt stock and market price are shown
in Fig. 10 for +25% increase of both procurement fraction and
release fraction simultaneously. The price is almost same for
both cases. The differences in the levels of govt. stock and
food supplies are also small.
Simulated population, food per capita and imported grain
unloading rate for simultaneous +25% increase of both
procurement fraction and release fraction are shown in Fig. 11.
These are almost same in both the cases.
System Dynamics '90
87
Fig.
Fig.
11.
FOOD SUPPLY (0,7 £6) MT
|
GOVT. STOCK (0,7 £6) MT
PRICE (0,7 £4) TAKA/TON
i 2 1 . 4
10.
SIMULATED PERIOD (MONTH)
Simulated food supply, govt. stock and market
price : policy planning mode ; 25%
simultaneous reduction of both procurement
fraction and release fraction, and ...... 25%
simultaneous increase of both procurement
fraction and release fraction.
RECEIVING AATC (0,7 ES) MT/ MONTH
POPULATION
RECEIVING RATE
PX Foon pe
cADITA
RATIO.
Lf i
Suuatee etnen downy
Simulated population, food per capita ratio and
imported foodgrain receiving rate : policy
planning mode; 25% simultaneous reduction
of both procurement fraction and release
fraction, and ....... 25% simultaneous increase
of both procurement fraction and release
fraction.
88 System Dynamics '90
The implications of the simulated policies are that market
price and food per capita are not sensitive to procurement
fraction and or release fraction. These are highly sensitive to
the different levels of foodgrain production. The amplitudes of
fluctuations are small but do not die out with time.
Procurement fraction significantly influences the importation
The higher is the procurement fraction, the lower is the amount
of the imported food grain for govt. stock. Again an increase
of release fraction demands for higher amount of imported
grain. Hence, the amount of imported food grain in the years of
good harvest can be either reduced or eliminated by improving
the internal procurement programme through increased local
procurement.
Procurement programme should be enhanced in the year of good
crop to raise the market price and at some time either to
reduce or eliminate the importation of the grain for
maintaining the minimum food reserve required. The case is
reverse for the bad year. Establishment of annual food plans
and importation of deficit will not assure domestic seasonal
price stabilization unless an active role is played in the
domestic grain procurement and release through rationing system.
For reasons of national security and to assure adequate
buffer stocks during the bad years, adequate importation should
be made. This would necessitate a timely and speedy unloading
of the imported grains. Hence, an adequate unloading facility
is an urgent need for domestic price stabilization during the
years of bad crops.
We have clearly indicated the implication of the model for
policy planning. This model would not replace policy planners
and grain storage managers, rather this model would provide
them greater insight and better understanding of the complex
grain storage management system. For practical application of
the model further requirements and elabortions of the model are
required and the success of the policies for the food grain
management system depends upon the systems for their effective
implementation also. In summary, the model is useful in
examining the alternative foodgrain storage management
policies.
ACKNOWLEDGEMENT
This work is a part of the research project Studies on
Traditional Storage Systems for Paddy in Bangladesh and
Development of Low Cost Efficient Storage Systems for Use at
the Farm Level supported by Bangladesh Agricultural Research
Council and the financial assistance by BARC is gratefully
acknowledged.
System Dynamics ‘90 &°
REFERENCES
Alamgir, M and L.J.J.B. Berlage. 1973. Foodfrain (rice and
wheat ) Demand, Import and Price Policy for Bangladesh. The
Bangladesh Economic Review 1(1) : 25-58.
Bala, B.K. 1989. Studies on Traditional Storage System for
Paddy in Bangladesh and Development of Low Cost Efficient
Storage Systems for Use at the Farm Level. Annual Report.
Department of Farm Power and Machinery, Bangladesh
Agricultural University, Mymensingh.
Bala, B.K., M.A. Satter and M.S. Alam. 1990. System Dynamics
Simulation of Foodgrain Procurement, Release and Import
System in Bangladesh. Presented at the 34th Annual Convention
of the Institution of Engineers, Bangladesh. Dhaka.
Berlage, L.J.J.B. 1973. An Application of Dynamic Programming
Models to Foodgrain Import and Storage Policy in Bangladesh.
The Bangladesh Economic Review 1(4) : 341-374.
Forrester, J.W. 1968. Principles of Systems. Cambridge, Mass :
Wright Allen Press.
Gibson, F.J. 1978. The Grain Management Programme Model. In
Agricultural Sector Planning : A General System Simulation
Approach, ed. G.E. Rossmiller. Agricultural Sector Analysis
and Simulation Projects, Department of Agricultural
Economics, Michigan State University. Ann Arber, Michigan.
Islam, R. 1980. Foodgrain Procurement, Import Subsidy and the
Public Food Distribution System in Bangladesh. The Bangladesh
Develpment Studies 8(1 & 2) : 89-120.
Mohiuddin, M.G. 1988. Development of Rural Storages in
Bangladesh. Presented at a Seminar on Rural Industrial
Technology, Dhaka.
Mohiuddin, M.G. 1989. Bangladesh Foodgrain Imports and the Port
Unloading Facilities for the year 2000. Presented at the 14th
Annual Bangladesh Science Conference. Gazipur.
Shahabuddin, Q. 1987. Interrelations in the Public Foodgrain
Distribution System in Bangladesh - An Econometric Analysis.
The Bangladesh Development Studies 15(1) : 57-81.
