Bus reliability and Planning TIme

If you ask people what they want from public transport, they usually say "Speed." But if you look at behavioral data, speed isn't the most important factor. Reliability is. I would rather take a bus that takes 40 minutes guaranteed, than a car that takes 20 minutes but might take 60 if there is traffic. Uncertainty is stressful. Knowing "I will arrive at 8:55 AM" is worth more than "I might arrive at 8:30 AM". This is why dedicated lanes (BRT/Tram) are so powerful. They might not be fast, but they are predictable. We don't need to break the sound barrier. We just need to keep a promise. Do you prefer a fast gamble or a slow guarantee? - Juan Mora Triana

When my bus was scheduled to arrive at 7:55, I had to catch the bus scheduled to arrive at 7:40, because the second bus was late in picking me up. I worked in an office, being 5m late wasn't critical--but it earned me an unfriendly glance from my boss. And so, I started taking the earlier bus. And so, what was nominally a 23m trip (7m walk, 13m bus ride, 3m walk) started to eat 40 minutes out of my day. 

Making it worse was the irregularity in what was nominally a 15m bus. The 7:40 bus would be on-time, or early. The 7:55 bus would be 7-10 minutes late, reliably. So reliably that I began to time my walk from my house to the bus stop to start when the 7:55 bus was supposed to arrive, knowing that the bus wouldn't have arrived before I reached the bus stop. But that was for a bus I took five days a week. 

When I was trying to take a new bus to someplace new, I'd stroll up to the stop, pull its schedule out of my pack of pamphlets, use the system map to check which routes it serviced, pull out the appropriate schedule for the route, and check my watch. Then I'd take half the headway and estimate the blocks, figuring I could make 8 SLC blocks (~1 mile) in 20m, and that the bus moved about 3x that fast. For a 15m bus, I'd estimate a 10m wait (half the headway plus some uncertainty). For a 30m bus, I'd estimate I'd be waiting 20 minutes--the actual wait time plus what I could have walked waiting for the bus. For the hourly bus--I'd start walking. The combination of wait time plus uncertainty plus lost walking time made it a bad bet--my standard estimate was that I was better off walking two miles than waiting for an hourly bus. 

Even today, any trip under 10m, I walk. It's less stressful, because it's less uncertain. Even with a bus with a 15m headway, odds are I'll make it a half mile before the bus comes. I've repeatedly had the frustrating experience of seeing the bus I would have liked to take pass me by... but not often. 

Rapid transit, it's the same calculation, except that I remove the mental 'penalty' for potential to be late. The station will be nicer, there will be places to sit, there is an arrival time count-down [2] and I know the vehicle will be on time. Riding TRAX in SLC, if the train was 15m late, I'd tell me wife "Someone is probably dead" [1]. 

This also tied into my willingness to make transfers. My rules of thumb were that it was always worth it to transfer between a 15m bus and rapid transit, sometimes worthwhile to transfer between two 15m buses, and worthwhile to take a 30m bus TO rapid transit (but not the reverse). Anything else, I was better off walking. So I did a lot of things like taking TRAX west then south and then walking east from a Trax station, because it was a better bet than just waiting for a 30m bus to take me directly south. The bus would either be delayed in arriving or delayed enroute, where TRAX was practically a guarantee.

Admittedly, I was young and fit, with plenty of time to kill, and Salt Lake is a very safe city. But I also didn't have any other options: No car and lack of plowed sidewalks/lanes made biking in winter in SLC prohibitive. 

[1] After enough highly reported crashes at railroad gates, people quit racing the train, but someone racing the train at a red light stills seems like an annual event. SUV drivers are used to being the biggest thing on the road, but in a car-eat-car world, a train is an apex predator.

[2] Transit Information systems that only provide scheduled GTFS arrival times, rather than real-time GTFS data about vehicle location are a cruel joke, so knowable false that they can only be considered a deliberate deception.

Awkwardly, rapid transit does almost nothing for drivers in the peak hour

 To be awkward, thanks to the Triple Convergence, rapid transit does almost nothing for drivers in the peak hour--trips just converge from other routes and other times. However, drivers on the shoulders of the peak (from whence trips migrate) do benefit. Rapid transit can still induce the long-cycle aspect of induced demand, when less-miserable commutes induces new development in peripheral areas. However, rapid transit continues to act as a congestion 'safety valve', such when the traffic congestion gets too bad, people switch from cars to rapid transit. But talking about rapid transit as a congestion reliever kind of misses the plot--parking is the real story, and rapid transit allows places to get denser (earlier, more cheaply) by reducing parking demand.

Rapid Transit Network Expansion

If you start with a small transit network, and build it out over time, it gets larger and more efficient, providing more access over time through the network effect. But at some point, the dynamics switch and the operator realizes that they can get more money out of running more service on the existing service than on further expanding the existing network. (Which is financially prudent). But it breaks an implicit political bargain between city and suburb, viz: "If you pay to build ours now, we will pay to build yours later". As long as those network expansions are promised, and voters think they might be able to use them (or make use of them or benefit from their use), they'll fund them. 

But if you are an exurban commuter, you'll never see any direct benefit, and you know it, but you are still paying for it. Which means sales taxes work basically in urban areas, where they can match the beneficiaries with the payers. People who drive into the county and buy things matter less politically, because they don't vote on referendums--but still do matter, as they still do lobby for things, as NJ and NY demonstrate with the congestion charge. 

Anyway, if the transit network stops expanding, and the transit constituency ceases to include those folks, the transit providers switches to concentrating on existing riders. And then all the capital and operating costs go into providing additional mobility (more tracks, faster loading at stations, more trains/hour). And the main trouble, I think, is that transit agencies that have made that transition once have a hard time switching back to the other regime. Once you've quit building subways, it takes herculean efforts to get things like CrossRail and the Second Avenue Subway built. On the flip side, LA did manage to get the Regional Connector done. 

Against Light Rail

Don't have a rail corridor handy? Don't build light rail. By the time congestion is bad enough for a metropolis to think about light rail, land costs are too high for any 'greenfield' corridor, and using existing right of way is the only feasible way. 

If the only existing right of way is public street, why spend the money to build a electrified railway? Electrified traction has undeniable perks over rubber-tired buses (fuel costs, acceleration, comfort), but the cost is rarely worth the marginal benefit.  

Neither ride comfort nor transit signal priority (TSP) nor dedicated/reserved guideway are specific to rail. Nor level boarding or off-board fare collection. With double-articulated buses, the rider capacity/driver ratio rail previously offered over bus has largely evaporated. 

So why do places like Portland keep building light rail? I suggest an availability heuristic: They have lots of old freight rail lines to convert to light rail. I seen symptoms of the same disease in Utah, and efforts to bring light rail to Utah County. 

The only actual advantage of rail over bus is political: its easier for a metro to create rail rapid transit than bus rapid transit. In the public imagination, light rail vehicles are like freight trains, whose necessary separation from cars is implicitly accepted. 

It is the separation from cars that the true virtues of rapid transit emerge - speed, safety and reliability. Rapid transit requires exclusive/dedicated guide way--exceptions generate accidents reliably, whether in travel lanes or turning lanes.  

So it's politically feasible for a light rail to claim a lane a lane on a congested arterial in a way no mere bus could ever attempt. That said, double-articulated buses seem to be able to make some of the same claims to exceptionalism and necessary special accommodations.  

Urban Phase Shift?

One of the comments on this article was so great I had to repost it:

When I took my Urban Economics course from Barton Smith, one of the observations Smith told us in class was that whenever prices for major goods and services in an urbanized area keep going up (and housing is certainly a big part of most people's budgets), eventually prices will reach a point where they will become a market signal to everyone that the urbanized area needs to stop growing. Population growth will either come to a halt, or people will start leaving until prices fall to an equilibrium where people can afford to live there again. - Neil Meyer 

I think it gets to the core of why city promoters like transportation improvements--without affordable housing, cities can't grow. Affordable housing includes both rents and access costs. As rents in one area rise, people can offset that by moving to nearby areas with lower rents (and higher access costs).
In pre-industrial (and third world slums) the only access available is by walking, and so some truly hideous densities result.

It's theoretically feasible that a city might stop growing***. But most cities don't--instead, they invest in transportation improvements. NYC built it's first subways in the name of "De-Congestion".

One of the ideas I'm kicking round is an urban scaling induced 'phase shift' in the effectiveness of transportation improvements. Namely, total metro population drives average metropolitan density. (As you get bigger, you naturally get denser). At small sizes and low densities, auto-mobility works: Land is cheap, parking is available, walking anywhere is madness. And it keeps working, as long as your addition of automobile capacity keeps up with congestion.

However, while travel is an 'derived demand' in terms of the number of trips made, it's a 'induced demand' in terms of the length of trips made. If you make traveling cheap and easy, people go further.* Urban form is 'set' by the dominant transportation mode at the time of construction**. So places developed under conditions of auto-mobility tend to be low-density with segregated land-uses. The large amount of travel required to get around for basic needs is 'baked-in' at that point.

This is problematic, if the metropolitan area keeps expanding.

from: https://people.hofstra.edu/geotrans/eng/ch6en/conc6en/cellular_automata.html

As the number of zones increases, the average distance from one zone to another also increases. Which means, all else equal, the amount of travel required to get to all the places you want to go also rises. Simply because things are scattered about: home in one place, work in another, groceries in a third, kids-school in a fourth, soccer in another, ballet-practice in another, dentist in yet another.

The big problem is that the increase in average distance between all of these things is non-linear. Every additional 'zone' you add to a city, the average distance rises more than it did for the last zone.
Even if the road lane miles per capita (and associated costs) remain constant, the amount of travel doesn't. As travel demand outstrips supply, congestion results.

Some cities try to fix this increasing congestion with massive increases in road-building. Failure is inevitable. Exponential increases in travel require exponential increases in road capacity, imposing exponential costs on a linearly growing population.

And so at some point, every city gets the 'rapid transit bug'. And it has to be rapid transit, because only rapid transit makes it possible to avoid congestion. Non-rapid transit, such as regular buses and streetcars, are undeniably cheaper to build than rapid transit (sometimes by an order of magnitude).

But that doesn't matter. Rapid transit is premium transit*****. It's transit for choice riders (people who could drive, but choose not to). And to get people to make that choice, it has to be better (in some way) that the driving alternative.

Non-rapid transit suffers from congestion. It's less convenient, less reliable, and less comfortable than  your personal automobile. You don't have a 'locker' to store things in, and hauling groceries is difficult. Thus, it can almost never compete with a private automobile (excepting when parking costs/hassles are enormous).****

Rapid transit is better than the personal automobile when it is a) faster, b) cheaper, and c) more reliable. 'a' only happens when congestion is fierce; transit vehicles make repeated stops. 'b' seems simple, but for most people a car is a 'stock' of mobility they can draw upon even with no money, while transit is a 'pay-per-use' thing, so it seems expensive, even if it is less expensive in aggregate. Again, 'c' is reliant on congestion to work.

In summary: Auto-mobility works until it doesn't. It stops working because average travel distances increase exponentially as the metro area expands, while population increases linearly. This makes it impossible to keep expanding roadway capacity to match demand. (Some try; all fail). As travel demand outstrips supply, congestion results.

The total amount of delay drivers experience is exponentially proportional to the amount of congestion (from: https://ops.fhwa.dot.gov/congestion_report/chapter2.htm) The AADT/C ratio is the ratio of traffic to capacity.

-----------------------------------------------------------------------------------------------------------------------

***This may be one of the reasons for competing cities, closer together. When a city reaches maximum walkable density, it makes a great deal of sense to go elsewhere. Of course, large cities have huge advantages in terms of access to resources and agglomeration economies, which (typically) more than offset the cost of congestion. I expect the only time you'd really see such a switch is for 'Twin cities' like Minneapolis-St. Paul, or the Texas MetroPlex.

*The reverse is also true. As congestion increases, trip length should start falling.

**The trouble being that urban form is fixed at date of construction, and is very difficult to retrofit for alternative transportation modes. Adding highways to NYC came at enormous financial and social cots. Retrofitting auto-dependent cities will likely be painful/costly as well.

****Which is why buses to downtowns, Universities and hospitals work--they are all places with terrible parking.

*****The political quid-pro-quo of rapid transit for dependent riders is that "Hey, you too can ride the premium transit!". The flip side to that is that the high cost of rapid transit means less of it is provided--its goes fewer places. Typically, bus operations get cut to pay for capital improvements, which leaves the transit-dependent population worse off.

Rail Transit

Rail Transit makes sense where there is an existing rail corridor. Outside of that, not sure it does. Last mile links might count. But those tend to grow, into their own corridors. Any end of line location needs a place to store vehicles (which is why Central bus depots are terrible). So a rail line call never naturally terminate in a CBD--it must pass through.

That said, for single line networks, the terminal station can store one train, and operate such that when a train on one track arrives, the train on the other track departs, so there is always a track available for an arriving train. Trax operated this way for years. Now days, UTA builds long tail track at termini to store trains. Which works at peripheral locations, but not for a CBD location.

LRT vs. BRT

To built more Rapid Transit, it should be BRT rather than LRT. Bus Rapid Transit is simply cheaper per mile, which means more of it for the same price.

LRT's advantages are as much political as technical.

First: LRT's primary advantage is that it can use rail corridors to achieve rapid transit status. Cars can't use railway right of way, so there is no competition with cars. This advantage is not limited to LRT. As both the BRT tunnels in Pittsburgh and the Orange line in LA have shown, BRT can use railway right of way quite well. Buses, driven by professionally trained drivers, can pass each easily in 22' (11' per bus). I recall the buses in Eugene, Oregon, coming scarily close. (Railway right of way is only ~22' wide at different points, which is too narrow for even two lanes of car traffic 10' lane + 10' lane + 8' safety area + 8' safety area). Word on the street is that international Fire Code mandates 26' of clear space as necessary for emergency operations.

Second, LRT vehicles are heavier, so they stop more slowly, which makes sharing a lane with cars much more dangerous, especially at high speeds, and consequently less likely to happen.A train in a highway median gets to keep its separated guide-way, and no crank(y) politician can change that. BRT 'Freeway Flyers' (BRT in an exclusive freeway lane) inevitably become BRT in a HOT lane, with consequent degradation in speed and reliability. Trains don't have the problem.

There may be one area where LRT has an actual advantage: Elevated track, with very right clearances. There have been experiments with 'guided' BRT, using things like optical sensors, guidewheels, or specially constructed guide-track. 'The Gap' between vehicle and platform can be an issue, and I'm not sure how well BRT does in that regard.

However, a 'train' portion (exclusive/separated) guideway is expensive. Denver, Seattle, and LA all achieved it by putting light rail in freeway medians. LA is increasing achieving it through the use of elevated lines (albeit at very high prices). Seattle has tunneled, at a cost of over a hundred million dollars a mile.

Previously, I've talked only about 'Rapid' transit in separated/exclusive guide-way. If we want to talk streetcars, it's only fair to compare them with their wheel counterpart, buses. The advantage of streetcar is (again) political. Chambers of Commerce like streetcars, due to their permanency. They are also cheaper and less politically fraught. A 'tram'-type light rail (AKA street-car) is cheap, because it runs in dedicated/mixed traffic right of way. So there is no need to take right of way from cars, or eliminate parking. But this type of right of way comes with costs. It makes the service slow and unreliable--it's like a bus on steel wheels.

Streetcars are also (politically) easier to build, because the FTA helps pay for them. Streetcars can get 50% FTA funding for capital expenses, such as vehicles, TSP, and station platforms. There is no reason a bus couldn't enjoy these same benefits.

The essence of BRT is dedicated guideway. BRT stands for Bus Rapid Transit. 'Rapid' is a technical jargon for exclusive right of way.

A BRT that shares right of way with cars is just a bus. Be it double-articulated, double-storied, multi-door boarding, off-board fare collection, significant stations, with 5-minute frequency. It may be a very nice bus. The FTA can call it 'BRT-lite'. Nonsense. Not BRT.

Contrast with Real rapid transit:
'Metro': Grade separated crossings, underground or elevated track (both exclusive).
'Pre-Metro': Grade-separated crossings, exclusive track
Light Rail Train: Time-separated crossings (railroad gates), exclusive track.
Light Rail Tram-Train: Time-separated crossings (intersections), exclusive track.

Nor Rapid Transit:
Light Rail Tram (Streetcar): Time-separated crossings (intersections) dedicated or shared right of way.

Getting away from 'dedicated' right of way, and obtaining exclusive right of way is hard. Demand for right of way for automobiles is constant, and most severe at intersections. The UTA Trax shares left-turn pockets with automobiles, to it's detriment, and to serious detriment of safety. And rare is the 'BRT' that doesn't have a side-running section where it sits behind right-turning cars.

Read the ITDP BRT Standard. When someone tries to sell you BRT, ask what 'grade' you are getting.

Grades of Guideway

Right of way comes in three grades: 'Separated', 'Exclusive', 'Dedicated'.

Dedicated is the lowest grade. A lane nominally belongs to one mode, an obligation more commonly honored in the breach. Think of HOV lanes and bicycle lanes.


Exclusive can be though of as 'excluding'; exclusive use is maintained through barriers. Exclusive guideway is never continuous--there are always gaps in it, typically at intersections.

Separated is similar to exclusive, except that it is continuous. Typically, this requires grade separation, if only where the guideway intercepts other rights of way.


Heavy railways (commuter rail, freight rail) have separated guideway, for the simple reason that trains can't stop quickly. Where they must cross other roads 'time-separation' (railroad gates) are used. Where this generates unacceptable levels of delay for the cross-streets (a train can take minutes to pass) grade-separation is used. Typically, this requires an overpass for cars traveling over the road. In cases where there are too many roads, the railroad may be buried. The 'cut and cover' model of a trench with a roof were how the first subways were built. (Actual tunnels came later). In some cases (Alameda Corridor, in LA), digging an uncovered trench for the railway is simpler than many bridges.


Light railways have a mix of Separated, Exclusive, and Dedicated. Most light railways make use of some portion of old freight track, which are already time-separated. Some light rail systems provide Separated guideway by running down the center of freeways (Portland) or on elevated structures (Los Angeles, Salt Lake). Light rail on streets (a train in in tram-mode) typically has Exclusive guide way, with intermittent sections of Dedicated guide way. San Diego and Buffalo, two or the first cities to build light rail, made use of this. Light rail vehicles weigh less than heavy rail vehicles, and so can stop more rapidly. Streetcar/Tram vehicles weight even less, so they can stop faster and are even more suitable for on-street use.

The decision to use either Exclusive or Dedicated right of way is one of cost and safety. Without barrier separation, conflicts with turning cars results in accidents. Sadly, even Exclusive guide way is no guarantee against human stupidity.





Higher barriers, which prevent drivers from crossing them, may be safer. The trade-off is that first responders/emergency vehicles lose the ability to make left-turns and U-turns across the street, due to the curb. To accomodate this, curb heights vary between systems.





There is a fourth grade of right of way: Shared. This is what regular buses have. It offers no speed advantage to the transit vehicle over a regular automobile. Once the time consumed with exiting the flow of traffic, boarding and alighting passengers, and re-entering the flow of traffic is considered, transit vehicles in shared right of way travel at about half the speed of an automobile.


In contexts where right of way is scarce or political will lacking, Shared and Dedicated guide way are more common, to the detriment of transit vehicle speed, reliability and safety.  A streetcar operating in shared guide way has minimal advantage over a bus. (The minimal advantage  is that is cannot leave the flow of traffic, and so loses less time entering/leaving the traffic lane).



BRT systems are cheap to build, and largely unregulated. Consequently, they vary wildly. They make use of every kind of guide way.

Grade-Separated 




Time-Separated 
(LA Orange Line), for a time. Drivers ignored them, causing crashes. It still experiences issues with drivers running red lights. Grade-separation is in the works.

Exclusive - Fenced


Exclusive - Curb




Dedicated - Median



Dedicated - Side

BRT and Congestion

The principle of equilibrium assignment suggests that it is unlikely that congestion will change much on the corridor. If BRT successfully reduces automobile congestion on the corridor, travel will be faster in that corridor, and Down’s ‘triple convergence’[1] from alternate routes, times and modes will occur. In that context, the amount of congestion experienced by automobile drivers on the BRT corridor is unlikely to change significantly. However, from a system user perspective, the BRT may provide substantial benefits by actually reducing the amount of diversion (and out of direction travel) that is currently occurring. If this is so, it would be reasonable to expect a drop in volumes along the diversion corridors. It seems likely that the combination of ITS features and dedicated transit guideway will serve to increase the overall capacity of the roadway, and that a drop in traffic volumes on the diversion corridors is a reasonable hypothesis.

However, if congestion increases, a ‘triple divergence’ to alternate routes, times, and modes will occur. How much diversion occurs will depend on how attractive the alternatives are. Assuming no significant addition in roadway capacity on alternate corridors, diversion to alternate routes will result in a slight worsening in overall congestion. Diversion to alternate times will make the ‘peak hour’ longer (AKA ‘peak spreading’). Diversion to other modes may or may not reduce 

congestion.

Buses in general traffic lanes reduce capacity and increase congestion, a phenomenon well asserted both by the literature and by experience. The core principle of making transit ‘rapid’ is removing transit vehicles from general traffic lanes. This serves to both remove the effect of their operations on automobile traffic, and remove the effect of automobile congestion on transit vehicles.

As a thought experiment, assume the BRT is very attractive (in terms of time or cost), and attracts a large number of riders. This reduces automobile congestion along the alignment, making it faster. Drivers diverge from other modes and other routes, and the corridor becomes congested again. But only for automobiles--due to exclusive guideway, the BRT is less affected, and remains an attractive alternative. For drivers on the BRT corridor, there is no net benefit. For transportation system users, there are two classes of beneficiaries: BRT riders, and drivers on the diversion corridors.

A caveat to the benefits to drivers: The benefits to drivers on the alternate routes is going to get ‘lost in the noise’. They will be dispersed over a large number of roads, and reflected in small changes in the duration of peak periods, or in minor traffic volumes in a large number of roads. Provo-Orem is a rapidly growing metropolitan area, with substantial development taking place both north and south of the study area. Any minor advantage from the BRT to drivers will be rapidly eroded by additional land use changes.

A caveat to the benefits for riders: ‘rapid transit’ implies exclusive guideway; most BRT systems are only ‘semi-rapid’. While provided with transit signal priority, time separation (at intersections) provides a reasonable analogue to rapid transit conditions. However, the Provo-Orem BRT has only 51% exclusive guideway. Where the BRT lacks dedicated guideway, it will be exposed to the effects of congestion. In ideal circumstances, this guideway will be placed in the most effective location; where congestion is most intense. Congestion also tends to be greatest near intersections. Thus, roadways tend to be widest at intersections, where the road shoulder is used to provide turn lanes. Many worthwhile BRT projects have been subjected to the ‘death of a thousand cuts’; minor sacrifices made in the name of preserving automobile capacity (or worse:maintaining on-street parking).

However, given the number of routes that the also service parts of the BRT corridor[2], it is unlikely that all of the delay induced by local buses will be eliminated. In the context, it seems likely that the corridor will stay at a very similar level of congestion. 

---

[1] https://escholarship.org/uc/item/3sh9003x#page-4

[2] http://www.rideuta.com/-/media/Files/System-Maps/2016/Utah-County-System-Map.ashx

Rapid Transit In North America

Montreal Metro
Toronto Subway
Vancouver

Chicago L Train
RTA RapidTransit
Baltimore Metro
Boston 'T'
NYC Subway
Philly Subway
PATCO Speedline
San Fran's Muni-Metro

LA Metro