Sources of Variation in Life Cycle Assessments of Desktop Computers


  • Paul Teehan,

  • Milind Kandlikar

Paul Teehan, Institute for Resources, Environment and Sustainability, University of British Columbia, 2202 Main Mall, Vancouver BC, V6T 1Z4, Canada.


Life cycle assessment (LCA) studies of desktop personal computers (PCs) are analyzed to assess the environmental impact of PCs and to explain inconsistencies and disagreements across existing studies. Impacts, characterized in this work in terms of primary energy demand and global warming potential, are decomposed into inventory components and impact per component in order to expose such inconsistencies. Additional information from related studies, especially regarding use-phase energy consumption, helps interpret the LCA results. The weight of evidence strongly suggests that for primary energy demand and contribution to climate change, the use phase is the dominant life cycle phase; manufacturing impacts are smaller but substantial, and impacts due to product transportation and end-of-life activities are much smaller. Each of the few LCA studies that report manufacturing impacts as being greater than use-phase impacts make unrealistically low assumptions regarding use-phase energy consumption. Estimates of manufacturing impacts, especially those related to printed circuit boards and integrated circuits, are highly uncertain and variable; such estimates are very difficult to evaluate, and more systematic research is needed to reduce these uncertainties. The type of computer analyzed, such as low-power light desktop or high-power workstation, may dominate the total impact; future studies should therefore base their estimates on a large sample to smooth out this variation, or explicitly restrict the analysis to a specific type of computer.


Information and communication technology (ICT) equipment has come under scrutiny in the past decade due to rising concern over their environmental impacts. According to the best recent estimates, the global ICT sector is responsible for about 2% of global anthropogenic greenhouse gas emissions; half of these emissions are due to personal computers (PCs) and peripherals (Climate Group et al. 2008). Quantifying the environmental impacts of PCs is important in order to understand both the aggregate impacts of the industry and the relative share of the impacts from residences and offices that can be attributed to PCs.

Several life cycle assessment (LCA) studies have focused on consumer electronics, including PCs. Unfortunately the results of these studies are not consistent with one another, with significant variation in both the absolute impact reported and the life cycle phase that dominates the impact, which is either the manufacturing or the use phase. This literature has been recently reviewed by several authors, including James and Hopkinson (2009), Malmodin and colleagues (2010), Yao and colleagues (2010), and, most thoroughly, Andrae and Andersen (2010). Each review noted the lack of consensus in the literature, but none attempted to provide a systematic and rigorous exploration of the source of the disagreement or identify which (if any) of the reviewed studies might be the most accurate, a task we undertake here. Focusing specifically on desktop PCs without displays in order to limit variation, we break down the existing LCA studies into life cycle phases, identify the data sources they have used for each phase, and assess the validity of these data sources and any assumptions they have made. This work has a significantly narrower scope than Andrae and Andersen and thus allows us to go in greater depth into understanding the results of existing studies of desktop PC impacts. We also survey associated literature that helps contextualize and interpret these results.

We first summarize the relevant studies in order to show their reported results in terms of total environmental impact, and relative impact by life cycle phase, and identify which components of the desktop PC's life cycle inventory are the largest contributors to the total impact of each life cycle phase. We then answer the following two questions:

  • 1To what extent and for what reasons do estimates from different studies disagree with one another?
  • 2Given the uncertainties in estimates for the environmental impacts of a desktop PC, and in light of the various studies, what are the best guess estimates or ranges for evaluating the impacts of desktop PCs?

Our approach to structuring the analysis and answering these questions is described in the following section.


Meta-analysis of LCA studies is challenging, because studies can differ in terms of unit of analysis, temporal and geographic scope, inventory data, impact factor data, and impact characterization scheme. These differences each add variation to the numerical result and must be accounted for where possible in order to facilitate a valid comparison.

Scope and Unit of Analysis

We have chosen to focus on desktop PCs, excluding other electronic devices like laptops and cell phones, because desktops have received the most attention to date from LCA researchers, and because they represent a relatively large share of the total impact of the ICT sector. Desktop PCs typically include a cathode ray tube (CRT) or liquid crystal display (LCD); peripherals such as a keyboard, mouse, and printer; and a central unit, often called the control unit. We have restricted the scope of our comparison in this review to control units only, excluding displays and peripherals. Displays in particular are known to represent a significant component of the total impact of a desktop PC system, but they are not treated consistently in the literature, with some studies assuming CRT displays, some assuming LCDs, many assuming a mixture of the two, and some excluding displays altogether. An accurate comparison of these studies would have to separate the display and control unit and compare them independently, and while a comparison of LCA estimates of displays would be valuable, we have not performed such an analysis here in order to limit the scope and complexity of this study. As such, when we refer to desktop PCs, we are referring to the control unit only. Desktop PCs come in different sizes and with different features, but most studies aim to measure a PC with characteristics representative of an average or typical product within this category; we use the same approach here.

Geographic scope does vary from study to study; manufacturing impacts usually occur in Southeast Asia, but use-phase impacts depend on the local electricity generation mix. To correct for this source of variation, use-phase impacts are compared in terms of kilowatt-hours (kWh)1 of electricity used rather than endpoint impacts. Temporal scope also varies; we report results according to their year of publication and give a slight preference to more recent studies, as they are more likely to be representative of modern products.

Definition of Environmental Impact

Impact can be measured in many different ways depending on the life cycle impact assessment (LCIA) method chosen by the practitioner. Two categories, global warming potential over 100 years (GWP100, in kilograms carbon dioxide equivalent [kg CO2-eq],2 per the Intergovernmental Panel on Climate Change [IPCC] standard) and cumulative primary energy demand (CED, in megajoules [MJ]),3 were used in a large number of studies. Some categories like ecotoxicity and eutrophication potential were used only in a few studies, and in some cases were measured with different units, making a comparison across studies in these categories much more difficult. Our study therefore reviews reported impact measurements in both GWP and primary energy demand.

Analytical Approach

Our central challenge is judging the quality of the LCA studies and commenting on their reasonableness. Our strategy for doing so is decomposition. We split reported results into (1) component life cycle phases; (2) the largest sources of impacts for each phase (i.e., decomposing the manufacturing phase into a listing of key modules such as mainboard, integrated circuits [ICs], and so on); and (3) an inventory listing of items, in kilograms for material parts and kilowatt-hours for electricity, and a corresponding impact factor, in megajoules per item unit and kilograms CO2 equivalent per item unit. Assumptions in each study are compared at this fine-grained level, highlighting key differences that are obscured in the total result.

In judging reasonableness, we employ a weight-of-evidence approach. It is often very difficult to independently verify reported results of LCAs, but some aspects of the desktop PC life cycle, notably use-phase electricity usage and product lifetime, have been studied outside the body of LCA studies surveyed here. We have surveyed this literature as well, assigned a data quality score to each study, and assumed that reasonably representative values lie toward the middle of the range reported by the highest-quality evidence; comparing against this broader context allows us to roughly judge whether the LCA studies make reasonable assumptions. For parts of the desktop PC life cycle without a large ancillary literature, we are less able to judge reasonableness. Instead, we identify obvious outliers or errors, and suggest that more weight should be given to higher-quality studies, as demonstrated by thorough reporting of methods, data sources, and results. Recognizing the limitations of this approach, we draw only cautious conclusions in such cases. In all cases, data heterogeneity precludes a rigorous statistical aggregation with probability distributions. Instead, we assign reasonable upper and lower bounds on impact measurements.

In answering these questions, we hope to mitigate some of the uncertainty surrounding desktop PC LCAs and provide the reader with the appropriate information and context needed to evaluate such efforts and accurately interpret their results.

Overall Results of Previous Studies

LCA studies of desktop PCs were identified through literature searches of Google Scholar and a number of academic databases, and through citations in other published studies and reviews. We excluded a 1993 study (MCC 1993) due to its age, and one study (Eugster et al. 2007) because the same results were republished in another study (Duan et al. 2009); to our knowledge, all other published LCAs of desktop PCs are included in our review. Details about the scopes and methodologies of the studies used for this work are available in the supporting information available on the journal's Web site. Results of studies included in this synthesis are shown in figure 1 (numerical data tables for all figures are also provided in the supporting information on the journal's Web site), where the left-most graphs show the relative impacts of each phase of the life cycle and the right-most graphs show absolute impacts, where available. Studies that measured impact in terms of cumulative primary energy demand in megajoules are on the top; studies that measured GWP in kilograms CO2 equivalent are on the bottom. Where possible, the results have been adjusted to exclude displays. Only one study, Tekawa and colleagues (1997), did not provide enough information to do so; its results thus include a CRT monitor. All other studies report the desktop PC control unit only.

Figure 1.

Summary of life cycle assessment studies showing the breakdown of life cycle energy use and carbon equivalent impacts. MJ = megajoules; kg CO2-eq = kilograms carbon dioxide equivalent.

The relative impacts on the left side of the figure suggest that the use phase is the dominant component in total impact, according to most studies. The studies measuring primary energy report average use-phase impacts of 62% of the total, with 35% of the total being due to manufacturing and very small portions for distribution and end of life. Studies measuring GWP similarly report an average of 58% for use phase, 39% for manufacturing phase, and very small portions for distribution and end of life.

We will not consider end of life or distribution in this analysis because of their relatively small impacts in terms of megajoules and CO2 equivalent. Small impacts in terms of primary energy and GWP should not, however, be taken as indications that the end-of-life life cycle phase is environmentally benign, as impact categories relevant to toxic releases may show higher scores for the end-of-life phase. In particular, informal recycling in lower-income countries is not captured in LCA databases and is difficult to accurately characterize, but has been shown to have significant negative health impacts on workers and local residents due to environmental contamination; see the work by Sepúlveda and colleagues (2010) for a review on this topic.

The remainder of this article will focus on the manufacturing and use phases, and will attempt to deduce the reasons underlying the disagreements across the studies shown in figure 1. In particular, we seek to understand why studies that report dominant impacts from the manufacturing phase, namely the studies by Choi and colleagues (2006), Masanet and colleagues (2005), and Williams (2004), disagree with the majority of the LCA studies that show the use phase as dominant.

Manufacturing and Production Phases

The manufacturing phase of the life cycle, which includes material extraction and processing, subassembly production, and final assembly, is particularly difficult to analyze because of the large number of highly complex processes involved. The analysis can nonetheless be made tractable by examining the impacts of a relatively small number of components. Figure 2 shows the results of our initial review, in which the impacts have been divided into component categories: mainboard ICs, mainboard other, power supply, hard drives and disk drives, and the metal or plastic casing. Graphics cards and other internal circuit-board-based peripherals are grouped with the mainboard. All additional items are grouped as “other.”

Figure 2.

Manufacturing impacts of desktop control unit components (relative). ICs = integrated circuits.

Our goal, to identify the quantity and impact per quantity unit of each component category, is challenged by a lack of published complete bills of materials or complete listings of impact factors, and by differences in the way the respective authors structure their analyses. For example, some studies consider ICs separately and others lump them in with the mainboard. Nevertheless, as shown in figure 2, it is possible to make observations about relative impacts of component categories: the mainboard including ICs is responsible for the largest impact, accounting for more than 50% of the impacts in all but one study, with the other components constituting small but nonnegligible proportions of the remainder. The proportions vary considerably, as do the absolute totals of the production phase on the right half of figure 1; in order to explore the sources of this variation, in the following sections we identify the quantity and impact of each component category.

Component-Level Impacts

Our approach here is to identify the quantity, in kilograms, of each component category, and its impact, in terms of primary energy (in megajoules per kilogram of component) and GWP (in kilograms CO2 equivalent per kilogram of component). Figure 3 shows the results of our survey. ICs are excluded from this figure because they have very high impacts and negligibly small mass, making a comparison inappropriate; they are examined separately in the following section. Product packaging such as cardboard and polystyrene foam is excluded from the total product mass. Note that only four full inventories for component categories were available: studies by Hikwama (2005), Hischier and colleagues (2007), IVF (2007), and Williams (2004). One of those studies, (Williams 2004), does not provide GWP, and another (Hikwama 2005) provides neither GWP nor primary energy. Few additional studies of these components are available.

Figure 3.

Mass and impacts of desktop personal computer components (excluding mainboard integrated circuits).

The studies roughly agree over the mass of the components, with no obvious outliers, yet they disagree on the impacts per part. Hischier and colleagues' results are consistently much larger than IVF's, roughly by a factor of two. Thus, for apparently similar inventories, the two studies report significantly different final results. The primary energy impacts from Williams vary significantly from both of these studies, but this is likely an artifact of the different way the Williams study was organized, as described in the supporting information on the journal's Web site. With a limited number of data points, no specific conclusions can be drawn except that the impacts may lie in the range shown on these graphs. Ideally we would like to benchmark these results against other external studies, as we do later in this article for use-phase energy consumption, but we are not aware of any such studies focusing on these particular component categories. Fortunately the literature for ICs is slightly richer; we draw upon this literature in the next section.

Integrated Circuits

The LCA studies measure semiconductor content in three ways: area of input silicon wafer, area of finished silicon die, and mass of packaged chip. In the latter case, a mass ratio of finished silicon die relative to the mass of the packaged chip is assumed. It is possible to convert between these units by making a few assumptions; calculations described in the supporting information on the journal's Web site result in a silicon wafer input of roughly 1.67 square centimeters per square centimeter (cm2/cm2) finished die, and a mass of 0.2 grams per square centimeter (g/cm2) finished die.4Table 1 shows the semiconductor content per desktop according to all of the studies that reported it. The area of the table labeled “adjusted inventory” shows our estimates of the total mass and area of finished die given the reported inventories and these calculated conversion factors.

Table 1.  Integrated circuit inventories and impacts per desktop mainboard, with originally reported inventory and adjusted inventory assuming 0.2 g/cm2 finished die
  Williams (2004) Yao et al. (2010) Hischier et al. (2007) Kemna (2005) IVF (2007) Williams et al. (2002) Krishnan et al. (2008) Boyd et al. (2009) Andrae and Anderson (2010)
  1. Note: cm2= square centimeter; Si = silicon; g = gram; MJ = megajoule; kg = kilogram; CO2-eq = carbon dioxide equivalent.

Reported inventory
 Input wafer (cm2/desktop)110        
 Finished die (cm2/desktop) 12       
 Packaged chip, 0.9% Si (g/desktop)  30      
 Packaged chip, 1% Si (g/desktop)    95.5    
 Packaged chip, 2% Si (g/desktop)  59      
 Packaged chip, 5% Si (g/desktop)   100.469    
 IC primary energy (MJ/desktop)1,992991,175 463.6    
 IC global warming (kg/desktop)  72.4 34.8    
Adjusted inventory
 Si mass (g/desktop)    
 Finished die (cm2/desktop)661219.425.122.0    
Impacts per finished die area
 Primary energy (MJ/cm2)30.28.360.6 21.0313381 
 Global warming (kg CO2-eq/cm2)  3.7 1.6  5.57.0

The table also shows the impacts per finished die area, measured in megajoules per square centimeter and kilograms CO2 equivalent per square centimeter. For the five LCA studies, the figure is obtained by dividing the total reported IC impact, where available, by the adjusted finished die area. In addition, three studies were identified that assessed the life cycle impacts of semiconductors alone; these results are included on the right side of the table. The latter, Boyd and colleagues 2009, reports a total of 91 MJ/die and 6 kg CO2-eq/die for all production phases, which corresponds to 81 MJ/cm2 and 5.5 kg CO2-eq/cm2 assuming an average 1.11 cm2 die in a 45 nanometer (nm) process, according to their data. Likewise, an algorithm from Huawei Technologies based on Boyd's work estimates 34.7 kg CO2-eq/g of die (Andrae and Anderson 2010), which would be 7.0 kg CO2-eq/cm2, assuming our estimated conversion rate of 0.2 g/cm2.

Overall, the finished die area ranges from 12 to 66 cm2—though the latter figure was criticized by Yao and colleagues (2010) as being an overestimate—with a median of 22 cm2. We believe that the impact estimate of Yao and colleagues is too low due to an apparent misapplication of Williams's (2004) methods, and that of Hischier and colleagues (2007) is too high due to an apparent error in calculating the die area of a packaged chip (see supporting information on the journal's Web site for details). Discarding these, reported primary impact estimates range from 21 to 81 MJ/cm2, and global warming impact estimates range from 1.6 to 7.0 kg CO2-eq/cm2. The study of Boyd and colleagues (2009), which reported impacts at the upper end of these ranges, is the most thorough and up-to-date LCA of semiconductor manufacturing, which suggests that other studies may be underestimating impacts due to semiconductors. Establishment of standard impact factors and inventory reporting schemes would help a great deal in removing some of this uncertainty.

Total Manufacturing Impact: Analysis

Studies at the level of component categories disagree significantly. For the component categories shown in figure 3, inventories are relatively consistent, indicating that much of the variation is due to different assumptions regarding the impacts of the various components, especially the mainboard and ICs. For ICs, both inventory and impacts were highly variable. Using the data from figure 3, we see a range of total desktop mass from 9.0 to 11 kg, a GWP ranging from 13 to 23 kg CO2-eq/kg desktop, and a primary energy consumption ranging from 180 to 590 MJ/kg desktop.

The data are not sufficient to determine a plausible range for the manufacturing impacts of a desktop PC. We will summarize the data by assuming that both total desktop mass and impacts per desktop mass may fall within the ranges reported in figure 3. Thus by multiplying these ranges together, we estimate that the total GWP can vary from 120 to 250 kg CO2-eq/desktop and the total primary energy consumption varies from 1,600 to 6,500 MJ.

Use Phase

The use phase is responsible for the greatest impact according to a majority of studies, as shown in figure 1. Impacts due to the use phase are simpler to measure and quantify than impacts in the manufacturing phase, as the only impacts are due to electricity consumed by the device during its lifetime. Nevertheless, the LCA studies vary to a surprising degree, with estimates of lifetime primary energy consumption due to the use phase ranging from a low of 580 MJ (Williams 2004) to a high of 16,800 MJ (Kemna et al. 2005), which is a 30-fold variation.

Total use-phase electricity consumption is a function of the power demand of the device, patterns of usage, and the lifetime of the device. Most devices have several different operating modes, such as active, standby, and off, which have different power demands. Using methods described by Kawamoto and colleagues (2001) and Roth and colleagues (2002), given a set of power modes (PM) such that, for mode iPM, the power draw is P(i) and the average time spent in mode i is t(i), the total lifetime energy consumption can be found using the following equation:


where energy consumption is in kilowatt-hours, lifetime is in years, P(i) is in kilowatts, and t(i) is in hours per year.

The total of the summation on the right side of the equation is sometimes called the unit energy consumption (UEC), measured in kilowatt-hours per year. Variation between studies occurs due to differing estimates for the lifetime; differences in the power draws and time shares, which cause variation in the UEC; and, to a lesser extent, differing assumptions regarding the conversion factors from kilowatt-hour to impact. Each of these three causes of variation is analyzed in turn below.

Unit Energy Consumption

In addition to the LCA studies already introduced, a significant body of literature has assessed the energy consumption of consumer electronics. Results from these studies are summarized in figure 4, with the annual unit energy consumption in kilowatt-hours per year on the y-axis (desktop control unit only, no display) and the year of publication on the x-axis. Studies have been assigned a data quality score, with large-n primary studies and meta-analyses marked “high,” smaller or less rigorous studies marked “medium,” and n= 1 studies or unreferenced assumptions marked “low”; the size of the marker corresponds to the data quality. The marker shape indicates whether the study measured use in a home setting (square), an office setting (circle), or both/not specified (triangle). Primary studies are shaded dark.

Figure 4.

Unit energy consumption of desktop personal computers (no display).

The visual presentation of the data is illuminating; most studies are clustered between 100 and 350 kWh/year, with an approximate average of 225 kWh/year. The highest data points are case studies of a high-performance office workstation (Roth et al. 2002) and a high-performance home gaming PC (Braune and Held 2006), highlighting the wide range of possible direct measurement results, and thus the need for a large n if a study is to be representative of a typical PC. The lower outliers all tended to make questionable assumptions regarding power consumption; for example, Kawamoto and colleagues (2001) assumed that residential PCs are off 91% of the time, but plug-load measurements by Porter and colleagues (2006) show this figure to be closer to 60%. Likewise, Williams (2004) assumed 3 hours (h) of daily use for a residential computer, but Porter and colleagues (2006) measured almost 8 h of daily use. Note that studies that report dominant impacts from the production phase rather than the use phase—Choi and colleagues (2006) and Williams (2004)—both report use-phase consumption at the low end of this range, indicating that differing assumptions are largely behind the disagreement. Consumption of 50 to 100 kWh/year as these studies assumed is not implausible for low-power desktops or for computers that are used infrequently, but seems to be well below the representative range of a typical desktop under typical usage patterns.

Product Lifetime

The lifetime of the product multiplied by its annual energy consumption determines the total energy consumed in the use phase. Measuring average product lifetime is deceptively difficult because computers can sit in storage for years after their useful life is over, and sometimes enjoy continued use in the secondary market. The latter activity should be included in the total lifetime, as the product is still consuming energy, but the former should not. Techniques to measure lifetime, which may include customer surveys, waste stream monitoring, or purchase monitoring, may not always be able to identify storage and reuse, leading to variations in lifetime estimates.

Studies measuring lifetime, and the estimates used in LCA studies, are shown in figure 5. The shape of the markers indicate whether the study was measuring first life only, or included reuse, or did not specify whether or not reuse was included. The most persuasive study of lifetime to date is Babbitt and colleagues (2009), which calculated product lifetime based on 20 years of procurement data in a university setting (n > 2,000/year) and documented a steadily declining trend in lifetime; unfortunately the last reliable data were for purchases made in 2000, when the average lifetime was 5.5 years. Other reported results came from a survey of consumer purchases in Japan (Williams and Hatanaka 2005), a Gartner survey cited in the work of Smulders (2001), and a series of studies in Japan (ESRI 2007; JEITA 2006; Oguchi et al. 2006), originally cited in the work of Yoshida and colleagues (2007, 2009) (the original studies, in Japanese, were not available to us, so we could not confirm these figures). Studies have been assigned a data quality rating, with larger primary studies ranking higher than secondary data, and direct measurements (such as through a procurement database, as in the study by Babbitt and colleagues [2009]) ranking higher than surveys. Notably, several LCAs rely on unreferenced assumptions.

Figure 5.

Lifetime of desktop personal computers (no display).

As shown in figure 5, estimates for PC lifetime range from 3 years to more than 8 years. The best-quality study, from Babbit and colleagues (2009), is probably close to an upper bound on first lifetime because it tracked computers on an academic campus with significant internal reuse. Downward trends in that dataset suggest an average lifetime of 5 years in the present day. The smaller result from Gartner (cited in the work of Smulders [2001]), a 3-year lifetime, was conducted in a business context that likely featured higher replacement frequencies than in an academic context. Second lives due to reuse will increase the average life span. According to Yoshida and colleagues (2009), the establishment of take-back recycling schemes increased reuse rates such that, in 2004, about 37% of discarded PCs in Japan were reused domestically, and an additional 25% were exported, though reuse rates are likely to be lower in countries such as the United States that do not have well-established recycling programs. The prevalence of reuse and the length of any secondary life are additional sources of uncertainty; very little high-quality data are available assessing these factors. Given the available evidence, our best judgment is that lifetimes between 3 and 6 years are reasonable estimates, depending on the context of use and the availability of infrastructure to support reuse.

Total Use-Phase Impact: Analysis

The impacts due to electricity consumption can be measured in megajoules of primary energy or kilograms CO2 equivalent by multiplying kilowatt-hours of electricity by the appropriate impact factors. Not all studies actually report the emissions factors used, but in some cases they can be derived by identifying reported total use-phase impact, in megajoules of energy or kilograms CO2 equivalent, and dividing by total reported lifetime electricity consumption in kilowatt-hours. These impact factors are shown in figure 6, alongside emissions factors for various electricity grids from the ecoinvent database (Dones et al. 2007), obtained using low-voltage at-grid supply data with cumulative energy demand (CED) and IPCC GWP100 LCIA schemes.

Figure 6.

Impact factors for electricity from life cycle assessment studies and ecoinvent database.

Most studies assume 1 kWh of electricity is equivalent to between 10 and 12 MJ of primary energy (except Williams [2004], who assumed 1 kWh electricity = 3.6 MJ primary energy; an apparent error), in order to account for losses in electricity generation. Likewise, emissions factors are usually between 0.4 and 0.6 kg CO2/kWh. Both of these figures are intended to encapsulate the impacts of the electricity generation infrastructure at whatever geographic locale is appropriate for the study, so some variation is to be expected. In order to remove this variation from our results, we assume constant factors of 11 MJ/kWh and 0.5 kg CO2/kWh, which are the averages of the reported factors, excluding the findings of Williams (2004). We note that these are slightly lower than most impact factors in the ecoinvent database, but have not been able to identify why, as rationale for impact factor choice is usually not presented. The large spread in country data shows that the geographic location of product use is an important determinant of the total impact of the product; for example, global warming impacts due to use phase will be 30 times higher in China than in Norway, as China's electricity system is coal based and Norway's relies almost exclusively on hydropower. This information is not unknown to LCA practitioners, but it is not always clearly communicated in the presentation of results.

The data in the previous sections suggest that plausible estimates for UEC range from 100 to 350 kWh/year, and lifetime ranges from 3 to 6 years. Multiplying UEC and lifetime together gives a possible range of 300 to 2,100 kWh. Converted to primary energy using the average emissions factors reported by the studies, this yields a range of 3,300 to 23,100 MJ; in GWP, impacts range from 150 to 1,050 kg CO2-eq. We have chosen absolute rather than probabilistic bounds because the ranges are defined by qualitative assessment of the data rather than rigorous statistical aggregation, the latter approach being inappropriate due to the heterogeneity of the data. Consequently the ranges are relatively large. We examine the implications of this result, and the result of the analysis for the manufacturing phase, in the next section.

Analysis of Overall Impact

Figure 7 shows the ranges identified in the previous two sections and their overall impact; also plotted are the results from other available studies. In interpreting these data, the weight of evidence suggests that the use-phase impacts of a typical desktop PC are more likely to occur at the middle of the range, seen in figure 7, and less likely to occur at the fringes. Studies that lie toward the bottom of the range, including those by Apple (2010a), Atlantic Consulting (1998), Masanet and colleagues (2005), and Williams (2004), have therefore calculated use-phase impacts that are likely well below those of a typical desktop PC. The other Apple study (2010b) is well above typical, though we note that these two studies examined a small system (the Mac Mini) and a high-powered workstation (the Mac Pro), neither of which are intended to be representative of a typical desktop PC. It is unfortunate that no methodological details are available for these studies, as this prevents any evaluation of their quality or reasonableness. Nevertheless, the very wide variation in overall impact for these two studies is intriguing; if the methods are at least internally consistent, then internal variation within the product category of desktop PCs may itself be large enough to overwhelm any methodological or impact-data-related variation.

Figure 7.

Overall impact, showing study results and our estimates of reasonable ranges for a typical desktop personal computer without display.

The emissions factors for use-phase electricity consumption used in this figure are slightly below the average of the impact factors of European Union for the Coordination of the Transmission of electricity (UCTE) country electricity grids according to the ecoinvent database. Use-phase impacts will proportionally increase or decrease according to the impact characteristics of the local grid where consumption occurs. In very low-carbon jurisdictions like Norway, manufacturing impacts (which occur in Southeast Asia) will far outweigh use-phase impacts.

The collective evidence for impacts due to manufacturing is less conclusive. Our previous analysis identified significant variation in impact factors of subcomponents, especially mainboards and ICs, which were a substantial source of variation in the overall impact due to manufacturing, but there was not enough evidence to identify a reasonable range for these impact factors. Methodological differences add more uncertainty: Masanet and colleagues (2005) used a top-down input-output method that is not easily comparable to those studies previously analyzed; the Apple studies (2010a, 2010b) did not report any methodological details. The complexity of the underlying methods and data makes independent verification of any of these studies next to impossible. We have defined a reasonable range of manufacturing impacts to span the range defined by those studies analyzed in the Overall Results of Previous Studies section of this article, on the basis that they provide the most detail, making an assumption that detail indicates quality, which is correlated with accuracy. Outlier studies could expand the range of reasonable results only if they are comparable in observable quality. This line of reasoning is admittedly less than ideal, but until standardization efforts reduce or eliminate uncertainty in product inventories and impact factors, accepting all high-quality studies as reasonable seems to be the only defensible option.


When measuring primary energy consumption and GWP, the impacts due to use-phase energy consumption of a typical desktop PC without display are the dominant impacts in the product life cycle, except in areas with very low-impact electricity systems; manufacturing impacts are smaller, but still significant, and distribution and end of life are both negligible (note, however, that end-of-life activities can cause significant damage to the environment and human health that is not captured in these two impact dimensions). A few studies have reported that manufacturing impacts exceed use-phase impacts, but these studies all report use-phase energy consumption at or below the low end of a reasonable assessment of a typical PC's energy consumption, which ranges from about 100 to 350 kWh/year for a lifetime of 3 to 6 years. Manufacturing impacts exhibit a particularly high variability, most of which is due to disagreement about the impacts per unit of the various components inside the PC, especially mainboards and semiconductors. Estimates of the physical contents of a PC by mass vary as well, but less so. Total impact, summarized in figure 7, might be expected to range from 270 to 1,300 kg CO2-eq and from 4,900 to 39,100 MJ of primary energy, with the most likely result toward the middle of these ranges. The use-phase impacts are strongly dependent on the local electricity grid where the product is used. In addition, variation of the functional unit, such as choosing a lightweight, low-power computer or a high-performance workstation, can have a dominating effect on the value of the impact, making it important for researchers to either conduct studies with a larger sample of products in order to smooth out this variation, or to explicitly limit the scope of their assessments to a specific subcategory. The latter approach has become increasingly important with new products being introduced and the desktop PC category becoming broader.

This exercise uncovered several inaccurate assumptions in published studies and highlights a general problem with LCAs: they are very difficult to evaluate, even for an experienced practitioner, as a full listing of data, methods, and assumptions used is rarely available, often for reasons of industrial confidentiality or proprietary data, and the correctness of the data may itself be difficult to determine. We notice a tendency for data reuse as newer studies build on older studies, but as the pool of primary research is small and difficult to verify, errors may have been propagating through the literature undetected, creating a risk that erroneous results may become established. Data on electronics manufacturing, especially for less-frequently researched components and processes, are particularly vulnerable to undetected errors, and current reported results are of an unknown quality.

Data quality issues will diminish when standardized impact reporting systems with participation from electronics manufacturers are in place. Until such time, practitioners should use caution when adapting results from previous studies and critically evaluate such results, regardless of their prominence and apparent acceptance by other researchers. Rich opportunities remain for future research to reduce uncertainty in LCAs of electronics.


We thank Hadi Dowlatabadi for feedback, and the anonymous reviewers, whose suggestions have greatly improved the quality of this article. This work was supported in part by the Natural Sciences and Engineering Research Council of Canada (NSERC) and the University of British Columbia's Bridge Program.


  • 1

    One kilowatt-hour (kWh) ≈ 3.6 × 106 joules (J, SI) ≈ 3.6 megajoules (MJ) ≈ 3.412 × 103 British thermal units (Btu).

  • 2

    One kilogram (kg, SI) ≈ 2.204 pounds (lb); carbon dioxide equivalent (CO2-eq) is a measure for describing the climate-forcing strength of a quantity of greenhouse gases using the functionally equivalent amount of carbon dioxide as the reference.

  • 3

    One megajoule (MJ) = 106 joules (J, SI) ≈ 0.278 kilowatt-hours (kWh) ≈ 239 kilocalories (kcal) ≈ 948 British thermal units (Btu).

  • 4

    One gram (g) = 10−3 kilograms (kg, SI) ≈ 0.035 ounces (oz); one square centimeter (cm2, SI) ≈ 0.155 square inches (in.2).

About the Authors

Paul Teehan is a Ph.D. candidate at the Institute for Resources, Environment and Sustainability at the University of British Columbia (UBC) in Vancouver, BC, Canada. Milind Kandlikar is an associate professor at the Institute for Resources, Environment and Sustainability and the Liu Institute for Global Issues, also at the University of British Columbia in Vancouver, BC, Canada.