On the morning of December 2nd 2014 the power went out for ~100 buildings in downtown Detroit. People were stuck in elevators and railcars. Flights were grounded as the city airport also suffered from the outage. Traffic lights stopped working. A few hospitals were without electricity leaving patients plugged to devices that weren’t functioning as required. All city fire stations were without power. The Wayne county jail (with its own backup power) kept running as usual. A spokesman for the Mayor said the outage was ‘caused by extreme heat, cable failure and routine maintenance — all combining causing system overload. System shock starting in one location leading to total system failure.
On August 14th 2003 about 100 power plants across Northeast US and Canada stopped functioning, resulting in a blackout that affected ~50M Americans and Canadians. 21 power plants providing power to ~10M people went down in 3 minutes. 3 minutes. 100 power plants went out over the course of the blackout with 50M people affected in total. People stuck in elevators, subway systems and without water due to non-functioning water pumps starved of electricity. All caused by a downed power line that was not reported because of a software bug that was not noticed by a control room engineer who couldn’t see there was a surge on the system. I take that back. All caused by our reliance on an electric grid with archaic machines, systems and processes unfit for the modern day requirements and capabilities of a power dependent populace. A non-adaptive system with one point of failure. System shock starting in one location leading to total system failure.
Right now (as I write this post) across the US there are at least 23 power outages affecting ~300k consumers (see image below) including Louisiana.
Fragility of the grid: Our grid is more fragile than we realize. This became more apparent to me after reading Gretchen Bakke’s ‘The Grid’. But this is an optimistic post. As fragile as the grid is, we already have the technology required to make it antifragile. It starts with seeing the possibility that occurrences like the acquisition of Solarcity by Tesla, even if it does not fully make financial sense right now, provide to the possibility of bringing grid resilience. But what is Anti-Fragility?
All credit to the concept of antifragility goes to Nassim Nicholas Taleb.
Seemingly abrasive and disdainful of people who falsely claim expertise, in his view, Nassim Taleb gave the world a new way to frame how systems that get stronger and benefit from chaos and volatility. He gave us a word for the opposite of ‘fragile’; Antifragility is beyond resilience or robustness. The resilient resists shocks and stays the same; the antifragile gets better. Even with this simple understanding of antifragility it’s easy to see that the electric grid system I’ve described in the two examples above is fragile. A problem with the approaches being taken to fix our fragile grid is that we are focused on building a resilient grid, on building a grid that is capable of recovering quickly from difficulties or failure. This is the wrong approach. We need to build a grid that is antifragile, one that grows stronger with the stresses and the chaos. But how does Tesla/Solarcity get us there?
Antifragile grid: Amidst all the noise about autonomous vehicles what the pundits miss is the antifragile system capacity that comes as a result of many autonomous vehicles operating for a common purpose that are not centrally controlled. This concept is called swarming. In laymen’s terms; every autonomous vehicle that is driving thousands of the miles on the road is gaining knowledge, through machine learning, that improves the collective knowledge of every autonomous vehicle out there. When one autonomous vehicle experiences a shock, say an accident, even though the vehicle might have a little damage, the collective whole (the autonomous driving knowledge database) learns and improves. One vehicle, a node in the system, might have failed, but the whole system (adaptive system) has grown stronger from the shock to that node.
Now replace ‘autonomous vehicle’ node in the transport system with a ‘Tesla roadster/Tesla Powerwall/Solarcity panel’ node on an antifragile grid. The same antifragility can be achieved, where a shock to one Tesla/Solarcity node does not cause the whole system to fail. In fact, the whole grid learns from the failure of that one node and grows stronger allocating resources dynamically.
The Tesla battery pack and Powerwall, the real reasons why Elon is working on a gigafactory, combined with the solar panels from Solarcity start to move us towards this antifragile grid. The whole system will be managed by software (built by partners and competitors) that knows when you need power, where to pull it from (whether battery, solar panel or Powerwall) because of the price signal, seamlessly ensuring that you get your hot coffee to start your day or hot chocolate on a cold night.
To take the example further, what this looks like in the future;
- One home with a Tesla/Solarcity system loses power.
- the antifragile grid identifies the reason for the failure (from its database) or adds the failure to its knowledge base.
- no other node on the grid fails, the system continues to function
- the antifragile grid learns from the failure and implements a fix/self-heals.
- the antifragile grid determines where there is excess capacity (a neighbor’s vehicle-to-grid enabled Tesla roadster) and powers up the failed node.
- The antifragile grid has become a stronger and smarter system waiting for the next failure to learn from…
In 1955 the first home microwave oven that was introduced to the market by the Tappan Stove company cost $1295 [$11,471.91 in 2016 $’s!!!].The price of a Tesla/Solarcity pack will come down to a point where it comes down to a price point that can be considered one for the mass market, it’s the path of most new consumer technology. The Tesla/Solarcity deal, which a lot of people who understand financial statements but do not understand the state of the grid are resisting and boohooing, starts to help our grid develop its ability to grow stronger from the shocks and chaos that comes with the adoption of more intermittent sources of energy like wind and solar. Yes, not everyone can afford a Tesla or solar panel right now, but we all know that will change.
The design is the flaw in the current utility system. What needs to happen is a full redesign of the grid to make it antifragile. What will we need, beyond the Tesla/Solarcity deal coming through to make this happen? We’ll need Elon Musk/Solarcity and his crew to open source some part of this system software, this will enable the tinkerers and developers out there to build AI and machine learning software at a scale that Tesla/Solarcity itself cannot build. We’ll need the utilities, who now have the chance at a new business model and a role to play in this antifragile grid future (where they didn’t have one before), to become a lot more transparent about how the infrastructure part of the current grid functions; this allows the Tesla/Solarcity’s of the world to ramp up their understanding of what it takes to be a provider of stable, secure and affordable power.
Bottom line? We all need this Tesla/Solarcity deal to come through, probably more than Elon Musk… For us to avoid system-wide power outages, that cripple humans and businesses, we’ll need companies that jump into providing the critical elements of renewable energy generation, storage and consumption management. At scale. But without the complexity of the present day grid system.
Considering the size of Elon’s ambitions, that’s hoping for a lot…