An activity simulation model for the analysis of the harvesting and transportation systems of a sugarcane plantation

Please download to get full document.

View again

of 18
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Similar Documents
Information Report



Views: 56 | Pages: 18

Extension: PDF | Download: 0

An activity simulation model for the analysis of the harvesting and transportation systems of a sugarcane plantation
  Computers and Electronics in Agriculture32 (2001) 247–264 / locate / compag An activity simulation model for the analysisof the harvesting and transportation systems of a sugarcane plantation Enrique Arjona  a, * ,1 , Graciela Bueno  a , Luis Salazar  b a Colegio de Postgraduados ,  Especialidad de Estadistica ,  Montecillo ,  Edo .  de Mexico  56230  ,  Mexico b Uni   ersidad Tecnologica del Valle del Mezquital  ,  Ixmiquilpan ,  Hgo .  42300  ,  Mexico Received 18 September 2000; received in revised form 4 April 2001; accepted 11 July 2001 Abstract We developed a discrete event simulation model of the harvesting and transportationsystems of a sugarcane plantation in Mexico that covers all processes from the burning of thecane to its unloading in the mill yard. The model was built to solve a problem with theamortization of machinery used in the plantation. The model was fit and validated usingplantation field data collected over an entire year. With model results we showed thatmachinery is underutilized and found possible solutions to the problem. The solutionsinvolve increasing the efficiency of machinery use, thereby allowing a reduction in theamount of machinery without increasing sugarcane processing times. © 2001 ElsevierScience B.V. All rights reserved. Keywords :   Sugarcane; Discrete simulation; Computer-aided analysis 1. Introduction The sugar industry plays an important role in Mexico, both as a basic productsupplier and a source of employment. The industry requires much machinery,including harvesters, loaders, trucks, and tractors. This machinery often does notbelong to the plantation and is rented from other farmers. To purchase machinery, * Corresponding author. Tel.:  + 52-5-955-5801; fax:  + 52-5-954-4294. E  - mail address : (E. Arjona). 1 arjona – / 01 / $ - see front matter © 2001 Elsevier Science B.V. All rights reserved.PII: S0168-1699(01)00168-5  E  .  Arjona et al  .  /   Computers and Electronics in Agriculture  32 (2001) 247–264  248 farmers obtain long-term loans at low interest rates from lending institutionsoperated by the Mexican government. Usually only one loan is granted to eachfarmer. In spite of the low interest rates of the loans, machinery owners are oftenlate in paying loans or default on them, a serious problem for the lendinginstitutions. For their part, farmers argue that profits from machinery use areinsufficient for timely repayment of loans. However, the managers of the lendinginstitutions say that the machinery is underutilized. To help address this problem,we developed a discrete event model that simulates the harvesting and transporta-tion systems of a specific sugarcane plantation faced with the difficulties describedabove. The objectives of the model are to estimate the effect of machineryreductions on daily total processing times, and to evaluate if machinery use isefficient enough to repay loans.Sugarcane harvesting and transportation are complex, and include daily planningof areas to harvest, and allocating labor and machinery for burning, trimming,heaping, loading, and delivering the cane from the plantation to the mill yard.Harvesting occurs simultaneously in separate, often distant sections of the planta-tion, making it hard to share machinery among sections. Machinery is transportedover local roads in riparian areas where few bridges exist; it must often be ferriedacross rivers. Finally, sugarcane must be harvested within certain periods of plantmaturity and, once trimmed, the cane should be milled within 24 h to preserveweight, saccharose content and juice quality.Modeling of sugarcane harvesting and transportation systems by Whitney andCochran (1976), Cochran and Whitney (1977) used queuing theory to predictdelivery rates. Their models are for a transport system of tractors and wagons andone continuous road between the plantation and the mill. It is assumed that loadingtimes follow an exponential distribution and that arrival times at the plantation arePoisson distributed. Outputs of the models are nomographs that can be used topredict the rate at which sugarcane can be delivered from the plantation. Crossley(1987) analyzed a series of mechanical transportation factors that limit vehiclespeed. His analyses focused on road upgrading costs and benefits. He developed acomputer program to simulate and predict the performance and yearly operatingcosts of sugarcane transport systems. Transport systems considered consisted of diverse types of tractors and wagons and continuous roads between the plantationand the mill. The data required by the program includes road geometry, surfacecharacteristics, and technical specifications of vehicles. Singh and Abeygoonawar-dana (1982), Singh and Pathak (1994) developed computer systems to simulatemechanical harvesting and transporting. The model embedded in these systemsconsiders one continuous road, one combine harvester and multiple trucks. Milltime is considered constant. The systems compute annual operating costs anddetermine the optimum number of trucks for the harvester. More recently, Salassiand Champagne (1998) developed a spreadsheet model that uses capacity and costestimates of equipment to evaluate alternatives for efficient use. In this model,harvesting is done mechanically using both combine and wholestalk harvesters.Transport units are tractors and wagons that deliver the cane to the mill or totransloading sites joined to the plantation by continuous roads. The model uses  E  .  Arjona et al  .  /   Computers and Electronics in Agriculture  32 (2001) 247–264   249 fixed and variable equipment costs to estimate the total cost of owning andoperating different equipment configurations. No random variables are consideredin this model.The model developed in this paper covers both mechanical and semi-mechanicalharvesting, as well as all harvesting and transportation processes from the burningof the cane to its unloading in the mill yard. The transportation system is integratedby trucks, ferries and multiple roads. Roads between the plantation and the millmay be continuous or disrupted by rivers. Mill processing is modeled in detail toanalyze bottlenecks at unloading areas. Most of the variables in the model arerandom. Distributions are fit from sampling data and no a priori assumption ismade on them. Machinery is not owned by the plantation and machinery costs areindependent of the number of units utilized.This paper is organized as follows: In Section 2, we describe the main character-istics of the plantation under study. Section 3 presents the amortization problemsthat face machinery owners. In Section 4, we discuss how a simulation model canhelp to solve these problems. Sections 5–8 are dedicated to the simulation model.These sections present details of the model, how data was acquired and fit, how themodel was validated, and the results of the experiments performed. Finally, inSection 9 we give some conclusions. 2. The plantation The model was developed and implemented for a sugar plantation in the state of Veracruz, on the Gulf Coast of Mexico. On average some 3200 ton / day of cane areharvested and milled on the plantation. The sugar-making season lasts about 161days. The plantation is divided into seven sections; each section integrates severalsmall, rural communities. Three of the roads that connect the sections and the millyard merge at a river crossing served by a ferry. Most of the cane is trimmedmanually; heaping and loading are wholly mechanized. The plantation actually usestwo mechanical harvesters, 27 loaders, 127 trucks and 54 tractors and wagons. Allthis machinery is rented with the exception of two tractors that belong to the millowners. Rental costs are determined on a job-by-job basis. The average labor forceduring the sugar-making season is about 1400. Each section of the plantation isassigned a weekly amount of cane to harvest, based on estimated saccharosecontent. Weekly plans for allocating labor and machinery are made accordingly. 3. Machinery use and amortization problems To estimate average machinery use and income of machinery owners, wecollected plantation field data for a year on machinery use and performance. Exceptfor a few cases, machinery was used only during the sugar-making season.The average amount of cane transported per truck during the study period was25.2 ton / day. The average cane loaded per loader was 118.5 ton / day. Rental costs  E  .  Arjona et al  .  /   Computers and Electronics in Agriculture  32 (2001) 247–264  250 for machinery during the study period were MXP$12.40 for each ton of canetransported from the plantation to the mill yard, and MXP$3.80 for each ton of cane loaded. The long-term loan prices (including interest and finance costs) of machinery during the study period were the following, MXP$205000 for a truck;MXP$292000 for a loader. Loans are paid in five equal annual payments of MXP$41000 for a truck and MXP$58400 for a loader.Considering a sugar-making season of 165 days (the length of the season in thestudy period), a machinery operating cost of 50% of the gross rental income, anannual inflationary increase of 20% in rental costs, and an annual interest rate of 24% for payments not covered, machinery owners face serious amortization prob-lems (Table 1). Net annual income of the owners is not enough to cover annualpayments and past debts. The machinery owners do not obtain any profit formachinery rentals during the amortization period, and at the end of that periodtruck owners still have a debt greater than one annual payment. 4. Solving amortization problems When trying to solve the above problems several difficulties arise. On one side,rental costs are regulated and may only be increased by inflation indices. So theonly way for machinery owners to improve their earnings is to increase the amountof cane loaded and transported. On the other hand, the amount of cane that can beloaded and transported is restricted by the actual capacity of the plantation; thus,the only way to increase the amount of cane loaded and transported by each loaderand truck is to reduce machinery in the plantation. But machinery reductions willonly be feasible if increases in daily processing times caused by the reductions are Table 1Actual amortization planYear 4 Year 5Year 3Year 2Year 1 Trucks Gross income 51 559 61 871 74 245 89 094 106 913Net income 25 780 30 935 37 123 44 547 53 45656 73649 30235 883Past debts 18 8730 41 000 41 00041 000 41 000Annual payments 41 000 − 15 220  − 28 938Balance  − 39 760  − 45 755  − 44 2800 0Profit / net income 000 Loaders 128 390 154 06789 159Gross income 106 99174 300 37 150 44 580Net income 53 496 64 195 77 034Past debts 0 26 350 49 811 67 847 76 94458 40058 40058 40058 400 58 400Annual payments − 21 250  − 40 170Balance  − 54 715  − 62 052  − 58 31000 000Profit / net income  E  .  Arjona et al  .  /   Computers and Electronics in Agriculture  32 (2001) 247–264   251 minimal. One objective of the simulation model presented in this paper is toestimate those increases for each proposed machinery reduction. 5. The simulation model The model was developed using the activity approach. (Poole and Szy-mankiewicz, 1977; Kreutzer, 1986). Components of an activity model are entities,waiting lines, activities, and entity flows. Entities represent the physical or logicalcomponents of the system and may have attributes. When an activity model isexecuted, entities move from waiting lines to activities and vice-versa. Entities inwaiting lines are inactive. At the beginning of an activity, the entities required forthe activity are taken from the waiting lines. When an activity is in progress, sometaken entities may be modified, others may be consumed, and new entities may begenerated. At the ending of an activity those entities that were not consumedbecome inactive and are stored in the waiting lines. Activities may start any time(whenever the entities required by them are available in the appropriate waitinglines).The number of entities required to start an activity may be constant or variable.There may be several choices of entities required in an activity and of waiting linesfrom which the entities are taken. Entities required for an activity may beconstrained to specific positions in the waiting lines or to meet attribute conditions.Attribute conditions may be absolute or relative to attribute values of othersentities required by the activity. The duration of an activity may be constant orvariable. Activity durations may depend on attribute values of the entities thatparticipate in the activity and on theoretical or empirical distributions. At the endof an activity, entities may be stored in any of several waiting lines and at specificpositions in those lines.As mentioned, the model developed covers both mechanical and semi-mechanicalharvesting, as well as all harvesting and transportation processes from the burningof the cane to its unloading in the mill yard. Fig. 1 depicts harvesting and Fig. 2transportation and part of processing (a system that sometimes causes bottlenecksin transportation). Rectangles represent activities, circles represent waiting lines,and directed arcs represent flows of entities. Each entity represents a physical orlogical component of the system (raw materials, machinery, workers, orders,products, etc.). The main entities, waiting lines, and activities of the model are givenin Table 2 (for clarity, components used for scheduling and information flows werenot listed. Some of those components appear in Fig. 1 as empty boxes and circles).Each activity has specific flow conditions, duration, and attribute modifications.For example, the activity ‘river crossing of loaded trucks’ (A10) requires thefollowing entities.   An inactive ferry in the plantation shore (WL29).   Three to five loaded trucks waiting to cross the river (WL8).These entities do not have to meet any condition on their attributes. At thebeginning of the activity they are taken from the first positions of the waiting lines.The number of trucks taken will depend only on the availability of trucks in WL8.
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!