Nov 11, 2018 | Benjamin Etgen, Professor of Mathematics
This is the time to think critically about the practical aspects of transit in Sacramento and to dream about a network that would entice our friends and neighbors to ride. I have been considering the lessons of the existing network, what we want out of a new network, the means to get there, and what that might look like.
I hope that you will take the time to consider where transit has been and where it could be going. The goal of the attached paper is to start a conversation on our terms, as advocates for better communities and transit riders.
No detail is too small. No dream is too big.
Keep on ridin',
The 31 Year Regional Transit Network Experiment
When light rail opened in 1987 a 31 year long experiment began. The “multi-destinational” network created under the leadership of Tom Matoff was built around timed-connections among buses and light rail trains at light rail stations. Almost all of the bus routes served the newly opened light rail stations, and were scheduled to allow a few minutes between the arrival of buses and the departure of trains. Beyond that, bus routes were very different from each other. These differences in alignment have become more and more important. Over the course of the following 31 years some routes have attracted larger numbers of riders while others have attracted smaller numbers. Today we can clearly see why this happens.
As revenues increased and fell over the past 31 years, Regional Transit has experienced periods of growth and periods of contraction. Regional Transit responded by cutting service to routes with lower ridership in periods of contraction and increasing service to routes with higher ridership in periods of expansion. As the system evolved, this interaction of ridership and service levels has resulted in successful routes and weak routes. What has not changed over the past 31 years is the alignment of the routes. The choice of sequence of roads that routes take through our community is largely the same as it was when light rail was launched. The residents of our region have selected for the successful routes through their willingness to ride some routes more than others. The 31 year experiment tells us how the alignment of routes helps to determine their success.
I find it very interesting that many of the successful routes parallel weak routes. While low ridership on Fulton means that evening service on the #26 ends at 7:00 pm, the parallel service on Morse, Watt and Whitney on the #82 has high ridership and now has half-hourly service on weekday evenings until 9:45pm. Weekend service on both roads has diverged equally dramatically. Similar to the evolution of increased ridership and service on the #82, the #93 on Hillsdale began as an hourly route in 1987. The #93 now offers half-hourly service. Before we compare the alignment of the weak #26 with the successful #82, let us look at an even starker example of parallel routes.
Alignment: El Camino and Marconi Avenue
Imagine you had a choice between living on Marconi Avenue or El Camino Avenue in the Arden and Carmichael communities of Sacramento County. Both are signal controlled 5-lane suburban roads with high-density housing and a smattering of schools, places of worship, parks, and small businesses.
Marconi Avenue has two libraries and two grocery stores. Density is slightly greater on Marconi Avenue than El Camino. El Camino has two grocery stores and a hardware store. One would imagine that Marconi Avenue would have among the best transit service and ridership. The reality is very different. Service on Marconi Avenue has been hourly for much of its existence. Even when Marconi Avenue finally gained half-hourly weekday service, every other bus terminates in the east at Fair Oaks Blvd and Manzanita Ave. Beyond Fair Oaks Blvd, service to Mercy San Juan Hospital and Sunrise Mall is hourly. Service ends at 7:00 pm. Saturday service is hourly. There is no service on Marconi Avenue or to Mercy San Juan Hospital on Sundays. As part of the 2010 budget cuts, Regional Transit cited low productivity of the route and proposed making all service on Marconi Avenue hourly again.
High ridership on El Camino has been matched by Regional Transit with increased service. Even before light rail, ridership was higher on El Camino than on Marconi. At that time, during the evening commute, extra buses were added to the #23 outbound to provide 15-minutely service. Buses on El Camino are half-hourly throughout the route, as far as Sunrise Mall. As the network evolved, the entire route on El Camino gained half-hourly Saturday service. Service ends at 10:45 pm.
Much of the difference in the success of transit service on El Camino and Marconi is due to their alignment. The two roads are served by routes with very different alignments.
Marconi Avenue is served by the #25. Heading west, the #25 goes straight to the Marconi/Arcade light rail station. It offers the most direct service to light rail in the Arden/Arcade area of Sacramento. I find the Marconi/Arcade light rail station to be a depressing place, especially after dark. There are no facilities or shops to pass the time when connecting between the bus and light rail. Heading east, buses are hourly beyond Fair Oaks Blvd to Mercy San Juan Hospital and Sunrise Mall.
El Camino is served by the #23. Heading west, the #23 turns away from light rail and serves Ethan Way before battling traffic on Arden Blvd to serve Arden Fair Mall. Restrooms, refreshment and entertainment are available at Arden Fair Mall. The route ends at the busy intersections of Arden Way and Del Paso Blvd at the light rail station. Heading east, the #23 follows Fair Oaks Blvd, serving more high-density housing, before ending at Sunrise Mall.
If riders rewarded direct routes to light rail, then the #25 on Marconi Avenue would have much higher ridership than the meandering #23 on El Camino. The #25 is designed to require transfers. The #25 bypasses major traffic generators, like American River College and Arden Fair Mall. To reach CSUS a rider must transfer once at Marconi/Arcade light rail station. The #25 passes within a mile of American River College, but bypasses it. Instead a transfer is required on Watt Avenue to reach American River College. Similarly, to reach Arden Fair Mall a rider on the #25 needs to transfer once. The #25 is aligned for linearity. In contrast, the #23 is aligned to serve major traffic generators, even when that means going out of its way.
In Sacramento’s hot summers and rainy springs, more people have chosen to become riders on El Camino than on Marconi Avenue. The alignment of the routes shows that serving major traffic generators is more valuable to many riders than connecting to light-rail.
As another example, consider Ful
ton Avenue and Morse Avenue. Both roads have sections with high-density housing. Fulton is a signal controlled 5-lane road. Morse is a 2-lane road with stop signs and speed bumps. Fulton Avenue is home to many of the region’s used car lots. Morse has half-hourly service all day and into the evening on route #82 as well as service on the combined #80 and #84.Service on Fulton ends before 7:00 pm on weekdays on route #26. Half-hourly service on the #82 continues until 9:45 pm on weekdays. Even with a longer, more meandering route and half-hourly evening service the #82 has higher productivity than the #26. What can we learn from these examples?
Riders Reward Service to Major Destinations without Transfers
The existing network was designed for timed-connections among light rail and bus routes. Those of use who rode the network in the 1990s, know how well it worked for trips to downtown. Connections were convenient and quick for these trips. Unfortunately, that meant that bus routes missed a nearby mall, college, or shopping center. The result is a network with many near-misses.
The Dreaded Transfer
Sacramento’s multi-destinational transit network was considered so unique when it was implemented in 1987 that academic papers explained how it worked. In the 1980s the phrases hub-and-spoke, waves, and connecting complex were as new as the experience of transferring from flight to flight at a major airline’s hub. Among transit professionals, transfers were widely acknowledged as one of the major impediments to providing attractive transit service. To design a network that required a “dreaded transfer” was unheard of. To address this, timed-transfers made changing routes less of a nuisance. The network was not built around transfers, but timed-transfers that were scheduled to connect.
As Sacramentan’s we can easily imagine timed-transfers and accept schedules that add a few minutes to our trip. However, they can only occur at one point on each route. Timed-connections anchor the schedule of each route at one location. The time at which the bus arrives at every other stop along its route is then determined. That is, once one location for a timed-connection is chosen, the rest of the schedule is set. There is no way to offer timed-connections with other routes that themselves have a schedule created for timed-connections elsewhere. Since almost all of the routes were scheduled to offer timed-connections with light rail, timed-connections could not also be offered elsewhere. Because of this, Sacramentans think of transit service as an excellent way to reach downtown. Regional Transit’s marketing emphasizes this by focusing on downtown events to this day. Reaching destinations that are closer requires long waits for unscheduled connections.
When transfers are not scheduled, the time needed to connect is essentially random. You may be required to wait a few minutes or almost the entire headway of the route. Thus, we can model the amount of time that a transfer requires as half the headway. Thus, an unscheduled connection to a 15-minutely route would be expected to take 7.5 minutes. If that seems too simplistic, consider the situation in which a bus arrives a few minutes before another. In this situation, the transfer in this direction is short and convenient, but transfers in the other direction are long and inconvenient. If you took 4 minutes to transfer between 15-minutely routes in one direction, transferring in the other direction takes 11 minutes. The average transfer time for the round-trip is still 7.5 minutes. Without timed-connections, every good connection for one rider is a terrible connection for another.
Again, the design of the routes is important, in this case as important as frequency.
In the experiment above, we had to depart at a given time. For both the two 15-minutely routes or the single 30-minutely route, we set out without consulting a schedule. That captures that fact that many trips depend on leaving or arriving at a given time.
Not all trips have this feature though. Because schools have libraries and malls have shops, if you consult the schedule, you can borrow a book or buy a coffee and make use of the time before departing. Here the direct 30-minutely route is preferable. Using the schedule, you get to the stop a few minutes before departure time, book or a covered coffee cup in hand, and minimize your wait time.
It is good to remember that transfers are not only temporal, they are also physical. A transfer requires collecting your belongings, leaving the bus, being outdoors, showing your pass or transfer to another driver, and then taking another seat. For many people a transfer also requires using the ramp twice and having their chair secured for a second time. In many places, a transfer may also require crossing a road. In the worst places, a transfer requires crossing unsignaled right-turn slip ways, highway ramps, or access roads to shops. Where stops are placed far from the intersection, a long trip is needed to reach the connecting bus. Together, this makes transfers both dangerous and inconvenient. Any transit rider has had the experience of seeing the connecting bus pass on the far side of the road while they wait for the signal. Adventurous riders set-off against the signal.
For all these reasons the alignment of routes determines the experience riders have of the network. Is the trip direct? If not, is the transfer-timed? If a transfer is required must the rider cross the street, cross highway ramps, or cross slip ways? Is the transfer point lit and patrolled? Are there other pedestrians around throughout the span of service?
In 1987, before the term “frequent grid” or even “frequency doubling” were in common usage, Regional Transit created frequent service within the “multi-destinational” network. Exploring the system as a kid, one of the first things that struck me was that there was a bus every 15 minutes from Arden/Del Paso Light Rail station to Arden Fair Mall on the #20, #22 and #23. These buses then took different roads to serve the, then busy, County Club Plaza shopping center. Over their separate routes 15-minutely service was offered as far as Country Club Plaza. This cleverly maintained timed connections at light rail, provided coverage on Arden Way, Cottage Way and El Camino, and provided 15-minutely service. This excerpt of the 1993 system map illustrates this. This also requires that light rail trains arrive at Arden/Del Paso in both directions at about the same time. The #23 was then, as now, 30-minutely. The #20 and #22 were both hourly, but scheduled to provide 15-minutely service to their shared destinations.
In the existing network the #80 and #84 provide a combined 30-minutely service on Watt Avenue between Fair Oaks Blvd and McClellan Business Park. Similarly, the #67 and #68 provide a combined 15-minutely service from Arden Fair Mall and 29th Street Light Rail. From 29th Street Light Rail to Florin Mall they provide a nearly 15-minutely service. (The existing #67 and #68 have schedules that differ from 15-minutely by 2 or 3 minutes between 29th Street and Florin Mall.)
Summary of the lessons of the 31-year experiment:
The “High Coverage Alternative”
The “high coverage alternative” resembles the existing network but with less geographical distribution of service and less frequency. This alternative represents deep cuts to an already decimated system. In many areas where there is high-ridership today, there is no service at all in this alternative. For example, there would be no service in North Highlands on Elkhorn between Watt and Hillsdale. Similarly, there would be no service at all on Antelope. The service in the northern portion of downtown and East Sacramento (the existing #34) would be gone. Service on Fair Oaks parallel to Sunrise would be gone. Express buses to the Pocket area would be gone. Significant frequency reductions are also evident. Service on San Juan and Fair Oaks parallel to Madison would become hourly replacing the current half-hourly service. Service on Sunrise north of Sunrise Mall would also be cut from half-hourly to hourly. Service on Rio Linda in North Sacramento would be similarly lose half of its service and become hourly. There is also some new service. For example, to the Amazon facility on Elkhorn near the airport and Delta Shores, in the City of Sacramento. There are also some potential improvements. Service on Fruitridge continues downtown, as it did before the light rail was extended south as the Blue Line. Service is restored on 24th Street south of City College. Service east of Sunrise Mall would connect with light rail in Folsom. Although many roads around Florin Mall and Cosumnes River College are served by different routes the changes in frequency do not effect either of these destinations. This alternative represents an overall decrease in frequency and coverage.
Near Misses: This alternative would create new near misses while only addressing one of the existing near-misses. For example, the #81, taking the place of the existing #87, still does not serve Arden Fair Mall. The #61 and #15 still do not serve City College. The #81 would continue from University/65thStreet Station to CSUS, patching one near miss in the existing system. New near-misses include the #67 on Franklin that would no longer serve Florin Mall. Similarly, the #65 would pass near, but not serve Florin Mall. The #67 on El Camino ends further than ever from American River College.
Direct Service: There would continue to be no direct service from City College to Florin Mall, from American River College to Arden Fair Mall nor from CSUS to Arden Fair Mall.
Number of Transfers: This alternative would do nothing to reduce the number of transfers needed to move through the network. New 15-minutely service from downtown to West El Camino and Azevedo is too far from Truxel to be of any use to riders on the #11. (How 15-minutely service on that section could be maintained by two routes with terminals as far apart as County Club Center and Sunrise Mall also requires further thought.) For most trips, transfers would take place exactly where they are now. New transfers would be required between the #67 and the #25 for riders from from El Camino to reach Sunrise Mall.
Safety of Transfers: Since most transfer locations would not change, the safety of transfers would also remain similar to today. New transfers would be required from the #67 to #25 on Marconi. Since the #67 would terminate in a loop, each trip would require crossing Marconi once.
Frequency Doubling: Existing frequency doubling between University/65th station and CSUS and between 29th Street station and Florin Mall would be lost with this proposal. (The existing #82 and #87 ave schedules that differ from 15-minutely by 1 or 2 minutes between University/65th and CSUS. The existing #67 and #68 have schedules that differ from 15-minutely by 2 or 3 minutes between 29th Street and Florin Mall.) New, 15-minutely service would be offered between downtown and West El Camino and Azevedo. However, this does not serve a major traffic generator. Using two branches of the #25 to serve Sunrise Mall reduces the frequency of service on the #25B and does not increase the frequency between Carmichael and Sunrise Mall. Splitting the #62 would provide hourly service in the Pocket area. Splitting the #81 before it reaches John F Kennedy High School reduces frequency.
Summary: In many ways the “high coverage alternative” is worse than the existing network when considering route design. Some new service is matched with a reduction in frequency on many routes and elimination of service in other areas. The service day of the remaining routes would be no better than the existing network. The most predominate failing of this proposal is the reduction in coverage and frequency. This proposal reduces the geographical distribution of service when compared with the existing network.
For comparison, the 1985 network could be considered a high-coverage network for Citrus Heights, Fair Oaks, Carmichael, Antelope, Arden, and Arcade in the north area and Oak Park, Tahoe Park, and Colonial Heights in the south area.
The 2006 network could be considered a high-coverage network for North Sacramento and Natomas. At that time, Main, Bell, Grand, Fairfield, Arcade, Fairbanks, Ford, Pinell all had service.
The “High Ridership Alternative”
The “high ridership alternative” redirects capacity to the Arden/Arcade areas of Sacramento. The emphasis of the proposal is on frequency for Arden/Arcade. Many routes are cut in other areas of Sacramento to provide the capacity for additional frequency in Arden/Arcade. Because of the design of the routes, many major traffic generators, even in Arden/Arcade, can only be reached after transferring, sometimes twice. Many of these transfers are in unsafe locations. The effect on other areas of Sacramento is devastating. Sunrise Mall, among the largest source of tax revenue for Sacramento County, would be served by only 8 buses per hour. Arden Fair, by comparison, would be served by 20 buses per hour. Florin Mall would be served by 16 buses per hour. The Pocket area would be served by 4 buses an hour, while today it is served by 20 buses per hour. Service in Oak Park would see just 8 buses per hour, instead of 12 today. North Sacramento would be served by 8 buses per hour, in place of the 10 buses per hour today. North Highlands and Antelope would be served by 8 buses per hour, instead of 14. By contrast, Arden/Arcade would be served by 48 buses per hour, where there are only 26 today.
Even within Arden/Arcade, where the amount service would be dramatically increased, some places with high ridership lose service entirely, while some places with low ridership receive twice the amount of existing service.
Frequent Service With Low Riders: Two excellent examples of locations that would receive 15- minutely service that have low ridership are Watt Avenue south of Arden and Fulton Avenue south of Northrop. Ridership on Howe north of Arden is lower than south of Arden, but it would be served by a 15-minutely route that terminates after only serving this stretch of road. Although the design of the routes make service on Arden Way necessary for rider from Fulton or Watt to transfer to reach Arden Fair Mall, there is little ridership on Arden today.
Service Day: While the “high coverage alternative” mirrors the existing service day and frequencies, this proposal not only increases frequency but also dramatically extends the service day. Most routes would run until 11:00 pm daily. This requires further explanation as to how this could be juggled with existing capacity. This option promises both higher frequency for Arden/Arcade and a longer service day. Why this is possible for the high frequency option but not with high coverage option is unexplained.
Bunching: High ridership, frequent lines have a natural tendency to bunch. Bunching means that two (or more) buses arrive within a few minutes and then a large gap in service follows. This happens when the leading bus is delayed and picks-up part of the following buses load allowing the following bus to catch up. Preventing bunching requires more precise schedules, active controllers and additional buses. As an example, bunching on the #51 was bad enough that to simply maintain schedules Regional Transit added two buses to the schedule in September of 2018 to address inadequate recovery time.
Headway or Schedule: Even if additional buses and controllers were available to ensure that buses did not bunch, schedules could drift. Even a difference of 1.5 minutes to the headway would mean that bus schedules would drift by 6 minutes an hour. For a 15-minutely route, this means that buses come as far from the scheduled time as is possible in 75 minutes. For those using transit to reach jobs or classes, missing a connection would not only add 15 minutes to their trip but could cause them to be late. If schedules were abandoned for headways, even without considering bunching, slight variations in headways could make connections unreliable. It is also valuable to be reminded that if schedules were abandoned for headways, then timed connections to light rail would also be lost. Connections with light rail would become as random as those between bus routes.
Canceled Trips: Those familiar with the “Shuttle” at Watt/I-80 are also aware of the fact that the buses used for that service are taken from the schedule of the #1. Canceled trips, like bunching, mean that frequency disappears and all that remains is a network that requires connections.
Near Misses: This alternative would create new near misses while addressing three in Arden/Arcade and Carmichael. The #81, taking the place of the existing #87, would serve CSUS and Arden Fair Mall. Finally, the #27, taking the place of some of the existing route of the #23 on El Camino, would serve American River College. Five new near-misses would be created. Service on the #27 on El Camino would no longer serve Arden Fair Mall, creating a new near-miss. The #67 on Franklin would no longer serve Florin Mall, creating another new near-miss. The #81 no longer serves John F Kennedy High School in the Pocket Area, creating yet another near-miss. Additionally, the #67 would terminate at Franklin Light Rail Station instead of using Center Parkway to serve Cosumnes River College. Many near-misses disappear with their routes. Only one bus route would remain near Sacramento City College. Only two routes would remain near Florin Mall.
Direct Service: There would be new direct service between CSUS and Arden Fair Mall, but direct service from American River College and CSUS is lost. There would still be no direct service between City College and Florin Mall or between American River College and Arden Fair Mall.
Number of Transfers: Because of the emphasis on frequency, even in Arden/Arcade where existing service would be more frequent, many major traffic generators would require transfers to reach.
American River College: Although American River College gains a second 15-minutely service, it would remain hard to reach. From Marconi Avenue, riders would have the choice of waiting for the 30- minutely #27 at Walnut Ave, two miles out of the way for those riders coming from Manzanita, or waiting for the 15-minutely #26 at Marconi and Fulton, adding 4 miles to their trip for those coming from Marconi east of Watt Avenue. From Watt Avenue, riders would have the choice of transferring to the 15-minutely #1 as they do today at Watt/I-80 light rail station, or to the 15-minutely #26 at Watt and Auburn. Worse, many riders who would need to transfer to reach American River College in this proposal do not need to do so today as the existing #82 follows the route of the #84 between Arden and Whitney. Service from American River College to Whitney Avenue, Morse and Northrop is lost entirely. Riders on Howe Avenue south of Arden would have the choice of traveling 3 miles out of their way to transfer to the #26 to reach American River College or transferring twice. Riders on Howe north of Arden would have to transfer twice, once to light rail at Marconi/Arcade Light Rail Station and again at Watt/I-80 light rail station.
Watt Avenue: Although #84 would offer 15-minutely service on Watt Avenue, the route does not serve any major traffic generators, since Country Club Plaza is nearly deserted and Kaiser Morse Hospital will be closing. Any trip from Watt Avenue would need to transfer once to reach American River College or Arden Fair Mall, or twice to reach CSUS.
CSUS: Since the link between American River College and CSUS would be lost with the elimination of the #82, traveling between American River College and CSUS would require a transfer for trips that do not require a transfer today. As mentioned above, trips between CSUS and Watt Avenue would require two transfers. From El Camino and Marconi riders headed to CSUS would be required to transfer twice. In the existing network only one transfer is required.
Sunrise Mall: Riders on El Camino and Marconi who currently enjoy direct service to Sunrise Mall would have to transfer. Because the proposed #25 meets the #1 in a loop, connections in one direction would require that the driver’s break time and the route’s recovery time delay riders. As shown, the #25 would connect with the #1 and then travel to Mercy San Juan Hospital, where the recovery time for the schedule would take place and drivers have their break. Riders connecting from the #1 could be asked to leave the bus while the driver takes their break. Thus, connections between the #25 and #1 add an exceptional amount of time to the rider’s trips.
Outside of Arden/Arcade: Since very few routes remain outside of Arden/Arcade, most connections and opportunities for trips are lost entirely. Where service would remain, one new transfer would be required for riders on West El Camino to travel downtown. In the existing network, riders can continue directly downtown. Another new transfer would be required for riders on the rerouted #67 to travel to Florin Mall. In the existing network, riders can travel directly to Florin Mall. The only major traffic generator on the rerouted #67 would be Arden Fair Mall, even though the majority of the route is in south Sacramento. The rerouted #67 would pass within a mile of Sacramento City College, but without any routes to connect it, requiring a long walk that crosses two highway ramps. Similarly, the rerouted #67 would end two miles from Cosumnes River College.
Oddly, two 15-minutely routes would be provided between light rail stations on the Blue Line and Arden Fair Mall. The #67 would serve Arden Fair Mall from Marconi/Arcade while the the #22 would also serve Arden Fair Mall from Arden/Del Paso. At Marconi/Arcade there would only be one other bus route to connect with, the #25. Similarly, 15-minutely service would be provided between Watt Avenue and American River College over two routes, the #1 from Watt/I-80 light rail station and the #26 from Watt and Auburn Blvd. Surprisingly, two of the three bus routes that would serve American River College, would connect with the same route on Watt Avenue, the #84.
Safety of Transfers: Not only would this proposal require riders to make many new transfers, the locations of those transfers are troubling.
Perhaps the worst would be the connection between the #26 and the #84 at Watt Avenue and Auburn Blvd. Unfortunately, this is the only location in which these two routes would connect. The result is that some riders would be forced to make this connection. None of the stops have shelters, lights or other pedestrian activity. One stop is on the far side of a highway ramp. Two stops require crossing a right turn slip lane as well as the intersection. One stops is between the access roads to a filling station.
Shelters: Several of the locations in which riders would have to connect are at bus stops without shelters. At these locations, there is no escape from the sun or rain for waiting riders: all four stops at Watt and Auburn, all four stops at Fulton and Arden, both stops on Truxel at Truxel and West El Camino, all four stops at Franklin and Florin, both stops on the 65th Street Expressway and the southbound stop on Stockton Blvd at 65th Street Expressway and Stockton Blvd.
Frequency Doubling: None of the routes would double frequency between major traffic generators.
Summary: Outside of Arden/Arcade the “high ridership alternative” eliminates most service throughout Sacramento. For the remaining routes, there is an emphasis on frequency. Because of the design of the routes, the advantage of frequency is lost because of the requirement that riders transfer in order to reach a major traffic generator. Many of those transfers are forced to occur at locations without amenities or facilities. Most of the new transfers take place in locations without shelters. One transfer would require riders to cross a highway ramp. Two would require riders to cross a right turn slip lane. The proposal would not create any frequency doubling. Many major traffic generators are not connected by a direct route. For those riders who would still have transit service, the option forces riders to transfer. This is the no-service or high-transfer option.
What is Next?
Sacramentans deserve a network that (1) eliminates near misses, (2) provides direct service between major traffic generators, (3) minimizes the number of transfers, (4) ensures that the locations of transfers take safety into consideration, and, perhaps most important, (5) create a frequent-grid by using frequency-doubling. The lessons of the 31 year experiment is clear. The new network must use frequency doubling to provide both coverage and frequency between major traffic generators.
The existing bus service in each area of Sacramento allows opportunities for both coverage and frequency. To illustrate, consider a network in Arden/Arcade that applies the lessons of the 31 year experiment.
The existing network, like the proposed “coverage” network, has a sparse grid of frequent routes that connect major traffic generators every 15 minutes. This is shown in the left diagram (“Coverage” Frequent Grid in Arden/Arcade). From Watt/I-80 the #1 serves American River College and Sunrise Mall (shown in red). From Arden Fair the #67 and #68 use Frequency Doubling to connect Arden Fair Mall with J-Street, 29th Street light rail station, and Florin Mall (shown in green). Finally, outside of Arden/Arcade, the #30 connects CSUS to downtown (shown in blue).
With a dramatic shift of bus service to Arden/Arcade, the “ridership” network creates a grid of frequent routes. This is shown in the middle diagram (“Ridership” Frequent Grid in Arden/Arcade with Additional Capacity Redirected from the Remainder of Sacramento). Many of the transfers would be on surface streets, shown as ÛThe emphasis on transfers creates a network that requires transfers to reach major traffic generators.
Vision for Sacramento
Using frequency doubling, existing bus service in Arden/Arcade can be redesigned to create a frequent grid among major traffic generators without redirecting capacity from other areas of Sacramento. This is shown in the right diagram (Frequent Grid in Arden/Arcade using Existing Capacity and Frequency Doubling). The resulting frequent grid emphasizes connections at major traffic generators, with pedestrian activity and facilities. Existing service is scheduled to provide timed-connections at light rails stations. Realigned routes would continue to have their schedules anchored at light rail stations to ensure that trips connecting with light rail as convenient as they are now.
Using the Lessons of the 31 Year Experiment to Create a Vision for Sacramento
Transit Service in Arden/Arcade using Existing Capacity and Frequency Doubling
Marconi/El Camino service: Half-hourly service on Marconi and El Camino would serve American River College (with frequent connections to Sunrise Mall and Watt/I-80), Arden Fair Mall (with frequent connections to Arden/Del Paso, CSUS and U65th), J-Street (with frequent connections to Mercy Hospital, CSUS and Downtown), and 29th Street (with frequent connections to Florin Mall, Downtown and Folsom). Both American River College and Arden Fair Mall have public restrooms and pedestrian activity. The Marconi and El Camino service would be essentially one 15-minutely route, with two branches in the middle. The routes would be timed to connect with light rail at 29th Street and could be interlined with the #67/68 there. Interestingly, going east these routes would offer service to a college with frequent connections to a mall and going west they would offer service to a mall with frequent connections to a university.
Commute Connections: If both branches of the Marconi and El Camino service are taken to Marconi/Arcade light rail station in the peak-hour in the peak-direction, then they can be scheduled to maintain a 15-minutely schedule before and after the connection. This could also allow the routes t have a "tighter" off-peak schedule that requires less buses.
Watt Avenue: Service on Watt Avenue would be 15-minutely north from Watt/I-80 to McClellan serving Walmart. Half-hourly service would replace hourly service on Don Julio, Elkhorn and Antelope. Instead of terminating at Greenback and Auburn, both branches would then continue to Sunrise Mall, via Auburn and Sunrise or Madison and Sunrise. Service south of Watt/I-80 would branch. One branch would serve Morse, Hurley, Bell, Collage and Ethan to Arden Fair Mall. The other branch would serve Grand and Northrop to Arden Del Paso light rail station and Arden Fair Mall. Again, the Watt Avenue services would be essentially one 15-minutely route, with bi-directional loops at the ends. The routes would be timed to connect with light rail and the #1 at Watt/I-80. Additionally, because of the alignment of routes no buses would need to change direction between trips, as they do today.
Fulton Avenue:Fulton Avenue would gain service to two major traffic generators with frequent connections (CSUS and American River College). It would also serve high-density housing on Northrop and Edison. Finally, it would provide a connection between CSUS and American River College. The route would be timed to connect with light rail at U65th. Since the Fulton Avenue service would continue to Sunrise Mall without a transfer, it would be as useful to riders as a connection between two 15-minutely routes.
Edison Avenue: In addition to returning the Fulton Ave service to Edison Ave, the density and necessity support taking a rerouted #19 from Watt/I-80 and Marconi/Arcade light rail stations via Edison Ave.
Ridership: Ridership of existing service is reflected in the alignment of the routes. If you refer to the ridership data, there is remarkably little ridership on Watt Avenue south of Country Club. Service on Watt Avenue south of Arden Way would continue to be served by a rerouted peak-only #29 and perhaps SacSTA’s planned demand responsive service from Country Club Plaza. Service south of Northrop on Fulton would be rerouted to CSUS.
Two Lane Roads: The high ridership on the exising #82 on two-lane roads, like Morse and Northrop, means that the Watt Ave service would reach Arden Fair Mall through high-density areas on two-lane roads on Hurley, Bell, Cottage, and Ethan. Similarly, the Fulton Avenue route would use Edison Ave to reach Whitney Ave and American River College.
Demand Responsive Service: SacSTA’s Demand Responsive Transit Service between Historic Folsom light rail station and Antelope is shown since this route could connect with frequent service at Sunrise Mall. Demand Responsive Transit Service is also planned for Country Club Plaza, that could connect with light rail via Watt Avenue at the Watt/Manlove station. The form of Demand Responsive Transit Service has not yet been discussed, but many existing Demand Responsive Transit Services have a timed-connection and fixed stops. In Union City, AC Transit’s FLEX service offers timed-connections to and from BART with demand responsive service to existing bus stops in Union City. In Pacifica, SamTran’s offers the FLX, that has a fixed route and schedule as well as offering deviations within one-half mile of the route to any rider who calls in advance.
Arden Fair Mall: The amount of connections at Arden Fair Mall would benefit riders in Arden/Arcade and those passing through. Service on Marconi, El Camino, and Watt would have frequent connections to CSUS and University/65th Street light rail. These connections would also allow frequent connections to CSUS and University/65th Street light rail for service on Rio Linda, Norwood, San Juan and West El Camino.
Sunrise Mall: Service on Auburn and north of Sunrise Mall on Sunrise would be combined. This allows for riders on Auburn to reach Sunrise Mall. Currently, there are three buses per hour between Sunrise Mall and the Gold Line. This allows for 20 minutely service from Sunrise Mall to Mather Field/Mills using existing bus service.
“Using the existing bus service in Arden/Arcade a mix of coverage and frequency can be created that does not require redistributing bus service from other areas. This allows other areas of Sacramento to also apply the lessons of the 31 year experiment and create a better bus network for all Sacramentans.”