T H R D E D T 0 N

Simulation Using
ProModel

Dr. Charles Harrell

Professor, Brigham Young University, Provo, Utah
Director, PROMODEL Corporation, Oram, Utah

Dr. Biman K. Ghosh, Project Leader

Professor, California State Polytechnic University,
Pomona, California

Dr. Royce O. Bowden, Jr.

Professor and head, Mississippi State University,
Mississippi State, Mississippi

Connect
i Learn
1 Succeed
C 0 N T E N T S

PART I 2.8 The Systems Approach 46

2.8.1 Identifying Problems and
Study Chapters
Opportunities 47
2.8.2 Developing Alternative
1 Introduction to Simulation 3 Solutions 47

1.1 Introduction 3 2.8.3 Evaluating the

1.2 What Is Simulation? 5 Solutions 48

1.3 Simulate? 6 2.8.4 Selecting and Implementing

Why
1.4 Simulation 8 the Best Solution 48
Doing
1.5 Use of Simulation 10 2.9 Systems Analysis Techniques 48
1.6 When Simulation Is 11 2.9.1 Hand Calculations 50
Appropriate
1.7 for 2.9.2 Spreadsheets 50
Qualifications Doing
Simulation 13 2.9.3 Operations Research
1.8 Conducting a Simulation Study 14 Techniques 51
1.8.1 2.9.4 Special Computerized
Defining the Objective 15
1.8.2 Planning the Study 18 Tools 54

1.9 Economic Justification of 2.10 Summary 55

Simulation 18 2.11 Review Questions 55
1.10 Sources of Information on References 56
Simulation 22
3 Simulation Basics 57
1.11 How to Use This Book 22
1.12 Summary 23 3.1 Introduction 57
1.13 Review Questions 24 3.2 Types of Simulation 57
1.14 Case Studies 25 3.2.1 Static versus Dynamic
References 30 Simulation 58
3.2.2 Stochastic versus
2 System Dynamics 33
Deterministic

2.1 Introduction 33 Simulation 58

2.2 Definition 34 3.2.3 Discrete-Event versus

System
2.3 Elements 35 Continuous Simulation 59
System
2.3.1 Entities 36 3.3 Random Behavior 61

2.3.2 Activities 36 3.4 Simulating Random Behavior 63

2.3.3 Resources 36 3.4.1 Generating Random

2.3.4 Controls 37 Numbers 63

2.4 37 3.4.2 Generating Random

System Complexity
2.4.1 Variates 68
Interdependencies 38
2.4.2 3.4.3 Generating Random
Variability 39
2.5 Performance Metrics Variates from
System 41
2.6 Variables 43 Common Continuous
System
2.6.1 Decision Variables 43 Distributions 71

2.6.2 Variables 3.4.4 Generating Random

Response 44
2.6.3 State Variables 44 Variates from

2.7 44 Common Discrete

System Optimization
Distributions 73
 Contents vii

3.5 Simple Spreadsheet Simulation 74 4.8 Summary 112

3.5.1 Simulating Random 4.9 Review Questions 113
Variates 75 References 114

3.5.2 Simulating Dynamic,

Stochastic Systems 79 5 Data Collection and Analysis 115
3.5.3 Simulation Replications and
5.1 Introduction 115
Output Analysis 82 5.2 Guidelines for Data 116
Gathering
3.6 Summary 83 5.3 Determining Data
3.7 Review Questions 83
Requirements 118
References 85
5.3.1 Structural Data 118
5.3.2 Operational Data 118
4 Discrete-Event Simulation 87
5.3.3 Numerical Data 119

4.1 Introduction 87 5.3.4 Use of a Questionnaire 119

4.2 How Discrete-Event Simulation 5.4 Identifying Data Sources 120

Works 88 5.5 Collecting the Data 121

4.3 A Manual Discrete-Event Simulation 5.5.1 Defining the Entity
Example 89 Flow 121

4.3.1 Simulation Model 5.5.2 Developing a Description

Assumptions 91 of Operation 122

4.3.2 Setting Up the 5.5.3 Defining Incidental

Simulation 91 Details and Refining Data

4.3.3 Running the Values 123

Simulation 94 5.6 Making Assumptions 124

4.3.4 Calculating Results 99 5.7 Statistical Analysis of Numerical
4.3.5 Issues 102 Data 125

4.4 Commercial Simulation 5.7.1 Tests for

Software 102 Independence 127
4.4.1 Modeling Interface 5.7.2 Tests for Identically
Module 102 Distributed Data 132
4.4.2 Model Processor 103 5.8 Distribution Fitting 134
4.4.3 Simulation Interface 5.8.1 Frequency
Module 103 Distributions 135
4.4.4 Simulation Processor 104 5.8.2 Theoretical
4.4.5 Animation Processor 104 Distributions 138
4.4.6 Output Processor 105 5.8.3 Fitting Theoretical
4.4.7 Output Interface Distributions to Data 142
Module 105 5.9 Selecting a Distribution in the

4.5 Simulation Using ProModel 106 Absence of Data 148

4.5.1 Building a Model 106 5.9.1 Most Likely or Mean
4.5.2 Running the Value 149
Simulation 106 5.9.2 Minimum and Maximum
4.5.3Output Analysis 107 Values 149
4.6 Languages versus Simulators 109 5.9.3 Minimum, Most Likely, and
4.7 Future of Simulation 111 Maximum Values 149
viii Contents

5.10 Bounded versus Boundless 7 Model Verification and

Distributions 151 Validation 193
5.11 Modeling Discrete Probabilities 7.1 Introduction 193
Using Continuous Distributions 151
7.2 Importance of Model Verification and
5.12 Data Documentation and
Validation 194
Approval 152
7.2.1 Reasons for Neglect 194
5.12.1 Data Documentation
7.2.2 Practices That Facilitate
Example 152
Verification and
5.13 Summary 155
Validation 195
5.14 Review Questions 155
7.3 Model Verification 196
5.15 Case Study 158
7.3.1 Preventive Measures 197
References 159
7.3.2 Establishing a Standard for

Comparison 198
6 Model Building 161
7.3.3 Verification
6.1 Introduction 161 Techniques 198
6.2 Converting a Conceptual Model to a 7.4 Model Validation 202
Simulation Model 162 7.4.1 Determining Model
6.2.1 Modeling Paradigms 162 Validity 203
6.2.2 Model Definition 164 7.4.2 Maintaining
6.3 Structural Elements 165 Validation 205
6.3.1 Entities 165 7.4.3 Validation Examples 205
6.3.2 Locations 167 7.5 Summary 209
6.3.3 Resources 169 7.6 Review Questions 210
6.3.4 Paths 171 References 210
6.4 Operational Elements 171
6.4.1 Routings 171 8 Simulation Output Analysis 211
6.4.2 Entity Operations 172
Introduction 211
8.1
6.4.3 Entity Arrivals 175
8.2 Statistical Analysis of Simulation
6.4.4 Entity and Resource
Output 212
Movement 177
8.2.1 Simulation
6.4.5 Accessing Locations and
Replications 213
Resources 178
8.2.2 Performance
6.4.6 Resource
Scheduling 180
Estimation 214
6.4.7 Downtimes and
Repairs 181
8.2.3 Number of Replications
6.4.8 Use of Programming
(Sample Size) 218
Logic 185
8.2.4 Real-World Experiments
6.5 Miscellaneous Modeling Issues 187
versus Simulation
6.5.1 Modeling Rare
Occurrences 187
Experiments 221
8.3 Statistical Issues with Simulation
6.5.2 Large-Scale Modeling 188
6.5.3 Cost Modeling 189
Output 222
8.4 Terminating and Nonterminating
6.6 Summary 190
Simulations 225
6.7 Review Questions 190
8.4.1 Terminating
References 192
Simulations 226
 Contents ix

8.4.2 Nonterminating More Than Two Alternative

Simulations 226 Systems 258
8.5 Experimenting with Terminating 9.4.3 Factorial
Design and
Simulations 227 Optimization 264
8.5.1 Selecting the Initial Model 9.5 Variance Reduction Techniques 265
State 228 9.5.1 Common Random
8.5.2 Selecting a Terminating Numbers 266
Event to Control Run 9.5.2 Example Use of Common

Length 228 Random Numbers 268

8.5.3 Determining the Number 9.5.3 Why Common Random
of Replications 228 Numbers Work 27 0
8.6 Experimenting with Nonterminating 9.6 Summary 271
Simulations 229 9.7 Review Questions 271
8.6.1 Determining the Warm-up References 272
Period 229
8.6.2 Obtaining Sample 10 Simulation Optimization 273
Observations 234
10.1 Introduction 273
8.6.3 Determining Run 10.2 In Search of the Optimum 275
Length 239 10.3 Combining Direct Search Techniques
8.7 Summary 240 with Simulation 276
8.8 Review Questions 241
10.4 Evolutionary Algorithms 277
References 242
10.4.1 Combining Evolutionary
Algorithms with
9 Comparing Systems 243
Simulation 278
9.1 Introduction 243 10.4.2 Illustration of an

9.2 Hypothesis Testing 244 Evolutionary Algorithm's

9.3 Comparing Two Alternative System Search of a Response

Designs 247 Surface 279

9.3.1 Welch Confidence Interval 10.5 Strategic and Tactical Issues of
for Comparing Two Simulation Optimization 281
Systems 248 10.5.1 Operational
9.3.2 Paired-1 Confidence Efficiency 281
Interval for Comparing Two 10.5.2 Statistical Efficiency 282

Systems 250 10.5.3 General Optimization

9.3.3 Welch versus the Paired-/ Procedure 282
Confidence Interval 252 10.6 Formulating Example
an

9.4 Comparing More Than Two Optimization Problem 284

Alternative System Designs 253 10.6.1 Problem Description 285
9.4.1 The Bonferroni Approach 10.6.2 Demonstration of the
for Comparing More General Optimization
Than Two Alternative Procedure 286

Systems 253 10.7 Real-World Simulation Optimization

9.4.2 Advanced Statistical Project 289
Models for Comparing 10.7.1 Problem Description 289
 Contents

10.7.2 Simulation Model and 11.7 Summary 320

Performance Measure 290 11.8 Review Questions 320
10.7.3 Toyota Solution References 320
Technique 291

10.7.4 Simulation Optimization 12 Modeling Material Handling

Technique 292 Systems 323
10.7.5 Comparison of Results 292
12.1 Introduction 323
10.8 Summary 294
12.2 Material Handling Principles 323
10.9 Review Questions 294
12.3 Material Handling
References 295
Classification 324
12.4 Conveyors 325
Modeling Manufacturing 12.4.1 325
Conveyor Types
Systems 299 12.4.2 Operational
11.1 Introduction 299 Characteristics 327
11.2 Characteristics of Manufacturing 12.4.3 Modeling Conveyor
Systems 300 Systems 328
11.3 Manufacturing Terminology 301 12.4.4 Modeling Single-Section
11.4 Use of Simulation in Conveyors 329
Manufacturing 303 12.4.5 Modeling Conveyor
11.5 Applications of Simulation in Networks 330

Manufacturing 304 12.5 Industrial Vehicles 330

11.5.1 Methods Analysis 305 12.5.1 Modeling Industrial
11.5.2 Plant Layout 306 Vehicles 331
11.5.3 Batch Sizing 308 12.6 Automated Storage/Retrieval
11.5.4 Production Control 309 Systems 331
11.5.5 Inventory Control 312 12.6.1 Configuring an

11.5.6 Supply Chain AS/RS 332

Management 313 12.6.2 Modeling AS/RSs 334

11.5.7 Production 12.7 Carousels 335

Scheduling 314 12.7.1 Carousel

11.5.8 Real-Time Control 315 Configurations 335
11.5.9 Emulation 315 12.7.2 Modeling Carousels 335
11.6 Manufacturing Modeling 12.8 Automatic Guided Vehicle

Techniques 316 Systems 336

11.6.1 Modeling Machine 12.8.1 Designing an AGVS 337
Setup 316 12.8.2 Controlling an AGVS 338
11.6.2 Modeling Machine Load 12.8.3 Modeling an AGVS 339
and Unload Time 316 12.9 Cranes and Hoists 340
11.6.3 Modeling Rework and 12.9.1 CraneManagement 340
Scrap 317 12.9.2 Modeling Bridge
11.6.4 Modeling Transfer Cranes 340
Machines 317 12.10 Robots 341
11.6.5 Continuous Process 12.10.1 Robot Control 341

Systems 319 12.10.2 Modeling Robots 342

 Contents xi

12.11 Summary 343 LI.4 Run-Time Menus and

12.12 Review Questions 343 Controls 368
References 344 LI.5 Simulation in Decision

Making 369
13 Modeling Service Systems 345 Ll.5.1 California Cellular 370

13.1 LI.5.2 ATM System 373

Introduction 345
13.2 LI.6 Exercises 376
Characteristics of Service
Systems 346
2 Building Your First Model 379
13.3 Performance Measures 347
13.4 Use of Simulation in Service L2.1 Building Your First Simulation
Systems 348 Model 379
13.5 Applications of Simulation in Service L2.2 Building the Bank of USA ATM
Industries 350 Model 387
13.5.1 Process Design 350 L2.3 Locations, Entities, Processing,
13.5.2 Method Selection 350 and Arrivals 392
13.5.3 System Layout 351 L2.4 Add Location 396
13.5.4 Staff Planning 351 L2.5 Effect of Variability on Model
13.5.5 Flow Control 352 Performance 397
13.6 Types of Service Systems 352 L2.6 Blocking 398
13.6.1 Service Factory 352 L2.7 Effect of Traffic Intensity on

13.6.2 Pure Service Shop 353 System Performance 401

13.6.3 Retail Service Store 353 L2.8 Exercises 402
13.6.4 Professional Service 354
13.6.5Telephonic Service 354 3 ProModel's Output Viewer 405
13.6.6Delivery Service 355 L3.1 The Output Viewer 405
13.6.7 Transportation Service 355
L3.2 Report Tables 407
13.7 Simulation Example: A Help Desk
L3.3 Column Charts 408
Operation 355
L3.4 Utilization Charts 409
13.7.1 Background 356
L3.5 State Charts 409
13.7.2 Model Description 356
L3.6 Time-Series Charts 412
13.7.3 Results 359
L3.7 Dynamic Plots 413
13.8 Summary 360
L3.8 Exercises 415
13.9 Review Questions 360
References 360
4 Basic Modeling Concepts 419

L4.1 Multiple Locations, Multiple

PART II
Entity Types 419
L4.2 Multiple Parallel Identical
Labs
Locations 421
L4.3 Resources 425
1 Introduction to ProModel 365 L4.4 Routing Rules 426

Ll.l L4.5 Variables 429

ProModel Opening Screen 366
LI.2 L4.6 Uncertainty in Routing—Track
ProModel Menu Bar 366
LI.3 ProModel Tool Bars Defects and Rework 432
367
xii Contents

L4.7 Batching Multiple Entities of L6.2 Cycle Time 479

Similar Type 434 L6.3 Sorting, Inspecting a Sample, and
L4.7.1 Temporary Rework 480

Batching 434 L6.4 Preventive Maintenance and

L4.7.2 Permanent Machine Breakdowns 483

Batching 436 L6.4.1 Downtime Using
L4.8 Attaching One or More Entities MTBF and MTTR

to Another Entity 438 Data 484

L4.8.1 Permanent L6.4.2 Downtime Using

Attachment 438 MTTF and MTTR

L4.8.2 Temporary Data 484

Attachment 439 L6.5 Shift Working Schedule 488
L4.9 Accumulation of Entities 443 L6.6 JobShop 491
L4.10 Splitting of One Entity into L6.7 Modeling Priorities 492
Multiple Entities 444 L6.7.1 Selecting among
L4.ll Decision Statements 445 Upstream
L4.11.1 If-Then-Else Processes 493
Statement 445 L6.8 Modeling a Pull System 494
L4.11.2 While... Do L6.8.1 Pull Based on

Loop 447 Downstream

L4.11.3 Do...While Demand 495

Loop 449 L6.8.2 Kanban System 497

L4.11.4 Do... Until L6.9 Tracking Cost 499
Statement 449 L6.10 Importing a Background 502
L4.11.5 Goto Statement 451 L6.11 Defining and Displaying
L4.11.6 Wait Until Views 503
Statement 452 L6.12 Creating a Model Package 506
L4.12 Periodic System Shutdown 453 L6.13 Exercises 508
L4.13 Exercises 456
7 Model Verification and
5 Fitting Statistical Distributions Validation 517
to Input Data 465
L7.1 Verification of an Inspection and
L5.1 An Introduction to Stat::Fit 465 Rework Model 517
L5.2 An Example Problem 467 L7.2 Verification by Tracing the
L5.3 Auto::Fit Input Data 470 Simulation Model 519
L5.4 Exercises 472 L7.3 Debugging the Simulation
Model 520
6 Intermediate Model Building 475 L7.3.1 Debugging ProModel
L6.1 Attributes 475
Logic 522
L7.3.2 Basic Debugger
L6.1.1 Using Attributes to
Track Customer
Options 523
L7.3.3 Advanced Debugger
Types 476
L6.1.2 Attributes and
Options 524
Using
L7.4 Exercises 525
Local Variables 478
 Contents xiii

8 Simulation Output Analysis 527 L10.2.2 Multiterm Objective

Functions 583
L8.1 Terminating versus
L10.2.3 Target Range Objective
Nonterminating Simulations 527
Functions 586
L8.2 Terminating Simulation 528
L10.3 Conclusions 588
L8.2.1 Starting and L10.4 Exercises 590
Terminating Conditions
(Run Length) 529 11
L8.2.2 Replications
Modeling Manufacturing
530
Systems 595
L8.2.3 Required Number of
Replications 534 Ll 1.1 Macros and Run-time
L8.2.4 Simulation Output Interface 595
Assumptions 534 L11.2 Generating Scenarios 599
L8.3 Nonterminating Simulation 536 L11.3 External Files 600
L8.3.1 Warm-up Time and Run L11.4 Arrays 604
Length 538 L11.5 Subroutines 607
L8.3.2 Replications or Batch Ll 1.6 Random Number Streams 611
Intervals 543 L11.7 Merging a Submodel 611
L8.3.3 Required B atch Interval L11.8 Exercises 614
Length 545
L8.4 Exercises 547 12 Material Handling Concepts 617

L12.1 Conveyors 617

9 Comparing Alternative Systems 549
L12.1.1 Single Conveyor 618
L9.1 Overview of Statistical L12.1.2 Multiple
Methods 549 Conveyors 618
L9.2 Three Alternative Systems 550 L12.1.3 Recirculating
L9.3 Common Random Numbers 553 Conveyor 620
L9.4 Bonferroni Approach with L12.2 Resources, Path Networks, and
Paired-? Confidence Intervals 554 Interfaces 621
L9.5 Analysis of Designed L12.2.1 Resource as Material

Experiments 557 Handler 622

L9.5.1 Full Factorial L12.2.2 Resource as Machine

Design Simulation Operator

Experiment 558 and Material
L9.5.2 Fraction Factorial Handler 625

Experiment 561 L12.3 Crane Systems 626

L9.6 Exercises 563 L12.4 Exercises 628
Reference 637
10 Simulation Optimization with
SimRunner 569 13 Modeling Service Systems 639

L10.1 Introduction to SimRunner 569 L13.1 Balking of Customers 639

L10.2 SimRunner Projects 572 L13.2 Table Functions 641
L10.2.1 Single Term Objective L13.3 Arrival Cycles 643
Functions 574 L13.4 User Distribution 646
xiv Contents

L13.5 Modeling a University Appendix B Critical Values for Student's t

Cafeteria 647 Distribution and Standard Normal
L13.6 Modeling a Call Distribution 674
Center—Outsource2US 651
L13.7
Appendix C F Distribution for a = 0.05 675
Modeling a Triage—Los Angeles
County Hospital 654 Appendix D Critical Values for Chi-Square
L13.8 Modeling an Office (DMV) 657 Distribution 676
L13.9 Exercises 662

Appendix A Continuous and Discrete Index 677

Distributions in ProModel 669