I'm part of a group that promotes light electric vehicles (hybrids between electric bikes and cars) and I'm also a huge user of deep learning technologies. As a part of that I am also involved in a fablab, where we often use things that are weirdly cataloged as low-tech despite being high-tech, like DIY electronics. Discussions about what's beneficial, what's compromise, what's something to avoid crop daily and I would like to clear a few points out. I hope it will be useful, and I hope it will bring some interesting discussions here.
Our transition to a sustainable society requires us to make choices, often tech choices, in a way that's aligned with the final objective. There is a general misunderstanding about the different types of accounting you need to do at the individual level, organizational level, national level and global level in order to achieve true sustainability on a global scale.
CO2-equivalent accounting (which is by the way not the only metric that matters, but is still a crucial one) is generally divided into 3 scopes:
- Scope 1 is the CO2 that's directly emitted by the subject. You burn fuel in a generator or in a thermal engine, that's scope 1 emissions.
- Scope 2 is the CO2 that's emitted by the energy that you are using, mostly electricity, but can be heat and cooling.
- Scope 3 is the CO2 that's emitted by your production chain. In other words, that's CO2 that you don't directly emit, but that through your activity, you make others emit. For instance, you're asking for the delivery of something. The CO2 emitted by the truck that brings it is scope 3.
Note that CO2 that's accounted in your scope 2 & 3 is actually somewhere in the scope 1 of someone.
If you on a personal level or on an organizational level you want to minimize your impact on global CO2 emissions you need to have all three into account and 1, 2, 3 is kind of a good priority order.
The tricky part is that as soon as you have a higher point of view, be it at the regional, national or global level, you should not add these different scopes because that makes you count emissions several times depending on the length of your supply chain.
Consider a paperclip factory. Let's say that extracting material to make one paperclip emits one gram of CO2, that the transport of the raw material to the factory emits another gram, and that the transformation uses electricity that emits one more gram. If we consider it's the same company that does the mining, the transport, the electricity production, and the transformation, it has a scope 1 of 3 grams of CO2. That is the actual real number of gas emitted.
Now imagine if the mining, the transport, the electricity production and the manufacturing factory are actually separated entities:
- Mining: 1g CO2 in scope1
- Transport: 1g CO2 in scope1
- Electricity produciton: 1g CO2 in scope1
- Transformation: 0g in scope1, 1g in scope2, 2g in scope3
Add all of this, through the magic of accounting, we have twice the amount of emissions! Now my point is not to debate whether this exists as a genuine tool to reach carbon neutrality or as a greenwashing tool to make fake savings easier. I think it has a purpose and a use but it needs to be used carefully, because a naive reading of that would be that we can cut CO2 total emissions by just concentrating companies into a few zaibatsus.
Especially when you are trying to decide if a specific technology could be part of a sustainable society on the longer term, only scope 1 actually matters: a sustainable society is a society where all scope 1 are at 0, which means it will automatically make all scope 2 and 3 at zero too. In a transitional period, sustainable tech will need to deploy with some scope 2 & 3 emissions, it is unavoidable but as long as it diminishes the total sum of scope 1 out there, it is a net benefit.
As an engineer, scope 1 is usually what I'm looking at. But it also often makes me blind to other paths of action. When I am looking at the above example, I'm thinking that the transformation step is non-problematic and that we should focus on the other three sectors (mining, transport, electricity) in order to have a sustainable society. Thing is, this example is an oversimplified reality. As a company or individual, you usually have a choice between several alternatives, especially when it comes to electricity production or transport. And you can decide to pay more for something that emits less. So there is a point into pressuring organizations to reduce their scope 2 and scope 3 levels as well.
However, when it comes to evaluate not a company, but a technology, one should only look at its scope 1. We can produce electricity, transport things and mine materials without emitting CO2. Therefore, if your production only uses electricity, raw materials and transport, it can be part of a sustainable society, at least from the CO2 point of view. It does not mean that the companies producing/deploying that tech will automatically be carbon-neutral (scope 1,2,3 = 0), especially if we demand them to optimize their costs in the current industrial ecosystem, but then it is the business/industrial practices that need to be attacked.
This is a paradox that is present in electric vehicles and basically anything that mostly consumes electricity for use or production. If you make the accounting on a personal or organizational level, you can't dismiss the fact that the production of your electric vehicles will have emitted a lot of CO2 during production (scope 2 and 3). However, it is often missed that the most important part of making an EV switch is that it brings down your own scope 1 dramatically. Your scope 2 and scope 3 emissions are usually more than offset by the savings your scope 1 brings into other people's scope 3.