The embodied carbon of a device is an important aspect of the whole carbon footprint. The ratio between emissions from use and production varies greatly depending on the type of device.
The carbon footprint for ICT, including TVs and other consumer electronics, rises to 1.2–2.2 GtCO2e (2.1%–3.9% of global GHG emissions) in 2020 with ca. 30% coming from embodied emissions and 70% from use phase emissions.
Across studies, roughly 23% of ICT’s total footprint is from embodied emissions, yet the share of embodied emissions for user devices specifically is ca. 50%. This is because, unlike networks and data centers, user devices are only used for parts of the day and use less electricity, but are exchanged often, especially in the case of smartphones.
The global embodied emissions of selected ICT user devices for 2020 were estimated to 180 Mtonne CO2e (excluding use and end of life), by combining the Supply Chain approach gathering data for the entire electronics sector and allocating it to user devices, with the LCA/PCF approach scaling product level data with annual sales/shipments.
For devices such as laptops, embodied carbon typically accounts for around 80% of the total life cycle emissions, while usage emissions make up only 20%. In contrast, for servers, the breakdown can be reversed, with embodied carbon constituting around 20% and usage emissions around 80% of the total. These proportions can fluctuate depending on the specific device model. Desktops tend to have a higher proportion of usage emissions compared to other user devices like laptops or smartphones.
Servers can be associated with high usage emissions compared to devices like laptops or smartphones. This is largely due to higher power consumption associated with more powerful processors, large amounts of RAM, and multiple hard drives to store and process data, as well as more continuous usage patterns. Typically, server usage carbon emissions contribute the larger portion of their overall carbon footprint.
FIGURE 1. Majority of computing’s carbon footprint comes from embodied emissions not operational energy use. From iPhone 3 to iPhone 12, the operational footprint reduced by 2.09 given efficiency optimizations while the embodied emissions rose by 2.85 due to higher manufacturing overheads. Between 2010 and 2018, the operational energy consumption of data centers increased by only 6% collectively while the number of compute instances, i.e., infrastructure overheads, increased by 6.
gupta.etal.2022.chasingcarbon (pg. 2)
In contrast, Figure 1 (right) shows the carbon footprint breakdown for data centers. At the data-center scale, we find operational emissions between 2010 and 2018 have increased by 6% while infrastructure capacity, which correspond to embodied hardware emissions, has increased by 6 .
Comment
Wrong. The figure doesn't show the carbon footprint breakdown of data centers. Instead it shows the energy consumption of data centers. The 6 % increase is electricity not operational emissions. The change in operational emissions is unknown based on the used source (Masanet et al. 2020).
gupta.etal.2022.chasingcarbon (image) (pg. 5)
FIGURE 3. Breakdown of carbon emissions for various Apple, Google, and Microsoft personal-computing platforms. As the top chart shows, hardware manufacturing dominates the carbon output for battery-powered devices (e.g., phones, wearables, and tables); most emissions for always connected devices (e.g., laptops, desktops, and game consoles) come from product use. The bottom chart shows the absolute carbon output of battery powered and always connected devices. Overall, carbon footprint (total, manufacturing, and use) is variable and scales with the platform.
In most cases, the “embodied emissions” (all stages excluding the use stage) of software are not significant compared with the overall emissions of the ICT system, particularly when the embodied emissions caused by development of the software are amortized over a large number of copies. In these cases, it is not necessary to carry out a detailed life cycle assessment of the software as part of a wider system.
While extending the lifespan of notebook computers is necessary for reducing the climate impact of the IT sector, not all of them are designed for a long life.
For example, a 2019 R640 Dell Server has an amortized embedded carbon cost of 320 kg CO2eq/year. It's also expected to consume 1760.3 kWh/year of electricity. The average carbon intensity in the EU was 0.276 kg CO2eq/kWh for 2019.
Therefore, the total carbon cost will be 320 + (0.276 * 1760.3) = 805 kg of carbon/year, of which 320 kilograms or about 40% is from the embodied carbon. Embodied carbon is a significant contributor to the total emitted carbon of servers.
— Principle 4: Embodied carbon - Training | Microsoft Learn