Developing a simplified eco-impact assessment tool to assess the environmental impact of ICT products more efficiently compared to standard life cycle assessment methodologies
by Thomas Okrasinski, Alcatel-Lucent; John Malian, Cisco; and James Arnold, iNEMI
Life cycle assessment (LCA) is becoming a fundamental methodology within a broader sustainability management structure for businesses to use materials, manufacturing processes and energy resources more efficiently. LCA is typically evaluated over four main stages of a product’s existence: manufacturing, transport, usage, and end-of-life treatment. While LCA can provide substantive results when used within its limitations, it is resource-intensive and challenging to apply properly. This is especially true in the information and communication technology (ICT) industry sector where products are complex and technology evolves rapidly.
The International Electronics Manufacturing Initiative (iNEMI) organized the Eco-Impact Evaluator for ICT Equipment Project to develop a methodology framework and simplified estimation tool that could be used to more efficiently assess the environmental impact of an ICT product.
The first phase of this project researched and defined a simplified methodology framework. In Phase 2, the project team is developing an estimator tool based on the framework. This tool is intended to more easily estimate the eco-impact for different types of ICT equipment while still providing sufficient accuracy. A key part of this second phase is the assessment of data needs and development of mechanisms for prioritizing and collecting pertinent data from the supply chain. Supply chain data must stay abreast of the rapid technological advancements within the ICT industry in order for the estimator tool to be useful to the LCA practitioner.
Estimator methodology — the basics
Methods and data are similar for most classes of products. About 90% of parts have common application in ICT product types/classes. iNEMI’s estimator tool is designed to use a “building block” approach to provide LCA-based eco-impact information for ICT components and equipment. It is centered around key parameters of ICT components that contribute significant eco-impact based on past LCA data. The project team plans to establish consensus with major ICT industry constituents to ensure that the tool will have updateable databases via cross-industry information sharing.
The iNEMI-defined methodology categorizes targeted components to provide a unified format for requesting LCA information from suppliers. It defines key elements within ICT product types based on their relative importance in contributing to overall eco-impact; and provides a simple, unified mechanism for evaluating eco-impacts, summarizing results, and communicating information within the industry and requesting information from suppliers. To further simplify its operations, only one eco-impact category — Global Warming Potential (GWP) – 100 years, as measured by greenhouse gas emissions in units of carbon dioxide equivalents (kg CO2e) — was included by the project team in the initial development of the LCA estimator tool.
Classification and categorization of ICT products
ICT products can be classified into distinct categories with common attributes that produce certain levels of eco-impact regarding their component makeup, assembly, usage, and design life. iNEMI defined the following major classification categories for ICT products:
- LAN (local area network) and enterprise telecom
- Service provider telecom
- PCs (personal computers)
- Printers
- Monitors
- Handheld ICT devices
These classifications were then sorted into component categories comprised of similar materials and manufacturing processes. The components were analyzed with regard to their respective contributions to the eco-impacts associated with raw materials extraction and processing, intermediate materials manufacturing, and component / subassembly manufacturing. The intent of categorizing these ICT components was to have a concise list that can be analyzed for common eco-impacting attributes, which can then be normalized and modeled to derive their level of eco-impact within an LCA estimator tool. Table 1 provides the major component / subassembly categories that the project team defined for ICT products.
Modeling manufacturing stage using common component characterization
The key parameters and metrics for assessing the eco-impact of components that comprise ICT products can be summarized for each of the categories listed in Table 1. Such parameters can represent the significant eco-impact contributors (per the defined boundary conditions and cut-off criteria) based on the analyzed datasets – internally available from within the ICT industry (e.g., integrated circuits) and externally available from other industry sectors (e.g., bulk metals and plastics).
An associated algorithm can be determined based on the Life Cycle Impact Assessment (LCIA) data available for the above parameters. For example, a linear regression equation of the following type can be prepared to determine GWP for the component category of printed circuit boards.
GWPPWB = AB [α + (β SF) + (γ BL)]
Where:
- GWPPWB is the total Global Warming Potential (100 years) for the printed wiring boards in the product / asset; expressed in kg CO2e
- AB is the area of the PWB; expressed in square meters
- α is the “intercept” constant for this linear regression equation
- β is the “PWB surface finish type” constant for this linear regression equation
- SF is the PWB surface finish type (e.g., HASL → SF = 1; ENIG → SF = 2)
- γ is the “PWB layer” constant for this linear regression equation
- BL is the number of layers in the PWB
LCIA data is available from databases worldwide. Nearly all of them have data on plastics and metals, which are important for the electromechanical and cabling portions that comprise electronic products. However, a little over half of the databases surveyed have data surrounding electronic components.
The challenge for the ICT industry will be in developing and maintaining databases that can provide global value, while allowing that data to be corroborated with current information from the ICT sector. Updates will need to be made periodically to assure that the databases remain valid and that the functional units for the components and materials in the databases are accurately defined. Finally, the databases should be made publicly available and not include any proprietary information, so that they can be open to external peer review and continual improvement. The European Reference Life Cycle Database is a major step in this direction.
Detailed LCA analyses conducted on ICT products have shown that the component types providing the greatest contribution of environmental impact are the bare printed wiring boards (PWBs) and the large integrated circuits (ICs). The bare PWB composed of the carbon bearing epoxy FR4 and the large integrated circuits such as plastic encapsulated Ball Grid Arrays (BGAs), together make up almost 90% of the manufacturing stage carbon content. Further investigation of the PWB and of the large devices reveal other relevant facts. There is a linear functional dependence between the carbon content per unit area of a PWB and its number of layers. Similarly, for the large ICs, the carbon content has a functional relationship with the number of pins and the IC package type. Algorithms derived from currently available data have been developed and incorporated into iNEMI’s LCA estimator proof-of-concept tool.
A key goal in developing the algorithms is that the end result should be within 15% of the result obtained from more detailed LCA methods for over 90% of the circuit pack assemblies investigated.
Modeling manufacturing stage using equipment parameterization
The equipment parameterization approach maps product characteristics to environmental impact through analysis of generic ICT products. The approach aggregates comparable, relevant data with temporal and spatial uncertainty and variation. The goal of this probabilistic triage is to identify those key drivers of impact so that data collection can be focused on the aspects of a product life cycle that matter and thereby conserve limited resources for data collection. The triage method relies on available data sources, but tries to accurately reflect the associated uncertainty that comes with the use of secondary data.
The modeling performed for equipment parameterization includes statistical contribution analysis to identify hotspots within the carbon footprint (Figure 2). These are based on the most significant contributors to total impact. Quantitative metrics are provided to understand the significance of each phase or module’s contribution to impact.
This is then followed by statistical regression analysis to map attributes to activities and impact. These regressions are based on existing data found in literature, industry data and disassembly data.
Based on elements of the analysis, which at this point are at the product class level only, the modeler determines the desired resolution between product classes. For example, a 13” LCD screen produced in a facility without extensive greenhouse gas emissions abatement equipment will have a higher impact than a 14” LCD screen produced in a facility with extensive abatement equipment, even though the latter employs a larger screen.
Summing the eco-impact for the manufacturing stage
For the manufacturing stage the LCA eco-impacts reflect the total of the ICT components’ manufacturing, transport of components and intermediate materials to final product manufacturing locations, product assembly and testing, and product packaging. Transport of the intermediate materials, components, and subassemblies from their respective manufacturing facilities to the manufacturing facilities for final assembly into finished products includes discrete shipments from a large number of nodes (facilities) to one or more final assembly nodes. Typically, weight of the shipment and distance between the manufacturing nodes are included. Because the intermediates are very low in weight and shipped in bulk, the summation of the total transport venues can be treated as an overall factor applied to the total eco-impact of the product for this LCA stage.
The eco-impact for final product packaging is based on the packaging types used to ship the finished products to its intended distribution facilities and end-use locations. Packaging of intermediate components and materials was excluded in this estimation, as these items are typically packaged in bulk amounts, and the packaging materials can be considered to contribute an insignificant amount to the total LCA manufacturing stage.
Assembly and testing of the intermediate materials, components, and subassemblies into finished products and assets includes processes such as surface mounting technology, thru-hole mounting technology, mounting of ICT product / asset, surface treatment (e.g., painting) for pre-manufactured cabinets, and testing of the ICT product. These parameters are treated as a collective summation of the total assembly and testing processes, and defined as an overall factor applied to the total eco-impact of the product for the manufacturing LCA stage.
Modeling the transport stage
The parameters for assessing the eco-impact of the transport stage include the total shipped product weight, distance shipped and type of transport such as air, marine, truck and rail. The appropriate carbon emissions factors are then applied to these values to derive the carbon emissions for the logistics portion of the ICT finished product.
Also included in the transport life cycle stage is the ICT product’s installation. The total eco-impact associated with the installation of an ICT product is highly dependent on its type. For small ICT devices that are designed for consumer premises (e.g., PCs, printers, IP phones, cable modems), few — if any — ancillary materials, parts, and resources are needed to complete the installation such that the eco-impact from installation can be considered negligible to the total eco-impact of the Transport LCA Stage.
On the other hand, for network servers and telecom products the ancillary materials, parts, and resources necessary to complete an installation at a customer’s premises may be more significant. Typically an assessment of these materials and resources would be needed to further determine the eco-impact related to the specific installation.
Modeling the use stage
The parameters for assessing eco-impact of the use of ICT products can be modeled by knowing the location where the product will be used (by region or country), the power consumption profile (typically a daily average), the availability of the product over the course of a year (subtracts downtime for repair or non-daily use), and the product’s intended operating life.
The product’s power consumption profile should be based on its typical configuration and feature set deployed, and should include on-power, power-save, idle-power and off-power periods. Some government agencies have developed such use profiles for certain ICT product categories, e.g., US EPA Energy Star Program.
The overall power consumption should also include any power needed to cool the equipment. Internal cooling includes fans and heat exchangers within an equipment cabinet or enclosure. External cooling should be included if there is a need to transfer heat, control humidity levels, and cool the surrounding equipment area (e.g., CRAC unit within a central telecom office / server facility).
Modeling the end-of-life stage
The parameters for assessing eco-impact of the end-of-life (EoL) of ICT products include the weights of the constituent materials, the disposition of the product (i.e. the percentage of the product’s materials that are recycled, reused, incinerated with energy recovery, or landfilled). Typically EoL treatment is provided locally (within region), so transport to such treatment, recycling and final disposition facilities can be included within factors developed for this LCA stage. There can be more sophisticated approaches taken to develop EoL models (e.g., European Life Cycle Data System).
Constructing the LCA estimator tool
Based on the modeling analyses for the four major LCA stages of an ICT product outlined in this article, a proof-of-concept LCA estimator tool can be developed. The project team is currently developing and testing the tool with an expected completion of June 2012.
Other aspects of the tool development include prototype testing, comparison of test results with known detailed LCA results (published and/or within the ICT industry), and development of data collection and maintenance guidelines.
Conclusion
The LCA estimator proof-of-concept tool defined in this article can provide the basis upon which LCA practitioners can assess the greenhouse gas (GHG) emissions of ICT products over their full life cycle – manufacturing, transport, use, and end-of-life treatment. The available LCIA databases and information on the GHG emissions of ICT products and components referenced and collected by the iNEMI project team provide a starting point for the development of algorithms that can be employed in the estimator tool. Additional approaches for collecting and developing LCIA data have also been presented.