The tech space brims with people keen to act on global warming. However, some efforts seem misaligned with where people in the energy transition space say to focus.

Our timeline is short. We need to target, as much as we can, where software will have the most impact on CO2 emission reduction.

This post gives a forest-for-the-trees view of the energy transition in 2020. It is serves as a starting point for understanding where software can help move the needle. It will cover three topics:

  • What are the elephant-in-the-room emissions sources?
  • What is the broad plan for transition?
  • What is the timeline?

The Elephants

There are many ways to view emissions sources.

The approach below groups emissions by sector. The electric grid has its emissions assigned to the sector using the power, rather than the plants generating it. This perspective brings into focus where we are consuming energy, and sets the grid apart as a special function. That aligns with proposed plans (see further down) of cleaning up the grid and moving almost all energy demand onto it.

EPA calls this perspective “Total Greenhouse Gas Emissions by Sector with Electricity Distributed”.

In that perspective, the world looks something like so:

() (s.d.). Retrieved from https://ourworldindata.org/emissions-by-sector

In other words, there are four large areas to keep in the back of your mind

  • Industrial (ex: steel, cement, petrochemical production)
  • Agriculture (ex: cattle methane emissions, deforestation to fit more cattle, NO2 from fertilizers to grow food for cattle, diesel for tractors)
  • Buildings (ex: heating, cooling, water heating, laundry, stoves)
  • Transportation (ex: cars, 18-wheelers, boats, airplanes)

Methane emissions from landfills and sewers gets an honorable mention.

Things like reducing the size of font files on your web properties is, frankly, like polishing the silverware during a house fire.

The plan

We know where the major sources of emissions are. Now what?

Projects like the Pathways to Deep Decarbonization ( & al.) & (n.d.). Pathways to deep decarbonization in the United States.. Retrieved from https://usddpp.org/downloads/2014-technical-report.pdf and Roadmap to 2050 ( & al.) & (n.d.). Roadmap to 2050: A Manual for Nations to Decarbonize by Mid-Century. Retrieved from https://roadmap2050.report from The United Nations’ Sustainable Development Solutions Network (UN SDSN) are typical sources to draw on here. They use a modelling approach called “backcasting” that starts with the end-goal, say net zero by 2050, and then works backwards to today.

Their models tell us we can meet climate goals with existing technology - if we act. Let’s go through the broad strokes.

Electrify everything

Emissions today are distributed: hundreds of millions cars, gas appliances, and industrial processes.

To make the problem tenable, we centralize these sources by putting them on the grid. Replace gas water boilers with heat pumps, gas vehicles with battery electric ones, and so on. Millions of small problems become a few thousand large problems: power plants. ( & al., p. 42) & (n.d.). Roadmap to 2050: A Manual for Nations to Decarbonize by Mid-Century. Retrieved from https://roadmap2050.report

Decarbonize the grid

With the bulk of our power use moved to the grid, the task then becomes to make the grid clean.

Detailed hour-by-hour modelling of the US grid shows we can run it on 90% wind+solar+batteries and 10% natural gas with existing technology ( & al.) & (n.d.). 2035 The Report. Retrieved from https://www.2035report.com/ .

The details of getting to a 100% clean grid apparently remain debated. Chris Nelder argues convincingly that we should start down the path to a 90% clean grid immediately, and work to solve the last 10% along the way. () (n.d.). Energy Transition Show: 5-Year Anniversary Show. Retrieved from https://xenetwork.org/ets/episodes/episode-130-5-year-anniversary-show/

Move non-electrifiable processes to green fuels

Some processes, like steel production and long-distance flying, are hard to electrify. For some of those processes the models suggest we use “green fuels”.

This gets complex and requires care. Green fuels are things like synthetic methane, synthetic methanol and hydrogen. ( & al., p. 42) & (n.d.). Roadmap to 2050: A Manual for Nations to Decarbonize by Mid-Century. Retrieved from https://roadmap2050.report Synthetic methane and synthetic methanol are biofuels, a fuel category fraught with prior disasters. Some biofuels that use food crops as inputs have caused catastrophic deforestation due to their high land use. The idea of biofuels using human food as input has been rightly criticised: Solving the problem of how to feed 9.7 billion people by 2050 is hard enough without a behemoth additional demand source on food crops.

The path ahead would instead include what’s being called “second-generation” biofuels and hydrogen. Second-generation biofuels are those that can be produced by non-food biomass, like biowaste from other industries. Hydrogen can be produced from water using electrolysis, which could run when there is excess supply of wind or solar.

The EU is investing half a trillion dollars into hydrogen projects in the next ten years. Still, given the history here and the novelty of some of this technology, green fuels seem an energy source of last resort.

Improve resource use efficiency

By making processes that cause emissions more efficient, we reduce emissions. This is primarily true in two areas: land use in agriculture and materials use across the economy. ( & al., p. 43) & (n.d.). Roadmap to 2050: A Manual for Nations to Decarbonize by Mid-Century. Retrieved from https://roadmap2050.report

A major portion of emissions in agriculture come from land use and associated deforestation. By finding ways to feed ourselves with less land - increasing yield per acre - we can feed more people without burning forests to create more agricultural land.

Materials is a wide area. Major items include things like metals, cement and plastics. ( & al., p. 43) & (n.d.). Roadmap to 2050: A Manual for Nations to Decarbonize by Mid-Century. Retrieved from https://roadmap2050.report For each of these, part of the solution is finding ways to reduce use in the first place, like better structural engineering software to make it easier to build safe buildings with less steel and cement. For steel and iron, improving recycling is another critical path. Recycling those metals can be done at temperatures within reach of electric arc furnaces, making them amenable to be powered by the clean grid. () (n.d.). Energy Transition Show: Hard to Decarbonize Sectors. Retrieved from https://xenetwork.org/ets/episodes/episode-127-hard-to-decarbonize-sectors/

The timeline

The Intergovernmental Panel on Climate Change, IPCC, gives targets for staying below 1.5C of warming and for staying below 2C of warming. They give the targets as max cumulative emissions, as how much human-caused CO2 can be added to the atmosphere.

To stay below 2C with 50-in-100 odds, we need to stay below 1,500GtCO2 emitted counting from Jan 1st 2018. To stay below 1.5C with 50-in-100 odds, we need to stay below 580GtCO2 emitted counting from Jan 1st 2018 (, p. 108) (n.d.). Special Report: Global Warming of 1.5 ºC. Retrieved from https://www.ipcc.ch/sr15/ .

Because the goals are as cumulative emissions, timelines end up depending on how rapidly we start reducing output. If we address major sources quickly, we buy ourselves more time, and vice versa. This is why you see different rates from different outlets - Climate Clock says we have 7 years because they assume we don’t reduce output at all and just run out the budget.

Reputable organizations like the Rocky Mountain Institute use a timeline from the IPCC 1.5C Special Report to have a 50-in-100 chance of staying below 1.5C. That timeline is as follows () (n.d.). Energy Transition Show: Hard to Decarbonize Sectors. Retrieved from https://xenetwork.org/ets/episodes/episode-127-hard-to-decarbonize-sectors/ :

  • We have until 2030 to reduce emissions by 50%
  • We then have until 2050 to reduce emissions to net zero

This is the timeline Roadmap to 2050 gives a blueprint for.

Summary

When you think of addressing global warming, the broad strokes in 2020 seem to be roughly:

  • Emissions come primarily from four places: Buildings, Vehicles, Agriculture and Industry
  • We know it is possible to meet this challenge by: Electrifying everything, decarbonizing the grid, moving hard-to-electrify processes to green fuels and improving resource use efficiency
  • We until 2030 to get halfway there. Then we have until 2050 to get to net zero.

Where does software fit?

Let’s circle back to where we started, with software. With the pathway by UN SDSN and others laid out above in mind, where does software fit?

Everywhere. Software written by companies like Camus Energy and Schneider Electric Solar Division will be critical to reaching 90%+ renewable penetration on the grid. Software from places like GreenFlux, Polestar and Tesla makes electrifying transport possible. Apps like CCLCalls grow the impact of activists pushing for local, regional and national policy change. Code in smart appliances like what NEEA is talking about here allows shaping energy demand to match wind and solar output.

This blog will cover these and many more, exploring where hackers fit as we mobilize to address global warming. Stay updated by subscribing to the RSS feed or to the mailing list.

If you have examples of software projects or ideas you think fit in the context of UN SDSN’s pathways, or you have thoughts, comments or found an error - please don’t hesitate to reach out. You can email me at jake at davis-hansson dot com.

P.S. If you want to dive deeper right now, some resources I’ve found helpful are:

See also