CASE STUDY OF LAPTOPS IN LIMA - Amazon S3

Proceedings Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium/ 2 - 6 October 2017 S. Margherita di Pula, Cagliari, Italy / © 2017 by CISA Publisher, Italy

REDUCING E-WASTE FLOWS BY EXTENDING LIFESPAN THROUGH REPAIR PRACTICES: CASE STUDY OF LAPTOPS IN LIMA M. GUSUKUMA*, R. KAHHAT*+ AND, J. LEPAWSKY**, AND E. ARAUJO** *Department of Engineering, Pontificia Universidad Católica del Perú. Av. Universitaria 1801, San Miguel. Lima 32, Perú. ** Department of Geography, Memorial University of Newfoundland St. John's, NL, Canada + corresponding author, email: [email protected]

SUMMARY: Repairing ICT equipments such as personal computers (PC) and laptops helps save energy and reduce greenhouse gases by extending their life cycle. Formal importers are not willing to do that in Peru since they prefer selling new equipment instead of reaconditioning the all ones. Consequently, this activity is run by formal and informal sectors distributed in clusters in the city of Lima. The main objective of this research is to study how repairing practices of information and communcations technologies (ICT) equipment, principally laptops, contributes to GHG emissions reduction by extending their lifespan vs. the alternative scenario of replacing the inoperative laptop by a new one. Additional socioeconomic benefits are also discussed. Because lack of data is the most difficult task for estimating GHG reduccions, a stochastic simulation based on Monte Carlo method was used, and the expected amount would be equivalent to 1.79 Gg CO2e.

1. INTRODUCTION

Laptop computers, as well as other electronic devices, have the chance to lengthen their life cycle by several processes. Whereas reconditioning only improves external part of devices, remanufacturing can return equipment to full functionality (Nnorom & Osibajo, 2010; Gutowsky et al; Fegade et al, 2015; 2011 Fatimah & Biswas, 2016), there are other situations in which reusing, recovering operative components from damage laptops, or recycle the complete laptop to return materials to manufacturing process are the only options (Nnorom & Osibajo, 2010; Fegade et al, 2015). On the other hand, Waste Electrical and Electronic Equipment (WEEE or e-waste) is a special kind of waste due to their particular composition. For instance, they contain not only dangerous substances such as mercury and lead, but also contain precious metals such as gold and silver. Another of their characterics is the energy consumed during their manufacturing phase is bigger than the energy used during their use phase due to higher energy intensity of manufacturing and rapid turnover (Williams, 2004). To determine whether lengthening the life cycle of a laptop is feasible, three conditions must be taken into

Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017

consideration: the equipment should have a residual value, there may exist the will of the manufacturer to recover it, and how expensive would be costs related with repairing a laptop. Fortunately, many electronic equipment, including laptops, meet these three conditions that may foster a market for this kind of service.

Remanufacturing laptops has several advantages to the environment and society (Fang, et

al, 2016): there are a reduction in material and energy consumption, as well as a reduction in solid waste generation (Gutowsky et al, 2011; Kakufu et al, 2015). From a social perspective, remanufacturing makes ICT products more accessible and brings jobs directly and indirectly through the supply chain (Fatimah & Biswas, 2016). Since income is an important factor of purchase, the coexistences of a new laptop and a second-hand laptop market is recommended (Ferrer, 1997). In addition, to understand impact of repairing computers and other appliances in Peruvian economy, according to INEI, agency in charge of statistical information, this sector is equivalent to 0.33% of Gross Domestic Product (GDP), and having his own category in Peruvian System of National Accounts (SNA) (INEI, 2016).

Figure 01: Principal Clusters for Repairing PCs, Laptops and other ICT devices

Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017 However, remanufacturing also has limitations: manufacturers may not be eager to perform

this activity. Although a third-party company would be in charge of remanufacturing laptops, they may deploy some actions that make remanufacturing more difficult. Examples include turning into inoperative equipment due to the lack of spare parts, planned obsolescence by small changes in design and giving used products as part of payment when a product is purchased (Nnorom & Osibajo, 2010; Gutowsky et al, 2011). Third-party companies must consider fluctuations between supply and demand as well as changes in labor intensity and the variety of products. Consequently, such operations are forced to design flexible processes for adapting to these fluctuations (Gutowskyet al, 2011). Finally, from consumers’ point of view, there is a perceived quality gap between a new product and a remanufactured one. Quality of a remanufactured laptop might be evaluated in terms of lifespan, features, performance and serviceability (Hazen et al, 2016). Thus, remanufacturing could be a valid option in developing countries where price is an important driver to buy a product.

Figure 02: Process of repairing a laptop, Leticia Street cluster

Remanufacturing in Peru has two options: Most of manufacturers of laptops offer technical

support whenever their products are under their period of warranty. Nonetheless, their service is considered expensive by some lower-income consumers, especially if a laptop is out of the period of warranty which is usually one year. On the other hand, an alternative market for repairing and selling second-hand laptops in Lima exists, and it doesn´t have usually a formal relation with laptop producers or importers. Clusters for repairing laptops are mostly concentrated in Downtown of Lima, and people from many part of Peru could come to theses places since they know they may find a solution for their problems. Each of these clusters has grown spontaneously and each offers several services such as repairing PCs, laptops, tablets, cellphones and other ICT devices, software troubleshooting, and selling spare parts and accessories. In addition, these clusters are places in which broken equipment that is not


possible to repair might turn into spare part for other jobs. Figure 01 shows the location of the most important clusters for repairing PCs, laptops and other ICT devices in Lima.

Understanding motivations of actors in laptop repairing process helps identify the most

desirable path from an environmental perspective (Fitzpatrick et al, 2014). Repairing a laptop in alternative markets is a process in which three actors participate: a customer, who has a laptop to be repaired, a repairperson, who diagnoses and repair laptop, and a collector who can have the spare part necessary for fixing the laptop. There is a possibility that the repairperson and collector may be the same person. Knowledge of electronics varies considerably in this sector. Some people have formal training in electrical engineering. Others have no formal training at all. Those in the latter situation may learn on-the-fly by trial-and-error methods, by formally or informaly apprenticing with more experienced technicians, and/or through self-teaching methods such as accessing videos in internet. Figure 02 shows the process of repairing a laptop in Leticia, which is one of the most important clusters for repairing devices in Lima.

Repairing PCs, laptops and other ICT devices in Peru is an activity that offers job positions

locally within Lima and avoids exporting the economic value of these devices for recycling in other countries like China, Germany, Sweden and Belgium (Kahhat & Williams, 2009; Kahhat & Williams, 2012). Repair also reduces environmental impacts by lengthening the life cycle of a device versus replacing it with a new one. Consequently, the main objective of this research is to estimate the reduction of greenhouse house emissions measured as Global Warming Potential in a scenario that would otherwise have scarce data. In addition, this research analizes how repairing processes, perfomed by formal and informal third-party sector in Lima, contribute to the reduction of greenhouse gases emission, and examine this activity in terms of the employment prospects that need some special knowledge and skills. To achieve these objectives, the study identified the principal clusters of repairing applicances in Lima and made an in-depth qualitative and quantiative assessment of one of these clusters to understand how the process of reparing a laptop is performed and finally, to estimate the amount of greenhouse gases, principally carbon dioxide (CO2) is reduced as result of the practice.

2. METHODOLOGY

First stage of this research consists in field work to identifiy principal clusters in Lima, number of stores their kind of services provided in these clusters, as well as understand in situ how a laptop can be repaired, so approximately 27 qualitative interviews were also conducted with repair technicians in one cluster of such businesses in Lima. These interviews inform our qualitative understanding of the sector, but our focus in this paper is the quantitative measurements using LCA.

Life Cycle Assesment (LCA) is a holistic environmental assessment tool that studies the impacts of a product or service through its life cycle (EEA, 1997; Matthews et al, 2015). The term life cycle refers to all activities that a product undergoes, from raw material extration, manufacturing and use stage and its end-of-life (EPA, 2006). Moreover, LCA compiles the inputs, outputs and environmental impacts of a product through one or more stages, from raw material extraction to final disposal, and it is useful to identify opportunities to improve environmental performance of a product, and to present environmental information to decision-makers (ISO, 2006).

Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017 As mentioned before, lack of data would be the main obstacle whenever a LCA is elaborated,

so application of Monte Carlo method may be suitable to deal with this limitation. Monte Carlo method is a numerical solution of a problem in which an attemp to understand the nature of system is performed by simulation of dynamics of its elements (Bielajew, 2001). This method provides approximate solutions for problems performing statistical simulations expressed as random number sequences (Pengelly, 2002). Thus, solution of a problem solved by Monte Carlo method could be expressed in some cases as a probabilistic distribution due to its stochastic nature instead of being shown as a single value as I happened in deterministic problems. Monte Carlo method is used in this research to simulate several cases in which all variables change according to a statistical distribution when a sensibility analysis is too complex to be performed due to the amount and complexity of variables.

Table 01. Global Warming Potential (GWP100), in kg CO2e/unit

Item Value unit Min Max Mean

Laptop 144 159 151.5 piece DVDROM 5.79 5.87 5.83 piece Hard Disk Drive 4.17 4.26 4.22 piece Integrated circuit, logic type 785 783 784 kg Integrated circuit, memory type 606 606 606 kg Wafer, fabricated, for integrated circuit 3.40E+04 3.56E+04 34800 m2 Printed wiring board, mainboard, laptop 276 278 277 kg Table 02. Dimensions of spare parts of laptops (Kasulatis et. Al, 2015)


Printed Circuit Board 0.270 0.393 0.329 kg

Wafer, in mainboard 3.90E-04 6.06E-04 5.04E-04 m2

Wafer, in processor 8.00E-05 1.12E-04 1.70E-04 m2 Integrated circuit, memory type 7.50E-03 7.65E-03 7.80E-03 kg Table 03. Data collected from Leticia St. Cluster


Total number of stores 179 stores stores dedicated to repairing laptops 51 stores

Number of days worked 252 days Laptops repaired 0 4 2 kg % repairs without changing spareparts 80 % % replacing mainboard 10 % % replacing DVDROM 6 % % replacing Hard Disk Drive 4 %

In order to to obtain an expected value of an environmental impact, data needed is get from the following sources: Global Warming Potential, 100-years time horizon (GWP100), are extracted from Ecoinvent 3, (Pré, 2016), and supplementary information of the paper of


Kasulaitis and others (Kasulaitis et al, 2015) gives information about laptop components characterization, whether there might be express in mass or on area.

There are some limitations to this research. For instance, the determination of environmental

impacts is limited to repairing process, excluding transport, use and end-of-life stages, so emissions of GHG in this activity are concentrated in the amount of GHG emitted during manufacturing of laptop components like motherboards, DVDROM and HDD. Because of lack of data, main sources were limited to searching in journals and other databases. Economic issues related to laptop repairing, such as the maximum amount of money a customer is willing to pay and the price difference between the latter and buying a new one, are ignored. The decision tree assumes that only one component of a laptop must be replaced per each reparing service. If more components require replacement, repairing would be so expensive that a consumer would prefer not to repair the laptop since it is possible to get a new laptop in Peru for US$300, and a used one for US$100 depending on the model and equipment conditions. This research is focused on the installed capacity of one of the clusters in Lima, so instead of offerring a single number of reduction emissions, it is offers a probability distribution that shows expected value (mean) and uncertainty (standard desvation), and the fact that a laptop would be operative but obsolete due to software issues are also ignored. Finally, this research is focused on GHG emissions and excludes other environmental impact indicators such as ozone depletion potential, human toxicity, or other possible negative environmental impacts.

Figure 03: Decision tree for repairing laptop process

To estimate environmental impacts, the following steps are followed to perform a simulation based on Monte Carlo method: § Collecting data: Developing the probability distributions. Since Monte Carlo is a stochastic

method, probability distribution of variables is necessary. This step depends strongly on the amount of data collected.

§ Building an equation about the “topic of interest”: This research is focused on determining reduction of environmental impacts, so a formula, which uses variables and their probability distributions mentioned above is used. The environmental impact for a single laptop, expressed as the amount of GHG emitted during manufacturing stage of laptops or their components, can be calculate as a dot product (Kahhat, Poduri, & Williams, 2011):


GWP100 = GHG emitted = BOA · (GWP100 / attribute) (1) Where:

o BOA (Bill of attributes): vector of product attributes (mass, area, piece and other information) of a laptop or its components.

o GWP100 / attribute: Vector of Global Warming Potential, 100-year horizon, expressed in kilograms of CO2 equivalent per attribute of a laptop or its components.

§ Reduction of the environment impact for a single laptop is calculated with the following equation:

ΔGWP100 = GWP100,laptop,new – (GWP100,repairing + 10% GWP100,MB

+ 6% GWP100,DVD + 4% GWP100,HD) (2)

Where: o GPW100,laptop: GHG emitted for manufacturing a laptop, kg CO2e o GPW100,repairing: GHG emitted for repairing tasks, kg CO2e (≈ 0) o GPW100,MB: GHG emitted for manufacturing a mainboard, kg CO2e o GPW100,DVD: GHG emitted for manufacturing a DVDROM, kg CO2e o GPW100,HD: GHG emitted for manufacturing a Hard Disk Drive, kg CO2e

Figure 04: Example of Monte Carlo for estimating GHG emissions of a Mainboard

§ Reduction of the environment impact per year at a specific cluster might be estimated with

this formula: ΔGWP100,cluster = ΔGWP100 · S· L· d (3)

Where:

o S: number of stores placed in the cluster of interest (Leticia) = 51


o d: number of days worked per year = 252 o L: average number of laptops repaired per day

§ Decision tree in figure 03 shows how the amount of GHG, expressed as Global Warming

Potential is calculates: A decision node, represented by a square, indicates that a decision must be take. On the other hand, a chance node, represented as a circle, indicates that some random event will happend, so the value that represents this node is the expected value is calculated as follows:

E[X] = Σ GWP100,i · pi (4)

Where is the probability an event occurs.

Figure 05: Estimating GHG reduction by Monte Carlo method for repairing a single laptop

§ Running the “roulette”. After getting the equation of interest, the process is quite simple, but

require several calculations: for every variable, a random number between 0 and 1 is generated, and then this number is used to calculated the inverse probability distributions. A huge number of runs is recommended to get better results. Accumulated triangular probability distributions are used due to lack of data. Figure 04 shows an example of a “roulette” to estimate GWP100 of a mainboard:

§ Processing results. The advantage of working with accumulated probability distribution is

there is no necessity to generate histograms. Instead, set of results are rearranged, from minimum to maximum. After that, parameters of position (mean) and dispesion (standard deviation), can be found.

3. RESULTS AND DISCUSSION

Figure 06 shows the accumulated probability distribution of the reduction of GHG, expressed in kg CO2e. Expected value of emissions reduction for a single laptop is [139.6 ± 6.27] kg


CO2e, 95% confidence interval. On the other hand, figure 07 shows the total reduction of GHG by the repair work done in the cluster of Leticia St. Expected value is [3.60 ± 3.05] Gg CO2e, 95% confidence interval. Due to the amount of simulations performed (n = 5000), only selected points are visible in figures 06 and 07 to be easier to see simulations (hollow blue circles) and the accumulated distribution (red continuous line). To understand magnitude of emission reduction, it is interesting to compared with some results of 2012 Greenhouse Gas Inventory of Peru1: expected emissions reduction can be compared with GHG emissions due to only manufacturing processes, excluding energy consumption, of ammonia (2.39 Gg CO2e) and lead (1.79 Gg CO2e). Moreover, it is also equivalente to 0.54% and 014% of GHG emissions for energy consumption of public (663.94 Gg CO2e), and business and residential sectors (2440.17 Gg CO2e) (MINAM, 2016).

Figure 06: GHG reduction for repairing a single laptop, n = 5000 µ = 139.6 kg CO2e and σ = 3.2 kg CO2e

Figure 07: Annual GHG reduction for repairing laptops in Leticia St., n = 5000 µ = 3.60 Gg CO2e and σ = 1.56 Gg CO2e

1 According to 2012 GHG inventory of Peru, emissions are equivalent to 171,309.57 Gg CO2e (MINAM, 2016)


4. CONCLUSIONS

Repairing laptops, and other ICT devices offers job positions, and contributes to reductions in the carbon footprint in Peru, and are equivalent to 3.60 Gg CO2e. Although it is fewer than emissions generated by energy consumption of public and business and residencial users, results were obtained for a specific activity, repairing laptops in one of the clusters, so it would be expected if other devices were included, as well as other clusters too, GHG emissions reduction would be greater than results obtained. This model could be improved as well as the data available increases, and using Material Flow Analysis (MFA) of laptops to get a better estimate of GHG emissions reduction. However, methodology applied maight be useful to estimated emissions and emissions reduction in other economic activities. Next topics to be investigated in next future might be related with extending lifespan of laptops by changing their current operative system (OS) for an open-source alternative, since most of them turn into obsolete due to software issues, and how deep can be the social impact of using Free and Open Source Software (FOSS) in repaired computers to increase access to ICT for people living in low income sectors.

AKNOWLEDGEMENTS

This research was supported by the Social Sciences and Humanities Research Council of Canada, Grant #435-2012-0673, and a special thank to all people: repairpersons, salesclerks, collectors and recyclers for agreen to be interviewed for this reseach.

REFERENCES

Bielajew, A. (2001). Fundamentals of the Monte Carlo method for neutral and charged particle transport. Retrieved from http://www-personal.umich.edu/~bielajew/MCBook/book.pdf EEA (European Environmental Agency), (1997). Life Cycle Assessment. A guide to approaches, experiences and information sources. Retrieved from http://www.eea.europa.eu/ publications/GH-07-97-595-EN-C EPA, US Environmental Protection Agency (2006). Life Cycle Assessment: Principles and Practices. Fang, H., Ong, S., & Nee, A. (2016). An integrated approach for product remanufacturing assessment and planning. Procedia CIRP 40, 262-267. Fatimah, Y. A., & Biswas, W. K. (2016). Sustainability Assessment of Remanufactured Computers. Procedia CIRP (40), 150-155. Fegade, V., Shirivatsava, R., & Kale, A. (2015). Design for Remanufacturing: Methods and their Approaches. Materials Today: Proceedings (2), 1849-1858. Ferrer, G. (1997). The economics of personal computer remaufacturing. RESOUR CONSERV RECY (21), 79-108. Fitzpatrick, P., Hickey, S., Schischke, K., & Maher, P. (2014). Sustainable life cycle engineering of an integrated desktop PC; a small to medium enterprise perspective. J CLEAN PROD. (74), 155-160. Gutowsky, T., Sahni, S., Boustani, A., & Graves, S. (2011). Remanufacturing and Energy Savings. ENVIRON SCI TECHNOL (45), 4540-4547. Hazen, B., Boone, C., Wang, Y., & Khor, K. (2016). Perceived quality of remanufactured products: construct and measure development. J CLEAN PROD (in press). INEI, I. N. (2014). Cuentas Nacionales 2007. Año Base 2007.


INEI, I. N. (2016). Cuadros de Oferta y Utilización, a Precios de Comprados (2007-2014). ISO, I. S. (2006). ISO 14044:2006. Retrieved from https://www.iso.org/obp/ui/#iso:std:iso: 14044:ed-1:v1:en Kahhat, R., & Williams, E. (2009). Product or waste? Importation and end-of-life processing of computers in Peru. ENVIRON SCI TECHNOL, 43(15), 6010-6016. Kahhat, R., & Williams, E. (2012). Materials flow analysis of e-waste: Domestic flows and exports of used computers from the United States. RESOUR CONSERV RECY, 67, 67-74. Kahhat, R., Poduri, S., & Williams, E. (2011). Bill of Attributes (BOA) in Life Cycle Modeling of Laptop Computers: Results and Trends from Disassembly Studies. Kakufu, J., Mat Saman, M., Yusof, S., Sharif, S., & Zakuan, N. (2015). Investment Decision Issues from Remanufacturing System Perspective: Literature Review and Further Research. Procedia CIRP (26), 589-594. Kasulaitis, B., Babbitt, C., Kahhat, R., Williams, E., & Ryen, E. (2015). Evolving materials, attributes, and functionality in consumer electronics: Case study of laptop computers. RESOUR CONSERV RECY, 100, 1-10. Matthews, H., Hendrikson, C., & Matthews, D. (2015). Life Cycle Assessment: Quantitative Approaches for Decisions that Matter. MINAM, Ministry of Environment (Peru) (2016). Inventario Nacional de Gases de Efecto Invernadero 2012 . Retrieved from: http://infocarbono.minam.gob.pe/annios-inventarios-nacionales-gei/ingei-2012/ Nnorom, I., & Osibajo, O. (2010). Overview of Prospects in Adopting Remanufacturing of End-of-Life Electronic Products in the Developing Countries. IJIMT, 1 (3), 328-338. Pengelly, J. (2002). Monte Carlo methods. Retrieved from http://www.cs.otago.ac.nz/cosc453/ student_tutorials/monte_carlo.pdf Pré. (2016). SIMAPRO [computer software] (Version 8.1.1.16). Williams, E. (2004). Energy Intensity of Computer Manufacturing: Hybrid Assessment Combining Process and Economic Input-Output Method. ENVIRON SCI TECHNOL (38), 6166-6174.

CASE STUDY OF LAPTOPS IN LIMA - Amazon S3

Documents