🔥RTN: can Africa go from energy laggard to climate leader?
+ the end of Moore's Law, humanoid robot factory, Tesla's manufacturing advantage
Can Africa be a leader in clean energy?
Now, I’m half Moroccan 🇲🇦 so please call me out on any biases in this piece.
A few weeks ago, Kenya hosted the inaugural African Climate Summit. Leaders gathered, to discuss climate change on the continent and to keep leaders of rich nations accountable for their actions. Namely, COP15’s promises of $100bn support for African Nations, which so far has failed to materialise.
Late Industrialisation. Africa’s late and less developed 20th-century industrialisation has been a detractor to economic prosperity (energy deployment up until 2010 has been tightly correlated to GDP growth). However, as the world’s industrial and energy base is rebuilt to be sustainable, might that turn into an advantage?
Africa, according to the United Nations, accounts for only 3 to 4% of the world’s carbon emissions but has been the continent most affected by global warming. Some 600m people in Africa are still without electricity, and getting electricity to them will cost $45bn per annum until 2030.
GHGs from richer countries cost the continent $15bn annually in climate-induced natural disasters.
Leapfrogging? Africa has abundant renewable energy potential, especially from solar, wind, hydro, minerals and geothermal sources. Importantly, the continent does not have some of the issues plaguing countries like the UK and US - centralised grids that require scale and huge CapEx, a grid at capacity, multi-year interconnect queues, regulatory permitting burdens and rampant NIMBYism.
Africa can “leapfrog” the installation of carbon-intensive energy generation, much in the same way that they leapfrogged landline infrastructure and went straight to mobile connectivity.
There are a few areas Africa can lead in:
Solar grids: It has 60% of the world's best solar resources but <1% of installed solar PV capacity currently. Africa has an annual average solar radiation between 1,800-2,500 kWh/m2 in North, East, and Southern Africa. The technical potential for solar PV generation is estimated to be 7.9 million TWh per year, over 250x current consumption
Minerals: Africa has large resources of critical minerals like cobalt and manganese will create new export revenues to fund clean energy projects, as these minerals are crucial for batteries, EVs, and other technologies. These have been the subject of much external political and economic influence from Russia and China
Hydro: Africa's hydropower potential is estimated at 1,850 TWh annually, with the largest resources in Central Africa along the Congo River. Only about 10% of this potential has been tapped so far. For example, the Grand Inga project in DRC could provide over 40,000 MW of capacity
Geothermal: There is an estimated 15 GW of geothermal power potential in Africa, located primarily along the Rift Valley system running through Ethiopia, Kenya, Uganda, and Tanzania
Private sectors like manufacturing, agriculture and mining are adopting renewables for self-supply, bypassing centralized grids. We’ve seen successful projects such as;
Kenya scaling up solar home systems and mini-grids thanks to mobile money and pay-as-you-go models, with over 500,000 solar systems installed
Morocco built the world's largest concentrated solar power plant, Noor complex, providing over 500 MW of dispatchable solar capacity
South Africa is a leader in this regard through the implementation of its Renewable Energy Independent Power Producer program – which mobilizes independent power producers to enhance energy access and security
There are a number of interesting startups also; Xlinks, who are building a pipeline from Morocco to the UK to deliver solar energy output. Opibus and ShiftEV are retrofitting ICE vehicles with electric motors, M-Kopa Solar helps the underbanked access and own solar.
Whilst much of the conference was focused on keeping Western leaders accountable, it’s clear that there’s a compelling, broader opportunity for the continent to achieve clean energy independence.
Top Stories
The EV car crash is a warning for Europe’s Industrial transition (FT, paywall)
Warehouse robotics startup, Dexory, raises funding extension (Dexory)
Frog like robot uses combustion to hop along (Spectrum IEEE)
Chinese EV maker, Nio, launches mobile handset (Bloomberg, paywall)
ZeroAvia raises late-stage financing round, partners with Airbus (Flight Global)
SoftBank leads Mapbox fundraising as it pushes into AI (FT, paywall)
Tesla’s ‘Gigacasting’ process marks a clear advantage in EV production (Electrive)
French battery company, Verkor, raises $2bn for gigafactory (Tech.eu)
Rishi Sunak rolls back key UK climate targets (The Guardian)
Redwood Materials acquires European battery recycler (TechCrunch)
PowerUs, a platform for electric and plant mechanics raises €24m Series B (StartBase)
Analysts predict that China will have 1Tw of solar production capacity in 2024 (PV Magazine)
Mobileye CEO discusses potential for '“ChatGPT moment” for Autonomous Vehicles (Mobileye)
⚡️Energy and Climate
Solar installation has increased massively this year, yet is still only 2% of energy generation
🦾 Manufacturing and Robotics
First humanoid robot factory, unfortunately not staffed by humanoids initially
Agility Robotics, just announced the first humanoid robot factory which will open later this year and will be able to produce 10,000 robots a year.
Whilst these multifunctional robots are not super useful today, recent progress has been accelerating
The ultimate goal might be for the factory to be staffed and operated purely by the humanoids themselves
Reviews and critiques of Walter Isaacson’s new book on Elon Musk are plentiful. However, this interesting insight into Elon Musk’s engineering and company-building achievements caught my eye (via slate star codex):
Is Moore’s Law dead and does scaling semiconductor wafer density matter anymore? Huang’s Law
We covered last week, Apple’s new 5nm chips eeked out a paltry ≈25% performance gains, from the prior 7nm node chips
Last week, Nvidia chief, Bill Dally gave a presentation on how Nvidia managed to improve performance of their GPUs over 10 years
He attributed a 16x improvement to number precision and a mere 2.5x to process improvement
Nvidia’s H100 is made on TSMC 5nm process node, same as Apple
This demonstrates an expectation that process improvements to TSMC’s 3nn (which btw is delayed until 202x) is unlikely to yield further marked improvements. The end of scaling is clear as day.