Literature Review

WFH Carbon Footprint > Methodology > Literature Review

Literature Review Scope

Over 100 articles of academic and grey literature were reviewed for this project. Such a volume was required to:
• fully understand the nuances of carbon accounting
• determine if there was a need for a work-from-home calculator
• review the elements included by currently available calculators 
• determine best practice methodology
• identify key elements for a working from home calculator.
It should be noted that the underlying assumption for this project, supported by all the papers reviewed, is that climate change is real, current, and attributable to the direct impact of human activity. Specifically, the release of an excessive volume of greenhouse gases (GHG) including carbon dioxide, methane, nitrous oxide, chlorofluorocarbons, and ozone (IPCC, 1992). 
For further details, see the Fundamentals pages: Greenhouse Gases and Carbon Footprint.
 

Introduction to Carbon Accounting

Carbon accounting is the process of itemising emission-generating activities and assigning them values, whether nominal or financial, and whether through obligatory or voluntary reporting. (Stechemesser and Guenther, 2012).

The seminal text in this field is the Greenhouse Gas Protocol developed in 2001 through a partnership of the World Resources Institute and the World Business Council for Sustainable Development. (Greenhouse Gas Protocol, 2016). Essentially, the GHG Protocol operates as an accounting guidance and reporting standard for corporations, governments, and non-government organisations to disclose their greenhouse gas emissions across their supply chain it terms of global warming potential. (Greenhouse Gas Protocol, 2015)

For clarity and commonality, the Protocol outlines the methodology to convert each gas emission into the “equivalent amount of carbon dioxide.” (Eurostat, 2023)
The euphemistic term of ‘carbon footprint’ identifies the sum of all greenhouse gas emissions for an individual, business/organisation, or community. (Carbon Trust, n.d.) This is expressed in terms of carbon dioxide (CO2) equivalence (abbreviated to CO2e) by weight in metric measurements. For example, 25 kgCO2e.

This is the most simplistic approach and is useful to assess options ‘at a glance.’ For example, an incandescent 100W light bulb uses around 172g CO2e over 4 hours, compared to 30g CO2e from an 18W bulb. (Clever Carbon, n.d.) Best practice requires a more detailed analysis to both fully account for factors and to report them in a meaningful manner.

Current Calculators

There are plenty of calculators online to map individual and business carbon footprints at this level. Individual examples tested include:

Typically, these ask for details on:

  • Diet and eating habits
  • Housing type and household size
  • Waste type and volume
  • Transportation and travel
  • Lifestyle spending
  • Electricity source

The more robust of these factor in the respondent’s location, which has a marked effect on total emissions due to local variations in electricity supply sources, household heating/cooling methods, and transport efficiency.(Greenpixie, 2024)

Of the six, only Clever Carbon allows the user to select “Works From Home” as an option for Transportation / Commuting.

None of the calculators reviewed assign any value to information and communication technology (ICT) devices, transmission networks or data centres even though this sector now rivals or exceeds the global aviation carbon footprint. (Copenhagen Climate Centre, 2020; Freitag et al., 2021; Mytton, 2023; and Molthan-Hill et al., 2023) In a 2021 Harvard study, Gupta et al, claims an optimistic expectation of the energy footprint for ICT to double by 2030, but that a realistic projection would be 20% of global energy demand.

Given the essential nature of ICT for business, it’s reasonable to assume this comes under business calculators. Yet the Irish Government’s ClimateToolkit4Business does not provide an answer. Its focus is on what seem to be the big ticket items such as electricity and business travel.

The UK Government refers businesses to a whitepaper on homeworking developed by the consultancy, Eco Act. (Skillet and Lindsay, 2020) The paper was sponsored by the banking institutions, Lloyds and NatWest, but the methodology it proposes—while basic—includes all assumptions and is relatively easy to implement from a business-perspective. That is, a carbon footprint allocation can be given to any employee working from home in the UK on a full-time equivalent basis.

This figure cannot be carried directly over to an Irish remote worker because it vastly under-reports the carbon emissions from the Irish electricity grid and home heating methods that typically include the direct burning of peat, coal and heating oil. Many of these figures, however, can be directly sourced from the Sustainable Energy Authority of Ireland (SEAI) or the Central Statistics Office (CSO).

But how robust is this methodology?

Best Practice Carbon Accounting

Under the Greenhouse Gas Protocol, there are three papers relevant to this project. The first is an updated Corporate Standard (2015), the second is the ICT Sector Guidance (2017), and the third is the Scope 3 Technical Guidance (2013).

None of these specifically deal with working from home, but the Corporate Standard includes: “Employee business travel” and “Employees commuting to and from work” under “Other Indirect Emissions”. The ICT Sector Guidance confirms greenhouse gas emissions from ICT should be accounted for within Scope 3.

Scope refers to how the GHG Protocol splits emissions, as follows:

  • Scope 1: emissions directly consumed by an entity.
  • Scope 2: relates to purchased electricity.
  • Scope 3: captures upstream activities of purchased materials and services and downstream activities such as waste disposal, product use and outsourced activities.

The GHG Scope 3 Technical Guidance outlines four calculation methods to assign a carbon equivalency to goods and services. These are:

  1. Scope 1 and 2 figures directly from a supplier
  2. Taking industry averages of mass data and applying it to a specific product or service.
  3. A hybrid of both above.
  4. Using industry averages and applying it per monetary value of a good or service.

The Eco Act methodology predominantly follows the second calculation approach by taking industry and national averages. Unfortunately, a scan of various ICT suppliers illustrates why this is necessary, since reporting is either at a global aggregate level or behind a provider paywall, and inevitably comes with several years’ time-lag.

Various educated estimates are available, but these rarely agree at either the macro or micro level. For example, the size of the ICT industry in 2020 ranged from 1.8%-3.9% (Molthan-Hill et al., 2023). Alternatively, transferring and storing a single gigabyte of data to a cloud server is estimated to use between 3.1 to 7kWh. (Adamson, 2017)

This variation is due to the complexity and opacity of supply chains—and the fact that Scope 3 reporting remains voluntary despite the fact it is, conceptually and in actuality, the largest source of emissions.

As a final example, take this claim by Vodafone UK (n.d.), that one gigabyte of data is enough to “send or receive about 1,000 emails.” As Salahuddin and Whit (2023) of Deloitte UK note, emails range from text-only (about 4g CO2e) to those with an image or file attached (about 50g CO2). Emails may then be forwarded, downloaded, saved, or replied to almost infinitely. At the email provider level, there is further replication because, as Thierry Joubert (2023) notes, Microsoft stores data on at least three different servers to ensure constancy of cover.

Are there any alternative approaches?

Jens Malmodin believes there is. In a paper presented to the 2023 International Conference on ICT for Sustainability, Malmodin explains how he attached smart plugs to his family’s ICT equipment, used a professional multimeter, and recorded the household’s activity over a six-week period. He then factored variables such as fibre broadband specifications, substation, and grid electricity data, and extrapolated an annual carbon footprint. But Malmodin is not only an industry-recognised leader in the ICT carbon accounting field, he is also an engineer at the Ericsson communications company with access to commercially-confidential information. He also provided a sample of one: a statistician’s worst nightmare.

His approach, however, follows the ICT Sector Guidance’s consideration of the sector splits of:

  • Data centres
  • Devices
  • Communication networks.

Malmodin also allows for embodied carbon (the emissions required to manufacture a product or deliver a service) across fibre cable deployment and for all devices. However, he also includes gaming devices, smart TVs and media streaming. For the purposes of homeworking, a business-centric perspective must be taken on ICT. (Zuppo, 2012). 

A 2023 study by Tao et al identifies the device selection necessary for desk-based work.

Working from Home Desktop Devices (Tao et al, 2023)

Tao et al’s study also identified that the “impacts of IT usage [were] negligible” when directly compared to the same set of people’s on-site office setup. Of more importance was the home office environment.

It should also be noted that most calculators and studies deal primarily in operational emissions. The Cloud Carbon Footprint website, an open-source initiative, notes that the optimal calculation includes capital expenditure in the form of embodied emissions, as per this formula:

Total CO2e = operational emissions + embodied emissions

Devices are particularly high in regards to embodied emissions. According to Gupta et al (2021), “most emissions related to modern mobile and data-center (sic) equipment come from hardware manufacturing and infrastructure.” For a battery-powered device such as a mobile phone or tablet, manufacture creates up to 75% of its lifetime carbon footprint. By comparison, the figure for a desktop is closer to 40%.

Therefore, a light data user who upgrades a mobile phone and watch annually may well have a larger carbon footprint for ICT than a heavy data user who replaces a desktop and monitor every five years.

Finally, while other considerations such as water and waste might be added in future, these are already accounted for in most individual carbon footprint calculators.

The Need for a Work-From-Home Calculator

One of the few benefits to the Covid-19 pandemic was the immediate and sustained drop in carbon emissions due to lockdowns, as shown in this chart from the World Bank.

This is attributed to the significant dent in global transportation emissions when remote working replaced commuting by as much as 50% in the USA (Thompson, 2023) alone. In Ireland, nearly 40% of employees worked from home during that period, up from 16% pre-pandemic (Central Statistics Office, 2022a and b).

This is a trend set to continue in Ireland with legislation passed this year granting Irish residents the legal right to request to work from home (Citizensinformation.ie; Walsh, 2024b), and a continuing high level of hybrid/remote-working job advertisements (Walsh, 2024a).

Apple Inc, is one of the first Big Tech companies explicitly factoring in working from home into Scope 3 reporting (2023).

Apple Inc Corporate Emission 2023

As noted previously, the UK Government is also providing reporting companies with FTE CO2e conversion factors for homeworkers.

But perhaps the strongest case for the benefit of calculating the carbon footprint of remote workers is to illustrate the carbon emissions avoided.

Project Drawdown, a “network of scientists, researchers, and fellows” (Project Drawdown, 2020) identified “telepresence” as an alternative to commuting in 2018 (cited by Molthan-Hill et al., 2023). A year earlier, the ICT Sector Guidance labelled this switch “the enabling effect,” that is, that teleconferencing and telecommuting can act as a substitute for transportation. It also warns, however, of a related concept: the rebound effect. This is when switching to a cheaper and less obviously an emissions producer, usage is likely to increase substantially. (Something anyone who has switched from film to digital photography will appreciate.)

Research by Tao et al confirms a further danger of this enabling effect. Their 2023 study, with access to anonymised data from Microsoft, compared homeworkers to on-site workers, and recorded:

substantially more total travel miles for remote workers to drop off/pick up friends, conduct recreational activities, visit healthcare facilities, visit friends/relatives, and exercise.

In an interview for Scientific American (2023), Yanqiu Tao, the lead author of the paper, noted “It’s not all about how many days you work from home. It’s about how well you live sustainably.”

Therefore, while providing a comparison to commuting can make a strong argument for an employee to justify more days working from home, it remains up to the individual to make this a genuine case.

Meanwhile, as this short data visualisation from Polyhedral (2024) shows, while our individual choices make a difference, the true global impact comes from holding tech giants, governments and industry accountable.

Want to know more? Select one of the links below.