HSR Should Go Where the People Are

(Rafael helped with some of the research for this post. The words and interpretations are mine, so don't blame him for any errors or controversial statements.)

In some ways, the California High Speed Rail project has been a bit too successful in selling itself to voters. The project's emphasis - and, at times, my own - has been on getting people from the SF Bay Area to Southern California in just over 2 1/2 hours. Although this blog has frequently discussed the other areas served by the trains, we haven't always given them equal weight in the basic framing of the project. HSR is and has always intended to be more than just connecting two endpoints. It also connects some of California's fastest-growing cities, particularly those in the San Joaquin Valley.

Not surprisingly, when some people want to find a place to criticize the CHSRA and the HSR project, they point to the route choice between San José and Los Angeles. Some argue that HSR should follow I-5 through the Valley and over the Grapevine. And many of those commenters argue that the CHSRA's failure to pick such a route is either a sign of their incompetence or their complicity with big bad sprawl developers.

These comments crop up often enough that it seemed worth devoting a post to debunking such nonsense.

The basic flaw with the "use I-5" claims, whether they refer to the San Joaquin Valley, the Grapevine, or both, is that they argue for bypassing between 2.5 and 3 million people. There is no good reason to do so. Given that the system needs all the riders it can get to pencil out financially, it is absurd to send the route through completely empty land along Interstate 5 and ignore the populations along the CA-99 or CA-14 corridors.

Especially when those corridors desperately need an intercity alternative. The Highway 99 corridor is notoriously congested throughout much of the Valley, as it is the region's primary transportation route. An increasing number of both trucks and cars use the route, and more will do so if and when economic recovery comes to the Valley. A 2005 estimate showed it would cost $25 billion to bring Highway 99 up to Interstate standards and handle the projected traffic loads. HSR can be built through the Valley for a lower cost but can provide for the movement of people (and improve the movement of goods, especially through projects like Fresno rail consolidation) in a way that can make the widening and upgrading of Highway 99 less necessary.

There's also an environmental reason: the San Joaquin Valley has some of the worst air quality in the nation. HSR will play a major role in addressing that.

The Central Valley is going to see increased population growth during this century. The question whether it'll be sprawl or whether it'll be dense urban infill. HSR can help support infill development by providing opportunities for transit-oriented development. City center stations would help pull growth inward instead of push it outward.

The CHSRA's route FAQ has some more good info on the CA-99 route:

The I-5 corridor has very little existing or projected population between the San Francisco Bay Area and Los Angeles. In contrast, according to the California Department of Finance, well over 3 million residents are projected to live between Fresno and Bakersfield along the SR-99 corridor by 2015, which directly serves all the major Central Valley cities. Residents along the SR-99 corridor lack a competitive transportation alternative to the automobile, and detailed ridership analysis shows that they would be ideal candidates to use a high-speed train system. The I-5 corridor would not be compatible with current land use planning in the Central Valley that accommodates growth in the communities along the SR-99 corridor.

Express trains in the SR-99 corridor would connect San Francisco to Fresno in just 1 hr and 20 min, and Fresno to Los Angeles in 1 hr and 24 min. This corridor would link San Francisco to Bakersfield in about 1 hr and 50 min, and Bakersfield to Los Angeles in 54 min. The SR-99 corridor was estimated to have 3.3 million more intermediate-market ridership (passengers to or from the Central Valley) per year than the highest I-5 corridor projections (CRA 1999). Therefore, while SR-99 corridor travel times would be 11 to 16 min longer than the I-5 alternatives between Los Angeles and San Francisco, overall ridership and revenue for the SR-99 corridor would be higher.

Similarly, the case for Palmdale and against the I-5/Grapevine alignment is compelling. In addition to the fact that Palmdale/Lancaster has just under 500,000 people right now, whereas hardly any live along I-5 north of Castaic, there is the inescapable geological fact that the Tehachapi Pass is flatter, less seismically risky, and cheaper to construct than the extremely hilly I-5 corridor. The I-5/Grapevine route would require individual tunnels of 6 miles in length, more overall miles of tunneling, and would come closer to the seismically unstable junction of the Garlock and San Andreas Faults, one of the more dangerous of California's numerous faults. Typically, you want to cross faults at-grade and not in a tunnel, which is very difficult on the Grapevine route. See more in the CHSRA Tunneling Report.

The time difference is estimated at 12 minutes between Grapevine and Tehachapi, which isn't nothing, but neither is it a huge sacrifice, given the fact that Tehachapi is cheaper, more seismically stable, and serves more people.

Combine all this with the fact that there are forecast to be 1 million people in the Antelope Valley by 2020 (even if that isn't reached, there will be more growth here, as in the San Joaquin Valley), and it makes it clear that here as well, HSR should go where the people are. Even if the Tejon Ranch housing development is actually built, there still won't be as many people as in the Palmdale/Lancaster area.

Quoting again from the CHSRA route FAQ:

The most significant difference in regards to potential environmental impacts between the Antelope Valley option and I-5 alignments is in regards to major parklands. The Antelope Valley alignment would not go through major parks. In contrast, the I-5 options would potentially impact Fort Tejon Historic Park, Angeles and Los Padres National Forests, Hungry Valley State Vehicular Recreation Area, Pyramid Lake and other local parks. The Antelope Valley alignment would also have a lower overall potential for water-related impacts, less potential impacts to wetlands and non-wetland waters, and was forecast to have less impacts on urbanized land and farmland conversion than the I-5 options (because the I-5 options would result in more growth in the Central Valley).

The Antelope Valley alignment traverses less challenging terrain than the I-5 options, which would result considerably less tunneling overall (13 miles 21 km of tunneling for the Antelope Valley option versus 23 37 km miles for I-5 options), and considerably shorter tunnels (maximum length of 3.4 miles 5.5 km for the Antelope Valley option versus two tunnels greater than 5 miles 8 km for the I-5 options) which would result in fewer constructability issues. Although the Antelope Valley option is about 35 miles longer than the I-5 alignment options, it is estimated to be slightly less expensive to construct as a result of less tunneling through the Tehachapi Mountains. In addition, due to its more gentle gradient, geology, topology and other features, the SR-58/Soledad Canyon Corridor offers greater opportunities for using potential high-speed train alignment variations, particularly through the mountainous areas of the corridor, to avoid impacts to environmental resources. In contrast, the more challenging terrain of the I-5 Corridor greatly limits the ability to avoid sensitive resources and seismic constraints. The alignment optimization system (Quantm) that was utilized to identify and evaluate approximately 12 million alignment options for each mountain crossing could only find one practicable alignment option through the Tehachapi Mountains for the I-5 Corridor.

For practical, financial, ridership, planning, and any number of other reasons, HSR should be built along the proposed route.

As we noted in the San Diego post last week, the purpose of HSR isn't to connect two points with a straight line. It's to move people. HSR serves people, not geometry. We need to find the balance between serving the most people possible and a sensible, non-circuitous route. I strongly believe the current CHSRA plan strikes that balance.

---

Reminder by Rafael:

Altamont Corridor Rail Project - Public Scoping Meeting
Today, Tuesday, Nov 17 3:00p to 8:00p
at Fremont Central Park Teen Center, Fremont, CA
39770 Paseo Padre Parkway, Fremont, CA, 94539
(510) 790-5541‎

If you attend and learn something new you feel would be worth sharing with readers of this blog, please email a summary to cruickshank at gmail dot com.

Tehachapi Residents Excited About HSR

Last week the California High Speed Rail Authority took its scoping meeting roadshow to Tehachapi, and judging by this report in the Tehachapi news the locals are excited and supportive - but they'd also like a stop:

“I believe in alternative transportation,” said architect Alex Kosich at the Sept. 16 meeting at the Stallion Springs Community Center. “I rode my bike six miles to this meeting.

“I think it's great. I think it's wonderful. It's a big engineering task. We have to start somewhere.”

Kosich said he and his wife Laila, who is French, are accustomed to riding the high-speed trains from Paris to Lyon.

“Why isn't there one here?” Kosich said. “What are we waiting for?”

The article includes some critical voices, including one guy who voted against the train out of financial concerns.

One of the main concerns raised was the possible impact of the train on a new hospital being built in Tehachapi, though project managers said that the tracks would be sited in such a way to avoid negative impacts. And this being Tehachapi, the article included a brief discussion of the famous Tehachapi Loop, one of the reasons there is currently no passenger rail service between Bakersfield and Los Angeles:

While the rail line in the Tehachapi Pass will follow the BNSF line, there will be no loop similar to the famous Tehachapi Loop at Keene that has been in use since it was built in 1876 and draws railfans from around the world.

With better technology, railroads do not need to adhere to the 2.2 percent grade restriction dictated by Abraham Lincoln's Pacific Railway Act of 1864.

“We can go steeper, to 3 and one-half percent,” said Roworth, the engineer.

In any case, he said after a moment of calculation, at 220 miles an hour, the train would need a circle 16 miles in diameter to make a loop.

Of course, whenever the Tehachapi Pass/Palmdale alignment comes up, we get commenters coming out of the woodwork to argue that this alignment is flawed, too costly, a sop to developers, etc. Never mind the fact that 500,000 people live in the area, a number projected to rise to 1 million by 2020 with or without HSR. The key factor in picking the Tehachapi alignment over I-5 is that, as anyone who has actually driven I-5 through the Grapevine knows, the I-5 route is very mountainous, whereas the Tehachapi Pass is much less so. The CHSRA reached that conclusion over 5 years ago when it eliminated the I-5 route as too challenging and costly for the significant amount of tunneling that would be required to serve that route - a route that wouldn't include any new riders.

How Important Is UPRR To California HSR?

Note: The Blogger service will be down for scheduled maintenance on Thursday 3/26 between 4:00pm and around 4:10pm. We apologize for any inconvenience.

---

On Tuesday, an article in the Hollister Freelance discussed the HSR project from the perspective of Gilroy and the cities that will be in the catchment area of the planned stop there. The piece is timely, since the High Speed Rail Authority will host a project-level EIR/EIS scoping meeting in Gilroy this week (cp. heads-up at the end of this post):

• Thursday, March 26
  
• 3-7pm
  
• Hilton Garden Inn, Ballroom A, 6070 Monterey Road, Gilroy
  

Three potential concerns are listed: the notion that HSR would bisect the community (cp. our post Grade Separations Done Right), the potential noise impacts (cp. our posts Thunder Alley and La Vitrine) and, the position of Union Pacific Railroad (UPRR):

1. Spokeswoman Zoe Richmond said her employer gets "very skittish" about freight trains running in close proximity to high speed rail, whose equipment is very light, [very] fast and carries [lots of] passengers. A UP train derailed and spilled coal onto an [adjacent] light rail line in [Littleton, Colorado on 11 December of 2007, causing the light rail train to derail as well. None of the 30 passengers on board was injured but] Mz Richmond characterized the incident Richardson as "too close for comfort", adding that safety is her company's primary concern.
   
   
2. She also raised concerns about continued access to the existing customer base and winning new business, because freight trains cannot easily and safely cross high speed rail lines at grade. It would be counterproductive if bullet trains ended up forcing freight onto the state's and the nation's roads.
   
   
3. In the same vein, Joseph Thompson, a transportation lawyer in south Santa Clara county claimed that CHSRA had not yet clarified how it would cross UPRR's tracks between Santa Clara (where they run west of UPRR's Alviso line) and Pacheco Pass, well east of Gilroy. He asserted that "Union Pacific's eminent domain trumps High Speed Rail's" because it had been delegated by Congress and President Lincoln.

As a first step, let's look on the map where CHSRA's preferred route would have bullet trains running on or next to UPRR's right of ways.


View Larger Map

The sections of concern are south SF-Gilroy, south Fresno-Merced, the northern approach to Tehachapi Pass, Mojave-Palmdale in the Antelope Valley and part of the Inland Empire route for the phase II spur to San Diego. The Merced-Sacramento section and the "HST/commuter overlay" that is "under consideration" would also raise concerns. All told, roughly 50% of the entire preferred HSR route requires appropriate agreements with UPRR.

Note that while the PCJPB owns the Caltrain ROW, the terms of the 1991 contract with SP - which UPRR acquired in the context of a merger a few years later - give UPRR limited but perpetual trackage rights and 30-minute windows during which it may run its trains "at commuter speeds". For more details, please see Clem Tillier's posts Freight on the Peninsual, Port Pork and Memorandum of Understanding. Legally, CHSRA only needs to deal with the PCJPB, but the latter needs to ensure UPRR's rights are upheld in the process. That could put Caltrain in the middle of a dispute between CHSRA and UPRR, so it should insist on three-way negotiations for issues related to operational safety.

Now, let's examine the concerns raised above in reverse order:

Re 3: In the 19th century, Congress declared that privately owned for-profit railroads were performing a public service by moving goods and passengers around the country. To that end, Congress delegated to them strictly limited powers of eminent domain for the purpose of expanding the public service. In practice, that referred to widening rights of ways, acquiring land for new turnoffs, sidings, yards etc. The idea was to protect railroads against speculators and other landowners who could otherwise exact extremely high prices because railroad alignments must meet certain minimum radii, maximum gradients etc.

The delegation of eminent domain was not intended to allow railroads to prevent competition or other public services from being delivered. Since Nov 4 2008, California HSR is arguably a public service in development. It is therefore not immediately clear that UPRR's powers of eminent domain would trump those of the state of California, let alone those of Congress. However, it is extremely unlikely that anyone will seek eminent domain against UPRR anyhow, even for air or ground rights needed to cross. The objective should be to negotiate in good faith.

Re 2: For now, CHSRA, Caltrain and others are drafting plans that ensure existing freight operations can continue unhindered. In practice, that may mean the monopoly dispatcher for a given corridor may need to instruct one or more bullet trains to slow down or stop to give a freight train the opportunity to cross over to the non-HSR tracks via a diamond. Alternatively, the split of HSR and regular tracks could be defined such that most freight movements are anyhow unaffected. For the remainder, one option would be grade separation between the HSR tracks and freight spurs off the main line. Another, possibly cheaper alternative would be to pay UPRR and its customer to stop using a given spur.

Re 1: UPRR has been in business for 146 years, during which time they've forgotten more about freight railroad operations than CHSRA can ever hope to learn. In particular, they are fully aware of the risk of derailments, which is very small but non-zero. Since US-style heavy freight trains can be up to a mile long, the engineer in charge of a train may not even notice the derailment of a single truck on a single car at first. Indeed, major derailments involving cars tipping or toppling over and fouling adjacent track are quite rare. Usually, a train can be brought to a full stop long before that happens.

UPRR's concern relates to the early detection of a derailment event and, to sending early warning to the operator of the service on the adjacent track - preferably via computer-to-computer messaging to avoid delays related to human-to-human interactions. Time is of the essence because a bullet train traveling at 300km/h can take 40 seconds to come to an emergency stop (less if traveling at e.g. 200km/h). During this time, it can cover well over a mile. With CHSRA planning up to 12 trains per hour each way (esp. on the network's trunk line in the Central Valley), the probability that a bullet train traveling at high speed would be within 60 seconds of the site of a freight rail derailment could be as high as 40% during peak travel periods. That's if the freight train has already derails, a very low probability event.

The upshot is that even if the bullet train's automatic/European/positive train control system were notified of a derailment on an adjacent track and applied the emergency brakes immediately, there would still be a high residual risk of a follow-on collision if the derailed freight train were to foul the bullet train tracks. If that were to happen at significant relative speed, the result could be catastrophic loss of life.

It is a fairly pathological scenario but one that is at least theoretically possible. In general, engineers define a hazard as the product of the probability of occurrence and the damage done. Indeed, Burlington North Santa Fe (BNSF) has not raised a red flag on this issue (at least not in public), perhaps because it perceives the hazard as much lower than UPRR does.

In any event, UPRR published a press release on June 4, 2008, giving notice that it had had no discussions with CHSRA on operational safety in two years and no interest in selling any of its ROW. Mehdi Morshed, the authority's senior engineer, responded with a terse press release of his own, stating that HSR would not share track with UPRR freight trains and citing the excellent safety record of HSR elsewhere in the world (cp. Union Pacific's HSR Games). Note that the circumstances of the 1998 Eschede disaster in Germany had nothing to do with freight trains and everything to do with the hubris of Deutsche Bahn's engineers - it is simply not germaine to the issues raised by UPRR.

That does not mean there is no hubris on CHSRA's part here. US freight railroads are private companies that must pay property taxes on their rights of way, while their competition - the trucking industry - gets a heavily subsidized ride on the nation's highways. Considering US freight rail operators are for-profit corporations, it is not surprising that they should try to make do without expensive active safety systems and keep maintenance overheads on their infrastructure and rolling stock as low as possible without compromising safety in the existing operational context. UPRR does not want to increase its cost of operations just to accommodate high speed rail.

In other words, if CHSRA wants to have any chance of sticking close to its preferred route, it will need to sit down with UPRR and discuss safety concerns regarding derailments in the California context. Unless and until UPRR's engineers are satisfied that these concerns are being taken seriously and adequate measures to keep the hazard acceptable are feasible, the business managers will not be willing to offer any part of the ROW. Moreover, they may raise a red flag with FRA even if no land is transacted. Considering a mile-long heavy freight train traveling at 70mph represents a vast amount of kinetic energy, the civil engineering approach ("add more concrete") may not be sufficient to prevent track fouling in the event of a freight train derailment.

Similar concerns apply to a freight train derailing and hitting a support column for HSR on an aerial structure or, fouling an open trench containing HSR tracks.



FRA has already done some work regarding the aerodynamic interactions between Amtrak Acela Express and freight trains on adjacent tracks. The largest impact was on empty, tall freight cars passed at a relative speed of 110mph. At higher speeds, the response was less pronounced in spite of the greater load pulses at the bow and stern of the passing train because those pulses also lasted less long. The aerodynamic shape of the Acela meant its interactions were less severe at 150mph than those produced by conventional Amfleet trains at 125mph. In any event, the interactions were not considered severe enough to cause a freight train to derail.

The inverse problem, i.e. the serious derailment of an HSR train - mercifully an extremely unlikely event, even in an earthquake (cp. Shake, Rattle and Roll) - could set the scene a follow-on accident with an approaching freight train. However, since there will be far fewer freight trains and they travel at lower speeds, this hazard is a secondary concern. Besides, life is risk, there is no such thing as 100% perfect safety in the transportation sector. The cost of safety measures has to be commensurate with the hazard reductions they achieve.

If CHSRA has not yet done so, it might want to consider hiring a recently retired senior US railroad operations manager with an engineering background, specifically to reach a technical understanding and mutual comfort level with UPRR. Worst case, CHSRA may find UPRR unreceptive even after good faith efforts to address safety concerns. If CHSRA can secure a brand-new ROW that is sufficiently removed from UPRR's, e.g. in the San Jose-Gilroy section, it might well still be possible to proceed without having to redo that portion of the program EIR/EIS.

Plan B?

Otherwise, the only remaining option would be to select a route that minimized or eliminated statewide interactions between UPRR and the high speed rail system, even if BNSF remains willing to share its own ROW in the Central Valley. The most significant impacts would be on the way out of the Bay Area, on the detour via Palmdale, on the spur up to Sacramento and, on the spur through the Inland Empire - four major aspects of the planned network.

Nevertheless, a solution would be possible, at least for the starter line, see the following map. Please note that the alignment implementation details (at grade vs. below grade) are only valid for the section west of Tracy.


View Larger Map

For argument's sake, I've assumed an HSR-capable link across the San Franciso Bay at Dumbarton will prove infeasible because of the Don Edwards National Wildlife Refuge. However, I have assumed an arrangement acceptable to all parties will be found for the entire Caltrain ROW but not for the section down to Gilroy where more freight trains operate. Under that specific set of assumptions, the south bay HSR station would probably have to be moved from SJ Diridon to SantaClara/SJC to secure run-through tracks to the I-880 median. The HSR tracks would cross underneath UPRR's Alviso line and skirt the vast Newhall yard VTA has reserved for BART to the east. It would make sense to reserve part of that yard for additional HSR platforms and/or a yard, especially since it would lie one level below the BART facilities.

The next problem would be crossing over to Pleasanton/Livermore. The BART extension to Fremont Warm Springs and beyond means that CHSRA's Altamont variations based on a route through Niles are no longer feasible. Instead, the most likely option would be to tunnel underneath CA-262 and all the way across to Haynes Gulch (Calaveras Road). Six miles long, it would only just be legal to construct this without a third service/escape bore. At least the Calaveras, perhaps even the Hayward fault would need to be crossed underground. Fortunately, approaching Sunol from the south means the new route could bypass both Pleasanton and Livermore by tunneling across to El Charro Rd, one of the alignments being considered for the BART extension to Livermore. The HSR alignment might need to remain underground to cross under both the UPRR Altamont Pass line and Livermore municipal airport. The lakes near El Charro Rd might have to be drained, at least during construction. I'm not sure what they are used for.

HSR would continue east in the I-580 median, deviating only once to keep the alignment sufficiently straight for high speed service. In the interest of keeping express line haul time down, the Tracy station would end up in the I-205 median, well north of downtown. Beyond it, the CA-120 median and sections across farmland would connect the new route to the BNSF alignment just south of Escalon. Modesto would be served at E. Briggsmore, Merced county at Castle Airport with a possible detour around the Merced town.

The BNSF ROW through Fresno is not straight enough for high speeds, even if adequate noise mitigation measures could be found. It might make more sense to construct a western bypass for HSR/BNSF/Amtrak (3-4 tracks) through farmland and, to run a new DMU-based light rail service on the old BNSF ROW through town. This would deliver passengers from downtown to basic "beet field" HSR stations near Gregg and Bowles that would each be served by 50% of the trains originally slated to stop in downtown Fresno.

Unfortunately, even with all these measures, switching to Altamont implies a line haul penalty of 8-10 minutes for SF-LA express trains, relative to Pacheco Pass. To compensate, the detour via Palmdale would have to be sacrificed (cp. Future's So Bright...) in favor of the technically more challenging but already studied alignment across the Grapevine, past Lake Castaic Wildlife Preserve. For LA county, this sacrifice would presumably not be acceptable unless at least Ontario airport were well served by HSR as early as possible - not an easy proposition if UPRR refuses to co-operate (cp. Quo Vadis: LA- San Diego).

Conclusion: CHSRA had better get into UPRR's good graces, or the entire project could potentially face massive changes to the route, with the fate of some portions (e.g. Stockton - Sacramento) unresolved. The relevant sections of the program EIR/EIS would then have to be re-done, setting the project back by several years. In particular, simply buying land from someone other than UPRR but very close to its ROW may not be sufficient: UPRR could still raise a red flag with FRA if it feels its concerns regarding integrated operational safety and by extension, liability for accidents and loss of revenue, are not adequately addressed.