PLACE IN RETURN BOX to remove this checkout from your record.

TO AVOID FINES mum on or before date due.

 

DATE DUE, DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

#

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Institution

“nod—ll...

COMPUTER SIMULATION OF
COMBINE HARVESTING AND HANDLING
OF SUGAR CANE IN BARBADOS

By

Winston O'Neale Harvey

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY

Department of Agricultural Engineering

1982

ABSTRACT
COMPUTER SIMULATION OF

COMBINE HARVESTING AND HANDLING
OF SUGAR CANE IN BARBADOS

BY

Winston O'Neale Harvey

The broad objective of this study was to improve the efficiency of
combine harvesting of sugar cane in Barbados. The harvesting process was
broken down into two subsystems: a field subsystem and a factory yard
subsystem. Two computer simulation models, structured in GASP IV simu-
lation language, were developed to model the operations involved in these
systems. Mbdel FIELDOP simulated the activities involved in the harvest-
ing and loading of cane in the field, and in its transportation to the
factory for processing. Model FACYARD simulated the weighing and unload-
ing activities performed on cane transport units at the factory.

Following validation of the models, four different factory yard
configurations and eight different field equipment combinations were
simulated. Model parameters varied were the number of factory yard
scales and the numberscnffield tractors and cane transport wagons assigned
to a harvester.

Output from the models included utilization factors for the various
component machines, daily cane delivery from the field system, and daily
amounts of cane handled by the factory yard system. This output was fed
into a cost program which calculated unit harvesting costs and total
annual cane delivery for the equipment combinations simulated.

Results indicated that a second scale at the factory can reduce the

humory residence time of transport units by 88 percent, increase combine

 

 

 

 

harvester utilization efficiency by 50-60 percent, increase daily cane
receipts at the factory by more than 30 percent, and eliminate milling
lost time due to lack of cane. Harvesting systems using two field
tractors, rather than one, were also shown to be capable of consistently
delivering 15-25 percent more cane per day.

The economic analysis demonstrated that harvesting cost per tonne
can be significantly reduced by either adding a second field tractor,
increasing the number of cane wagons assigned to a harvester, installing
a second weigh scale at the factory, or a combination of these.

A sensitivity analysis revealed that, as a single measure, adding
a second scale to the factory would be two to three times more effective

in reducing costs then would either of the other measures.

Approved

 

Major Professor

Approved

 

Department Chairman

ACKNOWLEDGEMENTS

The author wishes to express sincere gratitude to his Major Pro-
fessor, Dr. Merle L. Esmay, for his support and assistance throughout
this study, and to Dr. Ajit K. Srivastava, Dr. Robert H. Wilkinson and
Dr. J. Roy Black for serving on his guidance committee.

Special thanks are also expressed to the following:

Mr. R.A. Baynes of Barclays Bank Caribbean Head Office, who was
instrumental in the negotiationg of a loan to finance the research
phase of the study in Barbados.

Mr. G.B. Hagelberg of the Ministry of Finance and Planning,
Barbados, who negotiated some additional funds to help defray research
expenses.

Dr. John R. Ogilvie, Chairman, School of Engineering, University
of Guelph, Canada, who offered some valuable technical assistance.

Mr. Michael Gill of Ashbury Plantation, Mr. Geoffrey Skeete of
Edgecumbe Estates Ltd. and Mr. Geoffrey Armstrong of Sunbury, whose
harvesting operations were the main ones monitored during the data
collection exercise.

The Managements of Carrington, Foursquare and Andrews Sugar
Factories, who readily accommodated the monitors on their prOperties.

Ms. Joan Antrobus of the office at Carrington Factory for making
weekly factory statistics available.

Mr. Ishmael Roett of Barbados '0' Level Institute for his assist-

ance in selecting five students who turned out to be excellent monitors.

ii

The organizers of the Latin American Scholarship Plan for American
Universities, for their financial assistance in the form of a LASPAU
Fellowship.

The University of the West Indies, for granting the author leave
to take up the Fellowship.

Ms. Lolly Carrington for her friendship and encouragement while
she was also a student at Michigan State University.

Finally, the author wishes to express his deepest thanks and
appreciation to his wife, Ena, for her unfailing moral support and
encouragement and for typing a major section of the first draft of

the dissertation.

iii

2.4

4.3

TABLE OF CONTENTS

INTRODUCTION. . . . . . . . . . . .

Background. . . . . . . . . . . . . . . .
Objective of the Study. . . . . . . . .
Approach to the Study . . . . . . . . .

OVERVIEW OF THE BARBADOS SUGAR INDUSTRY . . .

Introduction. . . . . . . . . . . . . .
The Role of the Sugar Industry in the .
Barbadian Economy . . . . . . . . . . . . . .
Structure of the Industry . . . . . . . . . .
.3
3

2 .1 The Production Sector . . . . . . . .
2. 2

The Processing Sector . . . . . .

Markets for Barbados' Sugar . . . . . .

2.4.1 The European Economic Community . .
Market. . . . . . . . . . . . . .
2.4.2 The World Market. . . . . . . . . . .
2.4.3 The Domestic Market . . . . . . . . .
Review of Mechanical Harvesting of Sugar
Cane in Barbados. . . . . . . . . . . . . . .

LITERATURE REVIEW . . . . . . . . . . . . . .

Introduction. . . . . . . . . . . . . . . . .
Symbolic and Mathematical Models. . . . . . .
Computer Simulation Models. . . . . . . . . .
Sugar Cane Simulation Models. . . . . . . . .
Computer Simulation Languages . . . . . . .

METHODOLOGY AND MODEL DEVELOPMENT . . . . . .

Introduction. . . . . . . . . . . . . . . . .
Mbdel Development Approach. . . . . . . . .

4.2.1 An Overview of the GASP IV Simulation
Language . . . . . . . . . . . . . .

The Field Operations. . . . . . . . . . .

4.3.1 Objectives. . . . . . . . . . . . . .
4.3.2 Activities and Events . .

iv

Page

LDC-9H

. 10
. l4

. 16

17
. 19

20

. 21

27

27
28
37

. 42

48
52
52
53

55

. 60

. 60

62

4.4

4.5

CHAPTER 5

5.3

5.4

5.5

CHAPTER 6

6.1

6.2

TABLE OF CONTENTS (cont'd.)

The Factory Yard Operations. . . . . .

4.4.1
4.4.2

Obj actives O O O O O O O O O O O O O O 0
Activities and Events. . . . . . . . . . .

Data Acquisition and Analysis.

4.5.1 Data Acquisition . . . . . . . . . . . . .
4.5.2. Data Analysis. . . . . . . . .
4.5.3 The Data Analysis Program. . . . . .

FIELD AND FACTORY YARD COMPUTER SIMULATION
mnu s O O C O 0 O O O O C C O O 0

Introduction . . . . . . . . .
Field Operations Simulation Model. . . . . .

The System Environment . . . . . . . . . .
Entities . . . . . . Q . . . . . . . .
Endogenous Variables . . . . . . . . . .
Output Variables . . . . . . . . . . .

The Field Operations Simulation Program (FIELDOP).

5.3.1 General Flow through Program FIELDOP .
5.3.2 Input Data used for FIELDOP

The Factory Yard Simulation Model. . . . . . . . .

The System Environment . . . . . . . . . .
Entities . . . . . . . . . . . . . . . . .
Endogenous Variables . . . . . . . . ._. .
Output Variables . . . . . . . . . . . .

The Factory Yard Simulation Program (FACYARD). .

5.5.1 General Flow through Program FACYARD . . .
5 5 2

Input Data Used for FACYARD. . . . . . . .

RESULTS AND DISCUSSION . . . . . . . . . . . .
Introduction . . . . . . . . . . . . . . . . . . .

Results from the Factory Yard Simulation Model .

6.2.1 FACYARD Results for Current Yard
Configuration . . . . . . . . . .
6.2.2 FACYARD Results for Extended Work

Period . . . . . . . . . . . . .

Page
. 64

. 64
. 66

. 68

. 68
. 70

. 75

. 75
. 76

. 76

. 80
. 80

. 81

. 81

.104
.104
O 104
.106
.106
.110

.110
.115

.121
.121

.121

.124

.125

TABLE OF CONTENTS (cont'd.)

6.2.3 FACYARD Results for Configuration with
1 Weigh Queue and 2 Weigh Scales. . . . . .
6.2.4 FACYARD Results for Configuration with

2 Weigh Queues and 2 Weigh Scales . . . . . .
6.3 Results from the Field Operations Simulation Model. .
6.3.1 Validation of model FIELDOP . . . . . . . . .
6.3.2 Output from FIELDOP for Different Equip-
ment Combinations . . . . . . . . . . . . . .
6.4 General Discussion. . . . . . . . . . . . . . . . . .
6.4.1 Implications Associated with an Extended
work Per 10d 0 C O O O O O O O O O O O I O O 0
6.4.2 Implications Associated with Two-Scale
Configurations. . . . . . . . . . . . . . . .
6.4.3 Proposals for Re-organization of Carring-
ton Factory Yard. . . . . . . . . . . . .
6.5 Economic Analysis . . . . . . . . . . . . . . . . .
6.5.1 Calculation of Harvesting Machinery Costs .
6.5.2 Sensitivity of System Cost and Output to
Various System Parameters . . . . . . . . . .
6.6 The Projected Requirement of Cane Combine Harvesters.
6.6.1 The Theoretical Minimum Harvester
Requirements. . . . . . . . . . . . . . . .
6.6.2 Required Number of Harvesters for
Various Simulated Annual Output Levels. .
CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS . . . . . . . . .
7 O 1 COnClusionS O O O C O O C O O O O O O 0 O O O O 0 O I
7.2 Recommendations . . . . . . . . . . . . . . . . . . .
7.3 Suggestions for Future Work . . . . . . . . . . .
APPENDICES
Appendix I Data Collection Manual for Monitors . . . . .
Appendix II Fortran Listing of Data Analysis Program
DATANAL O O O O O O C O O O O O O O O O 0
Appendix III Fortran Listing of Factory Yard Operations

Program FACYARD . . . . . . . . . . . . . . .

vi

Page

.125
.128
.128
.130
.134

.142

.142
.145
.146
.151
.151
.153

.162

.162
.163
.165
.165
.166

.167

O 168

.174

.196

Appendix IV

LIST OF REFERENCES .

TABLE OF CONTENTS (cont'd.)

Fortran Listing of Field Operations
Program FIELDOP. . . . . . . . . .

vii

LIST OF TABLES

TABLE PAGE

3.1 Comparison of some contemporary simulation
languages. . . . . . . . . . . . . . . . . . . . . . . 51

4.1 Events monitored for the field operations. . . . . . . 65
4.2 Events monitored for the factory yard activities . . . 67
5.1 Entities and their associated attributes (FIELDOP) . . 82
5.2 Files used in model FIELDOP. . . . . . . . . . . . . . 83
5.3 Non-GASP user input variables used in FIELDOP. . . . . 85
5.4 Statistical distributions used for activity times

in model FIELDOP . . . . . . . . . . . . . . . . . . . 90

5.5 Input parameters for distributions used in FIELDOP . . 91
5.6 FIELDOP variables monitored by GASP IV subroutine

COLCT. O O O O O I O O O O O 0 O O O O I O O O O I I O .103
5.7 FIELDOP variables monitored by GASP IV subroutine

TIMSTO O O O O O O O O I O O I O O I O O O O O O O O O 103
5.8 List of non-GASP variables used in FACYARD . . . . . . 107
5.9 Files used in model FACYARD. . . . . . . . . . . . . . 108
5.10 Statistical distributions used for activity times in

in model FACYARD . . . . . . . . . . . . . . . . . . . 108
5.11 Attributes used in model FACYARD . . . . . . . . . . . 109
5.12 FACYARD variables monitored by GASP IV subroutine

TIMST. I O O O O O O O I O O O O O O O I O O O O O O O 117
5.13 FACYARD variables monitored by GASP IV subroutine

COLCT. o o o o o o o o o e e o o o o o o o o o o o o o 118
5.14 Input parameters and distributions used in FACYARD . . 119
6.1 Output from model FACYARD for current yard config-

uration at Carrington Factory (Simulation time =

700 minutes) . . . . . . . . . . . . . . . . . . . . . 123

viii

TABLE

6.2

6.3

6.4

6.8

6.9

6.10

6.11

6.12

6.13

6.14

LIST OF TABLES (cont'd.)

Comparison of observed and simulated values
for selected system performance indicators
for factory yard operations. . . . . . . . . . . . .

Output from model FACYARD for extended work
period at Carrington Factory (Simulation time =
960 minutes) I I I I I I I I I I I I I I I I I I I I

Output from model FACYARD for configuration
with l weigh queue and 2 scales (Simulation time
700 minutes) . . . . . . . . . . . . . . . . . . .

Output from model FACYARD for configuration with
2 weigh queues and 2 scales (Simulation time -=
700 minutes) . . . . . . . . . . . . . . . . . . . .

Output from model FIELDOP for 10:1:2:2: equipment
combination. . . . . . . . . . . . . . . . . . . . .

Comparison of observed and simulated values of
selected system performance indicators for field
operations I I I I I I I I I I I I I I I I I I I I

Equipment combinations simulated for field Opera—
tions I I I I I I I I I I I I I I I I I I I I I I I I

Machinery specifications and cost structure assumed
for combine harvesting of sugar cane based on 1982
figures I I I I I I I I I I I I I I I I I I I I I I I

Estimated cost of combine harvesting of cane for a
field system with 2 field tractors.(One factory
yard scale). . . . . . . . . . . . . . . . . . . . .

Estimated cost of combine harvesting of cane for
a field system with 1 field tractor.(0ne factory
yard scale) I I I I I I I I I I I I I I I I I I I I I

Estimated cost of combine harvesting of cane for a
field system with 2 field tractors. (Two factory
yard scales) . . . . . . . . . . . . . . . . . . .

Estimated cost of combine harvesting of cane for
a field system with 1 field tractor. (Two factory

yard seal e8) I I I I I I I I I I I I I I I I I I

Response of total system cost/tonne and total
annual output to changes in the number of wagons . .

ix

PAGE

124

. 126

127

129

. 131

. 132

. 135

154

156

. 157

. 158

159

160

TABLE

6.15

6.16

6.17

LIST OF TABLES (cont'd.)

PAGE

Response of cost per tonne of cane harvested
to an additional field tractor or factory
scale I I I I I I I I I I I I I I I I I I I I I I I I I I 161

Response of annual cane delivery from the field
system to an additional field tractor of factory
scale I I I I I I I I I I I I I I I I I I I I I I I I I I 161

Cane delivery rates and the number of harvesters
required for various equipment combinations . . . . . . . 164

FIGURE
5.9
5.10
5.11
5.12
5.13
5.14
5.15

6.1

6.2

6.3

6.4

6.5

6.6

6.7

6.8

6.9

LIST OF FIGURES (cont'd.)

Flowchart through subroutine SHTDN. . . . . . . . .
Flowchart through subroutine STRTUP . . .

Causal loop diagram for the factory yard system
Flowchart through FACYARD , . . , .
Flowchart through subroutine ARIVAL
Flowchart through subroutine ENDWGH , , , ,
Flowchart through subroutine ENDULD , . . . . . . .

Simulated daily cane delivery vs. number of wagons
in SyStem I I I I I I I I I I I I I I I I I I I I I

Simulated daily cane delivery vs. travel distance
to factory for mean factory residence time of 9.0
minUteS I I I I I I I I I I I I I I I I I I I I

Simulated daily cane delivery vs. travel distance
to factory for mean factory residence time of
73'98 m1nu£es o o e o o o o o o o o o o o e o o o o o

Harvester utilization vs. travel distance to factory
for mean factory residence time of 9.0 minutes,-,

Harvester utilization vs. travel distance to factory
for mean factory residence time of 73.98 minutes, ,

Road tractor utilization vs. travel distance to
facrory I I I I I I I I I I I I I I I I I I I I I I I

Current layout of, and current path of cane transport
vehicles through Carrington Factory Yard, , , , , , ,

Proposal #1 for re-organization of Carrington Factory
Yard I I I I I I I I I I I I I I I I I I I I I I I I I

Proposal #2 for re-organization of Carrington Factory
Yard I I I I I I I I I I I I I I I I I I I I I I I I I

xi

I I I I I I I I

PAGE

99

. 100

. 105

I 111

113

. 114

. 116

. 136

. 138

. 139

. 140

. 141

. 143

I

147

148

150

FIGURE
1.1
2.1

2.2

3.1

3.2

3I3

4.1

5.3
5.4
5.5
5.6
5.7

5.8

The geographic location of Barbados. .
The location of estate and small holder land .

Percentage of small holders producing various

LIST OF FIGURES

amounts of cane.

Symbolic model of closed circuit queuing system.

Generalized models representing the behavior of

systems. .

Regions in the (B1, BZ) plane for various distribu-

tions . .

Flow process chart of Operations performed on each

transport unit .

Basic modes of GASP IV control .

Functional flowchart of a GASP IV program.
Activities involved in the field subsystem .

The 2- and 2- parameter forms of the Gamma

distribution . .

Flowchart through data analysis program DATANAL.

Causal loop diagram for the field system .

Flowchart

Flowchart

Flowchart

Flowchart

Flowchart

Flowchart

Flowchart

through
through
through
through
through
through

through

FIELDOP. .

subroutine

subroutine

subroutine

subroutine

subroutine

subroutine

xii

I I I

ARFLD

STLD.

NDLD.

ARHTR .

ARFCT .

DPFCT .

PAGE

. 12

. 13

. 29

. 30

. 31

. 54

. 59

. 61

. 63

. 72

. 73

. 78

. 84

. 87

. 89

. 92

I 95

. 95

. 96

CHAPTER 1

INTRODUCTION

1.1. Background

Barbados, located 59° 37' east longitude and 13° 4' north latitude,
is the most easterly of the archipelago of islands in the Caribbean Sea.
It is a small island of 430 square kilometres, 43,000 hectares of land
and a population density of approximately 580 persons per square kilo—
metre. (Figure 1.1).

Colonized by the British for over 300 years, the island has an
economic history which is divisible into three main phases. The first
phase, 1627-1650, was characterized by a peasant economy in which an
inexcessive number of settlers pursued a relatively self-sufficient and
diversified economy based on the production of tobacco, cotton and
indigo.

The second phase, 1651—1950, saw the island transformed into a
rigid and 'lop-sided' export oriented plantation economy dependent
almost exclusively on sugar cane production. During this period
scarcely any effort was made to exploit what limited opportunities
existed for diversification, even within the dominant agricultural
sector. The third and current phase, which began in the 19503, has so
far been characterized by the political decline of the white planter

class and a concomitant increase in the number of somewhat more

.moomoumm mo nowumooa owemmumoow one ~.~ ounwfim

 

          
   

I. I'llo’l‘...
138,320»;

«$4.. .4550 .. . . . ., . «3233

I I IIIIIII
- .. II,
c w '0 I\ -
...—.....- .Ool o .
.I".l .a . I ...tl .aIlO.’
. l.- .O. . o
I at. e. . o. n o
.o O u u
l ... ..N . .. . . a . .t .
I ’ I0. I .I,.
0 . I. a .... .
| a. I . a o I . I
o .H I. c . I I. u

"‘2

8922-». 523:: .5

2 . <mm 258.50 A)“. .
goon. . .. 2.542353 ..H
assessoaseha» washout: ., .. . .
39:3» . ”washed? .rc . mm
. $2.41 $93%.$ use.

2:3...3 . k m]. 33.4.. m - .
1...... ......afl... .fia.u£.,n.£.£._. ...... ...-
3M“? 83.323. r34 .....sflu \ Ma

- . 4.3.3.

 

.. . l .. 1... 009m... “.0 "also
NWEEWWMPN z<moo 0:232 ,. 3.3.5.4.”... _

. . (‘30!

 

.
M

 

 

democratic black rulers. In pursuit of a more self-sustaining economy,
attempts have been made to promote optimum use of the country's limited
resources and to diversify its production structure. Measures taken
include human resource development, growth and development of a viable
fishing industry, development of vibrant tourism and manufacturing
sectors, diversification within the agricultural sector and, more re-
cently, on and off-shore oil and natural gas exploration.

Agricultural diversification has involved the development of live-
stock production, the establishment of non-plantation cash crops (mainly
food crops and Sea-Island cotton) and modernization of the dominant
sugar industry. Such modernization was initially restricted to mechani-
zation of the land preparation and cultural practices involved in the
sugar cane production process. Due to steadily declining harvest labour
availability and escalating harvest labour costs, however, recent
‘modernization efforts have concentrated on mechanization of the harvest-
ing and handling activities and it is on this very area that this study

is focussed.

1.2. Objective of the Study

 

This study is concerned with mechanical harvesting and handling of
sugar cane in general and, in particular, with the chopper combine
harvesting systems currently in use in Barbados. In 1979 there were 3
chopper harvesters in operation in Barbados. At that time, McGregor
et a1. (1979), after carrying out a detailed technical and socio-
economic evaluation of the industry, re-iterated the need for mechani-

zation but suggested a controlled increase in the number of chopper

harvesters from 3 to 25 over the 10 year period, 1980—1989.

To date, however, the acquisition of chopper harvesters has pro—
ceeded at a rate nearly 3 times that recommended. In the 3 year period,
1980-1982, 18 additional chopper harvesters and ancillary equipment have
been brought into the island. At an estimated cost of BDS $500,000 per
harvester package, this represents an alarmingly large capital invest-
ment (of the order of $9 million) in the harvesting operation over the
very short period of 3 years. Of particular significance also is the
fact that, instead of the annual output per machine increasing from
6,500 tonnes to 10,000 tonnes as suggested by McGregor et al. (1979),
it has fallen, apparently below the 6,000 tonnes level. This year har-
vesting machines have been operated at less than 30 percent of their
rated field capacities, idle time in the field has been of the order
of 30 percent of available working time, and factory retention time and
total turn-around-time of cane transport vehicles have both been rather
excessive (Harvey, 1982).

This study seeks to address the problem of gross under-utilization
of this already large and rapidly increasing fleet of sugar cane combine
harvesters and associated cane transport equipment. There seems to be
little point in continuing to inject large capital sums into the cane
harvesting and handling process until optimum or near optimum perform-
ance and efficiency levels are achieved with the mechanical equipment

already in existence.

1.3. Approach to the Study

 

The approach chosen to study the mechanical harvesting and handling
of sugar cane in Barbados (or more specifically, combine chOpper harvest-
ing) is to use a combination of computer simulation and Operations
Research techniques. This involves firstly, development of simulation
models to accurately represent existing systems and secondly, experiment-
ation with these models so as to simulate and evaluate alternative system
operating procedures.

The overall inefficiency of mechanical harvesting systems probably
has as much to do with the inefficiency of the cane transport and factory
yard operations as it has to do with under-utilization of combine har-
vesters in the field. Essentially, therefore, two computer simulation
models will be developed; a field operations model (FIELDOP) which will
simulate in-field operations as well as cane transport operations; and
a factory yard operations model (FACYARD) which will simulate the activ-
ities involved in the handling of the cane as it passes through the
factory yard. Activities included in the first model would be the load—
ing of cane transport vehicles and their travel to and from the factory,
while those included in the factory yard model would be queuing of
transport vehicles at the factory, weighing, unloading, taring (if done)

and vehicle departure from the factory.

CHAPTER 2

OVERVIEW OF THE BARBADOS SUGAR INDUSTRY

2.1. Introduction

 

Production of cane sugar has been a major activity in Barbados ever
since the days of colonization, when the British established the industry
primarily as a cheap source of sugar for the British empire. The country
was subdivided into a large number of production units called plantations,
each of which was managed by a British owner or his appointee and staffed
with cheap local labour. From its inception, therefore, the industry was
export oriented and for a long time, it was foreign owned and dominated,
with benefits from the sale of sugar accruing to Britain rather than to
Barbados.

Over the years, the ownership structure and technological and
economic character of the industry have changed significantly. More
than 99 percent of the sugar lands and all of the sugar factories are
now owned by Barbadians. Some of the plantations (now referred to as
'estates') have been amalgamated in order to take advantage of economies
of scale with regard to improved technology. Cane production activities
are almost fully mechanized and, in light of a declining harvest labour
force, efforts are now being made to mechanize the harvesting operation.
The final product of the industry is currently sold through a sophisti-
cated system of largely pre-negotiated markets and the returns from
these sales has made the sugar industry a major foreign exchange earner

for Barbados.

 

 

2.2. The Role of the Sugar Industry in the Barbadian Economy

Until about 1960, sugar was undeniably the mainstay of the
Barbadian economy. Since then, however, the industry's share of gross
domestic product and of exports has dwindled, due both to the decline
of sugar production and the growth and development of other sectors,
principally tourism and manufacturing. In the mid 19503, sugar's con-
tribution to GDP was 33 percent (Barbados Economic Report, 1960). This
fell to approximately 20 percent in the early 19603 and in the last
three years has been around 6 percent (Barbados Economic Report, 1961-
1981). As far as exports are concerned, the share of sugar and sugar-
based products (rum and molasses) has declined less sharply from around
95 percent in the late 19503 to about 40 percent in recent years
(Barbados Economic Report, 1950-1981). Despite this apparent decline
in national importance of the sugar industry, however, the industry is
still considered to be playing a very crucial role in the overall
economic activity of Barbados.

It has been estimated that in 1977 some 7,500 persons were gainfully
employed, on a regular or seasonal basis, by the sugar industry, repre-
senting about 7.5 percent of the total national work force (Dept. of
Labour, 1979). Beyond this there are considerable backward and forward
linkages, as well as multiplier effects, arising from the expenditure
of incomes earned in the sugar industry itself and from related activi-
ties (McGregor et a1., 1979). Persaud (1973) estimated that, in 1968,
20 percent of the materials and 60 percent of the services purchased by

the sugar sector as a whole, accrued to local incomes.

 

 

 

Certain firms engaged in heavy engineering and industrial activity
(for instance, the Barbados Foundry and Central Foundry), owe their
existence and development largely to the equipment and machinery needs
of the sugar industry, particularly of the processing activity. In
addition, the sugar industry generates a substantial local demand for
tractors and ancillary agricultural equipment, transportation and auto-
motive equipment, materials handling equipment, and repair and mainten-
ance services associated with such equipment. It is estimated that
such backward linkages accounted for approximately BDS $6 million of
local income in 1977 (Barbados Economic Report, 1978).

Direct forward linkages emanating from the sugar industry include
distilling and, to a lesser extent, food and beverage processing and
manufacture of livestock feeds. Out of a total production of 23 million
litres of molasses in 1976, 16 million litres were used to manufacture
rum and a small quantity went into animal feed manufacture. That year,
the two major Barbadian distillery companies produced 22 million proof
wine litres of rum of which 8 million litres, worth BDS $4.1 million,
Were exported (Barbados Economic Report, 1976).

Apart from direct backward and forward linkages, income is generated
by the multiplier effect arising from the Spending of income earned in
the industry. Persaud (1973) has reported that income generation has
tended to be relatively larger in the sugar industry than in other
sectors of the economy with the exception of the distribution sector.
Based On figures for the year 1968 (the most recent available), he
est“LInEited sugar's share of the direct value added in the final sector

Clem
and to be 67 percent in sugar cane production and 8 percent in cane

W, F——- %‘-r V M

 

 

 

 

processing, or about 52 percent for the sector as a whole. Comparative

figures for other sectors are 55 percent in distribution, 42 percent in
non-sugar manufacture, 35 percent in non-sugar agriculture, 33 percent
in construction and 19 percent in the tourist industry (Barbados
Economic Report, 1969).

As far as foreign exchange earnings are concerned, the sugar
industry stands second only to tourism (Barbados Economic Report, 1980).
In terms of stability, however, it seems to be the general opinion of
most: Barbadians that the sugar industry offers considerably more long-
term financial security and reliability than does the tourist sector.
While the former is well known to be vulnerable to volume and price
fluctuations, it seems somewhat far-fetched to conceive of a situation
in which the entire industry could become severely shrunken over the
short term. 0n the other hand, the tourist industry is easily inter-
ruPt ed in the short term, the inflow of tourists being extremely sensi-
tive to natural disasters and political disturbances not only in
Barbados, but in any of the neighbouring Caribbean countries, or merely
to the fear of such events occurring, and to economic trends in the
developed countries of the world.

Finally in assessing the role of the sugar industry in the Barbadian
economy, there are a number of socio-economic benefits that cannot be
ignored. For example, it may be argued that, whereas in the 300 years
of British colonial domination, the sugar industry, being geared towards
the needs of the British Empire, may have been a net beneficiary of
social and economic infrastructures. Today, it probably is a net contri-

bu .
tor’ mainly as a result of the imposition of various Government lev1es.

 

 

 

 

 

10

2.3. Structure of the Industry

The present structure of the Barbados Sugar Industry has evolved
from the plantation system of agriculture to which the country was
subjected during nearly 300 years of British colonization. In those
days, the main commercial unit of production was the plantation which

was characteristically owned and managed by a British master and staffed
by black field workers. Today the basic commercial land unit is essen-
tially unchanged but former plantations are now called 'estates' and
estate workers are represented by vibrant workers' trade unions.

Sugar cane processing was carried out by numerous small individual
sugar mills with little or no central control or direction. Marketing
too was a simple process, marketing activity being confined to trans-
porting the final product to British refineries. In recent years,
however, the structure of the industry has changed considerably and
today well defined and quite sophisticated production, processing and

marketing sectors are discernible.

2- 3 . 1 . The Production Sector

In Barbados there are two modes of sugar cane production: estate
production and small—holder production. An estate is defined as any
holding of 5 or more arable hectares, all other holdings being classi-
fied as small-holdings. Currently there are 134 estates cultivating
approximately 15,500 arable hectares (115 hectares per average estate
unit) and some 10,000 small-holdings cultivating approximately 3,200

h
ectares of cane (McGregor et a1., 1979)-

 

11

Unlike the case in several other Caribbean cane—growing territories,
there is little expatriate ownership in the Barbados sugar industry and,
in fact, only four estates (the equivalent of 620 arable hectares) are
(owned by non-Barbadians (McGregor et a1., 1979). Despite this, however,
zabsentee ownership is still a prominent feature of the industry and only
:25 estates (2,140 arable hectares) can be regarded as substantially
cywnerdmanaged. The attorney system, a legacy of the days of foreign
aibsentee ownership, has been retained as the core of the management
sitructure. An attorney represents the interests of the owners of an
(estate and presides over the financial and other policy matters of the

Gestate. Subordinate to him there is a manager responsible for the day—
tcr-day running of the estate. In Barbados, a single attorney is usually
tresponsible for several estates, rather than just one and, in fact, four
Etttorneys now control 34 estates, comprising some 6,500 arable hectares.
EStates and small-holdings are located throughout the country as shown
in Figure 2.1.

Small holders are essentially part-time cane farmers, yet, in
*aggregate, they account for a very significant 15 percent of total cane
Production in Barbados (Figure 2.2). They constitute the bulk of the
full-time estate labour force during peak labour demand periods, and
most of them find additional employment outside agriculture. In 1971,
the National Agricultural Census indicated that as much as 70 percent
of fill-time estate workers cultivated theirown cane that year. However,
a more recent survey conducted by McGregor et a1. (1978) has put this

eStimate at 50 percent.

-12-

11111]...

 

: r-
.I

 

SMALL. 0’! rm fl,

Elem: (see hey to: name)

[wwwos AGHDJLTURAL DEVELOMNI cow onmum

W‘CE: Compaed by Lard V‘dU-‘um W'su.,tfl<t'lnlwl l'i/h

 

 

 

 

Figure 2.1. The location of estate and small holder land.

-13-

2...

cc

.mcmu mo muasoam msownm> mafiozoouo mumoaos Hamam mo mmwmucmoumm .~.N muswwm

.3263 3:29am 2me «3.33m on. .0 3.303838

0cm... .0 mecca»
o. c

on o~
. - 1 -

 

X

 

 

wound

 

 

 

14

2.3.2. The Processinngector

The manufacture of cane sugar in Barbados began in the 16403
following the introduction of a model of a sugar mill and some skilled
artisans from Brazil (Barbados Sugar Review, No. 47, 1981). The early
factories were extremely inefficient, the output per factory in the
ciecade 1700-1709 averaging just 15 tonnes sugar per year with a tonnes
(:ane/tonnes sugar ratio of 18 to 20.

In 1709, there were 485 sugar mills, 409 of which were driven by
twind and 76 by animal power. Thereafter the processing of cane remained
t:echnologically stagnant until the 18408 when the first steam plant was
iaitroduced (Barbados Sugar Review, No. 47, 1981). This switch to steam-
powered plants generated a steady decline in the number of factories
and by 1958 only 26 mills remained.

Sugar factories in Barbados are still relatively small and much of
ttmdr equipment is outdated. Maintenance of the sugar mills has, there-
fcue, posed increasingly difficult problems over the last 20 years and,
618 a result, the number of mills kept in operation has been progressively
‘reduced from 26 in 1958 to 17 in 1969.

Prior to 1970, sugar factories were operated largely by individual
Owners or private individual companies and closures of factories were
based on individual financial decisions, rather than on the interests
of the industry as a whole. In June 1970, however, a single company,
Barbados Sugar Factories Limited (BSFL), was incorporated to collectively
own and operate all existing factories. With the birth of this company,
a rigorous rationalisation programme (aimed at planned restructuring of

the processing sector of the industry) ensued, leading to the further

 

 

La

15

closure of the smaller, marginal and uneconomic mills and the moderniza-
tion and improvement of those mills that were retained. During the
rationalisation process, the island was divided into three zones; North,
ICentral and South and two factories were allocated to serve each zone.
'10 date five of the original sugar factories are in operation; two each
in.the South and Central zones and one in the North. In addition a new
tnodern sugar factory has been constructed in the North zone and was com-
tnissioned during the latter half of the 1982 harvesting season. The
(erection of this new factory represents a major capital injection into
tihe processing sector since the last factory established was built as
IJDng ago as the year 1920. With these six factories in full operation,
tine theoretical throughput capacity of the processing sector is in the
rwegion of 480 tonnes per hour or 11,500 tonnes per 24 hour working day.

The Barbados Sugar Factories Limited holds majority shares in three
siervice companies associated with the processing activity: the Barbados

Idolasses Terminal Limited (97%), the Sugar Terminal Limited (56%) and
t1m28ugar Transport Limited (100%). The BSFL is primarily a grower-
<JWned processing company whose ordinary shares are tied to ownership of
Sugar cane lands. All but three estates hold shares in the company
(McGregor et a1., 1979).

The BSFL is the sole purchaser of all sugar cane grown on the
island. After deducting monies for industry-wide fixed expenses, the
factory company retains 26 percent of the revenue from the sale of sugar
and molasses to cover processing costs, and the rest is paid to the
growers.

In Barbados cane farmers are paid for their canes on the basis of

Weight (rather than sucrose content as is done in several other cane

 

l6

growing countries including Australia). However, at the end of the
season, each factory adjusts the final cane price on the basis of its
average sugar recovery from the cane. The final cane price, therefore,
differs between factories and it is not uncommon for this difference to
be in the region of $2 per tonne. During the 1982 harvesting season,
efforts were made to reflect the extraneous matter content of cane
delivered by a grower in the price received for that cane. Throughout
the season, extraneous matter determinations were done on random samples
(of cane delivered to each factory each day, so that appropriate penalties
could be included in the price paid for cane having an extraneous matter
content in excess of 3 percent by weight.

0n the international sugar scene, the Barbados Sugar Industry
Ireportedly stands high with respect to overall operating proficiency
11nd final product quality (McGregor et a1., 1979). In terms of energy
\ltilization, the industry's processing sector may even be unique.
lJnlike cane sugar factories in most other parts of the world, those of
IBarbados operate exclusively on bagasse (a by-product from the mills)
and some factories also manage to produce a considerable bagasse sur-
plus which has the potential of being used for the manufacture of soft

wood laminates.

2.4. Markets for Barbadian Sugar

Like most primary industries of countries emerging from extended
periods of foreign colonization and domination, the Barbados Sugar
Industry is primarily export oriented. 0f the average annual production

0f 112,000 tonnes over the five year period 1973-77, some 98,000 tonnes

17

(or 87%) were exported with only 15,000 tonnes (13%) being disposed of
on the domestic market (McGregor et a1., 1979). The main export markets

were the European Economic Community (EEC) and the World Market.

2 - 4 .1. The European Economic Market
At present, the European Economic Community is the main export
market for cane sugar produced by the African, Caribbean and Pacific
(ACP) group of countries, of which Barbados is a member (McGregor et
a1 - , 1979). Prior to 1975, there was a Commonwealth Sugar Agreement
(CSA) which fixed export quotas for Commonwealth sugar producers, thus
providing a guaranteed and sheltered market for most of the sugar pro-
duced. In February of 1975, the CSA was superceded by a Sugar Protocol
agreement signed at the Lome’ Convention of that year. Under the terms
Of the Protocol agreement, the EEC is committed to purchase from
Barbados 49,300 tonnes, white value (about 53,600 tonnes raw value) of
sugar annually (Barbados Sugar Review, No. 41, 1978). Since the Protocol
is Supposed to be valid indefinitely (unlike the Convention itself
which is subject to re-negotiation every five years) then, in theory,
it represents an assured outlet for Barbadian sugar.
The EEC is reportedly self sufficient in beet sugar and is, on
bal ance, a net exporter to the World market (McGregor et a1., 1979).
In ef fect therefore the cane sugar imported by the Community from
, ,

the ,
ACP countries pursuant to the Sugar Protocol of the Lome Convention,

is
‘2 e---exported. Despite this, the Protocol agreement has so far been
to

t:el:Lly honoured (probably for political reasons) and appears to be

re
la“hively reliable. It is worth noting, however, that a large British

‘

 

 

 

18

refinery concerned mainly with the refining of ACP produced sugar, has
recently ceased operation and initial implications are that some ACP
countries may have to seek alternative markets for a sizeable quantity

of their raw sugar.

African, Caribbean and Pacific countries exporting sugar under the
Ilcrmne’ Sugar Protocol receive a guaranteed price which is negotiated
EiIIIIUally within the price range obtained in the EEC while taking
c:c>risideration of all relevant economic factors (Barbados Sugar Review,

No - 41, 1979). Given normal market conditions, this price tends to be
substantially higher than free World market prices. However, since
Lome' prices are indexed to EEC price ranges, annual fluctuations in the
price received by ACP producers are more strongly influenced by the EEC
Supply and demand schedules than by production costs in the ACP coun-

tries themselves. The net result of this is that the guaranteed Lome’

Price has not risen as rapidly as production costs have in the Barbados

Inch-ISI:ry. The indication is, therefore, that it is essential for the

BarbEldos Sugar Industry to keep its production cost trend as closely
paralleled as possible to the movement of production costs in the EEC,
in oI‘Cier to maintain present profit margins. This implies increased
productivity of existing sugar lands and decreased production costs and,
Since harveSt labour costs are a major component of production costs,
inc )3 eased harvest mechanization.
The actual economic return to Barbados from sugar sales to the

Eu
1:9 Dean Common Market reflect not only the annually negotiated Protocol

Pr
iqe but also any premia obtained from individual purchasers within

the

Community (McGregor et a1., 1979). The largest purchasers of

 

 

19

Barbadian sugar within the EEC have been the United Kingdom and Ireland.

Sugar supplied to refineries in these two countries commands a premium

over the Lome’ negotiated price.

2.4.2. The World Market

In global terms, the free (World) market represents a residual out-
let for the relatively small proportion of the world's sugar which is
not consumed in the producing countries or traded under pre-negotiated
Inarket arrangements such as the Sugar Protocol of the Lome’ Convention.
Ihie to the residual nature of the world market, cyclical imbalances bet—
Ween supply and demand (mainly as a result of weather stimulated supply

\rairiations) often generate wide fluctations in the price of 'free'

Sugar.

Sale of sugar on the world market is governed largely by the

llriternational Sugar Agreement (ISA). Under the terms of reference of

tillis agreement, Barbados is regarded as a small exporter and is granted

E111 annual export entitlement of 70,000 tonnes of raw sugar. This en-

t:i‘tlement is not subject to quota adjustments, neither is Barbados
‘Piéquired to observe the ISA's stock provisions with which larger export-

ing countries must comply (Barbados Sugar Review, No. 41, 1979).

In recent years, the main world market outlets for Barbadian sugar

have been the Canadian and U.S. markets.
Not all of the sugar exported to North America is in the form of

'tlle crystalline final raw product. A significant portion of the exports

tx) the U.S. and the bulk of the exports to Canada are in the form of

fancy molasses, (a specialized intermediate product of the processing

20

activity) which normally commands a premium price over raw sugar
(Ministry of Finance, 1975-81).

The 1977 International Sugar Agreement established a target price
range of 24 - 47 U.S. cents per kilogram of raw sugar. The average
ISA free market spot price for raw sugar for the period 1977-1979 was
1 7 .90 U.S. cents per kilogram, a price which was below the average
f inancial costs of production of most producers in the world (McGregor
et a1., 1979). Owing to the large excess supply of world sugar over
demand, however, even the lower end of this target price range was not

reached until 1981 when the price reached 31.25 U.S. cents per kilogram.

2 - 4 . 3. The Domestic Market

The domestic consumption of sugar in Barbados averages 15,000
tonnes annually (McGregor et al., 1979) and, with a population of about
250,000, this works out to be an annual per capita consumption of 60
kilograms. This figure is high by world standards (OECD/OCDE, 1975,

19 79) and, with the demand for sugar in Barbados practically income in-
elastic, one cannot envisage any growth in domestic demand for sugar,
except a gradual response to total population increase.

Before 1976, factories received a fixed price of BDS $376.17 per
tonne 96° Pol, or 19 U.S. cents per kilogram, for all grades of sugar.
Since then, however, ex-factory prices for domestic sales have risen
BUbStantially. In 1977, price differentials between all grades of
Sugar (Browns, Yellows and Straws) were established to reflect product
(1qu ity and the general level of ex-factory prices even rose above the

Dr
iQeS obtained for foreign sales other than to the European Economic

(3Q
tun‘llnity.

¥

21

Currently confirmed potential sales of the Barbados Sugar Industry
total some 139,000 tonnes, raw value; (54,000 tonnes to the EEC; 70,000
tonnes through the ISA and 15,000 tonnes for domestic consumption).
McGregor et a1. (1979) have estimated that, if the world sugar prices

were to exceed 33 U.S. cents per kilogram, Barbadian exports to the

free market could well surpass the 70,000 tonnes allowed under the ISA

agrenent. Unfortunately, however, the Industry has not been able to

fully exploit the available sales opportunities over the last few years.

In 1979, 1981 and 1982 sugar production was only 112,000, 96,000 and

96 ,000 tonnes representing shortfalls in potential sales of 20 percent,

31 percent and 31 percent, respectively.

2.5. History of Mechanical Harvestig of Sugar Cane in Barbados

Mechanical harvesting of sugar cane was first introduced to Barbados

in 1968 when two machines, a Toft J-150 and a Crichton, were imported

frOm Australia. Both of these machines were wholestick harvesters

which cut the cane at the base and deposited it in piles for subsequent
pickrup either manually or by a mechanical loader. These machines were
operated in both burnt and green cane but, even in burnt cane (for
Which they were primarily designed) their performance was unsatisfactory.
Both machines were set on 1.5 m wheelbases which were not only narrower
that). the inter-row spacing of the cane but were inadequate for the
rolling terrain and stony conditions that existed in most of the fields
at the time. The machines were reportedly very little used (McGregor
at 1979) and, since no significant efforts were made to adopt

a1-

’

f i
Qld conditions and inter—row Spacing to the design constraints of

 

¥

 

‘.
turd

5‘!
H:

r.
(I)

u
'A
.

’_'-?

c)

22

the harvesters, it is doubtful whether they were ever really subjected

to a serious field evaluation.

In 1970, a Cameco wholestick harvester and a push-pile loader were

imported from the U.S.A. The harvester was large and proved to be too

cumbersome for operation in the small, irregularly shaped Barbadian

cane fields. The push-pile loader, however, generated considerable

interest among the larger cane growers.

In 1971, a decision was taken to switch from wholestick to chopper-

type cane harvesters. That same year, two chopper harvesters, a Toft
Cid-364 and a Don Mizzi 741, were imported from Australia. The follow-

ing year, a second Don Mizzi 741 was brought in and in 1973, a Don Solo

chopper harvester was imported. All of these machines were designed to

operate in burnt cane and controlled burning of cane was, therefore, a

Prerequisite for their use.
The Toft CH-364 was a rather large, self-propelled combined har-

vester, well furnished with engine power. It was a quick cutter but,

under the stony field conditions that existed at the time, numerous

problems were experienced, particularly with the machine's cane eleva-

tion system. Its wheelbase, though wider than that of the wholestick

haI“"esters, was still too narrow and even on relatively gentle lepes,
1El':et‘al stability was a problem.

The Don Mizzi harvester was essentially a cane base-cutting,
e1 e\’a.ting and cleaning mechanism attached to a reversed L—shaped frame,
on to which a standard 60 kw agricultural tractor was mounted. The
drive train of the tractor was linked to the main drive shaft of the

ha
rVesting mechanism by means of a duplex chain and sprocket coupling.

¥

Q.

.\

(Q

'h

23

Tractive power, as well as mechanical and hydraulic power for the moving
parts of the harvesting mechanism, were derived from the power unit of
the tractor. The resulting combination was a relatively compact, self-
propelled unit set on an extended wheelbase 3.7 m in width. Like the
Toft CH-364, the Don Mizzi 741 machine delivered cane to the left side
only, making two-directional cutting impractical in the smaller cane
fields.

The Don Solo chopper harvester imported in 1973 was a more compact,
integrally assembled, self-propelled version of the Don Mizzi with the
capability of delivering cane to either the right or left of its direc—
tion of travel. On account of this, it was theoretically more suited
for work in the smaller cane fields. However, because this machine
was set on a narrow wheelbase of just 1.5 m, and because most of the
smaller fields are located on sloping and undulating terrain, machine

stability during operation became a major problem.

Among the machines mentioned above, the Don Mizzi 741 was undoubt—
edly the most satisfactory, largely because of the extended width of
its wheelbase and because concerted efforts were made to prepare the

fields for mechanical harvesting according to the guidelines suggested
by Baxter, 1969. By the 1973 harvesting season, many of the teething
problems associated with this machine had been overcome, and a well
organized and technically efficient mechanical harvesting system had
been built around the harvester. Just then, however, two independent
St‘lciies conducted by Chase and Eavis (1972) and Hudson (1972) showed
that burning of canes at harvest time reduced the organic matter content

O
f the soil to alarmingly low levels and caused a significant reduction

 

k

 

24

in subsequent crop yields. In addition, reports from the Entomological

Division of the Ministry of Agriculture indicated that the populations
of predators, which were heavily relied upon for biological control of

economically important pests of sugar cane, were likely to be signifi-

cantly depleted if burning continued. Based on these findings, burning

of sugar cane prior to harvesting was outlawed.

Following this, efforts were made to adapt the Don Mizzi system to
green cane harvesting but the harvester's unsuitability to this task
prevailed and, in 1974, the Don Mizzi system was finally abandoned.
Nonetheless, the experience with the Don Mizzi program proved to be

very valuable. Not only did it demonstrate the applicability of chopper

lizalrvesting to Barbados, but it also provided a foundation of knowledge

firwom which more recent chopper harvester operators were able to benefit.
In 1975, two Toft 300 chopper harvesters, reported to be capable

cane

Of handling green cane, were brought into Barbados by individual
Producers. Based on the performance of these, two more chopper har—

vesters, one Toft 4000 and one Massey-Ferguson 205, were imported by
agricultural machinery distributors in 1979 for contract harvesting
operations. Between 1979 and 1982, 17 additional green cane chopper
harvesters, 1 Massey-Ferguson 305 and 16 Toft 6000 machines, have been
bI.Q"-Jl-ght into the country. In terms of field operating procedures, the
Mag Sey-Ferguson and Toft cane harvesters are quite similar but, in
terms of design, the Massey-Ferguson machines rely heavily on mechanical
drives and linkages while the Toft machines are almost exclusively
hydraulically driven.

The rapid increase in the number of chopper harvesters in Barbados

Q .
Vet the last three years has occurred without a corresponding rapid

¥

V a
u\u

25

development of machine management capabilities and without adequate re-
organization of cane receiving facilities at the factory yards. This
has resulted in rather inefficient use of most of the individual mechan-
ical harvesting units.

In the late 19603, the Barbados Sugar Producers Association, in-
spired by the need for a harvesting system specifically suited to
Barbados conditions, initiated development of a wholestick machine

system in collaboration with a British firm, McConnel Engineering. The

first prototype cutter began working in 1972. This machine, which sub-

sequently became known as the BSPA/McConnel Stage I, cut the cane at
the base and left it in a swath along the row for manual cleaning and
loading. Seven additional Stage I machines were introduced in 1973 and
by 1976 some 37 of these units were in private ownership. Owing to the
large number of workers required to work behind these machines as re-
trievers, machine productivity was very low and probably for this reason,
most of the machines were little used.
In an effort to eliminate the heavy labour requirement behind the
Stage I, a BSPA/McConnel Stage II machine was developed, the first one
being used in 1975. This was essentially a cleaner-piler which picked
up the swath formed by the Stage I cutter, removed most of the cane
tops and trash, and deposited the cane in piles for final pick-up by a
mechanical grab loader. By 1979, seven of these machines were privately

own
ed , mainly by the larger estates. They, too, were used very little

a
“(1 most of the Stage II owners and potential owners have now invested

1
t1 QEine combine harvesters.
The Stage I machine has been retained and further refined by a

1
Q in1 firm called CARIB Enterprises. The current model, now known as

k

.3”

av

r.

Si

26

the CARIB cane cutter, is a vast improvement on the original Stage I

prototype. The most major modification was mounting the cutting mechan-

ism on a reversed, 4-wheel drive tractor so that considerably more
*weight is now on the front end of the machine, and traction and stability

.are greatly improved. Based on observations during the 1982 harvesting

season, the CARIB machine seems to have a lot of potential for work in

‘the more hilly central and eastern areas of the country that are not

zaccessible to the larger chopper harvesters.

CHAPTER 3

LITERATURE REVIEW

3.1. Introduction

Tomlinson (1973), in an assessment of the problems of the Caribbean
ssugar industry, highlighted the need for greater efficiency in organiz-
ing and planning the use of harvesting equipment and proposed that the

techniques of Operations Research could be found useful in this respect.

Operations Research, as described by the Operations Research

Society of America, 'is concerned with scientifically deciding how to

best design and operate man-machine systems, usually under conditions

requiring the allocation of scarce resources' (Phillips et a1., 1976).
Broadly speaking, the application of Operations Research techniques

requires the construction of a model which incorporates all the impor-

tElnt characteristics of the system (Shamblin and Stevens, 1974). Boyce

(1972a) classifies three types of models; iconic, analogue and symbolic.

IcOnic models physically resemble the subject of inquiry and are char-

ac terized by some scaling effect. Analogue models are characterized by

the use of a convenient transformation of one set of properties for

ant) ther. Symbolic models are characterized by the representation of

a. system by mathematical or logical symbols and expressions.

27

28

3.2. Symbolic and Mathematical Models

Most harvesting systems can be represented by the symbolic model,

as illustrated in Figure 3.1, and can be generally grouped as closed

circuit transport syst-s (Boyce, 1972b). In general, a closed circuit

tranSport system can be broken down into two separate sub-systems, one

located at the field and the other at the installation (factory, mill

(Dr other facility). Each sub-system can be represented by a queuing

model, in which transport units queue at the field and factory awaiting

service. The length of time each arriving unit must wait in line

(queue) for service depends on the number of other units already in the

<1ueue, the service rate (or service time) and the rate of arrival at

the queue (Saaty, 1957). In a real life system, arrival and service

rates are not constant but subject to variation about a particular mean.
The successful use of mathematical models to represent real queuing
Systems, therefore, depends primarily on the identification of the form

0f the probability distributions which best represent the actual arrival

and service rate distributions (Page, 1972; Phillips, 1976; Hillier and

IJileberman, 1974) .
In a general discussion on the use of mathematical models to de-

Sc1':ibe phenomena in agricultural systems, Demichele (1976) proposed
tI'lli‘ee general forms which could be used, depending on the state of know-

ledge (Figure 3.2). Based on the system being studied, these forms may

be represented by different statistical distributions. Hahn and Shapiro

( 1967) suggested that, in the engineering context, there are often in—

sufficient grounds for choosing a Specific model. They indicated that

I“Zirgmre 3.3 (adapted from E.S. Pearson, University College, Brandon) may

 

 

29

QUEUING SERVICES

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Loading
-‘ ' f"""""1l
___% |.- - ..
,______T ------- 1‘ - l
___..| -- --..-.4 l
——-—— |r--'""--1' u n
‘_—_d‘. -
‘r v
ISON QUEUING SERVICES NON QUEUING SERVICES
Travel loaded ' Travel empty .
I
.-t. J

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hi

 

COMMON QUEUING SERVICE
Unloading

Figure 3.1 Symbolic model of closed circuit queuing system

3O

 

 

 

Probability of an
outcome 3

 

all possible x

A

1) The distribution function of a system where nothing is known

 

 

Probability of an
outcome x

 

all possible x

(L2) The distribution function of a system where some things are now known

Probability of an
outcome x

 

 

 

 

all possible x
(3 ) The distribution function of a system where most things are known

JPTlgure 3.2 Generalised models representing the behaviour of systems

§3<Durce: Demichele,(1976).

 

 

B 2
c ousmaunou ——-;—>

F igure 3 .3

 

45

0|

OI

q

31

  
   
  
  
 
 
 
 
 
   
   
 

 
    
  

   
      

‘1 h . .. 1 aw ".9 ‘ u
wrn‘tm| agi‘ fit?“ ‘3 5".“ .&# is“
4* W“

   

.nnM" 1.2.3,.
.. “W IMPOSSIBLE muss” “53""

'~.a- n.-
l
“W‘*W‘fi1"'blwi‘ , mwmmaf‘ .‘u t- ‘r.. 2

iv“: ...": ”1.12""“J‘HIE. ‘fi‘tflflwI I 21; a?” {NET-L: L¥*
... " “I I ’ r 1‘» mo ‘1‘
a» “Mes! ...... 3% \s.»
...... ..w v.6“ ° 11"7'3'u‘ asses;
,th ..-
....31ugI\.I‘I'I\1h “1‘31" .I‘ J!" I".|.-.H|I"I"1"'I‘:h"“‘l “.IJI'
"um. I W‘! 'IiI'!‘ “M“; 3'” :4‘;\ H " “L ‘J HHIJII {-T‘ ‘1'1'11‘
Ikéxw“\$_'¢ "‘ ‘1‘." 2;: .‘Is‘fl ‘4‘, 41M "3“; éu' WI".
"-I “WWW“? 3" «‘9‘ “51"”; -.'- fins-‘33,;

09’!“
. 3’13“ 1 1- Pt}? I.
' . Til? Lfef- -10‘ ‘&1|\\%.' “LN...“

‘M “Jr“? M?” Var-1‘15

A“? 1
r'.‘ ‘%Id '18

  
  

 

1x ‘ 41.1%.,“ Ilfl‘iw

 

       
  

     

    
 

      
 
     

      
 

       

 
  

1"

 

 

 

   
 

      

... w““‘hh1‘|§fi"c'$“h.

   

 

 

 

 

 

 

 

 

 

EXPONENTIAL
DISTRIBUTION

 

 

 

 

 

 

I 2 3 4
Bl

Regions in the (81,82) plane for various distributions

32

be used for such a selection. The values of B1 and 82, the square of
the standardized measure of skewedness and the standardized measure of
peakedness, respectively, can be estimated using equations (1) and (2)
below. The estimates so obtained are, however, very sensitive to a few
extreme observations and should, therefore, be used with caution,
particularly in situations where the number of observations is less
than 200. On account of these factors, it is generally desirable to
set up a frequency table to enable the fitted distribution to be com-

pared with the observed data.

3/2l2
Bl=b1=IM3/M2) (1)
2
32 = b2 = Mg/(Mz) . . . . . . . . . . . . . . . ... . . (2)
mahere
N _ 2
M2 = l/N 2 (X1 - x) . . . . . . . . . . . . . . . . . (3)
=1
N _ 3
M3 = l/N 2 (X1 - X) . . . . . . . . . . . o . . . . . (4)
1:
N _ u
M. = l/N 2 (X1 - x) . . . . . . . . . . . . . . . . . (5)
i=1

xi = 1th value of the variable x

x = sample mean

Hahn and Shapiro (1967) further pointed out that the initial selec-

tlzirbn of a model should be based on an understanding of the underlying

 

33

physical phenomena and that a distributional test should be used to
evaluate the adequacy of the physical interpretation.

Boyce (1976) indicated that the process of selecting suitable
statistical distributions and estimating the appropriate parameters
could, in some cases, involve a virtually impossible sampling task if
the real system were operating in an unsteady and/or erratic way. He
further suggested that the use of an optimistic, a pessimistic and a
Imost likely estimate of the particular quantity of interest, based on
.field samples, might provide a basis for estimating the parameters of
as selected distribution.

Dumont and Boyce (1972) described three methods of fitting three
ssuitable distributions to observed work time data for two agricultural

tmnit operations; tranSport loading and combine harvesting: These are as
onllows:
(1) the method of maximum likelihood for the Gamma
distribution,

(2) the method of matching moments for the Beta
distribution, and

(3) the method of matching percentiles for the Johnson
SB distribution.

The Chi-squared test statistic was used to test the goodness of fit
13‘:>2:r each set of parameters determined. Their general conclusion was
that, in practice, any one of the three distributions would be acceptable
‘3‘:’?t? use in simulation modelling to represent the distribution of transport
:I‘CDHEiding times, but that the Gamma distribution seemed incapable of pro-
xifii~cling an acceptable fit for the combine harvesting times data.

Bouland (1967) conducted a study of truck queues which formed at

(:CDIIntry grain elevators in the U.S.A. and found that transport unit

34

arrival rates could be described by Poisson distributions for rates of
less than 35 arrivals per hour, and by Uniform distributions for rates
exceeding 35 arrivals per hour. Service times of weigh scales and
unloaders were found to fit Erlang distributions. The probability

density functions of these distributions are as follows:

The Poisson distribution

 

f(x) e‘M M3 x = 0,1,2. . . . . . . . . . . . . . . . . . (6)

where

value of discrete variable

x
M 8 expected value = variance

Source: Bhattacharya and Johnson, 1977.

the Discrete Uniform distribution

f(x) l/N x = l,2,.....,N . . . . . . . . . . . . . . . . (7)

where

value of discrete variable

K

Source: Bhattacharya and Johnson, 1977.

L393§2:e Erlangidistribution

(Anx(n-l)e-Ax)/(n-l)|,x > 0 . . . . . . . . . . . (8)

f(x:n,A) =
“°qb1£ere
x = value of continuous variable
A = scale factor
n = shape factor (restricted to integers)

Source: Manetsch and Park, 1980.

35

Dumont and Boyce (1972) found that the Gamma distribution, the
more general form of the Erlang distribution, gave an adequate repre-
sentation of transport loading times for various field conditions. The

Gamma distribution has the following probability density function.

f(x:n,A,u) = An(x - u)n-1 e-A(x-U). . . . . . . . . . . . . . (9)
where
n > O; A > O; u < x < co; -m < u < w and G(n) is the Gamma

function given by

G(n)= 2°(x-u)”'1e'(x'“)dx................(10)
where
x = value of continuous variable
A = scale factor
n = shape factor
n = location parameter

Brooks and Shaffer (1971) reportedly developed a mathematical model
11C> predict the output for tipper trucks hauling dirt from an excavation
site to a landfill area (Ogilvie et a1., 1978). Their approach consisted

of six basic steps:

(1) determination of all possible combinations of shovel and
truck, each combination being called a state;

(2) measurement of the ability of the shovel to change each
state defining the rate of transition between states due
to the shovel by the mean service rate;

(3) measurement of the ability of the truck to change each

state defining the rate of transition between states

36

due to the trucks by the arrival rate of the trucks at
the shovel;

(4) determination of the net effect of changes by shovel and
by trucks;

(5) determination for steady—state conditions, of the per-
cent of time on average that an Operation was in any
particular state;

(6) calculation of the steady state output of the system
by multiplying the service rate by the Production
Index (PI), where PI = sum of the percentage of time

that the system was in each productive state.

The main objective of this model was to improve the opportunity for
the contractor to apply queuing theory in practice by providing a pro-
(zedure which did not require knowledge of the mathematical rigour of
caueuing theory, but at the same time provided a method of predicting

C>l1tput, which could be rapidly and easily applied. Based on their model,
Brooks and Shaffer (1971) reportedly developed a set of curves for var-
iOUS combinations of the mean service and arrival rates and the number

of trucks for simple single server and multiple server systems, as well
as systems in which the shovel had a hopper.

Audsley and Boyce (1973) used a mathematical approach of queuing
t".‘Il‘iory to develop models of cyclic transportation systems. In their
‘iICDIEIC, they assumed that service times were independent and identically

distributed and could, therefore, be represented by some form of Erlang
distribution. They obtained similar results to the production index

‘flilich Brooks and Shaffer obtained in 1971 (Ogilvie et a1., 1978).

37

3.3. Computer Simulation Models

 

An approach to the study of large and/or complex systems which is
rapidly gaining popularity is computer simulation. Phillips et al.
(1975) stated that simulation has become one of the most widely used
and accepted tools of systems analysis. Broadly speaking, simulation
is a problem-solving technique for defining and analysing a model of a
system (Dent and Blackie, 1979). To simulate means 'to duplicate the
essence of a system without actually attaining reality' (Rockwell, 1965).
Underlying any simulation is a mathematical abstraction of the system
which relates various system functions. In a paper on the use of simu-
lation methodology in agriculture, Rockwell (1965) suggested several
reasons for using simulation analysis. Most of these reasons have been

re-iterated in a more recent work by Naylor (1971). They are as follows:

(1) The system may be too complex for analytical solution
in which case, simulation can yield valuable insight
into which variables are more important in the system
and how these variables interact.

(2) With simulation, it is possible to build in time
delays, non-linearities, irregular distributions
and discontinuities into the system.

(3) Systems in which time is a critical factor are well
suited to computer simulation. For certain types of
stochastic problems, the sequence of events may be of
particular importance. Information about expected

values and moments may not be sufficient to describe

 

 

 

(4)

(5)

(6)

(7)

38

the process. In these cases, simulation methods may

be the only satisfactory way of providing the required
information.

Simulation can be used to experiment with new situa-
tions about which we have little or no information.

The simulator permits later decisions to be based on
earlier system output as in a dynamic programming
framework.

The process of designing a computer simulation model
forces the researcher or designer to explicitly describe
the system processes and the required data. The know-
ledge obtained in the design activity frequently
suggests changes in the system being studied. The
effects of these changes can then be tested, via simula-
tion, before implementing the changes on the actual
system.

The process of simulation design is in itself a valu-
able educational tool and it has been used by many
companies as a pedagogic device to help management
understand the characteristics of the systems which

they control.

Simulation permits us to experiment with systems that,
in reality, it would not be possible to experiment with.
Through simulation, one can study the effects of certain
informational, organizational and environmental changes
on the operation of a system by making alterations in

the model of the system and observing the effects of

(8)

(9)

(10)

(ll)

(12)

‘ (13)

(14)

 

39

these alterations on the system's behaviour.
Simulation can serve as a 'preservice test' to try out
new policies and decision rules for operating a system,
before running the risk of experimenting on the real
system.

Simulation can often lead to a training device such as
a management game. Since the simulation process gives
no valuable insights into the qualitative aspects of
human decision making, and because of its natural
realism, it often becomes possible to convert the
simulator into a simulation training device.
Simulation models usually provide better user accept-
ance than analytic models. The systems user can see
the reality of the simulation and thus will usually
have more faith in the conclusions from the simulation
output.

The interpretation of simulation results does not
usually demand a mathematical background on the part
of the user.

Mbnte Carlo simulations can be performed to verify
analytical solutions.

Simulation is well suited to sensitivity analysis in
which key system parameters are selectively altered to
test their contribution to overall system performance.
When new elements are introduced into a system, simula-

tion can be used to anticipate bottlenecks and other

 

*‘Ifl/ ’5

 

40

problems that may arise in the behavior of the

system.

Computer simulation, as defined by Pritsker (1974), is the estab-

lishment of a mathematical logical model of a system and the experimental

manipulation of the model on a computer. Development of the model re-

quires an in-depth analysis of the system in order to identify its

important characteristics and components. Such components include

entities, their attributes and events, which may be defined as follows:

Entities - objects within the boundaries of the system

being studied

Attributes - characteristics of entities within the

system

Events - occurrences which cause change in the status

of the system.

Simulation can be analogue or digital, discrete or continuous, with
or without a computer, with or without real time (fast or slow) and with
Or without a human decision maker in the simulated process (Rockwell,

31-5?€559. Digital simulation requires that for each instant of model time,

a Series of calculations be performed to produce a set of discrete out-

‘:>‘11:s. The reaction of the model to any input function is, therefore,

(l‘iilermined by repeating the series of calculations for each instant of

time for which the reSponse is required. Analogue simulation requires

that model variables simultaneously assume their appropriate values so

tjlat parallel recording of these values can reproduce all the important

\r a”)...

 

 

41

aspects of the system's performance (Link and Splinter, 1970). Generally,
the rapid development of high speed digital computer technology, and of
simulation languages based on such technology, has made digital simula-
tion preferable.

Hillier and Lieberman (1974) stated emphatically that simulation
models need not be completely realistic representations of the real
systems, since representing all of the minute details of real systems
often leads to excessive programming and use of excessive amounts of
computer time for the benefit of a small amount of additional informa-
tion. These authors further suggested that the behavior of system
elements is best represented by theoretical distributions which best
fit observed data from which random samples can be drawn. In light of
this, an important aspect of simulation modelling is the validation of
the model to show that it adequately represents the 'real-world' situa-
tion.

Model validation can be achieved by comparing observed 'real—world'

£3)rstem performance data with data generated by the model (Hillier and
Lieberman, 1974; Manetsch et a1., 1974). Standard statistical tests
E311(2has the Student's 't' test for the comparison of two means and the
X2 and F tests for inferences about variances, can sometimes be used to
Cletermine whether the two sets of data are statistically different.
These procedures are designed to make inferences about the values of
't:11€3 parameters 'u' and 's' that appear in the prescription for the
mathematical curve of the normal distribution and are collectively

1‘nown as 'normal-theory' parametric inference tests.

Another useful statistical test which may be used for the compari-

£3Onof two sets of data is the Wilcoxon Rank-Sum Test originally proposed

 

 

42

by F. Wilcoxon (1945). An equivalent alternative version of this test
was independently proposed by H. Mann and D. Whitney (1947) and is now
known as the Mann-Whitney U Test (Bhattacharrya and Johnson, 1977).
Neither the Rank-Sum nor the U Test is restricted by the assumption
that the data are normally distributed and they both test specifically
whether two samples drawn from different populations have the same dis-
tribution. These tests are non—parametric tests and they can be used
for both small and large sample sizes, although some test power is lost

as the sample size decreases (Bhattacharrya and Johnson, 1977).

3.4. Sugar Cane Simulation Models

Simulation models have been used for about 15 years to study farm
machinery systems. Specific areas of application have been: predicting
expected returns (Sowell et a1., 1967); analysis of specific cropping
systems such as cotton production (Stapleton, 1967); forage harvesting
(XZOupland and Halyk, 1969); the performance of field machines and trans-

I>C>rt units for a row crop planting system (Von Bargen and Peart, 1969);

tillea effect of the different harvesting system configurations on closed
Circuit cyclic transportation systems (Boyce, 1972); silage harvesting
<:Iinssel et a1., 1977) and sugar cane harvesting. The literature on
a‘F'Plication of simulation modelling to sugar cane harvesting is examined
th‘ some detail below.

Sorenson and Gilheany (1970) developed a model for testing different

StIrategies and decision rules governing the deployment of equipment on a
(lane plantation in the Caribbean. For this model, time for loading in

tile field, cane transport travel time for a given distance and unloading

 

 

43

time at the factory, were represented as constants, and statistical
models were developed for crop rotation, rain and trash, rained-out
field conditions, mill capacity and transport unit breakdowns. These
authors suggested that, with minor modifications, their program could
be used to optimize the length of the cane crop.

Farquhar (1972) discussed the potential of adapting the systems
approach in general and simulation in particular, to the analysis of
sugar cane production systems, and used a generalized model to highlight
some of the complexities of such systems.

Shukla, Chisolm and Phillips (1973) developed a computer program
for analyzing harvesting, loading and transportation of sugar cane.

The programme reportedly gave good results for the systems studied but
was somewhat locale-specific and depended heavily on the coefficients
derived from a time and motion study which had been on-going for several
years.

Early (1974) used an inventory model to simulate harvesting condi-

tiions in the Philippines. The specific objective of this work was to

g337nchronize field and factory operations under conditions of random
rainfall. He identified three limitations to his model based on the
Inetzhodology used in constructing the model, the data used in modelling
tlllee system response and the sequencing of rainfall events. Despite these
limitations, he concluded, after simulating 10 years of operations, that
'tillea existing policy of allocating a daily quota to all farmers, each in
I’lnaportion to their share of the milling capacity, could be improved by

adepting a system of reaping zones which showed better comparative yields

Siach month of the cropping period. He also estimated that this policy

‘Ehange, depending on the utilization and availability of irrigation,

44

could increase yields by between 12.5 and 16.5 percent.

A linear programming model was used by Tonsman (1974) to predict,
for different periods, which fields should be harvested, and the amount
and type of machinery that should be used at each location. This model
included in the optimization process, such agricultural characteristics
as seeding type and variety, soil preparation, irrigation regime, fertil-
ization, climatic conditions, water availability, machine field capaci-
ties, transportation equipment and the factory's production programme,
among others. Based on recommendations which resulted from this work,
improvements of 14 percent and 30 percent in the ratios of sugar pro-
duced to cane milled and sugar produced to harvested area, respectively,
were observed.

Hoekstra (1973, 1974, 1975) constructed three simulation models
for a single cane harvesting and handling system in South Africa. His
first model examined the effect of mill stoppages on transport units
which off-loaded directly into the factory. The results indicated that
the total number of transport unit hours lost over the cropping period,
for mill stoppages ranging from less than 10 minutes to 5 hours, was
negligible. In his second model, he examined mill yard operations. In
this model, four vehicle types were processed through one of two unload-
ing and each vehicle had to be weighed in and out. In general, the
simulation results showed that improvement in the cane flow through the
system could be achieved by improving the communications between the
mill yard and the different suppliers, and by increasing the amount of
cane delivered directly into the mill. In the third simulation model,
the influence of mill operations on the delay time between cutting and

milling was examined. In general, it was found that the delay between

45

cutting and milling was dependent only on the respective timing of the
cutting. Grinding over six days reduced both the mean and the Spread
of delay times. Hoekstra also indicated that the spread of delay times
about the mean was not due to the irregular grinding of deliveries but
rather to variances from a strict first-in first-out policy of milling
cane deliveries.

Cochran and Whitney (1975) examined the effect of different numbers
of transport units on field loader utilization. The overall results of
their work were combined with a theoretical analysis adopted from the
work of Melissa (1966) to develop a nomograph which permits graphical
prediction of delivery rates for given values of transport unit capabil-
ities, mean loading rate, total trip time and the number of transport
units. A cost model was also developed to facilitate selection of the
optimum number of transport units for any given system configuration.

Loader transport system costs were broken down into three categories.

(1) Labour cost - defined by equation 11
(2) Equipment fixed cost - defined by equation 12

(3) Equipment operating costs - defined by equation 13

The total loader transport system cost (TC) is obtained by summing
the three equations 11-13 to give equation 14. The equations are as
follows:

Systsn labour cost = Wlo + (Nt)(Wtr)/Del . . . . . . . . . . (11)

to

System fixed cost = Lfc + (Nco)(Cfc) + Z chi/Thui)

i=1

(l/Del) . . . . . . . . . . . . . . . . (12)

46

System operating cost = Loc/R + (Coc + Nt/Nc*Toc)

' (tP/(S)(Cc)) . . . . . . (13)
to
Total cost, TC = l/Del((Lfc + (Nco)(Cfc/Shu + Z chi/Thui
i=1
+ W10 + (Nt)(Wtr)) + (Loc/R + (Coc + Nc/Nc-
Toc-tP/((S)(Cc)). . . . . . . . . . . (14)
where
TC = Total transport cost $/tonne
Lfc = Loader fixed cost $/year
Cfc - Cart fixed cost $/year
chi = ith tractor fixed cost $/year
Thui = ith tractor hours of use per year hours/year
Shu = Season hours of use hours/year
Wlo = Labour cost for loader operator $/hour
er 8 Labour cost for tractor operator $/hour
Loc = Loader operating cost $/hour
Coc = Cart Operating cost $/hour
Toc = Tractor operating cost $/hour
tp = Round trip distance (field to mill) kilometres
Nt = Number of transport tractors in use
Nc = Number of carts in use
R = Loader rate for given conditions tonnes/hour
S 8 Average speed of transport units kilometers/hour
Cc = Cart capacity tonnes
Del = Transport system production rate to mill tonnes/hour

Nco

Number of carts owned

47

Nto = Number of tractors owned and reserved for
transport system use only during harvesting
season

One of the more recent applications of simulation modelling to sugar
cane handling was reported by Ogilvie et a1. (1978). These researchers
investigated the handling and transport of hand-cut whole stick sugar
cane at the Frame Cooperative Farm in Jamaica. Two computer simulation
models were constructed; one for the field operations and one for factory
yard operations. They reported that the models developed permitted the
testing of current and modified equipment arrangements and management
policies, which in turn resulted in higher throughput and optimum use
of transport and handling equipment. They also reported that the GASP
IV simulation language used for constructing the models not only proved
to be most adaptable to the system studied, but also provided the user
with excellent intrinsic filing and report formatting routines and cap-
abilities.

Loewer et a1. (1979) developed a computer model to optimally select
sets of equipment for 56 possible alternatives through the sugar cane
harvesting network in Brazil, and to compute the labour requirements
and fixed, variable, indirect and total costs. The results of the
simulation indicated that transportation costs accounted for 60 percent
of the total cost of harvesting sugar cane under Brazilian conditions.
It was concluded, therefore, that modification of the transportation

activity offered the greatest potential for reducing total harvesting

and handling costs.

48

3.5. Computer Simulation Languages

 

In implementing a systems model on a computer, the user has the
option of using a general purpose programming language or one of several
special purpose simulation languages. General purpose programming lan-
guages that may be of interest to the simulation modeller are FORTRAN,
ALGOL, BASIC, and PL/l. On account of their generality, these languages
may be used to construct simulation models of any type of system. How-
ever, the modeller is required to develop his own input-output routines,
set up his own time clock and switches within the model, and write his
own Special purpose routines such as normal pseudorandom number generators
(Dent and Blackie, 1979). Modelling using these languages requires con-
siderable programming ability and a reasonable knowledge of the computer
and its associated systems (Manetsch and Park, 1980).

Special purpose simulation languages, on the other hand, have
evolved in response to a need to reduce the programming skill and effort
required to program and use computer simulation models (Krasnow and
Merikallio, 1964). These languages contain specialised facilities, such
as automatic time-keeping routines and sophisticated output formatting
routines, which are convenient for modelling particular types of problems
(Teichrow et a1., 1966). Some of these languages are really supersets
of general purpose languages (for example CSMP is a superset of FORTRAN)
whereas others, such as GPSS, are self-contained (Tocher, 1965). In
any case, most of these languages were originally developed to satisfy
the requirements of specific problems and they, therefore, differ in the
type and range of their possible applications. Manetsch and Park (1980)

described a number of criteria by which simulation languages may be

49

evaluated. These are as follows:

Language type - Simulation languages are classified as either

 

'discrete event' or 'continuous flow' or both of these. Discrete event
languages are designed to simulate real world systems when a microscoPic
viewpoint, which considers individual objects or events as entities, is
appropriate. In contrast, continuous flow languages are designed to
simulate systans which are best characterized by aggregates or flows of
discrete entities. Models of such systems are normally structured

using differential and/or difference equations and continuous flow
languages usually contain integration packages designed to efficiently

obtain particular solutions for large sets of such equations.

Universality - This criterion describes the generality of a language

 

with respect to the range of computers with which it is compatible. Some
languages are very machine dependent and are useable only on limited com-

puter makes, models and sizes.

Higher Order Modularity - Situations are sometimes encountered in

 

which a complex sub-system structure occurs repeatedly in the structure

of the larger total system (for example, similar machines in a production
process). In such cases, it is often possible in simulation to design
one basic model 'higher order building block' which can be used repeatedly
in the overall model (with different inputs and structural parameter
values where necessary) to model the replicated sub—system whenever it
occurs. This capability of a language is termed 'higher order modularity'
since it involves modularity vis-a-vis a model component constructed

from more basic components.

50

Programmability - This criterion relates to the programming effort

 

required to program and operate models in a given language. The concept
is somewhat arbitrary since it is often highly dependent upon the partic-
ular problem at hand. Nevertheless, for some types of problems, some

languages are easier to work with than others.

Optimization Capability - Some simulation languages offer the

 

facility of running simulation models linked with optimization routines
(such as Complex or Powell's routines) which use search or other techni-
ques to optimize the performance of the model with respect to some

criterion.

Output Formats - Most special purpose simulation languages include

 

'canned' output formats which provide selected model outputs in tabular,
graphical (versus time), histogram or other forms. Such facilities can

significantly reduce model programming requirements.

The following is a tabulated comparison of a number of contemporary
simulation languages. The table was originally compiled by Manetsch and
Park (1979) for six languages, based on the six criteria described above,
but has been extended by the author to include an additional simulation
language 'GASP IV' and an additional descriptive criterion 'Error Diag-

nostics'.

 

 

Table 3.1 Comparison of some contemporary simulation languages
w____.
0
3 >. ~4
u o u m a
GH 14 U (0
° a r: .1 s g 2
15 m .4 u >. .o 'H >. no
. .3 '3 a: g 2:: .2 :2
no u :5 m 5H NH 9
a o s u a a -H-H u
a u #1 m .4 u .§.a : u
a: u u > .c a u: a a. o
“ 4'3 5 :1. 39'“ 8 ”“ “ t
.3 o 8 => =£ a. 8'3 5 n:
CSMP X Limited Yes Good Yes Good Good
number of
large computers
DYNAMO X Moderate for Yes Good No Good Good
large scale
machines
FORDYN X Excellent for Yes Fair Yes Fair Fair
most machines
FORTRAN XL X Excellent for Yes Poor Yes Poor Poor
most machines
GASP II X Excellent for No Fair Yes Good Good
most machines
GASP IV X X Excellent for Yes Fair Yes Very Good
most machines Good
GPSS X Limited number No Good No Good Good
of large machines
SIMSCRIPT .X X Good for large ? Fair Yes Good Fair
machines only
FORTRAN & X .1 Excellent for Yes Fair Yes Fair Fair
FORDYN 5 to
GASP 11 Good

 

 

 

 

 

 

 

 

 

 

i i___L_L_=

CHAPTER 4

METHODOLOGY AND MODEL DEVELOPMENT

4.1. Introduction

 

Most agricultural crop harvesting and tranSportation systems can
be represented by some form of closed circuit queuing system as shown
in Figure 3.1. Essentially, such a system consists of cutting the crop
and loading it into transport units which are then taken to a process-
ing, storage or other facility where unloading takes place (often after
the loaded tranSport units are weighed). Empty transport units are sub—
sequently returned to the field. Since a service facility (e.g., har-
vester, scale or unloader) is not always free on arrival of a transport
unit, service queues often build up within the system.

The calculation of the capacity of cyclic queuing systems can be
rather complex, especially in an agricultural context, due to the wide
variability of operating conditions that may prevail. One relatively
simple approach to finding approximate solutions for such systems is
through the use of computer simulation programs. In such programs, the
variability of operating conditions can be handled by using appropriate
statistical distributions to represent the operating times of the dif-
ferent pieces of equipment involved in the system (Dumont and Boyce,
1972). This chapter describes the simulation methodology used to study

mechanical harvesting and handling of sugar cane under Barbadian condi-

tions.
52

53

4.2. Model Development Approach

 

There are two distinct systems of mechanical harvesting of sugar
cane in use in Barbados: chopper harvesting and whole-stick harvesting.
In the chopper harvesting system, the cane harvested by a combine is
delivered directly into a wagon pulled alongside the harvester by a
field tractor. When a wagon is filled, it is moved from under the
delivery chute of the harvester and an empty wagon is drawn up in its
place in such a manner as to ensure continuous harvester operation. In
the case of the whole-stick system, the cane is cut by a mechanical
cutter and then loaded by a mechanical grab loader (in a separate opera-
tion) into a transport trailer pulled alongside the loader in the field.
. In either system, filled transport units are deposited in a queue at
the edge of the field and, when a complement of two full wagons or
trailers and a road tractor becomes available, a trip is diapatched to
the factory. The actual flow of individual operations performed on
each transport unit during the entire harvesting and handling process
is shown in Figure 4.1. Operations above the 'broken line' on the
Figure are associated with the field subsystem, while those below the
line are involved in the factory subsystem.

For the purposes of this study, the overall cane harvesting and
handling system was broken down into two subsystems: a field subsystem
and a factory subsystem -- a division not at all unrealistic since in
Barbados, the cane production and processing sectors are separately
owned and managed entities (see Chapter 2). Each subsystem was then
decomposed into its Specific activities, and events identified such

that each activity was bounded by a pair of sequential time events, one

54

 

 

 

 

WAGON WAITS TO BE LOADED

 

WAGON IS MOVED TO HARVESTER

 

 

 

WAGON IS LOADED

 

 

WAGON IS MOVED AWAY FROM HARVESTER

 

 

WAGON WATTS TO BE MOVED TO FACTORY

 

 

 

UNIT (2-WAGON TRAIN) IS MOVED TO FACTORY

 

 

UNIT WAITS TO BE WEIGHED

 

 

 

EACH WAGON IS WEIGHED SEPARATELY

 

 

 

 

UNIT IS MOVED TO UNLOADER

 

 

UNIT WAITS TO BE UNLOADED

EACH WAGON IS UNLOADED SEPARATELY

 

 

 

 

 

UNIT 18 RETURNED TO FIELD

 

 

 

 

 

Figure 4.1. Flow process chart of operations performed

on each transport unit

55

signalling its start and the other signalling its end. The time to
perform a given activity was calculated by finding the lapse time between
the relevant pair of sequential events.

Two computer simulation models were then developed: a field opera-
tions model (FIELDOP) to simulate the activities involved in the field
subsystem; and a factory yard operations model (FACYARD) to simulate
the activities involved in the factory subsystem. Based on a review of
the literature on simulation models and computer simulation languages,
the GASP IV simulation language was selected for constructing both models.

Details of this language are presented in the following section.

4.2.1. An Overview of the GASP IV Simulation Language (Source: Pritsker,
1974)

 

Simulation is a problem solving procedure for defining and analys-
ing a model of a system. Digital computer simulation, as defined by
Pritsker (1974), is "the establishment of a mathematical-logical model
of a system and the experimental manipulation of it on a digital com-
puter."

Simulation languages typically provide the structure and the term-
inology to facilitate the building of simulations and relieve the user
of a considerable amount of personal programming effort. GASP IV is
such a language. It helps the user to build computer simulation programs
that can be both the model of the system and the vehicle for analysing
the system.

The philosophical basis for the design of GASP IV is the concept
of modelling a system in two dimensions: the time dimension and the

state-space dimension. Fundamental to building a GASP IV simulation

56

model is the decomposition of time and state space into manageable ele-
ments. The decomposition in the time dimension requires the user to
define events and potential system changes generated by these events,
to specify the causal mechanisms by which events can occur, and to de—
fine the mathematical-logical relations that tranSpire when an event
occurs.

In the state-space dimension, GASP IV presumes that a system model
can be decomposed into its entities, which are described by attributes.
Attributes may be discrete or continuous. A discrete attribute is one
whose value remains constant between event times, while a continuous
attribute is one whose value may change between event times according
to a prescribed code of dynamic system behaviour.' Continuous attributes
are referred to as "state variables". In essence then, GASP IV provides
a formalized view of the world which specifies that the status of a
system be described in terms of a set of entities and their associated
attributes and state variables.

In GASP IV, an event occurs at any point in time beyond which the
status of a system cannot be projected with certainty. Events are
described in terms of the mechanism by which they are scheduled. Those
that occur at a specified projected point in time are referred to as
"time events", while those that occur when the system reaches a partic-
ular state are called "state events".

The behavior of a system model is simulated by computing the values
of the state variables at small time steps and the values of attributes
at event times. GASP IV automatically decomposes the time axis into

points at which events occur, based on the equation form of the state

57

variables, the time of the next event and accuracy and output require-
ments. The user is, therefore, relieved of the task of sequencing events.

When an event occurs, it can change the status of the system in
three ways: by altering the values of the state variables or the attri-
butes of the entities; by altering relationships that exist among enti-
ties or state variables; and/or by changing the number of entities pre-
sent. Methods are available in GASP IV for accomplishing each type of
change.

At each time step, the state variables are evaluated to determine
if the conditions prescribing a state event have occurred. If a state
event was passed, the step size was too large and is reduced. If a
state event occurs, the model status is updated according to the user's
state event subroutine. Step size is automatically set so that no time
event will occur within a step, by setting the step size so that the
time event ends the step.

Since time events are scheduled happenings, certain attributes are
associated with them. At the minimum, a time event must have attributes
that define its time of occurrence and its type. If the time event is
associated with an entity, then either the attributes of that entity must
be associated with it or the event must be able to refer to the attri—
butes of the entity. For example, if there is an end-of-service event
for an item, the attributes for that item must,:hnsome way, be associated
with the event. A filing system is provided for storing entities and
their associated attributes. System monitoring procedures, which can be
pre-scheduled or called as required, are also provided.

GASP IV is designed to provide eight specific functional capabili-

ties:

58

(1) Event control

(2) State variable updating using integration, if necessary
(3) System state initialization

(4) Program monitoring and event reporting

(5) Information storage and retrieval

(6) System performance data collection

(7) Statistical computations and report generation

(8) Random deviate generation.

The first four of these capabilities are primary functions which
constitute the basic modes of the language as shown in Figure 4.2. The
remaining four are support oriented.

GASP IV has two forms of program control. One, the executive
function, directs the system program into its various modes: initializa-
tion, state variable updating, monitoring, etc. The other, the event
selection function, operates within the simulation model and sequences
the execution of event routines. The modular structure of GASP IV
allows event logic to be changed relatively simply to investigate the
effect of changes in a system on selected measures of system perform-
ance, since each event is, in fact, a separate subprogram. A data pool
allows changes made in data inputs to be communicated throughout an
entire simulation model. The preparation of reports summarizing the
results of simulation runs is simplified by utilizing standard report
programs that obtain their information from the common data pool. Model
debugging is also facilitated by the provision of access points at which
program results can be sampled without interfering with the logic of

particular events.

 

 

 

 

 

 

 

 

59

 

 

 

 

 

 

 

 

l
l l .
8:53.39...- . ,2 - 7w: .4 .8»...
.....é- I E 3 II
s l l r

 

 

 

 

 

95.933... 62.2.. :35 8:. ...-=8

 

 

 

 

 

 

 

 

 

cos-22. 25326 L

a!”

 

 

 

 

 

 

 

Figure 4.2. Basic modes of GASP IV control.

60

In summary, GASP IV has five distinct features that make it partic-

ularly attractive as a simulation language:

(1)

(2)

(3)

(4)

(5)

The language is FORTRAN based and requires no separate
compiling system. It is easily maintained and can be
implemented on new computing systems and on the comput-
ing systems of different manufacturers.

GASP IV is modular and can be made to fit on all
machines that have a FORTRAN IV compiler.

GASP IV is easy to learn since the host programming
language is usually known and only the simulation
concepts need to be mastered. The implementation of
these concepts is apparent to the user.

GASP IV can be used for discrete, continuous, and com-
bined simulation modelling, and is the only well-docu-
mented simulation language with this capability (see
Table 3.1)

GASP IV can be easily modified and extended to meet

the needs of particular applications.

A functional flowchart of a GASP IV program is presented in Figure 4.3.

4.3 The Field Operations

 

4.3.1. Objectives

The main objective of simulating the field operations was to identify

changes in the cane harvesting and transportation procedures that could

lead to increased efficiency of utilization of the expensive chopper

CM manual

‘ STA!" )

l

 

 

61

"with! M UM" IV

 

 

“on wound.
“amino 7 III m
cum-6A9 vaults maul MO

 

 

 

Swimm- INTLC
Immin- non-GASP

 

 

 

 

 

 

I Mud-a GASP MUM
I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

canola. «on
warm and tun-mu I
ii,
I /|\
VII M 1
I 00"“ A
I u.
I ..
M a V“
tulle-m
I m? 1
No m Im-
l m: "on
Swot-um “AV! L L
Rm m I
Mum STATE I
than om. vacuum V“
ESuI-thm SCOND '
We non-m
Nahu-
WI-um u... an
* I
l
l
— fl
; l
m swan |
SUI-u WI. 7
L m I
1
twin (an: . . Em
Two | Type 2 Tu. N I
i I
Saul-twan- 0"!" I Stimu- SUNNY
W mod-m 7 but W m
a um“ mm 6! M
MO” I _—_——i.

 

Mum to START.
RESIN". or 510'

 

Figure 4.3. Functional flowchart of a GASP IV program.

combine

were as

l.

62

harvesters and their associated equipment. Specific objectives
follows:

Determination of the current levels of utilization

of the chopper harvesters and of the cane transporta—

tion equipment within the field subsystem.

Construction of a computer simulation model to accu-

rately reflect the Operational characteristics of the

field subsystem.

. Use of this model to determine the effect on potential

subsystem output of variations in the characteristics,
specifications and combinations of different subsystem

components.

. Use of the simulation model to test recommendations

which may be implemented to improve and/or optimize the
overall performance of the field harvesting and cane

transport operations.

4.3.2. Activities and Events

The principal activities involved in the field subsystem are the

loading

of empty cane transport vehicles and the dispatch of full trans-

port units to the factory for processing (Figure 4.4). Inevitably, road

tractors are not always available whenever a complement of two full

wagons is ready for transportation to the factory, neither are harvesters

always free when empty transport wagons return to the field. Queues of

full and empty wagons, therefore, form an integral part of the field

subsystem.

63

 

 

I
IFROM FACTORY .

 

 

 

 

 

 

 

r--——-——————-

 

 

 

 

 

 

 

 

wuoeo<m OH

 

$:I$_+______________M___________________________________________________c

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ _ I _. _=______________=__=______:H

 

 

 

 

 

 

 

 

 

 

 

 

 

_________________________________________________________________$_:_u

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

QNQ<OA qumm zoo¢3

 

 

 

 

lll

 

 

 

 

Figure 4.4. Activities involved in the field subsystem

64

The Specific sequential time events identified and monitored for

the field Operations model are detailed in Table 4.1 and the sets of

activity times derived from these time events are as follows:

1. TTHT:

2. LTPW:

3. TTFQ:

4. HIT:

5. TTTF:

6. ZATU:

travel time of field tractor and empty wagon
combination to harvester for loading (STLD -

STHK).

loading time per wagon, including turning time at
the end of each row when applicable (ENDLD - STLD).
travel time of a field tractor plus a loaded wagon
to the queue Of full wagons (ENDUH - ENDLD).
harvester idle time due to unavailability Of empty

wagons at the field (STLDn+ - ENDLDn).

1
travel time of a road tractor and its complement
of two full wagons from the field to the factory
(ARFCT - DPFLD).

inter-arrival time of cane transport units (a unit
being 1 road tractor plus 2 wagons) at the factory

l

7. RTFCT: - residence time for a transport unit at the factory

(DPFCT - ARFCT).

8. TTFLD: - travel time of road tractor plus two empty wagons

from factory back to field (ARFLD - DPFCT).

4.4. The Factory Yard Operations

4.4.1. Objectives

The main objective of simulating the factory yard operations was

to identify changes in the layout and organization of the yard which

65

Table 4.1. Events monitored for the field operations.

 

 

Event

Event
Code

Definition

 

Start hook

Start load

End load

End unhook

Depart field

Arrive factory

Depart factory

Arrive field

STHK

STLD

ENDLD

ENDUH

DPFLD

ARFCT

DPFCT

ARFLD

Time at which a field tractor stops
in front of an empty wagon in prep—

aration for hitching.

Time at which the first cane hits

the bottom of the wagon.

Time at which the rear end of the
loaded wagon passes the front end
of the harvester on its way to the

queue of full wagons.

Time at which a field tractor
starts moving, having deposited
a loaded wagon in the queue of full

wagons.

Time at which a road tractor plus
its complement of two full wagons

starts moving enroute to the factory.

Time at which the front wheels of
the road tractor cross the entrance

to the factory.

Time at which a road tractor plus
its complement of two empty wagons
or trailers first start to move
enroute back to the field, the cane

having been unloaded.

Time at which a road tractor and its
complement of two empty wagons stop
in the queue of empty wagons at the

field, having returned from the factory.

66

could lead to improved operating efficiency, before committing large

sums to major physical changes in yard configuration. Specific objec-

tives were as follows:

1.

Determination of the current levels and rates of utiliza-
tion of existing equipment and the overall output of the

factory cane handling system.

. Construction of a computer simulation model to accurately

represent the current configuration of a given factory

yard.

Use of this simulation model to investigate the effects
of changing the values of certain model parameters and

variables so as to achieve varying operational charact-

eristics of the factory yard processes.

. Minimization or reduction of the residence time of cane

transport equipment in the factory yard, since this
implies more efficient use of both the cane harvesters

and the transport vehicles.

4.4.2. Activities and Events

The principal activities involved in the factory yard process are

weighing and unloading of the cane transport vehicles. Since, however,

the service facilities (scales and unloaders) are not always free when

a tranSport unit arrives at the service area, two additional activities

"waiting in the queue to the scale" and "waiting in the queue to the

unloader" were also considered.

The sequential time events identified and monitored for the factory

yard Operations are listed in Table 4.2 and the activity times derived

67

Table 4.2. Events monitored for the factory yard activities

 

 

 

Event Event Definition
Code

Arrive factory ARFCT Time at which the front wheels of the
road tractor cross the entrance to the
factory.

Start weigh STWGH Time at which the front wheels of the
road tractor first make contact with
the weigh scale platform.

End weigh ENDWGH Time at which the rear wheels of the
second wagon or trailer (in a train)
leave the weigh scale platform.

Start unload STULD Time at which the chains of the unload-
ing crane or hoist first make contact
with a full wagon or trailer.

End unload ENDULD Time at which a road tractor and its

complement of two empty trailers or
wagons first starts to move en route
from the factory after the cane has

been unloaded.

 

68

from these events are defined below:

1. Weigh Q-time: - time road tractor and its complement of two
full wagons or trailers spend in the queue waiting to be
weighed (STWGH - ARFCT).

2. Weigh time: - time to weigh complement of two wagons or
trailers (ENDWGH - STWGH).

3. Unload Q-time: - time road tractor and its complement of
two full wagons or trailers spend in the queue waiting to
be unloaded (STULD - ENDWGH).

4. Unload time: - time to unload complement of two wagons or
trailers (ENDULD - STULD).

5. Factory Residence time: - total time road haulage tractor
and its complement of two wagons or trailers are retained

at the factory yard (ENDULD — ARFCT).

4.5. Data Acquisition and Analysis

 

The main data required for this study were duration times of the
various activities involved in the Field and Factory yard operations.
As stated in Section 4.2, these times were obtained by monitoring the
occurrence times of sequential events and then calculating the lapse
times between appropriate pairs of events. The event times monitored
for the field and factory yard systems have already been presented in

Tables 4.1 and 4.2, respectively.

4.5.1. Data Acquisition

 

The data collection for this study was undertaken during the period

February 24 to April 10 of the 1982 cane harvesting season in Barbados.

69

For this exercise, five monitors were employed, to of whom were assigned
to the field operations and three to the factory yard Operations.

In the case of the field operations, one monitor was responsible
for recording the event times associated with the loading process in-
volving interaction of the combine harvester, the field tractors and
the transport wagons. Event times associated with the arrival of empty
transport units at the field and the departure of full transport units
to the factory were recorded by the other monitor.

In the case of the factory yard operations, one monitor recorded
transport unit arrival times at the factory while a second monitor
recorded event times associated with the weighing activity for all trans-
port units as well as those associated with the unloading activities for
chopped-cane units. The responsibility of the third monitor was to re-
cord event times associated with the unloading of whole-cane tranSport
units.

All persons selected as monitors were required to study an instruc-
tion manual which explained the data collection process and described
how to identify the precise times at which the various events occurred
(see Appendix I). In addition, prior to the commencanent of the actual
data collection, the monitors were taken into the field and each was
shown exactly how to record his own data as well as the data for which
each of his co-workers was responsible. This precautionary measure
'was taken to ensure that monitors could be readily inter-changed as and
when the need arose.

Each monitor was equipped with a clip-board, an adequate number of
data recording forms, scoring pencils and pens and a digital quartz

watch with a continuous, 6-digit time display. In order to synchronize

70

events occurring at various points in the system, all watches used were
adjusted on the first day so that they all read the same time and kept
synchronized throughout the study.

Data recording consisted of entering the event times and the iden-
tity of the field tractors or road tractors (registration numbers) in
the appropriate columns on the forms. The comment number column was
used to identify activity times during which unusual events occurred.

A comment number consisted of two numbers; a row (observation) number
and a column (event) number which together indicated the exact event
time entry to which the comment referred. The actual comments were
written in the "comments" section provided at the bottom of the data

recording form. Samples of some actual data are given in Appendix 1.

4.5.2. Data Analysis

 

In order to model the specific events using the intrinsic distribu-
tion functions of the GASP IV simulation language, it was necessary to
select appropriate statistical distributions to represent the various
activity times and then determine suitable parameters for these distri-
butions. Based on the literature review presented in Chapter 3, activity
times for the loading, weighing and unloading processes, and waiting
times in the queue to the various servers can all be represented by
either the Erlang or the Gamma distribution. Since the latter is a more
general form of the former (which itself is a more general form of the
Exponential distribution) it was decided to use the Gamma distribution
to determine the appropriate parameters for the distributions Of the

activity times mentioned above.

71

Two forms Of the Gamma distribution exist: a 2-parameter version
which assumes that the minimum value for the dependent variable is
always zero and a 3-parameter version which allows the dependent vari-
able to take on minimum values greater than zero (Figure 4.5). Since,
for this study, the minimum values of service or activity times are
always positive, the 3-parameter Gamma was the version fitted to the

data in all cases.

4.5.3. The Data Analysis Program

 

Most of the program used was taken from a program originally
develOped by Dumont (1972). The original program accepted activity or
service times which it processed and analysed as statistical distribu-
tions, using the maximum likelihood method to fit the distribution. It
was later modified by Ogilvie et a1. (1978) to provide for input data
manipulation into a form suitable for curve fitting analysis. Minor
improvements in the computer code and general program flow have been
made by the author.

The general flow through DATANAL is shown in Figure 4.6. The
program accepts data in sequential event time form just as it is re-
corded in the field. Subroutine DACON converts the 6-digit event times
from the 'hour, minute and second' mode to a single decimal number,
single unit mode. In subroutine COLDIF successive event times are sub-
tracted from preceding event times so as to produce columns of single
unit decimal service times. These columns of service times are then
sorted in ascending order by the main program and used as input to sub-
routine SIMPLE which determines the values of the three parameters of

the GAMMA distribution for each column in turn. Output from the program

72

.comumnwuumwv mEEmw m5» m0 wEhOm nwuwemumm um can LN one .m.v wusmflm

x X
OS Qm 0.0 06 o.» o.~ o.—

 

 

0.0

 

 

 

xx 2p.” r. $2
$43.. I” .5qu A .x:

 

 

 

onhanmamHa dzz<w zomhamwmhwmo dzxmo
Hmz<mmmum «mamzmmmanm

 

cm 3

 

 

 

 

 

73

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

C 3

Figure 4.6. Flowchart through Data Analysis

Program

 

 

FUNCTION
DECIHR
INIT
- CONVERTED
I .——- DACON ; INPUT DATA
INPUT
DATA
TIME
COLDIF DIFFERENCES
DATA
FUNCTION
SA
SORTED
DATA FUNCTION
SB
FUNCTION
A(///’///A LIK
SIMPLE A» ITERgTIONS
SUMMARY
FUNCTION
w(/////// GAMMA
.I_ HISTOGRAM
FITA GRAPH 1,, PLOT ,
~\\\\\\\\ FUNCTION
RINT
HISTOGRAM
P: DATA
AYES MORE \/—
DATA
NO

(DATANAL)

74

includes: a list of input event data converted to single unit data, a
list of activity or service times, histograms of the distribution for
each activity time and parameter values calculated for each distribution.

The computer listing of program DATANAL is given in Appendix II.

CHAPTER 5

FIELD AND FACTORY YARD COMPUTER SIMULATION MODELS

5.1. Introduction

 

A system may be defined as a set of inter-connected elements or
components which interact, under a prescribed set of conditions, to
achieve a Specific goal or set of goals (Naylor et a1., 1968). In
general systems terminology, a number of different systems components
have been identified (Manetsch and Park, 1980). These are:

l. Exogenous or environmental variables: these are input

variables determined by factors completely independent
of or external to the system, and which together consti-
tute the system environment.

2. Entities: these are objects within the boundary of the
system.

3. Endogenous or dependent variables: these are internal or
output variables of a system which are generated or caused
by system inputs and/or interaction or other endogenous
variables.

4. System boundary: this is an imaginary line which sepa-
rates endogenous system components from the exogenous
variables or the system environment.

5. System output variables: these are variables caused by

a given system and are used either as inputs to another

75

76

system or as measures of performance of the given

system.

6. Controllable input variables: these are input factors

which can be used as a means of overtly altering system

behaviour, usually in a beneficial manner. They are

also known as management or decision variables.

The set of interactions of elements and related variables that

intervene in the causal chain linking system Outputs to system inputs

is referred to as the systems structure. Causal lOOp diagrams, in which

the system elements and the variables that interact among them are rep-

resented in two dimensions, are often used to illustrate system structure.

5.2.1. The

5.2. Field Operations Simulation Model

 

System Environment

 

In the case of the field operations model (FIELDOP), the system

environment consists of the following exogenous variables:

- Cane variety (its suitability to mechanical harvesting)

- The

- The

- The

- The

- The

- The

- The

conditions of the field

daily quota of cane fixed by the factory
factory yard conditions (degree of congestion)
road conditions (traffic congestion)

climatic conditions (particularly rainfall)
operational policies of management

travel distance between the field and the factory.

77

The first five factors may affect the output of the harvester
directly, while the remaining three would tend to affect the turn-
around time (total round trip time) of cane transport units (see Figure
5.1). For purposes of model simplification, however, a number of assump-
tions were made about these factors:

1. The travel distances between fields.andfactories are

known.

2. Variations in field, factory yard and road conditions
can be handled by using appropriate statistical distri—
butions of the work times for the varioUs operations
performed at these locations.

3. Since the daily quota of cane is established for six
consecutive days of a week at the beginning of the week,
over-supply on one day is deducted from the next day
and under-supply is added.

4. The factory yard operations (weighing and unloading)
can be represented as a single server operation for
the field operations model, since total round—trip time
is the time variables of interest here.

5. The field system operates from 7:00 a.m. to 7:00 p.m.,
six days a week with a break of one hour per day for
lunch. Further, based on short-term observation, approx-
imately one hour is utilized per day for repairs and
general maintenance. Effectively, then, 10 hours (two
5-hour shifts) are available for work per day.

6. All harvesters, field tractors, road tractors and wagons

in one location are identical and their work times can,

78

    

mmawu

HO

 

  

\

\|.._-I
\

\ waowmepz /
s
. Hana

”mo onus—u 3.

I’m.“ 255 x

I
"-“

 

 

mafi>umm

nuwsoa

  

Soumzw pamfiw mnu “Om Emuwowp aOOH Humnmo .H.m ounwwm

 

   

amozuma
mocmumfi.

 

zuouumw

    

  

-- /

+v\\maowwwz
. huaaw .
. mo wanna H
,5" mafia \
I, \

\

"- ‘

..-FH\
\\%66 MW

\
avoumu>umsx

 
    

 

muwumm>u.
memo
mo .o:

      

 
 

 

 
 

 

 

   

 

 

 

 

muouomuu I +
s I _
moon \ 06.3 / muouomuu +, 053 7
mo .0: “ H03 , mamfim I; Omanoyx
.. x L mo .0c 1!! - x
. 3.33. + /
I \ + K
/

 

mumwcnom amumam

 

mmanmfinm>
msocmwovcm IIII

wuoav
%HHMp

hamwum>
mama

79

therefore, be represnted by the same probability
distributions.
7. A transport unit always consists of a road haulage
tractor and a complement of two wagons.
8. Delays in field operation due to rainfall are infrequent
in occurrence and negligible in duration.
9. Transport units and individual wagons are treated on a
first-come, first-served basis throughout the system.
10. Agronomic practice dictates that, for a given field,

only one cane variety is used.

On the basis Of the above assumptions, the daily cane quota, the
transport unit capacity, the length of a shift, the length of the break
between shifts, and the operational policies of management are all con-

sidered to be constants in the system model.

5.2.2. Entities
The entities considered for the model are:
l. The mechanical (chopper) harvesters
2. The field tractors
3. The cane transport wagons

4. The road tractors

The operating personnel associated with the equipment (e.g., tractor
operators, the harvester operator) are considered to be integral parts
of the component equipment and do not affect the system as individuals.

They are, therefore, not treated as separate entities in the model.

80

5.2.3. Endogenous Variables

 

The endogenous variables generated within the system boundary are

as follows:

Time in the queue of empty wagons
Time in the queue of full wagons
Time to complete a round-trip

Tonnes of cane harvested per day

. Available work time.

5.2.4. Output Variables

 

The output variables of interest for the field subsystem are:

1.

2.

10.

The average total round-trip time
The average time an empty wagon waits at the field to be

loaded

. The average time a full wagon waits at the field to be

transported to the factory
The utilization of the harvester

The utilization of a field tractor

. The utilization of a road tractor

The number of trips completed per Shift

. The number of trips in transit at the end of a Shift

The number of full wagons in the field
The total tonnes of cane delivered to the factory from the

field subsystem.

81

5.3. Field Operations Simulation Program (FIELDOP)

As previously stated, in GASP IV each entity has one or more attri—
butes associated with it. The entities and their associated attributes
for the FIELDOP program are listed in Table 5.1. Nine files are used
in the program for storing information on entities and combinations of
entities at different points in time during the simulation. These files,
along with the entities stored in each and the attributes used to rank
entries in files, are shown in Table 5.2. Throughout the simulation,
the identity of each piece of equipment is maintained by assigning it
to the appropriate file. For example, if at the end of a shift, a
field tractor and empty wagon combination remains in the field, the com-
bination is dissembled and the field tractor put in the file for field
tractors (File 5) while the empty wagon is placed in the file for empty

wagons (File 2).

5.3.1. General Flow Through Program FIELDOP

 

The general flowchart for program FIELDOP is presented in Figure
5.2. The main program assigns appropriate values to the card reader and
printer units, reads in the initial values for non—GASP (user-generated)
variables (Table 5.3), prints out a list of these variables, and initi-
alizes the GASP IV variables declared in the Dimension, Common and
Equivalence statements before making a call to the GASP IV executive
subroutine, GASP. The GASP IV subroutines DATIN and INTLC (user-written)
are then called from subroutine GASP. DATIN enters input data read in

by the main program and subroutine INTLC initializes the start, end, and

operating work times, daily cane delivery quotas, report arrays and

 

82

Table 5.1. Entities and their associated attributes (FIELDOP)

 

 

 

 

 

Attribute Entity
Number Description Harvester Field Wagon Road
tractor tractor
1 Event time X X X X
2 Event code X X X X
3 Time of arrival in
queue of empty wagon X
4 Time of arrival in
queue of full wagons X
5 Time of departure
from field X X
6 Time of arrival of
a transport unit at
factory X X
7 Time of departure of
a transport unit from
factory X X
8 Harvester number X
9 Field tractor number X
10 Road tractor number X
11 Number of full wagons
for one trip to factory X
12 Trip completion indicator X X
13 Identification Of field
location (i.e., NEST) X X X X
14 Time of entry into
file 8 X X

 

83

Table 5.2. Files used in model FIELDOP

 

 

 

File no. (I) Entries KKRNK (I) IINN (I)
1 Future events 1 3
2 Queue of empty wagons at field 3 3
3 Queue of full wagons at field 4 3
4 Queue of road tractors l3 3
5 Queue of field tractors 13 3
6 Harvesters 13 3
7 Information on road tractors

processed at the factory 6 1
8 Field tractors and empty wagons

coordinates waiting to be loaded 14 1
9 Transport units in transit at

end of the morning shift

 

 

cameo
L

INIT

RDR
PRNT

 

 

 

GASP

 

 

 

 

 

 

 

 

 

 

 

 

** USER SUBROUTINE

f

 

b

EVNTS ‘-____‘

84

COMMON
DIMENSION

JEQUIVALENCE

OTHER SUB-PROGRAMS USED

FUNCTIONS

GAMA
ERLNG
RNORM

 

 

 

FILEM

 

 

 

 

INTLC

 

 

 

 

 

 

 

ARFLD

 

 

 

 

STLD

 

 

 

 

 

NDLD

 

 

ARHTR

 

 

 

 

 

 

ARFCT

 

 

 

DPFCT

 

 

 

 

 

 

SHTDN

 

 

 

 

 

 

 

 

STRTUP

 

 

* GASP IV SUBROUTINE (USER WRITTEN)

Figure 5.2. Flowchart through FIELDOP

SUBROUTINES

DATIN
TIMST
COLCT
FILEM
RMOVE
ERROR
DSTART**
REPRT**
OTPUT*

85

Table 5.3. NON-GASP user input variables used in FIELDOP

 

 

 

Symbol Definition
ISC System condition: ISC = 1: simulate
ISC = 0: end of simulation
NWG Number of wagons
NHT Number of harvesters
NFT Number of field tractors
NRT Number of road tractors
NEST Identifies estate simulated
QUOTA Estate delivery quota of cane for one day
CAP Amount of cane delivered by a transport unit per trip
(tonnes)
STM Time of starting field operations on first day
ENT Time of ending field operations on a work day
AMOUNT Amount of quota remaining
DWN Status of system: DWN = 0: system Operational
DWN = 1: system down
BREAK One hour lunch break for which harvesting stOps

between shifts

 

86

equipment combinations, and establishes performance criteria for the
simulation. A successful return to the executive routine GASP from
DATIN is indicated by a print-out of the initial values established for
GASP IV variables. The executive routine GASP then calls subroutine
EVNTS (IX) which passes control to individual user-written event sub-
routines ARFLD, STLD, NDLD, ARHTR, ARFCT, DPFCT, SHTDN and STRTUP, in
accordance with the value of the argument IX.

Subroutine ARFLD (Figure 5.3) is called when a transport unit,
consisting of 2 wagons and a road tractor, arrives at the field, having
delivered a load of cane to the factory. The road tractor is placed
in File 4 and the empty wagons in File 2. The amount of the day's quota
remaining is then calculated, as well as the round-trip time for the
delivery. If the quota is satisfied, a shut-down event is scheduled
and the road tractor is set idle. If not, the status of the harvesters
is checked and a start of loading event is scheduled for each available
harvester. If there is a complement of full wagons at the field, it
is combined with the road tractor and a trip is diSpatched to the factory.
Otherwise, the road tractor is set idle.

Subroutine STLD (Figure 5.4) handled the start of a loading event.
The harvester is combined with a field tractor and empty wagon combina-
tion and set busy. The end of loading is then predicted by generating
a deviate from the Gamma distribution of loading time using parameter
set #1 (see Tables 5.4 and 5.5).

Subroutine NDLD (Figure 5.5) simulates the activities in the field
at the end of loading. The wagon just loaded is placed in the queue of

full wagons (File 3) and the field tractor in File 5 (see Table 5.2).

87

 

C m)

Is
transport unit Yes Reduce quota

arriving from by capacity of
facfi::Z//””’r transport unit

Calculate ‘

    

   
 
 
 
 

 

No

round-trip time

 

 

J,

Park empty wagons lt

 

in queue 2

1

Park road tractor
in queue 4

 

 

 

   

  
     
 
 

Is
system
down

 

 

Yes { RETURN )

 

Y Schedule shutdown
es .
of operations for
the day

quota fulfilled

   

 

 

Check no. harvesters ( )
available J RETURN

Figure 5.3. Flowchart through subroutine ARFLD

88

 

 

 

 

 

 

Is
NO no. field tractors
CT. 0
Yes
Is Set no. hgrvesters
no. field tractors Yes to be used .EQ.
GT. ’ no. empty wagons
no. empty wagons available - N
Set no. harvesters to be
used .EQ. no. field
tractors available - N
Is
Yes N.EQ.O
No

 

Set up N field tractor and
empty wagon combinations
and schedule arrival

at harvester

 

 

 

  
  
 
 

  
 
 
   
  

Are
all wagon complements
available for dispatch
to factory

 

 

Set road tractors
idle, collect stats.

1
C W}

Figure 5.3. (continued)

 

 

 

 

 

 

 

 

 

Dispatch as many trips to
factory as there are road
tractors/full wagon
combinations available

 

1

Schedule arrival time at
factory for all units

dispatched

 

J

 

89

 

C 'sm 3

Record identity of_]

 

field tractor

 

 

Remove a field tractor/empty wagon
combination from file 6

 

 

 

collect statistics

Predict end of loading
for wagon

c m)

Figure 5.4 Flowchart through subroutine STLD

‘ Set harvester busy and

 

 

 

9O

 

axon any
wanOOH mo uumum mnu

 

pan comma H onmeH wo meHu OHOH

m wcmHum mouncHz one consume oaHu omemH amumm>umm

mcowm3 HHnw some: H a HHnm

o wOMHum wocoomm mo mnmnc Ou oEHu Hm>mua HOOOMHO pHmHm Ho>muu pHmHm

comma H a woman

m wOMHum mpcooom umumm>umn Ou mEHu Hm>mu9 nouomau pHOHm Hm>muu pHOHm

Mao: use moowns N a xuaam

a Hmsuoz monuOEOHHM pHmHm on women Hm>muH uouomuu omom Ho>muu vmom

huOuomm :H

msowma N w wsHmeHco

m maamo mounst %uOuumm um oaHu musmpHmom HOuomwu wood can wanwHoz

anon you muowms N w HHaw

N Hassoz mmuuoEOHHM muouomu Ou women Ho>mue wouomuu pmom Hw>muu vmom

H assoc mmuost sowm3 awe oaHu wSvaOH umumm>unm onvaH

Hanan: mom mama muHsa paumHSOHmO vo>Ho>sH mu > uo

umuoamumm OOHunnHuuan maHu >uH>HuO< >uHusm H H ¢

 

 

mOQHmHm Hmvoa CH mmaHu auH>HuOm pom pom: maOHunnHuumHv HOOHumHumum .¢.m mHan

91

 

 

 

Table 5.5. Input parameters for distributions used in FIELDOP
Paramter Dist.
Set A Min. Max. B type Entity
l 6.10 0.00 15.46 2.00 GAMMA Loading time per
wagon
2 12.74_ 6.44 24.62 5.89 NORMAL Travel speed to
factory
3 5.21 0.00 143.40 4.00 GAMMA Factory yard
residence time
4 23.27 19.31 32.18 4.02 NORMAL Travel speed to
field
5 0.02 0.00 429.00 1.88 ERLANG Travel time to
harvester
6 0.03 0.00 310.00 3.37 ERLANG Travel time to
full queue
7 0.06 0.00 102.30 1.00 ERLANG Harvester idle
time
8 7.89 6.05 18.00 3.20 Minimum observed
times1
9 3.00 5.00 7.00 Distance to

factory

 

lMinimum Observed times reading from left to right refer to minimum
loading time, minimum factory residence time, minimum travel time to
and from the harvester, and minimum harvester idle time, respectively.

92

 

C ....)
1

Record identification of field
tractor and hold in queue 5

l

7 Predict time of arrival of
full wagon in queue 3

 

 

 

 

Is
no. wagon/field tractor
combinations in queue 8
GT. 0

NO

Is
no. empty wagons in Yes Set field tractor ]
s.

 

 

field * idle, collect stat
LT. 1

 

 

No

Remove empty wagon and
schedule its arrival at

the harvester

1

Schedule start
of loading

 

    

   
       
 

Is a
road tractor and
two full wagons
available

 
 

 

J Dispatch a trip

RETURN g,/ I to the factory

 

 

 

Figure 5.5. Flowchart through subroutine NDLD

93

Since the loaded wagon spends some time travelling across the field
before it reaches the queue of full wagons, its time of arrival in the
queue is predicted by sampling the Erlang distribution of the travel
time to this queue using parameter set #6. This time is recorded as
attribute 4 (see Table 5.1).

If there is a field tractor and empty wagon combination waiting
in File 8, it is removed and set in motion towards the harvester. Its
time of arrival at the harvester is scheduled by sampling the Erlang
distribution for travel time to the harvester using parameter set #5.
If there is no such combination in File 8, the harvester is set idle.
The field tractor just released after depositing the full wagon is then
combined with an empty wagon (if there is one available) and a start
of loading event is initiated by setting the harvester busy.

If a road tractor and a complement of full wagons are available,
a trip is dispatched to the factory. Its time of arrival at the factory
is scheduled by sampling the Normal distribution for travel speed to
the factory and combining the deviate obtained with the known distance
to the factory to Obtain the travel time to the factory. Based on
numerous field observations, it takes an average of 3 minutes for a trip
to be prepared for departure from the field, so 3 minutes are added to
the current time TNOW when calculating the arrival time Of a transport
unit at the factory.

Subroutine ARHTR (Figure 5.6) simulates a field tractor and empty
wagon combination approaching the harvester while it is busy loading
another wagon. This occurs only if the number of field tractors in the

system is greater than one. Therefore, subroutine ARHTR is called only

94

if there are two or more field tractors assigned to a single harvester.
In this case, holding a field tractor and empty wagon combination in
Field 8 represents a situation where a wagon is being loaded and a
second wagon is waiting very close behind so that it can start receiv-
ing cane as soon as the first wagon becomes full, without the harvester
having to stop cutting.

Subroutine ARFCT (Figure 5.7) simulates the arrival of a transport
unit at the factory, predicts its residence time by sampling the Gamma
distribution of factory residence time using parameter set #3, and
schedules the departure of the emptied unit from the factory.

Subroutine DPFCT (Figure 5.8) predicts the arrival of an empty
transport unit at the field by sampling the Normal distribution of the
travel speed to the field using parameter set #4, and combining the
deviate Obtained with the known distance between field and factory to
arrive at a travel time. The returning transport unit is labelled as
having delivered its load (ATRIB (12) = 1.0) and information on it is
recorded in File 7.

Subroutine SHTDN (Figure 5.9) is used to close out all activities
at the end of a shift or a work day. All entities are returned to their
respective files and the future events file (File 1) is cleared in
preparation for a new start at the commencement of the next shift or
next day.

Subroutine STRTUP (Figure 5.10) is called to clear all of the GASP
IV statistical data collection arrays in preparationiknrthe new shift.
If full wagons and road tractors are available at this time, trips are
dispatched to the factory in accordance with the number of complete

transport units available.

C T“)

Put field tractor/empty wagon
combination into file 8

II
Cm)

Figure 5.6 Flowchart throxgh sibroutine ARHTR

 

 

 

 

 

‘ ARFCT )
Record time of arrival
at factory

Predict transport unit's
residence time at factory

 

 

 

 

Store residence time
data for histogram

 

 

« 1
CW 3

Figure 5.7 Flowchart through subroutine ARFCT

96

 

<:: Di[CT_:)

Record time of
departure from
factory

1

Label unit as having
delivered its load

I l #

[Schedule the arrival of \

 

 

 

 

unit back at the field
Store information on

road tractor processed
at the factory in file 7

1
(mm)

Figure 5.8 Flowchart through subroutine DPFCT

 

 

 

Cm)
1

[Set system status markerW]

 

 

    
     

  
 
 

Is
no. empty wagons
and field tractors
LE. 0?

   

 

Put empty wagons and
field tractors into their
respective files

Clear future events file
' (file 1) and sort out entities

 

 

 

 

[Set no. future events - N]

LIX-17

 

 

l_Calculate round-trip time]

 

 

Dark field tractors in file 21

 

 

v

Set road tractor idle
hold in file 4 and
collect statistics

CD + 0

Figure 5.9 Flowchart through subroutine SHTDN

 

 

 

98

 

[NR=NK+1

 

 

Yes

 

 

 

 

 

 

- Yes
LNNQ(9).I.E.oz—Jl

 

N

O
LN - we)?

 

 

 

 

 

 

    
   
 

 

Remove N entries from file 9 l
and store in file 1
quota satisfied? Yes __>

end-of-shift time?

 

Figure 5.9 (continued)

99

__.I

Schedule start T—schedule start
of next shift I of shift next day

a

 

 

 

 

Schedule and

Calculate amount of
next shift quota remaining

‘flv

 

 

 

of next shift

Schedule and
on next day

[ Schedule end time

 

 

 

 

Fix time of next
start up event

 

 

Fix time of next
shut down event

 

 

c m)

Figure 5.9 (continued)

100
C ...... D

Set sustem status marker
DWN - 0.0

1

Clear GASP IV data
storage arrays

1

set no. harvesters = N6 ]

 

 

 

 

 

 

 

 

 

 

Is

9
N6OEQoOo Yes {all ERROR (SD

IMO

Is .
NO. empty wagons = 0. Yes

 

 

 

No

 

from queue 2

1

Set time in queue I 0
and collect stats.

1

[ Remove N6 field tractors

 

L Remove N6 empty wagons ]

 

from queue 5

Figure 5.10 Flowchart through subroutine STRTUP

 

 

 

 

   
   

101

 

and collect stats.

1

[Schedule end of loading

l Set harvester busy

 

 

 

  
   
  
 
   
    
 

no.road tractors .LT.O?

 

 

 

Set no. full wagons = N

1

Calculate no. 2-wagon trains
available and set no. = K

 

 

 

 

 

no. 2-wagon trains .EQ.0.

Is

no. road tractors
.LT.

no. 2-wagon trains?

 

Call ERROR (6):)

 

 

 

Klano. road tractors
available

 

 

[Kl - no. 2-wagon trains available

 

 

Cb

Figure 5.10 (continued)

 

 

102

 

Remove Kl full wagons from queue 3,
set time in queue 3 = 0.0 and
collect statistics

l

Remove K1 road tractors
from queue 4

Fix time of departure of ]

 

 

 

 

 

 

 

a trip from the field

 

 

 

[Schedule time of arrival at factory]

 

 

Set road tractor busy
and.collect stats.

 

 

 

       

no. empty wagons LT.

0. field tractors Set no. empty

wagons = K

   

 

 

  

 

[Set no. field tractors-= K

e 1

Set K field tractor and empty
wagon combinations busy

_ 1

[Schedule arrival of the 1

 

 

 

combinations at harvester

Cm?“ 3

Figure 5.10 (continued)

 

 

 

103

Table 5.6. FIELDOP variables monitored by GASP IV subroutine COLCT

 

 

 

 

 

 

 

Variable D i ti
Symbol escr p on

RTRIP The total trip time of the road tractor from the field
to the factory and back.

TIQE The time spent by an empty wagon at the field waiting
to be loaded.

TIQF The time Spent by a full wagon at the field waiting to
be taken to the factory.

TIQW The time spent by a field tractor and wagon combination
waiting to be loaded when more than one field tractor
is used per harvester.

Table 5.7. FIELDOP variables monitored by GASP IV subroutine TIMST

Variable .

Symbol Entity Description
BUSHT (I) Harvester Utilization of harvester number I
defined by attribute 8.
BUSFT (J) Field tractor Utilization of field tractor number J
defined by attribute 9.
BUSRT (K) Road tractor Utilization of road tractor number R

defined by attribute 10.

 

104

5.4. The Factory Yard Simulation Model

5.4.1. The System Environment
The system environment for this model consists of the following

exogenous variables:

Number of cane harvesters in operation at any given time

Number of road tractors, field tractors and wagons operating

in locations from which the cane is diSpatched to the factory

The field conditions at these locations

- Road conditions along routes from the field to the factory.

All of these factors affect the total number of transport units dis—
patched to the factory which, in turn, affects the inter-arrival time
of transport units at the factory (see Figure 5.11 for causal loop

diagram).

5.4.2. Entities
The entities considered for this model are:
1. The weigh scale
2. The chopped cane unloader (crane)
3. The whole cane unloader (hoist)
4. The chopped cane transport units (2 wagons and 1 road
tractor)
5. The whole cane transport units (2 trailers and 1 road

tractor)

As was the case in the field operations model, Operating personnel of

transport units are considered to be integral parts of those units and,

105

seesaw was» muoOumm one How Emuwmfip eOOH Hmmnmo :Zwm ouanm

 

 

mumHos
mo
Hanan:

\‘I
i \ I

. manna /

. umHos

 

 

   

kHOuunm o
vmsoueeme
muHso menu

.383
mo .os

  

humvsnom seamen

    
    

  

mmHmOm

mo
Hones:

  
  

\ oSHu /
N HN>fiHHm a
. InousH .

a s
I uHcs

  

     

  

uouoem o
vacuumepr
muHsn ammo
commonu
mo .0:

     
  
  

 

  

GOGQHU

mo
Humane

    
  

 

106

as such, are not modelled individually. In addition, since a transport

unit is handled intact throughout the factory yard process, individual

wagons or trailers are not treated as entities.

5.4.3. Endogenous Variables

 

The endogenous variables generated within the system boundary are
as follows:
1. Time Spent in the weigh queue
2. Time spent by a unit in the chopped-cane unloader queue
(crane)
3. Time spent by a unit in the whole cane unloader queue

(hoist).

5.4.4. Output Variables

 

These include the endogenous variables listed above and the
following:

1. Average time a unit Spends in the factory yard subsystem

2. Total number of units through the system

3. Number of units processed by the scale server

4. Number of units processed by the crane unloader

5. Number of units processed by the hoist unloader

6. Maximum number of units in the scale queue

7. Minimum number of units in the chopped cane unloader

queue
8. Maximum number of units in the whole cane unloader queue

9. Utilization of the scale server

107

Table 5.8. List of NON-GASP variables used in FACYARD

 

 

 

Symbol Definition

XISYS Number in system

TISYS Time in system

XVW Number of units through yard

KJ Local assessment of type of unit

KL Local assessment of type of unloader
TIQ2 Time in queue for scale

TIQ3 Time in queue for crane (chopped cane)
TIQ4 Time in queue for hoist (whole cane)
BU82* Status of scale

BUS3* Status of crane (chopped cane unloader)
BUS4* Status of hoist (whole cane unloader)

 

*Values of 0.0 = free; 1.0 = busy for these variables.

108

Table 5.9. Files used in model FACYARD

 

 

 

File Number (I) Entities KKRNK (I) IINN (I)
1 Future events 4 1
2 Scale queue 4 3
3 Chopped cane unloader queue 4 3
4 Whole cane unloader queue 4 3

 

Table 5.10. Statistical distributions used for activity times in model

 

 

 

FACYARD
Distribution Parameter Random number
Event
type set stream

Inter-arrival Exponential 1 1

time

Weigh time Erlang 2 2
Unload time Erlang 3 3
(crane)

Unload time Erlang 4 3

(hoist)

 

109

Table 5.11. Attributes used in model FACYARD

 

 

 

Attribute number Description Value
1 Event time
2 Event code 1 - Arrival

2 - End of weighing
3 - End of unloading

3 Time into system

4 Time into weigh queue

5 Type of unit 1 - ChOpped cane
2 - Whole cane

6 Unloader type 1 - Chopped cane
2 - Whole cane

7 Unloader used 1 - Crane
2 — Hoist

8 Time into queue 3

(crane)
9 Time into queue 4

(hoist)

 

110

10. Utilization of the crane server

11. Utilization of the hoist server.

5.5. Factory Yard Simulation Program (FACYARD)

The system consists of an unlimited number Of cane transport units
which arrive randomly at the factory where they undergo two services:
weighing and unloading. Two types of transport units arrive: chopped-
cane wagons and whole-cane trailers. A transport unit consists of two
wagons or two trailers and a road tractor and has a nominal carrying
capacity of 10 tonnes. Current factory yard equipment consists of one
crane for tipping chopped-cane units directly onto the feeder table and
one hoist for unloading whole-cane and either stacking it or placing it

directly onto the feeder table.

5.5.1. General Flow through Model FACYARD

 

The flowchart for model FACYARD is presented in Figure 5.12. The
main program (user written) assigns appropriate values to the card
reader and printer units and initializes the variables declared in the
DIMENSION, COMMON and EQUIVALENCE statements, before making a call to
the GASP IV executive subroutine, GASP. Subroutine DATIN and INTLC
are then called from GASP to enter input data and establish initial
conditions and performance criteria for the simulation. A successful
return from DATIN is indicated by a print out of intermediate results
in which the input data and initial values established for the variables
are echoed. The executive routine GASP then calls subroutine EVNTS (IX)

from which calls are made to the individual user-written event subroutines

 

 

C m)

 

111

GASP IV SUB-PROGRAMS USED

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[ COMMON FUNCTIONS
DIMENSION
1 EQUIVALENCE ERLNG
INIT
RDR
PRNT
INTLC
# DATIN
/ ARIVAL
EVNTS ENDWGH
<:_ 'STOP ) '\\\\\\\\\\. ENDULD

 

 

 

 

 

 

* GASP IV SUBROUTINE
** USER SUBROUTINE

Figure 5.12.Flowchart through FACYARD

 

SUBROUTINES

DATIN
Tan
COLCT
FILEM
RMOVE
OTPUT*
ERROR

**

**

**

(USER WRITTEN)

112

ARIVAL, ENDWGH or ENDULD, based on the value of the argument IX. In
order to start the simulation, one arrival is scheduled for a chopped-
cane transport unit at simulation time 0.0 and one for a whole-cane
unit at time 0.5 minutes.

With IX - 1, subroutine ARIVAL is called to process the arriving
transport unit (Figure 5.13). The type of unit just arriving is re-
corded and the next arrival of this type is scheduled by generating a
random deviate from the Exponential distribution of inter-arrival times
using the distribution parameters stored in parameter sets #1 and #2.
The status of the scale is then checked. If free, it is set busy and
an end-Of-weighing event scheduled by generating a deviate from the
Erlang distribution of weigh times using parameter set #3. If the
scale is busy, the arrival is put into the queue to the weigh scale
(queue 2) and its time of entering the queue is recorded.

With IX = 2, subroutine ENDWGH is called to handle transport units
at the end of weighing (see Figure 5.14 for the flowchart of this rou-
tine). At the end Of weighing, the transport unit just weighed is sent
to either the crane or hoist unloader, depending on whether it is a
chopped- or whole-cane unit. If the unloader is free, an end-of-unload—
ing event is schedule by sampling the Erlang distribution using the
appropriate parameter set for the unloader service time (Table 5.14).
Otherwise, the weighed tranSport unit is put into the appropriate un-
loader queue and its time of entering the queue is recorded. The first
unit standing in the scale queue is then removed and an end-of—weighing
event scheduled for it. The time spent in the scale queue by this unit

is then calculated by a call to the GASP IV subroutine COLCT.

 

‘ SUBROUTINE ARIVAL >

 

Schedule next arrival

 

 

 

 

 

Add one to number in system
Collect statistics on number
in system

 

 

113

Yes

 

Is scale busy,

No

 

Set scale busy
Predict end—of-weighing

 

 

 

Set time in queue
Collect statistics on time
in queue

 

 

 

Put arrival in proper
queue. Record time in
queue

 

 

I.

 

 

 

:<:: RETURN ::)

Figure S.13.Flowchart Of the event ”arrival at factory"

 

114

< SUBROUTINE ENDWGH ‘:)

_Chopped cane Determine type Of cane 1 Whole cane
I Go to correct unloader I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Put arrival in Put arrival in
chopped cane queue ole cane queue
Set crane busy ’[Go to bring in Set hoist busy
Predict end-Of-service‘ next unit for Predict end-of-service
Collect statistics weighing Collect statistics

 

 

 

 

 

 

 

 

 

 

    
    
 

ERROR .g_____.Yes Queue2.1t.0.0

 

 

 

 

 

 

Remove lst entry
Yes_§ Set scale busy

Schedule end-of-wghng
Collect statistics

‘Queue2.gt.0.0

 

 

 

 

 

Set scale free _-_
Collect statistics RETURN

Figure 5.L4.Flowchart of the event ”end-of—weighing"

 

115

With IX = 3, subroutine ENDULD is called to handle units at the
end of unloading (see Figure 5.15 for flowchart). The type of tranSport
unit just unloaded is recorded and, since unloading is the last activity
performed on the unit, the time the unit spent in the system is calcu-
lated by a call to the GASP IV subroutine TIMST and the number of trans-
port units through the factory yard process is updated by one. The
first unit in the queue to the appropriate unloader is then removed and
an end-of-unloading event is scheduled for it. Subroutine COLCT is
then called to calculate the time spent by this unit in the unloader
queue.

The variables monitored by the GASP IV subroutines TIMST and
COLCT in program FACYARD are listed in Tables 5.12 and 5.13, respectively,

and a computer listing of the program is given in Appendix III.

5.5.2. Input Data for FACYARD

 

As mentioned earlier, the input data for FACYARD consists of the
statistical distribution parameters determined for the various activi-
ties of the Factory Yard system, using the data analysis program
(DATANAL) previously described. The input parameters used in the model
are given in Table 5.14. For the Erlang distribution, A represents
the scale factor and B the shape factor and the mean of the distribution
is given by A x B. The EXPONENTIAL distribution is a special form of
the ERLANG distribution in which the shape factor B is always equal to
1, so that A is, in fact, the mean of the distribution.

DATANAL not only outputs values for the scale and shape factors

of the distribution, but also generates a sorted list of the observed

116

 

<::_ SUBROUTINE ENDULD

 

V

 

 

Reduce number in system by 1

Collect statistics
Calculate time in system

 

 

 

V

 

 

Calculate number of
units through yard

 

 

_, Yes

Yes

 

Is

D

Yes

 

unit leavin
1\\\4¥fgj>//§I

 

ERROR

 

 

 

 

 

 

 

Ranove first entry
Schedule and of unloading

 

 

 

 

 

 

 

 

Nc

 

Set crane busy
Collect stats

 

 

 

 

Set hoist busy
Collect stats

No

 

 

V I

 

 

 

 

 

Calculate time
in queue 3

 

 

Calculate time‘
in queue 4

 

 

 

l

 

 

Set crane free

 

 

 

‘

 

 

 

Set hoist free

 

 

Collect stats

 

 

 

 

 

I I

C >

 

Collect stats

 

 

 

Figure 5415.Flowchart of the event "End—of—Unloading"

117

Table 5.12. FACYARD variables monitored by GASP IV subroutine TIMST

 

 

Variable Descri tion
Symbol P
XISYS The number of transport units in the
factory yard system at any given time.
BUSZ The utilization of the weigh scale
BUS3. The utilization of the chOpper-cane
unloader (crane)
BUS4 The utilization of the whole-cane

unloader (hoist)

 

118

Table 5.13. FACYARD variables monitored by GASP IV subroutine COLCT

 

 

 

Variable .
Symbol Description
TISYS The total time spent in the factory yard
system by a transport unit
TIQ2 The time spent by a transport unit in
the scale queue waiting to be weighed
TIQ3 The time spent by a chopped-cane transport
unit in the queue to the chopped-cane
unloader (crane) waiting to be unloaded
TIQ4 The time spent by a whole-cane tranSport

unit in the queue to the whole-crane

unloader (hoist) waiting to be unloaded.

 

119

 

 

nuoooHao amHo: oz<qmm H.m m.q w.H No.0 m

oooooeoo moons ozaamm m.~ H.m N.~ on.a s

oeoom ozaumw N.oH m.m No.0 om.o m

Ho>aooo oooouoaoez aaeezmzooxm a a.mm o.o o~.m N

Ho>aooo oooosoooooeo a<Hezmzomxm H o.om o.o oa.e H
.335 eoaommmooao m an: 5: < noowMMuom

 

 

Qm<wu<h oH new: msOHuoanume one

mumumamama unecH .qH.m OHan

120

data, thereby facilitating determination of the minimum and maximum
values. Specification of these minimum and maximum observed values in
the input data ensure the rejection of any random deviates, generated
during the execution of FACYARD, which may fall outside the desired

range.

CHAPTER 6

RESULTS AND DISCUSSION

6.1. Introduction

The standard output from GASP IV simulation models is in tabular
and/or graphical form. Statistics for time-varying variables (such
as the number of units through the system and the proportion of
time each server is busy) are collected as a standard function of
the GASP IV subroutine TIMST, provided that this subroutine is called
by the user at the appropriate times. Similarly, regular s.atistics
for time-persistent variables (such as the time spent in each queue
and the total time Spent by a unit in the system) are routinely
collected by appropriate calls to the GASP subroutine COLCT.

The GASP IV output function also provides detailed and summary
information on the entries in each file, a file being representative
of a Specific service queue in the real system. Statistics provided
on the files gives the maximum and minimum numbers of units in each
queue for the duration of the simulation, as well as the current

numbers of units in each queue at the end of the simulation period.
6.2. Results from the Factory Yard Simulation Model

Four alternative modes of factory yard operations were simulated

for Carrington Sugar Factory. The first one featured the yard in its

121

122

current configuration with a single weigh scale, a single queue to
this scale and two unloading facilities; a crane for tipping chOpped-
cane transport units and an overhead hoist for unloading whole cane.
The simulation time was 700 minutes, or the average length Of a
normal work day.

The second alternative was similar to the first, except that
the daily work period was extended from 11 1/2 hours (700 minutes)
to 16 hours (960 minutes).

The third alternative Simulated a yard configuration with two
weigh scales accepting cane from a Single arrival queue over the
normal work period of 700 minutes. Arriving cane transport units
proceed to either scale, depending on which one first became avail-
able.

Like alternative three, the fourth alternative simulated a con-
figuration with two weigh scales. In this case, however, each scale
had its own queue. Scale #1 was used exclusively for weighing chopped-
cane transport units and scale #2 for weighing whole-cane units.

In order to accommodate the additional scale in the case of
alternative 3, and the additional scale and queue in the case of
alternative 4, a number of changes were made in the coding of the
FACYARD subroutines ARIVAL and ENDWGH. Fortran listings of these
modified subroutines are given in Appendix III. The variables BUS5
and TIQS refer to the additional scale and queue respectively, and
the service time distribution of the second scale was assumed to be

identical to that of the first.

123

Table 6.1. Output from Model FACYARD for Current Configuration of
Carrington Factory (Simulation time - 700 minutes)

 

STATISTICS FOR VARIABLES BASED ON OBSERVATION

VARIABLE MEAN STD. DEV. CV MIN MAX OBS TIME IN
TISYS 73.98 25.61 0.35 6.05 119.40 210 System
TIQ2 64.71 25.52 0.38 0.00 113.80 212 Scale queue
TIQ3 0.30 0.77 2.61 0.00 4.61 83 Crane queue
TIQ4 0.24 0.58 2.43 0.00 2.61 128 Hoist queue

STATISTICS FOR TIME-PERSISTENT VARIABLES

VARIABLE MEAN STD. DEV. MIN MAX CUR. VALUE NAME
XISYS 23.81 7.02 OCOO 37.00 29.00 No. in system
BUSZ 1.00 0.06 0.00. 1.00 1.00 Scale

BUSB 0.41 0.49 0.00 1.00 0.00 Crane unloader
BUS4 0.51 0.50 0.00 1.00 1.00 Hoist unloader

ENTRIES IN FILE STORAGE AREA (SERVICE QUEUES)

VARIABLE MEAN STD . DEV . MAX CUR . NO . NAME

01 3.91 0.54 5.0 4.0 Future events
02 21.82 6.98 35.0 27.0 Scale queue
Q3 0.04 0.19 2.0 0.0’ Crane queue

Q4 0.04 0.21 1.0 0.01 Hoist queue

 

124

6.2.1. FACYARD Results for Current Yard Configuration

 

The output from model FACYARD for this configurationis shown in
Table 6.1. For purposes of model verification, values Obtained from
this simulation output were compared with actual, observed values for
selected system performance indicators. The comparison is presented in
Table 6.2 below:

Table 6.2. Comparison of observed and simulated values for selected
system performance indicators.

 

 

 

Performance Indicator Observed Simulated Z Difference
Mean factory residence times 67.86 73.98 8.27
Maximum factory residence times 128.41 119.40 7.55
Mean time in scale queue 63.27 67.41 6.14
Maximum time in scale queue 122.31 113.80 6.95
Mean scale service time 3.58 3.76 4.80
Maximun time in crane queue 5.01 4.61 7.80
Maximum time in hoist queue 2.86 2.61 9.20

 

In all cases the difference between the Observed and simulated
values of the system performance indicators was less than 10 percent,
verifying that the model adequately represented the real life system.

A study of Table 6.1 reveals that the main bottleneck of the factory
yard system, in its current configuration, is the weighing operation.

This is evidenced by the very high utilization factor (1.00) for the

scale, by the average time Spent by a unit waiting in the scale queue

125

(64.7 minutes) and by the length of the queue to the scale. The average
number of transport units in the queue to the scale at any time during
the day was 22; the minimum number of units in the queue was 5 and, at
the end of the working day, some 27 units were standing in line waiting
to be weighed. In contrast, the utilization factors of 0.41 for the
chopped-cane crane and 0.51 for the whole-cane hoist, indicate substan-
tial under—utilization of both of these unloaders. The total number of
transport units passing through the system daily was 210, of which 83

were chopped-cane units.

6.2.2. FACYARD Results for Extended Work Period

 

The output from the model for the extended work period is pre—
sented in Table 6.3. There was an increase in the total number of units
handled by the system to 289, 112 of which were chopped-cane units.

The scale was again busy 100 percent of the time and there were 37 trans-
port units awaiting service by the scale at the end of the day. Units
spent an average of 83 minutes in the system as a whole and an average

of 76.6 minutes in the scale queue. The chopped-cane unloader was util-
ized slightly less (39 percent of the time) and the whole-cane hoist

slightly more (53 percent of the time).

6.2.3. FACYARD Results for Configuration with l Weigh Queue and 2 Weigh
Scales

Table 6.4 shows the output from the model for this arrangement.
In this case a total of 236 transport units were serviced by the system,

141 whole-cane units and 95 chopped-cane units. Scale #1 was busy 55

126

Table 6.3. Output from model FACYARD for extended work period at

Carrington Factory (Simulation time = 960 minutes).

 

 

STATISTICS FOR VARIABLES BASED ON OBSERVATION

VARIABLE MEAN STD. DEV. CV MIN MAX OBS TIME IN

TISYS 82.83 28.20 0.34 0.05 143.70 289 System

TIQ2 76.63 28.55 0.37 0.00 138.20 291 Scale queue

TIQ3 0.27 0.74 2.74 0.00 4.61 112 Crane queue

TIQ4 0.28 0.60 2.15 0.00 2.61 178 Hoist queue
STATISTICS FOR TIME-PERSISTENT VARIABLES

VARIABLE MEAN STD. DEV. MIN MAX CUR. VALUE NAME

XISYS 27.53 8.84 0.00 45.00 39.00 NO. in system

BUSZ 1.00 0.05 0.00 1.00 1.00 Scale

BUS3 0.39 0.49 0.00 1.00 1.00 Crane unloader

BUS4 0.53 0.50 0.00 1.00 0.00 Hoist unloader

ENTRIES IN FILE STORAGE AREA

VARIABLE MEAN STD. DEV. MAX CUR. NO. NAME

Ql 3.91 0.53 5.0 4.0 Future events

Q2 2.53 8.81 43.0 37.0 Scale

03 0.04 0.18 2.0 0.0 Crane queue

Q4 0.05 0.22 1.0 0.0 Hoist queue

 

127

Table 6.4. Output from model FACYARD for configuration with 1 weigh

queue and 2 scales (simulation time - 700 minutes).

 

 

STATISTICS FOR VARIABLES VASED ON OBSERVATION

VARIABLE MEAN STD. DEV. CV MIN MAX OBS TIME IN

TISYS 9.00 3.54 0.34 3.85 21.32 236 System

TIQ2 0.93 1.57 1.69 0.00 9.15 239 Scales queue

TIQ3 1.63 2.89 1.77 0.00 11.06 95 Crane queue

TIQ4 1.75 2.39 1.36 0.00 10.99 143 Hoist queue

STATISTICS FOR TIME-PERSISTENT VARIABLES

VARIABLE MEAN STD. DEV. MIN MAX CUR. VALUE NAME

XISYS 3.07 1.96 0.00 9.00 3.00 No. in system

BUSZ 0.55 0.50 0.00 1.001 0.00 Scale #1

BUS3 0.47 0.50 0.00 1.00 1.00 Crane unloader

BUS4 0.57 0.49 0.00 1.00 1.00 Hoist unloader

BusS 0.65 0.48 0.00 1.00 1.00 Scale #2
ENTRIES IN FILE STORAGE AREA (SERVICE QUEUES)

VARIABLES MEAN STD. DEV. MAX CUR. NO. NAME

Q1 4.16 1.31 9.0 4.0 Future events

Q2 0.32 0.86 6.0 0.0 Scale queue

Q3 0.22 0.60 4.0 0.0 Crane queue

Q4 0.37 0.74 4.0 1.0 Hoist queue

 

128

percent of the time.and scale #2 65 percent of the time. The utilization
factors for the crane and hoist unloaders were 0.47 and 0.57, respectively.
The average time spent by a transport unit in the factory yard system
and in the queue to the scale both fell sharply to 9 minutes and 0.93

minites, respectively, and all transport units arriving at the factory

during the day were serviced.

6.2.4. FACYARD Results for Configuration with 2 Weigh Queues and 2 Weigh
Scales

 

Table 6.5 presents the results obtained for this configuration.
Some 236 transport units passed through the system daily. Of these, 95
were chopped-cane units. The average residence time at the factory for
a transport unit was 9 minutes, the average time spent by a chopped-cane
unit in the queue to scale #1 was 1.8 minutes; and the average time
spent by a whole-cane unit in the queve to scale #2 (95) was 1.2 minutes.
Scale #1 (the chopped-cane scale) was busy 45 percent Of the time and
scale #2 (the whole-cane scale) was busy 67 percent of the time. Utili-
zation factors for the:craneand hoist unloaders were 0.46 and 0.56,
respectively and, with this configuration, the factory yard system was
again able to service all arriving cane transport units within the normal
daily operating time of 700 minutes. In addition, at no time during the

day did the number of units in the system exceed nine.

6.3. Results from the Field Operations Simulation Model
The activities involved in the field operations system as modelled
are discussed in section 4.3 and the flow process of the operations are

illustrated in Figures 4.1 and 4.4. A computer simulation model of the

129

Table 6.5. Output from model FACYARD for configuration with 2 weigh

queues and 2 scales (simulation time - 700 minutes).

 

 

STATISTICS FOR VARIABLES BASED ON OBSERVATION

VARIABLE MEAN STD. DEV. CV MIN MAX OBS TIME IN

TISYS 9.61 4.24 0.44 3.85 24.00 236 System

TIQ2 1.82 2.96 1.63 0.00 11.75 95 Scale #1 queue

TIQ3 1.15 2.07 1.81 0.00 10.32 95 Crane queue

TIQ4 0.37 0.75 2.00 0.00 3.15 143 Hoist queue

TIQ5 3.10 3.75 1.21 0.00 14.90 144 Scale #2 queue

STATISTICS FOR TIME PERSISTENT VARIABLES

VARIABLE MEAN STD. DEV. MIN MAX CUR. VALUE NAME

XISYS 3.28 1.93 0.00 4.00 3.00 No. in system

BUSZ 0.45 0.50 0.00 1.00 0.00 Scale #1

BUS3 0.46 0.50 0.00 1.00 1.00 Crane unloader

BUS4 0.56 0.50 0.00 1.00 1.00 Hoist unloader

BUSS 0.67 0.47 0.00 1.00 0.00 Scale #2
ENTRIES IN FILE STORAGE AREA (SERVICE QUEUES)

VARIABLE MEAN STD. DEV. MAX CUR. NO. NAME

Q1 4.16 1.08 6.00 4.0 Future events

Q2 0.25 0.65 4.0 0.0 Scale #1 queue

Q3 0.16 0.42 2.0 0.0 Crane queue

Q4 0.08 0.28 2.0 1.0 Hoist queue

Q5 0.64 1.05 6.0 0.0 Scale #2 queue

 

130

system was develOped (in GASP IV simulation language) and validated
using an equipment combination of 10 wagons, 1 harvester, 2 field
tractors and 2 road tractors. The model was then used to Simulate oper-

ation of the system with 7 other field equipment combinations.

6.3.1. Validation of Model FIELDOP

 

In order to verify the model, values obtained from the simulation
(with the equipment combination of 10 wagons, 1 harvester, 2 field
tractors and 2 road tractors) for selected system performance indicators
were compared with values actually observed in the field for a similar
set of equipment. The output from program FIELDOP for this simulation
is shown in Table 6.6, and the comparison of simulated and observed
values is given in Table 6.7.

To further validate the model, one of the key system performance
indicators, round-trip time, was subjected to the non-parametric
Wilcoxon-Mann-Whitney Rank Sum test to ascertain whether the simulated
and observed values for this indicator had the same statistical distri—
bution. The test was conducted in accordance with the method given by
Steel and Torrie, 1980. The details are as follows:

1. The values for both the observed and simulated samples

of data for round-trip time were ranked together in
ascending order.

2. The ranks of the values from the smaller sample were
then added to give a total, T.

3. The rank sum T1 that would be obtained if the values
had been ranked in descending order was then cal-
culated according to the equation I = n (n + n2
+ l)-T; where n and n are the number 0 va ues
in the smaller and larger sample, respectively.

131

OHHnm wanm>m .wnHauoz n .2.m ..z.<

Ne uouomuu noon «0 nOHumNHHHuD I Nam:

He wouumau woos mo SOHueuHHHu: u Hem:
ooum>HHov mono mo mossoe u amazzoe
moowes HHom mo amnanz u uzmz

enemas homes no nooanz I 0302

ustm no one an uHmswuu OH meHaH u HHmMH
OQHuu unease Heuoa u mymmw

Ne nouomua vHOHm mo SOHumuHHHus u Nam:
He souumuu OHOHm mo OOHumuHHHua u Hem:
umumm>umn mo sOHumNHHHus u an:

ozone sowm3 HHom sH eaHH moHH

moose momma human :H oaHH MOHH

maHu eHuu mmmum>< I Hmmam

OMHnm mo maez u mmmz

woman: man I w<oz

 

ON 0« 0« 0m 0m 0m 0m
N N H N N m N
N m N w 0 N 0
H.H 0 H 0 H 0 H
w « « m m m m
mm.0 0.H 0.0 0.0 0.H 0.0 0.0
00.0 m.0 0.H 0.0 0.0 0.0 0.H
0m.0 m.0 «.0 m.0 m.0 m.0 m.0
um.0 m.0 m.0 m.0 0.0 m.0 m.0
H«.0 «.0 «.0 «.0 «.0 «.0 «.0
m.HN 0.0N 0.HN 0.0H N.0H N.NN m.HN
0.m0 0.0m H.Ho 5.0m «.HN 0.00 m.«m
N.0NH m.HOH «.nHH 0.mmH N.H«H 0.NNH m.NH
24m: .z.m .z.< .Z.m .Z.< .z.m .z.<
VHH<0 0 m «

q

MONOOOHHQOQNO
20H

° <fWH<fQ<fOO
z oo o
D-I

m

O
\T

[\NanOQQO‘
hO©OOOOO<¥N¢N

=3
.¢

ammzzoe
usez
ozmz
HHmme
memes
«we:
Hem:
Nee:
Hem:
em:

N eoHe
a mOHe
HH amma<

m
In
on
\T

HONC

$1

[\OlfiOOOOO

MQO‘OOOHHMHONO

HMI-I

MWNOOOOOMONMO

I-in
...;

0N000000M0¢ON0
:NN

\ONN¢\OMO\O\
v-lel-nl‘fi'oc
O
€¢HQO¢$O
mH©€©¢NG
O

H

.Z.m

=2
'<
z:
n.
>=I
<2

mmmz
H<Qz

N
Fl

 

 

SOHumanEoo unoaeano "NuNuHuoH How monHmHm Hupoa scum uneuao .o.o mHan

132

Table 6.7. Comparison of observed and simulated values of selected
system performance indicators for field operations.

 

 

Performance Indicator Observed1 Simulated1 Z Difference

 

Number of trips completed
per day 7 8 14.3

Average round-trip time 109.8 120.2 9.5

Utilization factor for
harvester 0.45 0.41 9.8

Utilization factor for
field tractor #1 0.65 0.57 12.3

Utilization factor for
field tractor #2 0.44 0.36 18.2

Utilization factor for
road tractor #1 0.80 0.89 11.3

Utilization factor for
road tractor #2 0.81 0.93 14.9

 

1Average of 6 consecutive days or the equivalent of one week's operation.

2One trip = 2 wagons = 10 tonnes of cane (nominal capacity).

133

l
4. The rank sum T for the smaller sample was then
compared with tabulated values at appropriate
probability levels.
The null hypothesis and the alternative tested were as follows:
Ho: Observed and Simulated values of round-trip times
for cane transport units over a distance of 3

kilometers between field and factory have the
same distribution.

H : The observed and simulated valued of round-trip
times have different distributions.
The data used for the test was as follows:
Simulated: (n1 - 12)
95.4; 98.3; 97.1; 103.6; 96.7; 110.4; 102.3; 106.6; 125.6; 101.4;

93.7; 138.5.

Observed: (n2 = 15)
105.4; 96.9; 81.1; 113.6; 142.1; 99.5; 111.3; 97.2; 91.1; 104.5;

109.7; 93.6; 125.9; 117.8; 112.0.

Ranked data: 0 = observed; S a simulated.

81.1(1); 91.1(2); 93.6(3); 93.7(4k 95.4(5);96.7(6); 96.9(7);
O O O S S S 0

97.1(8); 97.2(9); 98.3(10); 99.5(11); 101.4(12); 102.3(13);
S O S O S S

103.6(14); 104.5(15); 105.4(16); 106.6(17); 109.7(18);
S O O S O

110.4(19); 111.3(20); 112.0(21); 113.6(22); 117.8(23);
S O O O O

125.6(24); 125.9(25); 138.5(26); 142.1(27)
S O S O

4+5+6+8+10+12+l3+l4+l7+l9+24+26

F-l
II

158

134

T1 n 12(12 + 15 + l) - 158
= 178
TO 05 8 127
° Source: Table A.l9 Steel and Torrie (1980), p. 614.
TO 01 - 115

Based on the above test results, the null hypothesis HO was accepted
and it was concluded that the observed and simulated values of round-

trip times for cane transport units were similarly distributed.

6.3.2 Output from FIELDOP for Different Equipment Combinations

 

Following validation of the field Operations model, seven different
field equipment combinations (see Table 6.8) were simulated for travel
distances between the field and factory of 3, 5 and 7 kilometers, and
for transport unit factory yard residence times of 73.98 minutes and 9.0
minutes. These two times represented the mean factory residence times
obtained from program FACYARD for yard configurations having one weigh
scale and two weigh scales, respectively. The results of the simulations
are presented graphically in Figures 6.1 through 6.6. In the discussion
that follows (and on the figures) each equipment combination is identi-
fied by four numbers which sequentially represent the numbers of wagons,
harvesters, field tractors and road tractors assigned to a particular
operation. For example, the notation 10 : 1 : l : 2 indicates an
equipment combination consisting of 10 wagons, 1 harvester, 1 field
tractor and 2 road tractors.

Figure 6.1 shows daily cane delivery plotted against the number of

wagons assigned to the field system, for two different factory yard

135

Table 6.8. Equipment combinations simulated for field operations.

 

 

 

Field Road
wagons Harvesters Tractors Tractors
8 l l 2
10 1 l 2
12 l 1 2
14 l 1 2
8 1 2 2
10 1 2 2
12 1 2 2

l4 1 2 2

 

 

136

Mean factory residence time = 73.98 minutes

 

 

 

 

 

 

E 160-
c>
53
22
g“ 120-
5' 2 field tractors
S
(U 80'- 1 field tractor
V
p3
H
g /
40 l I . I 1 I
6 8 10 12 14
Mean factory residence time = 9.0 minutes

200' 2 field tractors
1'5 o

16 '-
E
S
g 1 field tractor
I-I
a:
c:
0 80'
S
H
.<
c:

40 I J I 1 I

6 8 10 12 14

NUMBER OF WAGONS

Figure 6.1 Simulated daily cane delivery vs. number of wagons

in system

137

residence times. Generally, cane delivery increased as the number of
wagons increased and outfits with two field tractors delivered 10 to
28 percent more cane than those with one field tractor. For the shorter
factory yard residence time, increasing the number of wagons caused a
large increase in the daily cane delivery to the factory. Systems with
less than 10 wagons show a disproportionately lower output than systems
with 10 or more wagons.
Figures 6.2 and 6.3 Show daily cane delivery versus travel distance
between the field and the factory. Indications are that:
1. Generally, the amount of cane delivered per day decreased
as the travel distance increased, and the decrease be-
came more pronounced as the distance got beyond 5 kilo-
meters and the number of wagons became less than ten.
2. For a mean factory yard retention time of 9 minutes,
the amount of cane delivered by systems with 2 field
tractors and 12 or more wagons was unaffected by in-
creases in travel distance until a distance of 5 kilo-
meters was exceeded.
The affect of travel distance on harvester utilization is presented
in Figures 6.4 and 6.5. These graphs are somewhat Similar in trend to
those obtained when cane delivery was plotted against travel distance.

The following observations were made:

1. Harvester utilization decreased as travel distance
increased.

2. For the mean transport unit factory residence time
of 74 minutes, harvester utilization varied from
52 percent (with a 14 : l : 2 : 2 equipment combin-
ation) to 38 percent, with a 8 : 1 : 2 : 2 combina—
tion.

3. Where the mean factory yard residence time was 9
minutes, harvester utilization varied from 78 to 71
percent when two field tractors were used, and from
65 to 50 percent when one field tractor was used.

138

 

 

 

 

 

 

14:1
32"
\ 12:1: -
160- °
:9.
ga .
hi .-
c:
g 8:1: °
5 80F
H
E
l l J L J
402 3 4 5 6 7
ZOOP
E
0—
,. 14:1: :
a:
E
g 10:1: :
t:
:3 8()" 8:1: :
H
3
l l J l J
402 a 4 5 6 7

TRAVEL DISTANCE (KILOMETRES)

Figure 6.2 Simulated daily cane delivery vs. travel distance to
factory for mean factory residence time of 9.0 minutes

N

139

 

 

 

 

 

 

 

 

 

A160-
In
E
E:
L—
53120
E 14:1:2:2
a
0 12:1:2:2
g 80-
U a 10 13232
8.1:2:2
S
2 l l l l I
D
" 402 3 4 5 6 7
A120F-
VJ
a:
z
z
o
5 \
80-
E \ ‘ 14:1:1:2
E. 12:1:1:2
51' 10:1:1:2
0 4o 8:1:1:2
N I—
i
0
E
3‘ o 1 l l 1 _J
2 3 4 5 6 7

TRAVEL DISTANCE (KILOMETERS)

Figure 6.3 Simulated daily cane deli-very vs. travel distance to factory
for mean factory residence time of 73.98 minutes.

HARVESTER UTILIZATION (PERCENT)

HARVESTER UTILIZATION (PERCENT)

Figure 6.4

100

80

60

 

12:1:2:
T“ 10:1:2:
4C) —_‘ :2:
l I l I J
202 3 4 5 6 7
80

60

4O

20

140

 

 

 

 

 

‘r-———______________‘~_—__——‘~—_—‘————_ 14:1:1:
12:1: '

 

 

 

 

 

\\ ‘_ 10:1:
‘“‘ 8:1‘1

l I I J J

3 4 5 6 7

TRAVEL DISTANCE (KILOMETRES)

Harvester utilization vs. travel distance to factory
for mean factory residence time of 9.0 minutes

l—i t—l
O. 0

NM

141

 

 

 

 

 

 

r
'2
I‘ll __
E
E: 12:1:2 2
g \
:51 ~10122
H
:4
H — \
S 8122
M
N
[—1
U)
m
>
g 1 1 l J A

 

TRAVEL DISTANCE (KILOMETRES)

Figure 6.5 Harvester utilization vs. travel distance to factory
for mean factory residence time Of 73.98 minutes

142

Road tractor utilization versus travel distance is shown in Table
‘6.6. The graphs reveal that for a mean transport unit factory residence
time of 74 minutes, all field equipment combinations resulted in high
levels of utilization (90 - 100%) for both road tractors. Conversely,
for the smaller factory residence time of 9 minutes, road tractor
utilization tended to vary with the amount of cane delivered to the
factory. Equipment combinations with one field tractor recorded 20 to
25 percent lower road tractor utilization than those with two field
tractors.

The high level of road tractor utilization recorded for the
factory yard configuration with one weigh scale reflect the fact that,
with this set up, a transport unit spent an average of 65 minutes in
the scale queue awaiting the weighing service (see Table 6.1) during

which time the road tractor, though non-productive, was in effect busy.

6.4 General Discussion

 

The results presented above indicate that an increase in the total
number of cane transport units serviced by the factory yard system can be
achieved by either extending the daily cane receiving period or adding a
second weighing facility to the system. There are a number of implications

associated with either policy.

6.4.1 Implications of an Extended WOrk Period
An extension of the daily cane receiving period at the factory
would necessitate an equivalent extension of the daily work period for

field operations. This implies harvesting and transportation of cane

143

Mean factory residence time = 73.98 minutes

100t-

 

 

 

 

 

 

 

 

 

/"f.'
7
80-
60P-
E
F: l J l 1 J
g 4O2 3 4 5 6 7
E:
2:
C)
El.“
51
S
H
e:
:a
a:
8
9’ Mean factory residence time = 9.0 minutes
2 100,. ~
3
M
80-
#—"."—'______,._.
60- ’5
4c) 1 41 I I J
2 3 4 5 6 7

TRAVEL DISTANCE (KILOMETRES)

Figure 6.6 Road tractor utilization vs. travel distance
to factory

HI—IH

G>c>haa~
HI—H—II—I

NNNN
00......
NNNN

14:1:2:2
12:1:2:2

10:1:2:2

8:1:2:2

144

during night-time hours, a change likely to be strongly opposed by workers
and their trade union representatives. If accepted, the extension would
bring about an immediate and significant increase in costs to the sugar
industry, in the form of 'over-time' wages. Eventually, however, this
increase should be offset by a contraction in the overall length of the
cane harvesting season.

Sugar factories in Barbados normally grind cane continuously through-
out the day for Six consecutive days before shutting down for cleaning and
maintenance. With a rated grinding capacity of 90 tonnes per hour, Carring-
ton Factory requires a minimum of 2,160 tonnes of cane per day to avoid
losing milling time due to lack of cane. Given the current factory yard
configuration of one weigh scale and an 11 1/2 hour work period, a maximum
of 210 transport units can be handled in a normal work day. With a nominal
transport unit capacity of 10 tonnes, this works out to a daily cane deli-
very of 2,100 tonnes, 60 tonnes short of the minimum requirement, and the
factory is likely to lose an average of one hour per day due to lack of
cane (see Table 6.1). When the work period is extended to 16 hours a day,
however, a total of 289 transport units can be processed at the factory
yard (Table 6.3) making an additional 790 tonnes of cane available to the
mill and eliminating mill lost time due to "grinding out".

If only chopped-cane transport units were received during the work
extension period, a minimum of 29 additional transport units would be
unloaded at the factory, yielding an increase in daily cane delivery of
290 tonnes, which would also be enough to eliminate mill lost time due
to lack of cane.

With the current arrival frequencies of cane transport units at the

factory and a single weighing facility, extension of the daily work

145

period is unlikely to have any positive effect on the efficiency of
utilization of the mechanical harvesting equipment in the field. In
fact, a comparison of Tables 6.1 and 6.3 reveals that the mean factory
yard residence for transport units actually increased from 73.98 to

82.83 minutes when the work period was extended.

6.4.2 Implications Associated with Two-Scale Configurations

 

The results presented in Tables 6.4 and 6.5 indicate that if either
of the two-scale configurations were implemented, the factory yard system
would be able to handle all the cane transport units currently dispatched
to it within the normal daily work period. Furthermore, since the utili-
zation factors for the service facilities (scales and unloaders) were all
less than 0.65, it is evident that, with two scales, the factory yard sys-
tem would be able to handle considerably more units than its current
allocation.

The existing factory yard layout provides only 30 metres between
the scale and the unloaders for queuing of transport units leaving the
scale. This restricts unloader queue lengths to two transport units and
often causes a back-up of weighed units on to the scale platform, result-
ing in stoppages Of the weighing process and increased traffic congestion
in the yard. The current queue positions also encroach on the area to
the north end of the feeder table, which should be used for stacking
whole cane.

With two scales in Operation, this problem is likely to be accent-
uated since transport units will be coming out of the weighing process

at about twice the current rate. The solution to the problem lies in

146

apprOpriate location of the second scale, together with re-routing of
the factory yard traffic and/or relocation of some of the existing fac-
tory yard facilities. Two alternative proposals aimed at achieving this

are presented in the following section.

6.4.3 Proposals for Re-organization of Carrington Factory Yard

 

Figure 6.7 shows a plan of the current layout of Carrington Factory
Yard and the path of cane transport units through the yard. Arriving
units travel along the western and southern borders of the general fac-
tory area and enter the factory yard at its south end through gateway
A-A. They then continue through a narrow section (10 metres) of paved
yard on to the weigh scale located near the north end of the work shop
building and, on leaving the scale, make a tight S-turn to get under
the boom of the crane or hoist unloader.
One proposed re-organization of the factory yard is presented in
Figure 6.8. Specific recommendations are as follows:
1. Remove the existing scale and relocate it, along with the
second scale to form a 2-scale parallel facility, at a point
B adjacent to the south border of the factory area as shown.
2. Let chopped-cane transport units, on leaving either scale,
proceed along the path shown through gateway A—A and on to
the crane unloader.
3. Let whole-cane units, on leaving the scale area, travel
behind the office and workshop buildings along the path
indicated to the hoist unloader.
4. Let empty transport units to be tared travel around the
main factory building back to the scale area as Shown. As
is currently done, units to be tared should be given priority
at the scales.

If adopted, the above proposal would provide more than enough space for

queues to be formed at all service facilities and would significantly

147

 

a mH ou so H ”onom

mHmoz I

/

one» zaouomm nouwsHuumo nwnounu mOHOHno>
uuoemcmuu menu «0 name assayed use .mo uSOAMH unouuoo N.o ouanm

 

~AHOUUWH Mandi/8H mUflgI
/ \ mflfiuwu How mUHGDI.II:.II

l...\\ muHsn msnOImHonz moanms

 

 

muHsn mnmulvoeeonu mop—363'

wsHumunm mann poveoqn.I.I.II

148

          
    

E mH On Eu H "OHeum

chow zuouomm GOumcHuumu mo nOHu

Emoz A||+Il

\ /

HcmeOIOH you He Humoaoum 0.0 ouanm

.33 new mnHwaH manal
waHumu wow muHsDI.II.I.

muHsp mmmOImHonz noanmB

man: annulvoaaonu omanw3.IIIIIII

wcHuoumu muHap munmoqn.l III

149

reduce traffic congestion in the central area of the factory yard.
Implementation of recommendation #3 above would necessitate construction
of a new road passing behind the office and workshop buildings. This
should be a paved road 12—15 metres in width. A gravel surface bound
with coal tar or other low-cost bituminous material should be adequate.
Proposal two, presented in Figure 6.9, is an attempt to avoid
having to relocate the existing weigh scale. The recommendations are
as follows:

1. Locate the additional scale (scale #2) along the southern
border of the factory area as in the previous proposal.

2. Separate arrivals into two queues, one for chopped-cane
transport units and one for whole-cane units.

3. Weigh chopped-cane units at scale #2 and allow them to
proceed, via gateway AHA, to the crane unloader.

4. Let whole-cane units by-pass scale #2 and proceed to scale

#1 via a new road passing behind the office and workshop

buildings. These units will now make their final approach

to the scale in a direction opposing that of the current

approach (i.e., from north to south).
If adopted, proposal two would increase the yard area available for
stacking whole cane (as would proposal one) and would provide additional
queuing Space for chopped-cane transport units. However, the area avail-
able for formation of the queue to the whole-cane hoist will remain
virtually unchanged and some backing up of weighed vehicles on to scale

#1 may still occur, particularly if there is a break in operation of the

hoist.

150

EA: oquH

Home.» .Couumm couwsHuumu mo COHumNHameOIou wow Ne Hemonoum 0.0 muanm

  

museum waH>me muHODI'II
wsHHMu now muH:DI.II.|
"OHmum / I \ 3ch oneOIOHOLB nostmfi

mann undulwomoono 025.3311
3.552

I

wnHuouao muHsn Hooves I I I I.

151

6.5 Economic Analysis

 

6.5.1. Calculation of Harvestingenachinery Costs

 

The cost of mechanical harvesting of sugar cane can be calculated
on the basis of the standard machinery cost accounting methods outlined
in the Agricultural Engineers Yearbook (ASAE, 1982). Component machin-
ery costs are defined as follows:

1. Fixed costs: Costs which do not depend directly on the

amount of machine use, such as depreciation, interest on
investment, taxes, insurance and storage, and which are
charged regardless of the extent of machine use.

2. Operating costs: Costs which depend directly on the

amount of machine use; namely, repairs, maintenance,
fuel and oil.

3. Labour costs: Costs which are directly associated with

machinery operators and any auxillary operating personnel.

4. Total cost: The sum of the fixed, operating and labour

COStS .

The total system fixed cost (TSFC) is given by summing the fixed
costs of the component machines involved in the harvesting operation,
as illustrated in the following equation:

TSFC = l/HCDR [(NHT x HTFC) + (NFT x FTFC) + (NRT x RTFC)

+ (NWG x WGFC)]. . . . . . . . . . . . . . . . . . . . (1)

where:

TSFC total system fixed cost ($/tonne)

NHT number of harvesters in system

NFT

NRT

NWG

HCDR

HTFC

FTFC

RTFC

WGFC

152

number of field tractors in system

number of road tractors in system

number of wagons in system

hourly cane delivery rate from system (tonnes/hr)
harvester fixed cost per hour

field tractor fixed cost per hour

road tractor fixed cost per hour

wagon fixed cost per hour

The total system operating cost is given by:

where:

TSOC -- l/HCDR [(UHT x HTOC) +FTOC (um + UFT2)
+RTOC(URT1 +URT2) + (ch xWGOC)]. . . . . . . . . . (2)
UHT = utilization factor for the harvester
UFTl = utilization factor for field tractor #1
UFTZ = utilization factor for field tractor #2
URTl = utilization factor for road tractor #1
URTZ = utilization factor for road tractor #2
HTOC = harvester Operating cost per hour ($/hr)
FTOC = field tractor operating cost per hour ($/hr)
RTOC = road tractor operating cost per hour ($/hr)
TSOC = total system operating cost ($/tonne)
Total system labour cost is given by equation (3) as follows:
TSLC - [(NHT x HTLC) + (NFT x FTLC) + (NRT x RTLC)]. . . . . . (3)

where:

TSLC = total system labour cost (S/tonne)

HTLC = harvester labour cost ($/hr)

153

FTLC = field tractor labour cost ($/hr)

RTLC = road tractor labour cost ($/hr)

The above equations formed the basis of a Fortran program which com-
puted the component and total costs for the eight field equipment combin-
ations simulated. The cost structure assumed for these computations is
presented in Table 6.9. Costs are based on 1982 figures and equipment
utilization factors were obtained from the simulation results previously
presented in Figures 6.1 through 6.6. Output from the cost program in-
cludes the total system cost per tonne and the total amount of cane
harvested per year for each equipment combination. In general, the cost
per tonne decreases as annual output increases. The details are shown
in Tables 6.10 through 6.13. The least cost combination consists of 2

field tractors, l4 wagons and 2 scales at the factory.

6.5.2. Sensitivity of System Cost and Output to Various System Parameters
The response of cost per tonne and annual cane delivery to successive
increments of two in the number of wagons is shown in Table 6.14. In
all cases, incrementing the number of wagons by two resulted in a decrease
in the cost per tonne and an increase in the annual cane delivery. In
addition, the response to the last increment (from 12 to 14 wagons) was
smaller than the responses generated by the first two increments. Gener-
ally, however, the response was neither linear nor uniform.
Table 6.15 shows the response of cost per tonne of cane harvested to
the addition of a second field tractor or weigh scale to the system, and
Table 6.16 shows the effect of these changes on annual cane delivery from

the field system. The figures indicate that the addition of a second

154

mN.H Hm.HH oo.m n¢.moa .mm\mamou nmme A<HOH

osN.H ~m~.HH ooh.m om~.HoH mumoo amass Hmaccm Hmuoa
NmH oeo.a cam mmw.a osum>
asap :wuufius Nm 9 uwuaonm a moamuamcH
mom qu.m omw.a mum.ma maam> caov couuauz so Nu e ammumuaH
oow omn.m coo.o coo.mm . Aummm\mv Ammasmmm ouou mo waam> wwm>ammv
Awesome mafia unwfimuumv aowumwompnmn

memou nmme
ooo.w ooo.~o ooo.wq ooo.omq A.mvm my mofiua compound HmHuHcH
0H m w o Amummmv mMHH magnoma voumaaumm
com cos com com Amazonv .mp5 0H so mama mauxuoa s as“:
mxwmz ea I commmm wcaumm>umn mo nuwcmq
ma «a as Aua\mmuuHHV coauaasmaou Haas
moccou m 33 cm . 33 on u:\mm:c0u mm mufiomamo vmumm

monH<UHmHummm mszU<z

 

zoo<3 MOHU<MH m0H0<MH Mmhmm>m<m
mz<o Q<Qm QAMHm

 

 

.Aumaaou m.o .m.: u umHHoe H .mumv .mmuamau Nmma co
momma memo umwSm mo wcfiumm>ums mcfinaoo pom bosomwm musuoauum umoo can maoaumoHMfioQO >umcfisomz .m.o manmh

155

 

on.o

mama com
um>o vmmuam Nw.qw ® mu:
N: All Homnum>o umw>umn
umoa mmv\mu: m 9 mzmv N +

mmmu em you zmv\uao: H

om.o mm.N
vowuma mono mop coo
um>o powwow un\mm.qw um
Hamsum>o uww>umn umoa

smc\mun w e mama OH +
mmmv om How zmv\mu=on q

usonma w.owcm£omz

 

II II II cm.m Aun\mw.~m e “mumm>umn
Hon NV Hmccomuma wcwumumno humaaaxa<
II Hm.m Hw.m mm.m momma m.u0umummo
mHmoo abom<q
mw.o oh.mH NH.NH mm.o~a m=\mHmou quH<mmmo A<Hoe
mw.o so.m mN.s NH.mm Aun\wv Amuufl umumaaumm um>o
umoo Hmfluaow Noav mocmcmuafima mam ufimamm
I- ms.o mm.o m~.~ Aun\mv «puss uma mm.~w a ago
II o~.m mm.m wq.oq Aua\.mvm my muufia pom um.ow e doom
mHmoo ozHH<mmmo
zoo<3 moau<ma moao<ma Mmymm>m¢m
mz¢o Q<Qm namHm

 

 

.v.u:oo .a.o manmw

Table 6.10. Estimated cost of combine harvesting of cane for a field
(One factory yard scale)

system with 2 field tractors.

 

 

No. of Wagons,

 

Variable 8 10 12 14
TRAVEL DISTANCE = 3 km
System fixed cost/tonne 22.48 19.99 13.54 13.76
System operating cost/tonne 13.09 12.51 8.79 9.49
System labour cost/tonne 5.59 4.89 3.26 3.26
TOTAL SYSTEM COST/TONNE 41.16 37.39 25.59 26.51
CANE DELIVERY (TONNES/YEAR) 6720 7680 11520 11520
TRAVEL DISTANCE = 5 km
System fixed cost/tonne 26.22 22.84 14.77 13.76
System operating cost/tonne 15.58 14.02 9.72 9.65
System labour cost/tonne 6.52 5.59 3.56 3.26
TOTAL SYSTEM COST/TONNE 48.32 42.45 28.05 26.67
CANE DELIVERY (TONNES/YEAR) 5760 6720 10560 11520
TRAVEL DISTANCE = 7 km
System fixed cost/tonne 26.22 22.84 18.05 18.34
System operating cost/tonne 14.67 13.88 11.60 12.04
System labour cost/tonne 6.52 5.59 4.35 4.35
TOTAL SYSTEM COST/TONNE 47.41 42.32 34.00 34.73
CANE DELIVERY (TONNES/YEAR) 5760 6720 8640 8640

157

Table 6.11. Estimated cost of combine harvesting of cane for a field
system with 1 field tractor.

(One factory yard scale)

 

 

No. of Wagons

 

Variable 8 10 12 14
TRAVEL DISTANCE 8 3 km
System fixed cost/tonne 21.18 18.86 17.05 15.60
System operating cost/tonne 12.07 11.91 11.13 10.55
System labour cost/tonne 5.59 4.89 4.35 3.91
TOTAL SYSTEM COST/TONNE 38.84 35.66 32.53 30.06
CANE DELIVERY (TONNES/YEAR) 6720 7680 8640 9600
TRAVEL DISTANCE = 5 km
System fixed cost/tonne 29.65 25.14 21.92 19.50
System operating cost/tonne 16.07 14.07 12.86 12.38
System labour cost/tonne 7.83 6.62 5.59 4.89
TOTAL SYSTEM COST/TONNE 53.55 45.73 40.38 36.77
CANE DELIVERY (TONNES/YEAR) 4800 5760 6720 7680
TRAVEL DISTANCE = 7 km
System fixed cost/tonne 29.65 30.17 25.57 22.29
System operating cost/tonne 15.34 16.63 14.10 12.77
System labour cost/tonne 7.83 7.83 6.52 5.59
TOTAL SYSTEM COST/TONNE 52.82 54.62 46.19 40.65
CANE DELIVERY (TONNES/YEAR) 4800 4800 5760 6720

 

Table 6.12. Estimated cost of combine harvesting of cane for a field
(Two factory yard scales)

system with 2 field tractors.

 

 

No. of Wagons

 

Variable 8 10 12 14
TRAVEL DISTANCE = 3 km
System fixed cost/tonne 13.11 10.66 9.56 8.25
System operating cost/tonne 7.07 6.51 7.21 7.04
System labour cost/tonne 3.26 2.61 2.30 1.96
TOTAL SYSTEM COST/TONNE 23.44 19.78 19.07 17.25
CANE DELIVERY (TONNES/YEAR) 11520 14400 16320 19200
TRAVEL DISTANCE = 5 km
System fixed cost/tonne 15.73 11.42 9.56 8.25
System operating cost/tonne 8.55 6.76 6.75 6.97
System labour cost/tonne 3.91 2.80 2.30 1.96
TOTAL SYSTEM COST/TONNE 28.19 20.98 18.61 17.15
CANE DELIVERY (TONNES/YEAR) 9600 13440 16320 19200
TRAVEL DISTANCE = 7 km
System fixed cost/tonne 17.48 12.30 10.16 8.69
System operating cost/tonne 9.49 7.39 6.58 7.28
System labour cost/tonne 4.35 3.01 2.45 2.06
TOTAL SYSTEM COST/TONNE 31.32 22.70 19.18 18.03
CANE DELIVERY (TONNES/YEAR) 8640 12480 15360 18240

Table 6.13. Estimated cost of combine harvesting of cane for a field
(Two factory yard scales)

system with 1 field tractor.

 

 

No. of Wagons

 

Variable 8 10 12 14
TRAVEL DISTANCE = 3 km
System fixed cost/tonne 12.36 11.60 10.23 9.18
System operating cost/tonne 7.19 7.17 7.68 7.31
System labour cost/tonne 3.26 3.01 2.61 2.30
TOTAL SYSTEM COST/TONNE 22.81 21.78 20.52 18.79
CANE DELIVERY (TONNES/YEAR) 11520 12480 14400 16320
TRAVEL DISTANCE 5 km
System fixed cost/tonne 13.48 12.57 10.96 10.40
System operating cost/tonne 7.70 7.88 7.76 7.98
System labour cost/tonne 3.56 3.26 2.80 2.61
TOTAL SYSTEM COST/TONNE 24.74 23.71 21.52 20.99
CANE DELIVERY (TONNES/YEAR) 10560 11520 13440 14400
TRAVEL DISTANCE = 7 km
System fixed cost/tonne 18.53 15.09 12.79 11.14
System operating cost/tonne 10.40 9.16 8.85 8.67
System labour cost/tonne 4.89 3.91 3.26 2.80
TOTAL SYSTEM COST/TONNE 33.82 28.16 24.90 22.61
CANE DELIVERY (TONNES/YEAR) 7680 9600 11520 13440

 

160

Table 6.14. Response of total system cost/tonne and total annual output
to changes in the number of wagons.

 

Number of Cost per Percent Annual output Percent
Wagons tonne change (tonnes/year) change

 

TWO FIELD TRACTORS, ONE WEIGH SCALE

8 48.32 5760
10 42.45 -12.15 6720 16.67
12 28.05 -33.92 10560 57.14
14 26.67 - 4.92 11520 9.09

ONE FIELD TRACTOR, ONE WEIGH SCALE

8 53.55 4800
10 45.73 -14.60 5760 20.00
12 40.38 -11.70 6720 16.67
14 36.77 - 8.94 7680 14.29

TWO FIELD TRACTORS, TWO WEIGH SCALES

8 28.19 9600

10 20.98 -25.58 13440 40.00
12 18.61 -11.30 16320 21.43
14 17.18 - 7.68 19200 17.65

ONE FIELD TRACTOR, TWO WEIGH SCALES

8 24.74 10560
10 23.71 - 4.16 11520 9.09
12 21.52 - 9.24 13440 16.67

14 20.99 - 2.46 14400 7.14

161

Table 6.15. ReSponse of cost per tonne of cane harvested to an additional
field tractor or factory scale.

 

 

PERCENTAGE CHANGE IN COST/TONNE DUE T0:

N0. of Additional Additional
Wagons field tractor factory scale
8 - 9.77 -41.68
10 - 7.17 -50.58
12 -30.53 -33.65
14 -27.47 —35.58

 

Table 6.16. Response of annual cane delivery from the field system to an
additional field tractor or factory scale.

 

 

PERCENTAGE CHANGE IN ANNUAL CANE DELIVERY DUE T0:

N0. of Additional Additional
Wagons field tractor factory scale
8 20.0 66.0
10 14.3 100.0
12 57.0 55.0

14 50.0 67.0

 

162

scale to the factory results in a much larger decrease in the cost per

tonne, and a much larger increase in annual cane delivery, than does

the addition of a second field tractor to the field system.

6.6. The Projected Requirement of Cane Combine Harvesters

6.6.1. The Theoretical Minimum Harvester Requirement

 

Approximately 18,700 hectares of sugar cane are grown annually in
Barbados of which 10,400 hectares, cultivated on slopes of 100 or less,
are considered suitable for combine harvesting (McGregor et a1., 1979).
At an average yield of 57.5 tonnes per hectare, this gives a total of
598,000 tonnes of cane, or 56 percent of the total crop, that can poten-
tially be harvested by combines.

Since the annual rainy season extends from May to December, it is
desirable to complete all cane harvesting operations during the 16 week
period, January lst to April 30th. On the basis of a 10 hour work day
and a 6 day work week, this makes a total of 960 hours theoretically
available for harvesting. It is acknowledged that, even in the dry
season, some work days will be rained out. However, in the absence of
historical data on "suitable days for field work", one is forced to work
with the theoretical figure.

On the basis of 100 observations made during the 1982 harvesting
season, the average time taken for a chopper harvester to fill a cane
wagon under Barbadian conditions is 11.98 minutes. Given a nominal wagon
capacity of 5 tonnes (computed from weigh scale data) this works out to
a theoretical cane harvesting rate of 25 tonnes per hour or 250 tonnes

per day. Putting all these figures together, the theoretical amount of

163

cane that can be harvested annually by a single machine is 24,000 tonnes.
Therefore, to harvest the entire 10,400 hectares considered suitable for
combine harvesting, a theoretical minimum of 25 harvesters would be re-

quired.

6.6.2. Required Number of Harvesters for Various Simulated Annual Output
Levels

 

Table 6.17 shows simulated hourly cane delivery rates for various
field equipment combinations, transporting cane over a distance of 5
kilometres between field and factory. Also shown are the minimum numbers
of harvesters required at each level of output to ensure complete harvest-
ing of the 10,400 hectares considered suitable for combine harvesting.
The figures reveal that, for any given field equipment combination,
significant reductions in the number of combine harvesters required can
be achieved by either adding a second field tractor, or a second factory
yard scale, or both.

Some of the better combine harvester operators each harvested
approximately 7,000 tonnes of cane during the 1982 season (Personal
Communication). This represents a very modest harvesting rate of 7.3
tonnes per hour. Table 6.15 reveals several equipment combinations that
could result in significant improvements on this performance, the best
arrangement being one with 2 field tractors, 14 wagons and 2 weigh scales

at the factory yard.

164

Table 6.17. Cane delivery rates and the number of harvesters required for
various equipment combinations. (Travel distance = 5 km)

 

 

No. of Harvesting rate Cane delivered No. of harvesters
Wagons tonnes per hour tonnes per annum required

 

1 FIELD TRACTOR, 1 WEIGH SCALE

8 5 4800 124
10 6 5760 104
12 7 6720 89
14 8 7680 78

2 FIELD TRACTORS, l WEIGH SCALE

8 6 5760 104
10 7 6720 89
12 11 10520 57
14 12 11520 52

l FIELD TRACTOR, 2 WEIGH SCALES

8 11' 10560 57
10 12 11520 52
12 14 13440 45
14 15 14400 41

2 FIELD TRACTORS, 2 WEIGH SCALES

8 10 9600 62
10 14 13440 45
12 17 16320 37

14 20 19200 31

 

CHAPTER 7

CONCLUSIONS AND RECOMMENDATIONS

7.1. Conclusions

 

On the basis of the close agreement between observed and simulated
values of selected system performance indicators, it is concluded
that the computer simulation models FIELDOP and FACYARD accurately
reflected the operational characteristics of the field and factory
yard systems, respectively.

The main bottleneck of the harvesting system is the weighing oper-
ation at the factory. Given the inter-arrival data used in this
study, factory yard systems having only one weigh scale are incap-
able of handling all of the cane transport units dispatched to them
in a normal work day.

Extension of the daily work period from 11% hours to 16 hours (700
to 960 minutes) would increase the utilization efficiency of com-
bine harvesters only if cane reception at the factory were restricted
to mechanically harvested cane during the period of extension.

The addition of a second weigh scale to the factory yard system

can result in an 88 percent reduction in the average time spent by
a cane transport unit in the yard.

With two scales in operation, the factory yard system would be able
to process all of the transport units currently dispatched to it,
without excessive buildup of service queues.

165

10.

166

In terms of efficiency of utilization of service facilities in the
factory yard, there is no significant difference between having a
single transport unit arrival queue supplying two scales, and having
a separate queue for each scale.

The introduction of a second weighing facility at the factory can
have a profound effect on field systems from which the factory
receives cane. With two scales in operation, a 50-60 percent in-
crease in combine harvester utilization can be achieved and the
overall output from the field system can be increased by up to

68 percent.

With two scales at the factory, field system output can be increased
by increasing the number of wagons assigned to the system until
either the throughput capacity of the harvester becomes limiting,

or the daily cane quota allocated to the field system is satisfied.
Field systems having only one field tractor per harvester can be
expected to deliver 15-25 percent less cane per day than field
systems having two field tractors per harvester.

As a single measure, the installation of a second weigh scale at
those factories having only one scale will generate the most signif-
icant reduction in combine harvesting costs, and the most signifi-
cant increase in the total amount of cane harvested annually by a

harvester.

7.2. Recommendations

 

Based on the results of this study, the following are recommended:

Installation of a second weigh scale at the factory yard in accordance

167

with the guidelines suggested in section 6.4.3., so as to avoid
excessive congestion of the central area of the yard.

2. Assignment of two field tractors to a combine harvester in all
cases.

3. An increase in the number of cane transport wagons assigned to a

harvester from the traditional 8 to 12 or even 14.

7.3. Suggestions for Further Work

 

Due to the unavailability of reliable data on the frequency and
duration of mechanical breakdowns, a fixed time loss of one hour per day
was assumed for repairs, while designing the models. Modification of the
models to represent breakdown frequency and duration as stochastic pro-
cesses may, therefore, be a valuable improvement.

Use of the models to simulate cane transportation systems using
wagons having larger carrying capacities (10 or 15 tonnes rather than 5)
may also be useful, since the reduction in the number of trips between
the field and the factory may be significant.

Finally, from the factory standpoint, it would be interesting to
use the models to investigate the feasibility of temporarily storing
combine harvested cane (probably in pits) for subsequent elevation on
to the feeder table of the factory. The current practice of unloading
combine harvested cane directly on to the feeder table prolongs the

residence time of cane transport units at the factory yard.

APPENDICES

APPENDIX I

DATA COLLECTION MANUAL

168

DATA COLLECTION MANUAL FOR MONITORS

 

Materials Required

 

Clip board.
Data collection sheets.
Scoring pencils/ball-point pens.

Digital watch.

Data Collection Procedures

 

Getting Ready

1. Pick up 10 sheets of the correct form for the task to

which you are assigned, and clip them to your clip board.

2. Fill in the headings on each sheet.
For example: -
Location: Asbhury, Field #1

Start time: Time of arriving on scene and taking
up observation position.

Page of Pages
Put the number 1 in the first blank space on the first
page and leave the second space blank.

Whenever you go to a new page, be sure to fill in the
first blank space with the appropriate page number.

Monitor: Print your name here.

Date: 7 Fill in the appropriate date: e.g., Tues.

Feb. 2.

II.

169

Taking Data

 

3.

4.

10.

Station yourself at the position suggested.

You will be recording times of more than one event
and events may happen very close together. Get the
time first and worry about the vehicle description

later.

. Record the time for each event by noting the EXACT TIME

that the event occurs.

. DO NOT TOUCH ANY OF THE BUTTONS ON THE DIGITAL WATCH!

Simply keep looking at your watch while keeping an eye
on what is going on, and write down the exact time that
the event occurs in the appropriate column on the data
sheet.

For Example:

Start Start Start
Hook Load Load
095706 100230 102255

The events are described on the last page of this manual.
Record the vehicle description and number of trailers/
wagons in the spaces provided on the data sheet.
Observation No. 2 begins when the second vehicle undergoes
the FIRST EVENT on your data sheet.
Add your comments of any happenings which are not normal.
For example: - Collisions, flat tyres, extra movements to
unblock roads, arguments between tractor
operators and field or factory personnel,

machine breakdowns or other abnormal stoppages.

170

You can obtain reasons for stoppages. If a
stoppage occurs and the reason is not immedi—
ately obvious to you, ask the field or factory
supervisor, machine operators or even mainten-
ance personnel what the problem is. These
people have been informed about the project and
about your tasks and are expecting you to ask

questions.

Comments refer to a time value to show that it is abnormal.

For example: - The end of unloading is later than normal be-

11.

12.

13.

cause an argument broke out between the driver
and the unloader operator. To indicate this,
we need a comment number made up of the obser-
vation number and-the column number.
Place the comment number at the end of the line in the
comments column and write the comment in the space provided
at the bottom of the data sheet.
Some typical data collection is attached. Note that only
one set of headings should be on each page or set of pages.
On the first day of work, we will not undertake any data
recording. Instead, we will tour the fields and factories
so that you can become familiar with operating procedures
and learn about the machinery. You will be able to observe
each type of machine closely and you should feel free to ask

as many questions as you like.

14.

15.

16.

171

On the second day, we will start recording observations
but this will only be a practice run. We will spend 30
minutes to 1 hour at each location and you will be shown
exactly how to record the required observations. Each of
you will get a change not only to do some practice record-
ing for your own tasks, but also to see what the tasks of
your co-workers involve.

From the third day, you will be on your own. I shall,
however, be coming around from time to time to make sure
that everything goes as planned.

Finally, a word of caution. Please make sure that you do
not get involved in arguments or other unfriendly discus—
sions with field or factory personnel, with machinery
operators or members of the public who may be on the scene.
The people are doing us a favour by accommodating us on
their properties and we should, therefore, be as polite as

possible.

Identification of Events

 

Vehicle Description

 

1.

2.

Type:

tractor or truck.

Identification: licence plate number.

No. of trailer.

Events in Field

ARRIVE FIELD Time at which tractor returning from factory

stops to unhook empty trailers.

START HOOK Time at which infield tractor stops in front

of empty trailer to hook it up.

II.

END LOAD

END UNHOOK

DEPART FIELD

Events at Factory

ARRIVE FACTORY

DEPART FACTORY

START WEIGH

END WEIGH

START UNLOAD

END UNLOAD

172

Time at which rear wheels of full trailer

pass harvester delivery chute, or loader.

Time at which field tractor moves after

depositing trailer.

Time at which tractor plus two full trailers

leave field for factory.

Time at which tractor and two full trailers
stop in queue at factory, or when front

wheels of tractor enter factory gate.

Time at which tractor and two empty
trailers leave factory or pass monitor at

factory exit gate.

Time at which first trailer or truck stops

moving on weigh scale.

Time at which rear wheels of second trailer

or truck leave the weigh scale.

Time at which truck or first trailer stops

in position for unloading.

Time at which truck or tractor and two

empty trailers leave unloading area.

APPENDIX II

FORTRAN LISTING OF
DATA ANALYSIS PROGRAM
"DATANAL"

o I I I I I I I I I I I I I o I I I I I I I I I I I I I I I I I I I I I I
JJd-‘JJJJ-fiJ...‘Jd-‘JJ‘d‘dJ‘J-‘c‘J‘d‘o-‘d‘o‘d‘r.

173

9/13/82 MSU HUSTLER 2 LSD 51.45 09/10/82 CYBER7SO
.35.09.Is75803

.35.09.uos READ- 09/13/82 11.34.19.
.35.09.HAPVEY.RGZ.JCSOO. 100.

.35.10.LAST ACCESS- B 13. L49 08/15/82
.35.10.RUNS- 0048 PM DOLLAR BALANCE 300107.55
.35.1o.000501 CARDS READ VALUE 50000000 40
.35.10.CP SEC. .000- .000 s .00
.35.I0.0ISPOSE. 'OUTPUT. PRINTERsPAGE. FORMATSELIT
.35.10.E. 151050.

.35.10.RP 00000006 000000000247 MAXFL 025100
.35.10.gggpp SEC. .008- .332 s .01

.35.10

35.33. 1.159 CP SECONDS COMPILATION TIME
35.33 RP 00000300 000000013236 MAXFL 052000
35.33.CP-PP SEC. 1.171- 27. 628 I .97
.35.33.HAL.

.35.33.ZZZZMPL - CYCLE 01. HAL 5. MPL
.35.33.FILE ATTACHED

.35.;2 HAL 5 139

. .ZZZZSOO ' CVCLE 01. HAL 5. SPL
.35.34.FILE ATTACHED

.35.34.RP 00000331 000000013625 MAXFL 025400
.35.34.CP-PP SEC. .214- 28. 592 S .0
.35.34.EXEC. HAL XEC

.35.34.LDSET. LIB= ZZZZSOO.

. 5.34.LGO.

.35.40.EXEC BEGUN. 11. 35. 40.

.35.41. STOD

.35. 2. 032300 FINAL EXECUTION FL.

.35.42. 0.862 CP SECONDS EXECUTION TIME.
.35.42.MAX FILES 0007 MAX PRUS 0001008.
.35.42.RP 00000474 000000014467 MAXFL 050000
.35.42. PP 30.810 SEC.

. 5.42.CP USE 2.317 SEC VALUES .45
.35.£2.PP USE 38.055 SEC VALUES .10
.35.42.CM USE 102 W- H VALUES .97

4a4a4...;.a-u.g.a.;.a...o4.0.3.3...A44‘4‘4444444a4440
O .

.35.42.TOTAL COMPUTE VALUE AT RG2 S 1.52

 

cAAAA
C

C

C****
CAAAA
c****

C

CAAAA
25

cAAAA

cAAAA

55

CAAAA

PROGRAM DATANAL

174

MAIN PROGRAM DATANAL

T A 3- PARAMETRE GAMMA DISTRIBUTION TO WORK TIMES
THE SIMPLEX MAMIMUM LIKELIHOOD METHOD

ENCE--AUDSLEY, E AND D. S. BOYCE, J. AGR. ENG. RES.

'EXACT SOLUTIONS FOR CYCLIC TRANSPORT SYSTEMS'

SEIGRAM5 DATANAL (INPUT. OUTPUT, TAPE5-INPUT, TAPE6-0UTPUT)
INTEGER 5FRT’(IO7. UNITS( )
COMMON /BLKI/NV, x 1000 NTS.
COMMON /BLK2/VALU 100, 26) .DI
R AD(RI) IGO. NUMB, NACT

I ( 0 “6. 50.0

I8:2I7-230

E
F
PA
TS(
GO

)2 .
Reguc' RAW DATA TO ANALYZABLE FORM
D£CON(NUMB)

N- AW DATA READM IN NUMBER. UNITS OF DATA AND FORMAT
5.2) Nv, UNITS,F

IN DATA ACCOR6INO NTO FORMAT (FMT)

5 FMT; (x(I), I-I,

(6 II N

v-

N
AGE
I I
G TO
C LL
G TO
I NO
R D
R D
R D(
W TE , ACT, IPAGE, NV, UNITS
I GE PAGE+I
NVI-N I
030! I'I, NVI
M-N

SORT VALUES INTO ASCENDING ORDER
DO 50 J-I.M

f(
O
A
F
EA
EA
EA
RI
PA

KE D) VALUES
), I-I, NV)
P AND V MATRICES

”—111
_I
m
A
0‘
d
N

(”"0

IE MAXIMUM LIKELIHOOD MATRIX

4111—32—21:! I I .LF‘C
curinfinggnm—><—~»I-—n1
x A‘f—WNHOO“.

A

I

E ET OF NON- RAW DATA, READ FROM NEXT SET
T. 0. AND. NTS. LE. NUMB) GO TO 40

ORE COLUMNS OF RAW DATA TO CHECK, RETURN TO SUBROUTINE
."ESO2 ANB. .NTS. LE. NUMB) GO TO 25

.5 M)

I, AAA 2T6§L"PAGE..",I2/IHO, ."NUMBER OF VALUES I ".
O, I"VALUES) IN ASCENDING ORDER..."//(IX, 9F8. 3))

:ém—

SUBROUTINE FITA

c****
c****

C****

C****

c****

6
200

20]

CAAAA

8

202

I"DEGREES OF FR

IRINT

I”ACTUAL NUMBER.

175

SUBROUTINE FITA)(P, L)

DIMENSION 9(5.

COMMON /BLKI/N, XIID
COMMON /BLK2/VALU IO
COMMON /BLK3/NYVALU(O
ET-P I,L
PL-PZ
EPS-P(3‘L)

BUNBRYI I;-
NYVALU I -
THNOII)-o.
No-zo
ND-ND+I
BUNDRYEI )-I IFIx(x
BUNDRY NB)- -IFIx(
FIND CLASS BOUND
SIZE-(BUNDRY(NB)
DO'BII I-2, ND

3U30RY(I)-BUNDRY(J)+SIZE

I-2

J-J+I

IF J. GT. Nv) CD To I
XIJ). GE. DUNDRY(I

NYVALUIl-I)-NYVALU(

GO TDA

I-I +I

5F(liGT. .NB) CD To 2

\600

O
O. O
0.0
O

)I
I-

GOT
I)+ O 5

I

GO TO“

gngLUINOI'NYVALU(NO)+NV- J+I

FIND THEORETICAL NUMBER IN CLASSES
CONST=(PL**ET)/GAMMA(ET)

DO 6 I-I, NO
J-I+I
A-BUNDRYEI ) -EPS
E-BUNERYM J -EPS

THNOEI -CONST*NV*SIZE*(RINT(A, PL,ET)+RINT(B. PL, ET)+A. 0*
CONTICUE .PL, ET) 6.0

HRI ITE(6E ZOO)

FORMAT(IHI/IHO, 5X "LOWER BOUND. ", X."UPPER BOUND.”,5X,
X, "THEORETICAL UMBER."/)

DO 7‘I-I, ,NO

J- I+
HRITE(6, ZOI)BUNDRY(é), ,BUNDRY(J).NYV AL (I ) ,THNO(I)
EORMATééH ,SX, F72 X, F7.2.I3X,IA,I I6X .F7 2)

ON T IN

FIND CHI SQUARED
CHIZ-O. O

IDF-O

I- -I.NO
DI-THNO I)—NYVALU(I)
IFITHNO I .LT I.O) GO TO 8
CHIZ-CHI2+(DI*DI)/THNO(|)
IDF-IDF+I
CONTINUE
IDF-IDF-A
NRITE(6 202) IDF. CHIZ
FORMAT(IH0/IHOEE06”VALUE DF CNI SQUARED FOR", I3, 2x,

O

M 3" FIG. 6
IPAGEBIPAGE+I
CALL GRAPH(NO)
RETURN
END

FUNCTION RINT
176

c****
c****
FUNCTION RINT(X.PL.ET)
timid:
IF(X.LE.0.0) 50 TO I
RINT-EXP((ET-I.O)*ALOG(X)-PL*X)
RETURN

I RINT-0.0
RETURN
END

SUBROUTINE GRAPH
177

Ctttt
c****

C**** PLOT ACTUAL AND THEORETICAL HISTOGRAM
SUBROUTINE GRAPH(NO)

Ctttt
INTEGER UNITSée)
COMMON /BLKI/ .x 1000).NTS,NACT(IS).IGO.IPAGE
COMMON /BLK2/VALU IOO.20).DIFF IOO,2O).UNITS
COMMON /BLK /NYVALU(50),BUNORY 50),THNO(50)
OIMENSION A INE(80)
DATA DOT,BLANK,AYE.TEA/"."." "."A"."T”/
BlG-NYVAL (I)
no I-2. 0
IF( YVALU IIiLE.BIG) GO TO IO
LE.BIG) GO TO 9

‘-VCAZ C

CONTINUE
9 FACT-BIG/G

d
,0
r

LINE(L).L-IO.7O.IO)
ORETICAL NUMBERS IN CLASS ..... ",

O
row——
M—o'

cz—Izoo—I-Izzwzwzqzzmzz I I zzr+w3—I:c-Iz~\-

wbmm

d

x AOANE(§;$L-II,7I,IO)
IKLINE%L .L-II.80)
.70 I)

I Amsa‘f" . o—Aav—

x—x——
..a—ao d. Va.

DMDMM
—-‘A-4MV—II-'IM

IO ' ININI' U'I-IINI

ALU I FACT+II.

O(I;/%£CT+II.5)S)
HHT- O

HHA- O

6

80
LANK
v

H

——mr|

M4)--

(I).(ALINE(L).L-II.80)
7OAI)

- x--

BUNDRY(J), (ALINE(L),L-II,80)

—mmrn m pmmm II—-rnm

z—-—
_Ou V—nm. dV—lV—ldO dvdnnxxovd

20k

2 AMP ZWE-Ar- 4W,- 1’ '4 'H “M'-

zmowI-OOIH‘I—Or'oowt-l-ofi'fiIII—r’o ' oomomro,10§
Amr'lCOM—r-lr'lmmdr'l- '

mxmt>ont>>o>omt>>o——33>>ogcm£m£>w-H'I

oax——

SUBROUTINE SIMPLE 178

c****
c****

C****

SUBROUTINE SIMPLE(NP. NPS. P, VFUN. L)

COMMON /BLKI/NV. .XEIOOO). .NTS, NACT(IS). IGO,IPAGE
COMMON IBLKZ/VALU I00 20), DIFF IOO ,26),UNITS
COMMON /BLK /NYVALU(NO)I BUNDRY 50) ,THNO(SO)
REMENSION P 5.9

{NTEGER M
8:0. 0 h
mm.
NPI-NPS+
NP2-NPS+2
%T.NPS+?
|5|TP AL sVALUES OF FUNCTION

NPS
I VFUN(J)-LIK(P, J. NP. NPS)

NRITE(6, I9

FORMAT(//, ' SIMPLEx ITERATIONS - (MAX 300 PRINT EACH IOTH)"/)
OO IOI NT-I. OO

FIND MAXIMUM AND MIMIMUN VALUE...

537VEUN(I)

[S-VEUNU)P

DO 2 J-
IF(VFUN(J). PLE. FB) 60 TO 3

H-J

FB-VFUN(H)

GO TO 2

IE}VEUN(J).GE.E5) GO TO 2
Es-VFUN(L)

2 CONTINUE

c****
Ckkflt

Cicicfdc

OQN

<vo—-<‘u£co<‘uomn——o—

6
C****
Ckktfi

12

u.-
vow-"n—tn-m

<
“AC“M“AMOOWAOOOW‘WO“MWAOMIO‘Uxomo

:.

(I,H) IS THE MAXIMUM POINT, P(I .L L)
NO CENngID P(I. NPS+I), NOT IN CL UOI
NPI)-o.o

,NPS
é M) OTO L
U)E-Pu. oNPI)+P(I. J)/NP

NT/IO*IO) .NE. 0) GO
.200) ((P(I 3%,l-I. .NP). 2VFUN(J). J-I ,NPS)
IMO/( x RE]

.90 .Eq. 0 GO TO 22

IS TH EM INIMUM
NG wII H).

(.5?

ZCMLC- I 'v—nm—I I A U-IA—IZ'UMQ

g-nv—ovomv

“CA-"0A0-
Ax—Az-r‘

pmmoobmz-z

mow)—

+1
.I5 AL T6?, ,"PAGE.. 6 I2/

Ex 5ITERA IONS CONTO

INT P( I.H) TO THE NEN POINT P(I, NP2).

MA)*P(I, NPI)-A*P(l. H)
,P2,NP.NPS)
OINT A MINIMUM?
EUN(L)) GO To 6

u3(3)) so T0 8

f"

'"I—PZ

2M A :—
I-i-‘Z' -‘-'-"<‘-N<-IZ‘ ml'"

AIN‘NANZ)—-OAIOO-*h
z

:—

'11—.

-ru
N
v

PAND P(I, NPZ) TO P(I, NP3).
)-G*P(I .NPI)

o-Mt zzv

'u+

:zwca:

'03: I ,-‘°<A-‘
2""

WM.

A C-
22'?-

-n

NV <>zo:
E:

2' d<v2o

SUBROUTINE SIMPLE

179

AXIMUM?

M
H)) GO TO I3

B)*P(I,NPI)

.0-

I

THANSTHE MAXIMUM?

(H)) 60 TO I6

'pNP

nXDFkU()rDVLFLUQ:PHXUIAUDVLFLr
09'ILUDth§FVSLIIln=UPLbDKFViI

ANPSA/NPS

J;*P
.I.OE-5) GO TO 10]

.NPs

,NPS

8'
-RHN*RMN

(P(|.J)+P(|.L))/2.0

100

I

DO 12 J-
RMN- MN+P

19 $02-$02+P
soz=(snz
RMN-RMN/NPS
IF(SDZ.GT
GO TO 20

20 wRITE

"PARAMETERS"
.EIS.6/,

{3?ELETM. u

e
R-

AVIU}OI!A
AANVD

FUNCTION LIK

180
Ctttt
Ctkid:
REAL FUNCTION LIK(P,NPP,NP.NPS)
Ctflet
DIMENSION P(5, )
COMMON /BLKI/N ,X(IOOO).NTS,NACT(IS).IGO,IPAGE
ET-P I,NPP
PL-P 2,NPP
EP-P 3.NPP
IF ET.LE.O.O.OR.ET.OT.2O.og GO TO I
IF PL.LE.0.0.0R.EP.GE.X(I) so To I
‘géNE:L*SA(EP)+NV*ALOG(GAMMA(ET))-NV*ET*ALOC(PL)-(ET-I.O)*
RETURN
I LIK-IOE6
RETURN
ENO

FUNCTION SA
181

c****
Cttkt

FUNCTION SAIEP)
Ctttt

COMMON /BLKI/NC.X(IOOO),NTS.NACT(IS).IGO,IPAGE
SUM-0.0

DO I I-I,NC
I SUMsSUM+(x(I)-EP)
SA=SUM
RETURN
ENO

FUNCTION SB
182

C****
c****

FUNCTION SBIPL)
Ctttk

ggnngNo/BLKI/NC,X(IOOO),NTS,NACT(15).IGO,IPAGE

DO I I-I,NC
I SUM=SUM+ALOG(x(I)-PL)

SB=SUM

RETURN

ENO

SUBROUTINE DACON
c****
Ctttt

Ctttt

CAAAA

C**** RAN DATA

20

Cttkt
C****

25

CAAAA

30

35
AD

“5

50

c****

183

SUBROUTINE DACONINC)

DIMENSION “I’°m ,RM(20)( 5(20)
INTEGER FMT m?

COMMON /DLKI/NRJ NIooo NTS.NACT(IS).IGO,IPAGE
COMMON /BLK2/VALU I00. 26), DIFF(IOO,2O).UNITS
IF THIS Is THE SECOND CALL TO THIS SUBROUTINE DO NOT READ

IGO. E -2 GO TO 2I
I30. E .-3 GO TO 22

O 2
.0 3
2

IF NTS. GT. I) GO TO A;
:| IGO. E§ -I; GO TO 0
I

5 H)R UNITS FMT '
wig II NACT IPACE, UNITS

-I IPADE+I
RT RAN DATA FROM HOURS. MINUTES AND SECONDS To
AL HRS. MINS, DR SECS
( .EMEZ (H(J),RM(J).S(J),J81.NC)
K)-DECIHR(H(K), RH(K), S(K), D)
VERTED DATA IN HRS, MINS DR SECS

(VALU(I, KK) KK-I .NC)
A5) .EQ. 0) GO ’TD 35

IINACT, IPAGE, UNITS

0

00
A (
I E
IT
I E
A E
N E
C M
o

A

Ir-Z)- A—MZF"

I2
.I
PAG
E
LDIF(NC)

(K, NTS) LE. 0) GO TO 50

(
K )sDIFF(M. NTS)

NTINUE

NRTKNUMDER OF VALUES EQUAL TO NUMBER OF DIFFERENCES FOUND

a“)! oc_a‘tr\a‘oC‘ XVI-

O
I
I

+-n

T
T
60
O
T
T
G
V
I
h
D
2
U
TI
T
T
M
T
T
G
T
L
g
K
I

2 FORMAT Ia, ,SAA, IOAA)
II FORMAT AX I, Ah, T65 ,"PAGE. " ,I2//T20. ”INPUT DATA - " ,SAA/I
SIEERMATX ."A RIVE“,3X, “START",SX, "END“,SX, "START",5X,

I0
I
I2

FORMAHAD x, "FACTORY",3X, "NEICM", Ax "NEICM", Ax. "UNLOAD",
£0AMAT(IXW I3, IO(F7. I .2x))
ENDU

SUBROUTINE COLDIF

ISA
CAAAA
CAAAA
C**** SUBROUTINE COLDIF(NC)
C**** THIS SUBROUTINE) CONVERTS RAW DATA TO DIFFERENCES BETWEEN COLUMNS
INTEGER UNITS

COMMON IBLKI/NR, ,XEIOOO). .NTS, NACT(I5L IGO, IPAGE
COMMON 6/DLI<2/VALU 100 20) DIFF(IOO, 26). UNITS

NRITE 6II)NACT, IPAGE. UNITS
NRITE 2,?)
NRITE . O)
IRACE-IPACE+I
DO A I-I NR
DO 2 KK-I NCG
IF( K. ӣ2 NC) 025
DIFF(I , K)-VALU(I, oKK+I)-VALU(I. KK)
DO TO 30

23 DIFF(I.KK)-VALU(I, NC)-VALU(I, I)
CONTINgE)

5 NRITE( .I2) (DIFF(I. KJ) KJ-I .NC)
IF(MOD(I.A5) ).EQ O) GOT 0A0
DO TO A;

A0 IPAGE-l AGE+I

AI NRITE(6.II)NACT,IPACE.UNITS
IFACE-IRACE+I

A CONTINUE

I FORMAT(IHI,I5AA, T65."PAGE..' '.|2//T20, "TIME DIFFERENCES -

I AA
9E0 RMAT(8X, "WEIGH”. AX. "WEIGH", AX. "UNLOAD”, AX. "UNLOAD",
9IAX. “FA ACT ORY”

IO ESE€AT(8X. ;9TIME", ,AX, "TIME”,SX, "QTIME”,5X. "TIME”, 6X.
I2 FORMAT(IX, I3. 2X, F6. I, 3X. F6. I .2(AX, F6. I), 5X, F6. I)
RETURN

END

FUNCTION DECIHR
185
Ctttk

Ctkfit
FUNCTION DECIHR(A.B.C,D)
Ctttt ‘

DECIHR 3 (A+B/60.0+C/3600.0)*D
2‘53”“

186

FACTORY YARD OPERATIONS - CARRINGTON FACTORY

\DQNO‘U‘PWN—I

ARRIVE
FACTORY

516.
517.
528.
530.
550-
559.
560.
567.
583.
587.
623.
625.
656.
662.
666.
667.
67k.
681.
689.
69k.
707.
710.
717.
720.
738.
7&9.
766.
762.

mummuwN—o-oaxaom-aoowwruobanana—OJ.-

INPUT DATA -

START
WEIGH

536.
583-
559-
553-
56k.
567.
573-
578-
596-
603.
636.
639.
670.
677.
680.
682.
689.
693.
700.
711.
722.
726.
728.
731-
7‘9-
76A.
776.
778.

wuvuormmm~mmomwo-rmmmo--mr

END
WEIGH

SAB-
5&6.
553-
555.
567.
570.

575-
581.

600.
606.
638.
691.
672.
679.
682.
685.
690.
697.
70h.
713.
726.
728.
731.
73k.
751.
766.

773-
780.

\DPOWrNNO‘mJ-‘NNO‘U‘ICDUCDCDPPO‘WOOU‘OQP

HINUTES

START
UNLOAD

566.
576.
591.

595.
600.

60k.
609.
616.
622.
628.
655.
66k.
678.
689.
697.
701.
708.
71k.
719.
725.
73“-
7A3.
753.
762.
766.
771.

796.
806.

mu t‘UWO‘NOW m-m-mo~1mma~a~u~momu

END

UNLOAD

573.
578.
593-

597-
601.

606.
612.
619.
626.
633.
659.
667.
681.
69k.
700.
705.
711.
716.
722.
729.
7A0.
7h9.
756-
763.
768.
775-
80A.
810.

--rnmrommuwumwora-mwwoommmm-

PAGE..

1

FACTORY YARD OPERATIONS - CARRINGTON FACTORY

omooua‘mz-wN—o

GNO‘mt’UN-dommNVO‘mJ-‘WN—i

WEIGH
QTIME

20.
26.
21.
22.
13.
13.
12.
11.
1h.
16.
13.
13.
13.
19.
IA.
1A.
15.
12.
11.
17.
15.
12.
11.
11.
10.
IA.

9.
15.

oomruuaswuowuOmwwmmomoa‘N—r-L—o‘mo

NN—NuN-'WNWWdNNNNNN'fiWNNNNNWWN
0 0 0 0 0 0 0 0 . 0 0 0 0 . 0 0 0 0 0 0 0 0 0 0 0 0 . 0
U-‘mwotam~¢r~m~m~urmmuwgerNo

TIME DIFFERENCES -

WEIGH
TIME

187

UNLOAD
QTIME

23.

30.

38.

AD.

33-

3k.
3A.
35-
22.
22.
16.
22.

6.

9.
1h.
16.
17.
16.
15.
11.

7.
15.
22.
27.
15.

”—0
menu:

\OOC’OWJPdrU‘meWWWNNWmeWNdeNU

UNLOAD
TIME ‘

meN-‘Na‘a‘rNNUU-INU‘NWUrWNNNdNN—a‘
00000000000.0000.0.00 00

0.0 00
N-‘mW-‘errU‘IO‘mJ-‘WNNOQWWWPNNOPUb

MINUTES

FACTORY
RES TIME

56.
61.
65.
67.
51.
52.
51.
51.
A3.
116.
36.
AI.

0 0 0 0 0 0 0
NMPNNNQNWO‘O‘U‘rmNt‘m-iHWNNN-‘OWNN

PAGE..

2

188

FACTORY YARD OPERATIONS - CARRINGTON FACTORY PAGE.. 3
NUMBER OF VALUES I 28 UNITS - MINUTES
VALUES IN ASCENDING ORDER...

9.550 10.117 11.200 11.267 11.550 11.567 12.167 12.717 12.733
12.983 13.083 13.383 13.533 13.867 13.967 16.233 1h.250 1h.350
1h.917 15.017 15.667 15.817 16.767 16.983 19.983 21.583 22.h33
26.500

SIMPLEX ITERATIONS - (MAX 300 PRINT EACH 10TH)

.A326682&E+01
.60591956E+01
.6368h037E+01
.h82275h8E+01

.A3635936E+01
.h7075387E+01
.A7hh2150E+01
.h5539517E+01

.1A862276E+02
.15635693E+02
.95992662E+01
.17163772E+02

.I5A16567E+02
.15752651E+02
.16h96565E+02
.15725989E+02

.15701730E+02
.15731708E+02
.15503870E+02
.15563633E+02

.15635696E+02
.15699715E+02
.156031A7E+02
.15603990E+02

.391A1692E+00
.21380639E+00
.382311h5E+00
.21678137E+00

.30660293E+00
.33926012E+00
.30068875E+00
.28238562E+00

.10169962E+01
.11822876E+01
.67618837E+00
.132973525+01

.1111“95&E+01
.11737h87E+01
.1223126AE+01
.11395738E+01

.116386th+01
.11536767E+01
.112719AAE+01
.11607773E+01

.11626069E+01
.11652659E+01
.1139898&E+01
.11h1690hE+01

.51165327E+00
.6088793AE+00
.h73365h5E+00
.52928951E+00

.532A1633E+00
.h7535705E+00
.h9975253E+00
.52199057E+00

.90519532E+00
.10662053E+01
.19037891E+00
.13098001E+01

.10087209E+01
.10726h31E+01

.11765533E+01
.10502308E+01

.10532025E+01
.10607811E+01
.10225719E+01
.103796h7E+01

.1065h322E+01
.10581755E+01
.1090682AE+01
.106193h2E+01

.91h68765E+02
.87820716E+02
.900A0380E+02
.9226368hE+02

.85755193E+02
.8h088528E+02
.83563633E+02
.86160918E+02

.7h867333E+02
.76019570£+02
.76385932£+02
.76675727£+02

.762182h9E+02
.762513AOE+02
.7A23b350E+02
.79209999E+02

.76193322E+02
.7A189937E+02
.76192895E+02
.79189871E+02

.76188561E+02
.7h1885h66+02
.7h188h76E+02
.7h18839hE+02

189

FACTORY YARO OPERATIONS - CARRINGTON FACTORY PAGE.. A
SIMPLEX ITERATIONS CONTD
.15658793E+02 .11573867E+01 .12101A56E+01 .7h17h175E+02
.15550187E+02 .11596221E+01 .13511326E+01 .7h162213E+02
.1552h591E+02 .11393102E+01 .111b1655E+O1 .7A178329E+02
.15653355E+02 .11529379E+01 .111A3372E+01 .76182551E+02
.1bOh3390E+02 .11h097OOE+01 .2516hhhse+01 .7h06779hE+02
.1A802711E+02 .11267927E+01 .16hzh112E+01 .7h107156E+02
.1h9633h7E+02 .11501671E+01 .17007AA6E+01 .7h111212E+02
.1533h500E+02 .11627370E+01 .158260686+01 .7h1h3612E+02
.97586099E+01 .98819558E+00 .A5h87319E+O1 .73687839E+02
.10507773E+02 .10301260E+01 .h3532h76E+01 .73739718E+02
.98961771E+01 .10096667E+01 .h619067OE+01 .73736131E+02
.10993ZAOE+02 .1036117AE+01 .396781A3E+01 .7376317AE+02
.72h2h132E+01 .78369717E+00 .512110A2E+01 .73157087E+02
.7992h506£+01 .7953988hE+00 .h6632835E+01 .73h323h0E+02
.75937138E+01 .76758372E+00 .u6h02607E+01 .73537239E+02
.873h3092E+01 .8817839OE+00 .b6318710E+01 .73h7163hE+02
.65929051E+01 .7802h585E+OO .62986A81E+o1 .7300931AE+02
.5712Ah92E+01 .7AA65937E+00 .6690u823E+O1 .72975775E+02
.66035938E+01 .76561257E+00 .58235230E+01 .73160722E+02
.72995936E+01 .852u6026E+00 .59030120E+01 .732h55h6E+02
.38227033E+01 .58110096E+OO .78601336E+01 .72h08230E+02
.35766093E+01 .52852789E+OO .80185A20E+01 .722527A2E+02
.27523969E+01 .50537113E+00 .8651h895E+01 .72h9OAISE+02
.A1717089E+01 .61925531E+00 .803035h65+01 .72020687E+02
.29993557E+O1 .A9289028E+00 .86133557E+O1 .72036800E+02
.25865198E+O1 .A6A93686E+00 .8859A8A1E+01 .72063825E+02
.2h021990£+01 .A3229298E+OO .899h2108E+01 .71981076E+02
.27156126E+01 .h6082722E+00 .86hh6305E+01 .72069775£+02
.23376925E+01 .A1265523E+00 .90171690£+01 .719556A7E+02
.26081033E+01 .uu890850E+OO .89190937E+01 .71965578E+02
.2uh7hs36E+OI .A3322308E+OO .89658820E+01 .71962829E+02
.25582h57E+01 .h38h0926E+00 .88786879E+01 .71960365E+02

190

PACTORY YARO OPERATIONS - CARRINGTON FACTORY PAGE.. 5

SIMPLEX ITERATIONs CONTO
.23913111E+01 .A1912101E+00 .9003hh61E+01 .7195033AE+02
.2h55h715E+01 .h27573u6E+00 .897h813hE+01 .71950622E+02
.2A370836E+01 .h2351829E+OO .89773193E+01 .71950510E+02
.23855AAZE+01 .h185h316E+00 .90236252E+01 .71950021E+02
.2h123606E+01 .62158679E+00 .90025036E+01 .719h9798E+02
.23960191E+01 .A1935178E+00 .90128h66E+01 .71969802E+02
.2h113725E+01 .A21302h5E+00 .8999721AE+01 .719h9806E+02
.23955260E+01 .A1975256E+00 '.90129986E+01 .71969813E+02
.2h0h0959E+01 .A208797OE+00 .90072285E+o1 .719h9772E+02
.2A083011E+01 .h21307A1E+00 .900A2A29E+01 .719h9769E+02
.2h066h31E+01 .h2116280E+00 .90063A19E+01 .719h9771E+02
.2h036hth+01 .h2071019£+00 .90073917E+01 .719h9770E+02

191

FACTORY YARD OPERATIONS - CARRINGTON FACTORY PAGE.. 6

MEAN VALUE OF FUNCTION AT MINIMUM- .260655E+01
STANDARD DEVIATION I .316780E-05

PARAMETERS AT MINIMUM....ALPHA(SHAP£ FACTOR) - .2h0830E+01
LAMBDA(SCALE FACTOR) OR 1/BETA - .A21307E+00
EPSIL0N(OFFSET) - .900A2AE+01

192

LOWER BOUND. UPPER BOUND. ACTUAL NUMBER. THEORETICAL NUMBER.
9.00 9.90 1 .69
9.90 10.80 1 2.16
10.80 11.70 A 3.03
11.70 12.60 1 3.3L
12.60 13.50 5 3.27
13.50 IA.A0 6 2.97
16.60 15.30 2 2.58
15.30 16.20 2 2.16
16.20 17.10 2 1.76
17.10 18.00 0 1.61
18.00 18.90 0 1.11
18.90 19.80 0 .87
19.80 20.70 1 .67

20.70 21.60 1 .51
21.60 22.50 1 .39
22.50 23.A0 0 .29
23.60 26.30 0 .22
26.30 25.20 0 .16
25.20 26.10 0 .12
26.10 27.00 1 .09

VALUE OF CHI SQUARED FOR 6 DEGREES 0F FREEDOM 8 9.2657

193

FACTORY YARD OPERATIONS - CARRINGTON FACTORY PAGE.. 7
HISTOGRAM OF ACTUAL AND THEORETICAL NUMBERS IN CLASS..... MINUTES

0.00 1.00 2.00 3.00 6.00 5.00 6.00

9.OO.AAAAAAAAAA

9.90.TTTTTTT

9.9O.AAAAAAAAAA

IO.80.TTTTTTTTTTTTTTTTTTTTT

10.80.AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
11.70.TTTTTTTTTTTTTTTTTTTTTTTTTTTTTT

11.70.AAAAAAAAAA

12.60.TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT
12.60.AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
13.50.TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT
13.50.AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
16.60.TTTTTTTTTTTTTTTTTTTTTTTTTTTTTT

16.6O.AAAAAAAAAAAAAAAAAAAA

15.30.TTTTTTTTTTTTTTTTTTTTTTTTTT

15.30.AAAAAAAAAAAAAAAAAAAA .
16.20.TTTTTTTTTTTTTTTTTTTTTT .
16.20.AAAAAAAAAAAAAAAAAAAA

17.10.TTTTTTTTTTTTTTTTTT

17.10.

18.OO.TTTTTTTTTTTTTT

18.00.

18.90.TTTTTTTTTTT

18.90.

19.80.TTTTTTTTT

19.80.AAAAAAAAAA

20.70.TTTTTTT

20.7O.AAAAAAAAAA

21.60.TTTTT

21.60.AAAAAAAAAA

22.50.TTTT

22.50.

23.60.TTT .
23.60. .
26.30.TT .
26.30. .
25.20.TT .
25.20. .
26.10.T

26.10.AAAAAAAAAA

27.00.T

194

123456789.123456789.123456789.123456789.123456789.123456789.123456789.12345
123456789
123456789

.23.
.23.
.23.
.23.
.23.

.28.
.28.
.28.

.28.

.I23456789.123456789.123456789.123456789.12345678
.123456789.123456789.123456789.123456789.12345678

09/13/82
09/13/82
09/13/82
. 09/13/82
09/13/82

SHEETS PRINT ‘4
LINES PRINT 1396
LINES READ 1383
TOTAL LINES PROCESSED 1389
COST AT RG3

1875603 000040
1875603 000040

1.36

1.11
3.03

APPENDIX III

FORTRAN LISTING OF
FACTORY YARD OPERATIONS PROGRAM

"FACYARD"

195

09/20/82 wSU HUSTLER 2 LSD 51.47 09/19/82 CYBER7SO
.10.37.31.IE75790
.10.37.31.UOB READ- 09/20/82 .10. 37. 08.
.10.37.31.HLRVEY RG2. UC1000. L100.
.10.37 31.LAST ACCESS- 5 18.35 09/17/82
.10. 7 31.RUNS- 0063 PN DOLLAR BALANCE 300074.45
.10.37 31.000348 CARDS READ VALUE 60000000. 28
.10.37.32.CP- PP SEC. .0 .000 S .00
.10.37.32.DISPOSE. -DUTPUT PRINTER-PAGE. FORMAT-ELIT
.10.37.32.E 151050.
.10.37.32.RP 000 00006 000000000247 MAxFL 025100
.1O.37.32.CP-P SEC. .012- .326 S .01
.10.37.32.HAL. GASPIv.
.10.37.33.ZZZZMPL - CYCLE 01. HAL 5. MPL
.10.37.33.FILE ATTACHED
.10.37.33.HAL 5.140
.1C.37.33.GASPIV - CYCLE 01. HAL-GASPIV
.10.37.33.FILE ATTACHED
.10.37.33.RP 00000026 000000000635 MAXFL 025400
.10.37.33.CP-PP SEC. .046- 1.243 S .05
.10.37.33.L18RARY CASPIV.
.10.37.34.CP-PP SEC. .047~ 1.325 S .05
.10.37.34.FTN.
.10.37.49. .301 t? SECONDS CDMFZLATION TIRE
.10.37.49.RP 00000244 000000007512 MARFL 052000
.10.37.49.CP° oPP SEC. .366- 20.533 S .55
.10.37.49.LDSET PREEET-ZERD.
.10.37.49.LCO.
.10.37.54.ExEc 856UN.10.37.54.

10.27.55. STOP

10.37.55. 046000 FINAL EXECUTION FL.
10.37.55. 0.295 CP SECONDS EXECUTION TIME.
10.37.55.MA/ FILES 0006 MAY PRUS 0001005.
10.3’.55.RP 00000435 000000010305 MAxFL 060000
10.37.55. PP 23.104 SEC.

10.37.55.CP USE .947 SEC VALUES .18
10.37.55.PP USE 32. 793 SEC VALUES .08
10.37.55.CM USE 3.13 w- H LUES .74
10.37.55.T0TLL COMPUTE VALUE AT vRC2 S 1.00

PROGRAM FACYARD

196
E**** MAIN PROGRAM FACYARD
C HANDLES FACTORY YARD OPERATIONS. NEIGHING AND UNLOADING

CAARR
PROGRAM FACYARD(INPUT.0UTPUT,TAPE5-INPUT,TAPE6*0UTPUT)

DIMENSION NSET(6000)
COMMON SET(6000 0)
COMMON/ COMI/ATR18(25).
1NAPO. NNAPT. NNATR, NNFIL,
2TTCLR,TTF1N .TTR13(25).
2COMMON/UCOMI/XISYS.TISY
EQUIVALENCE (NSET(1).QS

C
C**** SET VALUES FOR CARD READER

NCRDR-z
NPRNT-F

CALL GASP
STOP
END

FA. MFE(100). MLE(100). MSTOP, NCRDR.N
), NNTRY, NPRNT, PPARM(5O, 6), TNOW, TTBEG

JEVNT,M
NN2(100
STTIEZ TIQ3 TIQA BUSZ BUS3 BUSA XVW
E+(I)). 9 0 I D D

A

ND PRINTER

SUBROUTINE INTLC
197

CAAAA
C

EVNT,MFA,MFE(100).MLE(100),MSTOP,NCRDR,N
NgéIOO).NNTRY,NPRNT,PPARM(50,A),TNOW,TTBEG
,TIQZ

,TI03.7106.ausz,DUS3.aUS6.wi

2

C
C**** INITIALISE L

C****
XISYSIO.O
BU52-0.0
BUS =0.0
BUS -0.0
XVW-0.0
RETURN

END

 

SUBROUTINE EVNTS
198

cAAAA
C****

SUBROUTINE EVNTS(IX)
cAAAA

COMMON/GCOM1/ATRIB(25).JEVNT.MFA,HFE(100).MLE(100),MSTOP,NCRDR,N
1NAPO.NNAPT,NNATR,NNFIL,NNE(100).NNTRY,NPRNT,PPARM(50.6),TNOH.TTBEG
2.TTCLR.TTFIN.TTR18(25).TT ET

COHHON/UCOMF/XISYSQT'SYSQTIQZQT'QBOTIQh98U52gBUS308UShngw

cAAAA
IF(IX- )101,102.103

2
101 CALL ARI AL
GO TO 10A
102 CALL ENDWGH
GO TO 106
10 CALL ENDULD
10 RETURN

END

 

SUBROUTINE ARIVAL

199
c****

c****
E**** SUBROUTINE TO HANDLE ARRIVAL OF Z-NAGON TRAINS AT FACTORY

SUBROUTINE ARIVAL

cAAAA
COMMON/GCOM1/ATRIB

( T, MFA, MFE(100). MLE(100), MSTOP, NCRDR.N
1NAPO,NNAPT,NNATR,,NN

JEVN
NN2(100). .NNTRY, NPRNT, PPARM(50, A), TNOW, TTBEG
2TTCLR, TTFIN, TTRIB STTI

T
2COMM0N7UCON17XISY s

02, TIQ3.TIQA,BU52.BUS3,BUSA,XVW
LE FOR NEIGHING
WHOLE CANE OR CHOPPED CANE
IVAL

5).
11'
c 157
C**** DETERMIN C

KJ-IF
C**** DETE

R
IF
C**** SET T
B

ATRI 1 -TN0N+ERLNC(MJ.1)
ATRIBZ -1
ATRIB? -ATRIB(1)
CALL F FLEMII )
cAAAA

E:::: ADD ONE To NUMBER IN SYSTEM AND COLLECT STATS.
XISYS-XISYS+1.0
CALL T1MST(x15YS,TNOw.1)

ATRIB(3)-TNON
c****

2
F
2
,
mgr
a)

(
N
(T
£0

E:::: IF SCALE NOT BUSY, GO TO SCALE AND PREDICT END OF VEIGHING
IF(BUSZ) 111,106,105
106 BU 52' 1.

CALL TIMST(BUSZ. TNOW. 2)

2%;:TNOW+ERLNG(3. 32)
Fl LEM
0.

C

I-I (DID

H(1)
ME IN 02 AND COLLECT STATS.
N0LCT(T102, 2)

CAAAA

CAAAA

E:::: IF SCALE IS BUSY. PUT ARRIVAL IN Q2 AND RECORD TIME IN;

105 ATR18(6)-TNON
CALL F1LEM(2)
RETURN

111 CALL ERROR(87)
STOP
END

 

 

SUBR‘

c****
c****
cAAAA
CRAAA

CAAAA

c****
c****

c****

c****
Ctttt
c****

c2022
CAAAA
c****

1
c****
CAAAA
CAAAA
c****

102

CAAAA

CAAAA
Ckfitfi
CAAAA

CAAAA
CAAAA
c****

103

cAAAA
cAAAA
cAAAA

C****
cAAAA
aflflflk
cAAAA

106
C

cAAAA
cAAAA

cAAAA
CAAAA

C*#**
C****
Cfififla':

105

SUBROUTINE ENDWGH

200

SUBROUTINE TO HANDLE VEHICLES AT END OF WEIGHING

SUBROUTINE ENDWGH

COMMON/CCOM1/ATRIB(25).JEVNT, MFA. MFE(100). MLE(100). MSTOP, NCRDR,
1NAPO,NNAPT.NNATR.NNF1L,NN (160). NNTRY, NPRNT, PPARM(50, 6). TNON, TTBEC
2.TTCLR,TTF1N. TTR13(25).TT ET

COMMON7UCOM17x1SYS,TISYS.T102. T103. 7106. 0052. BUS3.BUS6, wi
DETERMINE TYPE OFOVEHICLE FOR WEIGHING

KJ-IFIX(ATRIB§a)+ g

DETERMINE IF IS I WHOLE CANE OR CHOPPED CANE
KL-IFIXIATRIB(6)+0.5)
SEND WEIGHED VEHICLE T0 CHOPPED CANE CRANE 0R WHOLE CANE HOIST
IF(KL.EQ.2) GO TO 2

CHECK CHOPPED CANE CRANE

IF(BUS3.GT.0.0) GO TO 103

IF CRANE NOT BUSY. SET CRANE BUSY AND
PREDICT END OF UNLOADING

ATRIB 1 -TNOV+ERLNC(6.3)
ATRIB 2 - .0

CALL F1LEM11)

as; CRANE BUSY AND COLLECT STATS.
$163: $135T(8US3, ,TNOW. 3)

CA L COLCT(TIQ3, 3)

GO TO BRING IN NEXT UNIT FOR WEIGHING
GO TO 201

AS BUS3-1. O, PUT ARRIVAL IN CHOPPED CANE QUEUE
ATR18383-TNOW

ATRIB

CALLO iLEM(3)

GOT

CHECK WHOLE CANE HOIST

2 IF(OUS6.CT.O.O) GO TO 105

IF HOIST NOT BUSY. SET HOIST TO BUSY AND
PREDICT END OF UNLOADING

ATRIB ITN W+ R NG
ATRIBNZ7W 8 E L (5 3)

ATR18(7)-2.O
CALL F1LEM(1)

gfithglgT BUSY AND COLLECT STATS.
CALL T1MST(DUS6, TNOW, 6)

T106- .0, 0

CALL COLCT(TIQ6. 6)

GO TO BRING IN NEXT UNIT FOR WEIGHING
GO TO 201

AS 8056-1. 0. 'PUT ARRIVAL IN NHOLE CANE QUEUE
ATRIBégg-TNOW

ATRIB
CALL F1LEM(A)
GO TO

SUBROUTINE ENDWGH

201
CAAAA

CAAAA

201
IF(NNQ(2)) IIZ.II3,IIh
IF 230. SET SCALE FREE AND COLLECT STATS.

113 BUS

C****
c****
CAAAA

IIA

c****

c****

112

80.0
CALL TIHST(BUSZ,TNON,2)
RETURN

r
mm

22. REMOVE IST ENTRY AND
IGHING

rub-143"?"
flPZ’Wf‘I
PM<
qzorrvvomn
mm "If" I I <

02)—
O
t
I
)
.4
:1
W
A
r

ERROR(92)

HMO :nn-Ennwmnx-a-nm—
1"-

2-11’ m>—>>c:
cor- dryer-Pm

Ctkfit
CAAAA

c****

C

c****
c****
Ctkkk

c****
C****

CAAAA
C****

C****
C****

c****
cAAAR

I
Cthtk
Cfikfix

51

c****
c****

52

CAAAA
C****

C****
CAAAA

B
INAPO. NNAPT, NNATR, ,g
S

SUBROUTINE ENDULD

202

SUBROUTINE ENDULD

COMMON/GCOMI/ATRI NT, MFA, MFE(IOO). MLE(IOO). MSTOP, NCRDR,N

E éééTéo)’ NNTRY, NPRNT, PPARH(50. 6). TNOH, TTDEC
Y .TIQ2. TIQ3.TIQ6.DU52.BUS3.DUS6.XVH

HHOLE CANE OR CHOPPED CANE

T

E

2TTCLR, TTFIN, TTR

5
I
2CONHON/UCOHI/XIS I

Uhdzm.

(
N
(
I

THIS IS FOR UNLOADING
NUMBER IN SYSTEM BY ONE

- —AZA
cnnznnua -Nv*flv

CALL TIMST(XI ISYS. TNOW ,I)

CALCULATE NUMBER OF VEHICLES THROUGH YARD
XVW=XVH+I.O

CALCULATE TIME IN SYSTEM AND COLLECT STATS.

TISYS-TNOH-ATRIB(?;
CALL COLCT(TISYS

CHECK TYPE OF UNIT JUST LEAVING UNLOADER
éKL.E8.2 GO TO 2
ATRI (; GT. I. 0) GO TO 53
US 3

OF UE UE
(NNQ(3 3))E0 .31
3-0. S T RANE FREE AND COLLECT STATS.

=0. O
L R;IMST(BUS3.TNOH,3)

—1 U1
cn—UKD

REMOVE IST ENTRY IN Q3 AND SCHEDULE END OF UNLOADING
CALL RMOVE(MFE( (g) .g)

ATRIB Ig=TNOW+E LN (A, 3)

ATRIB 2 '30

CALL FILEM(I)

SET CRANE BUSY AND COLLECT STATS.
GALE TIHST(DUS3. TNDN. 3)

CALCULATE TIME IN QUEUE.3

TIQ =TNOH- ATRI 8(8
CAL COLCT(TIQ3. 3
RETURN

2 GO TO 53

c****
c****

53
CARAA
c****

EA

CAAAA
CAAAA

55

C
C****
C****

Cflkflz':
Cflkfif:

CHECK STATUS OF UEUE 6
IF(NNQ(6))50. 56. 5

ESSA-00 SET HOIST FREE AND COLLECT STATS.
LREIMST(BUSA, ,TNOW, A)

U

LQA. GT.O REMOVE IST ENTRY IN QA AND SCHEDULE END-OF-UNLOADING
LL RHOVE( (MFE(A). A) I
RIBéI ;:TNOH+ERLN G(5, 3)
ATRIB 2

CALL FILEM(I)

BUSAHDIST BUSY AND COLLECT STATS.
CALL TIMST(BUSA, TNOW, A)

ALCULATE TIHE IN QUEUE 6
6=TNOH~ -ATRID( (a;

CA

TI IQ
CALL COLCT(TIQA.
RETURN

SUBROUTINE ENDULD
203

0 CALL ERROR
5 STOP (93)

C
END

SUBROUT I NE OTPUT

206
****
Cittkt
SUBROUTINE OTPUT
COHHON/OCOHI/ATRID(25).JEVNT,HEA, MFE(IOO), MLE(IOO), MSTOP, NCRDR,
INAPO, NNAPT, NNATR, ,NNFIL.NNg(IOO). .NNTRY, NPRNT, PPARM(50, 6), TNON. TTBEC
2, TTCLR, TTEIN. TTRIB(25).TT ET
c COHHON7UCOHI7AISYS,TISYS,TI02,TIQ3.TIQ6,DU52.BUS3.DUS6,xVH
WRITE(NPRNT§I 00) PPARM(I,I) PPARM(2,I) ’
ANOHT - PPAR M( I)*PPARM( LL
AULDTC - PPARM A.I;*PP ARM L. ;
AULDTH - PPARM ?,I *PPARM E,A
NRITE(NPRNT, IOI AULDTC, Au DTH.xVN
c RETURN
IOO FORMAT(/I§X, ,"HEAN INTER- ARRIVAL TIME FOR CHOPPED- CANE VEHICLES . "
gig ’MEAN INTER- ARRIVAL TIME FOR NHOLE- CANE VEHICLES - '.
IOI F RMAT(/I§X, ,“MEAN UNLOADING TIME FOR CHOPPED- CANE VEHICLES - "
I I; "MEAN UNLOADING TIME FOR NHOLE- CANE VEHICLES = “.
g??? 2{, ,Is "NUMBER OF CANE TRANSPORT UNITS THROUGH YARD - ",

END

205

mvump

mO+wOOOh. szuhh .0 uwwmhh O Noaomn w Hawmbfi
w0+wOO—n. pO+wOOmv. «0+wCva. OO+wOONm.
pO+wCOmu. p0+woovo. «0+wOONN. .O‘wOOmp.
n0+w0n0'. .O+w00mm. 00+w0050. OO+mOOvn.
90+w000'. N0+w0mnm. .O u0+wOONm.
.O+wOCC.. N0+m0wvm. .0 vO+wOO—h.

n

\

v

.0 I U u

.0 u 0 ~

.0 n .0.—

.O u .0.—

O r044u2 O umOmZZ O NoOwZZ 0009 thmzz

Om. n>mh22 n Nahmzz O uhaazz m Nzuazz O Nmulzz

thDfinN ZO~muw> >~ am<0 OOOOOOOOOOOOOOOuaDmJJ
w no . awmlbz 23a «mm. \m \h uh<o

>w>¢¢I >m CO. uwm132.h0wfioma Zo~h<423~m

nmpnv

w .3407?

IIn HI 0
FNOVID

n 0
Q Q

vamuhmdad
anmuhm<44
«mam-hm<44
m>m~xth<aa

vo~huum<44
manhuumdaa
Nouhuomtaa
m>m~huom<44

v Iauuzz
v udhmzz

wmnvv Hamma—

'
w

hwm
hum
hum
bum
hum

A.

Ac

“N0? 'NC‘)‘

Q05

Nauudfi
UQDhmZ

auhw2<m<a
awhw2<a<a
muhwidada
mubu£<mda
awhw2<u(a

v N 22-

. uxzuxx

.02
.02
.02

hm2~h
hm2~h
hm2~h
hmzuh

#0400
#0400
#0400
#0400

nuh<22
«#4022

206

.O+uoooN .

—0+w000..

>ba2w mu w4~u ml»

.0 Nzuhoo
.0 N 302k
v awmzaz wank no bDDhZHGQ

>pa2m mu wank th
.0 "Inhoo
.0 u 302»
n mwmzzz WJHu no FDOhZHQQ
>ha£w mu wank th
.0 uzuhoo

.0 N 302p
N ameDZ mama no haohzmua

.0 .0 .0+m000N.
p0+w000N. .0 00+w000m. p0+w000.. 00+w000m. N N >ah2w
.0 .0 .0+w000..
w0+w000.. .0 .0 p0+m000w. .0 N . >mh2w
ththOU wquu
.0 Nimhco
.0 u 3OZh

w awmznz w4~u no h00h2~ma

N N <wfl< wudmchm Ngum 2— mw~m52m no ameDZ EDI~x<S

oo .0 wzuh ha ASDD (mad m04m0hm mahu nw<o.9

207

00.0— H Du¢> IUDoth mh—ZD haOQMZQHh wZCU m0 uwmzzz
mO.N mw40~1u> w2<UIu4013 can wink 02—04042: 2<w2

nh.n u mw40u1w> wZ<UIOwQQOIU do; wtuh 02~0<0420 2<w8
0N.m N meU~Iw> deUINJOIJ not wish 4<>~mm<Iuth~ Zdwt
0—.h N mu40~1w> mZ<UIDwQQDIU not win» 44>~auduawh2~ Z¢w2

9cmh43mwa wh<~0ut¢uh2~¢o

208

QN.

N.N
0.N

mmo

.O+mooo,.
.o

.C+uooo..

n0+w00mm.

m04<> .aZU
«0+wv.0N.
—0+w000v.
n0+wmn.w.
00+w10'v.

thuxdl

m w4~u 2— aumEDZ SDI—Kc:
anm. 20~h<—>u0 ou<02<hm
wuww.n want 2— mmeDZ w0<aw><
n0+u0005. m0~hm—h<hm mom Douawn w£~h

00+wpmmm. utmboo

n0+w000>. u 302»
w mwmtaz wa—u no h00h2~ma

an N dud< wo<a0hm w4~u 2~ mwanZN no awmzaz SDI~X<£

..n0+uoooh. u2~h hd azno (wad mu<mo»m wank am<0..
n0+wOOOn. —O+wooo.. .o OO+mmmmv.
n0+wocoh. .O+wOOO—. .o 00+wv.mv.
n0+wocob. —O+mooo.. .o .OIum.mm.
n0+mocch. ~0+woohn. .o .O+wm—op.

4<>aubz~ mxuh tat~x<z sat—zuz >wo ohm
ocwwam<~u<> FZuhmnmuua-w:~» ecu wo-m~p<hw.o R

.O .O+wmnvw. .OIuow.m. oo.mmvmm.

.o .O+wo.mu. .OIwmwvw. 00+wm.h>.

.o 00+w.w>n. .O+m.mb.. m0+umvmw.

.O+wnmoo. 00+uvan. —O+mhmh.. ~0+mpme.

132—Zn: >0 z<uz no om >wo ohm

..20~h<>&wmmo 20 wa<m mwdm<~u<> mom mumhmnhapmoc

—0+m009n. .0+m000v. ¢0+w000¢. 00+m00Nm.
-0+w00mN. w0+w00wm. p0+w00NN. —0+w00m..
N0+wCNOp. .0+w00mm. 00+w0050. 00.w00vn.
.0+w000'. N0+w0mmm. .0 p0+w00Nm.
p0+u000.. N0+w00vm. .0 «0+m00.h.

00+w0005. u wank wamuDU
w no — awmtbz 20a Nam. \m \h wh<o
>w>m<I >0 00. ummzbz homfioaa zouhdaazum

..haoama >a<523m amdo.a

UN
'Nflvm

OO+un00m
00+wthv
00+umwmm
N0+u.mnu

z<uz
00+whOVN
00+whmmN
«0+wnvhw
N0+wmmnh

de!

hwm
hum
bwm

hwm

vam
anm
Nmam
m>m~x

vouh
non»
N0~h
m>m~h

awhwidudn
awpw2<a<n
uwhw2434n
awhwiduda
uuhw2<8<a

209

.0+w000w. wC+w000p. n0+w0000. m0+w00¢0. .0+w000p. m0+wmmww. N r. >a»Zw

0 0 00+u000N.
¢0+w000N. p0+w000N. n0+wmmm0. 00+wmmmw. p0+w000p. n0+w0hww. N N. >uh2w
.0 .0 .0+w000..
p0+w000.. .C+w000—. 00+wnmmw. n0+wmmmo. ~0+m000.. n0+w00¢®. U u. >ah2w
.0 .0 p0+w000N.
'0+w000N. .0+u000N. n0+umavw. n0+wmmvw. ¢0+w000.. 00+wmmmw. N 0. >uh2u
.0 .0 w0+m000..
.0+w000w. .0+w000.. nO+mpmvw. n0+wvmvw. .0+w000.. n0+wnmmw. n m >ah2m
.0 .0 p0+u000N.
.0+w000N. .09w000N. 00+wmhv0. n0+umhvw. p0+w000.. 00+wmmvw. U m >c»2w
.0 .0 .04w000w.
.0+u000.. p0+w000p. 90+wmocw. n0+umocm. .0+w000.. 00+w.mvw. N b >uh2w
.0 .0 .0+w000w.
.0+w000.. .0+w000.. 00+whwv0. 00+mbwvw. u0+w000.. 00+mmwfiw. u w >mh2w
.0 .0 .0+w000N.
¢0+w000N. .va000N. n0+wmnv0. 00+wwovw. v0+w000p. n0+mmhvw. N m >mh2u
.0 .0 p0+w000w.
¢0+w000p. u0+w000.. n0+wh¢n0. 00+whmnw. .0+w000¢. n0+whwvw. N v >ah2w
.0 .0 .0+w000N.
.0+w000N. w0+w000N. 00+wchnw. 00+wv>nw. .0+w000—. 60+wmnvw. N n >uh2w
.0 .0 .0+u000N.
.0+w000N. ¢0¢w000N. n0+wthw. n0+wthw. .0+w000.. n0+wv>nw. u N >ck2w
.0 .0 «0+w000N.
.0+w000N. .0+w000N. n0+wm>Nw. 00+wthw. .0+m000—. n0+mwth. a _ >a»2w
mFZNhZOU w4~m
mm wauu Zn awmzaz SDI~X<I
ppmm.@ Zo~h<~>wo oa<DZ<hm
00Nm..N wank 2~ ameDZ u6<uw><
00+u0005. muuhm~hthw cog Doumwa wimh
n0+mcmom. NS—hoo
60+w0005. n 302»
N cwmzaz mama no h30h2~ma
.0 .0 w0+w000N.
.0+w000N. p0+w000N. .0 m0+wrn0h. .0+u000.. 00+w—n05. n v >mh2m
.0 .0 '0‘w000N.
.0+w000N. .0+w000N. 00+uwaD. no+wthQ. .0+w000N. n0+me05. u n >awa
.0 .0 p0+w000N.
.0+w000N. u0+w000N. 00+wmme. 00+wmme. u0+u000n. 00+me05. U N >ab2w
.0 .0 w0+w000w.
.0+w000¢. u0+w000w. .0 00+uvp05. .0+w000p. n0+w9.0h. a . >¢h2w

ththOU wguu

 

210

.O+wooow.
.O+mooo..
.O+woooN.
¢O+wOOON.
.O+uoooN.
.O+woooN.
.O+wooo..
'O+woooN.
¢O+moooN.
00+uoooN.
.O+wooo..
.O+wooo..
'O+wOOON.

.O+wooo..

.ovuooop.
—O+wooo..
.O+woooN.
.O+woooN.
.O+wOOON.
.O+u000N.
.0+w000p.
—O+woooN.
.O+wOOON.
.O+m000w.
.O+mooo..
—0+w000w.
.04wOOON.

.O+wooo..

>hazw mu m4~u th

N wJ—u Zn mmeDZ 232~x<2
050.. 20~h4~>wo DfldDZdhm
.mnO. wJ—u Zu aumIDZ wo<mw>4

00+w000h. mouhm~h<hm aOu 00~awa wz~h

n0+wuhow. N2~h00
00+w0005. n 302w
n ammSDZ wJ—u no h00b2~mn

m0+m00mw. nO+uowmeo PO+uooovn
n0+mvomm. n0+wwomwuo p0+wooo.H
n0+wmmom. no+ummowuo .O+wooo.H
n0+wmbom. n0+wmhmmno .O+wooo.v
60+u'cow. n0+wvvmmuo —0+u000pfl
60+wN.om. 00+wN.me0 .0+w000¢H
n0+womh@. 00+womth0 .0+w000.H
n0+wmmho. n0+ummhwno .O+wooo.H
n0+wchho. n0+wvbbouo .O+uooo.u
n0+mbmbw. nO+mhmwao .O+mooovfl
n0+wnmbm. 00+mnmhono .0+w000.H
n0+wb.hw. 00+whphwuo ¢0+u000.fl
n0+m0hww. n0+w0>owuo v0+w000.”
n0+wmmww. n0+wmmwwHM .O+uooo.n

ooooooooooooood

.O+wooo..
n0+wv.0h.
.O+wooo..
n0+woomm.
—O+uoooN.
00+w.moh.
.O+woooN.
nO+wmmmm.
.O+uoooN.
nO+wmhmw.
.O+woooN.
n0+u.vmm.
.04wooo—.
n0+u.0mm.
w0+woooN.
n0+un.mo.
.O+moooN.
n0+wmmb0.
.O+moooN.
mO+uvbhw.
.O+mooo..
nO+wombo.
.O+wooo..
n0+unm>m.
.O+woooN.
n0+m>mbo.
.O+wooo..
nO+wh.hw.
.O+wooo..

ll

QN

mN

NN

wN

0N

mv

m.

by

0—

m.

vw

>ub2m
>¢h2w
>mh2w
>uh2w
>uh2w
>ab2w
>ah2w
>mh2w
>mh2w
>mh2w
>mh2w
>mh2m
>mh2w

>mh2w

 

211

mo; «

a... . a m5
m8
«mm

mm. a 2

003.000» .

Non »( »m00

Dummmooua mw2~4 4<»0»
040a muzua

»Z~mn wa—a

»2—un w»wwIm

>»a2m mu wquu wI»

N0\»N\00
N0\'N\mo
N0\.N\00
N0\.N\00
N0\»N\00

w4~u 2m aw0202 EDS~X<S
20~»¢~>w0 0m<02<»m
wduu 2m awmzaz w043w><

m0~»m~»<»m «Om oouuwa wtu»
00+w»o»0. u!~»00
00+0000». N 302»

Cwmznz mama no »00»2~¢u

.0..»0.»».
.0».»0.»».
.0..»0.»..
.0..»0.»u.
.0»..0.w..

Results for
Extended work period

Simulation time = 960 minutes

212

00.00N a 0¢(> 200001» m»~20 »10nm2<a» 0240 no cwmtaz
m0.N u mm40~Iu> w2<0Iw4013 co; wZH» 02~040420 2402
05.0 a mw40~1w> w2¢0I00au010 cog min» 02—01042: 21w!
0N.m - mw40~Im> deuquOII 105 min» 4(>~au(Imm»2~ 2(w!
0'.» I mmJU—Iw> NZCUIOmAQOIU 80$ 02—» 4(>~¢a¢Ium»2~ 24w:

.om»40mw¢ m»<—0w!um»2~cc

213

0».
N».
.mN
moN

W00

.0
»0+0000».
-0+0000..
N0+00000.

w34¢> .uao
.O+wv.wu.
.O+wwva.
n0+uuon..
n0+w»nc..

102~X¢2

n0+00000.
00+00000.
00+00000.
00+00000.

J<>uw»2~ wtu»
Iemwa0(~¢

.0
.0
.0
v0+00000.

t:x~2~x
«Izo~»<>uwm

.O+woo.n.
.O+uoomn.
uO+uOno..
.O+wooo..
.O+uooo..

» to —

>w>¢(1

.0+0000.. .0 00+wnmmv.
.O+wooo.. .0 00+mwmmv.
.O+wooo.. .o .OIw.»mv.
«0+w00mc. .o .O+mmnoc.
208~K<I zasnzut >00 0»m
<> »2w»m_mawd-wx~» can mu~»m~»<»moo
.o+wmv.u. .OIwmovv. 00+wmmmm.
.0+wmn»N. .OIwVOOh. 00+0N.Q».
OO+mmn»n. .O+wv»o.. «0+wmmmu.
OO+mv0vo. .O+ummo.. N0+w0now.
>0 24w: no om >wo o»m .
no 20 cum¢m mmam<~c<> can mo~»m~»<»ma.
.O+w00ov. .0+wooQ.. OO+mOONm.
.0+000.0. »0+000NN. .0+0000..
.O+w00mo. OO+woo»o. OO+moo'n.
«0+u0mnm. .o .O+woon.
«0+womvm. .o .O+uoo.».

00+00000. I win» »20¢¢00
¢w0202 20¢ N00. \0 \» w»(0
>0 00, 800202 »0000ca 20~»1402~m

Io»coamm >c<zzam nw<qu

00+wmwnm
00+mbnmn
00+umhmm
«0+wnm»N

24w:
00+mm»»u
00+waohw
No+mnwo»
u0+wnmuo

2(0!

"NOVID

»wm
»wm
»0m
»00
»wm

vm00
nm00
Nm00
m>m~x

V0~»
nO~»
Non»
m>m~»

aw»w!(u<a
mm»w!<u<a
um»w!(m<n
aw»02<u<n
cu»02<u¢a

Modified subroutines
ARIVAL and ENDWGH

for l-queue, 2-scale configuration

Ctttt
Cttit
Ctttk

Ctttt

Ctttt
c****
c****

Ctttt
Ctttt
Ctttt

c****
c****
Ctttt
Ctktt

IDA DU 52' I

c****

c****
Ctttk
Ctttt
IOS
CAARA
1068

c****
Ctttt

I07

III

INAPO NNAPT NNATR NNF

SQORQUT I NE ARIVAL

214

SUBROUTINE TO HANDLE ARRIVAL OF Z-HAGON TRAINS AT FACTORY
SUBROUTINE ARIVAL

COHHON/GCOHI/ATRIB(2fl-,NN JEVNT MFA ,MFE(IOO) ,HLE(IOO), MSTOP, NCRDR.N

#géIOO). .NNTRY, NPRNT, PPARH(SO, h), TNOW. TTBEG

SYS S.TIQ2. TIQB.TIQA.3U52,OU53.OUSA.Buss.wi
9F VEHICLE EOR HEICHINC

E

(a +0.?)

TSIS I NHOLE CANE OR CHOPPEO CANE
Ex

TTCLR. TTEIN, TTRIB(2?)W
2COHHON7UCONI7x SYS .T

ADD ONE To NUMBER IN SYSTEM AND COLLECT STATS.
XISYs-XISYS+I.O

CALL TIMST(XISYS, TNOW, I)

ATRIB(3)-TNOH

CHECK SCALE 2
IF SCALE NOT BUSY. GO TO SCALE AND PREDICT END OF WEIGHING

IF(BUSZ) III, IOA. I05

CALL TIHST(OU52 HTNO
ATR Rg I-TNOH+ER LNO( $3
T

EHII )
IN Q2 ANO COLLECT STATS.

I
O.
CO OCT(TIQ2. 2)

I
RI
L
I 2
L

IF SCALE 2 BUSY, CHECK SCALE I

IF(BUSE) III 106.10;
IE SCA EREE, SE BUSY ANO PREOICT ENO OE HEICHINC

CAL? TIMSTIBUSE TNOW65
ATRIB -TNON+ RLNO(6 52)~
ATRIB 2-
CALL FILEH(I)

TIoz- 0.0
CALL COLCT(TIQZ, 2)
RETURN

IE SCALE I IS BUSY. PUT ARRIVAL IN Q2 ANO RECORO TIME IN.
ATRIB(h)-TNON

CALL FILEH(2)
RETURN

CALL ERROR(87)
STOP
END

Ctttt
Ctktt
c****
C****

c****

INAPO, NNAPT NNATR, NNF

SWRDUT I NE ENDWGH

‘ 215

SUBROUTINE TO HANDLE VEHICLES AT END OF WEIGHING
SUBROUTINE ENDWGH

COHHON/GCOHI/ATRIB(2 EVNT6 MFA ,HFEIIOO). HLEIIOO) HSTOP NCRDR,N
"26
T

I.
I' 60). NNTRY, NPRNT, PPARN(§O. u). TNON. TTOEC

:‘ZL

TTCLR, TTEIN ,TTRIB(25

* 2COHHON7UCOHI7IIISVS,TI ISYS. IQ2. TIQ3. TIQA, BUSZ. BUS3.BUSA, BUSS. xVN

c****
c****

Ctktt
Ctttt
Ctttt

C****
Ctttt
CAAAR

ICLE FOR WEIGHING
WHOLE CANE OR CHOPPED CANE

OETERHINE TYPE OE VE
KJ-IFIXIATRIB )+ O.
OETERHINE IE IS I
KL-IFIX(ATRIB(6)+O. 5)

SENO WEIGHED VEHICLE TO CHOPPEO CANE CRANE OR NHOLE CANE HOIST
IF(KL.EQ.2) GO TO 2

CHECK CHOPPED CANE CRANE

I IF(BU53.GT.0.0) GO TO 103

Ctktt
CAAAA
C****
c****

102

C****

c****
c****
Ctttfi

CRAAR
c****
CRAAA

103

Cttkk
c****
c****

IF CRANE NOT BUSY, SET CRANE BUSY AND
PREDICT END OF UNLOADING

ATRIB I -TNOH+ERLNc(h.3)
ATRIB 2 - .O
ATRIB I - .0
CALL E LEN(II
SET CRANE BUSY ANO COLLECT STATS.

CAL?” TIHSTIBUS}, TNOW. 3)

TIQ W2-
CAL COLCT(TIQ3, 3)

GO TO BRING IN NEXT UNIT FOR WEIGHING

GO TO ZOI

AS BUSB-I. O, PUT ARRIVAL IN CHOPPED CANE QUEUE
ATRIBEB}:TNOW

ATRIB
EALL02 FILEH(3)

CHECK WHOLE CANE HOIST

2 IF(BUSA.GT.0.0) GO TO 105

C****
c****
Ctttt
c****

IOA

Ctttt
c****

CRAAR
Ctttt

c****
CAAAR
Ctkkt

I05

C****

IF HOIST NOT BUSY. SET HOIST TO BUSY AND
PREDICT END OF UNLOADI NG

ATRIB ITN W+ERLNG
ATRIBIZIW 8 (5 3)

ATRIB -2.
CALL RIIEN(I)

BUT HOIST BUSY AND COLLECT STATS.

S II.

CALh TIHSTIBUSA,TNOW,A)

Q -O.O
CALL COLCT(TIQA.A)

GO TO BRING IN NEXT UNIT FOR WEIGHING
GO TO 20]

AS BUSA'I. O. PUT ARRIVAL IN WHOLE CANE QUEUE

ATRIB{;;-TNOH
ATRIB
CALL EILEM(A)
GO T02 zoI

“BROUT I NE ENDWGH

C*§gl IE(NNQ(2)) II2.II3,IIA
C**** IE 32-0 SET SCALE EREE ANO COLLECT STATS.
II3 BUS -0.
CALL TIHST(ausz.TNON.2)
RETURN
CAM":
C**** IE VEHICLE IS IN 2. CHECK SCALES
C**** IE VEHICLE IS IN 2. REHOVE IST ENTRY AND
C**** SCHEDULE ENO OE w IGHING
can
IIA CALL HOVE(HEE(2).2)
ATRIB I}-TNOW+ERLNG(3. .2)
ATRIB 2 -2.0
CALL EILEH(I)
C**** EEEZSCAbE BUSY ANO COLLECT STATS.
CALL TIHST(BU52 TNOW 2)
c**** CALCULATE TIH B(IN QUEUE
TIQ2-TNOW-ATR
c CALL COLCT(TIQ2. 2
C**** IE SCALE 2 BUSY HECK SCALE I
IF(NN8(2)I I12.
I16 BUSE- 0
CAL TIHSTIBUSS, TNOW, 5)
RETURN
II7 CALL RHOVE<HEE(2). 2)
ATRIBilg-TNOW+ERLNG(6, ,2)
ATRIB 2 -2.
CALL EILEH(I)
BUSE-I.0
CAL TIMST(BUS§ .TNOW, 5)
TIQ2 -TNO W-ATR (A)
CALL COL LCT(TIQ2. 2
RETURN
c
c

II2

~216-

CALL ERRORISZI
STOP
END

 

217

Modified Subroutines
ARIVAL and ENDWGH

for 2-queue, 2-scale configuration

cAAAA
cAAAA
CAAAA

SUBROUTINE ARIVAL

218

SUBROUTINE TO HANDLE ARRIVAL OF 2-WAGON TRAINS AT FACTORY

* SUBROUTINE ARIVAL

C

c****
c****
CARAA

CAAAA
cAAAA
c****

C****
Ctttt

CAAAA
cAAAA
cAAAA

IOh

c****

c****
C****
CARRA

I05

CAAAA
C****
C

106
I078

cAAAA

c****
c****

108

III

C0HHON/CCOHI/ATRIBI25).JEVNT. HEA .HEE(I00). HLE(IOO). HSTOP NCROR, N

INAPL NNAPT. NNATR, ,NNFIL.NN2(IOO ). NNTRY, NPRNT, PPARH(5L A), TNOL TTBEG
TTCLR, TTEIN. TTRIBIZ?).TT ET

:COHHON7UCOHI7XISYS.T SVS.TI02.TI03.TI0A.TI05.ausz.ausa.OUSA.OUS5.

:10: dl-Il-I

T
O
IND OW+ERLNGIKJ, I)
EA

H‘I’RIBU)
CAL

A00 ONE TO NUHOER IN SVSTEH ANO COLLECT STATS.
XISYS-XISYS+I.O

CALL TIHST(xISvS. TNOL I)

ATRIB(3)-TNOW

SENO TRANSPORT UNIT TO CORRECT SCALE

.2) 00 TO 106

H PPEO CANE UNIT 00 TO SCALE #1

C LE NOT BUSY, 00 TO SCALE ANO PREOICT ENO 0E WEIGHING
) III. IOA, I05

HST(BU52 TNOL 2)

E-ENOW+ERLNG(3. 2)

"(1)

IN 02 ANO COLLECT STATS.

NOLCT(TIQL 2)

”n-IMO)>OID—

IE SCALE IS BUSY, PUT ARRIVAL IN 02 ANO RECORO TIME IN;
ATRIOIA)-TN0N ‘

CALL EILEH(2)

RETURN

IE WHOLE CANE UNIT. GO TO SCALE #2

IF(BUSS) 111,107,108

CAL TIHST(OUS ,TNOH 5)

ATRIBEI g-TNOW+ RLNG(S, 2)

ATRIB 2
CALL EILEH(

SET TIME IN 05 ANO COLLECT STATS.

“‘2'"?
gALURN COLCT(TIQS. 5)

IE SCALE 2 IS BUSY, PUT ARRIVAL IN 05, RECORO TIHE IN.
ATRIB(A)-TNON

CALL EILEH(5)

RETURN

CALL ERR0R(87)

STOP

ENO

Ctttt
Ctktt
c****
Ct ”A

Ctttt

cAAAA
CRAAA

C****

c****
c****
c****

c****
c****
c****

I
c****
c****
c****
c****

I02

Ctttt

c****
c****
Ctttt

C****
c****
c****

I03

Ctttt
c****
c****

SUBROUTINE ENDWGH

219

SUBROUTINE To NANOLE VEHICLES AT END OF NEICHINC

SUBROUTINE ENONCN

COHMON/GCOMI/ATRIB(2?),JEVNT.8 NEA ,NEE(IOO), NLEIIOO) NSTOP+ NCROR,N
INAPO NNAPT NNATR NNE L.NN2(IOO ), NNTRY, NPRNT, PPARN (5L A). TNOL TTBEG
2. TTCLR, TTEIN ,TTRIB(25).TT ET
I§8°M0N7UCOHI7XISYS.TISYS.TIQ2.TIQ3.TIQN.TIQ5,BUSZ.BUS3,BUSN.BUSS.

EHICLE FOR WEIGHING
g WHOLE CANE OR CHOPPED CANE

EHICLE TO CHOPPED CANE CRANE 0R WHOLE CANE HOIST
IE(KL.EQ.2) GO TO 2

CHECK CHOPPED CANE CRANE

IF(BUSB.GT.0.0) GO TO I03

IF CRANE NOT BUSY. SET CRANE BUSY AND
PREDICT END OF UNLOADI NG

ATRIB I-TNON+ERLNC(A. 3)

ATRIBZ

ATRIB7

CALL EILEN(o )

SET CRANE BUSY ANO COLLECT STATS.

CALE.l TIHST(BU53, TNOH, 3)

TIQ WE-
CAL COLCT(TIQB 3)

GO TO BRING IN NExT CHOPPED CANE UNIT EOR VEICNINC
CO TO 20) ‘

As BUS3-L O. PUT ARRIVAL IN CHOPPED CANE QUEUE
ATRl8283-TNOW

ATRIB
CALL oEILEN(3)
GOT

CHECK HHOLE CANE HOIST

2 IF(BUSN.GT.0.0) GO TO 105

c****
CAAAA
C****
c****

ION

CAAAA
c****

c****
C****

c****
CARAA
CARAA

I05

IF HOIST NOT BUSY. SET HOIST TO BUSY AND
PREDICT END OF UNLOADING

ATRIB I -TNOH+ERLNG .
ATRIBIZI'M (5 3)

ATRIB -
CALL NILEH(I)

35S HOIST BUSY AND COLLECT STATS.

h- I. 0
CALL. TIHST(BUSN, TNOW, A)

-o. 0
CALL COLCT(TIQA, A)

GO TO BRING IN NEXT WHOLE CANE UNIT FOR WEIGHING
GO TO 202

AS BUSh-I. 0. PUT ARRIVAL IN WHOLE CANE QUEUE
ATRIBE;;:TNOW

ATRIB
CALL oEILEN(A)
GOT

c****

201
cA*A*
c****

1138

c****
c****
Ctttt
Ctttt

11h

c****

C****

c****

C

202
c****

116

C
c****
117

nn

112

SUBROUTINE ENDWGH

220

IF(NNQ(2)) 112,113,11h

IF US92-06 SET SCALE FREE ANO COLLECT STATS.
CALL TIMST(BUSZ, TNON. 2)

RETURN

S IN
OF N

( FE(2). 2)
oN8N+ERLNO(3. .2)

2, REHOVE IST ENTRY AND

I
ND IGHING
H

EH (I)
E BUSY ANO COLLECT STATS.
s

T(BUSZ TNON 2)
E TINEB IN OUEU
ON-ATR (h;
LCT(T|Q2, 2

IF(NNQ(§)) II2, II6 II

EHPTY, SET 5 ALE 2 FREE ANO COLLECT STATS.
aus -O. O

CAL

RETUR

EEHST(BUSS.TNON.5)
E CANE UNITS ARE IN Q5. SENO ONE TO SCALE #2
OVE(NEE w(3)' NE)
g:TNO WE LN (6. 2)
LEH(I)

HST(BUS

E I

NH
L
I
IS 20
L

IF L
CAL H
ATR I
ATR 2
gAL I
CAL ,TNON, 5)
TI H- ATR (h

0
a?
F
T
_T
C CT(T|Q5.5

5 -IE
N

83°

cETURN

CALL ERROR(92)

STOP

ENO

Results for

Configuration with l queue and 2 scales

221

OO.wnN N ou¢> IUDORI» mh~22 baoam2<ah w240 no cwmIDZ
mm." U ww40~Iw> w2<ouw4013 mom wtuh 02—D<OJZD 24w!
“5.0 N mw40~1w> w2<UIOwnaDIU Bog wank 02~D<OAZD 2(u!
ON.m u mw40~1w> w2«UIm4013 no; w2~h 4<>~aadnuuh2— z<ut
O..h u mwaun1w> uZ<UIDwaaDIU men wzuh 4¢>~uu<Imth— 24w!

¢.mb43mwm upduowzmmhzuq.

222

nvw

man
now

mmo

pO+wOOO..
¢O+wOOO'.
.O+wOOO..
.O
uO+wOOOn.

w34<> .ano
uO+wmmo..
u0+moo.—.
.O+wmv.m.
n0+wun.u.

838~x<8

n0+uoooh.
n0+uooob.
nO+mOOOb.
n0+woooh.
n0+woooh.

4(>¢mh2~ mluh

—O+wwmon.

.O+wooo.. .o 00+thnv.
.O+wooo.. .o 00+umvmv.
.O+wooo.. .o oo.m.mmc.
.O+wooo.. .o 00+ww>mv.
.O+mooom. .o .0+wowm..
SDI—x4: z::~z~: >uo chm
camwum<~u<> FZubmumcua-m:_p no; munpmnp<pmSA

.O+w.mn.. 00+u>mm.. .O+wmmnu.
.O+wohh.. 00+w~mmn. .O+whmon.
.O+uomm.. 00+w>.o.. .O+wnsm..
00+u.nmn. 00+un0nw. .O+wonmn.
>0 Zau: no om ‘ >uo ohm

838~2~8

oozonh<>awmmo 20 mecm mw4m<~u<> can monbmuhdhch

uO+mO~o..
.O+woo.n.
.O+w00mn.
«0+w0uo..
.O+wooo..
.O+wooo..

— no w

>w>u¢1

—O+w00mm. OO+wOOhm. OO+wOO'n.
.O+w000v. .O+wOOm.. OO+wOONm.
.O+woovm. ¢O+mOONN. .O+wOOm—.
.O+wOOmm. OO+wOOhO. OO+wOOvn.
NO+wOmnm. .O .O+wOONm.
NO+wowvm. .O —O+mOO.b.

nO+wOOOb. A ulnh Fanunu
uwmiaz 23a «mm. \m \h whdo
>m DO. uwmtaz howfioun 20~h<432~m

cohuoawu >a¢823w athcc

"Nfl'lnm

OO+whwvw
00+w0n>m
OO+u.ohv
OO+wmmem
.O+mn>0n

z<wz
.O+wvmr—
.O+w.no.
OO+wwOmm
.O+wno0m

24w!

bum
hum
hmm
hum
hum
hum

mmDm
vam
anm
«mam
m>m~x

v0~b
non»
«Duh
m>m~h

mwhwzdcda
awhw2¢a<a
awhwzdada
cwhw2<m<a
awhwidudn
cmhwtflu<a

APPENDIX IV

FORTRAN LISTING OF
FIELD OPERATIONS PROGRAM

"FIELDOP"

01\
LOW

88

00000000

Mammmmmmmmqqqmmmmnnn44a.....A

mmummmuuumwmumwmmummmmmmmmmmmummmmmwwma
OOMNMMMMDMMMMNMMMO

. . . C . . C C C . . . . C ' . 0 . . C . C . . . . . . U . . . . . . . . .
‘di‘d““dd“‘c‘“““‘““‘J‘d‘d-‘d‘u‘dd

223

2 MSU HUSTLER 2 LSD 51.52 10/30/82
.1884610

.UOB READ- 11/01/82 .15.00.02.
.HARVEY.RG2.JC1000.L100.
.LAST ACCEss- B 17.12 10/29/32

.RUNs- 0101 RN DOLLAR BALANCE 300067.30
.001026 CARDS READ VALUE 30000000. 82
.CR- -PP SEC. .000- .000 S .00
.DISPOSE. ‘OUTPUT. PRINTER-PAGE. FORMAT-ELIT

.E.1SIDED.

.RP 000000000247 MAXFL 025100

DETNPP SEC. .011- .337 S .01
1.313 CP SECONDS COMPILATION TIME

.RP 00000376 000000016343 MAXFL 052000

.CP-PP SEC. 1.328- 44.740 S 1.18

.HAL.GASPIV.

.ZZZZMPL - CYCLE 01, HAL 5. MPL

.FILE ATTACHED

.HAL 5.149

.GASPIV - CYCLE 01. HAL'GASPIV

.FILE ATTACHED

.RP 00000416 000000016731 MAXFL 025400
.CP-PP SEC. .342- 45.728 S 1.22
.LIBRARY GASPIV

.CP -PP SEC. .343- 45.769 S 1.22
.LDSET. PRESET-ZERD.

IEXEC BEGUN.15.04.01.
ST OR

067100 FINAL EXECUTION FL.

. 0.134 CP SECONDS EXECUTION TIME.
.MAX FILES 0006 MAX PR US 0001008.

.RP 00000615 000000017624 MAXFL 067100
. PP 49.803 SEC.

.CP USE 1.838 SEC VALUES .36
.PP USE 47. 784 SEC VALUES .12
.CM USE 5. 235 V H ES 1.24
.TOTAL COMPUTE VALUE AT vRG2 S 1.72

PROGRAM F I ELDOP

 

 

 

224
c PROGRAM FIELDOP(INPUT.OUTPUT.TAPE5-INPUT.TAPE6-OUTPUT)
c--= x: ----- ssassaasssccczzzzzae
C THIS PROGRAM SIMULATES THE FIELD ACTIVITIES INVOLVED IN
C THE MECHANICAL HARVESTING OF SUGAR CANE. AND ALSO
C THE ACTIVITIES INVOLVED IN A CYCLIC TRANSPORT SYSTEM OF
C THE CANE BY VAGONS PULLED BY ROAD TRACTORS FROM FIELD
E TO THE FACTORY. AND BACK TO THE FIELD.
C
E**** LIST OF USER-CENERATED INPUT VARIABLES
C ISC - I :SIMULATE NAGONS
C - O :END OF SIMULATION
C NHT :NUMBER OF HARVESTERS
C NFT :NUMBER OF FIELD TRACTORS
C NRT :NUMBER OF ROAD TRACTORS
C NEST :IDENTIFIES LOCATION SIMULATED
C UOTA :ESTATE DELIVERY UOTA OF CANE FOR ONE DAY
C AP :AMOUNT OF CANE D LIVERED BY A TRANSPORT
C UNIT PER TRIP (TONNES)
C STM :START TIME ON THE FIRST DAY
C ENT :END TIME ON THE FIRST DAY
c BREAK :LUNCH PERIOD BETNEEN END OF FIRST AND
C START OF SECOND SHIFT
C NKTM :TIME AVAILABLE FOR NORK EACH DAY
C NRPRT. LE. I:COMPILE DATA FOR SHORT PEPORT
C .GT. I:COMPILE DATA FOR LONG REPORT
C LRPRT. LT.O:PRINT COMPLETE GASPIV RE
E . .O:PRINT FILE CONTENTS AND INPUT DATA ONLY
c
DIMENSION NSET(IOOOO)
COMMON SSET(IOOOO£
COMMON GCOMI/ AT |B(25LN .JEVN T MFA MFE(IOO ,MLE(IOO). MSTOP, NCRDR,N
INAPO, NNAPT,NNATR,NNFIL. N2(IO OOI, NNTRY, NPRN PPARM(5O A). TNON. TTBEG
2, TTCLR, .TTFIN,TTRIB(23) TT ET
COMMON /UCOMI/ ISC,8 OTA.AMOUNT ENTM BREAK STTM NRT, NHT, NFT, NNG,
ICAP. BUSHT(2).BUSFT( ).BUSRT(6) .TIQE, TIQF. RTRIP, SHFND, NKTM
COMNON /UCOM2/ DNN.SKLOM
COMMON /uCOMA/LRPRT Nz7, ,NSHF. NDAY, NEST, NRPRT. KL
COMMON /UCOM6/ ENT ST
EQUIVALENCE (NSET(I).QSET(I))
Citin’cit
C
IOO READ(g.I) ISCG
IF(Is EQ. OLGO zoo
READ .2 N G NHT. :NFT, NRT. NEST, NRPRT. LRPRT
READ .2 EUOIACA
READ . TM ENT. BREAK. NKTM
NCRDR-
NPRNT-

C
C**** THIS SEGMENT OF CODE PRINTS OUT THE INPUT DATA

IF (ISC. I) NRITE(6 IsAO) ISC NNG, NHT. NFT. NRT. NEST
filTEig. ”3 3 3UOTA, ,CAP éSTM. ENT. BREAK. NNTM'
mITE O RPRT LRPRT

CALL GA ASP
C GO TO IOO
I FORMAT 2|2;
2 FORMAT ZIz
z FORMAT FIO.N;
FORMAT AEIO.A
AD FORMAT "I”,///,T26."NON- GASPIV DATA INPUT. o",///,T AO,"ISC -".
II2./ TAO "NNG -".I2 / TAO, "NHT -' IL /.TA O,"NFT- ',|2.
2/ TNO."NRE -".I2,/, 3?, "NEST -H,
50 F6RMAT(T; ."QUBTA -" IO. 2 " TONNES" /. TAO. "CAP -",FIO.2,
I" TONNES'.// T O "ST M -' FIO. 2." MINS' ,TAO. "ENT -",FIO. 2.
2" HINS”. /. T38."BR REAK -" .FIO. 2." HINS”,/, T39."NKTM -",FIO. 2.

0END

SUBROUT INE INTL

Cfiktfi

nnnn

c 225

SUBROUTINE INTLC

THIS SUBROUTINE INITIALIZES THE E UIPMENT COMBINATION TO BE
SIMULATED AND RESETS HARKERS AND ATA STORAGE ARRAYS

 

 

c****

C
C****

E****
c****

IO
C
c****

200

210
c****

220

C
c****
c****

COMMON /GCOMI/ ATRIB(25). JEVNT MFA, MFE(IOO) .MLE(IOO). MSTOL NCRDR.N
INAPO. NNAPT. NNATR. NNFIL. NN (IOOI. NNTRY, NPRNT, PPARM(5O. h). TNON, TTBEG

2TTCLL TTFIN, TTRIB

ZEOMMON’ /UCOMI / IS Lia OTA; AMO NT,r ENTM +BREAK STTM NRT. NHT NFT. NNG.
ICAP. BUSHT(z). BUSFT(S )I WRT( TIQE. TIQF. RTRIP. SHFND. HKTM
COMNON /UCOM2 /

COMMON /UCOMA/ LRPRT ”N53 NSHF. NDAY NEST. NRPRT.K

COMMON /UCOH2/ TAB(2I L KARRAY(2I. 20L SUM(2I) .NSUM(2I)

COMMON /UCOM/ ENLS

INITIALIZE START.END AND OPERATING TIHES FOR DAY I

ENTM-ENT

STT ”-5

SHFND-STTM+NKTM

INITIALIZE DAILY QUOTA ASSIGNED TO THIS EQUIPMENT. RESET MARKERS
AMOUNT-QUOTA

KL-O

NDAY-I

NSHF-O

N '0
DIN-o

U

INITIALIZE DATA STORAGE ARRAYS USED FOR SHORT REPORT
DO IO I-I, ZI

KARRAY I
TAB(I. J)-O. 6

INITIALIZE WAGONS. FIELD TRACTORS. HARVESIERS AND ROAD TRACTORS

oJ)-O

C
CONTINUE
INITIALIZE WAGONS

ATRIBN?I;SEE§T
EEMIZ)
CONTINUE

SET UP LOADING OPERATION AND PREDICT END-OF LOADING
FOR EACH HARVESTER

INHT-NHT

DO 2 O I-I.N

g'LL RHgVE(MFE(2). .2)
CALL COLCT(TIQE.2)

IHSTIBUSFT(I), TNOW, (l+2))
T. GT .0)G 225
Ig-TN0w+20 .OT0

2

IL

230

It:
3

TR
S
L
(
R
R
L

EI‘I(I)

GREnITIIIE
so INTLC 226

225 INHT

.o
T TN ,
égSEEAékkI,I?gP$ARH(8,I)

C
C 230 CONTINUE
C**** SET SHUTDOHN TIHE FOR THIS SHIFT
300 ATRIBilg-SHFND
ATRIB 2

- .0
CALL FILE;(I)
hoo EETURN

SUBROUTINE EVNTS 227
c****
C SUBROUTINE EVNTS(|X)

THIS SUBROUTINE IDENTIFIES THE EVENTS TO BE PROCESSED BASED
ON THE VALUE or ATRI B(2 )

GI T0 (:88,200,300,h00.500,600.700.800),IX

nnn

MO
200
300
#00
500
Mm
700
800

 

 

mxnxnxnxnxnxnxnxg
odrdrdr4F4r4P4r4r
h l— ; l—l ‘— ‘— \— §—'
x x x a,» w x w
w
"1
n
4

2'“,

SUBROUTINE ARFLD 228

c****

nnnnn

C

c****
c****
CAAAA

C
C****
c****

IOO

EAAAA
c****

200

201
C
CAAAA

C
c****

C
CAAAA

C
C****

C
CAAAA

C
cAAAA

2IO

NN
u—ou-o
N—o

SUBROUTINE ARFLD

THIS SUBROUTINE SIMULATES THE ACTIVITIES AT THE FIELD WHEN
A TRANSPORT UNIT RETURNS

COMMON /GCOMI/ ATRIB(25).JEVNT MEA +MEE(IOO). MLE(IOO ). MSTOP, NCRDR,N
INAPO. NNAPT. NNATR,NNEIL.NN3(IOO$NN TRY. NPRNT, 'PPARM(5O, A), TN6N, TTBEG
2 TTCLR, TTFIN.TTRIB(23).TT ET

COMMON /UCOHI/ ISC.B OTA AnogNT, ,ENTN BREAK. STTM NRT, NHT NET, ch.
ICAP.8USHT(2) BUSFT( ).BU§RTé ).TI E. T E RTRIP SHEN6.N

COMMON /CCOM§/ IIEVT,IISED( %.JJB G JJ LR, MMNIT MMON .NNAME(%)’ NNOE
II NNDAY, NNPT, NNSET.NNPRJ,NNP M,NNRNS, NNRUN, NNSTR, NNYR SSEED6

C6MMON /UCOM2 2/ DHN.SMILE
CHECK STATUS OE ARRIVINC VEHICLE) IE ATRIBII2).CT.O REDUCE
QUOTA BY CAP. IE ATRIBIIz).EQ.O VEHICLE IS RETURNING EROM
LOCATION OTHER EACTORY - E6R EXAMPLE THE HORKSHOP.

NH
IF(ATRIB(IZ)) 300,200,100

CALCULATE AMOUNT OF DAY"S QUOTA REMAINING AND
COMPUTE ROUND- TRIP TIME

AMOUNT-AMOUNT-CAP
RTRIP-TNOH-ATRID (T;
CALL COLCT(RTRIP.
ATRI |B(I 2)-o.o
PARK EMPTY NACONS IN QUEUE 2
NII-INT(ATRIB(II))
002 201
ATRIB(?T-TNOH
CALL E LEM
CONTINUE

PARK ROAD TRACTOR IN QUEUE A
CALL FlLEM(h)

CHECK STATUS OE SYSTEM
IE(DHN.EQ.I.O) RETURN

CHECK WHETHER QUOTA IS FULFILLED. IF YES. SCHEDULE SHUTDOWN
IF(AMOUNT.LE.0.0) GO TO 250

CHECK NUMBER OF HARVESTERS IDLE

NI-NN (I)

DO 2 ,NI

HTOC-NFIND(2. 0 W8 I 2‘ I .0)

IE HTOC. GT.O
éF 3NQ(9). .CT. 0) GO 6To 7

TO A

L RMOVE‘HTOC.I)
L EILEM 9)

UE

(9)

-I,N

L Rngm éfMFE(?;, a)

ATRIB(2 99-N99-1

L FILEM (I)

CON NIgUE

K-NHT- -N9

PREDICT START OE LOADING OE HACONS EOR EACH HARVESTER AVAILABLE
CHECK SUEUEM OE EIELD NgRACTORS

IE :NNN%5 figs .NNQ(2)) CD To 211

CO NTSS 212

?E(NQEQ)O) GO TO 220

SUBROUTINE ARFLD 229

DO 2 |-I.N
C**** REMOV N WAGON R0 2M QZ AND A FIELD TRACTOR FROM Q5

C**** SCH E ARRIVAL 0F FIELD TRACTORA Ng gym WAGON AT HARVESTER

=TNOW+ERLNG(5. I)/60. O+PPARM(

2IA§ -TNON+ERLNC(5, I)/60. O+PPARM(8. 3)/60. O

.T

23
H5
‘32
2
I IL
C 215 CONTINUE

C**** CHECK NUMBER OF FULL WAGONS AND DISPATCH A TRIP IF A
E**** COMPLEMENT IS AVAILABLE

22 NII-INT(ATRIB(H )
IE (A). .CE. NII )GO TO 22I
gIg-IN T ATRIB

CALL RTTIMS2HBUSRT(NIO). TNOW. (&+NIO))
RETURN

O

C
C**** SINCE NNQI3).GT.NII, DISPATCH TRIP 0F FULL WAGONS TO FACTORY

c
22‘ Do CAEL RM6VE(MEE( 3)
TIQE -TNON-A BIL;

OLCTITIQF 3

CA
22 TI
éL RMOVE(HEE(A). A)
I
D

N
>n

33:"

CT A RIVAL TIME AT FACTORY
(I
Fl

ATLB {- ATR;D(5)+(PPARM(9. .NNRUN)/RNORM(2. 2))*60. O
N

c**** pR
C

c
E**** SHUT DOWN OPERATIONS EOR TODAY
25 ATRIBEI}-TNOW+I0.0
ATRIB 2 - .0
CALL FILE (n
c NIO-INT(ATRIB(IO))
E**** SET ROAD TRACTOR IDLE

BUSRT(NIO)-0
CALL TIMST(BUSRT(NIO). TNOW. (A+NIO))
GO TO 302

O

-o

§OIL CA fiRROREI 3
0 CA ERROR 2
02 fig

SUBROUTINE STLD

c****

230

SUBROUTINE STLD

 

nnnnnnnnn

Ctttt

C
c****

c****
Cttfit

THIs SUBROUTINE SIMULATEs THE START OE LOADING
I THE IDENTITY OE THE EIELD TRACTOR IS RECORDED As N
I THE HARVESTER Is COMBINED HITH THE ENTITIEs INVOLv D
THE LOADING PROCESS
3) THE END OE LOADING Is PREDICTED

COMMON /GCOMI/ ATRIB(25).J JEVNT MEA MEE(IOO), MLE(IOO).M$TOE, NCRDRN

INAPO. NNAPT. ,NNATR,NNFIL.NN2ET (IOOf, NNTRY, NPRNT, ’EEARM(5O.A). TNOH, TTBEG
2TTCLR, TTEIN WRIB(23).TT%

2COMMON’ /uCOMi/ Isc .2 OTA AH ENT ENTM BREAK STTM NRT, NHT NET,NNG.
ICAP. BUSHT(Z Bu SET (). Bué RT h; .TIQE. TIQE, RTRIP, éHEND .NNTM

RECORD IDEN 3T9 YOF EIE ELD TA

N9-ATRIB(9

HARVESTER IS COMBIN
CALL RHOVE(HFE(6).6

:é- N((ATR?B(8))
BAL MST(BUSHT(N8). TNON. N8

T
4.8 )

CT END OE LOADING BY SAMPLING THE GAMMA DISTRIBUTION
T3 LOAD TI IME
E

N

em

D WITH THE ENTITIES IN THE LOADING PROCESS

gg-TNOH+GAMA(I.1)+PPARH(8,I)
ILER(I)

mxn>>ovng

2m>—i 4*!”

U-fl‘fl” mt-
cr——-!Dr-I—:

 

SUBROUTINE NDLD

cAAAA

Csnsasusaxx

C

nan

C
c****

5 N2-INT(ATRIB) (9)

CAAAA
c****

C
c****
c****

IO

c****
CAAAA

C
CAAAA
CAAAA

'5

C
c****
ca***

25

26

C
CAAAA
c****

30

CAAAA

INAPO.NNAPT, N
2 TTCLR.TTFIN

II IbNNDAY .NNPT

231

SUBROUTINE NDLD
THIS SUBROUTINE SIMULATES THE ACTIVITIES AT THE FIELD
AT THE END OF LOADING

COMMON /GCOM ,MFA, MFE(IOO). MLE(IOO). MSTOP, NCRDR, N

T
O),NNTRY, NPRNT, PPARM(50, A), TNOW, TTBEG
N

T.ENTM, BREAK, STTM NRT, NHT, NET. NNG.
IEE. TISF RTRIP SHEND,H

B LR, MMNIT, MMON, ,NNAME(%)$ NNOE
NRNS, NNRUN. NNSTR, NNYR SSEED6

133?‘

C
V“ cm-‘m—wz’)

I

N

I

9 9T

COMMON /GCOM5 4‘
CMMON IUCOMi '

RECORD IDENTI

A'fl \Z\W\

I AT
CLL EILEM

PREDICT TIM OE ARRIVAL OE EULL HAGON IN QUEUE OE

EULL NAGONS (EILE 3)

ATRIB(A)-TNOH+ERLNG(6. I)/6O. O+PPARH(8. 3)/6O. O
A-ATRI BIA)
CALL EILEM(3)

CHECK NUMBER OF WAGON AND FIELD TRACTOR COMBINATIONS WAITING
FOR LOADING. IF NUMBER.GT.O SCHEDULE STLD FOR ONE

IF(NN8(8)) go
CALL MOVE( E
TIQN-TNOH-ATR
CALL COLCT(TI
ATRIBEI g-TNOW
ATRIB 2
CALL EILEM(I

CHECK NUMBER) OE EMPTY WAGONS AT EIELD AND SET UP
ARRIVAL OE ONE AT HARVESTER IE NUMBER.GT.O

IF(NN (2). E 0 TO 30
CAL MOVE( 2)

g%7.1)+PPARM(8,h)

THE CYCLE TIME IS NOT LESS THAN
HE QUEUE OE EULL NAGONS

;+ GO TO IS
+A-TNOW

(2))
MOVE(

7 I)+PPARM(3.A)
O 26
-TNOW

IF NUMBER OF EMPTY WAGONS AT FIELD. EQ. 0, SET FIELD
TRACTOR IDLE AND COLLECT STATS.

BUSFT(N3M

CALL TI ST(BUSFT(N§), ,TNOH, (2+N ))

Rgco RD ID OE HARVE TER AND HOL IN EILE 6
=INT(ATRIB(8))

SUBROUTINE NOLO

201

c:***
C****

202

20
ctttz
CAAAA
c****

c****

232

BUSHT(NB) mo

CALL TIMST BUSHT(NB), TNow, N8)
CALL EILEM 6

NII-INT(ATRI (II))

IF THERE IS A COMPLEMENT OF FULL WAGONS AND A ROAD TRACTOR
AVAILABLE. DISPATCH A TRIP TO THE FACTORY

IF§NNS(3).GE.NII.AND.NNQ(A).GT.O) GO TO 202

Do EOEL TIRMOVEIMEEQMD

L COLCT(TIQF, 3)

ROAD TRACTOR EROM QUEUE OE ROAO TRACTORS, Qh
VE(MFE h) h
OE DEPARTURE EROM EIELO

IME O ARRIVAL AT FACTOR
ATSIB(5)+(PPARM(9. .NNRUNI/RNORM(2. 2))*60. O

....
2
23m
4.

(I)
TRIB 0))
TRAC OR BUSY ANO COLLECT STATS,
I<BOSRT(NIO). TNOH. (A+NIo))

3383 III)
R 222)

zdofit—w Vmo,m
mo In"! I -II
:IU'I

SOBROUTINE ARHTR 233

 

 

Ctttt
SUBROUTINE ARHTR
Cictih'c
C THIS SUBROUTINE SIMULATES THE ARRIVAL OE A EIELO TRACTOR
C PULLING AN EMPTY VAOON AT THE HARVESTER HHEN IT IS BUSY.
E THE VEHICLE COMBINATION IS HELD IN FILE 8
COMMON /GCOMI/ ATRIB(25), JEVNT MEA,ME E(IOO). MLE(IOO). MSTOP,NCROR,N
INAPO. NNAPT. NNATR, NNEIL, NN2(IOOI, NNTRY Y.NPRNT, 'PPARM(5o’ A),TNOV.TTBEC
2 TTCLR, TTEIN TTRIB(2
COMMON’/ OMI/ ISC, .§UOTAg AMogNT ENTM. BREAK STTM NRT. NHT,NET,NNO,
ICAP. BUSHT(2). BUSFT( ) Bu SRT( ) TIQE.’ IQE,RIRIP, SHENO. NKTM

C
C**** PUT VEHICLE COMBINATION INTO FILE 8
ATRIB(III)-TNO
CALL FILEM(8)w
RETURN
END

 

SUBROUTINE ARECT 234

c****

C
C
C

c****

SUBROUTINE ARFCT

THIS SUBROUTINE SIMULATES THE ARRIVAL OF A TRANSPORT UNIT
AT THE FACTORY AND PREDICTS ITS FACTORY RESIDENCE TIME

2 -2-_____—__=

COMMON /GCDMI/ ATR|B(25). JEVNT MFA, MFE(IOO), MLE(IOO),MSTOP. NCRDR,N
INAPO. NNAPT. NNATR, NNFIL. NNQIIOOI. NNTRY. NPRNT, PPARM(SO,A), TNOW, TTBEG

TTCLR, TTEIN TTRIB(2)
26 OTA gAMOUNT+ENTH BREAK. STTM NRT,NHT, NET, NNO,

OMMON /UCO I/
ICAP. BUSHT(2F, BUSFT(2),B USRT(6) IQE. TIQE, RTRIP.SHEND,NKTM

ATRIB(6)-TNON

PREDICT UNITS RESIDENCE TIME AT THE EACTORY
ATRIB‘I ; EZNOH+GAMA(3, ,3)+PPARM(8, 2)

ATRIB

CALL EIL MI

ATR IB(I) -A RIB(I)-TNON

CALb :IS TO(ATRIB(I). I)

SUBROUTINE DPFCT

c****

n

235

SUBROUTINE DPFCT

=-=

THIS SUBROUTINE SIMULATES THE DEPARTURE OF A TRANSPORT UNIT
FROM THE FACTORY, PREDICTS ITS ARRIVAL AT THE FIELD, RECORDS
THE ARRIVAL IN FILE ;. ’ND LABELS THE UNIT AS HAVING DELIVERED
ITS LOAD:-'(ATRIB(I2 II

---:

 

I
2

I
I

nnnnnn

C

CAAAA
CAARR
Ctkkt

Ctkflk

Ctttk
Ctktk

COMMON /CCOMI/ ATRIB(25).JEVNT MFA.MFE(IOO%.MLE(IOO),MSTOP,NCRDR,N
NAPO,NNAPT.NNATR,NNEIL,NN2(IOOI,NNTRY,NPRN ,PPARM(5O.A),TNOV,TTBEC
TTCLR.TTFIN,TTRIB(28).TT ET .

COMMON /UCOMI/ ISC,B OTA AMogNT ENTM.BREAK STTM NRT,NHT NET,NVC,
CAP,BUSHT(2) BUSFT ) BUSRTé ).TI E,T|2F RTRIP SHFNO,VKTM

COMMON /CCOM§/ IIE .IISEO( ).JJB G JJ LR.MMNIT,MMON NNAME(3) NNOF
I NNOAY,NNPT.NNSET.NNPRJ.NNPRM.NNRNS.NNRUN,NN$TR.NNYR,SSEEO(6i
COMMON /UCOM2/ OVN.SMILE

SET TIME OF DEPARTURE FROM FACTORY

ATR|B(a)-TNOH

LABEL NIT As HAVING OELIVEREO ITS LOAO

ATRIB(I2 -I.O

PREOICT HE ARRIVAL OF THE TRANSPORT UNIT, VHICH IS JUST

LEAVING THE FACTORY, BACK AT THE FIELD
ATRIB213-TNOH+(PPARM(9.NNRUN/RNORM(A,2))*60.0)

ATRIB 2 -I.O

CALL FILEM(I)

SECRELENFORMATION ON ROAO TRACTOR, PROCESSEO AT FACTORY,

CALL FILEM(7)

RETURN

SUBROUTINE SHTDN 236

Ctttt

nnnn

SUBROUTINE SHTDN

THIS SUBROUTINE PROCESSES ALL SCHEDULED EVENTS AND RETURNS
THE ENTITIES ASSOCIATED WITH EACH TO THEIR RESPECTIVE FILES

 

C
c****
C

c****

20

C

c****
c****
C****

2A0

2A1

C
CARAA
CRAAA

I0

I00

IOI

I02

200

300

30I

COMMON /CCOMI/ ATRIB(25), JEVNT MFA, MFE(IOO) .MLE(IOO). MSTOP, NCROR,N
INAPO. NNAPT. NNATR, NNFIL. NN (IOOI, NNTRY, NPRNT, ’PPARMI5O. A). TNOV, TTBEG
2. TTCLR, TTFIN, ,TTRIBIZB), ,TT ET

COMMON /UCOMI/ ISC .2 OTA AMogm ENTM BREAK, STTM NRT, NHT NFT, NVC.
IESSASUSHT(2%5§USFT(W ) BUSRTI TIQE, TIQE, RTRIP. SHFND. VKTM

COMMON /UCOMA/ LRPRT, HN77. NSHF. NDAY, NEST, NRPRT

SET SYSTEM STATUS MARKER

.O
IF(NRPRT. EQ. I) GO TO 20
CHECK OOBEUE OF FIELD TRACTOR ANO EMPTY VACON COMBINATIONS
IF(NNQ(8).GT.O) GO TO 2A0
CO TO 5

PUT EMPTY HACONS ANO FIELD TRACTORS VAITINC IN THE FIELD INTO
THEIR RESPECTIVE FILES
N-NN2(8)
OO 2 I I-I,N

CALL RMOVE MFE(B).8)

CALL FILEM 2;

CALL FILEM 5

a-ATRIB(9)

B SFT(Na)'0.0

CALL TI STIBUSET(N9).TNOV,(2+N9))
CONTINUE

CLEAR THE "FUTURE EVENTS” FILE (FILE I) AND SORT OUT

THE ENTITIES
ngNNQII)
figLLN¥M2V§5MFE)(I). .I)
GO TO 100,200,,200m,200$SOO,SOO.102,102),N2
RTRIP-ATRIB(I)-
CALL COLCT(RTRIP,I
ATRIB(12)-0.0
NII-INT(ATRIB(II))
OO IOI I-I.NII
CALL FILEM(2)
CONTINUE
II?:+III6I“I%‘°”
CALL TIMSTEBUSRT(NIO), ,ATRIB(I), (A+NIO))
CALL EILEM A)
NK-NK+I
IF(NK. LE. N) GO TO IO
GO TO {06
CALL r LEM( )
EBEIIIISIRIE’P”
CALL TIMSTEBUSFT(N9). .TNOV, (2+N9))
CALL FILEM 5)
CO TO I02
CALL FILEM(g)
I: :I-IIISIRI 39”
CALL TIMSTEDUSFT(N9). .(TNOV+2. O). (2+N9))
CALL FILEM g)
ESEIIIN‘IIRA (8))
CALL TIHSTEBUSHT(N8). .(TNOH+2. O), N8)
CALL FILEM 6)
IF(ISC.ER.I) GO TO 102
AMOUNT= A OUNT- CAP
CALL Fl ILEM(2)

SUBROUTINE SHTDN 237

CO TO I02

500 IF(SHFND. LT.ENTM) GO TO 502
CALL FlLEMI?)
NII-INT(ATR B(II))
DO EDI I-I,NII

ALL FILEM(2)

50I CONTINUE
AMOUNTcAMOUNT-CAP
CALL FILEM(h)
NIo-INT(ATRIB(IO))
BUSRT(NIO4-O. .0
CALL TIMS (BU SRT(NIO), TNOH, (A+NIO))
CO TO I02

502 CALL FILEH(9)
GO TO I02

C
C**** THIS SEGMENT OF CODE COLLECTS DATA ON THE OPERATIONS
C**** UP TO THIS POINT FOR THE DESIRED REPORT

700 IF(NRPRT.EQ.I) CO TO 7IO
GO TO 20
c 7Io CALL R PRT
C**** THIS SEGMENT OF CODE SETS UP FOR THE START OF NEXT SHIFT
72o é3(NNQ(9).GT.O) CD TO 205
20 N-NN (_ )
5 006 6Q N

CALL RMOVE MFE
CALL FILE WI (9) 9)

8 F RMAT(/, zo(" ”),"AFTER- SHIFT- END DELIVERIES - ".
IF .2.20('* "))

C**** IF UOTA IS SATISFIED ORY END OF SHIFT TIME HAS COME. SUSPEND
C**** OPE ATIONS UNTIL NEXT DAY
9 IFEAMOUNT. LE. 0) GO TO 03
IF TNOH GE. ENTM) GO TO 3
C**** SET START OF NEXT SHIFT
ATRIB(I)-SHFND+BREAK
C**** SCHEDULE END OF NEXT SHIFT
SHFND-ATRIB(I)+HKTM
GO TO A
C**** SCHEDULE START F SHIFT ON THE NEXT DAY
3 ATRIB(I)-STTM+I ho

STTM-ATRIBII)
AMOUNT- UOTA+AMOUNT
C**** SCHEDUL END TIME ON NEXT DAY
ENTM-ENTM+Ihh0
SHFND=STTM+NKTM
A ATRIB(2)-8.
CALL FILEM(I)

C
C**** FIX TIME OF NEXT SHUTDOHN EVENT
ATRIBEI g-SHFND
2 'M' 0
CALL FILE (I)
730 EAIERERROR(3O)
RETURN

SUBROUTINE DSTART 238

 

C****

SUBROUTINE DSTART(I,K)
c— - TNTE SUBROUTINE RESETS THE TIME or ARRIVAL IN QUEUES 2 AND 3
E AT THE START or EACH SHIFT

 

 

COMMON /ccon1/ ATR|B(25).JEVNT MFA,MFE(IOO).MLE(IOO),MSTOP,NCRDR,N
INAPo.NNAPT,NNATR,NNrIL,NN2(Ioof,NNTRY,NRRNT,RRARN(50,A),TN0N.TTBEG
2,TTCLR,TTFIN,TTRIB(25).TT ET

Eg%:9éé{0) RETURN

I J-I,N
CALL RHOVE(HFE(I).I)
ATRIa(K)-TNow
CALL FILEH(I)
I CONTINUE
RETURN
END

SUBROUTINE STRTUP 239

 

 

 

 

c****
c SUBROUTINE STRTUP __
C THIS SUBROUTINE SIMULATES THE START UP ACTIVITIES AT THE START
C OF EACH SHIFT ANO RESETS SPECIAL MARKERS USED IN THE PROGRAM
COMMON /CCOMI/ ATRIB(25).JEVNT, MFA ,MFE(IOO). MLE(IOO). MSTOP, NCROR,N
INAPO. NNAPT. NNATR.NNFIL.NN (IOO), NNTRY. NPRNT, PPARM(5O, A). TNOH, TTBEG
2, TTCLR, TTFIN .TTRIB(25), T ET
COMMON /CCOM5/ IIEW W.IISED(6).JJBEG JJCLR, MMNIT. MMON, NNAME( ) NNOF
II. NNOAY. NNPT NNSET, NNPRJ,NNPRM NNRNS, NNRUN, NNSTR, NNYR, SSEEO6 6)
COMMON /UCOMT/ ISC, ,BUOTA AHOENT, ENTM, BREAK STTM NRT, NHT NFT. ch.
ICAP. BUSHT(2).BUSFT( )iBuSRT( ) TIQE, TIQF, RTRIP, SHFND. NNTM
COMMON /UCOM2/ OHN.SMIL
c COMMON /UCOMA/ LRPRT. N77, NSHF, NOAY, NEST, NRPRT
gAttt RESET SYSTEM STATUS MARKER
c URN-0.0
E**** RESET GASPIV OATA STORAGE ARRAYS
c CALL CLEAR
E**** RESET TIME OF ARRIVAL OF VEHICLES IN QUEUES 2.3 ANOA
CALL OSTARTz 2.2;
c CALL OSTART 3,
E**** SET UP LOAOINC OPERATION FOR EACH COMBINE HARVESTER AVAILABLE
N6-NN ‘
FN6. .Egfl OTO 500
IF NN8(|_I )E :0) CO TO IOI
00 IO
CALL RMOVE(MFE(2). 2)
TIQE- O.O
CALL COLCT TI2E2)
CALL RMOVE MF W?) ,5)
Na; INTIAT
SFT(N)
CALL TI ST}BUSFT)(N2), ,TNOH, (2+N9))
IO CALL RMOVEM Mg
N8-INT(AT B( ))
BUS HT N8T -I
ATRIB I)-TNOH+CAMA(I, ,I)+PPARM(8. I)
ATRIB 2 -g.o)
CALL FILE El)
CALL TIMST BUSHT(NB),ATRIB(1),N8)
c IOO CONTINUE
C**** OISPATCH As MANY TRIPS TO THE FACTORY AS THERE ARE COMPLEMENTS
E**** OF 2 HACONS PLUS A ROAO TRACTOR AVAILABLE
$85 IF(NN3(h)) 501,205,200
Nll-?N ééATRIB( II))
C**** CALCULA E NUMBER OF 2- NACON TRAINS AVAILABLE
€26 100,205.10
I02K FNNQ h) LT. K) GO TO 201
CO TO 202
201 NI-NN (A)
202 no 20 I-I NI
OO ZOEJ
$TEF _SH8VE(HFE(3) 3)
CALL COLCT(TIQF. 3)
203 CONTINUE

CALL RMOVEIMFEIA), A)

ATRIB 5 -TNON+ .O

ATRIB I =ATRIB 5)+(PPARM(9,NNRUN)/RNORM(2,2))*60.0
ATRIB 2: =3. .0

CALL FILE (I)

NIO=INT(ATRIB(I0))

BUSRT(NIO)-I .O

SUBROUTINE STRTUP 240

C?thTIMST(BUSRT(NIO),TNOW,(h+NIO))

c 20A CONT
Eziiim IF NFTi GT. NHT, SET UP FIELO TRACTOR ANO EMPTY NACON COMBINATIONS
C
205 IF§:N ngjfi $.0Tw GO TO ogoo
IF NN 00
IF NN .NNQ(5)) on TO 3Io
CO "TB

3}? SBNN?2(§)I
EALL: RH6VE(HFE)(2). 2)

TIQE '0. 0
CALL COLCT(TIE

wa NT(ATR
B WPT(N3)-
ALL TI ST (BUSFT(N2h TNON (2+Ng))
ATRIBElg-TN OH+ERLN .I)/6O O+ PARM(8, 3)/6O. O
ATRIB 2 -A.o
CALL FILEM(o )
I2 CONTINUE
OO IF(NRPRT.EQ. I) GO To 302
HRITE(6.I) AMOUNT. TN NO
FORMAT(/ O.”THE UOTA FOR TOOAY Is " .FIO. A .5x. "AT", EIO. A)
IF(LRPRT) 00, 502, 02
600 CALL PRNTQ(0)
GO TO 302
O CALL E ROR RIO)
0 CALL ERRORZ
OI CALL ERRORZ
oz RETURN
ENO

SUBROUTINE REPRT 241

 

 

 

Ctktt
Ckin'ci:
c SUBROUTINE REPRT
C THIS SUBROUTINE IS USED TO COMPILE OATA EOR A REPORT AT THE END
C OE EACH SIMULATION RUN. EACH RUN SIMULATES I2 CONSECUTIVE
E-- SHIFTS (OR ONE NEEK'S OPERA ION
COMMON /CCOMI/ ATR|B(25).JEVNT MEA, MEE(IOO). MLE(IOO).M$TOE, NCROR,N
INAEO, NNAPT. ,NNATR,NNFIL,NNE(IOO O), NNTRY. NPRNT, 'ERARM(5O,A),TNON. TTBEG
2TTCLR, TTEIN TTRIB(25).TT ET
:COMMON’ /CCOM§/ IIEVT,||SED(6),JJBEG JJCLR, MMNIT,MMON,NNAME( ), NNOE
NNDAY, NNPT NNSET,NNPRJ,NNRRM NNRNS, NNRUN, NNSTR,NNYR,SSEE06
COMMON /CCOM6/ EEN2(IOO) IINN(TOO ,KKRNK(I60), MMAxO(IOO),
188T|H(IOO)M ,SSOBV(2 ,3),S§TPV(2 ,6 ,VVN (I OO
MMON MI/ ISC,8 OTA AMOUN ’ENTMB EAK STTM NRT,NHT,NFT, NNC,
ICAP. BUSHT(2).BUSFT( ).au SRT(6) .TIQE, TIQE. RTRIP, SNENO,NKTM
COMMON /UCOM / JOT KL
COMMON /UCOM / LRPRT N73 NSNE, NDAY. NEST, NRPRT
c COMMON /UCOM5/ TAB(zi, S. KARRAY(21, 20), SUM(2IL NSUM(2I)
3T7N77. GT 0) GO To IO
6 N7-NN (7)
IO N -NN ()-N
zoN NAE-N Hz+l 77

KARRAY I, KL INDAY
KARRAY 2 KL INSHF

:?F(SSO§V(J, 3) .CT. O. O) CO TO 29
29G 0“:32“ KL)-ssoav(J I)/ssoav(J. 3)

SUM 2+J)-TAB(2+J. KL)+SUM(2+J)
3O CONTIN
DO AO J-I
IF(SSTPVJ J,$;-TTCLR. .LE.o) 60021
XS-SSTPV L +SSTPV(J.6)*(TNOW- S TEV(L 3))
xT-TNON- ~TTC

TAB(6:J:KL)'XS/XT

I TA832+ KL) I
32 SUM 6+Ji‘TAB(é+J, KL)+SUH(6+J)

KARRAY(] 7.KL)-N 7
NSUM(I7)-NSUM(I +KARRAY(I7, KL)
KARRAYé18,,KL)-NN
NSUM(I )-N SUM(I8)+2 ARRAY(I8. KL)
KARRAY(I9, KL)-NNQ
NSUM(I%)-N SUH(I§)+KARRAY(19, ,KL)
KARRAY 20.KL)-N WK
gsg9(20)-NSUM(ZO)+ ARRAY(20, KL)
TAB‘ I KL)-B*CAP
an? Zfii-SUM(21)+TAB(21, KL)
CE<SEF3°p GE. ENTM. 0R. AMOUNT. LE. 0) GO TO 60
60 NDAY-NgAY+I
NSHE-O
70 RETURN

END

SUBROUTINE OTPUT 242

 

 

 

Cinhh':
Chum
c SUBROUTINE OTPUT
C THIS SUBROUTINE IS USED TO PRINT OUT A TABLE OF SUMMARIZEO OATA
E AT THE END OE EACH SIMULATION RUN
COMMON /CCOMI/ ATRIB(25).JEVNT.FA,MFE(100) MLE(IOO). MSTOP, NCROR.N
INAPO,NNAPT,NNATR.NNFIL.NN8(100 ),N NTRY, NPRNT, ’PPARM(5O' A), TNON, TTBEG
2.TTCLR.TTEIN.TTRIB(25).TT ET
COMMON /OCOM5 IIEVT.IISEO(6).JJB EG JJCLR. MMNIT. MMON, NNAME(3), NNOE
II,NNOAY.NNPT NNSET,NNPRJ.NNPRM.NN NRNS, KNNRU N, NNSTR, NNYR. SSEED 0(6)
C MMO /CCOM6/ EEN2(100) IINN()003.%RNK(160). MMAqu(IO o).
I TIN(IOO).SSOBV(2 .a).sSTPV(25.6T .VVN OO
MON /UCOMI/ ISC,2 OTA.AMOUN ENT M. B EAK. STTM, NRT. NHT NEL NNC.
ICAP.BUSHT( ).BUSFT( ).Bus SRT(6).TIQE. IQE. RTRIP. SHENO, NK TM
MON /UCOM2/ JOT. KL
COMMON /U C PR RT,N Z5,N NSHE. NO Y. NE ST NRPRT
c COMMON /UCOM5/ TAB(ZI ). KARR AY( (2 .2 o), SUM(2I). NSUM(2I)
IE(NRPRT.NE.I) RETURN
Nc-KL
z-KL
KILOM-INT(PPARM(9. NNRUN))
NRITE 6. o)
NRITE 6. ) NEST, KILOM. .SUOTAM ,NNG, NHL NFT. NRT
NRITE 6. O) KARRAY gJ NE;
NRITE 6,15 KARRAY 2, 'J J-I. NC
VE-SUM(3 {z
NRITE(6,2O (TAB(3,J).J-I.NC).AVE
AVE-SUM(A){Z
WRITE(6.2? (TAB(A.J).J-I.NC).AVE
AVE-SUM( {z
NRITE(6, O (TAB(S. J). J-I.NC).AVE
AVE=SUM( )/z
IF(NFT GT .I) NRITE(6. 35) (TAB(6.J).J-I.NC).AVE
I00 00 2I I-I.NHT
AVE-SUM(6+I I)/z
21 VRITE(6.AO) I, (TAB(6+I, J). J-I ,Nc). AVE
00 21 I=I.NET’
VE=SUM(8+I m)
31 NRITE(6.A5) I. (TAB(8+|, J). J-I .NC). AVE
OO AI I=I.NRT'
AVE=SUM(IO+I)/
AI NRITE(6.5O) I,(T AB(IO+I. J). J-I .NC). AVE
200 NAVE-NSU M(I7)/NC
NRITE(6. 55) (KAR RAY(I7. J). J-I. NC). NAVE
NAVE=NSU UM(18)/Nc
NRITE(6,6O) }KARRAY(18. J). J-I .NC). NAVE
NAVE-SU éI9) NC
NRITE(6. 5 (KAR RAY(I9. J). J-I .NC), NAVE
NAVE=NSU UM(2O)/NC
WRITE(6,ZO (KAR RAY(2O. J). J-I NC). NAVE
3OO AVE=SUM( I /z
WRITE(6,Z5 (TAB *gzI. J) -I. .NC)A
90 EORMAT(T O.80("w /J20. "N" 50,"ENO OE SHIET REPORT“. T99. "N".
)/T20.80(”*"))
NRITE(6.80)
80 EORMAT(/. T5,"NOAY_ - OAY NUMBER". T60. "URT# = UTILIZATION OE ROAO
TRACTOR #"/, 5 "NSHF -SHIET NUMBER". T60. "TRPTS- UMBR
2" TRIPS COMPLE EO TH SHIFT“/Ta WAT PT- Av E RI P TI ME“. T60.
£"TRPIT - NUMBER OF TRIPS IN TRA SIT AT ENO OE SHIET"/T5."TI
AVE TIME AN EMPTY NACON NAITS AT THE EIELO' '.T6o ."N ENG . N MBER".
" OE EMPTY NACONS AT IELO"/T12."TO BE LOADED". T60.“NENC=
NUMBER OE EULL NACONS AT EIELO"/T5 5."TIQE - AVE TIME A EULL NACON
7W$:¥S(ET8;HE FIELD”)
85 EORMAT(T3.TUHT# - UTILIZATION OF HARVESTER #". T60. "TNNESO=
OTAL TO NES OELIVEREO "/T5. "UET# = UTILIZATION OE EIELO TRACTOR
5 EORMAT(//.TIO."ESTATEN ." .Ton "s" I3. I.TA2m "D'ISTANCE TO EACTORY = "
" KILOMETERS" WTSS. IL ' TONNES". //. TA5.
2 E UIPMENT COMBINA TION” J “60 A "N"A .I3))
Io EORMAT /Ix.“NOAY “.2OI )
I5 EORMAT /Ix."NS HE “.20I " MEANS")
20 EORMAT /Ix."ATRPT ".2I 5. I)
25 EORMAT /Ix."TI QE “.2IE5.i)

SUBROUTINE OTPUT 243

o EORMAT /Ix,"TI E ",Zng.I;

g EORMAT /lX."T| w " 21F .I
EORMAT /lx."UH " IT,Ix,2IE .I

o EORMAT /Ix,"URT".II.Ix.2IE .I
EORMAT /lX,"UFT" II,Ix 2IE .I

g EORMAT /Ix."TRPT ",Ix.iII
EORMAT /lx."TRPlT" Ix zII

6 EORMAT /Ix,"NEwC".Tx.2II

; EORMAT /Ix."NENC" Ix zII

5 EORMAT /Ix."TNNEsb",Tx,2 E5.O)
RETURN

END

244

NON-GASPIV DATA INPUT:

ISC
NWG
NHT
NFT
NRT
NEST
QUOTA
CAP

dNN—Om—n

200.00 TONNES
10.00 TONNES

660.00 HINS
60.00 HINS
300.00 HINS

ENT
BREAK
WKTM
NRPRT I I
LRPRT - I

245

mo¢w0vv.

FO+w000m

 

nO+uOO...
.OAuOCnn.
.O+wOOO..
.oouobnm.
.C¢w##¢..
.O+wmm0q.
.o.ucCOv.
.04wowmm.
.oowOOOn.

u0~III n0+wOO...

0 u0<4u2 0
u>u#22 v Na#m22 0

nbzafinn 20—mmw> >~ amdu

#

>w>m41 .3

u2~u## .0

«.va n0.nv
n000## 0 ”Dav—u .
.0+00000. .040000m.
N0+w000.. #0+w0m00.
n0+unn0.. .0
n0.w00,n. .0
n0+womnv. .0
n0+mm.wn. m0+w.nm#.
m0+uvn... .0
N0+wn0vn. .0+w0vvw.
N0+umvm.. .0
0 — . n n
0 v. 0 . n— n—
u304II m? u4wUZZ
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
um0w22 0 "00022 0000’
u#4n22 m NSQQZZ #

awmxbz 20“ N0

«000??

.0+uOOOn.
.O¢w0#m#. n
.OIwoowm. u
.OIwownn. u
.OIuohmp. u
~O+w#wnw.. u
.O+wwONm. u
n0+wvbnw. u
.0+woo'o. n

n

0'

UCJUCJUCJUCJUCJ

H~~~~~H~~~

u#mm22
um—IZZ

.o..00000000000ma:m44
\m. \ow mpqo

>0 00. ameDZ #0010ua 20~#<402~m

_0nvv

# nu4011

"NNQLDLDFOO’?

n n
v m
#0u.#auIm<44

.m.»am:mu»m<44
.mvkamamupm444
.v.»um:mu»m<44
.n.#aw:mupm<44
An.»um:mupm<44
...#um:mu»m<44
.u.pum:muhm<44
..vpumamu#m<44
.vaxmamupm444
..vpzmnmu»m¢44

)0—#M00<44
u0~#u00<44
w0~#u00<44
a~a#muum<44

0 N4~u22
0w u<#m22

.mpnm noum__
e ”4:2.
# MQO#mI
#um aw#w£<&<Q
#uw auhuzqaca
#uw mu#m5<aqa
#ufi Uw#w§du<a
#wm mu#ui<aqa
#mv mw#u2<a<a
#wm aw#wzcada
#um au#widada
#um uw#w2<u(a
A. v 22-
A. v Nxzuxx
# .02 0#m~I
0p .02 #m£_#
m .02 #m2~#
m .02 #m2~#
h .02 #m2_#
0 .02 #m2_#
m .02 #m2~#
v .02 #m2~#
n .02 #mS—#
N .02 #m2~#
# 02 #m!—#
V 02 #0400
n .02 #0400
N 02 #0400
w 02 #0400
v# nm#<22
v N#4022

LIST OF REFERENCES

ASAE (1982). Uniform Terminology for Agricultural Machinery Management.
ASAE Standard: ASAE S 322. Agricultural Engineers Yearbook, pp.
207-208. ASAE, St. Joseph, Michigan.

Audsley, E. and D.S. Boyce (1973). Exact Solutions for Cyclic Transport
Systems. J. Agric. Engrg Res. 18: 217-230.

Barbados Sugar Review, No. 41 - September (1979). Barbados Sugar
Producers Association (Inc.), Warrens, St. Michael.

Barbados Sugar Review, No. 47 - December (1981). Barbados Sugar
Producers Association (Inc.), Warrens, St. Michael.

Baxter, S.W.D. (1969). Preparing for Mechanical Harvesting. Sugar
J. Azucar, September, pp. 29-34.

Bhattacharyya, G.K., and R.A. Johnson (1970). Statistical Concepts
and Methods. Publ. John Wiley & Sons, New York, pp. 233-275,

Bouland, H.D. (1967). Truck Queues of Country Grain Elevators.
Operations Research 15: 649-659.

Boyce, D.S. (1972a). Systems and Symbolic Models in Operational Re-
search and Systems Engineering. Proc. of Sym. of Systems Appli-
cations in Agricultural Engineering, Silsoe, England, pp. 1-14.

Boyce, D.S. (1972b). Probablistic Simulation of Field Harvesting
Systems as Closed Circuit TransPort Systems. Proc. of Sym. of
Systems Applications in Agricultural Engineering. Silsoe,
England, pp. 67—83.

Bridges, T.C., 0.J. Loewer, J.N. Walker, and D.G. Overhults (1979).
A Computer Model for Evaluating Corn Harvesting, Handling,
Drying, and Storage Systems. Transactions of the ASAE, Vol. 22
(3), pp. 618—621, 629.

Brooks, A.C., and L.R. Shaffer (1971). Queuing Models for Production
Forecasting on Construction Operations. Civil Engineering
Studies. Construction Research Series No. 16, Dept. of Civil
Eng., University of Illinois (as quoted by Ogilvie et 31., 1978).

246

 

247

Chase, K., and B.W. Eavis (1972). Studies on the Effects of Burning
Sugar Cane. Barbados Ministry of Agriculture, Science and
Technology, Bulletin No. 53.

Cochran, B.J., and R.W. Whitney (1975). Economies of Transporting
Sugar Cane to Mill. Sugar J. 38: 10-14.

Coupland, G.A., and R.M. Halyk (1969). Critical Path Scheduling of
Forage Harvesting Systems in Quebec. ASAE Paper No. 69-678,
St. Joseph, Michigan.

Demichele, D.W. (1978). Potential Uses of Mathematical Models in
Agricultural Research and Production. ASAE publication, "An
American Success Story - Increasing Agricultural Productivity,"
pp. 59-77, St. Joseph, Michigan.

Dent, J.B., and M.J. Blackie (1979). Systems Simulation in Agriculture.
Applied Science Publishers Ltd. London.

Dumont, A. and D.S. Boyce (1972). Methods of Fitting Statistical Dis-
tributions to Observed Work Times for Agricultural Unit Operations.
J. Agric. Eng. Res. 17: 295-302.

Early, A.C. (1974). Digital Simulation of a Sugar Cane Producing-Pro-
cessing District in the Philippines to Test Alternative Irrigation
Policies, Harvesting Policies, and Irrigation Capacities. 15th
Congr. ISSCT, pp. 849-862.

Farquhar, R.H. (1972). Computer Modelling for Improvement of Sugar Cane
Agriculture. Proc. 39th Conference of Queensland Society of Sugar
Technologists, pp. 74-80.

Hahn, G.J. and 5.5. Shapiro (1967). Statistical Models in Engineering,
Publ. John Wiley and Sons, Inc.

Harvey, W. O'N. (1982). Research into Mechanical Harvesting and Handling
of Sugar Cane in Barbados. Memorandum submitted to the Sugar
Policy Research Unit, Ministry of Finance and Planning, Barbados.

Hillier, E.S., and G.J. Lieberman (1974). Operations Research. Publ.
Holden Day, Inc., pp. 378-434, 521-656, San Francisco.

Hoekstra, R.G. (1973). Use of Computer Simulation in Predicting the
Effect of Mill Stoppages on Cane Transport Fleet Utilization for a
Cane Handling Installation Using a Spiller. Proc. SASTA, pp. 61-69,
South Africa.

Hoekstra, R.G. (1974). Simulation of Mill Yard Operations by Computer.
Proc. 15th Congr. ISSCT, pp. 1736-1753.

Hoekstra, R.C. (1975). How Delay Times are Affected by Various Opera-
tional Conditions. Proc. SASTA, pp. 29-40, South Africa.

248

Hudson, J.C. (1972). Fire, Water and Sugar Production in Barbados.
Barbados Sugar Producers Association (mimeo).

Krasnow, Howard 8., and Reino A. Merikallio (1964). The Past, Present,
and Future of General Simulation Languages. Management Sci. 11:
236-267.

Link, D.A., and W.E. Splinter (1970). Survey of Simulation Techniques
and Applications to Agricultural Problems. Trans. A.S.A.E. 13:
837-843, St. Joseph, Michigan.

Loewer, Otto J., L.A. Balastreire, A.P. Cobra, and D.A. Ometto (1979).
Optimization of Sugar Cane Harvesting Systems. ASAE Paper No.
79-1532. St. Joseph, Michigan.

Manetsch, Thomas J. and Gerald L. Park (1980). System Analysis and
Simulation with Applications to Economic and Social Systems.
Dept. of Electrical Engineering and Systems Science. Michigan
State University, E. Lansing, Michigan.

Meliza, R.S. (1966). Theory of Queues Applied to Measuring Production
Rates of Pusher-Scraper Fleets. Dept. of Civil Engineering.
University of Illinois at Urbana.

McGregor, Andrew, S.W.D. Baxter, G.B. Hagelberg, and Malcolm McGregor
(1979). The Barbados Sugar Industry - Problems and Perspectives.
Report to the Director of Finance and Planning, Government of
Barbados Bridgetown.

Naylor, Thomas H., et a1. (1966). Computer Simulation Techniques.
John Wiley and Sons, New York, pp. 1-20.

Naylor, T.H. (1971). Computer Simulation Experiments with Models of
Economic Systems. Wiley, New York.

Nurse, James 0.J. (1978). The Future of Sugar Cane Production in
Barbados. Master of Professional Studies Project Report, Cornell
University, New York.

OECD/OCDE (1975). Food Consumption Statistics, 1955-1973. Paris,
France.

Ogilvie, J.R., Colin 0. Lee, and Herman Zijp (1978). Simulation Model-
ling of Whole Sugar Cane in Jamaica. ASAE Paper No. 78-5028.
St. Joseph, Michigan.

O'Shea, J.B., G.N. Slutkin, and L.R. Shaffer (1964). An Application of
the Theory of Queues to the Forecasting of Shovel—Truck Fleet
Productions. Civil Engineering Studies, Construction Research
Series, No. 3, Dept. of Civil Engineering, University of Illinois.

Page, E. (1972). Queuing Theory in OR. Crane Russak & Inc., New York.

249

Persaud, B. (1973). An Economic Study of the Barbados Sugar Industry.
Ph.D. Diss., University of Reading, England.

Phillips, D.T., A. Ravindran, and J.J. Solberg (1976). Operations
Research Principles and Practice. John Wiley and Sons, Inc., pp.
1-11, 289-320, 359-418.

Pritsker, A.A.B. (1974). The GASP IV Simulation Language. John Wiley
and Sons, Inc., New York.

Pritsker, A.A.B., and Phillip J. Kiviat (1969). Simulation with GASP
II. Englewood Cliffs: Prentice Hall.

Rockwell, T.H. (1965). Use of Simulation Methodology for Solution of
Operational System Problems. Sym. Proc. Systems Engineering in
Agriculture. ASAE Proc. - 567, pp. 4-8, St. Joseph, Michigan.

Russel, D.G., K.W. Lievers, and J. Lovering (1977). Effects of Change
in Rate of Harvest and Selected Management Variables on Timothy
Silage Yield, Quality and Net Return at Charlottetown. C.S.A.E.
publication. Canadian Agricultural Engineering 19: 29-36.

Sasty, T.L. (1957). Resume of Useful Formulas in Queuing Theory. The
J. of the Operations Research Society of America 5: 161-200.

Shukla, L.N., T.S. Chisholm and A.L. Phillips (1973). Computer Program
for Analyzing Harvesting, Loading and Transportation of Sugar Cane.
TRANSACTIONS of the ASAE, 16(3): 420-425.

Sorenson, E.E., and J.F. Gilheany (1970). A Simulation Model for Har-
vest Operations Under Stockastic Conditions. Management Science
16: B549-B565.

Sowell, R.S., T. Liang, and D.A. Link (1967). Simulation of Expected
Crop Returns. ASAE Paper No. 67-614.

Stapleton, H.N. (1967). Crop Production System Simulation. ASAE Paper
No. 67-562, St. Joseph, Michigan.

Teichrow, D. and J.F. Lubin (1966). Computer Simulation - Discussion
of Techniques and Comparison of Language. Communications of A.C.M.
9: 723-741.

Tocher, K.D. (1965). Review of Simulation Languages. Operations
Research Quarterly 16: 189-218.

Tomlinson, F. (1973). The Caribbean Sugar Industry: Planning for the
Future, Proc. W.I.S.A. Sugar Technol., Barbados, pp. 35-51.

Tonsman, J.A. (1974). The Sugar Cane Scheduling System. Unpubl. IBM
del Peru, South America.

250

Younger, J.A. (1979). The Development and Use of Cane Harvesters in
Australia and Other Major Sugar Producing Countries. ASAE Paper
No. 79-5513.

GENERAL REFERENCES

Bagchi, J.P., and J.G.C. Templeton (1972). Numerical Methods in Markov
Chaine and Bulk Queues. Springer-Verlag Berlin. Herdelberg. New
York.

Barbados Sugar Review. Nos. 38, 39, 43, 44, 45, and 46 (1978 and 1980).
Barbados Sugar Producers Association (Inc.), Warrens, St. Michael.

Barbados Sugar Industry Review. Nos. 14 through 35 (1972 through 1978).
Barbados Sugar Producers Association (Inc.), Warrens, St. Michael.

FAO Production Year Book (1979). FAO, Rome. Vol. 33, pp. 167-169.

Horst, A.E., and Helmut Von Frazer (1977). Operations Research Handbook
(standard algorithms and methods). Walter de Gruyter, Berlin. New
York.

Steel, Robert G.D., and James H. Torrie (1980). Principles and Procedures
of Statistics. A Biometrical Approach. McGraw-Hill Book Company,
New York.

Sugar Year Book (1970). The International Sugar Organization, 28 Hay-
market, London, S.W.L., pp. 350-358.