Where’s the uncertainty in making a sustainability solution work? Where is the uncertainty that an innovation process might address?

Sustainability Frameworks

What are examples of sustainability problems? I always look first for a framework. What types of problems are there? What matters? Is there a taxonomy?

One way to approach this would be to survey the “low carbon” investments of major energy companies. Another would be to survey the characteristics of ESG programs of pledging companies.

Top down, I would expect to find a few taxonomies - from US government sources, the EU, the UK, the UN, and from advocacy organizations. The EU taxonomy, at first glance, looks both rigorous and quite burly.

To look at categories of climate impact, “Our World in Data” charts global greenhouse gas (GHG) emissions by sector, provides discussion and a figure. The sustainability problem of course is not just GHG emissions – it’s other forms of harmful emissions and contaminants, and resource depletion, and indeed the whole concept of unquestioned growth.

Is there something simple? We all know the 3 R’s - Reduce, Reuse, Recycle - which always made sense but now seem strikingly insufficient. My R’s for today are: Reduce (consumption, waste products), Replace (non-renewable sources), Restore (systems that sequester atmospheric carbon), and Recover (disruption and disaster response).

From a product perspective, understanding how markets frame value always seems to be what is most salient. For the framework to use here, I’ll rely on how an influential tech investor has characterized the climate-tech sector.

YCombinator, the VC firm, has put out a call for climate tech startups. They elaborate on the investor David Rusenko’s five sector climate-tech framework and diagram.

Five climate tech sectors

I’ll comment on both Rusenko’s framework and YC’s framing in terms of what I see as challenges for executing on innovation in these sustainability spaces. Ruskeno’s description of his framework thumbnails each sector, founder archetypes, and typical risk profiles for investment. Successfully making something work is necessary but not sufficient for an investor success in climate tech. What’s a view of the uncertainty in making something work?

  • This is quite the umbrella, with the unifying idea that component technologies may be already be there - we know how to do this - the problems are getting capabilities deployed.
  • To my reading of Ruskeno’s diagram and from the one from Our World in Data, there are four “Energy” sub-categories:
    • Generation - Replace - New ways to produce and store energy.
    • Transmission - “the grid”, or really the network of grids and operation under dynamic supply and demand.
    • In-place Consumption - by buildings and industry. Conversion to electric. Reduce - process and usage efficiency. Replacing, as retrofitting to new power systems.
    • Mobile consumption - transportation. Vehicles, batteries, distribution of charging systems. Compensating for the energy density of portable fossil fuel with a less energy-dense replacement is a limiting challenge here.
  • Rather than technology, the challenges appear to be execution - market based (cheaper always wins), regulatory, and political – this last in the group decision making sense. If it were software, I’d say the issues were system integration and system deployment. But it’s more: projects with lots of moving parts controlled by different parties. Coordination among organizations along a supply chain. In the case of transmission lines it’s multiple administrative domains of governmental regulation and interest; multiple community and advocate interests of preservation and employment; interests of supply and demand utility companies. In some cases, it’s politicized vs policy-centered processes. I think this is all what Rusenko calls adoption risk.
  • The execution uncertainty of coordinated action is one of achieving alignment of interest (and distribution of returns and impacts). The B2B perspective is “how do independent rational actors agree on mutual self-interest”. Perhaps a way to reduce uncertainty through is pilots – limited projects establishing trust and norms among actors to accommodate new operating models. Understanding the decision making of (and constraints on) utility companies seems to me to be central to making progress here.

“Science required”

  • This is where breakthroughs need to happen. The technology needs to be invented and or scaled. New energy sources. Breakthroughs in generation, consumption, transmission, and the grail, carbon sequestration.
  • Investment stakes are high, $100M to bring to a technology market, and multiple years.
  • Uncertainty - what will it take to get it work? Managing risk here is a classic problem of having enough funding to advance incrementally, and incrementing shrewdly. I think of risk management practice within pharmaceuticals as a potential model for accelerating “private” science.
  • Dispassionately, this kind of investment seems to me to be naturally a component of a long term strategic portfolio of a big company, or to take the form of government-funded research consortia. Emotionally, the idea of impacting a fundamental break through is a model of patronage for this time.

“Climate Adaptation”

  • Rusenko’s first emphasis is on sensors, measurement and prediction.
  • The current use cases – including predicting storms, informing decisions around crops, detection of and impact of trends like habitat loss - all depend on data and processing at peak scales and throughput capabilities. Part of the potential is extending the reach of climate observation and micro forecasting – the global south is under observed relative to the US and Europe.
  • Predicting the weather has always been uncertain. The new uncertainty is that predicting the weather is harder. This is the uncertainty of performing statistical data analysis with a world’s worth of incomplete observations. I wonder that the technology of climate science is a well-developed but not widely distributed capability.
  • Adaptation, further out, includes remediation, hardening, and disaster response. One uncertainty is not so much the technology but enabling the scale of potential deployment. We don’t know what the problems may be yet.
  • Perhaps the uncertainties around progress on adaptation are structural. Where does government-private sector partnership fit? Who will fund these capabilities? What are the commercial framework for extracting sufficient “value” from the scale of measurement and processing needed - to what degree are “adaptation” services a public good? To the extent that large scale organizations have climate information needs or resilience needs that lead or can complement government services, I suppose we will see the market drive.

“Green Fintech”

  • Rusenko puts it: “innovate at the intersection of traditional finance, public finance (e.g. Green Bank), and technology to create purpose-built risk, underwriting, or business models centered around specific “green” project needs and characteristics.” I’ll call that enabling market-based mechanisms for sustainability investment.
  • Organizing a market place for seeking funding for innovation on one hand and investment on the other seems technically straightforward; the challenges would seem to be getting participation and pricing the exchanges right. What makes “green” different?
  • I view ESG as both qualifying access to investment and as organizing investment opportunity. This works in existing exchanges like the LSE.
  • Within firms, there is the question of how can advocates make business case to management for a sustainability project. In fintech, there’s the question of how does a sponsor of a sustainability innovation seek funding outside the firm - how does one make the business case to an investor outside the firm? Are there different “arguments” to be made inside vs outside? VC investment in climate tech is one case of that. Difference between regulated investment (eg w/in purview of SEC) and what VC’s might consider. How does fintech for extraction - mining, energy - represent a model for “Green”?
  • I see uncertainty here as a question of adequately framing the prospects of any investment - measurement, estimation, prediction - where the investment uncertainty reflects the risks of investment of “Energy-related”, or “Science Required”, or “Climate Adaptation” projects. VC’s don’t throw money away, but perhaps they assign as small 1% chance of a return on any one investment. Firms performing well on ESG (George Serafeim writes) perform marginally but notably better than non-ESG committed firms – his argument is that an ESG investment is a well understood prospect. Broadly, I view ESG execution as “reduce” and “replace” using current technologies. In that sense, prospective incremental ESG investments seem much easier to “price” than radical “Science required” projects. To me, the broader uncertainty challenge of “Green Fintech” is one of producing credible estimates of structurally complex and uncertain phenomena – any robust operating model would depend on these.

“Carbon Accounting & Offsets”

  • I understand this sector as enabling market-based mechanisms for measurement and exchange of carbon. To me this looks like “fintech” except with treating carbon as an economic commodity.
  • As well, to the extent that there is a futures component, this looks to me like a mixture of economics and financial engineering and engineering analysis over supply and delivery chains. Perhaps ESG measurement and estimation will impact this area and create some standards and norms.
  • The uncertainty here is the product of causal estimation of unknowns and of procedural and algorithmic and measurement complexity. What’s the correct amount of carbon to offset for a given stage in a given process? What factors are material? What’s salient? What’s real? How much does it matter to get all the details right? On one hand, reporting here seems like a technical eventuality. On the other hand, early-to-market solutions here could be really complicated and not quite explainable. Unlike large language models in machine learning, there’s not an obvious set of baseline or test cases to validate against. But, it may be that technical rigor is not pre-requisite to a market success.

Reflecting on uncertainty in sustainability innovation

It would interesting to look at the climate impact of a given investment sector relative to implementation uncertainty. Are the obstacles ones of technology discovery, lack of information, complexity coordination, absent incentives, missing imagination?

Engineers like me look at potential solutions. Investors look at potential markets. YCombinator - whose motto “make something people want” is to the point - suggests that a climate-tech startup focus on B2B relative to B2C, both in terms of scale of potential impact and as a reflection of how B2B customers frame their needs.

A signature challenge of “Energy Related” sector is that different people (organizations) want different and apparently inconsistent things. As an engineer, it can be easier to conceive of technology breakthroughs than it is getting groups of self-interested actors - with partial and selective views of a situation - to agree on a solution of mutual good. A structured decision making process – like Decision Analysis – is a way through this quandary when participants are willing to engage.

Innovation is moving forward under uncertainty toward the creation of value.

One dimension for innovation is technology-centric. Creating that first instance of a solution to a hard problem, and then finding ways to scale that solution in a viable way. We can estimate and balance the costs of exploration against the benefits of discovery. Tools can help.

The other dimension for innovation is group decision processes. How can groups of mixed interests and diverse understandings come to a consistent view of a problem in a ways that allows for progress to be made in good time? This sounds like public policy-making – and that challenge there is how can a participatory and rigorous process also proceed equitably and expeditiously? Tools and facilitation can help, but this is a problem of mindsets and skillsets. It’s a process that depends on all actors bringing the ability to consider interests, uncertainty, and outcomes from multiple perspectives. This also depends on an idea of shared values, for the win/win. Perhaps the shared value is how do we all not lose/lose under a degraded environmental future.