FrontRunner Double-tracking and electrification

I've generally been negative regarding FrontRunner double-tracking and electrification. Partially because for most of my riding life, it's been an hourly commuter rail with PnR access. Not real useful for those living the car-free or car-less lifestyle. More recently I realized what double-tracking and electrification would do: Transform FrontRunner into something much more like the PATH train. Viz: an electrified heavy-rail system with (potentially) 15 frequencies. FrontRunner was built on the cheap (not a criticism), including a lot of single track, so in many segments the only place two trains can pass each other is at stations. So one train has to wait for the other, so if one train is late, the other train is made late. It's also a diesel train, so the amount of torque it can produce is limited, and hence how fast it can accelerate. (Diesel trains are great for long distances--for cheaper than stringing catenary for the whole distance, and rolling-friction efficiency of trains aside, a better option than carrying batteries the whole way). FrontRunner is also the only way the whole Wasatch Front can be on one train system. TRAX is great, but it's light rail, which means it can't share track with freight trains, which means it's limited to being on it's own ROW. (TRAX does some 'time separation', which is why the blue line will never have an 'after hours' train home for people who want to drink and not-drive.)

Rail to Logan?

North Carolina (barring Charlotte) lacks rail transit. But they do have state sponsored Amtrak, the Piedmont service. Makes me reflect on commuter rail and passenger rail. In Europe, the two represent a continuum rather than a dichotomy. Inter-regional passenger trains and commuter trains share the same track, just at different frequencies.

Utah has FrontRunner, running from Provo to Ogden. And it faces regular calls to extend FrontRunner, even unto Logan (Utah's other Metro region). Efforts to extend FrontRunner north haven't really worked out--simply not enough demand. But there might be enough demand for something less than the hourly commuter rail frequency, but something more like state sponsored Amtrak. There is an existing rail line between Ogden and Logan, it is very indirect, running an S-curve around the mountains.There is also an existing highway (HWY-91) that runs directly between Brigham City and Logan. So it would be very difficult for the train to compete with driving.But there might be a demand for passenger service regardless, for the non-driving population--Logan is home to Utah State University, with 27,000 students.

That said, that said, if it will be feasible for rail, it should be feasible for Commuter Express Buses.  And if it’s worth $100m in rail investment, it’s worth $100m of asphalt investment in dedicated lanes. (No implication that 100% of the lane need to be dedicated--just the most congested section). 

The political deviance of BRT

I love looking at transit plans. And I love looking at good maps. WFRC's website satisfies about 75% of them.

But today, I need to call out an egregious failure:

This is a planned BRT, stretching from downtown Salt Lake to Sandy, a large suburb some 15 miles south. East from the FrontRunner station in Salt Lake, South along 700/900 east, and West on 9400 South, and then on 102nd south to Frontrunner. Except for this  weird little jog.

I know why--there is a relatively dense residential cluster at the 106th South Trax Station. But there is also a Trax station at 9400 South. Connecting the two Trax stations by BRT makes no sense. Connecting the (low-rise) residential makes no sense. Why not just continue west on 94th, and south along Sandy Expo Parkway? Or South along States Street to 102nd South? This routing adds miles and minutes to a BRT route...and for what?

There must be some political schenanigans going on.

Medium Capacity Metro Systems

I was just reading up on Medium Capacity Metro systems, and one column listed capacity per hour as a criteria (20,000-30,000 persons per direction per hour) and it made me wonder how the UTA Trax stacks up. Years ago, a Portland planner explained the virtues of SLC's long blocks--we can run four car trains, when Portland can only run two-car trains. So, how does SLC's capacity stack up?

The Siemens S70 can hold 225 people, at crush loads. SLC can run four cars per train. And I know there are segments of the network that run trains every five minutes (where the blue, red and green lines share track), which is 12 trains per hour.  So: 4*225*12 is 10,880 persons per direction per hour. Which suggests that Trax is not at it's max, it is pretty close to it (10,000-12,000). 12000/4/225 is 13.3 trains per hour, or a train every 4.5 minutes. So Trax is (in certain sections) very close to capacity. And in sections with curves/turns (900 S to 400 S) probably at capacity, and starting to generate delays.

UTA should probably be looking at either an alternate alignment (400 west) through downtown for the green line, or a transit tunnel under the 400 south intersection.   

Always the bridesmaid, never the bride

This has been a perpetual project on the LRTP for as long as I've been watching the LRTP--Trax along 400 South to Frontrunner. If I was a better man (or a real estate investor), I'd have gone broke waiting for it. At this point, I'm frankly uncertain of the need for it. If the point is to connect FrontRunner to the University/hospital (without requiring a transfer between Trax lines, there seem to be better ways to do it: Either a bus direct from FrontRunner along 200 S, or a 'Black Line' direct from the Airport to the U, connecting to FrontRunner along North Temple. 

If we're going to spend $75 million on transit, this is NOT how we should spend it.

Perhaps it's just dumb planning? A holdover from a decade past? Or perhaps Crayonista hubris, wanting all the Trax lines to connect to a central station? The latter seems unsupportable: rail road 'Union Stations' were a questionable American hubris when they were built: European cities seem to do fine with a variety of stations at railroad stations at the edge of their periphiques (where the old city stops). And the modern metropolis is multi-polar: bringing everything together at a single point is foolish, and any regional agency should oppose it.  In the past couple of years, UTA seems to be getting the idea of a 'transfer network' composed of high frequency bus routes (rather than the historically-based idiocy of trying to bring everyone downtown for a transfer). And I don't see how this project fits into it.

$75 million buys a lot of bus service hours. Buys even more bus service ($150m) 10 years out.  Investing in 'Crayonista' Trax doesn't seem like a very good idea....

Free Fares for SLC

Reading the SL Trib today, the news article was on free fares, suggested as a way to bolster transit ridership. UTA has tried a couple of 'free fare days' as a pilot program, which boosted ridership by about 16%. Article said fares make up 11% of UTAs budget. The usual rhetorical questions about the cost of collecting fares were offered (with no quantification of how much fare actually actually costs--my bet being minimal). Fundamentally, UTA can't do away with the fare collection infrastructure in SLC without doing away with it system wide. And that's not happening.

More and more, my mind turns to SLC needing it's own transit agency (on top of UTA).  No one in UTA is happy when that gets brought up--if SLC breaks off, I imagine their budget goes to bits. UTA allocates service to cities in proportion to how much sales tax revenue the city contributes to UTA. If SLC ceased contributing, UTA would lose...perhaps a quarter of it's sales tax funding? And all the transit-service feasible destinations are in SLC, so without those destinations, none of the existing UTA routes would be feasible. Leave SLC out, UTA's ridership collapses, and so does the agency.

But there is a conflict: SLC wants a kind of service that UTA is not providing: Urban shuttle. Short trips on fast, frequent and reliable transit. Can't blame UTA--that's a very expensive type of service to provide. But if you are traveling within SLC, it's the only kind of service you care about: hop-on, hop-off trips, for (otherwise) walkable distances: 20 minute walks. SLC effectively wants another 400 S. Trax corridor....rather it wants several of them.

SLC's future is in being composed of 'Transit Oriented Corridors' (TOC). TOD is for suburbia: dense, mixed use, walkable nodes set amidst the suburban sprawl. TOC is for urban locations where station areas overlap: where both walking and riding between station are feasible. Portland gets it--at this point, Portland's CBD is so saturated with fixed guideway transit stations (LRT and streetcar) that like 90% of the whole thing is within walking distance of transit. (Paris does even better--98% within the Periphique, IIRC).

Not that I suggest the urbanization of all of SLC--I like the bungalow neighborhoods, and I'd like them to stay. But I recognize that SLC is growing, and growth means a need for more space, and more spacing needs building. But I'd like to see that building take the form of high-density (elevators required) next to transit stations, rather than scattered about in 'missing middle' penny-packets.

To circle back: SLC wants transit capable of supporting car-free living. (Any central city that tries out out-suburb suburbia is playing a mugs game.) UTA doesn't want to provide it. There is an essential conflict there, and one that is not going away. One way for SLC to achieve its desires (at great expense) is to build a geographically limited streetcar network, and make those streetcars free. The other way is to build a free-fare bus network.

Why fare-free? Because most of the bus trips made within SLC are short. Walkable distances. Distances where it's not worth paying the $2.50 for a ticket, but too far to walk. Electronic passes solve a lot of the issues of fare-free transit. I recall the weird rigamarole of 'pay when board' vs 'pay when exit', depending on where you boarded, and the complications of entry/exit from bus doors. With electronic passes, you tap on, and you go.  SLC is already partially fare-free, through various pass systems: U students, high school students, city employees, things like the Hive pass. It might be possible to just go one step further, and make it free to all residents. Then transit becomes a utility, a fee for service: Each parcel gets assessed on the basis of use (# of units), and then every address is mailed a transit pass.

Given that half of SLC's density is employment (actual pop is only 165k or so, IIRC), might also extend the pass program to employees, through an employer-parking cashout program. 

FrontRunner and Joint Development

Longterm utility of FrontRunner largely out of UTA's ambit. If we build office towers next to stations, it will be great forever. If we just create Station Parks and Ebays, it's a near-criminal waste. UTA likely needs to get into the Joint Development game (albeit with a VERY limited geographic context--6 minute walk/.3 mile), with unlimited densities. Using the most proximate land for surface-level purely for bus loops is crazy.  

More rail?

I can't see much room for more rail in Utah. I spend a lot of time looking at it, and I'd like to use that old freight rail in Utah County, but the connection to Trax along the east bench seems sillier and sillier. Not sure how much the network interconnect with Trax weighs against the cost of making the interconnection, but likely not worth it. A combination of FrontRunner and BRT could likely do the job better/cheaper.  UTA could easily spend a decade just double-tracking/electrifying FrontRunner, and extending it N/S.

Transit does best where roadway alternatives are worst: Downtown, the U, etc. I worry about Red line ridership when Mountainview reaches I-80. (I also worry what happens to I-80. )  

Connected and Autonomous Transit Vehicles

More and more research suggesting that complete automation is a pipe dream, and CAV are going to need special guideway--this makes CAV a great match for transit, specifically for line-haul transit. Especially so for BRT, which already has 50%+exclusive guideway. I know UTA is getting in CAV, but only in the context of first/last mile functions, as 'driverless shuttles' running on educational/corporate/medical campuses. But I think it's worth considering an automated shuttle for line-haul service. I'd like to propose travel demand modeling (MAG-WFRC model) to determine how good a line-haul transit CAV might be. Do some scenarios with tripled headway and the like. I think it could be done pretty simply, by playing with model coefficients. A transit CAV could break the link between labor costs and headway that limits the frequency UTA can provide services, a potentially transformative outcome. Ideally, I could piggy-back on an existing EA, AA, or EIS, (such as the proposed Taylorsville or Bountiful BRTs), using already prepared model data, and adding another alternative. Alternately, the use of connected and automated transit vehicles (CATV) could be explored in a 'shuttle' function, such as connecting Sandy Civic Center Trax with the Ebay FrontRunner, (Project 32b in the RTP) or connecting said FrontRunner station with the Phase 3 Blue line Trax Extension at Highland and I-15, via the Prison site. (Project 34b).

FrontRunner - on the perils of commuter rail

This morning finds me less than enraptured of FrontRunner. Not because what it is not (fast, frequent, reliable), but rather because of what it is: Commuter rail. Read somewhere (likely Twitter) criticising transit agencies for dumping billion of dollars into creating high-cost but lightly ridden suburban commuter rail networks. Today, looking at Denver, this sort of strikes home.


In a related vein, the Transport Politic had a good article up on "What Kind of Network to Build?" The relevant part of the article was this:

"Denver, Minneapolis, and Portland are developing primarily radial networks, focusing on expanding access into their downtowns. Their lines—not only those that already exist, but also those under construction and proposed—are widely spaced across the region. On the other hand, Atlanta, Montréal, and Toronto are largely pursuing a grid of new lines that focus on their respective regions’ densest areas. This approach is likely to increase overall transit use more effectively, though it may not provide as useful an alternative to regional traffic. Los Angeles and Seattle are pursuing transit investment programs that tow the line somewhat between the two."

Seems like Denver, Minneapolis and Portland are developing light rail networks that mimic commuter rail networks, sprawling out to the suburbs in all directions, while places with existing heavy rail networks are focusing on many (shorter, lower capital cost) lines within the already built up dense areas. 

Part of the appeal of Commuter Rail comes from the east coast, where commuter rail networks (LIRR, etc) are the outgrowths of profitable passenger service from years ago. Today, I question that they can be replicated: Their genesis is as railroads providing access to the walkable urban cores of (pre-automotive) towns. While small, such towns were based on a monocentric (pre-automotive) automobile model. When I visit FrontRunner stations, all I see are Park and Ride lots, and a bus depot. The PnR makes sense--it's the only way that enough people can be concentrated within walkable distance of the FrontRunner station. Parking has a density of about 1 rider/400 SF. Achieving the same thing with residential density (assuming one rider/household, and 1000 SF/household), requires a FAR of 2.5. At a 50% lot coverage ratio that implies something like 5 stories of residential. Nowhere is that present along FrontRunner.

So FrontRunner is not creating transit-oriented polycentric density. It's just congestion relief. Correlational research out there suggests that a mile of transit travel (a passenger mile) replaces 3 miles of automobile travel, and effect called the Transit Leverage Effect. Empirical evidence bears this out: When a transit strike ocurrs, traffic congestion gets truly gnarly. Research suggests that this is because it's the people with the worst automobile commutes (long commuters along congested roads) that tend to take transit. When deprived of that alternative, they drive anyway. 

What alarms me is the idea that commuter rail is actually supporting sprawl: For a few people, it makes otherwise infeasible automobile commutes possible. Because those people drive a long way (consuming a lot of road capacity), doing so frees up road capacity for others. Via the triple convergence (from other times, other routes, and other modes) that capacity gets used up, until the road system again reaches an equilibrium of barely bearable congestion. 

Which makes me wonder: What does FrontRunner actually achieve? What does any commuter rail actually achieve? To me, it seems like it only makes life better for a few extreme commuters, whose auto commutes would otherwise be awful. 

I'm sure it also benefits a few transit dependent people, for whom the alternative is not making the trip. But I seriously question if the benefit of this small number of people (and it is a small number--FrontRunner ridership numbers are not large) is actually worth the millions of dollars it costs. I feel like we'd be far better creating the 'Freeway Flyers' network of BRT, and making a freeway lane into a HOT lane, then spending millions on commuter rail. 

A decade ago, I supported FrontRunner as an 'express route', making is possible to travel between points that would take far too long to access by Trax. Not once (even while transit dependent) have I ever used FrontRunner for such a purpose. The frequency is too low to make it feasible as part of a transfer network.  FrontRunner is not part of the Frequent Transit Network. 

Why waste so much breath on this issue? Because UTA (and WFRC) are getting ready to spend hundreds of millions double-tracking and electrifying FrontRunner. Double-tracking so that FrontRunner can actually become more reliable (check UTA's twitter for records of it's perpetual delay), and electrification to reduce pollution. (The dirty secret of FrontRunner being that it actually produces more pollution, per rider, and an equivalent number of cars). Part of the pollution is also caused by the single-tracking, which means that the FrontRunner is constantly starting and stopping, waiting for other trains to pass so that it can proceed. 

Here is what is planned. I don't think I like it. That bit of the FronRunner at the south is $51 million dollars for 6.8 miles. Happily, a lot of the stuff on the plan is corridor preservation. Corridor preservation is the essence of planning: Buying cheap today what you are going to need tomorrow. 

SLC Free Fare Zone

UTA's free fare zone map leaves something to be desired.
 It shows the TRAX, but not the bus. When I look at the bus map for downtown SLC, there is a lot of service there! UTA might do better just combining the two maps, given their near-identical extents. Here is my MS-PAINT take on such a map: (Gray line with FFZ acronym)

 On a proper mapping program, I'd use a specially dashed line, perhaps using arrow symbols. Using a shader or a hash would obscure the information inside the box.

If UTA wanted to move away from a free fare zone, they could designate certain stops or stations, rather than an area.

SLC Transit Master Plan: More Executive Summary

Looking at the Executive Summary again.

It says:

Why a Grid Network? Salt Lake City’s existing, centralized hub model is effective for regional connections but is inefficient for some local trips. Currently, many of UTA’s routes terminate at Central Station, which provides good connectivity to commuter rail service, but creates challenges for people who need to travel to other destinations throughout the city, necessitating multiple transfers and/or indirect trips. The FTN builds on Salt Lake City’s strong street network grid. 

This concerns me a bit, because it displays a lack of understanding about transit operations. Many routes terminate at the central station, because that's where UTA has a facility.  Bus drivers, like other employees, get breaks and need to use the bathroom. UTA has a break room at Central Station for that reason.

Secondly, anyplace you have an 'end of line' you need someplace to store transit vehicles. This is especially true of places with high 'peak' loads (morning/evening/event). So many buses wind up at the Central Station because the Central Station has a place to store them. I asked someone at UTA how much land it takes for a bus center, like the ones at TRAX stations, and was told 2-3 acres. That's a lot of land.

Third: Transfers are actually a good thing, as Jarrett Walker point out. Essentially, with a fixed budget, you can choose longer waits but direct service, or shorter waits with a transfer. Research shows that transit riders hate waiting even more than they hate transferring.

I am all for the Frequent Transit Network (FTN). Higher frequency means shorter wait times. But frequent transit is always going to be in limited supply, because high frequency comes at the cost of coverage. That means only a few corridors are going to get it, which means most streets aren't going to get it. The street grid is irrelevant in this regard.

Waiting For Transit

"Service frequency determines the average wait time for transit, and thus much of the overall travel time for a trip"

Q: Reallly? People spend that much time waiting for transit? 

A: Not average trip time, average wait time. Wait time, however, is a substantial portion of many transit trips, and the part of the trip people hate worst. The lit says that riders treat a minute of wait as the same as multiple minutes of in vehicle time. Many UTA buses (especially in Utah County) run at 20, 30 or 60 minute headways. Once you include the amount of time you have to wait when the bus is late, this can mean substantial waits. Given that people choose transit based largely on its time competitiveness to the automobile, this is a significant issue. It's most significant for short trips.  TRAX is frequent and (generally) reliable,. A 10 mile bus journey on a local bus is pretty miserable. WFRC likes to use 5-10 minutes for the wait, but I don't that that effectively reflects how miserable the experience is. It assumes people know where the bus stop is, how long it will take to walk there, and how long the wait will be. That is valid if you are taking the bus to work, once per day, but doesn't reflect the experience of using the bus for general transportation. The longer the headway, the higher the time-cost of missing a bus, so the more likely you are to leave your current location early to make sure your make it. If you have to use an (unreliable) bus connection to make that connection (such as bus to FrontRunner), there is even more waiting, because you can't control if you arrive 10 minutes early or 2 minutes too late, so you have to be 25 minutes early to be guaranteed a connection. Hence, I prefer to assume random arrival, at 1/2 headway, for any bus. This becomes increasingly important as UTA moves away from the '1-seat ride' model and toward the more efficient 'transfer network'. 

9-Line BRT

Looks like UTA is planning to put some transit in place along 800 South/Indiana Avenue. Or so it appears from their ROW purchase plans. My best guess at the alignment for the BRT would be:

Start at 900 South Trax Station, West along 800 South, to Navaho Street.
Option A: Continue West along Indiana Avenue to Redwood Road
Option B: Head south along Navaho Street to Glendale 'Rose', then southeast to California Avenue.
-b1: Return to 1300 S. Trax
-b2: Continue West to Redwood Road.

And then hence south along Redwood Road, terminating at one of many TRAX stations, viz:

• Redwood Junction
• West Jordan City Center 
• Sandy Civic (at 10600 S. and State, following route 218)
• 114/118/Pioneer Road in Draper

Extending the 9-Line Streetcar

 In an earlier post, I discussed adding a streetcar to the existing 9-Line linear park. In this post, I'm going to suggest a possible extension and second phase for the proposed streetcar. It would start at the 9-line Navaho Street Station, proceed south along Navaho Street, around the edge of the Rose Park 'Rose', and into Glendale Shopping Center. It would hence go southward along Glendale Drive, terminating at California Avenue, near the schools. It would only make sense to do so if the long, deep single family parcels along Navaho Street could be redeveloped, which would almost certainly require the use of eminent domain, and thus actions by the Redevelopment Authority (RDA). I'd estimate it would require acquiring about 2.6 acres on each side of Navaho, from 32 different parcels, for an area of about 700' by 150'.


View 9-Line in a larger map

"UTA only knows how to build Light Rail'

I was looking at the illustrations for Sugarhouse Streetcar, and the criticism that UTA only knows how to build light rail sings true. Sugarhouse Streetcar looks like a TRAX line--a dedicated corridor, with a gravel bed and fences along both sides. I've ridden streetcar systems in Portland and Barcelona. Like the name suggests, streetcars run in the street...sometimes on the edge, sometimes down a center median.

UTA built most of the TRAX system using old railway corridor (Sandy--SLC--Mid-Jordan) and seems to be comfortable and familiar doing so. It certainly makes sense to do so. Building in an existing corridor reduces utility conflicts, and makes the property acquisition for right of way relatively simple.

Cost-wise, I know that the per-mile costs associated with running TRAX light rail through Salt Lake City on city streets was fairly high compared to the initial Sandy-Salt Lake stretch, and certainly more complex in terms of traffic engineering, and the need to come to negotiated agreements with UDOT and Salt Lake City.

The next phase of the Sugarhouse Streetcar will have to be along roadway (east on 21st, or north on 1100 East), so UTA is going to have learn how to build 'real' streetcar, and not just TRAX-lite.

TRAX shares right of way with automobiles only at intersections (typically left turn arrows). The rest of the street right of way is protected by curbs or jersey barriers separating the train from auto-traffic. That wasn't originally so, but a two or three cars got themselves t-boned as they tried to make (illegal) turns across the TRAX right of way). Hence, curbs.

But that's not something that is going to work with streetcar, which is going to have to run 'in traffic', right with the cars. That it is possible to do so is really the big difference between the two vehicles. A light rail car weights about 98,500 lbs, while a streetcar weights about 30,000 lbs, or about 1/3 as much. Thus, faster, more responsive stopping capacity.

Embark

http://letsembark.com/
"Free, fast and simple mobile apps for mass transit riders worldwide"
Pretty please, UTA?!

UTA 2 Go Trip Planner

The UTA trip planner remains as frustrating to use as ever. It's a great idea with (unremarkably) poor implementation.  The UTA trip planner relies on recognizing addresses*. The software that runs the trip planner relies on matching the address you enter to an address in its database.

The geodatabases used for the software are typically derived from U.S. postal files--great for keeping track of where to send mail, less useful for finding a physical address. Further, differences in notation may confuse the software. "150 S State Street" may be parsed as:

• 15000 South, State Street
• 150 South, State Street
• 150 South State Street

But UTA is hardly alone--this is characteristic of most transit planning tools.
I recall Portland's as being remarkably different, with one reason: It did not rely on addresses, but on intersections. I recall it as being something like this:

START: "23rd @ Burnside"
END: " Burnside @ NE 6th AVE"

Needless to say, there are vastly fewer intersections than addresses, making it much easier to for the software to find a match, making finding a start and end location much faster.


 *The trip planner also recognizes some landmarks. Finding out which ones are acceptable is a matter of trial and error. "U of U" is not, while "University of Utah Bookstore" is.