Featured post

Textbook: Writing for Statistics and Data Science

If you are looking for my textbook Writing for Statistics and Data Science here it is for free in the Open Educational Resource Commons. Wri...

Thursday 3 October 2019

Offsetting the carbon emissions of the blog, and then some.

Worrying about climate change is wearing down my sanity, and I'm know I'm not alone. I wanted to find a way let others reduce the amount of CO2 and methane in the air that didn't cost them money. One way is to buy carbon offsets with money from advertisements, like the ones that roughly 10% of you see at the top and side of this blog.

Is that futile?

How much carbon emissions are produced by a visit to a website like this? How does that compare to the cost of offsets? This is going to be a rabbit-hole of citations and unit conversion, so buckle in.

How much carbon emissions need to be offset?

According to this document by IEEE, https://ieeexplore.ieee.org/abstract/document/7828166 , the internet was directly responsible for 5% of the world's electricity consumption in 2012, and that the amount was growing by about 10-12% year-over-year. So how does that translate into an individual webpage?

According to Website Carbon Footprint Calculator (WCFC), at https://websitecarbonfootprint.com/ , "basic websites are responsible from 0.02 to 0.2 lbs of CO2 emissions per SECOND of viewing". 

That's... distressing, first because that would make carbon-offsetting a blog like this prohibitively expensive. Second, because it would imply that a smartphone or laptop being used for surfing could be putting out as much emissions as a mid-sized car on the highway.  WCFC cites "Harvard physicist Alex Wissner-Gross", and there seem to be many articles online confirming this figure, but with the same citation.

Except, there's this article by Tech New World, https://www.technewsworld.com/story/Harvard-Prof-Sets-Record-Straight-on-Internet-Carbon-Study-65794.html , which includes a quote from that same professor saying that he was previous misquoted, and that his estimate is 0.02g per second, NOT 0.02 lb. He also didn't give a range, just that "20 milligrams per second" as an average.

So that's only 1/454 as bad. What do we take from this? That people don't fact-check? That American universities are so out of touch with the common people that they use metric?

We need a new source. How about the Energy Information Administration, of the US Government? According to the article here https://www.eia.gov/tools/faqs/faq.php?id=74&t=11 on the carbon impact of electricity usage, electricity generated in the United States varies a lot in emissions based on its source, but averages around 1 lb per kWh. In places that are predominantly coal-powered, that's a tad over 2 lbs per kWh.

Therefore, if we can get a sense of the electricity used in web surfing, we have carbon emissions for a typical user.

According to http://energyusecalculator.com/electricity_laptop.htm , a typical laptop uses 20-100 watts, or 60 watts on average. That means 1 kilowatt-hour is enough energy to power a laptop for 1000/60 hours, or 16.67 hours (about 60000 seconds).  Using the Wissner-Gross figure of 0.02g/s for 60000 seconds implies 120 grams of emissions for that 1 kWh that the laptop uses, or 1/4 lb. But Wissner-Gross considered the whole pipeline, including the web server, intermediate infrastructure, and (I think) a tiny portion of the emissions from manufacturing all that technology, which might explain the other 3/4 lb.

Ultimately, the 0.02 grams per second figure is in line with the EIA's emissions estimates based on general electrical usage. So, let's go with that.

One more issue, the Wissner-Gross study was published in 2012. The web was viewed on more energy-efficient phones much less often then. Servers have gotten more efficient, and their waste heat is more often repurposed. Websites have also gotten a lot more bloated too. There's going to be some drift of that figure, but in which direct I don't know.

We're left with one other source, again from 2012: "The Megawatts Behind Your Megabytes" at  https://aceee.org/files/proceedings/2012/data/papers/0193-000409.pdf . This document is especially useful because it breaks down the energy costs by component. It uses that breakdown to estimate energy costs for not only web surfing, but other online activities like large downloads. It also offers suggestions for reducing the carbon footprint of users (close your unused tabs), and of creators (minimize gifs/images/bloat). 

Did we collectively decide that this calculation was done once and for all in 2012?

There is another website carbon calculator, at https://www.websitecarbon.com/ , that uses the data from this study to calculate emissions per view of a webpage that you send it. According to it www.stats-et-al.com … 

produces 0.94g of CO2 emissions per view. 

That's in line with the other estimates, if an average view is 47 seconds long.
But, that's just the main page, which is mostly text and doesn't have the added images that are in most articles, and it might not load the ad scripts because it's not viewing the page as a human would. We also want any offset to over-compensate for the emissions caused, and by a lot because offsets have their own limitations. Just to be sure, I've decided to 

offset 50.00g of CO2 equivalent per view.

It's "equivalent" because I'm starting with methane reduction until there's no more of those offsets left. Reducing methane has the most impact per dollar.

How much is that going to cost?

The Bonneville Environmental Foundation, at b-e-f.org , offers provable carbon and carbon-equivalent offsets for $10 USD / metric ton. So $1 USD is going cover 2000 page views, which is enough that the few ads you see are enough to cover or meaningfully subsidize the cost of the offset. So to answer the original question, no, it's not futile to try and make a blog carbon negative. It isn't even hard.

As of 2019-10-03, there have been about 64,000 pageviews, and I can only buy offsets in units of 1 ton. So...

... 3 tons offset. It's a start. It's really not much.

Did that seem like a grand virtue signal for something that didn't actually cost a lot? It should! It's easy and cheap to offset the carbon footprint of your activities, and places like BEF show their work in order to prove that the money given to them directly causes a reduction in greenhouse gases. (That's right, I used the c-word, CAUSE)

The price of $10 USD/ton is cheap. A typical Canadian emits 10-20 tons in a year, and they could nullify those emissions for less than 1% of their after-tax income. These offsets are low-hanging fruit, and it's going to get harder, the more we reduce, but it's start. By contrast, emission permits on the ECX (European Carbon Exchange) sell for about 25 Euros/ton in 2019. 

These particular offsets are so cheap because they're using the money to capture methane from landfills, which is startlingly simple given the fact it isn't done everywhere yet. If the low-hanging fruit of carbon offsets run out, I'll maintain 50 grams / view by funding clean energy or reforestation. If that reaches its natural maximum, I'll find a way to put it into carbon capture and sequestration. We can't be satisfied with slowing carbon emissions; we must reverse them.

Not alone.

Here's a list of websites that have a similar setup, in that merely visiting these webpages moves the needle farther away from human extinction. You can help by expanding this list. Contact me at jackd@sfu.ca if your website is provably carbon negative and I'll add it. This list will be repeated in future carbon offset updates.

Visiting any of these will help, and none of them will cost you money.

https://www.ecosia.org/   - A search engine that donates 80% of its profits to tree-planting.

https://tab.gladly.io/ - A browser extension that donates its ad revenue to the user's choice of charity. 

https://www.greengeeks.com/ -  A web hosting company offsets 300% of their emissions.

https://ecoally.co -  An environmentalism blog that uses Greengeeks.

plantthepeace.comA trivia game with ad revenue going towards tree-planting.


Follow-up questions

Q: What about bot views?
A: About 20% of my views are bots. These will be counted as real views for carbon offset calculation and will continue to be unless this percentage increases by a lot.

Q: Do I have to offset 50 grams per view to get on the list? 
A: No. That's more than some people make through AdSense, and the only reason I can justify that much is because this blog itself is like one big advertisement of my work to, say, people looking for consultants, or journals looking for copyeditors. Do what makes sense for your own monetization.

Q: Why not just give all your money to fight climate change?
A: Because this is in part a marketing ploy, part for my own mental health, and part because I feel like I can do more total good by setting an example and trying to pressure / incentivize others into following suit. When Ecosia started spending more on development and advertising, the total number of trees they were able to plant increased by a factor of 100+.

Q: Why only advertise to 10% of viewers?
A: The power law (a.k.a. the 80-20 law or 90-10 law) applies to ad revenue. The AdSense account on this blog is set to reject the lowest paying 90% of advertisements, leaving a blank space on the sidebar instead. After experimenting for six months, this has led to almost no drop in ad revenue while filtering out all the trashy "one weird trick" ads.

No comments:

Post a Comment