In order to have a market-driven high-speed rail system, there has to be a logical starting point. During meetings with AMT and other rail providers, I’ve consistently heard that what they’re looking for is a starting point: a short segment of track that could carry a large number of people and provide TOD opportunities. Such a line would serve to ‘prove’ the idea, which would allow future systems to be financed at more affordable rates.
In Houston there’s one candidate that I believe is the strongest ‘pilot’ for high-speed rail service: the South Corridor. Originally I was more interested in going Southeast or North, thinking how helpful it would be to me personally if there was a rail connection to the airport. However, Mike Rozell (Director of Economic Development for Harris County) suggested a different approach. His idea was to go south from the Texas Medical Center to Pearland. He cited the significant traffic congestion in that corridor and observed that, unlike other corridors in Houston, the people in Pearland overwhelming work in one location – the Med Center. Additionally, this route is short, and it offers enormous Transit-Oriented Development (TOD) opportunities as it travels through one of the least developed quadrants of the city on its way to a thriving suburb.
All these arguments convinced me, and I now would argue that this is the best place for a system to start as well. I’ve illustrated my conceptual alignment in the simplified graphic to the right, you can also download a much better PDF (34 MB) with underlying aerial photo.
The Alignment
The line would start in the Med Center, traveling from the center to the VA Hospital at Almeda Road. From there it would run down Almeda Road to Holly Hall. This initial segment could open and begin collecting fares as a shuttle service while the rest of the system is being built.
From Holly Hall the line would continue south, passing through the mostly vacant south 288 corridor where there are huge opportunities for TOD. After that it would continue to FM 518 in Pearland. After the first segment is operational the line could easily be extended into Downtown, and eventually the line could continue to new development sites in Manvel.
The train could be designed with a top speed of 150MPH. The interesting thing about a commuter alignment, however, is that the acceleration ends up being the biggest limiter to the performance of the train. At a moderate acceleration rate of .15g, the train needs about 1500 meters to reach full speed. That means when stations are less than ~2 miles apart the train can’t even reach full speed before it has to decelerate again. The stop time also makes a big impact on the total time. All that said, the train is still blazing fast. Here’s a table of the travel times from point to point:
| Station 1 | Station 2 | KM | MI | Minutes | Seconds |
|---|---|---|---|---|---|
| Downtown | Baldwin | 1.57 | 0.98 | 1 | 25 |
| Baldwin | Wheeler | 1 | 0.62 | 1 | 12 |
| Wheeler | Binz | 1.51 | 0.94 | 1 | 24 |
| Binz | TMC-E | 1.5 | 0.93 | 1 | 24 |
| TMC-E | HollyHall | 2.31 | 1.44 | 1 | 39 |
| HollyHall | South City | 3.81 | 2.37 | 2 | 25 |
| South City | City Park | 4.21 | 2.62 | 2 | 31 |
| City Park | BW8 | 2.7 | 1.68 | 1 | 46 |
| BW8 | Pearland | 4.55 | 2.83 | 2 | 36 |
So, in total, the trip from Pearland (FM 518) to Downtown would take 16 minutes and 23 seconds. From Pearland to the Medical Center (Central) would take 12 minutes and 24 seconds.
As mentioned before, AMT’s technology is built on the idea of a larger number of small trains running close together. So, in the mornings and evenings there could be a train every few minutes. When the line is complete trains would alternate between Downtown and TMC-Central as their final destination, but all the trains would stop at TMC-West. In the middle of the day, as traffic demand is less, trains could run about every ten minutes in the Downtown to Pearland corridor, with more frequent (shuttle) service traveling between TMC-Central and TMC-West.
The Economics
The cost for AMT’s system is (based on the test track they’ve built) is $15-17 million per mile, so lets assume $17 to get a conservative cost estimate. At $17M per mile and 15.4 miles of track (Phase 1 and 2) we get a system cost of about $260M. That’s significantly less than the projected cost of any of the planned light-rail lines. At that cost the system would require about $52M in private equity to be built, would require about $8M to fund that equity (at 15% discount rate) and about $15M to cover debt service (at 6.5% interest). That means about $23M per year in income to cover construction.
Now remember, this first project is intended to be a demonstration line, and also to be an asset to joint development of several TODs along the way. So the goal is simply to fund the construction loan during the first year or two while chunks of the TOD sites are being developed and/or sold. Those big chunks of income from land development are what is eventually going to pay off the debt early and allow the train to be profitable in the long run.
So the goal in the first few years is just to cover the capital cost (which includes the rolling stock) and keep the ticket price low enough that ridership will grow – then the operating costs are covered and profit is made as the TOD pays off the debt. (UPDATE: For more on the operating costs, see comment #2 below). The key is to prove the technology works and demonstrate ridership numbers.
According to TxDOT’s Statewide Planning Map, there were somewhere between 90k and 120k daily trips made from the vicinity of Beltway 8 and SH-288 in 2008. There were about 140k daily trips along 288 from Reed Road to the Loop, and about 175K trips made from the loop into Downtown. If we assume that a maglev train which is much faster than driving can capture about 10% of the daily traffic in this corridor, that would mean that around 15,000 daily trips might be served by the maglev. This doesn’t seem surprising when we consider that the Main St. Light Rail carries about 40,000 people a day.
If we assume that 15,000 passengers on weekdays and 5000 on weekends, that’s about 4.4M passengers per year, and a debt service of $5.30 per passenger. Let’s round up to $5.50 for good measure. The important question now is, would you pay $5.50 for a day-pass on a high-speed train that would take you from FM 518 in Pearland to the middle of the TMC in 12 minutes? By comparison, that’s 30-40 minutes in traffic.
That fare is a little high, but it’s less than the commuter bus service METRO offers in other corridors. Plus, there are two major things that could push it down quickly.
First, if there were higher ridership (due to a combination of marketing, future development and increased demand), the fare could be less. If we assume around 15% of the trips to the TMC and Downtown are captured, the ridership would be around 5.2M per year, or about $3.00 per passenger to cover costs. That’s a big, big difference.
What if we stick with the original ridership estimate but use a $50M in grants (from corporate sponsors, city, county, and state sources) instead of private equity. With $8 million less in amortized capital costs each year, the debt service per passenger would be about $4.00.
What if we had both higher ridership and grants? Now we’re looking at about $2.35 per passenger. Round it up to $2.50 and you’re still under half the price of the commuter bus service ($3.25 one way, $6.50 round trip).
Now, I can’t tell you which scenario would happen – that will have a lot to do with who is involved in a deal and what level of support the local government chooses to provide. The point is, this line makes sense, and is probably viable even without any public-sector concessions (except permission to use existing ROW without obstructing traffic).
Looking Ahead
I believe this concept is a great starting point for a prototypical commuter line employing the same basic technology that I would propose for inter-city rail. This “first-step” is critical, because the best way to make inter-city rail practical is to double-up with regional commuter service. Also, for a project of this size to be taken on by private enterprise it needs to be broken up into manageable chunks.
Over the next few weeks I’ll continue to look at these segments, first with some of the commuter framework in Houston, then in other Texas Cities, and finally how it all connects together into the mini-triangle plan I’ve already outlined.
I won’t cover the other segments in as much detail because it takes forever (it took me a week to write this post), and frankly I’m doing this on my own time. However, the same basic logic would apply to all the segments. Once you’ve got a starting point, each additional segment is incrementally more useful because its part of a bigger and bigger network. Also, if you’re making money on the first line, the rest become much easier to develop.
I’m interested to hear your thoughts and feedback on this idea, let me know what you think by leaving your comments!
Categories: featured, move
Tags: 288, AMT, high speed rail, Houston, maglev, South Corridor, Texas
Share:
Topics
License and Privacy
- neoHOUSTON design by Andrew Burleson
All neoHOUSTON articles are licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.- For commercial uses of the works listed on the site, please contact the post author to obtain permission.
- For privacy information, please click here.
13 Comments
Permalink
Your fare of $5.50 only covers cost of construction funds. What about operating expenses?
Permalink
Thanks for asking rather than ranting
As I was saying above, it’s much more difficult to model the operating expenses as they’re variable (price of electricity) and strongly tied to demand. However, because this system is driverless and largely computer controlled (there would be a human operated command center but the computers would do most of the work and the staff is primarily to monitor things and respond if there’s a problem), the operating costs are quite low. Also, remember that the trains are rolled into the capital cost and amortized, they’re not part of the operations budget.
So, it’s reasonable to assume that you could cover the operating costs through a number of other sources such as advertisements, services (like restaurants in the train stations or fees from car-shares attached to the stations), and other means. For instance, the rail conduit could fairly easily double as a high-voltage electric transmission line, in which case the service could realize some income from that.
Lastly and most importantly, if this is to be done as a private-sector endeavor then the train operator will have a stake in the TOD that happens around the stations, and that will generate a significant chunk of money that would cause the debt to be paid off very early. Without the debt load the fares would be going 100% to operations and profit – and expanding the system.
So, to sum up, I didn’t model the operating costs because they’re really complex and hard to model accurately, because the ways in which they could be recovered are numerous, and because in all honesty I don’t have time.
What’s critical to me is to find out how much it would cost per passenger to pay for the construction. In a worst case scenario you can add to the fare if you need to – you’re already talking about a fare that’s at least $0.75 less than what METRO charges for commuter bus service, so there’s room to raise the fares if needed.
Permalink
if you want TOD wouldnt it be a better idea to stay a little farther away from 288. Kirby extends all the way down to shadow creek(Pearland) with gaps. Future Kirby has more undeveloped land with possibilities for TOD access from both sides, a transit line adjacent to 288 would only have TOD potential in one direction. Land adjacent to 288 is almost always going to have more value serving 288. When looking into this possibility (yours and mine) since the light rail alignment is already so close to being able to serve pearland wouldnt it make more sense to extend it giving us transferless service to the med center and downtown.
In your first post in this series, your response to the anti rail folks was to remember that the development enviroment will change in response to infrastructure investment. With that in mind this southern alignment seems ideal a population center in pearland with concentrated commuting to the employment center, with lots of developable land in between.
Does your cost estimate account for any necessary land ROW purchases or planning costs?
Permalink
A few alignment comments:
-A single station for the TMC means that a lot of jobs will not be within walking distance, which then leads to transfers. Transfers and long walks neutralize fast operating speeds pretty quickly. A technology that offers fast acceleration can take advantage of that by adding some more stations.
-The Downtown station near the Toyota Center is nearly useless: it needs to be closer to jobs. But Downtown property owners might not be thrilled about an elevated structure in front of their buildings.
-The Downtown extension proposes elevated double track maglev down residential streets. If I lived there, I wouldn’t like that idea.
Cost comments:
-The cost of construction is heavily dependent on the context. Building a guideway in an open field is much easier and much less expensive than building a guideway in the middle of an active street.
-An elevated guideway is not immune from the difficulty of building in an urban environment. What if you’re putting pylons down the middle of the street and there turns out to be a sewer line right in the middle of the street?
-The cost of stations is a large component of the cost of any transit system but particularly an elevated grade separated system (where each station is essentially a 2-story building). I’m not sure these costs take that into account: the presentation that American Maglev has on their website for Orlando explicitly does not include station costs.
-The cost of maintenance facilities isn’t trivial, either. And those facilities (which take a fair ammount of land) need to be on the first operating segment.
-Any time a new technology is implemented operationally for the first time there are unexpected costs. We can’t assume that a system that works at 35 mph will scale to 150. Transit has a really bad track record for new technologies — they almost always take much more time and money than expected.
-Drivers are a part of operating costs, but only part. Fare collection costs money. Cleaning costs money. Security costs money. Maintenance costs money (and getting rid of wheels doesn’t make vehicles maintenance-free). The control systems required for automated operation involve maintenance costs, too. So far no advance in transit technology has resulted in systems covering their operating and capital costs. In places where transit does make money (i.e. Japan) it’s because of economic systems and demographics — all modes of transit make money. Data point: the fully automated SkyTrain system in Vancouver costs over $1 a trip to operate.
Permalink
Christof,
Regarding Alignments:
- There are two TMC stations, between the two of them the vast majority of the TMC is within a 1/4 mile radius. No, not every single job is within that radius, but at some point you need a circulator bus or other similar services to reach every single thing in such a district. Also, it’s foolish to assume that nobody will ever walk more than 1/4 mile if it’s a pleasant walk or an indoor walk – in the TMC nearly all the buildings are interconnected with skywalks etc, and the sidewalks are decent enough that you probably can get better walking distances than in the random middle of suburbia.
- Service from Pearland would go directly to the TMC Central station, which means no transfers.
- The Downtown Station is merely access point #1 – specifically located in a redevelopment area so the train builders can capitalize on TOD at the beginning. It’s also close to St. Josephs hospital (increasing the value of the rapid connection to TMC). Look to the next post in this series to see next stage of a downtown configuration.
- Also, it’s absurd to assume that the existence of the stations doesn’t dramatically change daily life in their vicinity. Not only would there be rapid development around the stations (esp. with a private sector option as the development of the station areas is a primary financial concern), but major shifts such as changes in METRO bus routes (circulator service) and private services (like the REV taxis downtown or car-shares etc etc) will very quickly react to new opportunities created by rapid commuter service to help bridge that last mile gap. Those things aren’t as ideal as having the train station stop directly in front of someone’s office, but it can’t stop at every building, and the reaction of the urban fabric to the presence of a station is not an insignificant contributor to the success of the system.
- As for whether people would like maglev above their street, nobody ever likes anything to happen on their street. At a minimum, the maglev is extremely quiet compared to normal bus service, and doesn’t obstruct traffic. If and when someone actually tries to build something, they can divert the route to avoid NIMBYs if they must. I’m going to stick to drawing what I believe to be the most effective route.
Cost:
Your observations on cost are valid. I don’t know how many stations they could include in a figure of $17M / mi, but since their price targets range from $15-$17 I’d guess that some basic platforms could be squeezed in there. Nicer stations would have to be absorbed on the TOD side.
Operations costs are not trivial, but frankly I don’t have the time or the data to model them. The capital cost is the more immediate problem, and my only goal was to give a guess as to what the capital cost per rider would be. The purpose of that illustration is to show that this idea is within the realm of possibility, not that it’s already perfected.
I do think that there are huge opportunities for revenue generation outside of fares that could probably account for the operating costs in the short term until the proceeds from station area development sufficiently offsets the capital cost of the track. The correct way of designing the station, from a private sector point of view, would be to develop enough stuff around the stations to more than pay off the capital cost of the track. At that point all the fares are going to operations and profit, and the economics are totally different.
This isn’t a proforma, nor is it pretending to be. The models involved in predicting the performance of this kind of thing would be extremely complex, and frankly I’m not getting paid to develop them, so I’m comfortable sharing some of the simple back-of-the-envelope numbers and sparking a discussion. That’s what makes this a blog, not a business plan.
Thanks for your feedback.
Permalink
Another cost of driving instead of taking the rail in this case is the cost of parking for those that don’t have reserved spots. I don’t know how common this is at the Med Center.
It would be interesting to look at some statistics for those going from Pearland to the Med Center area and looking to see what % would be within x meters from the drop off point(s). Downtown Park and Rides show that people are definitely willing to walk a few blocks from a bus stop, I assume there is similar mobility in the Med Center area as far as easy walkways to get from place to place.
Permalink
I was cheering your plan until you mentioned funding from city, county, and state sources. All that would amount to is tax payers subsidizing the transportation costs of those who live in Pearland. Remove that part and I really like your plan.
Permalink
Brian,
Let me emphasize that I offered that example only as a hypothetical possibility in the same category as corporate sponsorship or benevolent donations (like a wealthy individual purchasing naming rights to a station). I certainly don’t think that any serious system should be dependent on grant funding, and I would consider the private sector a far more productive target for fund-raising efforts. However, with all the money that’s flying around for “shovel-ready” projects and the like, if a public entity decided they wanted to contribute a grant to support the system, no private enterprise would rationally say no to free money.
The reason I included that remark is simply to illustrate the impact any outside contribution would make on the financing cost of the project, and to show how significantly that could lower the minimum fare.
I hope that adds some clarification to my thought process, and I’m glad you liked the idea otherwise.
-Andrew
Permalink
One of the main problems I see with this line is that the usage is going to be mostly unidirectional, morning and afternoon rush hour. It seems to get the most out high speed rail, you need to have destinations at both ends of the line and in between. Unless you had some unidirectional type of rail that went rush hour, but that seems like a waste of money compared to just expanding HOV.
Permalink
@ Andrew -
in any HSR system, the cost of electricity and especially, the cost of maintenance of both rolling stock and infrastructure are decidedly non-zero. You cannot just ignore them because you’ve got a tech fetish with this AMT maglev system.
You should also be honest enough to admit that this South Corridor would implement regional/commuter transit service, not HSR. If there are tightish curves in your right of way, you’ll need to bank the track severely or slow down. Based on your table above, my guess is that equipment rated for 80mph top speed would achieve nearly the same transportation value but possibly at substantially lower capital and operations cost.
You’ll need some ridership analysis to see which price points will be feasible. If the system is built using public funds, expect pressure to minimize fares such that nearby roads are effectively decongested. That means running longer trains every few minutes, especially during rush hour. Standees limit the feasible acceleration/deceleration rates and also mean extra attention has to be paid to jerk (change in acceleration, ideally a continuous function of time).
If a private investors steps up to the plate, which is unlikely without taxpayer operating subsidies, fares will have to be higher and the city/county may not achieve the hoped-for road decongestion. For the higher ticket price, passengers will expect a somewhat higher level of service in every respect. Things like cleanliness and seat availability rates matter at least as much as line haul time.
Permalink
Rafael,
On operating costs, please see the previous reply to Christof.
This piece isn’t HSR, it’s regional commuter service. I am not pretending this is the same as HSR, I’m proposing that if a private-sector HSR effort were to be successful its starting point would need to be a short, high-traffic, commuter corridor.
The goal is to demonstrate a private-sector concept, not a public sector concept. Also, congestion reduction is a myth, any traffic absorbed by the rail will be replaced by latent demand soon enough. The purpose of a rail system is to provide an alternative mobility system. It can’t and won’t eliminate automobile congestion, so it shouldn’t pretend to try.
The rest of your points on standees, acceleration, and quality of service are valid. I used an acceleration factor of .15g, which is the best estimate I could find for typical subway operation in major cities. To be honest, I had a hard time finding a good source of data on what was acceptable acceleration. If you’ve got a good source of data on transit acceleration rates, please share it.
Permalink
Acceleration at 1.5 m/s^2 is very ambitious. Subways top at 1, and can’t maintain that rate for too long – they drop to about 0.5 soon after. Most systems can’t even get much above 0.5 to begin with.
Japan’s maglev system has been tested at 1.47, but the test track has a U-shaped vertical profile, with inclines of 3-4% at both ends; gravitational acceleration is free in terms of both energy and passenger comfort.
Permalink
When it comes to calculating a fare, I would look at a customer centric value and see how that support operational costs.
At $3.00 gallon for gas and 25 mpg, that 12 cents per mile so a 15 mile each way commute has a real variable cost of $3.60 just in fuel to the consumer. Factor in a very low parking rate of $2.00 for the day and u get $5.60.
Yes, there are costs like insurance, depreciation and maintance but I think if the value is built on daily out of pocket expense, you can avoid alot of argument.
As far as time savings, be careful because all a person cares about is door to door time so the rail alternative has to account for
home to station drive
parking and waiting time for train
train transit time
destination station to office transit time
the high speed of rail buys some time but a system has to provide value in total time to win over opponents.
I am all for well executed rail and not a repeat of the toy train Houston ended up with.