Tweet thread with my thoughts on 100% renewable vs 100% zero carbon modeling.
1. All models are wrong. Some are useful (Box).
2. Having followed the modeling discussion about our low carbon future I am increasingly persuaded that the current modeling approach in the literature is reaching peak “usefulness”.
3. By that I mean that they are not particularly helpful in a) predicting what will actually happen and b) furthering productive discourse.
4. My rationale: a) Most models are greenfield, b) mainly driven by assumed generation costs and c) assume for simplicity that the system under study is centrally planned.
5. My assumption is that this is because the point of the exercise I to see if a particular low carbon mix can deliver a historic demand profile at a reasonable price.
6. This is a reasonable approach if you are trying to determine if a particular generation mix/demand profile can be implemented and explore variations to either the mix or the demand profile.
7. Spoiler alert: Yes it can – just depends on the underlying assumptions.
8. Now we are using the same models to discuss which low carbon mix is the most suitable/desirable/most cost effective. I am personally interested in the discussion about whether nuclear will be “required” in the low carbon mix.
9. In particular, the concept that the aggressive deployment of wind and solar PV now, may be detrimental to the overall transition because it will crowd out the space that nuclear “needs” and delay nuclear deployment to “later”.
10. I don’t believe this can be resolved by continuing the current modeling approaches, which appear to be entrenching two camps. I’d suggest that it is timely to consider alternative (modeling) frameworks that might assist in furthering the current discussion.
11. My assumptions:
a) We operate in a “market”, not a system.
b) We are in a transition (brownfield not greenfield. Existing actors will fight to survive.
c) The transition is not coordinated. The actors respond to a variety of market signals at the current moment in time.
a) We operate in a “market”, not a system.
b) We are in a transition (brownfield not greenfield. Existing actors will fight to survive.
c) The transition is not coordinated. The actors respond to a variety of market signals at the current moment in time.
12.
d) Responses are not always economically rational.
e) The technology choices of the past do not mandate the choices of the future.
f) The market will drive innovation.
g) Innovation will out.
d) Responses are not always economically rational.
e) The technology choices of the past do not mandate the choices of the future.
f) The market will drive innovation.
g) Innovation will out.
13. I would like introduce the concept of “weedy technologies”. These are technologies that are scaleable (distributed), fast to deploy, enjoy double digit cost reductions year on year, have low operating costs and capture the public’s imagination.
14. Weedy technologies will out compete all others in the market. In a market based environment, it is not possible to slow the deployment of weedy technologies once they get hold – just ask Nokia or Blackberry.
15. Weedy technologies will proliferate until they dominate their market niche. The market will inevitably adapt and innovate around this proliferation, particularly in response to market price signals (such as low midday generation prices).
16. It is difficult to predict what this innovation/adaption will be. “Difficult to predict” is not “leaving things to chance”. It is an understanding that we exist in a technologically and politically fluid environment that cannot be adequately described by building models.
17. Right now, what will help most in understanding the low carbon transition is to identify current weedy technologies (such as wind, solar PV and batteries) and model the likely extent of their deployment and effect on the market.
18. This should enable us to identify new potentially weedy technologies that will be part of the innovation matrix. Technologies that are scaleable, rapid to deploy and at the beginning of double digit cost reductions – like H2 electrolysis.
19. Taking this view: Wind and solar solve the generation problem, storage and demand management manage intermittency and H2/NH3 strategic reserve secure seasonal variation. End
Currently cruising the Straits of Malacca with limited bandwidth, so apologies if replies are delayed. Weediness is of course a concept used in other scientific disciplines (especially biology) so I have just translated the idea across disciplines.
