Skip to content

We meet Dan Lewis-Toakley and talk about public cloud carbon emissions

Green Tech South West are looking forward to welcoming Dan Lewis-Toakley (Green Cloud Lead) at ThoughtWorks North America to talk about understanding, measuring and reducing the energy and carbon emissions of public cloud use.

Dan will be joining us on Thursday July 8th at 1pm and we caught up with him ahead of his talk.

How big a problem are heavy cloud carbon footprints?

Dan: The Information and Communication Technology (ICT) industry accounts for around 3% of global greenhouse gas emissions, which is on par with the aviation industry. Within ICT, data centers consume around 1% of greenhouse gas emissions and global electricity usage.

Despite rapidly increasing demand for internet services and cloud technologies, these carbon emissions have remained fairly consistent over the past decade due to improved efficiency of data centers, particularly hyperscale data centers used by major public cloud providers. But there is no guarantee this trajectory will continue, especially as more organizations are digitizing due to the COVID-19 global pandemic — so public cloud customers have a responsibility to measure, understand and optimize their cloud usage if we’re to meet the goals of the Paris Climate Agreement.

How are cloud carbon footprints currently measured?

Dan: Currently, most cloud providers don’t disclose energy or carbon emissions from cloud usage to their customers (at an aggregate or individual level), which can be a challenge for organizations who want to baseline and reduce their carbon footprint.

But there are a number of tools that exist to help measure cloud carbon emission. Microsoft Azure’s Sustainability Calculator, Hubblo’s open source tool Scaphandre and the Green Web Foundation’s Green Cost Explorer tackle this problem in different ways.

Recently, ThoughtWorks also released Cloud Carbon Footprint, an open source tool that allows individuals or organizations to measure and optimize cloud energy consumption and carbon emissions. It currently supports Amazon Web Services, Google Cloud and Microsoft Azure.

The Cloud Carbon Footprint tool from Thoughtworks looks amazing, how does it work and who should use it?

Dan: Cloud Carbon Footprint is for any tech/IT organizations and sustainability team(s) who want to reduce their cloud costs, energy consumption and carbon emissions. More traditional sustainability efforts might look back over the last year and attempt to quantify cloud carbon emissions to report on this, and we hope the tool can help make this process more efficient and accurate. However, we’re also hoping it will help software development teams make better architectural and cloud infrastructure decisions on a day-to-day basis, closer to real time, to help reduce the carbon emissions that are generated in the first place.

Cloud Carbon Footprint works by pulling usage data (compute, storage, networking, memory) from major cloud providers and calculating estimated energy (kilowatt hours) and greenhouse gas emissions expressed as carbon dioxide equivalents (metric tons CO2e). We display these visualizations in a dashboard for developers, sustainability leaders and other stakeholders in an organization to view and take action.

It’s designed to be a starting point which can be extended and customized for individual organizations’ needs. Every organization will have a different cloud context and tech stack, so we are using domain driven design to separate the estimation logic from both the data input source (e.g. cloud APIs, on-premise or co-located data centers) and the output source (e.g front-end dashboard, CSV, etc) so new inputs and outputs could easily be added.

It’s great that the tool is free and open source, what was the thinking behind that?

Dan: ThoughtWorks has been a top contributor and strong advocate for open source software since we were founded over 26 years ago, so in some sense it was a bit of a no-brainer. We have built open source products in the past and offered premium paid features, however this model isn’t the future direction for our company, and if we’re to see more widespread adoption of green cloud practices in the ICT sector, we wanted to make the barrier to entry as low as possible. Also, given the experimental nature of the project, we really want to build a strong community of contributors and users to help shape the product direction and supported featureset, which is more effective with an open source and permissive license. We very much welcome all feedback and ideas — many of the recent features we’ve released have come from the open source community.

What has the response been so far on the CCF tool?

Dan: Overall very positive! There has been a lot of interest in the project from a wide range of organizations — from digital scale ups to enterprises and academic institutions. It’s still early days so the community is small, but it’s growing and we’ve seen an increasing level of engagement from people who become aware that Cloud Carbon Footprint exists. I hope that the Green Tech South West community gets involved and helps contribute to the software!

What are the plans for the product going forward?

Dan: As I mentioned above, our roadmap is often shaped by the needs of the community, but at a high level we have a few different feature sets on the roadmap. First, we are working on adding support for a recommendations engine so users can understand ways in which they can optimize their cloud usage and better reduce cost and carbon emissions. One thing we’ve often heard is “I want to reduce my cloud carbon footprint, but where do I start?” — and we hope these features will help more organizations take the first steps.

We’re also going to look into adding support for energy/carbon emissions from on premise or collocated data centers, which could help organizations quantify the potential emissions reduction from a cloud migration or modernization effort.

Lastly, we’re interested in providing an option for users of the software to anonymously share some of their energy/carbon metrics along with some other data points like size of the organization, or type of industry. If we were able to collect this data it could enable us to benchmark what good (or not so good) looks like for a given organization profile, so users of the software would compare themselves and we could potentially consider having a Green Cloud “score” that could be improved upon over time.

Thank you Dan for the chat.

Dan will take the virtual stage at our Green Tech SW meetup on Thursday 8th of July at 1pm. Head over to the meetup page to RSVP for free. (https://www.meetup.com/GreenTech-South-West/events/277181161/)

To top