Last time I brought up the Boring Company, I discussed some of the issues with the silver bullet thinking behind the project, similar projects that failed in the past, and the effect that the project could have on its urban environment if it were built. Since then the plans for the first phase of development have been revealed to the public. Today, I’d like to delve into the technical reasons why the Boring Company can’t solve LA’s traffic woes, and what the service is likely to look like in the future assuming it is built as advertised.
The tunnels have low capacity
We can start by estimating the ideal capacity of a single tunnel. There are two things we need to know to figure this out—the sled speed and a safe stopping distance that should be kept between adjacent sleds. Stopping distance is dependent on the sled speed and the g-force that riders are willing to sit through. (Normal stopping distances as on a freeway also need to include human reaction time, which we presumably don’t have to worry about here.) Mr. Musk claims that the sleds have a top speed of 120 mph, and for the following discussion I’ll assume that the sleds will stop by decelerating at a constant 1 g. This means that our sleds could go from top speed to complete stop in a bit under 5.5 seconds, and our stopping distance is 480 feet.
For now, we will also ignore how cars enter and leave the tunnel. We are just looking at maximum possible capacity given an operating speed of 120 mph and without having to make passengers uncomfortable (by potentially stopping very quickly) or unsafe (by not leaving the full stopping distance between adjacent sleds). 120 mph is 176 feet per second, which means that with a distance of 480 feet between sleds, we can put one sled through the tunnel every 2.73 seconds, for a total of 1,320 cars per hour per tunnel. Compare this against realistic estimates of freeway capacity: Caltrans estimates that one lane on a multi-lane freeway has a capacity of around 2,300 cars per hour.
There are other complications we haven’t considered yet in order to make the math easier. Our model basically assumed cars appear at one end of the tunnel, go straight, and disappear at the other end. In real-world use, cars will need to be loaded into the tunnels and taken out of them, and these processes are likely to take longer than 2.7 seconds. It’s also not clear from the plans whether cars all enter and leave at the designated “stops” or whether there are frequent exits and on-ramps, as in current freeway design. If it’s the latter, we are going to have big issues with merging and this is going to reduce the sled speed from the promised 120 mph. If it’s the former, then not only this is not the panacea for traffic that Mr. Musk claims it is, but it’s also going to have all of these problems anyway when the system expands beyond the two “stations” planned in the first phase. The supply and distribution of sleds will also be difficult; bikesharing companies are already running into similar issues today and need to do out-of-band redistribution of bikes to ensure that people won’t arrive at a dock and find that there is no bike to take, or that sleds have somewhere to go when they arrive. This means the company will probably have to deadhead sleds around, which reduces the capacity even further.
Mr. Musk claims that we can fix this problem simply by building more and more tunnels. One mile of six-lane urban freeway costs around $11 million according to ARTBA. To get similar capacity out of Boring Company tunnels, we would need to build at least 11 tunnels. The 405 which Mr. Musk wants to build under can have as many as 12 lanes. Even if he achieves his goal of reducing costs by a factor of 10 and boring tunnels at $100 million per mile per tunnel, that’s still over $2 billion for one mile of equivalent capacity. (I estimate that the proposed network covers about 60 miles, putting the cost in the hundreds of billions even in the best case.) Given that the 405 is already way over capacity, we are going to be making absolutely massive capital expenditures for little ability to move people around. But is Mr. Musk actually trying to solve issues of mobility?
Users pay for entry either in money or in time
High-speed, low-capacity transportation isn’t a new idea for Musk as he has proposed essentially the same thing with Hyperloop. This technical discussion of capacity issues on Hyperloop was an inspiration for the post you’re reading right now. If we open up a new system which allows you to get somewhere twice as fast as the current system, we should expect that lots of people will want to use it. Given the state of LA’s freeways and in particular the 405 as stated above, I think it’s safe to assume this demand will quickly surpass the low capacity of the tunnels.
Freeways already have capacity problems and work around them in one of two ways—either by charging tolls, or by forcing people to wait to get on and sit in traffic when they have managed to do so. In other words, given that the freeway has limited supply and high demand, we can either raise the cost or “ration it” and pay in time instead. If the 120 mph ride would save people an hour, they’ll be willing to wait almost that long if the price was low. Given that Mr. Musk first proposed this idea because he was frustrated about sitting in traffic, he would probably bristle at his company gaining a reputation for simply exchanging one type of waiting for another. So, we should expect the price of entry to a Boring Company tunnel to be very high.
This might be a great business plan, but it is a very bad way to move around large numbers of Angelenos.
Personal vehicle use and high capacity are at odds
The ultimate issue here is, of course, that we are still trying to move people around in cars. At 1.3 people per vehicle, a freeway lane moves about 3,000 people per hour. Boring Company tunnels will move around 1,700 people per hour in absolutely ideal conditions.
Our other option is obviously real high-capacity transit, which LA Metro is currently studying along the corridor that Mr. Musk wants to build his tunnels under. Metro is considering both light and heavy rail for this corridor. The Kinkisharyo P3010 has 68 seats and Metro assumes a load ratio of 1.3, so they would expect a typical 3-car train to hold 265 people. During commute hours, the number of people per car can be higher. At a full load, we might expect 150 people per car; at crush load, perhaps 200. Since this corridor is so heavily traveled, we can assume they would run on headways of at most 5 minutes, bringing us to 12 trains per hour. This is 5,400 people per hour at full load and 7,200 at crush load. Without building any extra tunnels or track, we can trivially decrease headways to 2 minutes, putting us over 13,000 people per hour. If Metro goes with heavy rail, train length goes up by a factor of three, and capacity goes up by at least that much.
Mr. Musk thinks that the big issue with grade-separated transportation is the cost of building tunnels. Even if he achieves the cost savings he promises, we would still get less capacity per dollar than we can now. Ironically, in (correctly) identifying that our current capital projects waste a lot of money, he has managed to find a way to make them even worse.
The tunnels combine the worst features of freeways and railways
We’ve already gone over how these tunnels are quickly going to run into capacity constraints, like freeways. They also contain vehicles that move on rails, and incur all of the maintenance issues that come along with that. The most obvious is rail maintenance; they will probably have to shut down the tunnels every night until there are several of them, at which point they might be able to stay open with reduced capacity.
Additionally, the Boring Company will have to design freeway structures on rails, such as interchanges. If they don’t want sleds to stop, they are going to have to build interchanges at every junction. This is much more complicated if they plan to have multiple “lanes” because they will either have to create flyovers connecting each “lane” to one in the perpendicular direction, or will have to allow sleds to move between “lanes,” creating all of the problems freeways currently have with people merging into the right lane to take an exit. This has further implications for capacity and speed, especially since sleds can’t merge at arbitrary points. They have to wait for a switch, just like a train.
They will also have to build station cavern equivalents to accommodate all of the cars moving between the surface roads and the tunnel network. Because the headways are so short and the dwell times of the sleds are comparatively long, these station caverns will have to be many times the size and complexity of current rapid transit station caverns. Commuter rail terminals for large networks already suffer from this issue, as showcased by Penn Station for the Long Island Rail Road. There is no through-running and trains dwell for a long time, so they come in faster than they leave during morning peak hours. As a result, Penn Station has to have an incredible number of tracks to handle the load. Additionally, Musk wants to build tunnels layers deep when he needs to increase capacity. Station caverns are perhaps the most expensive and time-consuming components of building subway lines right now, and they only get more expensive and complicated to build when you make them bigger and deeper.
We shouldn’t expect good ideas from someone who owns a car company
If Mr. Musk truly thought these tunnels were a cost-effective and sustainable way to get people around LA, it would show that he hasn’t done his due diligence. (Perhaps he’s too busy getting into Twitter fights with people who actually study and work on transportation issues.) But while I personally don’t consider most of Mr. Musk’s ideas that interesting or forward-thinking, I have a hard time believing that he’s stupid, and he does have good business sense. It should be clear that the aim of the Boring Company is not solving urban transportation problems. It is a member of the car industry digging its heels in and trying to stay relevant in a world where people are realizing cars can’t provide us with the transportation we need.
Mr. Musk owns Tesla Motors, which manufactures electric cars. Tesla’s claim to fame is that the cars have a good deal of range, and a lot of people like them. To ensure the success of Tesla, Mr. Musk has to ensure that the car industry continues to survive—it’s now well known that he has a personal hatred of public transit, but he also has a powerful financial incentive to stem the tide of transit development in places like LA. This is also why he is investing so heavily in self-driving technology. I’ve talked about this in a prior post: self-driving cars would not help move more people around, either, but it would make people willing to tolerate much longer commutes. The owner of a car company is trying to come up with “new” ideas for transportation, and it should come as no surprise that he thinks the biggest problem is that he’s not selling enough cars.
Consider that one of the ways Musk plans to reduce drilling costs is by reducing the diameter of tunnels to 14 feet. This makes them unusable for mass transit vehicles, which are almost that tall on their own. If we try to use the tunnel as a busway, we will probably need around 11 feet of flat “lane” for the bus to drive over. The bus itself is also at least 8 feet wide. A reasonable estimate of tunnel wall thickness would be 8 inches, and so this works out to a usable clearance of at most 8 feet, which is not enough for a bus. Narrowing the lane to 10 feet provides an extra 8 inches of clearance, but this is still not enough, and trolleybuses are right out. The aforementioned Kinkisharyo P3010 cars stand 12.5 feet tall and along with the tunnel walls, floor, track, and overhead wires, they will easily exceed the 14 foot diameter of the tunnel. (For comparison, Central Subway tunnels meant for similar light rail vehicles are 20 feet in diameter). Breda A650 heavy rail cars stand at 12 feet and will not be able to fit, either. These tunnels will only ever be useful for cars.
In the past, private transportation infrastructure could be acquired and run by the municipal government, increasing the availability of affordable rapid transit. Many New York City Subway lines were at one time run by the private Interborough Rapid Transit (IRT) Company and Brooklyn-Manhattan Transit (BMT) Corporation. Mr. Musk has effectively prevented the same thing from happening to the infrastructure he wants to build, ensuring that it won’t have the opportunity to become widely useful.
The Boring Company claims it will be able to provide cheap, fast, and convenient transportation for everyone. This isn’t possible given its technical details, but it was never the point, either. Musk is in this for the money. He wants people to continue to buy his cars, and to pay lots of money to ride around in his tunnels. The obvious consequence is that we will have a two-tiered transportation network where the vast majority of Angelenos won’t benefit, but a small group of very wealthy folks will be able to move around at will.
If you are not one of these wealthy people, you will not see any improvement in mobility if the Boring Company builds out their proposal. Elon will make lots of money while the streets fill up with even more cars. We risk keeping LA from getting the quality mass transit system it desperately needs if we let him distract us any more.