Layout evaluation of large capacity warehouses

Facilities Emerald Article: Layout evaluation of large capacity warehouses José Ignacio Huertas, Jenny Díaz Ramírez, Federico Trigos Salazar Article information: To cite this document: José Ignacio Huertas, Jenny Díaz Ramírez, Federico Trigos Salazar, (2007),"Layout evaluation of large capacity warehouses", Facilities, Vol. 25 Iss: 7 pp. 259 - 270 Permanent link to this document: http://dx.doi.org/10.1108/02632770710753307 Downloaded on: 11-04-2012 References: This document contains references to 18 other documents To copy this document: permissions@emeraldinsight.com This document has been downloaded 3075 times. Access to this document was granted through an Emerald subscription provided by COLLEGE OF PROFESSIONAL AND CONTINUING EDUCATION LIBRARIES For Authors: If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service. Information about how to choose which publication to write for and submission guidelines are available for all. Additional help for authors is available for Emerald subscribers. Please visit www.emeraldinsight.com/authors for more information. About Emerald www.emeraldinsight.com With over forty years' experience, Emerald Group Publishing is a leading independent publisher of global research with impact in business, society, public policy and education. In total, Emerald publishes over 275 journals and more than 130 book series, as well as an extensive range of online products and services. Emerald is both COUNTER 3 and TRANSFER compliant. The organization is a partner of the Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation. *Related content and download information correct at time of download. Layout evaluation of large capacity warehouses Jose´ Ignacio Huertas, Jenny Dı´az Ramı´rez and Federico Trigos Salazar School of Engineering and Architecture at ITESM, Toluca, Mexico Abstract Purpose – The purpose of this article is to present a model to estimate and evaluate the operational costs of alternative layouts for large capacity warehouses or distribution centers with a large variety of goods. Design/methodology/approach – The proposed model is based on time and resources studies per each of the basic activities on a warehouse operation. For validation purposes, the proposed model was applied on a perishable goods warehouse in Mexico. The output data was compared to actual data. Performance measures were operational costs and average picking time. Findings – It was found that the proposed model is robust, flexible, simple and easy to be implemented. The model was used to evaluate two new alternatives of layout and operations of the same warehouse. It was found that the option with the layout with docks on long opposite sides of the warehouse and the operation without a separate picking zone minimizes operational costs. Research limitations/implications – The richness of the model is strongly supported by the information the warehouse has about its operation. With knowledge of the process, it is required to distinguish deterministic from stochastic basic activities and develop distance computations that depend on the layout being studied. Practical implications – The approach used to model warehouse operations was to estimate the movements and resource consumption per commodity. This allows the model to be used in every operational context when the complexity of the system is strongly dependent on and proportional to the volume of operations. In addition, it is particularly adequate as a tool to compare average performance measures of different scenarios for the same system. Originality/value – The model proposed here provides a simple way to estimate particularly operational resource consumptions and picking times as proxy measures for efficiency and efficacy of a warehouse. It uses distance computations, time information and unit occurrence frequencies of basic activities over a single commodity in the system. Keywords Warehouses, Distribution centres, Operating costs, Plant layout, Facilities, Mexico Paper type Research paper Introduction In an attempt to benefit from scale economies, some companies produce big size lots of their different products. To satisfy market demand, they store inventory production in large capacity facilities located at strategic places. This is the case of companies that produce wheat-based commodities where lots are usually large in volume. The The current issue and full text archive of this journal is available at www.emeraldinsight.com/0263-2772.htm This work was financed by La Moderna S.A. and it was applied to the Palmillas warehouse during 2005. The authors are grateful for the involvement of the engineers Francisco Monroy, Narciso Manzola and Isidro Segura from La Moderna. They are also are grateful for the contributions of ITESM students Paulina Lara y Alejandra Vernet. Large capacity warehouses 259 Received November 2006 Accepted March 2007 Facilities Vol. 25 No. 7/8, 2007 pp. 259-270 q Emerald Group Publishing Limited 0263-2772 DOI 10.1108/02632770710753307 operation of these warehouses frequently represents high costs for the respective supply chain. However it is considered as of crucial importance for the customer service strategy, since it constitutes the last link in the chain before reaching the customer. The facility layout plays an important role in the business success of the company (Johnston, 1995). The most appropriate warehouse layout depends on its particular operational conditions, and characteristics such as modularity, adaptability, compactness, distribution of movements, accessibility, and flexibility (Hassan, 2002). As an example, Figure 1 presents a warehouse layout classification based on docks location for reception and shipment (Bartholdi and Hackman, 2005). Table I lists some of the metrics used to evaluate warehouses performance (Tranfield and Akhlaghi, 1995; Coyle et al., 2003; Tompkins et al., 2003; Bartholdi and Hackman, 2005). Layout design is problem-dependant, i.e. there is no best design, methodology or policy for all problems under consideration. Selecting the best layout for a given case is not trivial because of the diversity of factors influencing a warehouse operation such us docks location, aisles access, racks types, racks access, etc. (Tompkins et al., 2003; Carranza et al., 2004; Bartholdi and Hackman, 2005). Though there exists specialized Performance measures of a warehouse Quality Accuracy in storage Accuracy in picking Inventory Finances Operational costs Total storage costs per unit Cycle time Commodity cycle time Order cycle time Productivity Labor productivity (employees/moved unit) Resource consumption (space, equipment, labor) Flow (moved units through the system in a given period) Throughput volume (moved units/day) Productivity ratio (handled units per day/working hours per day) Table I. Performance evaluation metrics for warehouses Figure 1. Layout classification according reception and shipment docks location in warehouses F 25,7/8 260 commercial software for this purpose (e.g. simulation software), these are expensive and hard to use, as they require a lot of time and energy (Muroff, 1993). One of the most studied warehouses problems in the last 30 years is the layout design and product allocation to increase floor space utilization, and decrease traveling time, material handling and costs (Larson et al., 1997; Canen and Williamson, 1998; van den Berg, 1999; Shouman et al., 2005). However, the available design configurations methodologies do not include specifically the decision problem of storage partition. Warehouses are usually partitioned into reserve and picking areas to facilitate the operations and reduce movements. Partitions depend on factors such as the demand, the size and type of unit loads (Hassan, 2002; Garfinkel, 2005). In addition, layout design frameworks and steps for implementations are intended for the case when the warehouse has not been built yet (Benefield, 1999; Hassan, 2002; Hales, 2006). However, warehouse managers, who recognize that the warehousing operations can be improved by changing one or several characteristics of the current layout, do not easily find methods to evaluate these changes using consistently the current operational data, Thus, there is a need for an easy-to-use model to evaluate performance measures of a given layout, based on its particular operational conditions. In response to this need, this work presents a performance evaluation model of the warehouse operations, based on time and other resources that are associated to each basic activity carried out within a warehouse. Initially a description of the proposed model is presented. Later on, the model applicability is evaluated for the case of a large capacity warehouse in Mexico. Finally, the best configuration for this warehouse is identified by comparing three alternatives with different storage partition and docks location decisions. Description of the proposed model The objective of the proposed model is twofold: first, to evaluate the operation of large capacity warehouse based on the estimation of some key performance measures obtained from method and time studies of the current operation. Second, to allow an easy comparison of different layout configurations. Input data must describe the warehouse layout, processes characteristics (i.e. time study output), and demand or flow (e.g. monthly flow of material). Since normally commodities are of different size and weight, and their units are not unique (e.g. carton, pack, piece, etc.), it is required to convert them into one single unit. The proposed model uses “pallet load” or just “pallet” as a volume unit, meaning the amount of product that typically is stored over a pallet in the rack. Output data are indicators reflecting operational cost (e.g. $/month) and performance (e.g. picking time). Methods and time study Method and time studies identify the basic activities of a process and determine the times required to develop each of them. It is required to register the time required to develop each activity several times. In addition, it is necessary to make a “work rhythm valuation”, and to assign supplement times according to the activity type (Garcı´a Criollo, 1998). Studies of this kind in warehouses have shown that: Large capacity warehouses 261 . Every commodity goes through the same basic activities (Hassan, 2002). These identified basic activities are illustrated in Figure 2. . The most time consuming activity is transportation (e.g. a picker spends more than 50 percent of his/her time in moving the products from place to place (Bartholdi and Hackman, 2005; Tompkins et al., 2003)). . Every basic activity can be associated to a worker, and to some resources used during its execution. Operational cost of a warehouse Table II shows a typical operational cost composition of a warehouse. The total operational costs strongly depend on the layout and the operational methods used (Bartholdi and Hackman, 2005). Table II stands for those costs that depend on the warehouse size and operation. Remaining costs stay approximately invariant with layout and operation changes. Hence, for performance comparison purposes of different layouts of a given warehouse, it is only required to consider the operational variable costs (C): C ¼ Ce þ Cmo ð1Þ where: Ce Equipment utilization and maintenance cost. Cl Labor cost. Equipment utilization and maintenance costs estimation Costs of materials handling equipment are basically composed of equipment utilization cost (e.g. purchase, depreciation, leasing, etc.), and maintenance cost (e.g. internal, external, parts, etc.). The model requires average data from a representative period of time under normal working conditions: Figure 2. Each good unit in a warehouse goes through the same cycle of basic activities Managerial Operational Fixed Security – safety Labor Space utilization Variable Opportunity costs Fleet Taxes Labor Equipment utilization and maintenance Utilities (power, water, etc.) Materials Table II. Cost structure for a warehouse F 25,7/8 262 Ce ¼ j X q X Cejq ð2Þ Cejq Monthly equipment utilization and maintenance cost of equipment q required to do the basic activity j. Labor cost estimation The total monthly labor cost is given by: Cmo ¼ j X njOj ð3Þ where: nj Minimum required number of type j workers. Oj Montly cost of type j worker. The minimum required number of type j workers is computed as the smallest integer larger than the ratio between the sum of all times required to develop all the basic activities done by type j worker and the total available time of type j worker in a shift. nj ¼ k P i P f ikDtijk Tj ð4Þ where: Dtijk Time of activity i done by type j worker per unit (pallet) of commodity k. Tj Monthly available time per shift time of worker type j. fik Number of monthly pallets k that are handled during activity i. Tj has into account the supplements for labor delays, slowness for tiredness and other contingent delays. Besides, to compute Dtijk working rhythm valuations are considered (Garcı´a Criollo, 1998). When the activity is a transportation activity, its time is given by: Dtijk ¼ dijk=Vijk ð5Þ where: Vijk Velocity of type j worker moving pallet k during activity i. It depends on the equipment used to transport the pallet (e.g. none, forklift, lift truck, etc.). dik Distance to be traveled during activity i moving type k pallet. Simplification of labor costs model. To make the labor costs model described above feasible, the following simplifications and/or assumptions were done: . The time required to do a basic activity is independent of the commodity being handled. This means that the velocity does not depend on the weight being Large capacity warehouses 263 moved. Also it means the worker adjusts his pulling effort to his walking rhythm. This is: Vijk ¼ Vij ð6Þ Dtijk ¼ Dtij ð7Þ . The traveled distance during activity i moving pallet k can be expressed as an average distance that is independent of the commodity being moved. That is: dik ¼ di ð8Þ However, to make this assumption valid, other considerations to compute di are required as it is explained below. Average distances estimation. Figure 3 illustrates how the distance traveled by any pallet inside the warehouse depends on the chosen docks at reception and shipment areas (probabilistic events) and the assigned location for the pallet at the storage and picking areas (deterministic events). The model assumes that these locations are fixed. Follow-up distances will refer to Euclidian distances. For deterministic activities as transportation between storage and picking areas, the average distance is given by: dap ¼ i X f idapi= i X f i ð9Þ where: dap Average traveled distance by all commodities between storage and picking areas. Figure 3. Computation of average distances traveled by a commodity unit within a warehouse F 25,7/8 264 fi Average monthly flow of type i pallets. dapi Distance between geometric centroides (follow-up just centroides) of the fixed assigned spaces for commodity i in storage and picking areas. For probabilistic transportation activities, such as transportation between reception and storage area, the distance is given by equation (10), where: dda Average traveled distance by all commodities between reception and storage areas. fij Average monthly flow of type i pallets entering to the warehouse through dock j. dcj Distance between dock j and the centroides at reception area. dcai Distance between reception area and the centroides of type i pallet at storage area. dda ¼ i P f iddai i P f i ¼ i P j P f ijdcj i P f i þ i P f idcai i P f i ð10Þ When dock assignment at reception area is random between the available docks, it can be shown by using the Monte Carlo Simulation technique (Law and Kelton, 1982, Huertas, 1997) that the first term of the right hand side of equation (10) is zero. Thus, the average distance between reception docks and centroides at storage area dda is equivalent to the flow-weighted average distances from the reception docks and the centroides at storage area of each commodity. The same reasoning justifies the use of centroides at the storage area when computing average distances. Following this methodology, the average distance between picking and shipment areas is given by: dpe ¼ dpp þ dpq þ dqe þ dep ð11Þ where: dpe Average traveled distance by all commodities from picking area to shipment area. dpp Average traveled distance between commodities at picking zone. dpq Average traveled distance by commodities from their centroides at picking are a to packing centroides. dqe Average traveled distance from packing centroides to shipment docks centroides. dep Average traveled distance between picking and shipment areas. Model validation To evaluate the model applicability, it was implemented in a 10,700 m2 warehouse with a capacity of 17,000 pallets of more than 170 wheat-based commodities. Large capacity warehouses 265 A time study was developed during three months. It was found that the warehouse receives on average 600 ton of commodities daily. Trucks, usually loaded with a single commodity, are unloaded at the receiving area where there are 11 docks. Figures 4(a) and (b) show the trucks arrival frequency at reception and shipment areas. As illustrated in Figure 5 (alternative 1), commodities are stored and are moved to a picking area closer to shipment docks, where workers pick the commodities according to the customer orders. The shipment area has 12 docks, which are randomly chosen, as it can be seen in Figure 4(c). Figure 4(d) shows that the average loading time is 1.5 hours. The warehouse works continuously and delivers orders within 72 hours after the order is placed. The study also reported the average percentage of pallets going to the packing zone and the average number of commodities that are packed in a pallet. Finally, the average distances were estimated. For the picking function, where it is possible to handle smaller units than a pallet, it was used three-months historic data, with the assumption that the orders are picked sequentially as they appear in the customer order. The model was applied to the current layout of the warehouse three is illustrated in Figure 5, alternative 1. The output data was compared with current operational data. Results are shown in Table III, rows 1 and 2. The under estimation of the current situation is explained because the model considers only the basic activities of a warehouse, and no correction factor is included (e.g. for rework). It was found that: . any layout can be represented in the model through average distances; . resource estimations per unit allow any operation level; Figure 4. Time study results used to validate the proposed model F 25,7/8 266 . the model considers all the basic activities in a warehouse. It allows to state the times per cycle that each activity is to be done, according to the specific warehouse process; and . the variety of input parameters allow sensitivity analysis in the estimation of resources, time and operational costs. Therefore, it can be concluded that the model is flexible and easy to be implemented. Layouts comparison The model was applied to analyze the convenience of changing the current configuration of the warehouse described before to another one requiring less equipment utilization (specifically lift trucks). Since the warehouse is very long and narrow, alternative 1 (current layout) splits the place into two areas (storage area and picking area) for making the picking activity faster. Alternative 2 eliminates the picking area in exchange of a larger time assembling orders Alternativ