Data

Greenalytics calculates the carbon impact of websites by matching Google Analytics (GA) statistics with environmental research data. Greenalytics' calculations are approximations and are based on a number of assumptions. We have open and transparent approach for avoiding misinterpretation. The data used are based on the following references:



A. Users

The user impact of the website is the potential environmental impact generated by the users while browsing the site. In the current version this includes the electricity consumed by their computer. The application makes use of GA detailed information about the users: their exact number, location and duration of visit. The total CO2 is calculated by aggregating the impact per country, which is the total time visitors from that country spent on the site multiplied by how much electricity computer consume multiplied by the electricity factor (how much carbon dioxide is emitted per electricity unit) of that country

Assumptions:

• We assume that the users’ time is spent exclusively browsing the website, so all the electricity use for that time is allocated to the site.

• The type of computer and screen size is not possible to know. We assume a mix between laptops and desktops: 45 per cent using laptops consuming 19W and 55 per cent using desktops consuming 60W. Electricity consumption information is taken from IVF (2007) and the 45/55 distribution from IDC (2008).

• The carbon factors for the electricity are taken from carma.org through their API. These factors are calculated for a whole year and do not reflect the specific energy mix when the users visited the site. When the country is unknown, the global average is used.


B. Servers and Internet infrastructure

The server and infrastructure refers to the electricity use of servers, data storage and network infrastructure. It is calculated using the total data traffic generated by the site and an approximation of the energy used by internet per data unit. GA doesn’t provide a direct way of getting the total data traffic, so an approximation is calculated by aggregating the total traffic per page (the size of the page per the number of visits it has), multiplying with an energy use per data unit factor (3.5 kWh/GB, taken from Weber et. al (2010)) and finally multiplying with the electricity factor of the country where the server is situated for getting the CO2 value:

Assumptions:

• The carbon factor of the country where the server is situated is used for the whole calculation, even if part of the infrastructure is global.

• The energy factor of 3.5 kWh/GB is generic and contains itself many assumptions. Weber et. al (2010) presents 7 KWh/GB for 2008 and argues that this amount is halving every two years. This number is the biggest uncertainty of Greenalytics but it seems to be consistant with other data as from Malmodin et al (2010).

• The data traffic is calculated by the size of the pages, and it doesn’t take into account other factors such as caching. Dynamic content as streaming media or AJAX based interactions are not included. Using just traffic to measure the impact does not include factors as server efficiency or websites that include heavy data processing.


References:

Weber, C. L., Koomey, J. G. and Matthews, H. S. (2010), The Energy and Climate Change Implications of Different Music Delivery Methods. Journal of Industrial Ecology, 14: 754–769. doi: 10.1111/j.1530-9290.2010.00269.x

Koomey, J.G. 2007a. Estimating total power consumption by servers in the U.S. and the World. Lawrence Berkeley National Laboratory and Stanford University.

Koomey, J.G. 2009. Sorry, wrong number: Real-life lessons for responsible use of data and analysis in decision making

Malmodin, J., Moberg, Å., Lundén, D., Finnveden, G. and Lövehagen, N. (2010), Greenhouse Gas Emissions and Operational Electricity Use in the ICT and Entertainment and Media Sectors. Journal of Industrial Ecology, 14: 770–790. doi: 10.1111/j.1530-9290.2010.00278.x

IVF Industrial Research and Development Corporation. 2007. European Commission DG TREN Preparatory studies for Eco-design Requirements of EuPs. Personal computers (desktops and laptops) and Computer Monitors. Final Report (Task 1-8)

IDC (2008). Worldwide Quarterly PC Tracker.

Zapico, J.L., Turpeinen, M., Brandt, N. (2010) Greenalytics: A tool for mash-up life cycle assessment of websites. In Proceedings of the 24th International Conference on Informatics for Environmental Protection. Shaker Verlag, Aachen, Germany.

Disclaimer! The data generated by this services is just an approximation. Our intention is to reveal the hidden impact of websites, not to provide exact calculations (even if we try our best!).

Do you have better data? Suggestions? Questions? Comments?:

Info at greenalytics dot org