UBET Comments on Integration and Sequencing of Transit Scenarios

  • 0
  • November 06, 2003

Response to November 5, 2003 Community Participation and Information Committee Meeting on Sequencing and Integration

Comments on the Integration and Sequencing of Transit Scenarios for a Shared Solution for Transportation Improvements in South Davis County, Utah

Response to: November 5, 2003 Community Participation and Information Committee Meeting (CPIC) on Sequencing and Integration

Submitted by: Roger Borgenicht and Carol Werner, Utahns for Better for Transportation, and
Nina Dougherty and Marc Heileson, Sierra Club

Contact: Roger Borgenicht, 218 East 500 South, Salt Lake City, UT 84111
(801) 355-7085 email: future@xmission.com
Nina Dougherty, 638 Sixth Avenue, Salt Lake City, UT 84103
(801) 322-4610  ninadou@earthlink.net

We appreciated the opportunity to participate in the CPIC November 5, 2003 meeting on the critical transportation issues of sequencing and integration of transportation investments.  We also appreciate the one week extension given to us in making these comments.

Unfortunately, the meeting revealed significant inadequacies in the approaches and methodologies that have been used thus far to formulate and analyze an optimal transit system (also referred to as “extraordinary expansion of public transit” and “robust transit”).  We discussed many issues relevant to this topic in our scoping comments and in our previous comments during the CPIC process. 1   Although we had hoped to submit more detailed comments, the federal agencies refused to share copies of detailed data presented in the slide show during the November 5 meeting, on grounds that the information in the slides was “preliminary,” and despite the fact that it was disclosed in an open public meeting.  As a consequence, we can only address inadequacies identified during this process at a general level.2

  1. The ridership numbers calculated for the “robust transit” scenarios seriously underestimate transit potential in the corridor.  Table 1 shows automobile traffic and transit projections from the DEIS, FEIS, and more recent estimates for the SEIS.  The bottom section, which is from modeling output numbers shown at the November 5 CPIC meeting, shows transit ridership for peak hour, peak direction travel varying from 4.1% to 5.5%.  Study of this table shows that the number of persons forecast to take advantage of transit under the “robust transit” scenario in 2020 is lower than the number of riders allocated to transit’s existing capacity as identified in the FEIS three years ago.  It is also significantly lower than the transit ridership predicted using any of the five methods used to forecast future transit in the corridor in the FEIS. The outputs are one third of the transit riders predicted by the method ultimately relied upon in the FEIS, the financial constraint method.Since the FEIS, new funding for transit was passed by voters in Weber, Davis and Salt Lake Counties and the 2030 Long Range Transportation Plan assumes greatly increased funding for transit projects to be implemented sooner. This suggests that applying a financial constraints method using this increased funding would generate an even higher estimate of transit ridership. Even without additional transit funding, the numbers presented are lower than the anticipated commuter rail capacity at the 2007 start date (UTA plans to run trains every twenty minutes as recently reported in the Deseret News).3Illogically, the modeling suggests a negative relationship between transit ridership and transit improvements and investment.  In other cities with rail systems, transit carries much larger market shares of peak-direction/peak-hour flows.
  2. The history of the model and the results for this exercise suggest that the model used to forecast transit is not reliable, at least as it is apparently being applied.  The model has evolved over time, but has a history of underestimating the potential of transit.  The fact that Salt Lake County’s TRAX light rail is running about 37,610 boardings on the average workday (October 2003), significantly above the predictions of the model, shows the model’s inability to forecast transit ridership well.  Of the five methods used to forecast transit in the FEIS, the model produced the smallest increase of all the methods.  Models used in other metropolitan areas have shown dramatic increases in transit ridership when similar improvements and policies were forecasted.  An obvious example of where the model is failing, illustrated by the results, is in the model’s insensitivity to changes in travel demand resulting from changes in the transportation mix.  The three results presented at the CPIC meeting (and shown at the bottom of Table 1) show that travel demand does not change at all when different mode mixes are modeled.  This is contrary to elementary transportation theory and contradicts models and real world results across the country.We are bothered not only by the low numbers, but by the huge swing in ridership forecasts as well. Such large swings in estimates raise serious doubts about the reliability of forecasts. Without access to the technical inputs and methods used to generate forecasts, it is impossible for us to address possible sources of error.  For example, model run results cannot be analyzed without knowing how many households and employees were moved in the transit supportive land use scenario, and how transit access was improved.  We ask, therefore, that the agencies release for our review, the detailed reports and documentation of all of the inputs and other assumptions used for the model runs reported at the CPIC meetings. However, given the initial results, it appears to us that the models used to produce estimates are unreliable.
  3. There is no indication that light rail has been acknowledged as a reasonable or practicable alternative.  If light rail or other forms of transit, e.g. bus rapid transit, in combination with commuter rail are practicable, they must be evaluated.  Despite our past comments asking for the incorporation of light rail and further explanation of why it was eliminated so early in the CPIC process, light rail’s potential contribution to satisfying travel demand in the corridor was again ignored. This is especially troubling since a study supported by the south Davis County Mayors and others has just begun to evaluate the most effective transit investments for south Davis County.  This study focuses on the high population corridors east of I-15, not west, indicating leaders’ belief that transit will be successful there. Davis County leaders have frequently supported light rail extensions northward from the SL TRAX system as the “ideal” solution, even if a temporary interim strategy is some configuration of bus rapid transit
  4. Sequencing analysis is key.  The real world (both logic and actual experience along the Wasatch Front and elsewhere around the country) suggests that the sequence in which transportation improvements are made alters the use of transit. Expanding road capacity via a new highway provides no incentive for changing the predominantly single occupant vehicle mode, and provides an excess supply of road capacity that induces more and longer auto trips.  On the other hand, implementing a robust transit system first would provide other viable mobility choices for some trips and would act as an infrastructure investment that would attract transit oriented development that would further reduce single occupant vehicle trips. Transit should be maximized through sequencing as well as integration.  These are not separate analyses but rather synergistic strategies.  Observation of shifts to transit in other communities such as Dallas, San Diego, and the San Francisco Bay Area as well as empirical research (Brown, Werner & Kim, 2003) suggests that people are more apt to try transit if automobile use becomes unattractive.  Furthermore, once they become accustomed to transit use, they value the cost savings, the time transit provides for reading and relaxing, and the freedom from congestion and parking hassles it provides, meaning that even after a new road capacity is built, transit would retain much of this new growth in ridership.  Every community leader at the CPIC meetings stressed that they want transit to be successful.  The sequencing and integration process should include time lags between the availability of transit and introduction of new roads so that people are induced to try transit.
  5. How Transit Oriented Development (TOD) (or Transit Supportive Development) is being handled in both the Integration and Sequencing analyses is extremely unclear and of high concern to us.
    Transit supportive development around transit stations is essential for successful transit use–and for happy transit users.  It appears that there are two streams being pursued about TOD and the analysis of it: 1) No go in Davis County, for the most part, and 2) Yes, it’s assumed in the scenario analysis (but what are the details of the assumed TOD?)At several meetings, Davis County political leaders and consultants have said that modeling should be based on local community plans for density and development.  At the CPIC Integration and Sequencing TOD Subcommittee meeting we learned that Farmington and possibly West Bountiful would welcome TOD developments but the other communities said they have no available space.  The report from the TOD meeting at the November 5th CPIC meeting was that recommendations should be based on reasonably anticipated plans and policies and not on speculative futures.  However, at the final CPIC meeting, we were told that transit supportive land use is an assumption in the sequencing analysis–but what do these TOD’s look like? The information white-out we are experiencing doesn’t allow us to know exactly what is happening with the modeling or assumption of TOD’s around transit stops.

    Not only are TOD’s popular elsewhere, but once transit stations are being built there could be a market demand that would result in the release of zoning caps on density around transit stations. Accepting current density laws means that high density developments around transit stations (TODs or transit oriented developments) will not be available to many people who wish to live close to transit.  Moreover, such developments will not be available to absorb a high proportion of the anticipated population growth.  And yet, in communities outside of Davis County, TODs are popular and desirable; density caps are frequently changed in order to take advantage of their attractiveness.  Explicitly precluding the possibility of high density housing around transit stops serves to undermine the potential of TODs to increase transit use in a synergistic way.  Building a road into wetlands may seem to be the only “practicable solution” if we accept the idea that for some reason, Davis County residents are different from other people.  The fact that other communities in the West, including Salt Lake City, have smaller lot sizes (and therefore higher densities) shows that more compact neighborhoods can provide attractive and acceptable housing options.  Given current population growth rates and limited developable space in Davis County, higher housing densities are inevitable in the future.  Why sacrifice wetlands now in order to preserve low density housing when in thirty years, high density is extremely likely.

  6. Need for Optimal Capture of Land Use/Transportation Synergism 
    The results of the model suggest that the model does not adequately take into account the synergistic relationship between land use and transportation policies. Examples of this synergy include the increased developments of TODs, strengthened urban cores, and increased relocation of residences and places of employment served by the transit infrastructure.4  Also, it is questionable whether large-scale regional models are capable of capturing the fine-grained effects of TOD or TDM strategies on travel demand.
  7. An expert panel should be implemented to insure that modeling results are realistic. Backcasting can set goals which can be achieved through planning and development. We had recommended previously that alternative methods be used in light of the problematic nature of forecasting a robust transit alternative.  One way to assure integrity of the forecast is to rely on a panel of outside experts to review and guide the process of forecasting modal splits for a robust transit solution.  An independent, expert panel should oversee results of model outcomes to insure the scientific integrity required by NEPA. This is essential, particularly when the results are as seemingly spurious as those from the current model. Necessary adjustments could be made by such a panel.Another possibility is to employ back-casting.  This would involve setting a benchmark (e.g., a transit modal split of 25% at the screenline) and then working backwards to determine the model inputs (e.g., TOD densities, transit frequency, feeder bus quality) that would be needed to achieve the desired target.  Additionally, it would be wrong to pass judgment on m modal options for this corridor solely on the basis of screenline counts.  This simply perpetuates the tradition of planning and designing for the speedy movement of cars, a practice that has been repudiated as placing the movement of vehicles ahead of the mobility of people. VMT per capita should be used as a performance indicator when evaluating various transportation options.
  8. Redwood Road and Other Improvements to the Current Road System 
    Currently there are several problems on I-15 in Davis County that create chronic congestion during rush hour (e.g., the I-15 Beck Street overpass bottleneck).  We were told that the model assumed that these impediments had been corrected, so that sequencing began with higher capacity.  These improvements are important and we appreciate that the sequencing begins with them in placeThe upgrading of Redwood Road north of I-215 to 500 South in Bountiful/Woods Cross, linked to a new arterial connector to Parrish Lane along the utility corridor and tied to frontage roads along I-15 between Centerville and Farmington would provide a safety valve function (in the event of disruptions along I-15) and provide a continuous alternative route in the study area. As part of the sequencing plan, we ask that an expanded and enhanced Redwood Road facility, coupled with smart-corridor improvements on I-15 (e.g., interchangeable message signs, auxiliary lane improvements, frontage roads), be explored as an improvement prior to any Legacy Parkway construction.  A surveillance program could then be created to see how well the expanded Redwood Road facility, along with the Robust Transit Solution, handles traffic during peak hours and non-recurring incidents.  We believe this is a viable option that poses far less environmental threats that should be included in the sequencing analysis.


Table 1
Summary of Screenline Numbers and Transit Share in DEIS, FEIS & SEIS for
Transportation Improvements in Davis County

Prepared: Roger Borgenicht (801) 355-7085

Data are drawn from sources indicated, using the headings in the original source.  The FEIS summarized the increase of 5,620 (24,110-18,490) as 3,370 pcph due to model changes, and 2,250 from a change in the assumed directional split from 55/45 to 60/40.

Draft EIS – September 1998

Total 2020 Corridor Demand 18,490 vehicles
Existing Transit 800 vehicles, 4%
Future Transit 1,100 vehicles, 6%
Future ITS, TSM, TDM 600 vehicles, 3%

Final EIS – June 2000

2020 Demand at Woodscross Screenline 24,110 pceph*
Future ITS, TSM, TDM 600  pceph
Existing Transit Capacity 1,360 pceph, 6%x1.15 =1,564 people
Maximum Reasonable Future Transit Capacity 1,550 pceph, 6% =1,782
Transit 2020 total          2,910 pceph, 12%=3,346

* Pceph (passenger car equivalents, peak hour, peak direction)

Supplemental EIS – November 2003
Presented at 11/5/03 CPIC Meeting on Sequencing and Integration by Jerry Walters, Fehr and Peers

FEIS Summary

2020 Demand at Woods Cross Screenline 24,110 pceph
FEIS Highway and roadways 18,295 pceph
Transit 2,910
TDM 600
Unmet demand 2,305

Finalist Robust Transit Integration Scenarios – Nov 5, 2003

2020 Shared Solution in FEIS Inte Scen A Inte Scen B
Total pceph 19,614 19,615 19,613
Auto 18,689 18,461 18,341
Transit 809, 4.1% 1,038, 5.3% 1,096, 5.5%
Walk, bike, local 116 116 176

1. We incorporate these comments by reference and ask the consultants give them their fullest consideration.

2. These comments may be supplemented with more specific comments in the future once more specific information is provided to us.

3. Geoffrey Fattah, “Commuter Rail Ready to Roll by ‘07” Deseret News (Nov. 9, 2003).

4. Perhaps this panel may also be supplemented with output for the region’s UrbanSim model, which will soon be operable.  Results from a similar MEPLAN model in Sacramento and elsewhere show a much more realistic representation of the relationship between transportation improvements and policies, land use changes, and resulting shifts in travel demand and mode choice.


Leave a comment

Your email address will not be published. Required fields are marked.