Carbon Agile.
I wanted to document away-from-my-other-blog (mariosblog.co.uk) the focus I’ve been applying to the study & learning of carbon emissions & the subsequent calculation of those emissions.
I will be offering up all the data as-I-learn on my website, carbonagile.com which is COMING SOON.
My work started with an idea of calculating the carbon footprint of an Agile delivery based project. This came in turn from work I’d done for a retail client & we’d done a sustainability assessment around how the cloud compute services were run & at what time, so as to best be carbon efficient.
Microsoft, in particular offer some great guidance, technical support & a sustainability API which you can plug into, when running any of their Azure services. You can choose when to run your cloud service, or compute load, whether you’re pulling in some data once a day or churning through some data model & it tells you the best time to run the workload. You can see the carbon you’d saved if you’d elected to run the workload at the recommended time, or your original time & you can see the offset. All straight forward. You see this empirically on a great dashboard, all useful stuff.
You see, we understand quite well compute loads, the power usage & the details of how data centres impact the environment, the cost to run them etcetera etcetera. But what we don’t understand seemingly well is the ‘scope 3’ (more on scopes in a mo) end of the calculation puzzle. It all becomes a bit of a quasi finger-in-air science.
Scopes
As I started to think about how to calculate the carbon use of fleshy human beings in the context of delivering a project, I started to learn about ‘Scopes’. To measure carbon, you need to get your head around the over-arching principles of how it’s done. You’re mean’t to measure not just carbon but other gases such as Methane, (CH4) & Nitrous Oxide (N20) but for this experiment, I’m only focusing on Carbon (CO2).
The Greenhouse Gas Protocol or the GGP remains the sole global standard for emissions reporting. I found it a really useful resource to get to grips with some fundamental science. They in turn define the following scopes when calculating carbon emissions;
Scope 1 — these are emissions which are directly controlled by a company
Scope 2 — these are emissions which come from the generation of power, heat & steam & cooling. Things companies ‘buy in’
Scope 3 — these are emissions linked to the value chain. Whether that’s upstream involving the two scopes above, or downstream, which dovetails into the lifecycle of the product or services it offers, alongside those fleshy human beings!
Scope 3 I’ve learnt is a nightmare to try & calculate. For a number of reasons I wanted to explain. Scopes 1–2 are ‘easier’ in that they are quantifiable directly from the energy companies themselves. You can literally go onto any energy providers website, & find out the cost & standing charge calculations for any tariff you need. This makes plugging that data into a calculation model reasonably easy.
EDF for example, looking at the screenshot above, offer not only data but apply that to real-world case studies to allow you to understand an average carbon cost for a type of person. Its really useful & each energy supplier offer something similar.
Setting a baseline with Scope 3
I had an idea of a carbon calculator tool. Research showed me their were many already. Some frustratingly behind paywalls. Here’s some I found
Some others behind subscription paywalls but seemingly have a lot of publicity & clout behind them;
Those three above look like they promise the earth (no pun intended) but I don’t see how they will be driven by any more accurate data than with which you can find useful. Finally, for me, my go-to source was the UK Gov which not only offers calculation tools but actual quantifiable data I could use as a starting point for my model. — here. If you dont know who Sir David MacKay was, you can read about him here. He’s been involved in energy & climate change discussions for a very long time!
The site offers a comprehensive Microsoft Excel version of their online calculator tools which you can download here
The problem though it was still an unintelligible spreadsheet, so I kept looking, poured some coffee & found SSE Energy’s offer where they offer two types of calculator from their home page, the one I’m using as my starting point can be found here
The vision & approach
So what was I trying to do? Well, I was trying to add into my project tool kit, the ability to think about, understand & calculate carbon use within any given project delivery. I’m always thinking about cost, budget, time, but now I wanted to add the notion of ‘carbon saved’ at the bottom of the invoice. I also wanted to create a calculator, ‘yeah yeah I know everyone else is doing one’ but they aren't really granular or accessible enough for say, a project manager.
So there were going to be some concessions straight out the box. Namely;
- approximate values only
- ‘best endeavour’ in terms of data accuracy’ & grounded in common-sense
- No attempt at trying to be all things to all people & authoritative in the doing so
I wanted to also create a grounded data source I could then codify later down the line into an online calculator. But that’s for another post.
Data
So if I have a data source I can add to, I need to understand HOW to calculate carbon emissions. Then after that, what to include in my model.I learnt as-I-went that calculating carbon emissions is not easy. As I understand it currently, you calculate over the course of 12 months, (with an aim of comparing one year to the next) but I didn't want to work at that level. I wanted to be more granular.
It *should* be possible to make some arbitrary calculations if I applied the same approach. I needed to work out how & what to calculate. So to that end;
- 1 person (this could be scaled up to the size of the scrum team but wanted to start on a 1 person average)
- Their typical travel route & type, (distance, method)
- Their daily energy consumption across HVAC, Food consumption (by type & method), liquid consumption & broad energy demands per hour & day across;
- Appliances (such as TVs, kettles, toasters & Microwaves)
- Compute (inc. laptops & other IT peripherals with variable high power demand)
- Meetings, (inc taking the person above) plus any Video conferencing gear
If the above was the person, and by mapping them I had appliances & IT, I then only had to consider office vs home. Treating each one as a persona in it’s own right, so I considered;
- Home or office
- Size of home, its energy posture & standing rate, (leveraged from local council data)
- Size of office, its energy posture & standing rate, plus number of average appliances by type, HVAC profile, lighting & other considerations.
With all of that I should be able to create some broad formulae, which loosely look like;
To calculate a person(s) carbon emissions based on an 8 hour working day;
Person / location / compute x appliances x time %perday% / 8
To calculate a home carbon use;
%energytariff% / %knownstandingrate% x appliances / 8
To calculate an office carbon use;
%energytariff% / %knownstandingrate% x appliances / 8
Alexandra Shimo-Barry, author of “The Environment Equation,” has come up with a formula that explains how to calculate your carbon footprint at home, too! Simply follow the below steps, and voila.
- Multiply your monthly electric bill by 105
- Multiply your monthly gas bill by 105
- Multiply your monthly oil bill by 113
- Multiply your total yearly mileage on your car by .79
- Multiply the number of flights you’ve taken in the past year (4 hours or less) by 1,100
- Multiply the number of flights you’ve taken in the past year (4 hours or more) by 4,400
- Add 184 if you do NOT recycle newspaper
- Add 166 if you do NOT recycle aluminum and tin
- Add 1–8 together for your total carbon footprint
According to the above an “ideal” carbon footprint (or a “low” footprint) is anywhere from 6,000 to 15,999 pounds per year. 16,000–22,000 is considered average. Under 6,000 is considered very low. Over 22,000? You may want to treat your ‘carbon skid’ with a few paper towels! I ran the above with some dummy figures based on myself & I was hitting way over 22,000.
So I started creating entries into my spreadsheet using a broad process looking something like,
- Figure out the ‘thing’ I wanted to measure
- Apply some common sense into how I’d measure it
- Check a couple of sources and compare my numbers
- Use some kind of calculator & check again
It’s not an exact science, & I wasn't confident with the different answers I was getting from multiple sources,
I ended up going to multiple sources to understand the ‘average’ carbon cost of as many appliances as I could find. I did appliances, vehicles & helpfully the tech companies all provide data around VC carbon use, so those numbers are quite granular.
I ended up with a list of numbers I’m ok with to use as a basline for further codification. Now I needed to build a calculator with some Excel-fu. But that’s for the next post!