<%BANNER%>

Simulation Modeling and Benefit-Cost Sensitivity Analysis for Technology Adoption on Puerto Rico-United States Tomato Su...


PAGE 1

SIMULATION MODELING AND BENEFIT-COST SENSITIVITY ANALYSIS FOR TECHNOLOGY ADOPTION ON PUERTO RICO-UNITED STATES TOMATO SUPPLY CHAIN By JIAOJU GE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Jiaoju Ge

PAGE 3

This thesis is dedicated to my l oving husband Qiyong and our expected son.

PAGE 4

ACKNOWLEDGMENTS First, I express my deepest appreciation to all my committee, Dr. Allen Wysocki, Dr. Bruce Welt and Dr. Lisa House. Foremost, I would like to express my sincere gratitude to my committee chair, Dr. Allen Wysocki, for his understanding, guidance and support throughout my masters study in University of Florida. I also would like to thank my other committee members for their participation. I thank Dr. Lisa House for her advisement, valuable ideas and kindness. I thank Dr. Bruce Welt for his advice and the knowledge he shared. I wish to thank Dr. Cecilia do Nascimento Nunes and Sharon Dea for their kindness and cooperation on data collection. I would also like to thank my fellow graduate students in the Food and Resource Economics Department for their support, suggestions and all the unforgettable memories. I thank my families and my friends for their support. Finally, the greatest thank you goes to my husband, Qiyong Xu, for his understanding, encouragement and love. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................iv LIST OF TABLES ............................................................................................................vii LIST OF FIGURES .........................................................................................................viii CHAPTER 1 INTRODUCTION........................................................................................................1 1.1 Overview of Tomato Supply Chain........................................................................2 1.2 Problem Statement..................................................................................................5 1.3 Objectives...............................................................................................................6 1.4 Testable Hypotheses...............................................................................................6 1.5 Research Scope.......................................................................................................7 2 LITERATURE REVIEW.............................................................................................8 2.1 Physical Conditions and Tomato Quality...............................................................8 2.2 Simulation System Modeling...............................................................................10 2.3 Economic Analysis of Technology Adoption and Price Asymmetry...................11 3 METHODS.................................................................................................................14 3.1 To Find Tomato Keeping Quality Equation.........................................................14 3.2 To Build Tomato Supply Chain Simulation Model..............................................15 3.3 To Conduct A Benefit-Cost Sensitivity Analysis.................................................16 4 DATA COLLECTION...............................................................................................17 4.1 Production Related Data Analysis........................................................................18 4.2 Processing Time Related Data Analysis...............................................................19 4.3 Labor Related Data Analysis................................................................................20 4.4 Temperature and Shelf Life Related Data Analysis.............................................21 4.5 Financial Related Data Analysis...........................................................................23 5 SIMULATION MODELING SPECIFICATION AND RESULTS...........................27 5.1 Keeping Quality Equation and Results.................................................................27 v

PAGE 6

5.2 Tomato Supply Chain Simulation Model and Results.........................................29 5.2.1 Simulation Model before Technology and Results....................................29 5.2.2 Simulation Model after Technology and Results Comparison...................32 6 BENEFIT-Cost SENSITIVITY ANALYSIS AND RESULTS.................................37 6.1 Analysis of Tomato Growers Benefits with Increased Exports..........................37 6.2 Analysis of Tomato Grower/Shippers Costs with Increased Exports.................42 6.3 Benefit-Cost Analysis of Grower/Shippers and Wholesalers with Shrinkage Reduction...............................................................................................................44 7 CONCLUSIONS........................................................................................................47 7.1 Summary...............................................................................................................47 7.2 Conclusions...........................................................................................................49 7.3 Implications..........................................................................................................50 APPENDIX A PICTURES OF TOMATO SUPPLY CHAIN DISTRIBUTION SYSTEM..............52 B SIMULATION MODELING PARAMETERS..........................................................56 Before Technology Adoption.....................................................................................56 After Technology Adoption........................................................................................58 C SIMULATION MODELING PROGRAMMING......................................................60 D BENEFIT-COST SENSITIVITY ANALYSIS SiMeTAR MODELing....................64 Model Input ((for 6.1 &6.2)........................................................................................64 Model Output 1 (for 6.1 &6.2)...................................................................................65 Model Output (2) Sensitivity Analysis for Benefit (for 6.1 &6.2).............................66 Model (for 6.3)............................................................................................................68 LIST OF REFERENCES...................................................................................................70 BIOGRAPHICAL SKETCH.............................................................................................73 vi

PAGE 7

LIST OF TABLES Table page 4-1 Processing Time Data..................................................................................................20 4-2 Labor Input Data..........................................................................................................21 4-3 Tomato Shelf Life and Temperature...........................................................................22 4-4 Temperature Distribution of Marine Container ( o C)...................................................23 4-5 Tomato Operation and Production Cost (Average Per Acre)......................................24 4-6 U.S. Average Monthly Retail Price and f.o.b. Shipping Point....................................26 5-1 Statistics of Reaction Rate...........................................................................................27 5-2 Predicted Results of Reaction Rate.............................................................................28 5-3 Summary Statistics of Simulation Model Before Technology....................................31 5-4 Summary Statistics of the Simulation Model after Technology..................................33 6-1 Probabilities of Benefit-cost Ratio for All Scenarios..................................................42 6-2 Result of Technology Cost Sensitivity Analysis.........................................................44 6-3 Growers and Wholesalers Benefits Analysis Regarding to Shrinkage Reduction......45 vii

PAGE 8

LIST OF FIGURES Figure page 1-1 Top Factors in Selecting Primary Supermarket.............................................................1 1-2 Tomato Supply Chain Distribution Process..................................................................5 2-1 Tomato () Keeping Quality and Temperature Relationship......................................10 2-2 The Relationship Between Three Distinct Research Fields In a Dynamic Simulation Model.....................................................................................................11 3-1 Puerto Rico-United States Tomato Supply Chain Components..................................15 4-1 Export Tomato Size Distribution.................................................................................18 4-2 Export Tomatoes Grade Distribution..........................................................................19 5-1 Supply Chain Simulation Model before and After Technology Applied....................30 5-2 Probability of Tomato Keeping Quality before Technology Adoption.......................34 5-3 Probability of Tomato Keeping Quality After Technology Adoption........................35 5-4 Mean Comparison of Tomato Keeping Quality..........................................................36 6-1 Probabilities of Benefit Analysis (1)...........................................................................39 6-2 Probabilities of Benefit Analysis (2)...........................................................................41 A-1 Tomato Dumping Into the Packing Line....................................................................52 A-2 Tomato Grading Line.................................................................................................52 A-3 Tomato Packing Line (25LB/box)..............................................................................53 A-4 Tomato Palletizing Operation (80 boxes/ pallet)........................................................53 A-5 Tomatoes Stored in an Ethylene Room......................................................................54 A-6 Tomato Loading in Santa Isabel, Puerto Rico............................................................54 viii

PAGE 9

A-7 Truck Shipments Leaving Santa Isabel, Puerto Rico.................................................55 A-8 Trucks Arriving at the Quincy FL Distribution Center..............................................55 ix

PAGE 10

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science SIMULATION MODELING AND BENEFIT-COST SENSITIVITY ANALYSIS FOR TECHNOLOGY ADOPTION ON PUERTO RICO-UNITED STATES TOMATO SUPPLY CHAIN By Jiaoju Ge August, 2006 Chair: Allen F. Wysocki Cochair: Bruce A. Welt Major Department: Food and Resource Economics Supply chain coordination and shipping conditions have a substantial impact on food quality of produce supply chains, especially when supply chains cross international borders. After harvesting, shipping conditions are the primary determinants of food quality for highly perishable commodities such as tomatoes. In addition, based on food marketing research, food quality for highly perishable commodities is becoming one of the most important issues in todays food markets. The objective of this project is to develop a tool to help identify opportunities to improve food quality and their marketing implications on the Puerto Rico-United States tomato supply chain after new technology has been adopted. With the help of a discrete-event simulation software (SIMUL8 ), a simulated supply chain distribution model, including tomato harvesting, packing, shipping and distributing, was developed to analyze overall supply chain performance, and Simetar x

PAGE 11

was used to estimate associated costs and benefits for each participant in this supply chain (e.g., from tomato grower to wholesaler). Finally, benefit-cost ratios were simulated to assess the adoption of temperature controlling technology on tomato supply chain marketing implications. The results show that tomato keeping quality can be improved from 2.7 days to 5.4 days on average when temperature range within marine containers is controlled better from 11.2 0 C (52.5 o F) to 31.4 0 C (88.5 o F) down to from13.3 0 C (56.0 o F) to 15.6 0 C (60.0 o F). However, whether or not tomato grower/shippers should adopt the temperature controlling technology to increase exports and higher margin possibilities depends not only on the benefit-cost ratio, but also on the actual cost of technology. xi

PAGE 12

CHAPTER 1 INTRODUCTION Extending shelf life and quality are two keys for increasing tomato imports to the United States. Food quality is directly impacted by the quality of transportation being used, especially when food crosses international borders. Therefore, examining the worthiness of investing in technology is becoming more important. In the Caribbean Basin, fruit exports have declined from $553 to $272 million from 1991 to 2001 (FAO Stat, 2003). The decline in exports is due in part, to food quality issues. Food Industry Review 2003 (Figure 1-1) highlights that 87% of American consumers choose their primary supermarket based on high quality fruits and vegetables. Consumers demand produce that has desirable appearance, taste and texture. To meet consumers expectations, high quality fruits and vegetables must be provided. Clean Neat StoreHigh Quality Fruits & VegetablesLow PricesUse-before/sell-by date markedHigh Quality Meats88%87%83%82%81% To p Fa c t o r s In Se l e c t in g Pr i m a r y Su p e r m a r k e t Clean Neat StoreHigh Quality Fruits & VegetablesLow PricesUse-before/sell-by date markedHigh Quality Meats88%87%83%82%81% To p Fa c t o r s In Se l e c t in g Pr i m a r y Su p e r m a r k e t Clean Neat StoreHigh Quality Fruits & VegetablesLow PricesUse-before/sell-by date markedHigh Quality Meats88%87%83%82%81% To p Fa c t o r s In Se l e c t in g Pr i m a r y Su p e r m a r k e t Clean Neat StoreHigh Quality Fruits & VegetablesLow PricesUse-before/sell-by date markedHigh Quality Meats88%87%83%82%81% To p Fa c t o r s In Se l e c t in g Pr i m a r y Su p e r m a r k e t Figure 1-1 Top Factors in Selecting Primary Supermarket Tomatoes, along with potatoes and lettuce, are the most accepted and consumed fresh produce in America (ERS, USDA 2004). The tomato industry estimates that fresh-market tomato retail value may exceed $4 billion (ERS, USDA 2004). Tomatoes are prized for their good nutrition, wide acceptance and high consumption. However, as 1

PAGE 13

2 supply chains grow, improving quality is becoming more challenging for all tomato supply chain participants. For international distribution systems, physical and environmental conditions are the primary determinants of food quality for highly perishable commodities. Among these determinants, temperature is extremely important for maintaining high tomato quality, referred to in this research as tomato keeping quality. Tomato imports now account for about 39 percent of U.S. tomato consumption, an increase from approximately 20 percent in the early 1990s (ERS, USDA 2004). This trend shows potential for increases in tomato exports from Caribbean basin. Therefore, Puerto Rico could be one of the beneficiaries. Tomato grower/shippers in Puerto Rico might have the opportunity to increase their tomato shipments to the United States if they can, continuously and consistently, increase their tomato quality to meet American consumers needs. One way to improve tomato quality is to control temperature conditions after harvest. 1.1 Overview of Tomato Supply Chain Tomatoes are harvested at the mature green stage in Puerto Rico for shipment to the United States. Harvesting occurs from January until the end of March for tomatoes that are exported to the United States and the remaining harvests are consumed on the island. During the peak of the harvesting season there may be 500 to 600 workers in the tomato fields every day. By the end of the season there may be as few as 300 workers. In each harvesting team, there is a crew leader, pickers, a truck driver and a bucket dumper. The driver is in charge of punching the cards of the pickers every time they bring a bucket. Typically there are 5 to 7 supervisors who are responsible for managing the

PAGE 14

3 pickers and taking care of mulching, spaying, fertilization and pesticide applications. Most of the employees that work in the field have been with their respective tomato companies for several years. The pickers are mostly Puerto Rican women and are paid per hour or per bucket depending on what they prefer. They are paid $5 per hour and they work 8 hours/day excluding weekends ($40/day). They can make more money than $40/day if they harvest a number of buckets that is greater than the average number of buckets that results in $40/day. The price of each bucket may vary between 50 cents and $1 depending on seasonality. Normal daily production averages approximately 30,000-25 pound boxes of tomatoes. The price per box may vary from $13 to $15 during March and from $37 to $40 when Florida production is low in January and February. All boxes can be traced using a number and a series of information including field, variety, picker, and number of boxes picked in that field. When tomatoes arrive at the packinghouse from the field, they are sorted by color and defects, and graded by size into 4 different grades: Beefsteak Grade 1; Flavor ripe Grade 1; Homegrown Grade 2; Garden Grade 3. All tomatoes showing a little pink color are separated from the green and sold to local markets. Only green tomatoes are exported to the USA. Tomatoes are then washed and waxed, and then packed into 25 lb boxes. At the end of the packing line, tracking codes are stamped on the boxes. There are about 150 to 160 workers in the packinghouse, who work a 4 to 6 hr/day, depending on the tomato volume and grades to be packed.

PAGE 15

4 After packing, tomato boxes are immediately palletized and placed into ethylene treatment rooms. The medium-large size tomatoes stay three nights inside the ethylene room and the extra large size tomatoes stay four nights. Temperature of the ethylene rooms are approximately 68.0F (20.0C). Tomatoes to be exported are placed inside refrigerated marine containers (e.g., Horizon Lines) after the 3-day period in the ethylene treatment. Containers are shipped on the same day from Santa Isabel (south part of the island) to the port of San Juan (north part of the island). Depending on shipping schedules, containers are then loaded to ships and transported to the port in Jacksonville, Florida. During the peak of the tomato season, an average of 50 to 60 marine containers are shipped every week to the United States. By the end of the season, an average of 15 to 25 containers are shipped each week. Each marine container holds 18 pallets of tomatoes and each pallet holds eighty 25 pounds tomato boxes. Average temperature inside refrigerated container is about 13.3C (56.0F). Upon arrival to Jacksonville, the containers are unloaded and then taken by truck to a distribution center in Quincy, FL. Upon arrival, containers are immediately unloaded and placed in a large cold room with a temperature setting of 58.0F (14.4C). Each container has a specific place designated by a number in the cold room. Tomatoes are then stored for a short period of time (no longer than 1 day) before leaving the warehouse to be re-distributed to diverse wholesalers around the U.S. (Florida, Tennessee, California) and Canada (Toronto). Figure 1-2 illustrates the steps involved in the tomato supply chain used in this research from field to wholesalers.

PAGE 16

5 Field Hot water + chlorine treatment Sorting Washing Waxing Packing Ethylene chambers Marine containers Port of San Juan, PR Port of Jacksonville, FL Quincy, FL Containers unloaded Distributed in the USA and Canada Figure 1-2 Tomato Supply Chain Distribution Process 1.2 Problem Statement Any technology that could be used to increase tomato quality would be welcomed by Puerto Rico-United States tomato supply chain participants. Tomato supply chain participants are always looking for ways to improve their profitability. This study will identify the effect of temperature-controlling technology used by Puerto Rican tomato grower/shippers on tomato quality for all tomato supply chain

PAGE 17

6 participants and what the possibility is for tomato grower/shippers to adopt this technology from financial point of view. 1.3 Objectives The general objectives of this study are to develop a tool to help identify potential tomato quality improvements as a result of the adoption of new temperature-controlled technology, and to discuss their marketing implications for the Puerto Rico-United States tomato supply chain. The specific objectives are: To identify an established relationship between tomato keeping quality, temperature and time in the distribution system; To build a tomato supply chain simulation model to analyze tomato keeping quality changes as they relate to the changes in temperature and time in the tomato supply chain; To conduct a sensitivity analysis of benefit-cost ratio for tomato supply chain participants. 1.4 Testable Hypotheses Effects of temperature and time on fruit keeping quality are well documented in the literature. This information is applied in this study. While the impact of temperature on tomato keeping quality is well documented, the quantity impact of increasing keeping quality is not as clear. But its clear that tomatos shrinkage might be reduced by an increased keeping quality. An assumption is made in this study that higher quality usually results in more tomato exported to the United States market. Therefore, the hypotheses of the study are the following: H 1 : Simulated software can be used to analyze the complexities of Puerto Rico-United States tomato supply chain distribution system; H 2 : The adoption of temperature-controlling technology will prolong tomato keeping quality;

PAGE 18

7 H 3 : The adoption of temperature-controlling technology will result in increasing probabilities of the benefit-cost ratio being greater than 1. 1.5 Research Scope Discrete-event simulation software, such as Arena and Simul8 are commonly used to analyze complex systems with stochastic model. Simul8 was used in this study to build a tomato supply chain distribution model. Results from previous fruit keeping quality studies, which are highlighted in the next chapter, are used in this tomato supply chain model, to analyze the effect of temperature controlling technology on Puerto Rico-United States tomato supply chain participants. Even through benefit-cost analysis has not proven to be particularly useful in capturing the intricacies of complex scientific problems (Whittington and Grubb, 1984), it is widely accepted in the area of policy effect and project determination. Therefore, benefit-cost analysis is utilized to analyze the possibility of adopting the selected technologies for tomato supply chain participants. The benefit is derived from the overall revenues generated by supply chain participants. The costs of adopting temperature controlling technologies are obtained from financial cost data. Since the project is still on-going, the actual benefit and cost data will be collected in late 2006 or early 2007. Estimated benefits and costs are used in this study to conduct a sensitivity analysis.

PAGE 19

CHAPTER 2 LITERATURE REVIEW While there is a vast body of research related to technology adoption, much less research has been done studying the economic impact of technology adoption. This study will concentrate on previous studies in three different categories. The first category reviews literature related to how physical conditions affect tomato quality. The second category lists previous studies which used discrete event simulation models to solve problems in supply chain distribution systems. The last category is comprised of research focused on economic implication of new technology adoption and price asymmetry. 2.1 Physical Conditions and Tomato Quality The reaction rate of a decrease in quality (of all fresh fruit) as it relates to temperature, is identified in Arrheniuslaw: )11)*((expabsrefaTTRErefKK (2.1) where K is reaction rate of the decrease in quality; T is the temperature; E a and R are both Constant (Chang, 1981). This study showed that temperature has an effect on reaction rate (the speed at which keeping quality changes based on differences in temperature and time) for all different kinds of perishable fruits and vegetables. In addition, one study (Robertson, 1993) explained the quality of most foods and beverages decreases with storage or time. The relationship between shelf life and time can be expressed by the zero order equation. This research will use a common form of the pseudo first order equation, which is shown in (2.2) 8

PAGE 20

9 )*(0*TKEXPQQ (2.2) where Q is the remaining shelf life at the end of the time period, Q 0 is the shelf life at the beginning of the time period, K is the reaction rate and T is the time period. Fruit keeping quality and shelf life are often highly correlated concepts. Tijskenss study provided the definition of keeping quality (Tijskens et al., 1994). According to this study, keeping quality can be defined as the minimal quality necessary for a consumer to accept the product. Tijskens and Polderdijk (1996) extended the research on fruit keeping quality with the application of Changs theory to the study of keeping quality of vegetable produce during storage and distribution. Their results showed that a change of keeping quality (shelf life) is negatively impacted depending on temperature. This relationship between shelf life, keeping quality, and temperature is illustrated in the following reaction rate equation: )1(infQQKQdtdQ (2.3) where dtdQ is the change of the quality, K is the reaction rate, Q is the keeping quality and Q inf is the quality maximally possible at infinite time (Tijskens and Polderdijk, 1996). One useful result from the 1996 Tijskens and Polderdijks study, for this research, is the use of a discrete generic model to predict tomato keeping quality and temperature, shown in figure 2.1. It indicated that when temperature is around 15.0 0 C, tomato will have longest keeping quality, at around 15 days.

PAGE 21

10 Figure 2-1 Tomato () Keeping Quality and Temperature Relationship In summary of this category of literature, for tomato, keeping quality can be written as a function of both reaction rate K and time period t. Furthermore, it is understood that the optimal shelf life for tomato is a result of maintaining a temperature range of 12.0 0 C to 18.0 0 C. 2.2 Simulation System Modeling One particularly relevant study (Schepers et al., 2004) showed an example of building a dynamic simulation model for analyzing the complexity of the mango supply chain distribution system. Mangoes and tomatoes share many of the same physiological and distribution characteristics, and both are highly perishable fruits. Three distinct research fields (consumer science, quality management and chain science) and the links among them are analyzed by this study. The relationship between these research fields are illustrated in Figure 2-2. This approach is useful for studying and understanding a wide range mango supply chain management issues. However, this analysis is relatively less detailed for each of the three fields. Our current tomato supply chain study will be only concentrated on two of the three fields, but in more detail.

PAGE 22

11 Promotefirst usage 1. Consumer science:Adoption dynamicsFirst and repeat usageLoss of interestdynamics DemandProductLikingRealisequality 2. Quality management & Consumer sciencePost-harvest product handling & logistics (ripening)Biological variation in physico-chemical product propertiesSensory perception and evaluation (Liking) Cost of productloss & handlingSharecost 3. Chain science:Collaborative marketingCost sharingPricing Profits perplayer Figure 2-2 The Relationship Between Three Distinct Research Fields In a Dynamic Simulation Model In summary, the literature related to supply chain modeling analysis is less extensive. However, the Schepers et al. study provides a good example of using modeling software to analyze supply chain complexities. Furthermore, our study could be considered an extension of this earlier research, especially the economic analysis section 2.3 Economic Analysis of Technology Adoption and Price Asymmetry Deming (1967, 1972), Juran and Gryna (1970) and Juran (1974) emphasized that most quality problems are management or systems problems requiring statistics in their solution. Following this study, Marquardt (1984) pointed out that product quality is high visibility in todays economic environment. The article explores the respective roles of business philosophy, management systems and technology systems. It also pointed out

PAGE 23

12 that the quality technology systems require new directions and new emphases in statistics, software engineering, and other disciplines in order to be cost-effective. But how quality is affecting the food industry? Caswell et al. (1998) explained how quality management meta-systems affect the food industry. They indicated that food quality meta-systems are adopted by system participants for a variety of reasonsdepending on the internal benefits and costs of adoption This study utilized benefit-cost analysis to identify the adoption of a meta-system on one company. The internal benefits pointed out by this study are the revenues from sale of products, and costs are captured through production, transaction and regulatory compliance. The result of this study showed that 71% of small firms expected adoption to increase their market share, while only 48% of large firms had the same expectation. For the current study, price asymmetry in retail, wholesale and shipping point price (growers price) is also a key component that needs to be considered. Several studies have examined vertical price transmission in fresh produce. All of the studies indicated that short-run supplier price changes precede retail price changes (Ward, 1982; Powers, 1995; Heien, 1980). Furthermore, a statistical test finds that tomato retail prices respond more to free-on-board (f.o.b.) price increase than to f.o.b. price decrease (ERS, 1999). In the short-run, prices are affected more by supply. An increase (decrease) in supply causes f.o.b. prices to fall (rise) which, in turn, leads to decreases (increases) in retail prices (ERS, 1999). This study also indicated that on average, the shipping point price for fresh field grown tomato is about one-fourth of the retail value (ERS, 2004).

PAGE 24

13 In conclusion, technology is one of the most important factors for determining food quality. Technology can be adopted by a system or a private company to improve food quality and the effect can be evaluated by benefit-cost analysis. Moreover, for the whole tomato supply chain distribution system, if the shipping point price and its trend could be known, the wholesale and retail price and their trends should possibly be calculated through the vertical market markups.

PAGE 25

CHAPTER 3 METHODS The objectives of this study are to develop a tool to help identify potential tomato quality improvements that lead to less shrinkage and more exports, to discuss their marketing implications for the Puerto Rico-United States tomato supply chain with the consideration about new temperature controlling technology being adopted, and to conduct a sensitivity analysis of the benefits and costs of technology adoption. Several different, but connected, methods are used in this study. 3.1 To Find Tomato Keeping Quality Equation To find the impact on tomato keeping quality for temperature, the reaction rate (equation 3.1) of how keeping quality responds to the change of temperature is calculated by the following function, which is a transformation of equation (2.1): )1()(10absTKLn (3.1) where K represents the reaction rate and T abs is the absolute temperature. The keeping quality equation (3.2) can be derived based on the above calculated reaction rate K and equation (2.2): )*(1*TKtttTEXPQQ (3.2) where Q represents tomato keeping quality, t and t-1 represent time periods, K represents the reaction rate and T represents the time between time period t and t-1. 14

PAGE 26

15 This equation shows the relationship between temperature and tomato keeping quality, which is the most important input for the tomato supply chain simulation model in this study. 3.2 To Build Tomato Supply Chain Simulation Model A specific simulation model is built for tomato supply chain between Puerto Rico and the U.S. by using SIMUL8 software, a widely used simulation software. The dynamic nature of Simul8 allows the user to show the creation of processes and outputs as moving objects, while other software programs are static in nature. This tomato supply chain model features the tomato keeping quality equation as well as all other stages showed in Figure 3.1. Wholesaler:Distribution center for storageTruck for shipping Puerto Rico Tomato Grower:Packing House for Packing Ethylene Room for StorageShipping by Truck and Sea Retailers:Stores for retailing U.S.A. U.S.A. & CanadaShipping-pointPriceWholesale priceRetail price Wholesaler:Distribution center for storageTruck for shipping Puerto Rico Tomato Grower:Packing House for Packing Ethylene Room for StorageShipping by Truck and Sea Retailers:Stores for retailing U.S.A. U.S.A. & CanadaShipping-pointPriceWholesale priceRetail price Figure 3-1 Puerto Rico-United States Tomato Supply Chain Components In conclusion, the tomato supply chain simulation model includes all the above components. The key output of the model for this study is tomato keeping quality as it is affected by temperature changes in the tomato supply chain.

PAGE 27

16 3.3 To Conduct A Benefit-Cost Sensitivity Analysis Simetar (Simulation & Econometrics to Analyze Risk) software will be used to conduct benefit-cost sensitivity analysis for the adoption of temperature controlling technology on tomato supply chain participants. This section will consider the sensitivity for both increased export quantity and reduced shrinkage generated by the increased tomato keeping quality. The steps in this sensitivity analysis are: Modeling key output variables by using financial data with different distributions. Total revenue = f(export price, quantity exported) Total cost =Production Cost + Transportation Cost Benefit-cost ratio = B / C ratio = total revenue / total cost Benefit-cost ratio analysis for tomato grower with four possible increased export scenarios. Sensitivity cost analysis for highest amount money that tomato growers would pay for a given technology regarding to four increased export scenarios. Benefit analysis for other supply chain participants (wholesalers) for three possible industry mark-ups between each level of participation regarding to reduced shrinkage. The results of benefit-cost analysis are used to provide recommendations about whether or not tomato grower/shippers in Puerto Rico should adopt this temperature controlling technology.

PAGE 28

CHAPTER 4 DATA COLLECTION The data needed for this study is diverse as a result of the methods mentioned in the previous chapter. Most of the required data on operational and financial data are regarded as non-public information by most companies. Therefore, the data collected in this study came from four major sources. One source was collected from one typical tomato operating firm, Gargiulo, at both field and packing house in Santa Isabel, Puerto Rico, as well as distribution center in Quincy, Florida in March of 2005 and 2006. The second source was collected from the tomato export port at San Juan, Puerto Rico and the tomato import port at Jacksonville, Florida in March of 2005 and 2006. The third source of data was collected through telephone interviews in Florida during October of 2005 and March of 2006. The remaining data came from secondary on-line data sources. Five types of data were collected from both Puerto Rico and the United States: 1) production related data, such as daily production; 2) processing time related data, such as packing time and truck moving time; 3) labor related data, such as how many workers and their salaries; 4) temperature and shelf life related dat; 5) financial related data, such as costs and prices. Data from first three categories are used to build the Puerto Rico-United States tomato supply chain model. Category 4 data is used to build the tomato keeping quality equation which is an integral part of the simulated supply chain model. Finally, category 5 data and the results from the model will be used to conduct benefit-cost sensitivity analysis. 17

PAGE 29

18 4.1 Production Related Data Analysis As an integrated tomato company, Gargiulo has tomato fields in Puerto Rico and also a distribution center in Quincy, Florida. Around 750,000 pounds of tomato can be picked within one day and dumped into the packing house. Among all of these tomatoes, on average 25% never make the grade and are disposed of. The remaining 75% goes to both the Puerto Rican local market and the U.S. export market. For all tomatoes exported to the U.S., there are three different sizes and four different grades. Size one represents the extra large tomatoes, which constitute around 25% of all export tomatoes (Figure 4-1). Size two represents the large tomatoes, at about 60%. Finally, size three represents the medium tomatoes that are 15% of the total crop on average. 25 60 15 0102030405060 Size 1 (5*6)Size 2 (6*6)Size 3 (6*7) Figure 4-1 Export Tomato Size Distribution Tomato exports to the U.S. include grades 1, 2, 3 and 4 (figure 4-2). Grade 3 tomatoes make of the majority of tomatoes (about 60%), while Grade 1 and 2 each have approximately 12.5%. The 4 th Grade share approximately 15% of the tomatoes exported.

PAGE 30

19 12.512.56015 Grade 1 Grade 2 Grade 3 Grade 4 Figure 4-2 Export Tomatoes Grade Distribution 4.2 Processing Time Related Data Analysis The processing time is the time period from the start to the end of one task. In this research, the most important time period is the processing time for packing one box of tomatoes, palletizing one pallet, shipping, loading and unloading trucks and boats. Another important concept for the processing time related data is the distribution of the time. Not all the processing times are fixed in actual practice. Instead, most processing times vary over a predictable probability distribution. Considering this range in model building will increase the models prediction ability. For example, in table (Table 4-1), the shipping time from Santa Isabel to San Juan is 120 minutes (2 hours). This means that a truck driven from the packing house at Santa Isabel to the export port at San Juan takes around 2 hours. The corresponding distribution for this shipping time is considered to be a bounded normal distribution, which means that it takes half an hour to 4 hours to ship depending on other conditions such as traffic and weather. Other processing times can be explained in the same way.

PAGE 31

20 Table 4-1 Processing Time Data Task Processing Time (Minutes) Distribution (Minutes) Harvesting 1.4 Exponential Packing (Per Box) 0.016 Exponential Palletizing (Per Pallet) 2-5 Uniform (2,5) Ethylene Room Storage (Size 1) 5040 Fixed Ethylene Room Storage (Size 2&3) 3600 Fixed Loading & Unloading Truck (Per marine Container) 5 Average (5, 1.25) Shipping (Santa Isabel to San Juan) 120 Bounded Normal (120,10,90,240) Shipping (San Juan to Jacksonville) 2160 Bounded Normal (2160,240,2880,1440) Shipping (Jacksonville to Quincy) 180 Bounded Average (180, 240,120) Loading & Unloading Ship (Per marine Container) 15 Average (15, 3.75) Cold Room Storage (Quincy) 1440 Average (10, 2.5) 4.3 Labor Related Data Analysis For many industries, labor is a fundamental and required resource. Regarding the tomato supply chain between Puerto Rico and the United States, especially among tomato grower/shippers, the labor data is an integral component of the tomato supply chain simulation model. Labor data includes two types of data. One is the number of laborers for each task, such as picking and packing. The other is the working hours which includes what are the shifts and how long each shift is. For this study, field pickers of tomatoes averaged around 500 to 600 people daily (Table 4.2). They worked from 6am until 2pm everyday without a shift change. For

PAGE 32

21 packing house workers, including sorters, graders, packers and people who palletize, the total number ranges from 150 to 160 people. These laborers work from 10am to 2pm or 4pm depending on the amount of tomatoes in the packing house. The reason that packing house workers start later than field pickers is they have to wait until the tomatoes are dumped into the packing line after being picked in the field. Table 4-2 Labor Input Data Labor Type Numbers Working Hours Shifts Field Pickers 500-600 8 (6am-2pm) Day shift (6am-2pm) Packing House Workers 150-160 4-6 (10am-2pm or 4pm) Day shift 4.4 Temperature and Shelf Life Related Data Analysis The basic and central idea of the tomato supply chain simulation model is to find the ways to increase tomato keeping quality, which should generate more total revenue for the whole industry. Therefore, obtaining the temperature and shelf life data and better understanding the temperature changes among different stages are extremely important. Shelf life here is synonymous with keeping quality. The data in Table 4-3 are collected from the tomato supply chain between Puerto Rico and the United States, with the exception of shelf life, which is obtained by a secondary source. The T in the table 4-3 represents temperature. For instance, in ethylene room 1, when tomatoes stay inside for 3600 minutes and under the temperature of 68.0 o F, the shelf life is approximately 8.0 days. Most importantly, from the above data, it shows that the longest shelf life is about 14.8 days and the shortest shelf life is about 5.5 days due to different temperature conditions, treating humidity and other physical condition as constants. Therefore, tomato keeping quality can be increased to as long as 14.8 days.

PAGE 33

22 Therefore, temperature distribution outside of the ethylene room, and particularly, inside the marine container is considered as a critical condition for maintaining tomato keeping quality. These distribution data is showed in Table 4-4. It indicates that within a marine container, the temperature varies considerably from 11.2 o C (52.5 o F) to 31.4 o C (88.5 o F), following the uniform distribution with the min of 11.2 o C (52.5 o F) and the max of 31.4 o C (88.5 o F). (Source: experiments of Agricultural Biological Engineering Department at the University of Florida). From the secondary data about the shelf life, when the temperature is 11.2 o C (52.5 o F), the shelf life would be 7.5 days. However, when temperature increases to 31.4 o C (88.5 o F), the shelf life of tomato would be as low as only 2 days which is critical for tomato keeping quality (Figure 2-1). Table 4-3 Tomato Shelf Life and Temperature Items Storage Time (Minutes) T ( o F) T ( o C) T ( 0 K) Shelf Life (Days) Packing 120 76.5 24.7 297.9 5.5 Ethylene Room1 5040 68.0 20.0 293.2 8.0 Ethylene Room2 3600 68.0 20.0 293.2 8.0 Marine Container (Truck) 120 56.0 13.3 286.5 14.0 Marine Container (Ship) 2160 56.0 13.3 286.5 14.0 Marine Container (Truck) 180 56.0 13.3 286.5 14.0 Cold Room (Distribution Center) 720 58.0 14.4 287.6 14.8 However, since only green tomatoes are exported to the United States, they need to stay inside the ethylene room for enough time to begin the ripening process. If one

PAGE 34

23 assumes that ethylene room best practices are followed, the shelf life within ethylene room in the above table will not be considered as a critical condition for overall tomato quality. Table 4-4 Temperature Distribution of Marine Container ( o C) Marine Container Min Temperature Max Temperature Distribution Bottom 11.2 31.4 Uniform Center 13.7 25.7 Uniform Up 14.9 24.6 Uniform Over all (Min & Max) 11.2 31.4 Uniform 4.5 Financial Related Data Analysis Financial data are necessary to derive benefit-cost ratio sensitivity analysis. Two types of data were collected including costs related data (Table 4-5) and revenue related data (Table 4-6). Cost data are collected both from an on-line source and from a Puerto Rican tomato grower. It represents production cost of one typical tomato grower in Florida. This research compared production cost data for a particular Puerto Rican tomato grower through interviews. Based on these interviews, total operating cost for this tomato grower in Puerto Rico was $4,491.14, which was lower than the average Florida growers operating cost of $4,984.15 (Table 4-5). The fixed cost was $1,623.06, lower than the $1,658.01 average in Florida. However, the harvesting and marketing cost were similar. Therefore, the total production cost for Puerto Ricos tomato grower used in the model was $11,858.20(Table4-5) per acre, which is lower than the total production cost on average for Florida ($12,386.16). It is easy to see that tomato growers in Puerto Rico have an advantage compared to growers in Florida based on production cost. However, this advantage is not that large

PAGE 35

24 Table 4-5 Tomato Operation and Production Cost (Average Per Acre) Category Florida Gargiulo OPERATING COSTS Dollars Transplants 480.00 456.00 Fertilizer 358.50 376.43 Fumigant 550.00 577.50 Fungicide 241.00 253.05 Herbicide 20.77 21.81 Insecticide 459.93 482.93 General Farm Labor 140.63 140.63 Machinery Variable Costs 858.59 858.59 Tractor Driver Labor 193.55 203.23 MISCELLANEOUS String and Stake Disposal 123.42 129.59 Scouting 45.00 42.75 Pruning 79.86 83.85 Plastic Mulch Disposal 163.35 171.52 Level Land 145.00 0.00 Farm Vehicles 27.33 0.00 Drive Stakes 81.31 85.38 Cut Cross Ditches 27.20 0.00 Tie Plants 145.20 152.46 Drip Tube 145.20 0.00 Plastic Mulch 315.00 330.75 Tomato Stakes 90.00 94.50 Plastic String 28.75 30.19 Interest on Operating Capital 264.56 0.00 Total Operating Cost 4,984.15 4,491.14 FIXED COSTS Land Rent 500.00 475.00 Machinery Fixed Costs 208.15 218.56 Farm Management 678.47 644.55 Overhead 271.39 284.96 Total Fixed Cost 1,658.01 1,623.06 HARVEST AND MARKETING COSTS Tomato Cartons 1,200.00 1,200.00 Sell 240.00 240.00 Pick and Haul 1,360.00 1,360.00 Pack 2,800.00 2,800.00 Various Association Fees 144.00 144.00 Total Harvest and Marketing Cost 5,744.00 5,744.00 TOTAL COST 12,386.16 11,858.20 Total Fixed Cost 1,658.01 1,623.06 TOTAL Variable Cost 10,728.15 10,235.14

PAGE 36

25 since the production cost difference is only 5%. Tomato growers in Puerto Rico have to pay additional shipping costs averaging about $0.1 per pound. Revenue data were collected from two sources. One is the data collected from the Gargiulo firm. During the export season in Puerto Rico, the export price (the f.o.b. shipping point price) varied from $37 to $40 per 25 pounds during January and February (148 to 160 cents per pound) and from $13 to $15 per 25 pounds during March (52 to 60 cents per pound). However, from the data collected from USDA on Table 4-6, the f.o.b. shipping point prices during January and February ranged from 18.4 to 116.0 cents per pound with an average of 39.2 cents per pound for the year 1990 to 2004. For March, prices ranged from 21.2 to 81.7 cents per pound with the average of 44.3 cents per pound during the same period. This indicates that tomato growers in Puerto Rico have a great advantage for revenue on the shipping point price since the average price is much lower than Puerto Rican growers prices during the Puerto Rican shipping season. On the other hand, the Table 4-6 also indicates that along with the change of f.o.b. shipping point price, the retail price has changed in the same direction for the past 15 years during the January to March season without considering the abnormal price shocks brought about by weather, etc. This also suggests that the retailers should have the same price trend as the tomato growers. If tomato grower/shippers export price doesnt change, the retail price is likely to remain unchanged as well. In summary, all category data were needed for the next chapters tomato supply chain simulation model and benefit-cost ratio sensitivity analysis. The first 4 category

PAGE 37

26 data would be used for simulation model. The result of simulation model and the last category data would be used for benefit-cost sensitivity analysis. Table 4-6 U.S. Average Monthly Retail Price and f.o.b. Shipping Point Jan. Feb. Mar. Jan. Feb. Mar. Year Retail price (cents per pound): F.o.b. shipping-point price (cents per pound: 1990 173.5 236.1 176.5 116.0 97.6 32.3 1991 91.2 84.0 94.8 23.1 31.6 44.0 1992 93.6 143.0 172.9 40.5 76.0 80.7 1993 114.1 109.8 88.0 38.3 21.9 21.2 1994 160.4 111.2 91.4 41.5 19.3 24.5 1995 132.3 130.0 108.1 41.1 29.8 37.1 1996 110.3 108.4 146.7 18.4 40.0 81.7 1997 121.3 131.4 165.4 32.1 45.9 57.4 1998 145.2 135.6 151.5 26.4 44.0 34.0 1999 190.4 147.6 139.5 33.5 23.4 22.3 2000 144.3 128.6 136.4 21.4 21.1 33.0 2001 141.4 131.3 133.6 43.8 29.1 56.4 2002 145.1 129.8 129.2 38.2 28.0 41.7 2003 171.1 156.5 161.9 50.9 31.7 55.6 2004 147.2 151.0 152.9 34.5 36.3 42.2

PAGE 38

CHAPTER 5 SIMULATION MODELING SPECIFICATION AND RESULTS Based on the previously described methods and the data collected, a specific tomato supply chain simulation model between Puerto Rico and the United States was built and the details building this model and the specifications used are explained in the following paragraphs. The tomato keeping quality equation is discussed first, since it is an important input for the tomato supply chain simulation model. 5.1 Keeping Quality Equation and Results Based on the temperature and shelf life related data discussed in chapter 4, the reaction rate, which explains the keeping quality changes according to the change of the temperature, is calculated by the following equation: )1()(10absTKLn (5.1) where K represents the reaction rate and T abs is the absolute temperature. A linear regression is used to get the result and presented in table 5-1. Table 5-1 Statistics of Reaction Rate Intercept 24.22 F Ratio 250.26 Slope -7277.48 Prob. (F) 0.000018 R-Square 0.98 T-Test -15.82 S.E. 460.03 Prob. (T) 0.0004 From the probability of the F ratio and T-test, the statistics for intercept and slope are both statistically significant. Therefore, the reaction rate can be calculated as 27

PAGE 39

28 )1(48.727722.24)(absTKLn by using the above statistics as 22.240 and 48.72771 The predicted K is showed in table 5-2. Within a marine container, the predicted reaction rate averages 0.31 when the temperature is at the average of 56.0 0 F (13.3 0 C). However, if the range of temperature in the marine container is 52.5 0 F (11.2 0 C) to 88.5 0 F (31.4 0 C), the predicted reaction rate will be ranged from 0.25 to 1.38 (Table 5-2), where the range (1.13) is 3.7 times of the mean. It indicates that within a marine container, the wide change of temperature will have a huge impact on tomato keeping quality. Table 5-2 Predicted Results of Reaction Rate Items Temperature ( o F) Absolute Temperature ( 0 K) Predicted Reaction Rate (K) Packing 76.5 297.9 0.811612748 Ethylene Room1 68.0 293.2 0.547656726 Ethylene Room2 68.0 293.2 0.547656726 Marine Container (Min) 52.5 284.4 0.254924292 Marine Container (Mean) 56.0 286.5 0.307248107 Marine Container (Max) 88.5 304.5 1.38375325 Cold Room (Distribution Center) 58.0 287.6 0.338899278 After reaction rate is predicted, all the results for K will be used substitute into the following keeping quality equation: )*(1*TKtttTEXPQQ (5.2) where Q represents tomato keeping quality in days, t and t-1 represents time period in days, K represents the reaction rate and T represents the time between time period t and

PAGE 40

29 t-1 This keeping quality is used to plug into the following simulation model as a key input to get the result of the final keeping quality. 5.2 Tomato Supply Chain Simulation Model and Results The tomato supply chain simulation model between Puerto Rico and the United States was built by using simulation software, SIMUL8 SIMUL8 Corporation develops markets and supports business simulation software for enterprise-wide use in business, government, education, and any organization that handles flows of orders, people, transactions or products (SIMUL8 homepage, 2006). One of the most popular uses of SIMUL8 is for supply chain management (SIMUL8 homepage, 2006). There are some successful cases studied by the company which proved the potential use of SIMUL8 in this study. 5.2.1 Simulation Model before Technology and Results The model is very visual (Figure 5-1) and easy to see all the components of tomato supply chain between Puerto Rico and the United States. It includes tomato grower/shippers in Puerto Rico (light blue areas in Figure 5-1), tomato distributors (dark blue area in Figure 5-1) and tomato wholesalers and retailers in the United States and Canada (brown area in Figure 5-1). Graphs in the model represent all different facilities within tomato supply chain. For instance, harvesting represents being harvested and dumped into packing house. Packing represents tomatoes being packed into a 25lb-box and so on. Parameters (Appendix B) and computer programs (Appendix C) of each graph or facility are plugged into the model to allow the model to simulate the real world tomato supply chain distribution system. The data collection period for the model (Figure 5-1) is 3 months, which is exactly the period from January to March that tomatoes in Puerto Rico would be exported into

PAGE 41

30 Figure 5-1 Supply Chain Simulation Model before and After Technology Applied the United States. When model starts to run, both green and red tomatoes would be picked in the field and then dumped into packing house through harvesting. Then

PAGE 42

31 tomatoes would be sorted by four different grades and colors. Following, the red balls represent red tomatoes sold in Puerto Rico local market and the green balls represent green tomatoes being packed and exported to the United States by following all the processes and data that were collected. This tomato supply chain simulation model was built by varying the temperature conditions and holding other conditions such as humidity constant. Therefore the before-technology applied model (Figure 5-1) represents the supply chain as it is without any temperature controlling technology being adopted by any supply chain player, especially tomato growers. Table 5-3 Summary Statistics of Simulation Model Before Technology Keeping Quality (Days) Grade Mean SD CV Min Max Range 1 2.6 12.5 70.4 0.2 7.1 6.9 2 2.4 13.2 81.9 0.2 7.0 6.8 3 2.9 12.8 66.2 0.2 8.6 8.4 4 3.0 12.5 62.2 0.3 7.0 6.7 MEAN 2.7 12.7 70.2 0.2 7.4 7.2 The results of this before-technology model indicate that tomato keeping quality is relatively low and varies widely. The result summary statistics are given for all four different grades in Table 5-3. The mean of tomato keeping quality is 2.7 days, which is low compared to the normal condition where tomatoes are kept for around 14 days. In addition, the range of tomato keeping quality, which is from 6.8 to 8.4, is almost three times of the mean (2.7). Moreover, the coefficient of variation for tomato keeping quality

PAGE 43

32 is pretty high, on average at 70.2. These facts suggest that the variation of tomato keeping quality need to be reduced. In summary, the lower unacceptable mean with wide variation for tomato keeping quality indicates problems exist in this tomato supply chain between Puerto Rico and the United States. This is likely the result of wide temperature changes within the marine containers which is the key factor for changing tomato keeping quality. Therefore, the temperature controlling technology within marine container can be adopted by tomato shippers to increase tomato keeping quality and generate more exports and higher revenues. 5.2.2 Simulation Model after Technology and Results Comparison Before temperature within marine container was controlled more closely, the temperature range was from 11.2 0 C (52.5 o F) to 31.4 0 C (88.5 o F). This temperature range results in a wide range of keeping quality and tomato shelf life. Therefore, if better temperature control can be obtained by use of technology in the marine container, tomato keeping quality should increase. The literature review and data collected from this study indicate that a temperature of 14.4 0 C (58.0 o F) is the optimal temperature for ideal tomato keeping quality. It is reasonable to assume that temperatures within marine container can be controlled between 13.3 0 C (56.0 o F) and 15.6 0 C (60.0 o F). The results of the model after-technology will be compared to the results of the model before-technology to see the impact of the adoption of temperature controlling technology on tomato keeping quality. The new model, tomato supply chain simulation model after-technology applied is visually similar as the model before-technology model (Figure 5-1) since all the components of this supply chain from Puerto Rico to the United States are the same.

PAGE 44

33 However, the parameters for the temperature distribution within marine containers are different (Appendix B) to reflect the change of temperature. The result of tomato keeping quality after temperature controlling technology is adopted within marine container by tomato grower/shippers are presented in table 5-4. By comparing the coefficient of variance (CV) and the range to the result before-technology adopted model, it is clear that the coefficient of variation decreases substantially from 70.2 (Table 5-3) to 11.56 on average (Table 5-4), which reduce the variation by 85%. Moreover, the range is also decreased from 7.2 (Table 5-3) to 4.74 on average (Table 5-4). These suggest that temperature controlling technology adoption is important for reducing the variation of tomato keeping quality. It is interesting to know what the impact of this temperature controlling technology adopted by tomato shippers is by comparing the results between temperature controlling technology adopted before and after. Table 5-4 Summary Statistics of the Simulation Model after Technology Keeping Quality (Days) Grade Mean SD CV Min Max Range 1 5.37 4.00 11.02 3.99 7.05 3.07 2 5.48 4.44 12.00 4.28 6.99 2.70 3 5.45 4.51 12.24 0.35 8.63 8.27 4 5.43 4.03 10.99 1.95 7.00 5.05 MEAN 5.43 4.24 11.56 2.64 7.42 4.77 In the simulation, before marine container temperature was controlled, the tomato keeping quality at the end of supply chain averaged 2.7 days (Table 5-3) at temperature of 14.8 0 C (58.0 0 F). However, after temperature was controlled, tomato keeping quality

PAGE 45

34 increased to 5.4 days. Figure 5-2 shows that the probability of tomato keeping quality less than 2.7 days is 55%, more than 5.4 days is only 12%, and between 2.7 days and 5.4 days is 33%. This indicates that the chance for tomatoes to have lower keeping quality is 55% (less than 2.7 days) and the chance to have higher keeping quality is 45% (greater than 2.7 days). StopLight Chart for Probabilities Less Than 2.700 and Greater Than 5.4000.550.330.120%10%20%30%40%50%60%70%80%90%100%Before Tech Keeping Quality (Days) Figure 5-2 Probability of Tomato Keeping Quality before Technology Adoption However, after using technology to control the temperature, the model shows substantial improvement. The probabilities of tomato keeping quality after-technology adoption are illustrated in Figure 5-3. Figure 5-3 indicates that the probability for tomatoes to have a lower keeping quality less than 2.7 days is 0%, to have a higher keeping quality higher than 5.4 days is 51%, and to have keeping quality between 2.7 days and 5.4 days is 49%. Compared to tomato keeping quality before-technology, after temperature is being controlled more closely, the chance that tomato has less than 2.7

PAGE 46

35 days decreases from 55% to 0%. However, the chance for tomatoes to have keeping quality higher than 5.4 days increases from 12% to 51%. This suggests that if temperature in marine containers could be controlled, tomato keeping quality could be improved by 2 times. StopLight Chart for Probabilities Less Than 2.700 and Greater Than 5.4000.000.490.510%10%20%30%40%50%60%70%80%90%100%After Tech Keeping Quality (days) Figure 5-3 Probability of Tomato Keeping Quality After Technology Adoption Furthermore, it is helpful to compare the overall average keeping quality for tomatoes before and after temperature controlling technology is being adopted by tomato grower/shippers. On average, tomato keeping quality has been improved by over 2 times. This is also true for all four individual grades (Figure 5-4). For example, before-technology for grade 3, the average tomato keeping quality at retail level is 2.9 days. However, after temperature is being controlled, tomato keeping quality on average is 5.5 days, which is extended by 1.9 times more than before (Figure 5-4).

PAGE 47

36 5.45.55.55.42.62.42.93.00.01.02.03.04.05.06.0Grade 1Grade 2Grade 3Grade 4GradeTomato Keeping Quality (Days) After Tech Before Tech Figure 5-4 Mean Comparison of Tomato Keeping Quality In summary, for the tomato supply chain between Puerto Rico and the United States, if marine container temperature can be controlled more closely from a range between 11.2C (52.5 o F) and 31.4C (88.5 o F) down to between 13.3C (56.0 o F) to 15.6C (60.0 o F), tomato keeping quality on average will be improved from 2.7 days to 5.4 days, which is 2 times longer than before. Furthermore, the probabilities for tomato to have less than 2.7 days keeping quality drops to 0% and the probability of having greater than 5.4 shelf-life days increases from 12% to 51%. Additionally, the variation of tomato keeping quality decreases sharply by 85% from the calculation of the average coefficient of variation. All these figures imply that tomato supply chain participants total revenue should increase since all supply chain players and consumers prefer better quality product in the United States.

PAGE 48

CHAPTER 6 BENEFIT-COST SENSITIVITY ANALYSIS AND RESULTS Chapter 5 showed that proper application of temperature controlling technology can lead to improved tomato keeping quality. However, participants in the Puerto Rico-United States supply chain also want this question answered: Is this technology adoption worth it from a financial point of view? The benefit-cost sensitivity analysis results presented in this chapter are generated from the information and results from previous chapters. Based on a thorough review of the literature and interviews with actual supply chain participants, production and transportation costs represent the costs and the total revenue represents the benefits respectively for this analysis. Financial risk analysis software, Simetar will be used to build a financial risk analysis model to obtain the results and graphs for the analysis. 6.1 Analysis of Tomato Growers Benefits with Increased Exports Benefit was calculated as total revenue, using export prices and export quantity data collected in Puerto Rico. Cost was calculated as total cost, captured by both production and transportation cost. Based on the results from tomato supply chain simulation model in Chapter 5, one key assumption is made here. When the temperature conditions within marine containers are controlled better, tomato growers will partially regard this as a potential incentive to export more tomatoes to the United States, not only the green ones but also some pink ones. Recall that tomato grower/shippers normally only ship green tomatoes because of shelf life limitations. It is reasonable to assume that if shelf life can be improved through temperature-controlling technology, tomato 37

PAGE 49

38 grower/shippers may be willing to send a percentage of those turning-pink tomatoes that are normally reserved for local Puerto Rican markets. Therefore, its appropriate to assume that improved tomato keeping quality will eventually result in more tomatoes exported to the United States market. The sensitivity of total revenue is analyzed for various possible increased export scenarios, which are exports increased by 5%, 10%, 15% and 20%, as opposed to a zero increase in exports before the adoption of temperature controlling technology. It should be noted that these percentage increases apply only to the current level of tomatoes being exported, not to the entire Puerto Rican tomato crop. For example, before technology adoption, tomato exports to the Unites States are 52.5% of the total crop. An increase of 20% of the 52.5% results in 63% of the total crop now being exported, not 72.5%. These total revenue scenarios were compared to the total cost to find out the possibilities of increased benefit-cost ratio for tomato grower/shippers. The Simetar model is presented in Appendix D. The results of benefit sensitivity analysis based on four different scenarios which indicate varying levels of increased tomato exports (resulting in increased revenues) are illustrated in Figures 6-1 and 6-2. Figure 6-1 shows the probability of benefits over costs for no increases in tomato exports because there was no temperature controlling technology adoption, increased tomato exports by 5% and 10% with technology adoption. The Y axis represents the probabilities and the x axis represents total revenue related to the scenarios of increased export volume. The $4,155,496.64 (Figure 6-1) is 100% of tomato grower/shippers total cost and the number $4,571,046.30 (Figure 6-1) represents 110% of tomato grower/shippers total cost, when there is no tomato export

PAGE 50

39 increasing before technology. The other four numbers in figure 6-1 represents 100% and 110% of tomato grower/shippers total cost when tomato export increased by 5% and 10%, respectively. Therefore, the results indicate that without temperature controlling technology adoption, benefits have a 15% chance less than costs (at normal export levels) and an 85% chance between 100% and 110% of cost, which means benefit-cost ratio has 15% chance of being less than 1.This suggests that there is an 85% chance for tomato grower/shippers to break even However, benefit has 0% chance of being higher than 110% cost-level, which means benefit-cost ratio has no chance to be higher than 1.1, also indicating no chance at all to get higher than a 10% margin. StopLight Chart for Probabilities Less Than 4,155,496.640 and Greater Than 4,571,046.3000.150.850.000%20%40%60%80%100%TR (Q=100%) StopLight Chart for Probabilities Less Than 4,173,215.390 and Greater Than 4,590,536.9300.000.920.080%20%40%60%80%100%TR (Q=105%) StopLight Chart for Probabilities Less Than 4,190,934.140 and Greater Than 4,610,027.5500.000.240.760%20%40%60%80%100%TR (Q=110%) Figure 6-1 Probabilities of Benefit Analysis (1) Finally, with temperature controlling technology adopted by tomato grower/shippers, when tomato export increases by 5%, the benefit have 0% chance of being less than cost, a 92% chance between 100% and 110% of cost, and 8% chance to be higher than 110% of cost. It means that the benefit-cost ratio has a 92% chance to be

PAGE 51

40 between 1.0 and 1.1, and an 8% of being higher than 1.1. This suggests some positive margin exists up to 10%. However, the chance to get higher than a 10% margin is only 8%. When tomato export increases 10%, a benefit have 0% chance to be less than cost, a 24% chance between 100% and 110% of cost and a 76% chance of being higher than 110% of cost, which means that the benefit-cost ratio has 24% chance to be between 1.0 and 1.1, and a 76% chance to be higher than 1.1. These suggest that the margin has 24% chance to be up to 10% and a 76% chance to be higher than 10%. If tomato export increases by 15% and 20%, the chances for increased benefits would be higher, resulting in more margins for tomato grower/shippers. Figure 6-2 shows this situation. The number of $4,610,027.55 (Figure 6-2) is 110% of tomato grower/shippers total cost and $5,029,120.97 (Figure 6-2) represents 120% of tomato grower/shippers total cost, when tomato exports are at the 110% level. The red area in Figure 6-2 represents the possibilities of the benefit-cost ratio being less than 1.1, the yellow area represents the possibilities of the benefit-cost ratio being between 1.1 and 1.2 and the green area represents the possibilities of the benefit-cost ratio being higher than1.2 for all different increased tomato export scenarios. If tomato exports increase by 10%, the benefit has 76% chance of being between 110% and 120% of cost, which means the benefit-cost ratio is between 1.1 and 1.2. However, if tomato exports increase by 15%, the benefit has a 95% chance to be between 110% and 120% of cost, which means the benefit-cost ratio has increased to 95% from 76% for benefit-cost ratios between 1.1 and 1.2. It suggests that tomato grower/shippers have a 95% chance to obtain margins between 10% and 20%. Moreover, if tomato exports increase by 20%, the benefit will have 63% chance to be higher than 120% of

PAGE 52

41 cost, which means the benefit-cost ratio has 63% chance to be higher than 1.2. It suggests that tomato grower/shippers have a 63% chance to reach margins higher than 20%. StopLight Chart for Probabilities Less Than 4,610,027.550 and Greater Than 5,029,120.9700.240.760.000%20%40%60%80%100%TR (Q=110%) StopLight Chart for Probabilities Less Than 4,629,518.180 and Greater Than 5,050,383.4700.000.950.050%20%40%60%80%100%TR (Q=115%) StopLight Chart for Probabilities Less Than 4,649,008.800 and Greater Than 5,071,645.9700.000.370.630%20%40%60%80%100%TR (Q=120%) Figure 6-2 Probabilities of Benefit Analysis (2) In summary, Table 6-1 shows that if temperature controlling technology is adopted, the probabilities for tomato grower to have benefit-cost ratio between 1.0 and 1.1 are 92% and 24% when tomato exports increase by 5% and 10% respectively. And the probabilities for tomato grower/shippers to have a benefit-cost ratio between 1.1 and 1.2 are 8%, 76%, 95% and 37% when tomato exports increase by 5%, 10%, 15% and 20%, respectively. Also the possibilities for tomato grower/shippers to have benefit-cost ratio higher than 1.2 are 5% and 63% when tomato exports increase by 15% and 20%. It suggests that there is 100% chance for tomato grower/shippers to break even and get more margins if they adopt this temperature controlling technology.

PAGE 53

42 Table 6-1 Probabilities of Benefit-cost Ratio for All Scenarios Probabilities Increased Export (0%) Increased Export (5%) Increased Export (10%) Increased Export (15%) Increased Export (20%) B/C < 1.0 15% 0% 0% 0% 0% 1.0 =< B/C < 1.1 85% 92% 24% 0% 0% 1.1 =< B/C < 1.2 0% 8% 76% 95% 37% B/C >=1.2 0% 0% 0% 5% 63% 6.2 Analysis of Tomato Grower/Shippers Costs with Increased Exports All of the above benefit results imply that when temperature controlling technology is adopted by tomato grower/shippers, tomato keeping quality would be improved, which will create more possibilities to obtain higher benefits. However, on average for each tomato export quantity increased, how much could be spent on technology adoption to insure a given margin? Table 6-2 provides the answer. Total revenue on average (Table 6-2) represents the average benefit from the average export price with different tomato export scenarios. Highest possible margin is the highest margin that tomato grower/shippers can get without paying anything for temperature controlling technology. When temperature technology is not adopted by the tomato grower/shippers, the highest margin that tomato growers can have is 2.5% on average (Table 6-2). However, if temperature controlling technology is adopted by tomato grower/shippers, the highest margin tomato growers can have varies from 7.2% to 21% depending on the different levels of tomato export quantity increased. For instance, when tomato exports increase by 5% due to the technology adoption, the highest possible margin for tomato grower/shippers is 7.2%. If tomato

PAGE 54

43 growers/shippers wish to obtain a 5% margin instead of 2.5%, they still have 2.2% margin ($87,947) left paying for the cost of technology. However, if tomato export increases by 10%, the highest margin could be 11.8%. If a 5% margin meets tomato grower/shippers lowest expectation, they can use all or part of the remaining 6.8% margin, which is $273,141 to pay for technology adoption. Or they can choose to pay $70,228 for technology cost and get higher margin at 10%. This suggests that tomato grower/shippers can decide to choose how much they would like to pay for technology cost by reducing the margin from 10% all the way down to 5%. In addition, if a 5% margin is the tomato grower/shippers lowest expectation, when tomato export increases by 15% and 20%, tomato grower/shippers would have 11.4% margin ($458,335) and 16.0% ($643,528) to pay technology cost. However, this might be too much for the temperature controlling technology. Therefore, tomato growers can get at least 10% margin and still have money left to pay for the technology cost. When tomato export increases by 15%, tomato grower/shippers have $246,199 left to pay technology cost. When tomato export increases by 20%, tomato grower/shipper have $422,169 left to pay technology cost. If this is still too much, tomato grower/shippers can decide to have at least 15% margin, with some money left to pay for technology cost for these two last scenarios. In summary, the table 6-2 shows that under any situation, the benefit and cost ratios average greater than 1. However, after temperature controlling technology is adopted, tomato grower/shippers can get far higher margins than before. Even more, for each tomato export increasing scenario, they can choose a different margin level they like and use the rest of them to pay the cost of technology.

PAGE 55

44 Table 6-2 Result of Technology Cost Sensitivity Analysis Categories No Change in Exports (0%) Exports Increased (5%) Exports Increased (10%) Exports Increased (15%) Exports Increased (20%) Total Revenue on Average $4,261,163 $4,474,221 $4,687,279 $4,900,337 $5,113,395 Highest Possible Margin 2.5% 7.2% 11.8% 16.4% 21% Margin (5%) $213,058 $223,204 $233,349 $243,495 Total Possible Cost After Tech $4,261,163 $4,464,075 $4,666,988 $4,869,900 Total Cost Before Tech $4,155,497 $4,173,215 $4,190,934 $4,208,653 $4,226372 Temperature Technology Cost $87,947 $273,141 $458,335 $643,528 Margin (10%) $426,116 $445,485 $464,854 Total Possible Cost After Tech $4,261,163 $4,454,852 $4,648,541 Total Cost Before Tech $4,190,934 $4,208,653 $4,226372 Temperature Technology Cost $70,228 $246,199 $422,169 Margin (15%) $639,174 $666,965 Total Possible Cost After Tech $4,261,163 $4,446,430 Total Cost Before Tech $4,208,653 $4,226372 Temperature Technology Cost $52,510 $220,059 6.3 Benefit-Cost Analysis of Grower/Shippers and Wholesalers with Shrinkage Reduction Within the tomato supply chain between Puerto Rico and the United States, besides tomato grower/shippers, tomato wholesalers are also important participants. If temperature controlling technology is adopted by tomato grower/shippers, one possible benefit for tomato supply chain is to have tomato quantity exported increased, which has

PAGE 56

45 been discussed above. The other possible benefit for the supply chain is to have tomato shrinkage reduced. The benefit and cost associated with this shrinkage reduction is analyzed in this section. Not all tomatoes exported to the United States are sold to retail stores or foodservice operations. Some tomatoes are thrown away due to unacceptable keeping quality. For this portion of the analysis, we are assuming when tomato keeping quality is less than 2 days at wholesale level, these tomatoes will be considered a loss, which reduces total revenue of tomato supply chain. However, after technology is adopted, tomato shrinkage is reduced. Table 6-3 shows that before technology adoption, the tomato supply chain experiences a 41.4% shrinkage and after technology adoption, tomato has only 0.1% shrinkage, which means tomato shrinkage has been reduced by 99.8%. Table 6-3 Growers and Wholesalers Benefits Analysis Regarding to Shrinkage Reduction Before Tech After Tech Difference Shrinkage 41.4% 0.1% 99.8% Value of tomatoes with 2 days or less keeping quality At grower prices $1,764,121 $4,261 $1,759,860 At grower price + 10% $1,940,533 $4,687 $1,935,846 At grower price + 20% $2,116,945 $5,113 $2,111,832 At grower price + 30% $2,293,357 $5,539 $2,287,818 With the new technology, shrinkage is reduced, thereby increasing revenues for those further down the marketing channel that were previously having to dispose of spoiled tomatoes. Before technology adoption, with spoilage assumed at 2 days keeping

PAGE 57

46 quality, shrinkage is 41.4%. At grower prices, the value of this quantity of tomatoes is $1,764,121. However, after technology adoption, the value of tomatoes at 2 days or less keeping quality is only $4,261, a difference of $1,759,860. When tomato shrinkage reduces from 41.4% to 0.1% (Table6-3), the decrease in the cost of spoilage is $1,759,860 in grower prices (Table 6-3). However, wholesalers place a higher value on the tomatoes. Table 6-3 identifies the value of the tomatoes under 2 days keeping quality using 10%, 20%, and 30% mark-ups at the wholesale level. When mark-up between growers and wholesalers is 10%, the possible decrease in spoilage is valued at $1,935,846. When the mark-up is 20%, the possible decrease in spoilage is valued at$2,111,832 and when the mark-up is 30%, the decrease in spoilage is valued a t $2,287,818. In summary, when temperature controlling technology is adopted by tomato supply chain between Puerto Rico and the United States, tomato keeping quality is prolonged, which on one side increase the probability of tomato export to the United States, and on the other side reduce tomatos shrinkage. When shrinkage is reduced, the increased revenue will benefit not only tomato growers, but also other participants, such as tomato wholesalers. Moreover, wholesalers benefit even more than growers from the adoption of the technology. Hence, technology cost could be shared with the supply chain. However, whether the technology should adopt or not depends on the actual cost itself for tomato supply chain between Puerto Rico and the United States.

PAGE 58

CHAPTER 7 CONCLUSIONS 7.1 Summary This research focused on the overall tomato supply chain performance between Puerto Rico and the United States when new temperature controlling technology could be adopted. Among food issues today, food safety and food quality are paid considerable attention. Among all physical conditions, temperature is the most changeable and controllable condition for perishable food quality. Research has shown that American consumers desire higher quality products, and at the same time, Caribbean basin tomato growers, especially Puerto Rican growers are looking for ways to increase their tomato exports to the United States. The primary objective of this research was to find out if the temperature controlling technology is adopted by tomato grower/shippers, can improve tomato keeping quality, leading to increase exports and revenue. Researchers sought to understand the tomato supply chain between Puerto Rico and the United States by gathering as much information as possible. As a result, five different categories of data were collected from tomato firms in Puerto Rico and the United States. Secondary data was collected from the internet and from telephone interviews. To achieve this understanding, a tomato supply chain simulation model between Puerto Rico and the United States was developed using simulation software, Simul8 to find out how tomato keeping quality changes with the adoption of temperature controlling technology. The results indicate that if the temperature range within marine 47

PAGE 59

48 containers is narrowed from 11.2 0 C (52.5 o F) to 31.4 0 C (88.5 o F) down to13.3 0 C (56.0 o F) to 15.6 0 C (60.0 o F), tomato keeping quality would be improved by a factor 2 and shelf life could be extended from 2.7 days to 5.4 days on average. In addition, the variation of tomato keeping quality could be reduced by almost 85%. Finally, based on the results from the tomato supply chain simulation model, a benefit-cost sensitivity analysis was conducted to provide some guidelines for tomato grower/shippers about whether or not adopting this temperature controlling technology would be cost effective. Before temperature controlling technology was adopted, the results indicated that, on average, a tomato grower/shippers benefit is 1.025 times of their cost, which means the highest margin for tomato grower is 2.5%. Furthermore, tomato grower/shippers have a 85% chance to obtain a benefit that is greater than their cost which means benefit-cost ratios are between 1.0 and 1.1, keeping in mind they still have 15% chance that the benefit-cost ratio will be less than 1. If temperature controlling technology is adopted by tomato grower/shippers, the results show that, tomato grower/shippers have a 100% chance of reaching a benefit-cost ratio higher than 1.0, which means 100% chance to break even. The probabilities for tomato grower/shippers to have a benefit-cost ratio between 1.0 and 1.1 are 92% and 24% when tomato export increases by 5% and 10% respectively. And the probabilities for tomato grower/shippers to have a benefit-cost ratio between 1.1 and 1.2 are 8%, 76%, 95% and 37% when tomato export increases by 5%, 10%, 15% and 20%, respectively. Also the possibilities for tomato grower/shippers to have a benefit-cost ratio higher than 1.2 are 5% and 63% when tomato export increases by 15% and 20%.

PAGE 60

49 Further more, if tomato grower/shippers expect to earn a 5% margin, they have $87,947, $273,141, $458,335 and $643,528 to pay for technology cost when tomato quantity exported increased by 5%, 10%, 15% and 20%, respectively. If tomato grower/shippers expect to obtain a 10% margin, they have $70,228, $246,199 and $422,169 to pay for technology cost when tomato quantity exported increased by10%, 15% and 20, respectively. Even If tomato grower/shippers expect to get 15% margin, they have $52,510 and $220,059 to pay for technology cost when tomato quantity exported increased by 15% and 20%, respectively. The benefit-cost sensitivity analysis regarding to tomato shrinkage reduction by technology shows that when tomato shrinkage reduces 99.8% by the adoption of technology, the revenue for tomato growers increases by 70.5%. It also shows that this increase will benefit both growers and wholesalers, which means that the actual total revenue for tomato growers and wholesalers would be increased too depending on different mark-up level. Further study of the specific the benefit-cost impacts for distributors, wholesalers, and retailers are needed due to the difficulty of obtaining cost data. 7.2 Conclusions The results of tomato supply chain simulation model suggest that supply chain complexities can be analyzed by simulation software. Controlling temperature within marine container is one of the effective ways to increase tomato keeping quality as well as reducing its variation when other physical conditions are held constant such as humidity. Furthermore, increased tomato keeping quality means tomatos shrinkage will be reduced. This will generate more volume in the market from Puerto Rico to the United States, which also means that potential tomato export increasing exists.

PAGE 61

50 Whether or not a tomato grower/shipper should adopt temperature controlling technology depends on the benefit-cost relationship and financial performance. The results of benefit-cost sensitivity analysis suggest that whether or not tomato grower/shippers should adopt this temperature controlling technology depending not only on the possibility of achieving a better benefit-cost ratio, but also on the cost of the technology itself. The bottom line is tomato grower/shippers and other tomato supply chain participants will have a greater chance to earn more money and enlarge their market if temperature technology is adopted within this tomato supply chain between Puerto Rico and the United States. Overall, this research suggests that the opportunity for Puerto Rico tomato grower/shippers to increase tomato keeping quality, leading to more volume and shrinkage reduction of tomatoes exported to the United States, resulting in higher revenues and margins, if tomato grower/shippers decide to adopt the temperature controlling technology within marine containers. 7.3 Implications The volume and value of tomatoes sold and consumed in the U.S. food system is substantial Not only because their significant nutritional value, but also because tomatoes are one of the most consumed fruits and vegetables in the United States. Therefore, the opportunity to put more tomatoes into the United States market from Puerto Rico and other Caribbean Basin countries exists. However, one key thing importers will have to pay attention to is tomato quality since American consumers desire high quality products. This research quantifies the costs and benefits associated with effective controlling of temperatures, especially during marine shipment for Puerto Rican tomato grower/shippers. If tomato keeping quality is improved, exports of tomatoes into the

PAGE 62

51 United States will increase along with the possibility for increased margins. This method can also be adopted by all other tomato grower/shippers from other Caribbean Basin countries. The simulation model and benefit cost sensitivity analysis can be applied to other perishable goods. This analysis can be used to find the temperature bottle-neck for their supply chain and temperature controlling technology can be applied at critical points. The model can be expanded to measure the impact of other variables related to food quality other than just temperature, such as humidity. Finally, further research is warranted using the simulation and benefit-cost approach to studying food supply chains. Innovations would include the utilization of more detailed financial information for tomato distributors, wholesalers and retailers. This approach to studying food supply chains could be applied not only to state to state trade within the U.S., but also to perishable foods other than tomatoes.

PAGE 63

APPENDIX A PICTURES OF TOMATO SUPPLY CHAIN DISTRIBUTION SYSTEM Figure A-1 Tomato Dumping Into the Packing Line Figure A-2 Tomato Grading Line 52

PAGE 64

53 Figure A-3 Tomato Packing Line (25LB/box) Figure A-4 Tomato Palletizing Operation (80 boxes/ pallet)

PAGE 65

54 Figure A-5 Tomatoes Stored in an Ethylene Room Figure A-6 Tomato Loading in Santa Isabel, Puerto Rico

PAGE 66

55 Figure A-7 Truck Shipments Leaving Santa Isabel, Puerto Rico Figure A-8 Trucks Arriving at the Quincy FL Distribution Center

PAGE 67

APPENDIX B SIMULATION MODELING PARAMETERS Before Technology Adoption Parameters for Clock Properties: Time unit: minute; 7 days / week; start from 00:00 and 24:00 hours per day; Collection period: 3months (90*24*60 =129600 minutes) Parameters for Preferences: Distance: travel time = 0 Continues: units measure = tons ( 1ton=2200 lbs ) Labels and Distributions: Quantity: 1 tons = 2200 lbs Grade: lbl_GradeMix with probability profile distribution of dist_GradeMix (1-20%; 2-5%; 3-60%; 4-15%) Size: lbl_Size with probability profile distribution of dist_Size (1-25% extra large; 2-60% large; 3-15% medium) Image: Image:Tomato ( 1 red tomato for local; 2 green tomato for US) Color: lbl_ColorMix with probability profile distribution of dist_ColorMix (1-30% red; 2-70% green) Parameters for Harvesting: Inter-arrival times (minutes): average 1.4 (480/341) with exponential distribution Label Actions: Quantity set to fixed value 2200 with fixed distribution Lbl_ColorMix set to the value of dist_ColorMix Image:Tomato set to the value of lbl_ColorMix Parameters for Sorting: Timing: average 1.4 (480/341) with exponential distribution Resources: picker and packer pool (number required from 60 to 80) Label Action: lbl_GradeMix set to the value of dist_GradeMix lbl_ColorMix set to the value of dist_ColorMix lbl_Size set to the value of dist_Size Routing Out: Percentage (25% disposal and 75% for packing ) 56

PAGE 68

57 Parameters for Packing: Timing: average 0.016 (1.4*25/2200 ) with the distribution of exponential Resources: picker and packer pool (number required from 20 to 25) Routing In: Collect 25, assemble, do not collect until all available Routing Out: label ( lbl_ColorMix; 1red to local; 2-green to US) Label Actions: lbl_GradeMix set to fixed value of 1, 2, 3 and 4 for 4 packing lines; lbl_ColorMix set to the value of the distribution dist_ColorMix Image:Tomato set to the value of fixed value of lbl_ColorMix Parameters for Palletizing: Timing: 2 to 5 minutes with the distribution of uniform Resources: picker and packer pool (number required from 2 to 4) Routing In: Collect 80, assemble, do not collect until all available Routing Out: label ( lbl_Size; 1extra large to room 1; 2-large to room2; 3-medium to room 3) Label Actions: Lbl_Size set to the value of the distribution dist_Size lbl_GradeMix set to the value of the distribution dist_GradeMix Parameters for Ethylene Room: Room 1: min wait time 5040 minutes (12*7*60 ), high volume Room 2 and 3: min wait time 3600 minutes (12*5*60), high volume Parameters for Truck Loading: Timing: average 5 minutes with the distribution of average Resources: truck loader (number required from 3 to 5) Routing In: Collect 18, assemble Routing Out: on complete VL of quality equation and Travel time Label Actions: lbl_GradeMix set to the value of the distribution dist_GradeMix Lbl_Size set to the value of the distribution dist_Size Labels, Distributions and Information Stores Labels: lbl_Quality, lbl_TimeStamp Distributions: dist_TravelTimeTrPR (bounded normal; 120, 10, 240, 90) dist_TravelTimeTrUS (bounded average; 180, 240, 120) dist_TravelTimeShip (uniform; 2880, 4320) dist_DispatchtoEndConsumers (1-5%-Canada; 2-5%-Food Services 3-14%-Groceries; 4-76%-Wholesalers) dist_Tabs (Uniform; 284.39, 304.52, min & max Tabs in marine container) Information Stores: int_TravelTimePR (number), int_TravelTimeShip (number), int_TravelTimeUS (number), int_K (number; reaction rate within marine container) int_Tabs (number; Tabs within marine container) ss_Temperature (spread sheet for input temperature data)

PAGE 69

58 ss_Log (spread sheet for outputting) gbl_LogRow (number for ss_Log use) Parameters for Truck Unloading PR Timing: average (5) Resources: Parameters for Ship Loading Timing: average (15) Resources: Routing Out: on complete VL of ship travel time Parameters for Ship Unloading Timing: average (15) Resources: Routing Out: on complete VL of quality equation Parameters for Truck Loading US Timing: average (5) Resources: Routing Out: on complete VL of Truck Travel Time Parameters for Truck Unloading US Timing: average (5) Resources: Routing Out: on complete VL of quality equation Parameters for Cold Room Timing: average (10) Resources: Routing Out: on complete VL of quality equation, On exit VL of the log of tomatoes out, Percent: 5% (Canada); 5% (Food Services); 14% (Groceries); 76% (Wholesalers) After Technology Adoption Compared to parameters before technology, except the following parameters have been changed, all other parameters for tomato supply chain simulation model after technology keep the same.

PAGE 70

59 Labels, Distributions and Information Stores Labels: lbl_Quality, lbl_TimeStamp Distributions: dist_TravelTimeTrPR (bounded normal; 120, 10, 240, 90) dist_TravelTimeTrUS (bounded average; 180, 240, 120) dist_TravelTimeShip (uniform; 2880, 4320) dist_DispatchtoEndConsumers (1-5%-Canada; 2-5%-Food Services 3-14%-Groceries; 4-76%-Wholesalers) dist_Tabs (Uniform; 286.48, 288.71) Information Stores: int_TravelTimePR (number), int_TravelTimeShip (number), int_TravelTimeUS (number), int_K (number; reaction rate within marine container) int_Tabs (number; Tabs within marine container) ss_Temperature (spread sheet for input temperature data) ss_Log (spread sheet for outputting) gbl_LogRow (number for ss_Log use)

PAGE 71

APPENDIX C SIMULATION MODELING PROGRAMMING VL SECTION: fnc_Quality Function 'fnc_Quality Function SET lbl_Quality = lbl_Quality*EXP[0-[[ss_Temperature[6,int_Row]]*[[Simulation Time-lbl_TimeStamp]/[1440]]]] VL SECTION: fnc_Tiime Stamp SET lbl_TimeStamp = Simulation Time VL SECTION: Reset Log Logic1 'Obeyed just after all simulation objects are initialized at time zero 'Resets the ss_log spreadsheet LOOP 1 >>> int_Row >>> gbl_LogRow LOOP 1 >>> int_col >>> 6 SET ss_Log[int_col,int_Row+1] = "" SET gbl_LogRow = 0 VL SECTION: Packing 1 Quality Equation Work Complete Logic 'fnc_Quality Calculation at Packing SET lbl_Quality = 100 SET int_Row = 42 CALL fnc_Quality Function SET lbl_TimeStamp = Simulation Time VL SECTION: Packing 1 Time Stamp Route In After Logic CALL fnc_Tiime Stamp VL SECTION: Packing 2 Quality Equation Work Complete Logic 'fnc_Quality Calculation at Packing SET lbl_Quality = 100 SET int_Row = 42 CALL fnc_Quality Function SET lbl_TimeStamp = Simulation Time VL SECTION: Packing 2 Time Stamp Route In After Logic CALL fnc_Tiime Stamp VL SECTION: Packing 3 Quality Equation Work Complete Logic 'fnc_Quality Calculation at Packing SET lbl_Quality = 100 60

PAGE 72

61 SET int_Row = 42 (Disabled) SET int_Pre Shelf Life = ss_Temperature[5,4] (Disabled) SET int_Aft Shelf Life = ss_Temperature[5,4] CALL fnc_Quality Function SET lbl_TimeStamp = Simulation Time VL SECTION: Packing 3 Time Stamp Route In After Logic CALL fnc_Tiime Stamp VL SECTION: Packing 4 Quality Equation Work Complete Logic 'fnc_Quality Calculation at Packing SET lbl_Quality = 100 SET int_Row = 42 CALL fnc_Quality Function SET lbl_TimeStamp = Simulation Time VL SECTION: Packing 4 Time Stamp Route In After Logic CALL fnc_Tiime Stamp VL SECTION: Palletizing US Action Logic1 IF lbl_Size = 1 SET lbl_route = 1 ELSE SET lbl_route = 2 VL SECTION: Pelltizing US Quality Equation Work Complete Logic 'fnc_Quality Calculation at Palletizing SET int_Row = 42 CALL fnc_Quality Function SET lbl_TimeStamp = Simulation Time VL SECTION: To Loading Route-In After Logic CALL fnc_Tiime Stamp VL SECTION: (After Ethylene Room) To Loading Quality Equation Work Complete Logic SET lbl_TimeStamp = Simulation Time VL SECTION: Truck Loading PR Time Stamp Route In After Logic CALL fnc_Tiime Stamp VL SECTION: PR Truck Travel Time and Quality Equation Work Complete Logic 'Truck Travel Time from Santa Isabel to San Juan SET int_TravelTimePR = dist_TravelTimeTrPR Set Travel Time Truck Loading PR int_TravelTimePR Queue for Truck Unloading PR

PAGE 73

62 'fnc_Quality Calculation at Palletizing SET int_Tabs = dist_Tabs SET int_K = EXP[[0-[7277.48*[1/int_Tabs]]]+24.223] 'Calculate Quality according to Temperature distribution within Marine Container SET lbl_Quality = lbl_Quality*EXP[[0-[int_K]]*[[Simulation Time-lbl_TimeStamp]/[1440]]] SET lbl_TimeStamp = Simulation Time VL SECTION: Ship Loading Time Stamp Route In After Logic CALL fnc_Tiime Stamp VL SECTION: Ship Travel Time Work Complete Logic 'Ship Travel Time from San Juan to Jacksonville SET int_TravelTimeShip = dist_TravelTimeShip Set Travel Time Ship Loading int_TravelTimeShip Queue for Ship Unloading VL SECTION: Ship Unloading Quality Equation Work Complete Logic 'fnc_Quality Calculation at Ship Unloading SET int_Tabs = dist_Tabs SET int_K = EXP[[0-[7277.48*[1/int_Tabs]]]+24.223] 'Calculate Quality according to Temperature distribution within Marine Container SET lbl_Quality = lbl_Quality*EXP[[0-[int_K]]*[[Simulation Time-lbl_TimeStamp]/[1440]]] SET lbl_TimeStamp = Simulation Time VL SECTION: Truck Loading US Time Stamp Route In After Logic CALL fnc_Tiime Stamp VL SECTION: Truck Unloading US Quality Equation Work Complete Logic 'fnc_Quality Calculation at Truck Unloading US SET int_Tabs = dist_Tabs SET int_K = EXP[[0-[7277.48*[1/int_Tabs]]]+24.223] 'Calculate Quality according to Temperature distribution within Marine Container SET lbl_Quality = lbl_Quality*EXP[[0-[int_K]]*[[Simulation Time-lbl_TimeStamp]/[1440]]] SET lbl_TimeStamp = Simulation Time VL SECTION: US Truck Travel Time Work Complete Logic 'Travel Time from Jacksonville to Quincy SET int_TravelTimeUS = dist_TravelTimeTrUS Set Travel Time Truck Loading US int_TravelTimeUS Queue for Truck Unloading US VL SECTION: Truck Unloading US Time Stamp Route In After Logic CALL fnc_Tiime Stamp

PAGE 74

63 VL SECTION: (Out of Distribution Center) Dispatch to End Consumers Quality Equation Work Complete Logic 'fnc_Quality Calculation After Distribution Center Cold Room SET int_Row = 48 CALL fnc_Quality Function SET lbl_TimeStamp = Simulation Time SET lbl_route = dist_DispatchToEndConsumers VL SECTION: Log of Tomato Out of Distribution Center on Exit Logic SET gbl_LogRow = gbl_LogRow+1 'To record the number of days since the tomato were harvested SET ss_Log[1,gbl_LogRow+1] = [Simulation Time-lbl_Time Harvested]/1440 SET ss_Log[2,gbl_LogRow+1] = lbl_GradeMix SET ss_Log[3,gbl_LogRow+1] = lbl_Size SET ss_Log[4,gbl_LogRow+1] = lbl_Quality SET ss_Log[5,gbl_LogRow+1] = Quantity

PAGE 75

APPENDIX D BENEFIT-COST SENSITIVITY ANALYSIS SIMETAR MODELING Model Input ((for 6.1 &6.2) A B C D E F G 4 Production 5 Unit Mean Min Max Remarks 6 Daily Production LB 75,000 7 Export period (1) days 59 January and February 8 Export period (2) days 31 March 9 Total Export Period days 90 =SUM(D7:D8) 10 Total Production in Export Period (1) LB 4,425,000 =D6*D7 11 Total Production in Export Period (2) LB 2,325,000 =D6*D8 12 Total Production in Export Period LB 6,750,000 =SUM(D10:D11) 13 14 Quantity Sold 15 Disposal % 25% 16 U.S. Export % 52.5% =D18*70% 17 Local Market % 22.5% =D18*30% 18 Available Production % 75% 19 Quantity Export to U.S. in Period (1) LB 2,323,125 =D10*D16 20 Quantity Export to U.S. in Period (2) LB 1,220,625 =D11*D16 21 Total Quantity Export to U.S. LB 3,543,750 =SUM(D19:D20) 22 23 Export Price 24 Export Period (1) $ / LB 1.54 1.48 1.6 $37-$40 / box (25LB) 25 Export Period (2) $ / LB 0.56 0.52 0.6 $13-$15 / box (25LB) 26 27 Costs 28 Total Fixed Cost $/Acre/Year 1,623.06 29 Total Variable Cost $/Acre/year 10,235.14 64

PAGE 76

65 30 Pl anted Acres Acres 2000 31 Total Production Cost $/Year 23,716,400.0 0 =(D28+D29)*D30 32 Total Production Cost for Export Period $/90days 5 ,847,879.45 =(D31/365)*90 33 Transportation Cost $/LB 0. 1 0.1 34 Total Transportation Cost for Export Period $ 3 54,375.00 =D33*D21 Model Output 1 (for 6.1 &6.2) 36 Quantity Export to U.S. in Period (1) LB 2, 323,125 =D19 37 Export Price in Period (1) $/LB 1. 54 =UNIFORM(E24,F24) 38 Total Revenue in Period (1) $ 3, 577,612.50 =D36*D37 39 Quantity Export to U.S. in Period (2) LB 1, 220,625 =D20 40 Export Price in Period (2) $/LB 0. 56 =UNIFORM(E25,F25) 41 Total Revenue in Period (2) $ 6 83,550.00 =D39*D40 42 Total Revenue from Export (100%) $ 4 ,261,162.50 =D38+D41 43 Total Cost of Export $ 4, 155,496.64 =D32*(D16+0.5*D15) +D34 This is the Expected Value model, wh ich means that unit D36 and unit D39 are mean value1.54 and 0.56. However, D36 and D39 are stochastic variables with the uniform distribution with the min and max as the parameters. If the key of Expected Value be pressed again, the model will be appeared in Stochastic Value model, which means the number in unit D36 and unit D39 both can be changed any time according to their distribution. The output model (2) is also the Expect ed Value model like the above. All the revenues are expected mean values. If the st ochastic model is chosen, all the revenue numbers will be changed due to the stochastic price.

PAGE 77

Model Output (2) Sensitivity Analysis for Benefit (for 6.1 &6.2) 66 A B C D E F G H 48 Export Quantity Q(LB) Period 1 Remarks Period 2 Remarks Total Quantity Remarks 49 Q=105% 2,439,281 =$D$36*105% 1,281,656 =$D$39*105% 3,720,938 =C49+E49 50 Q=110% 2,555,438 =$D$36*110% 1,342,688 =$D$39*110% 3,898,125 =C50+E50 51 Q=115% 2,671,594 =$D$36*115% 1,403,719 =$D$39*115% 4,075,313 =C51+E51 52 Q=120% 2,787,750 =$D$36*120% 1,464,750 =$D$39*120% 4,252,500 =C52+E52 53 54 Revenue ($) Period 1 Remarks Period 2 Remarks Total Revenue ($) Remarks 55 Q=105% 3,756,493.13 =C49*$D$37 717,727.50 =E49*$D$40 4,474,220.63 =C55+E55 56 Q=110% 3,935,373.75 =C50*$D$37 751,905.00 =E50*$D$40 4,687,278.75 =C56+E56 57 Q=115% 4,114,254.38 =C51*$D$37 786,082.50 =E51*$D$40 4,900,336.88 =C57+E57 58 Q=120% 4,293,135.00 =C52*$D$37 820,260.00 =E52*$D$40 5,113,395.00 =C58+E58

PAGE 78

67 A B C D E F G H 61 100% Cost 110% Cost 120% Cost 62 Total Cost of Export ($) Q=100% 4,155,496.64 4,155,496.64 4,571,046.30 4,986,595.97 =$D$32*($D$16+0.5*$D$15)+$D$34 63 Total Cost of Export ($) Q=105% 4,173,215.39 4,173,215.39 4,590,536.93 5,007,858.47 =$D$32*($D$16+0.5*$D$15)+G49*$D$33 64 Total Cost of Export ($) Q=110% 4,190,934.14 4,190,934.14 4,610,027.55 5,029,120.97 =$D$32*($D$16+0.5*$D$15)+G50*$D$33 65 Total Cost of Export ($) Q=115% 4,208,652.89 4,208,652.89 4,629,518.18 5,050,383.47 =$D$32*($D$16+0.5*$D$15)+G51*$D$33 66 Total Cost of Export ($) Q=120% 4,226,371.64 4,226,371.64 4,649,008.80 5,071,645.97 =$D$32*($D$16+0.5*$D$15)+G52*$D$33

PAGE 79

68 The followings are the input and output model for 6.3 in Chapter 6. The model shows in the Expected Value style so all the numbers represents the mean value. However, this model can be showed by Stochastic Style in Simetar. Model (for 6.3) A B C D E F G 2 Input Data 3 4 Export Quantity 5 Quantity Export to U.S. in Period (1) LB 2,323,125 6 Quantity Export to U.S. in Period (2) LB 1,220,625 7 Total Quantity Export to U.S. LB 3,543,750 8 9 Export Price Mean Min Max 10 Export Period (1) $ / LB 1.54 1.48 1.6 $37-$40 / box (25LB) 11 Export Period (2) $ / LB 0.56 0.52 0.6 $13-$15 / box (25LB) 12 13 Shrinkage 14 Before Tech % 41.4% Shelf Life less than 2 days 15 After Tech % 0.1% Shelf Life less than 2 days A B C D E F G 18 Before Technology Model 19 20 Actual Quantity Sold 21 Actual Quantity Sold to U.S. in Period (1) LB 1,361,351 =D5*(1-D14) 22 Actual Quantity Sold to U.S. in Period (2) LB 715,286 =D6*(1-D14) 23 Total Actual Quantity Sold to U.S. LB 2,076,638 =SUM(D21:D22) 24 25 Export Price 26 Export Period (1) $ / LB 1.54 =uniform(E10,F10) 27 Export Period (2) $ / LB 0.56 =uniform(E11,F11) 28 29 Actual Revenue 30 Actual Revenue in Period (1) $ 2,096,481 =D26*D21 31 Actual Revenue in Period (2) $ 400,560 =D27*D22 32 Total Actual Revenue $ 2,497,041 =SUM(D30:D31)

PAGE 80

69 A B C D E F G 35 After Technology Model 36 37 Actual Quantity Sold 38 Actual Quantity Sold to U.S. in Period (1) $ 2,320,802 =D5*(1-D15) 39 Actual Quantity Sold to U.S. in Period (2) $ 1,219,404 =D6*(1-D15) 40 Total Actual Quantity Sold to U.S. $ 3,540,206 =SUM(D38:D39) 41 42 Export Price 43 Export Period (1) $ / LB 1.54 =uniform(E10,F10) 44 Export Period (2) $ / LB 0.56 =uniform(E11,F11) 45 46 Actual Revenue 47 Actual Revenue in Period (1) $ 3,574,035 =D43*D38 48 Actual Revenue in Period (2) $ 682,866 =D44*D39 49 Total Actual Revenue $ 4,256,901 =SUM(D47:D48) 50 51 Total Actual Revenue increased $ 1,759,860 =D49-D32 52 Total Actual Revenue increased % 70.5% =(D49-D32)/D32 53 shrinkage Reduced % 99.8% =(D14-D15)/D14

PAGE 81

LIST OF REFERENCES Chang, R. (1981). Physical Chemistry with Applications to Biological Systems. Macmillan Publishing Co., New York. Caswell, J.A., Bredahl, M.E., and Hooker, N. H. (1998). How Quality Management Metasystems Are Affecting the Food Industry. Review of Agricultural Economics. Vol.20, No. 2, pp. 547-557. Deming, W.E. (1967). What Happened in Japan? Industry Quality Control. Vol. 24, pp. 89-93. Deming, W.E. (1972). Report to Management. Quality Process. Vol. 5, No. 2-3, pp. 41. Economic Research Service. United States Department of Agricultural. (2004). Commodity Highlight: Fresh Tomatoes. http://www.ers.usda.gov/Briefing/Vegetables/vegpdf/FrTomatoHigh.pdf, Date visited, 02/15/2006. Economic Research Service. United States Department of Agricultural. (2006). the F.o.b.-Retail Price Relationship For Selected Fresh Vegetables. http://www.ers.usda.gov/Briefing/Tomatoes/tomatopdf/FOBRetailPriceVeg.pdf, Date visited, 02/15/2006 Food and Agricultural Organization. (2003). FAOSTAT Trade Data. http://faostat.fao.org/site/342/default.aspx Date visited, 10/25/2004 Food Marketing Institute. (2003). Food Industry Review 2003. Food Marketing Institute, Washington, D.C. Heien, D. M. (1980). Markup Pricing in a Dynamic Model of the Food Industry. American Journal of Agricultural Economics. Vol. 62, pp. 11-18 Institute of Food and Agricultural Science. (2003-2004). Florida Tomato Growers Production Cost. http://www.agbuscenter.ifas.ufl.edu/cost/cop03-04/SWTomatoSC.doc, Date Visited, 06/05/2005 Juran, J. M. (1974). Quality Control Handbook. McGraw-Hill. New York. Juran, J. M., and Gryna, F. M. Jr. (1970). Quality Planning and Analysis. McGraw-Hill. New York. 70

PAGE 82

71 Marquardt, D.W. (1984). New Technology and Educational Directions for Managing Product Quality. The American Statistician. Vol.38, No. 1, pp. 8-14. National Oceanic & Atmospheric Administration. (2005). 30 Years Average Temperature. http://www.ncdc.noaa.gov/oa/climate/online/ccd/nrmavg.txt, Date Visited 10/05/2005 Nunes, Cecilia (2005). Tropical and Sub-Tropical Agricultural Research Tomato project process report. October, Gainesville, Florida Powers, N. J. (1995). Sticky Short Run Prices and Vertical Pricing: Evidence from the Market for Iceberg Lettuce. Agribusiness. Vol. 11, No. 1, pp. 57-75 Robertson, G. L. (1993). Food Packaging. Marcel Dekker Inc. New York, NY. pp. 338-352. Science Made Simple. (2005). Temperature Conversions (From Fahrenheit to Kelvin). http://www.sciencemadesimple.net/temperature.php, Date Visited 10/05/2005 Schepers, H.E., Henten E. J. V., Bontsema, J. and Dijksterhuis, G.B. (2004). Tactics of Quality Management and Promotions: Winning Consumers for Fresh Exotic Produce. Dynamics in Chains and Networks. Wageningen Academic Publishers. Wageningen. The Netherlands. pp. 568-582. SIMUL8 Corporation. (2006). About SIMUL8 Corporation. http://www.simul8.com/about/index.htm, Date Visited 05/15/2006. SIMUL8 Corporation. (2006). SIMUL8 Case Studies. http://www.simul8.com/products/standard/casestudies.htm, Date Visited 05/16/2006. Tijskens, L. M. M., Sloof, and M. & Wilkinson, E. C. (1994). Quality of perishable produce. A Philosophical Research. In proceedings COST94 Workshop, October, Wageningen, The Netherlands. Tijskens, L. M. M., and Polderdijk, J. J. (1996). A Generic Model for Keeping Quality of Vegetable Produce During Storage and Distribution. Agricultural Systems. Vol. 51, No. 4, pp. 431-452. Ward, R.W. (1982). Asymmetry in Retail, Wholesale and Shipping Point Pricing for Fresh Vegetables. American Journal of Agricultural Economics. Vol.64, No.2, pp. 205-212 Whittington, D. and Grubb, W.N. (1984). Economic Analysis in Regulatory Decisions: The Implication of Executive Order 12291. Science, Technology, & Human Values. Vol. 9, No. 1, pp. 63-71

PAGE 83

72 Worth, Thomas. (1999). The F.O.B. Retail Price Relationship For Selected Fresh Vegetables. Economic Research Services. United State Department of Agricultural http://www.ers.usda.gov/Briefing/Tomatoes/tomatopdf/FOBRetailPriceVeg.pdf, Date visited. 02/15/2006

PAGE 84

BIOGRAPHICAL SKETCH Jiaoju Ge was born in Hubei province, P.R. China. After she obtained her Bachelor of Science degree in investment economics from Sichuan University in 1999, she worked for LG Electronics for 3 years as a marketing product manager. In August 2004, Jiaoju Ge entered the food and resource economics Master of Science program and specialized in marketing in University of Florida under the direction of Dr. Allen Wysocki. She will continue her Ph.D. program in the Food and Resource Economics Department beginning on August 2006. She was married to Qiyong Xu on May 21 st 2002. 73


Permanent Link: http://ufdc.ufl.edu/UFE0015882/00001

Material Information

Title: Simulation Modeling and Benefit-Cost Sensitivity Analysis for Technology Adoption on Puerto Rico-United States Tomato Supply Chain
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0015882:00001

Permanent Link: http://ufdc.ufl.edu/UFE0015882/00001

Material Information

Title: Simulation Modeling and Benefit-Cost Sensitivity Analysis for Technology Adoption on Puerto Rico-United States Tomato Supply Chain
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0015882:00001


This item has the following downloads:


Full Text












SIMULATION MODELING AND BENEFIT-COST SENSITIVITY ANALYSIS FOR
TECHNOLOGY ADOPTION ON PUERTO RICO-UNITED STATES TOMATO
SUPPLY CHAIN















By

JIAOJU GE


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2006

































Copyright 2006

by

Jiaoju Ge
































This thesis is dedicated to my loving husband Qiyong and our expected son.















ACKNOWLEDGMENTS

First, I express my deepest appreciation to all my committee, Dr. Allen Wysocki,

Dr. Bruce Welt and Dr. Lisa House.

Foremost, I would like to express my sincere gratitude to my committee chair, Dr.

Allen Wysocki, for his understanding, guidance and support throughout my master's

study in University of Florida. I also would like to thank my other committee members

for their participation. I thank Dr. Lisa House for her advisement, valuable ideas and

kindness. I thank Dr. Bruce Welt for his advice and the knowledge he shared.

I wish to thank Dr. Cecilia do Nascimento Nunes and Sharon Dea for their kindness

and cooperation on data collection. I would also like to thank my fellow graduate

students in the Food and Resource Economics Department for their support, suggestions

and all the unforgettable memories.

I thank my families and my friends for their support. Finally, the greatest thank you

goes to my husband, Qiyong Xu, for his understanding, encouragement and love.
















TABLE OF CONTENTS



A C K N O W L E D G M E N T S ......... .................................................................................... iv

LIST OF TABLES ........................ ........ .. ...... ............. vii

LIST OF FIGURES ............ .......... ............................... viii

CHAPTER

1 IN TR OD U CTION ............................................... .. ......................... ..

1.1 Overview of Tomato Supply Chain........................................................ .......... 2
1.2 Problem Statem ent .................. ..................................... .. ........ ....
1.3 O bjectiv es .................................................................... 6
1.4 Testable Hypotheses ................................................................ ...............
1.5 R research Scope .................................................................... 7

2 LITER A TU RE REV IEW ................... .... .......................... .. ....... .......... .... ....

2.1 Physical Conditions and Tomato Quality............... ............................................ 8
2.2 Sim ulation System M odeling ........................ ... ............. ............... .... 10
2.3 Economic Analysis of Technology Adoption and Price Asymmetry .................11

3 M ETHOD S ...................................... ................................. ........... 14

3.1 To Find Tomato Keeping Quality Equation .....................................................14
3.2 To Build Tomato Supply Chain Simulation Model............................................15
3.3 To Conduct A Benefit-Cost Sensitivity Analysis.....................................16

4 D A T A C O L L E C T IO N ................................................................. .....................17

4.1 Production Related D ata A nalysis............................................... .................. 18
4.2 Processing Time Related Data Analysis.................................... ............... 19
4.3 L abor R elated D ata A analysis .................................... .... ..................... ........... 20
4.4 Temperature and Shelf Life Related Data Analysis ...........................................21
4.5 Financial Related D ata A nalysis...................................... ......................... 23

5 SIMULATION MODELING SPECIFICATION AND RESULTS...........................27

5.1 K keeping Quality Equation and Results ...............................................................27









5.2 Tomato Supply Chain Simulation Model and Results ......................................29
5.2.1 Simulation Model before Technology and Results ..................................29
5.2.2 Simulation Model after Technology and Results Comparison .................32

6 BENEFIT-Cost SENSITIVITY ANALYSIS AND RESULTS..............................37

6.1 Analysis of Tomato Growers' Benefits with Increased Exports ........................37
6.2 Analysis of Tomato Grower/Shippers' Costs with Increased Exports................42
6.3 Benefit-Cost Analysis of Grower/Shippers and Wholesalers with Shrinkage
R e d u ctio n ............ .................................................................. 4 4

7 C O N C L U SIO N S ....................... .... .... ...................... ................ ...... ......... 47

7.1 Sum m ary ............................................................................................... 47
7 .2 C o n c lu sio n s ..................................................................................................... 4 9
7 .3 Im p licatio n s ................................................................5 0

APPENDIX

A PICTURES OF TOMATO SUPPLY CHAIN DISTRIBUTION SYSTEM ...........52

B SIMULATION MODELING PARAMETERS ........................................................ 56

B before T technology A adoption ............................................................................... 56
After Technology Adoption....................................... ......... 58

C SIMULATION MODELING PROGRAMMING ............... ...............................60

D BENEFIT-COST SENSITIVITY ANALYSIS SiMeTAR MODELing .................64

M odel Input ((for 6 .1 & 6 .2) .................................................................................. 64
M odel O utput 1 (for 6.1 & 6.2) ............................................... ...................65
Model Output (2) Sensitivity Analysis for Benefit (for 6.1 &6.2) ..........................66
M o d e l (fo r 6 .3 )...................................................................................................... 6 8

LIST OF REFERENCES ..................................... ..................70

B IO G R A PH IC A L SK E T C H ....................................................................................... 73
















LIST OF TABLES


Table page

4-1 P processing T im e D ata.......................................................................... ............... 20

4-2 Labor Input D ata .............. ......... ........ ............................. .... ...... 21

4-3 Tom ato Shelf Life and Tem perature ........................................ ....................... 22

4-4 Temperature Distribution of Marine Container (C) ................................................23

4-5 Tomato Operation and Production Cost (Average Per Acre)............................... 24

4-6 U.S. Average Monthly Retail Price and f.o.b. Shipping Point............................... 26

5-1 Statistics of R action R ate ......... ................. ...........................................................27

5-2 Predicted Results of Reaction Rate ........... ............ ........... ....... ........ 28

5-3 Summary Statistics of Simulation Model Before Technology................................31

5-4 Summary Statistics of the Simulation Model after Technology...............................33

6-1 Probabilities of Benefit-cost Ratio for All Scenarios.............. ......... ............... 42

6-2 Result of Technology Cost Sensitivity Analysis ............... ............................... 44

6-3 Growers and Wholesalers Benefits Analysis Regarding to Shrinkage Reduction......45
















LIST OF FIGURES


Figure p

1-1 Top Factors in Selecting Primary Supermarket....................... ................. 1

1-2 Tom ato Supply Chain D distribution Process ........................................ .....................5

2-1 Tomato (x) Keeping Quality and Temperature Relationship..................................10

2-2 The Relationship Between Three Distinct Research Fields In a Dynamic
Sim ulation M odel ...................... ....... ............... .. .......... .. .......... 11

3-1 Puerto Rico-United States Tomato Supply Chain Components..............................15

4-1 Export Tom ato Size D istribution.......................................... ........................... 18

4-2 Export Tom atoes Grade Distribution ............................................... ............... 19

5-1 Supply Chain Simulation Model before and After Technology Applied....................30

5-2 Probability of Tomato Keeping Quality before Technology Adoption....................34

5-3 Probability of Tomato Keeping Quality After Technology Adoption ......................35

5-4 Mean Comparison of Tomato Keeping Quality .................................. ...............36

6-1 Probabilities of B benefit A analysis (1) ........................................ ....................... 39

6-2 Probabilities of B benefit A analysis (2) ........................................ ....................... 41

A-i Tomato Dumping Into the Packing Line ........................................ ............... 52

A -2 T om ato G reading L ine ........................................................................ ...................52

A -3 Tom ato Packing Line (25LB/box).................................................................. ...... 53

A-4 Tomato Palletizing Operation (80 boxes/ pallet).......................................................53

A-5 Tomatoes Stored in an Ethylene Room ........................................... ............... 54

A-6 Tomato Loading in Santa Isabel, Puerto Rico .................. .................................54









A-7 Truck Shipments Leaving Santa Isabel, Puerto Rico ...........................................55

A-8 Trucks Arriving at the Quincy FL Distribution Center .........................................55















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

SIMULATION MODELING AND BENEFIT-COST SENSITIVITY ANALYSIS FOR
TECHNOLOGY ADOPTION ON PUERTO RICO-UNITED STATES TOMATO
SUPPLY CHAIN

By

Jiaoju Ge

August, 2006

Chair: Allen F. Wysocki
Cochair: Bruce A. Welt
Major Department: Food and Resource Economics

Supply chain coordination and shipping conditions have a substantial impact on

food quality of produce supply chains, especially when supply chains cross international

borders. After harvesting, shipping conditions are the primary determinants of food

quality for highly perishable commodities such as tomatoes. In addition, based on food

marketing research, food quality for highly perishable commodities is becoming one of

the most important issues in today's food markets. The objective of this project is to

develop a tool to help identify opportunities to improve food quality and their marketing

implications on the Puerto Rico-United States tomato supply chain after new technology

has been adopted.

With the help of a discrete-event simulation software (SIMUL8), a simulated

supply chain distribution model, including tomato harvesting, packing, shipping and

distributing, was developed to analyze overall supply chain performance, and Simetare









was used to estimate associated costs and benefits for each participant in this supply

chain (e.g., from tomato grower to wholesaler). Finally, benefit-cost ratios were

simulated to assess the adoption of temperature controlling technology on tomato supply

chain marketing implications.

The results show that tomato keeping quality can be improved from 2.7 days to 5.4

days on average when temperature range within marine containers is controlled better

from 11.20C (52.5F) to 31.40C (88.50F) down to froml3.30C (56.0F) to 15.60C (60.0F).

However, whether or not tomato grower/shippers should adopt the temperature

controlling technology to increase exports and higher margin possibilities depends not

only on the benefit-cost ratio, but also on the actual cost of technology.















CHAPTER 1
INTRODUCTION

Extending shelf life and quality are two keys for increasing tomato imports to the

United States. Food quality is directly impacted by the quality of transportation being

used, especially when food crosses international borders. Therefore, examining the

worthiness of investing in technology is becoming more important. In the Caribbean

Basin, fruit exports have declined from $553 to $272 million from 1991 to 2001 (FAO

Stat, 2003). The decline in exports is due in part, to food quality issues. "Food Industry

Review 2003" (Figure 1-1) highlights that 87% of American consumers choose their

primary supermarket based on high quality fruits and vegetables. Consumers demand

produce that has desirable appearance, taste and texture. To meet consumers'

expectations, high quality fruits and vegetables must be provided.



Clean Neat Store

High Quality Fruits & Vegetables

Low Prices

Use-before/sell-by date marked 82%

High Quality Meats 81%


Figure 1-1 Top Factors in Selecting Primary Supermarket

Tomatoes, along with potatoes and lettuce, are the most accepted and consumed

fresh produce in America (ERS, USDA 2004). The tomato industry estimates that fresh-

market tomato retail value may exceed $4 billion (ERS, USDA 2004). Tomatoes are

prized for their good nutrition, wide acceptance and high consumption. However, as









supply chains grow, improving quality is becoming more challenging for all tomato

supply chain participants. For international distribution systems, physical and

environmental conditions are the primary determinants of food quality for highly

perishable commodities. Among these determinants, temperature is extremely important

for maintaining high tomato quality, referred to in this research as tomato keeping

quality.

Tomato imports now account for about 39 percent of U.S. tomato consumption, an

increase from approximately 20 percent in the early 1990's (ERS, USDA 2004). This

trend shows potential for increases in tomato exports from Caribbean basin. Therefore,

Puerto Rico could be one of the beneficiaries. Tomato grower/shippers in Puerto Rico

might have the opportunity to increase their tomato shipments to the United States if they

can, continuously and consistently, increase their tomato quality to meet American

consumers' needs. One way to improve tomato quality is to control temperature

conditions after harvest.

1.1 Overview of Tomato Supply Chain

Tomatoes are harvested at the mature green stage in Puerto Rico for shipment to

the United States. Harvesting occurs from January until the end of March for tomatoes

that are exported to the United States and the remaining harvests are consumed on the

island.

During the peak of the harvesting season there may be 500 to 600 workers in the

tomato fields every day. By the end of the season there may be as few as 300 workers. In

each harvesting team, there is a crew leader, pickers, a truck driver and a bucket dumper.

The driver is in charge of punching the cards of the pickers every time they bring a

bucket. Typically there are 5 to 7 supervisors who are responsible for managing the









pickers and taking care of mulching, spaying, fertilization and pesticide applications.

Most of the employees that work in the field have been with their respective tomato

companies for several years.

The pickers are mostly Puerto Rican women and are paid per hour or per bucket

depending on what they prefer. They are paid $5 per hour and they work 8 hours/day

excluding weekends ($40/day). They can make more money than $40/day if they harvest

a number of buckets that is greater than the average number of buckets that results in

$40/day. The price of each bucket may vary between 50 cents and $1 depending on

seasonality.

Normal daily production averages approximately 30,000-25 pound boxes of

tomatoes. The price per box may vary from $13 to $15 during March and from $37 to

$40 when Florida production is low in January and February.

All boxes can be traced using a number and a series of information including field,

variety, picker, and number of boxes picked in that field.

When tomatoes arrive at the packinghouse from the field, they are sorted by color

and defects, and graded by size into 4 different grades: Beefsteak Grade 1; Flavor ripe

Grade 1; Homegrown Grade 2; Garden Grade 3. All tomatoes showing a little pink color

are separated from the green and sold to local markets. Only green tomatoes are exported

to the USA. Tomatoes are then washed and waxed, and then packed into 25 lb boxes. At

the end of the packing line, tracking codes are stamped on the boxes. There are about 150

to 160 workers in the packinghouse, who work a 4 to 6 hr/day, depending on the tomato

volume and grades to be packed.









After packing, tomato boxes are immediately palletized and placed into ethylene

treatment rooms. The medium-large size tomatoes stay three nights inside the ethylene

room and the extra large size tomatoes stay four nights. Temperature of the ethylene

rooms are approximately 68.0F (20.0C).

Tomatoes to be exported are placed inside refrigerated marine containers (e.g.,

Horizon Lines) after the 3-day period in the ethylene treatment. Containers are shipped

on the same day from Santa Isabel (south part of the island) to the port of San Juan (north

part of the island). Depending on shipping schedules, containers are then loaded to ships

and transported to the port in Jacksonville, Florida.

During the peak of the tomato season, an average of 50 to 60 marine containers are

shipped every week to the United States. By the end of the season, an average of 15 to 25

containers are shipped each week. Each marine container holds 18 pallets of tomatoes

and each pallet holds eighty 25 pounds tomato boxes. Average temperature inside

refrigerated container is about 13.3C (56.0F).

Upon arrival to Jacksonville, the containers are unloaded and then taken by truck to

a distribution center in Quincy, FL. Upon arrival, containers are immediately unloaded

and placed in a large cold room with a temperature setting of 58.0F (14.40C). Each

container has a specific place designated by a number in the cold room.

Tomatoes are then stored for a short period of time (no longer than 1 day) before

leaving the warehouse to be re-distributed to diverse wholesalers around the U.S.

(Florida, Tennessee, California) and Canada (Toronto).

Figure 1-2 illustrates the steps involved in the tomato supply chain used in this

research from field to wholesalers.










Field


Hot water + chlorine treatment


Sorting


Washing


Waxing


Packing


Ethylene chambers


Marine containers


Port of San Juan, PR


Port of Jacksonville, FL


Quincy, FL


Containers unloaded


Distributed in the USA and Canada

Figure 1-2 Tomato Supply Chain Distribution Process

1.2 Problem Statement

Any technology that could be used to increase tomato quality would be welcomed

by Puerto Rico-United States tomato supply chain participants. Tomato supply chain

participants are always looking for ways to improve their profitability.

This study will identify the effect of temperature-controlling technology used by

Puerto Rican tomato grower/shippers on tomato quality for all tomato supply chain









participants and what the possibility is for tomato grower/shippers to adopt this

technology from financial point of view.

1.3 Objectives

The general objectives of this study are to develop a tool to help identify potential

tomato quality improvements as a result of the adoption of new temperature-controlled

technology, and to discuss their marketing implications for the Puerto Rico-United States

tomato supply chain. The specific objectives are:

* To identify an established relationship between tomato keeping quality,
temperature and time in the distribution system;

* To build a tomato supply chain simulation model to analyze tomato keeping quality
changes as they relate to the changes in temperature and time in the tomato supply
chain;

* To conduct a sensitivity analysis of benefit-cost ratio for tomato supply chain
participants.

1.4 Testable Hypotheses

Effects of temperature and time on fruit keeping quality are well documented in the

literature. This information is applied in this study. While the impact of temperature on

tomato keeping quality is well documented, the quantity impact of increasing keeping

quality is not as clear. But it's clear that tomato's shrinkage might be reduced by an

increased keeping quality. An assumption is made in this study that higher quality usually

results in more tomato exported to the United States market. Therefore, the hypotheses of

the study are the following:

Hi: Simulated software can be used to analyze the complexities of Puerto Rico-United
States tomato supply chain distribution system;

H2: The adoption of temperature-controlling technology will prolong tomato keeping
quality;









H3: The adoption of temperature-controlling technology will result in increasing
probabilities of the benefit-cost ratio being greater than 1.

1.5 Research Scope

Discrete-event simulation software, such as Arena and Simul8 are commonly

used to analyze complex systems with stochastic model. Simul8 was used in this study

to build a tomato supply chain distribution model. Results from previous fruit keeping

quality studies, which are highlighted in the next chapter, are used in this tomato supply

chain model, to analyze the effect of temperature controlling technology on Puerto Rico-

United States tomato supply chain participants.

Even through benefit-cost analysis has not proven to be particularly useful in

capturing the intricacies of complex scientific problems (Whittington and Grubb, 1984),

it is widely accepted in the area of policy effect and project determination. Therefore,

benefit-cost analysis is utilized to analyze the possibility of adopting the selected

technologies for tomato supply chain participants. The benefit is derived from the overall

revenues generated by supply chain participants. The costs of adopting temperature

controlling technologies are obtained from financial cost data.

Since the project is still on-going, the actual benefit and cost data will be collected

in late 2006 or early 2007. Estimated benefits and costs are used in this study to conduct a

sensitivity analysis.















CHAPTER 2
LITERATURE REVIEW

While there is a vast body of research related to technology adoption, much less

research has been done studying the economic impact of technology adoption. This study

will concentrate on previous studies in three different categories. The first category

reviews literature related to how physical conditions affect tomato quality. The second

category lists previous studies which used discrete event simulation models to solve

problems in supply chain distribution systems. The last category is comprised of research

focused on economic implication of new technology adoption and price asymmetry.

2.1 Physical Conditions and Tomato Quality

The reaction rate of a decrease in quality (of all fresh fruit) as it relates to

temperature, is identified in Arrhenius'law:

E 1 1
( )*( )
K Kefexp R Tf Ta (2.1)

where K is reaction rate of the decrease in quality; T is the temperature; Ea and R are both

Constant (Chang, 1981). This study showed that temperature has an effect on reaction

rate (the speed at which keeping quality changes based on differences in temperature and

time) for all different kinds of perishable fruits and vegetables.

In addition, one study (Robertson, 1993) explained the quality of most foods and

beverages decreases with storage or time. The relationship between shelf life and time

can be expressed by the zero order equation. This research will use a common form of the

pseudo first order equation, which is shown in (2.2)









Q =Q EXP(K*T) (2.2)

where Q is the remaining shelf life at the end of the time period, Qo is the shelf life at the

beginning of the time period, K is the reaction rate and T is the time period.

Fruit keeping quality and shelf life are often highly correlated concepts. Tij skens's

study provided the definition of keeping quality (Tijskens et al., 1994). According to this

study, keeping quality can be defined as the minimal quality necessary for a consumer to

accept the product.

Tijskens and Polderdijk (1996) extended the research on fruit keeping quality with

the application of Chang's theory to the study of keeping quality of vegetable produce

during storage and distribution. Their results showed that a change of keeping quality

(shelf life) is negatively impacted depending on temperature. This relationship between

shelf life, keeping quality, and temperature is illustrated in the following reaction rate

equation:

= -K(1-Q) (2.3)
dt Qmf

dQ
where is the change of the quality, K is the reaction rate, Q is the keeping quality
dt

and Qinf is the quality maximally possible at infinite time (Tijskens and Polderdijk, 1996).

One useful result from the 1996 Tijskens and Polderdijk's study, for this research,

is the use of a discrete generic model to predict tomato keeping quality and temperature,

shown in figure 2.1. It indicated that when temperature is around 15.00C, tomato will

have longest keeping quality, at around 15 days.









Keeping Quality (days)
15








0 5 10 15 20 25
Temp ("C)

Figure 2-1 Tomato (x) Keeping Quality and Temperature Relationship

In summary of this category of literature, for tomato, keeping quality can be written

as a function of both reaction rate K and time period t. Furthermore, it is understood that

the optimal shelf life for tomato is a result of maintaining a temperature range of 12.00C

to 18.00C.

2.2 Simulation System Modeling

One particularly relevant study (Schepers et al., 2004) showed an example of

building a dynamic simulation model for analyzing the complexity of the mango supply

chain distribution system. Mangoes and tomatoes share many of the same physiological

and distribution characteristics, and both are highly perishable fruits.

Three distinct research fields (consumer science, quality management and chain

science) and the links among them are analyzed by this study. The relationship between

these research fields are illustrated in Figure 2-2.

This approach is useful for studying and understanding a wide range mango supply

chain management issues. However, this analysis is relatively less detailed for each of the

three fields. Our current tomato supply chain study will be only concentrated on two of

the three fields, but in more detail.












2. Quality management & Consumer science
*Post-harvest product handling & logistics (ripening)
S*Biological variation in physico-chemical product properties
Realise -Sensory perception and evaluation (Liking)
quality

Cost of product Product
loss & handling Liking

Demand
3. Chain science: 1. Consumer science:
Collaborative marketing *Adoption dynamics
Cost sharing First and repeat usage
Share *Pricing Loss of interest Promote
cost dynamics first usage


Profits per
player


Figure 2-2 The Relationship Between Three Distinct Research Fields In a Dynamic
Simulation Model

In summary, the literature related to supply chain modeling analysis is less

extensive. However, the Schepers et al. study provides a good example of using modeling

software to analyze supply chain complexities. Furthermore, our study could be

considered an extension of this earlier research, especially the economic analysis section

2.3 Economic Analysis of Technology Adoption and Price Asymmetry

Deming (1967, 1972), Juran and Gryna (1970) and Juran (1974) emphasized that

most quality problems are "management" or "systems" problems requiring "statistics" in

their solution.

Following this study, Marquardt (1984) pointed out that product quality is high

visibility in today's economic environment. The article explores the respective roles of

business philosophy, management systems and technology systems. It also pointed out









that "the quality technology systems require new directions and new emphases in

statistics, software engineering, and other disciplines in order to be cost-effective."

But how quality is affecting the food industry? Caswell et al. (1998) explained how

quality management meta-systems affect the food industry. They indicated that "food

quality meta-systems are adopted by system participants for a variety of

reasons...... depending on the internal benefits and costs of adoption......" This study

utilized benefit-cost analysis to identify the adoption of a meta-system on one company.

The internal benefits pointed out by this study are the revenues from sale of products, and

costs are captured through production, transaction and regulatory compliance. The result

of this study showed that 71% of small firms expected adoption to increase their market

share, while only 48% of large firms had the same expectation.

For the current study, price asymmetry in retail, wholesale and shipping point price

(grower's price) is also a key component that needs to be considered. Several studies

have examined vertical price transmission in fresh produce.

All of the studies indicated that short-run supplier price changes precede retail price

changes (Ward, 1982; Powers, 1995; Heien, 1980). Furthermore, a statistical test finds

that tomato retail prices respond more to free-on-board (f.o.b.) price increase than to

f.o.b. price decrease (ERS, 1999). In the short-run, prices are affected more by supply.

An increase (decrease) in supply causes f.o.b. prices to fall (rise) which, in turn, leads to

decreases (increases) in retail prices (ERS, 1999). This study also indicated that on

average, the shipping point price for fresh field grown tomato is about one-fourth of the

retail value (ERS, 2004).






13


In conclusion, technology is one of the most important factors for determining food

quality. Technology can be adopted by a system or a private company to improve food

quality and the effect can be evaluated by benefit-cost analysis. Moreover, for the whole

tomato supply chain distribution system, if the shipping point price and its trend could be

known, the wholesale and retail price and their trends should possibly be calculated

through the vertical market markups.














CHAPTER 3
METHODS

The objectives of this study are to develop a tool to help identify potential tomato

quality improvements that lead to less shrinkage and more exports, to discuss their

marketing implications for the Puerto Rico-United States tomato supply chain with the

consideration about new temperature controlling technology being adopted, and to

conduct a sensitivity analysis of the benefits and costs of technology adoption. Several

different, but connected, methods are used in this study.

3.1 To Find Tomato Keeping Quality Equation

To find the impact on tomato keeping quality for temperature, the reaction rate

(equation 3.1) of how keeping quality responds to the change of temperature is calculated

by the following function, which is a transformation of equation (2.1):


Ln(K) = /, + /,( ) (3.1)
abs

where K represents the reaction rate and Tabs is the absolute temperature.

The keeping quality equation (3.2) can be derived based on the above calculated

reaction rate K and equation (2.2):

Q, Q, EXP( T,*T) (3.2)

where Q represents tomato keeping quality, t and t-1 represent time periods, K represents

the reaction rate and T represents the time between time period t and t-1.









This equation shows the relationship between temperature and tomato keeping

quality, which is the most important input for the tomato supply chain simulation model

in this study.

3.2 To Build Tomato Supply Chain Simulation Model

A specific simulation model is built for tomato supply chain between Puerto Rico

and the U.S. by using SIMUL8 software, a widely used simulation software. The

dynamic nature of Simul8 allows the user to show the creation of processes and outputs

as moving objects, while other software programs are static in nature.

This tomato supply chain model features the tomato keeping quality equation as

well as all other stages showed in Figure 3.1.

Tomato Grower:

Puerto Rico *........ Packing House for Packing
II Ethylene Room for Storage
II Shipping by Truck and Sea Pr......- Shippin oi
PII rice
II
II
U Wholesaler:
U.S.A. *.--
S Distribution center for storage
II Truck for shipping -..* Wholesale price
II

Retailers:
U.S.A. & Canada ...............
Stores for retailing -- Retail price

Figure 3-1 Puerto Rico-United States Tomato Supply Chain Components

In conclusion, the tomato supply chain simulation model includes all the above

components. The key output of the model for this study is tomato keeping quality as it is

affected by temperature changes in the tomato supply chain.









3.3 To Conduct A Benefit-Cost Sensitivity Analysis

Simetarc (Simulation & Econometrics to Analyze Risk) software will be used to

conduct benefit-cost sensitivity analysis for the adoption of temperature controlling

technology on tomato supply chain participants. This section will consider the sensitivity

for both increased export quantity and reduced shrinkage generated by the increased

tomato keeping quality. The steps in this sensitivity analysis are:

* Modeling key output variables by using financial data with different distributions.

Total revenue = f(export price, quantity exported)

Total cost =Production Cost + Transportation Cost

Benefit-cost ratio = B / C ratio = total revenue / total cost

* Benefit-cost ratio analysis for tomato grower with four possible increased export
scenarios.

* Sensitivity cost analysis for highest amount money that tomato growers would pay
for a given technology regarding to four increased export scenarios.

* Benefit analysis for other supply chain participants (wholesalers) for three possible
industry mark-ups between each level of participation regarding to reduced
shrinkage.

The results of benefit-cost analysis are used to provide recommendations about

whether or not tomato grower/shippers in Puerto Rico should adopt this temperature

controlling technology.














CHAPTER 4
DATA COLLECTION

The data needed for this study is diverse as a result of the methods mentioned in the

previous chapter. Most of the required data on operational and financial data are regarded

as non-public information by most companies. Therefore, the data collected in this study

came from four major sources. One source was collected from one typical tomato

operating firm, Gargiulo, at both field and packing house in Santa Isabel, Puerto Rico, as

well as distribution center in Quincy, Florida in March of 2005 and 2006. The second

source was collected from the tomato export port at San Juan, Puerto Rico and the tomato

import port at Jacksonville, Florida in March of 2005 and 2006. The third source of data

was collected through telephone interviews in Florida during October of 2005 and March

of 2006. The remaining data came from secondary on-line data sources.

Five types of data were collected from both Puerto Rico and the United States: 1)

production related data, such as daily production; 2) processing time related data, such as

packing time and truck moving time; 3) labor related data, such as how many workers

and their salaries; 4) temperature and shelf life related dat; 5) financial related data, such

as costs and prices.

Data from first three categories are used to build the Puerto Rico-United States

tomato supply chain model. Category 4 data is used to build the tomato keeping quality

equation which is an integral part of the simulated supply chain model. Finally, category

5 data and the results from the model will be used to conduct benefit-cost sensitivity

analysis.










4.1 Production Related Data Analysis

As an integrated tomato company, Gargiulo has tomato fields in Puerto Rico and

also a distribution center in Quincy, Florida. Around 750,000 pounds of tomato can be

picked within one day and dumped into the packing house. Among all of these tomatoes,

on average 25% never make the grade and are disposed of. The remaining 75% goes to

both the Puerto Rican local market and the U.S. export market.

For all tomatoes exported to the U.S., there are three different sizes and four

different grades. Size one represents the extra large tomatoes, which constitute around

25% of all export tomatoes (Figure 4-1). Size two represents the large tomatoes, at about

60%. Finally, size three represents the medium tomatoes that are 15% of the total crop on

average.




60

50

40

30

20

10

0-
Size 1 (5*6) Size 2 (6*6) Size 3 (6*7)


Figure 4-1 Export Tomato Size Distribution

Tomato exports to the U.S. include grades 1, 2, 3 and 4 (figure 4-2). Grade 3

tomatoes make of the majority of tomatoes (about 60%), while Grade 1 and 2 each have

approximately 12.5%. The 4th Grade share approximately 15% of the tomatoes exported.












15 12.5

12.5







60



U Grade 1 U Grade 2 O Grade 3 O Grade 4


Figure 4-2 Export Tomatoes Grade Distribution

4.2 Processing Time Related Data Analysis

The processing time is the time period from the start to the end of one task. In this

research, the most important time period is the processing time for packing one box of

tomatoes, palletizing one pallet, shipping, loading and unloading trucks and boats.

Another important concept for the processing time related data is the distribution of

the time. Not all the processing times are fixed in actual practice. Instead, most

processing times vary over a predictable probability distribution. Considering this range

in model building will increase the model's prediction ability.

For example, in table (Table 4-1), the shipping time from Santa Isabel to San Juan

is 120 minutes (2 hours). This means that a truck driven from the packing house at Santa

Isabel to the export port at San Juan takes around 2 hours. The corresponding distribution

for this shipping time is considered to be a bounded normal distribution, which means

that it takes half an hour to 4 hours to ship depending on other conditions such as traffic

and weather. Other processing times can be explained in the same way.









Table 4-1 Processing Time Data
Tk Processing Time Distribution
(Minutes) (Minutes)
Harvesting 1.4 Exponential

Packing (Per Box) 0.016 Exponential

Palletizing (Per Pallet) 2-5 Uniform (2,5)
Ethylene Room Storage 5040 Fixed
(Size 1)
Ethylene Room Storage 3600 Fixed
.3600 Fixed
(Size 2&3)
Loading & Unloading Truck 5 Average
(Per marine Container) (5, 1.25)
Shipping (Santa Isabel to
Shipping (Santa Isabel to 120 Bounded Normal (120,10,90,240)
San Juan)
Shipping (San Juan to 2160 Bounded Normal
Jacksonville) (2160,240,2880,1440)
Shipping (Jacksonville to 180 Bounded Average
Quincy) (180, 240,120)
Loading & Unloading Ship 15 Average
(Per marine Container) (15, 3.75)
Cold Room Storage
(Quincy) 1440 Average (10, 2.5)
(Quincy)


4.3 Labor Related Data Analysis

For many industries, labor is a fundamental and required resource. Regarding the

tomato supply chain between Puerto Rico and the United States, especially among tomato

grower/shippers, the labor data is an integral component of the tomato supply chain

simulation model.

Labor data includes two types of data. One is the number of laborers for each task,

such as picking and packing. The other is the working hours which includes what are the

shifts and how long each shift is.

For this study, field pickers of tomatoes averaged around 500 to 600 people daily

(Table 4.2). They worked from 6am until 2pm everyday without a shift change. For









packing house workers, including sorters, graders, packers and people who palletize, the

total number ranges from 150 to 160 people. These laborers work from 10am to 2pm or

4pm depending on the amount of tomatoes in the packing house. The reason that packing

house workers start later than field pickers is they have to wait until the tomatoes are

dumped into the packing line after being picked in the field.

Table 4-2 Labor Input Data

Labor Type Numbers Working Hours Shifts

Field Pickers 500-600 8 (6am-2pm) Day shift (6am-2pm)
Packing House
g H e 150-160 4-6 (10am-2pm or 4pm) Day shift
Workers


4.4 Temperature and Shelf Life Related Data Analysis

The basic and central idea of the tomato supply chain simulation model is to find

the ways to increase tomato keeping quality, which should generate more total revenue

for the whole industry. Therefore, obtaining the temperature and shelf life data and better

understanding the temperature changes among different stages are extremely important.

Shelf life here is synonymous with keeping quality.

The data in Table 4-3 are collected from the tomato supply chain between Puerto

Rico and the United States, with the exception of shelf life, which is obtained by a

secondary source. The T in the table 4-3 represents temperature. For instance, in ethylene

room 1, when tomatoes stay inside for 3600 minutes and under the temperature of 68.0F,

the shelf life is approximately 8.0 days. Most importantly, from the above data, it shows

that the longest shelf life is about 14.8 days and the shortest shelf life is about 5.5 days

due to different temperature conditions, treating humidity and other physical condition as

constants. Therefore, tomato keeping quality can be increased to as long as 14.8 days.









Therefore, temperature distribution outside of the ethylene room, and particularly,

inside the marine container is considered as a critical condition for maintaining tomato

keeping quality. These distribution data is showed in Table 4-4. It indicates that within a

marine container, the temperature varies considerably from 11.2C (52.5F) to 31.4C

(88.5F), following the uniform distribution with the min of 11.2C (52.5F) and the max

of 31.4C (88.5F). (Source: experiments of Agricultural Biological Engineering

Department at the University of Florida). From the secondary data about the shelf life,

when the temperature is 11.2C (52.5F), the shelf life would be 7.5 days. However, when

temperature increases to 31.4C (88.5F), the shelf life of tomato would be as low as only

2 days which is critical for tomato keeping quality (Figure 2-1).

Table 4-3 Tomato Shelf Life and Temperature
Storage
Storage Shelf Life
Items Time T (F) T (C) T (OK) e
(Minutes) (Days)
Packing
120 76.5 24.7 297.9 5.5

Ethylene Room
5040 68.0 20.0 293.2 8.0

Ethylene Room2
3600 68.0 20.0 293.2 8.0

Marine Container
(Truck) 120 56.0 13.3 286.5 14.0

Marine Container
(Ship) 2160 56.0 13.3 286.5 14.0
Marine Container
(Truck) 180 56.0 13.3 286.5 14.0

Cold Room
(Distribution Center) 720 58.0 14.4 287.6 14.8

However, since only green tomatoes are exported to the United States, they need to

stay inside the ethylene room for enough time to begin the ripening process. If one









assumes that ethylene room best practices are followed, the shelf life within ethylene

room in the above table will not be considered as a critical condition for overall tomato

quality.

Table 4-4 Temperature Distribution of Marine Container (C)
Marine Container Min Temperature Max Temperature Distribution
Bottom 11.2 31.4 Uniform
Center 13.7 25.7 Uniform
Up 14.9 24.6 Uniform
Over all (Min & Max) 11.2 31.4 Uniform


4.5 Financial Related Data Analysis

Financial data are necessary to derive benefit-cost ratio sensitivity analysis. Two

types of data were collected including costs related data (Table 4-5) and revenue related

data (Table 4-6).

Cost data are collected both from an on-line source and from a Puerto Rican tomato

grower. It represents production cost of one typical tomato grower in Florida. This

research compared production cost data for a particular Puerto Rican tomato grower

through interviews. Based on these interviews, total operating cost for this tomato grower

in Puerto Rico was $4,491.14, which was lower than the average Florida grower's

operating cost of $4,984.15 (Table 4-5). The fixed cost was $1,623.06, lower than the

$1,658.01 average in Florida. However, the harvesting and marketing cost were similar.

Therefore, the total production cost for Puerto Rico's tomato grower used in the model

was $11,858.20(Table4-5) per acre, which is lower than the total production cost on

average for Florida ($12,386.16).

It is easy to see that tomato growers in Puerto Rico have an advantage compared to

growers in Florida based on production cost. However, this advantage is not that large









Table 4-5 Tomato Operation and Production Cost (Average Per Acre)
Category Florida Gargiulo
OPERATING COSTS Dollars
Transplants 480.00 456.00
Fertilizer 358.50 376.43
Fumigant 550.00 577.50
Fungicide 241.00 253.05
Herbicide 20.77 21.81
Insecticide 459.93 482.93
General Farm Labor 140.63 140.63
Machinery Variable Costs 858.59 858.59
Tractor Driver Labor 193.55 203.23
MISCELLANEOUS
String and Stake Disposal 123.42 129.59
Scouting 45.00 42.75
Pruning 79.86 83.85
Plastic Mulch Disposal 163.35 171.52
Level Land 145.00 0.00
Farm Vehicles 27.33 0.00
Drive Stakes 81.31 85.38
Cut Cross Ditches 27.20 0.00
Tie Plants 145.20 152.46
Drip Tube 145.20 0.00
Plastic Mulch 315.00 330.75
Tomato Stakes 90.00 94.50
Plastic String 28.75 30.19
Interest on Operating Capital 264.56 0.00
Total Operating Cost 4,984.15 4,491.14
FIXED COSTS
Land Rent 500.00 475.00
Machinery Fixed Costs 208.15 218.56
Farm Management 678.47 644.55
Overhead 271.39 284.96
Total Fixed Cost 1,658.01 1,623.06
HARVEST AND MARKETING COSTS
Tomato Cartons 1,200.00 1,200.00
Sell 240.00 240.00
Pick and Haul 1,360.00 1,360.00
Pack 2,800.00 2,800.00
Various Association Fees 144.00 144.00
Total Harvest and Marketing Cost 5,744.00 5,744.00
TOTAL COST 12,386.16 11,858.20
Total Fixed Cost 1,658.01 1,623.06
TOTAL Variable Cost 10,728.15 10,235.14









since the production cost difference is only 5%. Tomato growers in Puerto Rico have to

pay additional shipping costs averaging about $0.1 per pound.

Revenue data were collected from two sources. One is the data collected from the

Gargiulo firm. During the export season in Puerto Rico, the export price (the f.o.b.

shipping point price) varied from $37 to $40 per 25 pounds during January and February

(148 to 160 cents per pound) and from $13 to $15 per 25 pounds during March (52 to 60

cents per pound).

However, from the data collected from USDA on Table 4-6, the f.o.b. shipping

point prices during January and February ranged from 18.4 to 116.0 cents per pound with

an average of 39.2 cents per pound for the year 1990 to 2004. For March, prices ranged

from 21.2 to 81.7 cents per pound with the average of 44.3 cents per pound during the

same period. This indicates that tomato growers in Puerto Rico have a great advantage

for revenue on the shipping point price since the average price is much lower than Puerto

Rican growers' prices during the Puerto Rican shipping season.

On the other hand, the Table 4-6 also indicates that along with the change of f.o.b.

shipping point price, the retail price has changed in the same direction for the past 15

years during the January to March season without considering the abnormal price shocks

brought about by weather, etc. This also suggests that the retailers should have the same

price trend as the tomato growers. If tomato grower/shippers' export price doesn't

change, the retail price is likely to remain unchanged as well.

In summary, all category data were needed for the next chapter's tomato supply

chain simulation model and benefit-cost ratio sensitivity analysis. The first 4 category










data would be used for simulation model. The result of simulation model and the last

category data would be used for benefit-cost sensitivity analysis.

Table 4-6 U.S. Average Monthly Retail Price and f.o.b. Shipping Point


Jan.


Feb.


Mar.


Retail price (cents per pound):


173.5
91.2
93.6
114.1
160.4
132.3
110.3
121.3
145.2
190.4
144.3
141.4
145.1
171.1
147.2


236.1
84.0
143.0
109.8
111.2
130.0
108.4
131.4
135.6
147.6
128.6
131.3
129.8
156.5
151.0


176.5
94.8
172.9
88.0
91.4
108.1
146.7
165.4
151.5
139.5
136.4
133.6
129.2
161.9
152.9


Feb.


Mar.


F.o.b. shipping-point price (cents per


pound:
116.0
23.1
40.5
38.3
41.5
41.1
18.4
32.1
26.4
33.5
21.4
43.8
38.2
50.9
34.5


97.6
31.6
76.0
21.9
19.3
29.8
40.0
45.9
44.0
23.4
21.1
29.1
28.0
31.7
36.3


32.3
44.0
80.7
21.2
24.5
37.1
81.7
57.4
34.0
22.3
33.0
56.4
41.7
55.6
42.2


Year

1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004


I I














CHAPTER 5
SIMULATION MODELING SPECIFICATION AND RESULTS

Based on the previously described methods and the data collected, a specific

tomato supply chain simulation model between Puerto Rico and the United States was

built and the details building this model and the specifications used are explained in the

following paragraphs. The tomato keeping quality equation is discussed first, since it is

an important input for the tomato supply chain simulation model.

5.1 Keeping Quality Equation and Results

Based on the temperature and shelf life related data discussed in chapter 4, the

reaction rate, which explains the keeping quality changes according to the change of the

temperature, is calculated by the following equation:


Ln(K) = /, + ( ) (5.1)
abs

where K represents the reaction rate and Tabs is the absolute temperature.

A linear regression is used to get the result and presented in table 5-1.

Table 5-1 Statistics of Reaction Rate
Intercept 24.22 F Ratio 250.26

Slope -7277.48 Prob. (F) 0.000018

R-Square 0.98 T-Test -15.82

S.E. 460.03 Prob. (T) 0.0004



From the probability of the F ratio and T-test, the statistics for intercept and slope

are both statistically significant. Therefore, the reaction rate can be calculated as









1
Ln(K) = 24.22 7277.48( ) by using the above statistics as /0 = 24.22 and
Tbs

A = -7277.48.

The predicted K is showed in table 5-2. Within a marine container, the predicted

reaction rate averages 0.31 when the temperature is at the average of 56.00F (13.30C).

However, if the range of temperature in the marine container is 52.50F (11.20C) to 88.50F

(31.40C), the predicted reaction rate will be ranged from 0.25 to 1.38 (Table 5-2), where

the range (1.13) is 3.7 times of the mean. It indicates that within a marine container, the

wide change of temperature will have a huge impact on tomato keeping quality.

Table 5-2 Predicted Results of Reaction Rate
Items Temperature Absolute Predicted Reaction
(O___F) Temperature (OK) Rate (K)
Packing 76.5 297.9 0.811612748

Ethylene Rooml 68.0 293.2 0.547656726

Ethylene Room2 68.0 293.2 0.547656726

Marine Container (Min) 52.5 284.4 0.254924292

Marine Container (Mean) 56.0 286.5 0.307248107

Marine Container (Max) 88.5 304.5 1.38375325
Cold Room (Distribution 58.0 287.6 0.338899278
Center)


After reaction rate is predicted, all the results for K will be used substitute into the

following keeping quality equation:

Q, = Q, EXP(K *T) (5.2)

where Q represents tomato keeping quality in days, t and t-1 represents time period in

days, K represents the reaction rate and T represents the time between time period t and









t-1. This keeping quality is used to plug into the following simulation model as a key input

to get the result of the final keeping quality.

5.2 Tomato Supply Chain Simulation Model and Results

The tomato supply chain simulation model between Puerto Rico and the United

States was built by using simulation software, SIMUL8. SIMUL8 Corporation

develops markets and supports business simulation software for enterprise-wide use in

business, government, education, and any organization that handles flows of orders,

people, transactions or products (SIMUL8 homepage, 2006). One of the most popular

uses of SIMUL8 is for supply chain management (SIMUL8 homepage, 2006). There

are some successful cases studied by the company which proved the potential use of

SIMUL8 in this study.

5.2.1 Simulation Model before Technology and Results

The model is very visual (Figure 5-1) and easy to see all the components of tomato

supply chain between Puerto Rico and the United States. It includes tomato

grower/shippers in Puerto Rico (light blue areas in Figure 5-1), tomato distributors (dark

blue area in Figure 5-1) and tomato wholesalers and retailers in the United States and

Canada (brown area in Figure 5-1). Graphs in the model represent all different facilities

within tomato supply chain. For instance, harvesting represents being harvested and

dumped into packing house. Packing represents tomatoes being packed into a 251b-box

and so on. Parameters (Appendix B) and computer programs (Appendix C) of each graph

or facility are plugged into the model to allow the model to simulate the real world

tomato supply chain distribution system.

The data collection period for the model (Figure 5-1) is 3 months, which is exactly

the period from January to March that tomatoes in Puerto Rico would be exported into











Tomatb Prower
S DisposalPR
42



Harvesting Soing iueue for Packg
206 42 161







Dumping In


Packing 1 eueLocal Palletizing Paltizing Local To Local
3% 0% 0


Packing Pailetizng US
SEthylene Room 1

Packing 3 '-Que'e Us Palhtiz ng

3>% A% 3' e / 1n
3%- r '"" -luck Loading PR
sIore I ^ To Loading 0%

Packlng Pa,

,i Packinq Shipping

Santa Isabel

San Juan'


Ship Unloading Ship Loading
0% 0%
ruck Loading US Truck Unloading PR
0%0%




Jacksonville


Distributor Quincy



Truck Unloadi g US Cold Room
S 0% a Dispatch to End Consunr'J

C. ...


Wholesalers & Retailers
To Canada
0


US Food Services
0
-N

US Groceries



US WholeSalers



All US & Canada locations


Labor Resource

Fied Pickers
100%

6


SuperiBor
0%

a
8


Packing House Workers
100%
isO~


Figure 5-1 Supply Chain Simulation Model before and After Technology Applied

the United States. When model starts to run, both green and red tomatoes would be

picked in the field and then dumped into packing house through harvesting. Then









tomatoes would be sorted by four different grades and colors. Following, the red balls

represent red tomatoes sold in Puerto Rico local market and the green balls represent

green tomatoes being packed and exported to the United States by following all the

processes and data that were collected.

This tomato supply chain simulation model was built by varying the temperature

conditions and holding other conditions such as humidity constant. Therefore the "before-

technology applied model" (Figure 5-1) represents the supply chain as it is without any

temperature controlling technology being adopted by any supply chain player, especially

tomato growers.

Table 5-3 Summary Statistics of Simulation Model Before Technology
Keeping Quality (Days)
Grade
Mean SD CV Min Max Range

1 2.6 12.5 70.4 0.2 7.1 6.9

2 2.4 13.2 81.9 0.2 7.0 6.8

3 2.9 12.8 66.2 0.2 8.6 8.4

4 3.0 12.5 62.2 0.3 7.0 6.7

MEAN 2.7 12.7 70.2 0.2 7.4 7.2



The results of this before-technology model indicate that tomato keeping quality is

relatively low and varies widely. The result summary statistics are given for all four

different grades in Table 5-3. The mean of tomato keeping quality is 2.7 days, which is

low compared to the normal condition where tomatoes are kept for around 14 days. In

addition, the range of tomato keeping quality, which is from 6.8 to 8.4, is almost three

times of the mean (2.7). Moreover, the coefficient of variation for tomato keeping quality









is pretty high, on average at 70.2. These facts suggest that the variation of tomato keeping

quality need to be reduced.

In summary, the lower unacceptable mean with wide variation for tomato keeping

quality indicates problems exist in this tomato supply chain between Puerto Rico and the

United States. This is likely the result of wide temperature changes within the marine

containers which is the key factor for changing tomato keeping quality. Therefore, the

temperature controlling technology within marine container can be adopted by tomato

shippers to increase tomato keeping quality and generate more exports and higher

revenues.

5.2.2 Simulation Model after Technology and Results Comparison

Before temperature within marine container was controlled more closely, the

temperature range was from 11.20C (52.50F) to 31.40C (88.50F). This temperature range

results in a wide range of keeping quality and tomato shelf life. Therefore, if better

temperature control can be obtained by use of technology in the marine container, tomato

keeping quality should increase.

The literature review and data collected from this study indicate that a temperature

of 14.40C (58.0F) is the optimal temperature for ideal tomato keeping quality. It is

reasonable to assume that temperatures within marine container can be controlled

between 13.30C (56.00F) and 15.60C (60.0F). The results of the model after-technology

will be compared to the results of the model before-technology to see the impact of the

adoption of temperature controlling technology on tomato keeping quality.

The new model, "tomato supply chain simulation model after-technology" applied

is visually similar as the model "before-technology model" (Figure 5-1) since all the

components of this supply chain from Puerto Rico to the United States are the same.









However, the parameters for the temperature distribution within marine containers are

different (Appendix B) to reflect the change of temperature.

The result of tomato keeping quality after temperature controlling technology is

adopted within marine container by tomato grower/shippers are presented in table 5-4. By

comparing the coefficient of variance (CV) and the range to the result before-technology

adopted model, it is clear that the coefficient of variation decreases substantially from

70.2 (Table 5-3) to 11.56 on average (Table 5-4), which reduce the variation by 85%.

Moreover, the range is also decreased from 7.2 (Table 5-3) to 4.74 on average (Table 5-

4). These suggest that temperature controlling technology adoption is important for

reducing the variation of tomato keeping quality.

It is interesting to know what the impact of this temperature controlling technology

adopted by tomato shippers is by comparing the results between temperature controlling

technology adopted before and after.

Table 5-4 Summary Statistics of the Simulation Model after Technology
Keeping Quality (Days)
Grade
Mean SD CV Min Max Range

1 5.37 4.00 11.02 3.99 7.05 3.07

2 5.48 4.44 12.00 4.28 6.99 2.70

3 5.45 4.51 12.24 0.35 8.63 8.27

4 5.43 4.03 10.99 1.95 7.00 5.05

MEAN 5.43 4.24 11.56 2.64 7.42 4.77



In the simulation, before marine container temperature was controlled, the tomato

keeping quality at the end of supply chain averaged 2.7 days (Table 5-3) at temperature

of 14.80C (58.00F). However, after temperature was controlled, tomato keeping quality









increased to 5.4 days. Figure 5-2 shows that the probability of tomato keeping quality less

than 2.7 days is 55%, more than 5.4 days is only 12%, and between 2.7 days and 5.4 days

is 33%. This indicates that the chance for tomatoes to have lower keeping quality is 55%

(less than 2.7 days) and the chance to have higher keeping quality is 45% (greater than

2.7 days).


StopLight Chart for
Probabilities Less Than 2.700
and Greater Than 5.400

100%
90%
80%
70% 0.33
60%
50%
40%
30%
20%
10%
0%
Before Tech Keeping Quality (Days)


Figure 5-2 Probability of Tomato Keeping Quality before Technology Adoption

However, after using technology to control the temperature, the model shows

substantial improvement. The probabilities of tomato keeping quality after-technology

adoption are illustrated in Figure 5-3. Figure 5-3 indicates that the probability for

tomatoes to have a lower keeping quality less than 2.7 days is 0%, to have a higher

keeping quality higher than 5.4 days is 51%, and to have keeping quality between 2.7

days and 5.4 days is 49%. Compared to tomato keeping quality before-technology, after

temperature is being controlled more closely, the chance that tomato has less than 2.7










days decreases from 55% to 0%. However, the chance for tomatoes to have keeping

quality higher than 5.4 days increases from 12% to 51%. This suggests that if temperature

in marine containers could be controlled, tomato keeping quality could be improved by 2

times.


StopLight Chart for
Probabilities Less Than 2.700
and Greater Than 5.400

100%
90%
80%
70%
60%
50%
40%
30%
0.49
20%
10%
0%
After Tech Keeping Quality (days)


Figure 5-3 Probability of Tomato Keeping Quality After Technology Adoption

Furthermore, it is helpful to compare the overall average keeping quality for

tomatoes before and after temperature controlling technology is being adopted by tomato

grower/shippers.

On average, tomato keeping quality has been improved by over 2 times. This is

also true for all four individual grades (Figure 5-4). For example, before-technology for

grade 3, the average tomato keeping quality at retail level is 2.9 days. However, after

temperature is being controlled, tomato keeping quality on average is 5.5 days, which is

extended by 1.9 times more than before (Figure 5-4).












Grade 4


Grade 3 5,










Tomato Keeping Quality (Days)
Figure 5-4 Mean Comparison of Tomato Keeping Quality

In summary, for the tomato supply chain between Puerto Rico and the United








States, if marine container temperature can be controlled more closely from a range
Grade 2


Grade 1








between 110.0 1.0 2.0 3.0 4.0 5.0 (6.0) to 15.6C

(60),Tomato Keeping quality (Days)on average will be improved from 2.7 days to 5.4 days,
Figure 5-4 Mean Comparison of Tomato Keeping Quality

In summary, for the tomato supply chain between Puerto Rico and the United

States, if marine container temperature can be controlled more closely from a range

between 11.2C (52.5F) and 31.4C (88.5F) down to between 13.3C (56.0F)to 15.6C

(60.0F), tomato keeping quality on average will be improved from 2.7 days to 5.4 days,

which is 2 times longer than before. Furthermore, the probabilities for tomato to have less

than 2.7 days keeping quality drops to 0% and the probability of having greater than 5.4

shelf-life days increases from 12% to 51%. Additionally, the variation of tomato keeping

quality decreases sharply by 85% from the calculation of the average coefficient of

variation. All these figures imply that tomato supply chain participants' total revenue

should increase since all supply chain players and consumers prefer better quality product

in the United States.


I / / /














CHAPTER 6
BENEFIT-COST SENSITIVITY ANALYSIS AND RESULTS

Chapter 5 showed that proper application of temperature controlling technology

can lead to improved tomato keeping quality. However, participants in the Puerto Rico-

United States supply chain also want this question answered: "Is this technology

adoption worth it from a financial point of view?" The benefit-cost sensitivity analysis

results presented in this chapter are generated from the information and results from

previous chapters. Based on a thorough review of the literature and interviews with actual

supply chain participants, production and transportation costs represent the "costs" and

the total revenue represents the "benefits" respectively for this analysis. Financial risk

analysis software, Simetarc, will be used to build a financial risk analysis model to obtain

the results and graphs for the analysis.

6.1 Analysis of Tomato Growers' Benefits with Increased Exports

Benefit was calculated as total revenue, using export prices and export quantity

data collected in Puerto Rico. Cost was calculated as total cost, captured by both

production and transportation cost. Based on the results from tomato supply chain

simulation model in Chapter 5, one key assumption is made here. When the temperature

conditions within marine containers are controlled better, tomato growers will partially

regard this as a potential incentive to export more tomatoes to the United States, not only

the green ones but also some pink ones. Recall that tomato grower/shippers normally

only ship green tomatoes because of shelf life limitations. It is reasonable to assume that

if shelf life can be improved through temperature-controlling technology, tomato









grower/shippers may be willing to send a percentage of those turning-pink tomatoes that

are normally reserved for local Puerto Rican markets. Therefore, it's appropriate to

assume that improved tomato keeping quality will eventually result in more tomatoes

exported to the United States market.

The sensitivity of total revenue is analyzed for various possible increased export

scenarios, which are exports increased by 5%, 10%, 15% and 20%, as opposed to a zero

increase in exports before the adoption of temperature controlling technology. It should

be noted that these percentage increases apply only to the current level of tomatoes being

exported, not to the entire Puerto Rican tomato crop. For example, before technology

adoption, tomato exports to the Unites States are 52.5% of the total crop. An increase of

20% of the 52.5% results in 63% of the total crop now being exported, not 72.5%. These

total revenue scenarios were compared to the total cost to find out the possibilities of

increased benefit-cost ratio for tomato grower/shippers. The Simetarc model is presented

in Appendix D.

The results of benefit sensitivity analysis based on four different scenarios which

indicate varying levels of increased tomato exports (resulting in increased revenues) are

illustrated in Figures 6-1 and 6-2. Figure 6-1 shows the probability of benefits over costs

for no increases in tomato exports because there was no temperature controlling

technology adoption, increased tomato exports by 5% and 10% with technology

adoption. The Y axis represents the probabilities and the x axis represents total revenue

related to the scenarios of increased export volume. The $4,155,496.64 (Figure 6-1) is

100% of tomato grower/shippers' total cost and the number $4,571,046.30 (Figure 6-1)

represents 110% of tomato grower/shippers' total cost, when there is no tomato export









increasing before technology. The other four numbers in figure 6-1 represents 100% and

110% of tomato grower/shippers' total cost when tomato export increased by 5% and

10%, respectively. Therefore, the results indicate that without temperature controlling

technology adoption, benefits have a 15% chance less than costs (at normal export levels)

and an 85% chance between 100% and 110% of cost, which means benefit-cost ratio has

15% chance of being less than 1.This suggests that there is an 85% chance for tomato

grower/shippers to break even However, benefit has 0% chance of being higher than

110% cost-level, which means benefit-cost ratio has no chance to be higher than 1.1, also

indicating no chance at all to get higher than a 10% margin.


StopLight Chart for StopLight Chart for StopLight Chart for
Probabilities Less Probabilities Less Than Probabilities Less Than
Than 4,155,496.640 4,173,215.390 and 4,190,934.140 and
and Greater Than Greater Than Greater Than
4,571,046.300 4,590,536.930 4,610,027.550


100% 0 100% 0.100%

80% 80% -80%
60%- .85 60% 60% 0.76
60% 0.85
0.92
40% -40% 92 40%
20% -20% -20% -
0.24
0% 0% 0%
TR (Q=100%) TR (Q=105%) TR (Q=110%)


Figure 6-1 Probabilities of Benefit Analysis (1)

Finally, with temperature controlling technology adopted by tomato

grower/shippers, when tomato export increases by 5%, the benefit have 0% chance of

being less than cost, a 92% chance between 100% and 110% of cost, and 8% chance to

be higher than 110% of cost. It means that the benefit-cost ratio has a 92% chance to be









between 1.0 and 1.1, and an 8% of being higher than 1.1. This suggests some positive

margin exists up to 10%. However, the chance to get higher than a 10% margin is only

8%. When tomato export increases 10%, a benefit have 0% chance to be less than cost, a

24% chance between 100% and 110% of cost and a 76% chance of being higher than

110% of cost, which means that the benefit-cost ratio has 24% chance to be between 1.0

and 1.1, and a 76% chance to be higher than 1.1. These suggest that the margin has 24%

chance to be up to 10% and a 76% chance to be higher than 10%.

If tomato export increases by 15% and 20%, the chances for increased benefits

would be higher, resulting in more margins for tomato grower/shippers. Figure 6-2 shows

this situation. The number of $4,610,027.55 (Figure 6-2) is 110% of tomato

grower/shippers' total cost and $5,029,120.97 (Figure 6-2) represents 120% of tomato

grower/shippers' total cost, when tomato exports are at the 110% level. The red area in

Figure 6-2 represents the possibilities of the benefit-cost ratio being less than 1.1, the

yellow area represents the possibilities of the benefit-cost ratio being between 1.1 and 1.2

and the green area represents the possibilities of the benefit-cost ratio being higher

thanl.2 for all different increased tomato export scenarios.

If tomato exports increase by 10%, the benefit has 76% chance of being between

110% and 120% of cost, which means the benefit-cost ratio is between 1.1 and 1.2.

However, if tomato exports increase by 15%, the benefit has a 95% chance to be between

110% and 120% of cost, which means the benefit-cost ratio has increased to 95% from

76% for benefit-cost ratios between 1.1 and 1.2. It suggests that tomato grower/shippers

have a 95% chance to obtain margins between 10% and 20%. Moreover, if tomato

exports increase by 20%, the benefit will have 63% chance to be higher than 120% of









cost, which means the benefit-cost ratio has 63% chance to be higher than 1.2. It suggests

that tomato grower/shippers have a 63% chance to reach margins higher than 20%.


StopLight Chart for StopLight Chart for StopLight Chart for
Probabilities Less Than Probabilities Less Than Probabilities Less Than
4,610,027.550 and 4,629,518.180 and 4,649,008.800 and
Greater Than Greater Than Greater Than
5,029,120.970 5,050,383.470 5,071,645.970

100% 0. 100% 100%

80% -80% -80%
0.63
60% 0.76 60% -60%
0.95
40% -40% 40%

20% 20% 20% 0.37

0% 0% ---0%
TR(Q=110%) TR(Q=115%) TR(Q=120%)


Figure 6-2 Probabilities of Benefit Analysis (2)

In summary, Table 6-1 shows that if temperature controlling technology is adopted,

the probabilities for tomato grower to have benefit-cost ratio between 1.0 and 1.1 are

92% and 24% when tomato exports increase by 5% and 10% respectively. And the

probabilities for tomato grower/shippers to have a benefit-cost ratio between 1.1 and 1.2

are 8%, 76%, 95% and 37% when tomato exports increase by 5%, 10%, 15% and 20%,

respectively. Also the possibilities for tomato grower/shippers to have benefit-cost ratio

higher than 1.2 are 5% and 63% when tomato exports increase by 15% and 20%. It

suggests that there is 100% chance for tomato grower/shippers to break even and get

more margins if they adopt this temperature controlling technology.









Table 6-1 Probabilities of Benefit-cost Ratio for All Scenarios
Increased Increased Increased Increased Increased
Probabilities Export Export Export Export Export
(0%) (5%) (10%) (15%) (20%)

B/C < 1.0 15% 0% 0% 0% 0%

1.0 =< B/C < 1.1 85% 92% 24% 0% 0%

1.1 =< B/C < 1.2 0% 8% 76% 95% 37%

B/C >=1.2 0% 0% 0% 5% 63%



6.2 Analysis of Tomato Grower/Shippers' Costs with Increased Exports

All of the above benefit results imply that when temperature controlling technology

is adopted by tomato grower/shippers, tomato keeping quality would be improved, which

will create more possibilities to obtain higher benefits. However, on average for each

tomato export quantity increased, how much could be spent on technology adoption to

insure a given margin? Table 6-2 provides the answer.

Total revenue on average (Table 6-2) represents the average benefit from the

average export price with different tomato export scenarios. Highest possible margin is

the highest margin that tomato grower/shippers can get without paying anything for

temperature controlling technology. When temperature technology is not adopted by the

tomato grower/shippers, the highest margin that tomato growers can have is 2.5% on

average (Table 6-2). However, if temperature controlling technology is adopted by

tomato grower/shippers, the highest margin tomato growers can have varies from 7.2% to

21% depending on the different levels of tomato export quantity increased.

For instance, when tomato exports increase by 5% due to the technology adoption,

the highest possible margin for tomato grower/shippers is 7.2%. If tomato









growers/shippers wish to obtain a 5% margin instead of 2.5%, they still have 2.2%

margin ($87,947) left paying for the cost of technology. However, if tomato export

increases by 10%, the highest margin could be 11.8%. If a 5% margin meets tomato

grower/shippers' lowest expectation, they can use all or part of the remaining 6.8%

margin, which is $273,141 to pay for technology adoption. Or they can choose to pay

$70,228 for technology cost and get higher margin at 10%. This suggests that tomato

grower/shippers can decide to choose how much they would like to pay for technology

cost by reducing the margin from 10% all the way down to 5%.

In addition, if a 5% margin is the tomato grower/shippers' lowest expectation,

when tomato export increases by 15% and 20%, tomato grower/shippers would have

11.4% margin ($458,335) and 16.0% ($643,528) to pay technology cost. However, this

might be too much for the temperature controlling technology. Therefore, tomato growers

can get at least 10% margin and still have money left to pay for the technology cost.

When tomato export increases by 15%, tomato grower/shippers have $246,199 left to pay

technology cost. When tomato export increases by 20%, tomato grower/shipper have

$422,169 left to pay technology cost. If this is still too much, tomato grower/shippers can

decide to have at least 15% margin, with some money left to pay for technology cost for

these two last scenarios.

In summary, the table 6-2 shows that under any situation, the benefit and cost ratios

average greater than 1. However, after temperature controlling technology is adopted,

tomato grower/shippers can get far higher margins than before. Even more, for each

tomato export increasing scenario, they can choose a different margin level they like and

use the rest of them to pay the cost of technology.









Table 6-2 Result of Technology Cost Sensitivity Analysis

No Change Exports Exports Exports Exports
Categories in Exports Increased Increased Increased Increased
(0%) (5%) (10%) (15%) (20%)

Total Revenue on
Total Revenue on $4,261,163 $4,474,221 $4,687,279 $4,900,337 $5,113,395
Average
Highest Possible
2.5% 7.2% 11.8% 16.4% 21%
Margin
Margin (5%) $213,058 $223,204 $233,349 $243,495
Total Possible
Cost After Tech $4,261,163 $4,464,075 $4,666,988 $4,869,900
Cost After Tech
T ee $4,155,497 $4,173,215 $4,190,934 $4,208,653 $4,226372
Tech
Temperature
Temperature $87,947 $273,141 $458,335 $643,528
Technology Cost

Margin (10%) $426,116 $445,485 $464,854
Total Possible
Total Possible $4,261,163 $4,454,852 $4,648,541
Cost After Tech_
Total Cost Before $4,190,934 $4,208,653 $4,226372
Tech
Temperature $70,228 $246,199 $422,169
Technology Cost___

Margin (15%) $639,174 $666,965
Total Possible
Cost After Tech $4,261,163 $4,446,430
Cost After Tech
Total Cost Before
Tol Cost Before $4,208,653 $4,226372
Tech
Temperature $52,510 $220,059
Technology Cost


6.3 Benefit-Cost Analysis of Grower/Shippers and Wholesalers with Shrinkage
Reduction

Within the tomato supply chain between Puerto Rico and the United States, besides

tomato grower/shippers, tomato wholesalers are also important participants. If

temperature controlling technology is adopted by tomato grower/shippers, one possible

benefit for tomato supply chain is to have tomato quantity exported increased, which has









been discussed above. The other possible benefit for the supply chain is to have tomato

shrinkage reduced. The benefit and cost associated with this shrinkage reduction is

analyzed in this section.

Not all tomatoes exported to the United States are sold to retail stores or

foodservice operations. Some tomatoes are thrown away due to unacceptable keeping

quality. For this portion of the analysis, we are assuming when tomato keeping quality is

less than 2 days at wholesale level, these tomatoes will be considered a loss, which

reduces total revenue of tomato supply chain. However, after technology is adopted,

tomato shrinkage is reduced. Table 6-3 shows that before technology adoption, the

tomato supply chain experiences a 41.4% shrinkage and after technology adoption,

tomato has only 0.1% shrinkage, which means tomato shrinkage has been reduced by

99.8%.

Table 6-3 Growers and Wholesalers Benefits Analysis Regarding to Shrinkage Reduction

Before Tech After Tech Difference

Shrinkage 41.4% 0.1% 99.8%



Value of tomatoes with 2 days or less keeping quality

At grower prices $1,764,121 $4,261 $1,759,860

At grower price + 10% $1,940,533 $4,687 $1,935,846

At grower price + 20% $2,116,945 $5,113 $2,111,832

At grower price+ 30% $2,293,357 $5,539 $2,287,818


With the new technology, shrinkage is reduced, thereby increasing revenues for

those further down the marketing channel that were previously having to dispose of

spoiled tomatoes. Before technology adoption, with spoilage assumed at 2 days keeping









quality, shrinkage is 41.4%. At grower prices, the value of this quantity of tomatoes is

$1,764,121. However, after technology adoption, the value of tomatoes at 2 days or less

keeping quality is only $4,261, a difference of $1,759,860.

When tomato shrinkage reduces from 41.4% to 0.1% (Table6-3), the decrease in

the cost of spoilage is $1,759,860 in grower prices (Table 6-3). However, wholesalers

place a higher value on the tomatoes. Table 6-3 identifies the value of the tomatoes

under 2 days keeping quality using 10%, 20%, and 30% mark-ups at the wholesale level.

When mark-up between growers and wholesalers is 10%, the possible decrease in

spoilage is valued at $1,935,846. When the mark-up is 20%, the possible decrease in

spoilage is valued at$2,111,832 and when the mark-up is 30%, the decrease in spoilage is

valued at $2,287,818.

In summary, when temperature controlling technology is adopted by tomato supply

chain between Puerto Rico and the United States, tomato keeping quality is prolonged,

which on one side increase the probability of tomato export to the United States, and on

the other side reduce tomato's shrinkage. When shrinkage is reduced, the increased

revenue will benefit not only tomato growers, but also other participants, such as tomato

wholesalers. Moreover, wholesalers benefit even more than growers from the adoption of

the technology. Hence, technology cost could be shared with the supply chain. However,

whether the technology should adopt or not depends on the actual cost itself for tomato

supply chain between Puerto Rico and the United States.














CHAPTER 7
CONCLUSIONS

7.1 Summary

This research focused on the overall tomato supply chain performance between

Puerto Rico and the United States when new temperature controlling technology could be

adopted. Among food issues today, food safety and food quality are paid considerable

attention. Among all physical conditions, temperature is the most changeable and

controllable condition for perishable food quality. Research has shown that American

consumers desire higher quality products, and at the same time, Caribbean basin tomato

growers, especially Puerto Rican growers are looking for ways to increase their tomato

exports to the United States.

The primary objective of this research was to find out if the temperature controlling

technology is adopted by tomato grower/shippers, can improve tomato keeping quality,

leading to increase exports and revenue.

Researchers sought to understand the tomato supply chain between Puerto Rico and

the United States by gathering as much information as possible. As a result, five different

categories of data were collected from tomato firms in Puerto Rico and the United States.

Secondary data was collected from the internet and from telephone interviews.

To achieve this understanding, a tomato supply chain simulation model between

Puerto Rico and the United States was developed using simulation software, Simul8, to

find out how tomato keeping quality changes with the adoption of temperature

controlling technology. The results indicate that if the temperature range within marine









containers is narrowed from 11.20C (52.50F) to 31.40C (88.50F) down tol3.30C (56.00F)

to 15.60C (60.0F), tomato keeping quality would be improved by a factor 2 and shelf life

could be extended from 2.7 days to 5.4 days on average. In addition, the variation of

tomato keeping quality could be reduced by almost 85%.

Finally, based on the results from the tomato supply chain simulation model, a

benefit-cost sensitivity analysis was conducted to provide some guidelines for tomato

grower/shippers about whether or not adopting this temperature controlling technology

would be cost effective. Before temperature controlling technology was adopted, the

results indicated that, on average, a tomato grower/shipper's benefit is 1.025 times of

their cost, which means the highest margin for tomato grower is 2.5%. Furthermore,

tomato grower/shippers have a 85% chance to obtain a benefit that is greater than their

cost which means benefit-cost ratios are between 1.0 and 1.1, keeping in mind they still

have 15% chance that the benefit-cost ratio will be less than 1.

If temperature controlling technology is adopted by tomato grower/shippers, the

results show that, tomato grower/shippers have a 100% chance of reaching a benefit-cost

ratio higher than 1.0, which means 100% chance to break even. The probabilities for

tomato grower/shippers to have a benefit-cost ratio between 1.0 and 1.1 are 92% and 24%

when tomato export increases by 5% and 10% respectively. And the probabilities for

tomato grower/shippers to have a benefit-cost ratio between 1.1 and 1.2 are 8%, 76%,

95% and 37% when tomato export increases by 5%, 10%, 15% and 20%, respectively.

Also the possibilities for tomato grower/shippers to have a benefit-cost ratio higher than

1.2 are 5% and 63% when tomato export increases by 15% and 20%.









Further more, if tomato grower/shippers expect to earn a 5% margin, they have

$87,947, $273,141, $458,335 and $643,528 to pay for technology cost when tomato

quantity exported increased by 5%, 10%, 15% and 20%, respectively. If tomato

grower/shippers expect to obtain a 10% margin, they have $70,228, $246,199 and

$422,169 to pay for technology cost when tomato quantity exported increased by 10%,

15% and 20, respectively. Even If tomato grower/shippers expect to get 15% margin,

they have $52,510 and $220,059 to pay for technology cost when tomato quantity

exported increased by 15% and 20%, respectively.

The benefit-cost sensitivity analysis regarding to tomato shrinkage reduction by

technology shows that when tomato shrinkage reduces 99.8% by the adoption of

technology, the revenue for tomato growers increases by 70.5%. It also shows that this

increase will benefit both growers and wholesalers, which means that the actual total

revenue for tomato growers and wholesalers would be increased too depending on

different mark-up level. Further study of the specific the benefit-cost impacts for

distributors, wholesalers, and retailers are needed due to the difficulty of obtaining cost

data.

7.2 Conclusions

The results of tomato supply chain simulation model suggest that supply chain

complexities can be analyzed by simulation software. Controlling temperature within

marine container is one of the effective ways to increase tomato keeping quality as well

as reducing its variation when other physical conditions are held constant such as

humidity. Furthermore, increased tomato keeping quality means tomato's shrinkage will

be reduced. This will generate more volume in the market from Puerto Rico to the United

States, which also means that potential tomato export increasing exists.









Whether or not a tomato grower/shipper should adopt temperature controlling

technology depends on the benefit-cost relationship and financial performance. The

results of benefit-cost sensitivity analysis suggest that whether or not tomato

grower/shippers should adopt this temperature controlling technology depending not only

on the possibility of achieving a better benefit-cost ratio, but also on the cost of the

technology itself. The bottom line is tomato grower/shippers and other tomato supply

chain participants will have a greater chance to earn more money and enlarge their

market if temperature technology is adopted within this tomato supply chain between

Puerto Rico and the United States.

Overall, this research suggests that the opportunity for Puerto Rico tomato

grower/shippers to increase tomato keeping quality, leading to more volume and

shrinkage reduction of tomatoes exported to the United States, resulting in higher

revenues and margins, if tomato grower/shippers decide to adopt the temperature

controlling technology within marine containers.

7.3 Implications

The volume and value of tomatoes sold and consumed in the U.S. food system is

substantial Not only because their significant nutritional value, but also because tomatoes

are one of the most consumed fruits and vegetables in the United States. Therefore, the

opportunity to put more tomatoes into the United States market from Puerto Rico and

other Caribbean Basin countries exists. However, one key thing importers will have to

pay attention to is tomato quality since American consumers desire high quality products.

This research quantifies the costs and benefits associated with effective controlling

of temperatures, especially during marine shipment for Puerto Rican tomato

grower/shippers. If tomato keeping quality is improved, exports of tomatoes into the









United States will increase along with the possibility for increased margins. This method

can also be adopted by all other tomato grower/shippers from other Caribbean Basin

countries. The simulation model and benefit cost sensitivity analysis can be applied to

other perishable goods. This analysis can be used to find the temperature bottle-neck for

their supply chain and temperature controlling technology can be applied at critical

points. The model can be expanded to measure the impact of other variables related to

food quality other than just temperature, such as humidity.

Finally, further research is warranted using the simulation and benefit-cost

approach to studying food supply chains. Innovations would include the utilization of

more detailed financial information for tomato distributors, wholesalers and retailers.

This approach to studying food supply chains could be applied not only to state to state

trade within the U.S., but also to perishable foods other than tomatoes.















APPENDIX A
PICTURES OF TOMATO SUPPLY CHAIN DISTRIBUTION SYSTEM


Figure A-i Tomato Dumping Into the Packing Line


Figure A-2 Tomato Grading Line
































Figure A-3 Tomato Packing Line (25LB/box)


Figure A-4 Tomato Palletizing Operation (80 boxes/ pallet)







54






























Figure A-5 Tomatoes Stored in an Ethylene Room


r~;~i: *r;.* -"*


Figure A-6 Tomato Loading in Santa Isabel, Puerto Rico
































Figure A-7 Truck Shipments Leaving Santa Isabel, Puerto Rico


Figure A-8 Trucks Arriving at the Quincy FL Distribution Center














APPENDIX B
SIMULATION MODELING PARAMETERS

Before Technology Adoption

Parameters for Clock Properties:
Time unit: minute; 7 days / week; start from 00:00 and 24:00 hours per day;
Collection period: months (90*24*60 =129600 minutes)

Parameters for Preferences:
Distance: travel time = 0
Continues: units measure = tons ( Iton=2200 lbs)

Labels and Distributions:
Quantity: 1 tons = 2200 lbs
Grade: lbl_GradeMix with probability profile distribution of distGradeMix
(1-20%; 2-5%; 3-60%; 4-15%)
Size: lbl_Size with probability profile distribution of distSize
(1-25% extra large; 2-60% large; 3-15% medium)
Image: Image:Tomato ( 1 red tomato for local; 2 green tomato for US)
Color: lblColorMix with probability profile distribution of distColorMix
(1-30% red; 2-70% green)

Parameters for Harvesting:
Inter-arrival times (minutes): average 1.4 (480/341) with exponential
distribution
Label Actions: Quantity set to fixed value 2200 with fixed distribution
Lbl ColorMix set to the value of dist ColorMix
Image:Tomato set to the value of lbl_ColorMix

Parameters for Sorting:
Timing: average 1.4 (480/341) with exponential distribution
Resources: picker and packer pool (number required from 60 to 80)
Label Action: lbl GradeMix set to the value of dist GradeMix
lbl ColorMix set to the value of dist ColorMix
lbl Size set to the value of dist Size
Routing Out: Percentage (25% disposal and 75% for packing)









Parameters for Packing:
Timing: average 0.016 (1.4*25/2200 ) with the distribution of exponential
Resources: picker and packer pool (number required from 20 to 25)
Routing In: Collect 25, assemble, do not collect until all available
Routing Out: label (lbl_ColorMix; 1- red to local; 2-green to US)
Label Actions:
lbl_GradeMix set to fixed value of 1, 2, 3 and 4 for 4 packing lines;
lbl ColorMix set to the value of the distribution dist ColorMix
Image:Tomato set to the value of fixed value of lbl_ColorMix

Parameters for Palletizing:
Timing: 2 to 5 minutes with the distribution of uniform
Resources: picker and packer pool (number required from 2 to 4)
Routing In: Collect 80, assemble, do not collect until all available
Routing Out: label
( lbl_Size; 1- extra large to room 1; 2-large to room2; 3-medium to room 3)
Label Actions:
Lbl Size set to the value of the distribution dist Size
lbl GradeMix set to the value of the distribution dist GradeMix

Parameters for Ethylene Room:
Room 1: min wait time 5040 minutes (12*7*60 ), high volume
Room 2 and 3: min wait time 3600 minutes (12*5*60), high volume

Parameters for Truck Loading:
Timing: average 5 minutes with the distribution of average
Resources: truck loader (number required from 3 to 5)
Routing In: Collect 18, assemble
Routing Out: on complete VL of quality equation and Travel time
Label Actions:
lbl GradeMix set to the value of the distribution dist GradeMix
Lbl Size set to the value of the distribution dist Size

Labels, Distributions and Information Stores
Labels: lbl_Quality, lbl_TimeStamp
Distributions: dist TravelTimeTrPR (bounded normal; 120, 10, 240, 90)
distTravelTimeTrUS (bounded average; 180, 240, 120)
distTravelTimeShip (uniform; 2880, 4320)
distDispatchtoEndConsumers (1-5%-Canada; 2-5%-Food Services
3-14%-Groceries; 4-76%-Wholesalers)
distTabs (Uniform; 284.39, 304.52, min & max Tabs in marine container)
Information Stores: intTravelTimePR (number),
intTravelTimeShip (number), intTravelTimeUS (number),
int K (number; reaction rate within marine container)
intTabs (number; Tabs within marine container)
ss_Temperature (spread sheet for input temperature data)









ssLog (spread sheet for outputting)
gblLogRow (number for ss_Log use)

Parameters for "Truck Unloading PR"
Timing: average (5)
Resources:

Parameters for "Ship Loading"
Timing: average (15)
Resources:
Routing Out: on complete VL of ship travel time

Parameters for "Ship Unloading"
Timing: average (15)
Resources:
Routing Out: on complete VL of quality equation

Parameters for "Truck Loading US"
Timing: average (5)
Resources:
Routing Out: on complete VL of Truck Travel Time

Parameters for "Truck Unloading US"
Timing: average (5)
Resources:
Routing Out: on complete VL of quality equation

Parameters for "Cold Room"
Timing: average (10)
Resources:
Routing Out: on complete VL of quality equation,
On exit VL of the log of tomatoes out,
Percent: 5% (Canada); 5% (Food Services); 14% (Groceries); 76% (Wholesalers)

After Technology Adoption

Compared to parameters before technology, except the following parameters have

been changed, all other parameters for tomato supply chain simulation model after

technology keep the same.









Labels, Distributions and Information Stores
Labels: lbl_Quality, lbl_TimeStamp
Distributions: dist TravelTimeTrPR (bounded normal; 120, 10, 240, 90)
distTravelTimeTrUS (bounded average; 180, 240, 120)
distTravelTimeShip (uniform; 2880, 4320)
distDispatchtoEndConsumers (1-5%-Canada; 2-5%-Food Services
3-14%-Groceries; 4-76%-Wholesalers)
distTabs (Uniform; 286.48, 288.71)
Information Stores: intTravelTimePR (number),
intTravelTimeShip (number), intTravelTimeUS (number),
int K (number; reaction rate within marine container)
intTabs (number; Tabs within marine container)
ss_Temperature (spread sheet for input temperature data)
ss_Log (spread sheet for outputting)
gblLogRow (number for ss_Log use)














APPENDIX C
SIMULATION MODELING PROGRAMMING

VL SECTION: fncQuality Function
'fnc_Quality Function
SET lbl_Quality = lbl_Quality*EXP[0-[[ss_Temperature[6,intRow]]*[[Simulation
Time-lbl_TimeStamp]/[1440]]]]

VL SECTION: fnc_Tiime Stamp
SET lbl_TimeStamp = Simulation Time

VL SECTION: Reset Log Logicl
'Obeyed just after all simulation objects are initialized at time zero
'Resets the sslog spreadsheet
LOOP 1 >>> intRow >>> gbl_LogRow
LOOP 1 >>> int col >>> 6
SET ss_Log[intcol,intRow+1] = ""
SET gbl_LogRow = 0

VL SECTION: Packing 1 Quality Equation Work Complete Logic
'fnc_Quality Calculation at Packing
SET lbl_Quality = 100
SET int Row = 42
CALL fnc_Quality Function
SET lbl_TimeStamp = Simulation Time

VL SECTION: Packing 1 Time Stamp Route In After Logic
CALL fnc_Tiime Stamp

VL SECTION: Packing 2 Quality Equation Work Complete Logic
'fnc_Quality Calculation at Packing
SET lbl_Quality = 100
SET int Row = 42
CALL fnc_Quality Function
SET lbl_TimeStamp = Simulation Time

VL SECTION: Packing 2 Time Stamp Route In After Logic
CALL fnc_Tiime Stamp

VL SECTION: Packing 3 Quality Equation Work Complete Logic
'fnc_Quality Calculation at Packing
SET lbl_Quality = 100









SET int Row = 42
(Disabled) SET intPre Shelf Life = ss_Temperature[5,4]
(Disabled) SET intAft Shelf Life = ss_Temperature[5,4]
CALL fnc_Quality Function
SET lbl_TimeStamp = Simulation Time

VL SECTION: Packing 3 Time Stamp Route In After Logic
CALL fnc_Tiime Stamp

VL SECTION: Packing 4 Quality Equation Work Complete Logic
'fnc_Quality Calculation at Packing
SET lbl_Quality = 100
SET int Row = 42
CALL fnc_Quality Function
SET lbl_TimeStamp = Simulation Time

VL SECTION: Packing 4 Time Stamp Route In After Logic
CALL fnc_Tiime Stamp

VL SECTION: Palletizing US Action Logicl
IF lbl Size = 1
SET lbl route = 1
ELSE
SET lbl route = 2

VL SECTION: Pelltizing US Quality Equation Work Complete Logic
'fnc_Quality Calculation at Palletizing
SET int Row = 42
CALL fnc_Quality Function
SET lbl_TimeStamp = Simulation Time

VL SECTION: To Loading Route-In After Logic
CALL fnc_Tiime Stamp

VL SECTION: (After Ethylene Room) To Loading Quality Equation Work
Complete Logic
SET lbl_TimeStamp = Simulation Time

VL SECTION: Truck Loading PR Time Stamp Route In After Logic
CALL fnc_Tiime Stamp

VL SECTION: PR Truck Travel Time and Quality Equation Work Complete Logic
'Truck Travel Time from Santa Isabel to San Juan
SET int TravelTimePR = dist TravelTimeTrPR
Set Travel Time Truck Loading PR, intTravelTimePR, Queue for Truck
Unloading PR









'fnc_Quality Calculation at Palletizing
SET int Tabs = dist Tabs
SET intK = EXP[[0-[7277.48*[1/intTabs]]]+24.223]
'Calculate Quality according to Temperature distribution within Marine Container
SET lbl_Quality = lbl_Quality*EXP[[0-[int K]]*[[Simulation Time-
lbl_TimeStamp]/[1440]]]
SET lbl_TimeStamp = Simulation Time

VL SECTION: Ship Loading Time Stamp Route In After Logic
CALL fnc_Tiime Stamp

VL SECTION: Ship Travel Time Work Complete Logic
'Ship Travel Time from San Juan to Jacksonville
SET intTravelTimeShip = distTravelTimeShip
Set Travel Time Ship Loading intTravelTimeShip, Queue for Ship Unloading

VL SECTION: Ship Unloading Quality Equation Work Complete Logic
'fnc_Quality Calculation at Ship Unloading
SET int Tabs = dist Tabs
SET intK = EXP[[0-[7277.48*[1/intTabs]]]+24.223]
'Calculate Quality according to Temperature distribution within Marine Container
SET lbl_Quality = lbl_Quality*EXP[[0-[int K]]*[[Simulation Time-
lbl_TimeStamp]/[1440]]]
SET lbl_TimeStamp = Simulation Time

VL SECTION: Truck Loading US Time Stamp Route In After Logic
CALL fnc_Tiime Stamp

VL SECTION: Truck Unloading US Quality Equation Work Complete Logic
'fnc_Quality Calculation at Truck Unloading US
SET int Tabs = dist Tabs
SET intK = EXP[[0-[7277.48*[1/intTabs]]]+24.223]
'Calculate Quality according to Temperature distribution within Marine Container
SET lbl_Quality = lbl_Quality*EXP[[0-[int K]]*[[Simulation Time-
lbl_TimeStamp]/[1440]]]
SET lbl_TimeStamp = Simulation Time

VL SECTION: US Truck Travel Time Work Complete Logic
'Travel Time from Jacksonville to Quincy
SET int TravelTimeUS = dist TravelTimeTrUS
Set Travel Time Truck Loading US, intTravelTimeUS Queue for Truck
Unloading US

VL SECTION: Truck Unloading US Time Stamp Route In After Logic
CALL fncTiime Stamp









VL SECTION: (Out of Distribution Center) Dispatch to End Consumers Quality
Equation Work Complete Logic
'fnc_Quality Calculation After Distribution Center Cold Room
SET int Row = 48
CALL fnc_Quality Function
SET lbl_TimeStamp = Simulation Time
SET lblroute = dist DispatchToEndConsumers

VL SECTION: Log of Tomato Out of Distribution Center on Exit Logic
SET gblLogRow = gblLogRow+l
'To record the number of days since the tomato were harvested
SET ssLog[l,gblLogRow+1] = [Simulation Time-lbl_Time Harvested]/1440
SET ssLog[2,gblLogRow+l] = lbl_GradeMix
SET ss_Log[3,gblLogRow+l] = lbl_Size
SET ssLog[4,gblLogRow+l] = lbl_Quality
SET ssLog[5,gblLogRow+l] = Quantity
















APPENDIX D
BENEFIT-COST SENSITIVITY ANALYSIS SIMETAR MODELING

Model Input ((for 6.1 &6.2)

B C D E F G
Production
Unit Mean Min Max Remarks
Daily Production LB 75,000
Export period (1) days 59 January and February
Export period (2) days 31 March
Total Export Period days 90 =SUM(D7:D8)
Total Production in4,425,000 D6D7
LB 4,425,000 =D6*D7
Export Period (1)
Total Production in
LB 2,325,000 =D6*D8
Export Period (2)
Total Production in
Total Production in LB 6,750,000 =SUM(D10:D11)
Export Period

Quantity Sold
Disposal % 25%
U.S. Export % 52.5% =D18*70%
Local Market % 22.5% =D18*30%
Available Production % 75%
Quantity Export to U.S. in LB 2,323,125 =D10*D16
Period (1)
Quantity Export to U.S. in LB 1,220,625 D11*D16
Period (2)
Total Quantity Export LB 3,543,750 =SUM(D19:D20)
to U.S.


Export Price
Export Period (1) $ / LB 1.54 1.48 1.6 $37-$40 /box (25LB)
Export Period (2) $ / LB 0.56 0.52 0.6 $13-$15/box (25LB)


Costs
Total Fixed Cost $/Acre/Y 1,
1,623.06
ear
Total Variable Cost $/Acre/y 10,235.14
ear










30 Planted Acres Acres 2000
31 Total Production Cost 23,716,400.0(D28D29)D30
$/Year 0 =(D28+D29)*D30
32 Total Production Cost
32 Total Production Co $/90days 5,847,879.45 =(D31/365)*90
for Export Period
33 Transportation Cost $/LB 0.1 0.1
34 Total Transportation $ 354,375.00 =D33*D21
Cost for Export Period 3 7


Model Output 1 (for 6.1 &6.2)

36 Quantity Export to U.S. in LB 2,323,125 D19
Period (1)
37 Export Price in Period (1) $/LB 1.54 =UNIFORM(E24,F24)
38 Total Revenue in Period $ 3,577,612.50 D36*D37
(1)
39 Quantity Export to U.S in LB 1,220,625 D20
Period (2)
40 Export Price in Period (2) $/LB 0.56 =UNIFORM(E25,F25)
41 Total Revenue in Period $ 683,550.00 D39*D40
(2)
42 Total Revenue from
$Eport ( ) S 4,261,162.50 =D38+D41
Export (100%)
43 Total Cost of Export 4,155,496.6D32(D16+0.5D15)
$ 4,155,496.64 +D34
+D34


This is the "Expected Value" model, which means that unit D36 and unit D39 are

mean valuel.54 and 0.56. However, D36 and D39 are stochastic variables with the

uniform distribution with the min and max as the parameters. If the key of "Expected

Value" be pressed again, the model will be appeared in "Stochastic Value" model, which

means the number in unit D36 and unit D39 both can be changed any time according to

their distribution.

The output model (2) is also the "Expected Value" model like the above. All the

revenues are expected mean values. If the stochastic model is chosen, all the revenue

numbers will be changed due to the stochastic price.












Model Output (2) Sensitivity Analysis for Benefit (for 6.1 &6.2)


Total
Export Quantity Q(LB) Period 1 Remarks Period 2 Remarks otal Remarks
Quantity
Q=105% 2,439,281 =$D$36*105% 1,281,656 =$D$39*105% 3,720,938 =C49+E49
Q=110% 2,555,438 =$D$36*110% 1,342,688 =$D$39*110% 3,898,125 =C50+E50
Q=115% 2,671,594 =$D$36*115% 1,403,719 =$D$39*115% 4,075,313 =C51+E51
Q=120% 2,787,750 =$D$36*120% 1,464,750 =$D$39*120% 4,252,500 =C52+E52


Total
Revenue ($) Period 1 Remarks Period 2 Remarks TolRemarks
Revenue ($)
Q=105% 3,756,493.13 =C49*$D$37 717,727.50 =E49*$D$40 4,474,220.63 =C55+E55
Q=110% 3,935,373.75 =C50*$D$37 751,905.00 =E50*$D$40 4,687,278.75 =C56+E56
Q=115% 4,114,254.38 =C51*$D$37 786,082.50 =E51*$D$40 4,900,336.88 =C57+E57
Q=120% 4,293,135.00 =C52*$D$37 820,260.00 =E52*$D$40 5,113,395.00 =C58+E58














E

100% Cost


F

110% Cost


G

120% Cost


Total Cost
Total Cot Q =$D$32*($D$16+0.5*$D$15)+
of Export Q=100% 4,155,496.64 4,155,496.64 4,571,046.30 4,986,595.97 $D$32*($D$16+.5*$D$15)+
$D$34
Total Cost
=$D$32*($D$16+0.5*$D$15)+
of Export Q=105% 4,173,215.39 4,173,215.39 4,590,536.93 5,007,858.47 $D$32*($D$16+*$D$15)+
G49*$D$33
Total Cost
=$D$32*($D$16+0.5*$D$15)+
of Export Q=110% 4,190,934.14 4,190,934.14 4,610,027.55 5,029,120.97 $D$32*($D$16+5*$D$15)+
G50*$D$33
Total Cost
=$D$32*($D$16+0.5*$D$15)+
of Export Q=115% 4,208,652.89 4,208,652.89 4,629,518.18 5,050,383.47 $D$32*($D$6+5*$D$15)+
G51*$D$33
Total Cost
Total Cot Q~=$D$32*($D$16+0.5*$D$15)+
of Export Q=120% 4,226,371.64 4,226,371.64 4,649,008.80 5,071,645.97 $D$32*($D$16+5*$D$15)+
G52*$D$33
($)jZkp~kj










The followings are the input and output model for 6.3 in Chapter 6. The model

shows in the "Expected Value" style so all the numbers represents the mean value.

However, this model can be showed by "Stochastic" Style in Simetar.

Model (for 6.3)

A B C D E F G
2 Input Data
3
4 Export Quantity
Quantity Export to U.S. in
5 Period (1) LB 2,323,125
Quantity Export to U.S. in
6 Period (2) LB 1,220,625
7 Total Quantity Export to U.S. LB 3,543,750
8
9 Export Price Mean Min Max
10 Export Period (1) $/LB 1.54 1.48 1.6 $37-$40/ box (25LB)
11 Export Period (2) $ / LB 0.56 0.52 0.6 $13-$15 /box (25LB)
12
13 Shrinkage
14 Before Tech % 41.4% Shelf Life less than 2 days
15 After Tech % 0.1% Shelf Life less than 2 days


A B C D E F G
Before Technology
18 Model
19
20 Actual Quantity Sold
Actual Quantity Sold to U.S. in
21 Period (1) LB 1,361,351 =D5*(1-D14)
Actual Quantity Sold to U.S. in
22 Period (2) LB 715,286 =D6*(1-D14)
Total Actual Quantity Sold to
23 U.S. LB 2,076,638 =SUM(D21:D22)
24
25 Export Price
26 Export Period (1) $ / LB 1.54 =uniform(E10,F10)
27 Export Period (2) $ / LB 0.56 =uniform(E 11,F11)
28
29 Actual Revenue
30 Actual Revenue in Period (1) $ 2,096,481 =D26*D21
31 Actual Revenue in Period (2) $ 400,560 =D27*D22
32 Total Actual Revenue $ 2,497,041 =SUM(D30:D31)












A B C D E F G
After Technology
35 Model
36
37 Actual Quantity Sold
Actual Quantity Sold to U.S.
38 in Period (1) $ 2,320,802 =D5*(1-D15)
Actual Quantity Sold to U.S.
39 in Period (2) $ 1,219,404 =D6*(1-D15)
Total Actual Quantity Sold
40 to U.S. $ 3,540,206 =SUM(D38:D39)
41
42 Export Price
43 Export Period (1) $ / LB 1.54 =uniform(El 0,F10)
44 Export Period (2) $ / LB 0.56 =uniform(E 11,F1 1)
45
46 Actual Revenue
47 Actual Revenue in Period (1) $ 3,574,035 =D43*D38
48 Actual Revenue in Period (2) $ 682,866 =D44*D39
49 Total Actual Revenue $ 4,256,901 =SUM(D47:D48)
50
Total Actual
51 Revenue increased $ 1,759,860 =D49-D32
Total Actual
52 Revenue increased % 70.5% =(D49-D32)/D32
53 shrinkage Reduced % 99.8% =(D14-D15)/D14















LIST OF REFERENCES


Chang, R. (1981). Physical Chemistry with Applications to Biological Systems.
Macmillan Publishing Co., New York.

Caswell, J.A., Bredahl, M.E., and Hooker, N. H. (1998). How Quality Management
Metasystems Are Affecting the Food Industry. Review of Agricultural Economics.
Vol.20, No. 2, pp. 547-557.

Deming, W.E. (1967). What Happened in Japan? Industry Quality Control. Vol. 24, pp.
89-93.

Deming, W.E. (1972). Report to Management. Quality Process. Vol. 5, No. 2-3, pp. 41.

Economic Research Service. United States Department of Agricultural. (2004).
Commodity Highlight: Fresh Tomatoes.
http://www.ers.usda.gov/Briefing/Vegetables/vegpdf/FrTomatoHigh.pdf, Date
visited, 02/15/2006.

Economic Research Service. United States Department of Agricultural. (2006). the
F.o.b.-Retail Price Relationship For Selected Fresh Vegetables.
http://www.ers.usda.gov/Briefing/Tomatoes/tomatopdf/FOBRetailPriceVeg.pdf,
Date visited, 02/15/2006

Food and Agricultural Organization. (2003). FAOSTAT Trade Data.
http://faostat.fao.org/site/342/default.aspx, Date visited, 10/25/2004

Food Marketing Institute. (2003). Food Industry Review 2003. Food Marketing Institute,
Washington, D.C.

Heien, D. M. (1980). Markup Pricing in a Dynamic Model of the Food Industry.
American Journal of Agricultural Economics. Vol. 62, pp. 11-18

Institute of Food and Agricultural Science. (2003-2004). "Florida Tomato Growers'
Production Cost". http://www.agbuscenter.ifas.ufl.edu/cost/cop03-
04/SWTomatoSC.doc, Date Visited, 06/05/2005

Juran, J. M. (1974). Quality Control Handbook. McGraw-Hill. New York.

Juran, J. M., and Gryna, F. M. Jr. (1970). Quality Planning and Analysis. McGraw-Hill.
New York.









Marquardt, D.W. (1984). New Technology and Educational Directions for Managing
Product Quality. The American Statistician. Vol.38, No. 1, pp. 8-14.

National Oceanic & Atmospheric Administration. (2005). 30 Years Average
Temperature. http://www.ncdc.noaa.gov/oa/climate/online/ccd/nrmavg.txt, Date
Visited 10/05/2005

Nunes, Cecilia (2005). Tropical and Sub-Tropical Agricultural Research Tomato project
process report. October, Gainesville, Florida

Powers, N. J. (1995). Sticky Short Run Prices and Vertical Pricing: Evidence from the
Market for Iceberg Lettuce. Agribusiness. Vol. 11, No. 1, pp. 57-75

Robertson, G. L. (1993). Food Packaging. Marcel Dekker Inc. New York, NY. pp. 338-
352.

Science Made Simple. (2005). Temperature Conversions (From Fahrenheit to Kelvin).
http://www.sciencemadesimple.net/temperature.php, Date Visited 10/05/2005

Schepers, H.E., Henten E. J. V., Bontsema, J. and Dijksterhuis, G.B. (2004). Tactics of
Quality Management and Promotions: Winning Consumers for Fresh Exotic
Produce. Dynamics in Chains and Networks. Wageningen Academic Publishers.
Wageningen. The Netherlands. pp. 568-582.

SIMUL8 Corporation. (2006). About SIMUL8 Corporation.
http://www.simul8.com/about/index.htm, Date Visited 05/15/2006.

SIMUL8 Corporation. (2006). SIMUL8 Case Studies.
http://www.simul8.com/products/standard/casestudies.htm, Date Visited
05/16/2006.

Tijskens, L. M. M., Sloof, and M. & Wilkinson, E. C. (1994). Quality of perishable
produce. A Philosophical Research. In proceedings COST94 Workshop, October,
Wageningen, The Netherlands.

Tijskens, L. M. M., and Polderdijk, J. J. (1996). A Generic Model for Keeping Quality of
Vegetable Produce During Storage and Distribution. Agricultural Systems. Vol. 51,
No. 4, pp. 431-452.

Ward, R.W. (1982). Asymmetry in Retail, Wholesale and Shipping Point Pricing for
Fresh Vegetables. American Journal of Agricultural Economics. Vol.64, No.2, pp.
205-212

Whittington, D. and Grubb, W.N. (1984). Economic Analysis in Regulatory Decisions:
The Implication of Executive Order 12291. Science, Technology, & Human
Values. Vol. 9, No. 1, pp. 63-71






72


Worth, Thomas. (1999). The F.O.B. Retail Price Relationship For Selected Fresh
Vegetables. Economic Research Services. United State Department of Agricultural
http://www.ers.usda.gov/Briefing/Tomatoes/tomatopdf/FOBRetailPriceVeg.pdf,
Date visited. 02/15/2006















BIOGRAPHICAL SKETCH

Jiaoju Ge was born in Hubei province, P.R. China. After she obtained her Bachelor

of Science degree in investment economics from Sichuan University in 1999, she worked

for LG Electronics for 3 years as a marketing product manager.

In August 2004, Jiaoju Ge entered the food and resource economics Master of

Science program and specialized in marketing in University of Florida under the direction

of Dr. Allen Wysocki. She will continue her Ph.D. program in the Food and Resource

Economics Department beginning on August 2006. She was married to Qiyong Xu on

May 21st, 2002.