Will TESLA’s model 3 crash the grid and the internet?
by Ted Lewis
As I write this, Tesla Motors has racked up 325,000 preorders for its model 3 all-electric semi-automatic, digital car. Because preordering requires the buyer to deposit $1,000, consumers lent $325M, free of interest, to a company that is still two years away from delivering its first model 3. Consumer demand is so strong that Tesla now faces the challenge of building and selling cars worth approximately $14B. Compare this with Toyota, a car company that sold 207,372 Prius hybrids in 2014, a 11.5% decrease from 2013. Sales of the Chevy Volt don’t come close to matching Tesla pre-orders, and the Nissan Leaf lacks the range and power to be a contender. It seems like a landslide victory for Tesla.
But, here is my main concern — if Tesla ramps up deliveries too fast, it may bring down the Internet. What happens to the Internet when Tesla delivers?
The Internet-Grid-Car Entanglement
Everything digital depends on the electric power grid, and soon electric cars will too. My smartphone consumes a watt or two of electricity, but my Tesla consumes thousands of watts. Adding cars to the grid can have a much more dramatic impact than adding a smartphone. In fact, adding too many electric cars too fast, can bring the grid down. If the grid fails for more than about a day, so does your digital ecosystem. I call this the Internet-Grid-Car entanglement, for obvious reasons.
Let us assume Internet users like me plug our desktop, laptop, and smartphone into an outlet to either power them directly or recharge them at least once per day. So, if power runs out for more than a day, we are likely to be banished from Facebook, Twitter, Google, and Amazon. Even a brown out lasting for 24 hours has a detrimental effect on the Internet, because our Internet devices can’t be recharged. [I am not worried about the big servers in the cloud because they have diesel backup power that will keep them running for more than a day or two].
Most of us take for granted that there is, and always will be, an unlimited and continuous power supply. Adding a one or two smartphones or laptops to the load doesn’t make much of an impact, because digital devices use very small amounts of power. Even adding millions of digital devices have a relatively small impact on the grid as compared to a car with a big battery. Add a car like the Tesla, with its 70–90 kWh battery to the grid and you have increased the load by a factor of 100. Multiply this increase by 325,000, and you have a problem, because electric cars can consume all of the available power, making it impossible to recharge your smartphone and laptop batteries, which in turn leaves you sitting in the dark without Internet access. This kind of entanglement leads to the kind of unanticipated consequences that journalists write about after they happen, but nobody predicted.
The Back of the envelope
Now for the numbers. It is realistic to assume most Tesla sales will be in California. The HOV lanes of California’s freeways are already jammed with electric cars, and Californians want to be as green as possible. The Western Power Grid, with maximum generating capacity of approximately 160 million kilowatts, powers these cars, computers, and air conditioners. Brown outs and spot congestion problems begin to occur when the load reaches 140 million or more kilowatts. So we can use this number as a starting point for a back-of-the-envelope calculation on the likely impact of adding a massive number of electric cars to the load.
Now assume each Tesla model 3 requires an overnight charge of 70 kWh. The load on the grid will depend on the recharge rate, how empty each battery is, and how many cars are simultaneously recharging. If charging takes place over 2 hours, then the drain is 70 kWh per 2 hours or 35 “kWh per hour”. Thus, four million Tesla cars, recharging during the same 2-hour window, can consume all of the available electrons in the grid. Let us further assume that only half as many plug in simultaneously. Two million may seem like a lot, but it is less than 10% of the 28 million automobiles registered in California in 2014. Furthermore, perhaps less than ½ or 1 million Tesla cars will require refueling every 24 hours. On the conservative side, suppose only the pre-ordered 325,000 Tesla model 3 cars attempt to recharge at the same time. What happens?
Recharging 325,000 model 3 cars in two hours, each with 70 kWh batteries consumes about 8% of the 140 million kW available from the Western Power Grid. Most power engineers agree that a 10–20% surge in load can put the grid on the edge of a tipping point between stable and unstable operation — depending on when the surge happens. [Daily variations of demand on the order of 80–100% are not uncommon, but these variations are within the bounds of the 140 million kW capacity. A variation that exceeds 140 has historically crashed the grid].
If Tesla delivers all 325,000 cars at once, and the owners simultaneously plug them into the grid to recharge on the first evening after driving them home, expect a massive black out in some regions. But, Tesla is unlikely to deliver them all at once. In fact, Tesla has historically been slow to ramp up production of the model S, suggesting that the model 3 ramp will take years. This case is the most unlikely.
Assume a second scenario (case 2) where Tesla delivers 100,000 model 3 cars per year, for 5 years. That is an increase in load of approximately 4.3% per year. Utility companies currently struggle to keep pace with population increase amounting to 1.0–1.5% per year. To meet the challenge presented by Tesla, they have to ramp up the production of electrons by a factor of 3. This is pushing the limits of the 100-year old western power grid. It can fail due to spot congestion, heavy up and down loading, and unanticipated demand. In short, it is rather easy to crash the grid. Adding 7,000 mw of additional load every year over the next 5 years is going to be a problem.
Adding electric cars to the grid too quickly can bring down the electric power grid — an undesirable side effect of going green. How feasible is this scenario, and can it be avoided? Several other scenarios are possible (case 3 and 4). First, solar panels are replacing coal and gas-fired power plants rather quickly in California. The state has a mandate to become 50% green by closing power plants that emit CO2 over the same time frame that Tesla is filling preorders. A final scenario (case 4) involves an ambitious plan to get entirely off the grid. Tesla sells a battery for domestic use, which can smooth demand. If consumers own both panels and storage battery, they may be able to become self-sufficient and sever all connections with the grid. If grid demand declines by 4–5% per year due to these mitigations, then Tesla can install even more green cars in driveways across the West. Otherwise, Tesla consumers may inadvertently crash the grid, which has a domino effect on the Internet and everything digital. I invite you to do your own back-of-the-envelop calculation, and then go out and buy solar panels!
Edits: this version has been improved based on feedback.
Dr. Ted Lewis is the former director of the Center for Homeland Defense and Security and an expert on Critical Infrastructure Protection.