State of the art in life cycle assessment of laptops and remaining challenges on the component level: The case of integrated circuits

Ran Liu¹, Siddharth Prakash¹, Karsten Schischke², Lutz Stobbe²

¹Oeko-Institut, Germany; ²Fraunhofer Institut, Germany

The current hype for Product Carbon Footprints of IT, policy initiatives, such as the European ecodesign directive and several research activities recently lead to the publication of numerous LCA results of laptop computers. A comparison of the various studies unveils a broad variance among the results, which cannot be explained solely by technical differences: LCA for IT products still faces severe shortcomings and methodological uncertainties due to the complexity of the products, and assumptions to be made. The paper provides a comparison of latest studies and in particular points out, that the methodology used for the policy assessments under the ecodesign directive dramatically underestimates the manufacturing related global warming potential compared to results published by brandname OEMs.

In this context, a research project, commissioned by the Federal Environment Agency of Germany (UBA), on creating a dataset on notebooks, was initiated in March 2010. This paper documents the initial challenges of this research project in terms of identifying a consistent methodological approach: This paper addresses the component level in the case of integrated circuits (ICs) to reveal the challenges in the data collection, since ICs are the basic components in all electronical applications, but also ICs are one of the main contributors to the GWP. One key aspect in this regard is a clear definition of a reference unit along the life cycle data of ICs. However, a wide range of reference units are applied within the data on ICs, and hence, hindering the transferability and comparability.

The paper discusses the various approaches under consideration as reference units for IC-datasets, in order to contribute towards creating a harmonised reference unit of IC dataset. Chosing the right reference unit depends not only on what corresponds best with given technical specifics, but needs to consider also the intended use of any such IC data model: Some highly confidential data might make it into the parameterised model, if a manufacturer is supposed to provide PCF or LCA data as a “black box” model, but is rather not useful, if an LCA is solely based on reverse engineering where the key parameters of the model, such as potentially yield and number of mask layers, is not known to the LCA practitioner. Pros and cons of the various methodological approaches and potential reference units will be discussed to provide a sound basis to chose the right approach for the intended use.

Product carbon footprint (PCF) assessment of a Dell OptiPlex 780 Desktop – Results and recommendations

Markus Stutz

Dell, Germany

Purpose

Dell recognizes that climate change is real and must be mitigated, and we support efforts to reduce global greenhouse gas (GHG) emissions to levels guided by evolving science. We are also committed to reducing GHG emissions beyond our own operations.

To do this, we have adopted a strategy that takes into account the GHG impacts of our products and our suppliers. We look at each stage of the product life cycle — from developing, designing and sourcing through manufacturing and operations, order fulfilment, customer use and product recovery.

By assessing the carbon footprint of a desktop, we are able to identify areas for improvement to reduce overall GHG emissions and also help customers do the same.

Results

In research conducted in 2010, Dell determined the carbon footprint of the OptiPlex 780 Mini Tower, a typical high-volume, mainstream business desktop that is representative of a range of similar desktop products.

The total carbon footprint of a Dell OptiPlex 780 desktop is approximately 800kg CO2eq when used in the US, 720kg CO2eq when used in Europe and 1230kg CO2eq when used in Australia. The main reason for the differences between the three scenarios is the amount of emissions associated with the differing power generation modes in the three regions, although transports in the assembly and distribution chain also account for a significant part of the differences.

Conclusions

The total PCF for three target markets and a life time of four years has been determined. It has been compared with other carbon footprints, e.g. from driving or from food. The distribution between manufacturing and use has been assessed to determine where the focus on environmental improvement needs to be, manufacturing or use. Further the key components in terms of impact in the manufacturing phase are studied. Also the impact of logistics, specifically the role of air transport has been assessed.

LCM of metals supply to electronics: Tracking and tracing "conflict minerals"

Steven B. Young, Goretty Dias

University of Waterloo, Canada

The supply-chain performance of metals used in electronic products is a living laboratory for life cycle management (LCM). The electronics and metals sectors, as well as various governments and civil society actors, are performing widespread socio-environmental assessments and influencing the practices of miners, traders, smelters and others involved at various stages of the metal supply chain. The purpose of our study is to use a life cycle management (LCM) framework to analyse the effectiveness of sustainability-driven activities in this significant multi-sector global supply-chain.

Ten years ago human rights and environmental concerns were raised about coltan mining in the Democratic Republic of Congo (DRC).NGO groups have raised these and similar concerns, and have used various tactics to reach industry, government and consumers. Their efforts have led to several actions. Over the past two years, two industry associations, the Global e-Sustainability Initiative (GeSI) and the Electronic Industry Citizenship Coalition (EICC), have studied how brand-name multinational companies can address global socio-environmental performance associated with mining of metals used in electronics end-products. In 2011 governments are legislating to control the supply-chain of unsustainable “conflict minerals”, with particular focus on human rights issues in the DRC. The US, for example, is putting into effect a law that “will require reporting on all products containing tin, gold, cobalt and tantalum.”

The Life Cycle Initiative UNEP/SETAC LCM framework, including its definition and guidance, is employed in our study. We analyze activities in the metals/electronics chain and identify areas for continual improvement via supplier selection and management. Several results emerged: companies are undertaking “choice editing” via the “removal of unsustainable goods and services” from their supply-chain while civil society groups are “choice influencing”, as they run “marketing and awareness-raising campaigns to enable and encourage customers and consumers to choose and use goods and services” (UNEP/SETAC 2009) that are more sustainable.

We also present science-based analysis on the distinction between tracing vs. tracking of materials. Metal/electronics chains can exhibit at least nine stages. Tracking follows a unit of material downstream whereas tracing looks upstream. Given that different materials need different LCM strategies, we conclude that cellular materials (e.g., cotton) are more easily traced in commodity chains compared to minerals and metals, thus complicating the metals/electronics chain. Opportunities, pros and cons of tagging, verifying, reporting and labelling metals supplies are considered. We further conclude that the US legislation is a flawed scheme for LCM.

A product attribute to impact algorithm to streamline ICT environmental footprinting using a triage approach

Elsa Olivetti¹, Melissa L. Zgola¹, Christopher Weber², Sarah Boyd³, Eric Williams⁴, Jeremy Gregory¹, Randolph E. Kirchain¹

¹Massachusetts Institute of Technology, United States of America; ²Carnegie Mellon University, United States of America; ³University of California at Berkeley, United States of America; ⁴Arizona State University, United States of America

While cost, quality and performance remain the primary drivers for product decision making within information technology (IT) firms, changing market dynamics stemming from volatile energy prices, pressure from consumers and private groups, and rapidly developing labeling efforts are raising firm focus around quantifying environmental performance as well. However, executing quantitative measurement and identifying primary drivers of environmental impact present particular challenges within this industry due to the complexity, and dynamics of products and supply chains. This presentation describes the activities of an ongoing partnership among academia, IT companies, as well as governmental and non-governmental organizations convened to develop an efficient approach to life cycle assessment of IT. The project’s first phase aims to develop a near term, quantitative approach for environmental performance evaluation to support strategic decision-making and labeling. The approach initially intends to resolve the global warming impact among product types (as opposed to SKU-level) for laptops. We aim to determine impact with minimum data collection and minimum user input while providing actionable insight. Data collection is minimized through probalistic triage, leveraging existing data and its associated uncertainty to estimate impact and then targeting data refinement on the highest impact activities. To minimize user input, product attribute to impact algorithms (or PAIAs) are developed linking product attributes to components, components to the bill-of-materials and process (or bill of activities, BOA), and the BOA to impact. This presentation will focus on the methods used to identify and target key drivers, and thereby minimize data collection, through a triaged approach that incorporates uncertainty. The results of the triage are then used to develop algorithms that determine changes in environmental performance based on relevant attributes. This PAIA development provides a form of streamlined LCA because the method enables an input, for example, of screen size rather than a full LCD bill of materials. The relevant attributes we identify are those that have a significant impact on the results, are viewed as critical by stakeholders, and are measureable at “low” cost. Results indicate the feasibility of this triaged approach and that PAIAs effectively limit user input.

Quantifying the life cycle assessment uncertainty in the information and communication technology sector

Fredrik Guldbrandsson, Jens Malmodin, Anna Bondesson

Ericsson AB, Sweden

Conducting a Life Cycle Assessment (LCA) study of Information and Communication Technology (ICT) products and systems is a challenging task. Dealing with complex products and a significant amount of data involves many sources of uncertainty that will affect the accuracy of the results. In this paper identified parameter uncertainties significant for the calculations of the total Ericsson carbon footprint in 2010 are studied. Uncertainties have been estimated on an aggregated level for activities included in the carbon footprint and classical rules of error propagation have been applied to obtain the uncertainty of the total carbon footprint. The parameter uncertainty estimations of Ericsson's life-cycle carbon footprint for the included life-cycle stages indicate a combined uncertainty from around +/-6 percent for Ericsson activities to +/-30 percent for the supply chain with a level of confidence of approximately 95 percent. In addition different scenario and model uncertainties are discussed, uncertainty related to lifetime of the product being a major source of scenario uncertainty.

European LCA standardization of ICT: Equipment, networks and services

Anders Sven Gunnar Andrae

Huawei Technologies Sweden AB, Sweden

Problem addressed:

Both globally and in a European context major standardization bodies like International Telecommunication Union (ITU) and European Telecommunications Standards Institute (ETSI) are pioneers. In recent years carbon footprint and life cycle assessment (LCA) calculations have emerged as an important topic within the Information Communication Technology (ICT) industry. Especially the aim is to promote the probable so-called positive enabling effects of ICT Services. However, several pieces of evidence exist to strengthen the hypothesis that the lack of harmonization and transparency for LCA is a cause of concern in the ICT sector.

Project undertaken:

To handle the LCA challenges several Work Items were started in 2008 and are ongoing in ETSI and ITU with finish in 2011. International Electrical Commission (IEC) is also starting LCA standardization for Electrical Equipment. This paper will describe the core elements of the first edition of the ETSI LCA standard for ICT (DTS/EE-00014). Building entirely on ISO 14040/44, several important stakeholders (Network Operators together with Manufacturers of Network Equipment, End-user Equipment, and Parts) have jointly agreed on such core elements as system boundaries, recommended/optional life cycle phases, unit processes, functional units, life time, allocation methods, data quality evaluation, and cut-off rules. DTS/EE-00014 is especially intended to benefit small and medium sized companies who might not have the knowledge of multinationals in conducting LCA.

DTS/EE-00014 covers all kinds of ICT Equipment and “all” relevant environmental impact categories as well as enabling effects of ICT Services. Consequential LCA is out of scope at the moment. To show the added value of the performed work in relation to ISO 14040/44, DTS/EE-00014 will also be applied to common ICT Equipment and Networks

Results:

DTS/EE-00014 clearly makes the LCA process more streamlined and occasionally faster as the practitioner gets an “exact” guide what is recommended to include if applicable, and what is optional. LCAs done according to DTS/EE-00014 have consistent result presentations. This is shown for common ICT Equipment and Networks.

Conclusions:

Consistent result presentations lead to convenient quality comparisons of different ICT LCAs. Comparisons of absolute values between studies performed and presented by different individuals using different software/databases are beyond the scope of DTS/EE-00014, as such comparisons would require that the assumptions and context of each study are exactly equivalent.