Dynamic, Integrated Model System: Sacramento-Area Application, Volume 1: Summary Report (2014)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

59 Analysis of Policies and Alternatives of Interest to Planning Agencies A set of policy tests was conducted to demonstrate that the C10B integrated model is capable of analyzing the types of policies and alternatives that are part of typical urban trans- portation planning processes. The objective was to produce reasonable results in a real-world environment for typical transportation planning policies. With this in mind, it was decided that SACOG would perform the analyses at its offices, using its own hardware and staff, with assistance from other team members. The idea was to get an idea of the type of effort that would be required for a planning agency to per- form these types of analyses using the integrated model. A discussion of SACOG’s experience in performing these analy- ses is provided in Chapter 5. Team member Fehr and Peers provided assistance with some tests, and additional assistance was provided by Cambridge Systematics, Inc., and the Uni- versity of Arizona. Originally, eight tests were planned. The final number of tests completed by SACOG was five, as described in the first section of this chapter. Three tests related to transit service changes, and two related to highway system changes. Because of the differences in the types of system changes, the perfor- mance measures used to evaluate the tests varies among the alternatives. These are described in more detail in the second section of the chapter. For each of the five tests, the results of a particular scenario that related to a change in the transportation system were compared with the results from a baseline scenario, which was the same for all tests. The baseline represented year 2005 conditions in the Sacramento region. All scenarios were run using the C10B integrated model; most scenarios were also run using the original SACSIM model validated for the region. Limitations on project resources resulted in some short- cuts taken in the analyses and in the preparation of the C10B integrated model. These issues are described in the follow- ing list. It is hoped that further research with this type of integrated model can assist in assessing the effects of these issues and their practical implications. The issues include the following: • Perhaps the most significant issue was the limited valida- tion of the C10B integrated model, as described in the Chapter 3 section on model testing. This resulted in some significant differences in the baseline scenario results between the C10B model and SACSIM; some of the differ- ences were in the vicinity of the transportation system changes under study, making comparison of the model results difficult in some cases. • Another limitation was the level of convergence achieved in DynusT, also described in the Chapter 3 section on model testing. The test results implied that there was still substantial noise in some C10B integrated model results which affected the ability to fully evaluate the test results. There is, of course, also noise in SACSIM since it includes an activity-based demand model that simulates individual travel behavior. But there is more noise in the C10B inte- grated model since it includes SACSIM as well as the traf- fic and transit simulation components in DynusT and FAST-TrIPs. It would have been desirable to have tighter convergence in DynusT, but efficiency considerations pre- vented this. • Another issue was that each test was run only once with each model. Ideally, simulation models should be run multiple times to get a handle on the level of noise in the results. SACOG has done this with its own validated version of SACSIM, but it was not possible within the project schedule to run each test several times. As the results presented later show, some of the results appear questionable due to the noise level in the C10B inte- grated model, which is greater than in SACSIM because it includes the additional traffic and transit simulation components. C h a p t e r 4

60 test Scenarios The set of five policy and investment alternative scenarios analyzed were defined by SACOG. While the scenarios are realistic and typical of the types of policies and scenarios that SACOG analyzes in its transportation planning function, it must be made clear that the scenarios are not actual projects under consideration in the Sacramento region. Each test was performed using the C10B integrated model, and most tests were also performed using SACSIM. In all tests, the effects of a project scenario (labeled “Scenario 1,” “Scena- rio 2,” etc.) reflecting the specific transportation system change were compared with a baseline no-project scenario. So, for example, in the results for Test 1, the results of Scenario 1 were compared with the baseline scenario; for Test 2, the results of Scenario 2 were compared with the baseline scenario; and so on. For each test, the effects were compared with one or more hypothesized or expected results for direction or sign of effect and for magnitude of effect relative to random effects. Table 4.1 shows the five policy test scenarios. Test 1 was per- formed using only the C10B integrated model since SACSIM does not have the capability of analyzing this type of transit service change. The other four scenarios were analyzed using both the C10B model and SACSIM, and the results from the two models were compared. An additional test to analyze signal coordination on a major arterial corridor was initially run. This test was abandoned after examining the level of noise in the C10B model results— this policy change was too subtle to test using the C10B model without multiple runs. As discussed earlier, due to run time and resource limitations, the test results discussed in this chapter reflect only one run per scenario. The limited tests indicate that the number of iterations in the DTA and DaySim models play a significant role in the results and may need to be tailored to each particular scenario and level of congestion to obtain the level of sensitivity required for comparisons. The lower the level of random variability in a dynamic assignment model and the more its convergence approaches what is customarily found in static assignment models, the higher the substitutability between dynamic and static models in an integrated modeling framework will be. testing and performance Metrics For each scenario, up to three general categories of testing were conducted: • Demand testing focused on the number, mode, destination location, and timing of person trips or person tours. Demand testing utilized the standard person tour and person trip seg- ment output files of DaySim05, which were common to both the C10B and SACSIM travel demand models. Some aspect of demand testing was included in all of the tests. • Traffic assignment testing focused on vehicle volumes and vehicle speeds in and around the test segments, for the two scenarios involving roadway projects (2 and 3). The DynusT traffic assignment results and output files pro- duced using the C10B model for these tests include much more temporal detail than SACSIM model output files; the policy testing includes comparisons of the level of detail provided using the C10B model. • Transit assignment testing focused on total line passenger boardings, passenger boardings by time period, and vehicle loading by time period. Testing included the routes changed in the scenarios, plus several nearby routes likely to be affected by changes in service on the tested routes. Transit assignment testing was performed for three transit-related scenarios (1, 4, and 5). FAST-TrIPs produces much more temporal detail than SACSIM; and, like the traffic assign- ment testing, the transit policy testing includes compari- sons of the level of detail provided using the C10B model. Table 4.2 shows the test metrics used for the policy testing. To manage and evaluate random variation and “noise-to-signal” issues in the policy testing, the following process was used: • For test effects on demand, recent evaluations by SACOG of randomness and noise on the SACSIM model were used. Testing of random variation in SACSIM resulted in a range of expected variation for various metrics. For SACSIM, the expected range of variation was used to perform a simple “greater than/less than” test for each test effect. No testing of random variation was done for the C10B model, but variation was significantly higher than with the SACSIM model. For test effects on demand for the C10B model runs, three times the SACSIM variation was used as a proxy for the policy testing; these effects should be considered cautiously due to their very approximate nature. Table 4.1. Policy Test Scenarios Test Description Extend transit service coverage Test of extending the end of transit service for a bus route from 6:00 p.m. to midnight Improve inter- change design Test of operations-oriented interchange improvement project Relieve freeway bottleneck Test of adding fourth general purpose lane to a heavily congested freeway river crossing connecting to downtown Increase transit frequency Test of reducing service headways from 30 min to 10 min on a well-used bus line Delete bus line Test of deleting a well-used bus line

61 • Other than the limited tests described in Chapter 3, no evaluations of random variations for traffic or transit assignments were available for either the SACSIM or C10B models. For these tests, no evaluations of random variation or the likelihood of test effects exceeding random variation were performed. test results In summary, the policy testing results were conditioned by two general themes or patterns: (1) The C10B integrated model is more noisy than SACSIM since it includes simula- tion in the traffic and transit assignment components in addi- tion to the simulation that is characteristic of DaySim and (2) the tests performed were, in general, too fine-grained to dis- tinguish the test effects from random variation. Repeated runs of the baseline and test scenarios would be necessary, with statistical aggregation of the multiple runs. Because of run time considerations, multiple runs were impractical for this project, and policy testing results were inconclusive for most of the tests. • Although the C10B integrated model often reports feasible and reasonable results, other times the results are counter- intuitive. Overall, the project team concludes that the pol- icy test results are inconclusive and attributes this finding to the higher level of noise found in the C10B model com- pared with SACSIM. The results indicate that the C10B model is sensitive to the policy scenarios tested, and an integrated ABM and DTA model shows promise if the issues of noise and convergence are quantified and better understood. • The C10B integrated model provides a staggering degree of detail and flexibility in its outputs; the basic outputs, though inconclusive for the policy testing performed for this proj- ect, were, in the main, reasonable. The detail and flexibility is mainly based on the treatment of time in the modeling outputs. For purposes of these tests, the C10B integrated model outputs were aggregated to 1-h time slices, but the model would support much smaller time intervals, as well. Table 4.3 summarizes the demand testing results for all five scenarios. There are four columns for each for the results Table 4.2. Test Metrics Used for Policy Testing of C10B, SACSIM, or Both Metric Test Notes1 2 3 4 5 Demand Person trips by mode Both Both Both Both Both Changes in mode choice Person trips by depart time Both Both Changes in depart time Person tours by tour destination Both Both Changes in location of activities Traffic Assignment Segment daily volume Both Both Both Both Both Compare changes in volumes across models Segment volume by hour C10B C10B Changes in vehicle volumes and timing of trips Segment volume by aggregate time period SACSIM SACSIM Changes in vehicle volumes and timing of trips Segment average speed by hour C10B C10B Changes in speed over time Segment average speed by aggregate time period SACSIM SACSIM Changes in speed over time Vehicle-hours of delay by aggregate time period Both Freeway only, using 35 mph as threshold speed Transit Assignment Transit line daily passenger boardings C10B Both Both Compare changes across models Transit boardings by hour C10B C10B C10B Changes in volumes and timing of trips Transit boardings by aggregate time period SACSIM SACSIM SACSIM Changes in volumes and timing of trips Maximum load on vehicles by hour C10B C10B C10B Effects of vehicle capacity

62 Table 4.3. Policy Testing—Demand Effects Summary Hypothesis/Expectation SACSIM C10B Test Effect Random Effects (/) “Correct” Sign? Test Effect > Random? Test Effect Random Effects (/) “Correct” Sign? Test Effect > Random? Test 1. Transit Service Coverage—Test Extends Service on Rte. 11 from 6:00 p.m. to Midnight Increase in transit person trips n/a n/a n/a n/a +1.3% Unknown Yes Unlikely Increase in walk person trips n/a n/a n/a n/a +1.6% Unknown Yes Likely Decrease in private auto mode person trips n/a n/a n/a n/a -0.07% Unknown Yes Unlikely Increase in transit person trips after 6 p.m. n/a n/a n/a n/a -1.5% Unknown No n/a Test 2. Operations-Oriented Interchange—Test Removes Interchange Improvements from I-80/Douglas Decrease in tour destinations around interchange -0.12% 0.04% Yes Yes +0.57% Unknown No Likely • Roseville West +0.45% 0.04% No Yes +0.48% Unknown No Likely • Roseville East -0.74% 0.04% Yes Yes +1.04% Unknown No Likely • Granite Bay +0.11% 0.04% No Yes +1.97% Unknown No Likely • Citrus Heights +0.14% 0.04% No Yes -0.08% Unknown Yes Unlikely Decrease in total number of tours -0.02% 0.04% Yes No -0.05% Unknown Yes Unlikely Decrease in % of local area trip destinations during peak hours -0.34% Unknown Yes n/a -9.56% Unknown Yes n/a • Work trips -0.40% Unknown Yes n/a -13.27% Unknown Yes n/a • All other trips -0.32% Unknown Yes n/a -8.19% Unknown Yes n/a • Work trips change less than nonwork trips n/a Unknown No n/a n/a Unknown No n/a Test 3. Freeway Bottleneck—Add Lanes to Congested Segment Increase in tour destinations around bridge +0.18% 0.04% Yes Yes -0.59% Unknown No Likely • Downtown +0.41% 0.04% Yes Yes -1.31% Unknown No Likely • North Sacramento +0.03% 0.04% Yes No -0.81% Unknown No Likely • Arden Arcade +0.14% 0.04% Yes Yes -1.31% Unknown No Likely (continued on next page)

63 Table 4.3. Policy Testing—Demand Effects Summary (continued) Hypothesis/Expectation SACSIM C10B Test Effect Random Effects (/) “Correct” Sign? Test Effect > Random? Test Effect Random Effects (/) “Correct” Sign? Test Effect > Random? • East Sacramento +0.001% 0.04% Yes No -0.68% Unknown No Likely • Carmichael +0.24% 0.04% Yes Yes -0.14% Unknown No Likely • Antelope/North Highlands +0.19% 0.04% Yes Yes +1.53% Unknown Yes Likely Increase in total number of tours +0.03% 0.04% Yes No +0.28% Unknown Yes Likely Increase in % of local area trip destinations during peak hours +0.18% Unknown Yes n/a -7.51% Unknown No n/a • Work trips +0.04% Unknown Yes n/a -8.16% Unknown No n/a • All other trips +0.25% Unknown Yes n/a -7.12% Unknown No n/a • Work trips change less than nonwork trips n/a Unknown Yes n/a n/a Unknown No n/a Other Changes—No hypothesis/expectation Change in transit person trips +3.63% 1.06% n/a Yes +5.64% Unknown n/a Likely Change in bike/walk person trips -0.03% 0.30% n/a No +0.07% Unknown n/a Unlikely Change in private auto mode person trips +0.14% 0.04% n/a Yes +1.05% Unknown n/a Likely Test 4. Transit Route Frequency (30-to-10-minute headway) Increase in transit person trips +3.15% 1.06% Yes Yes +5.12% Unknown Yes Likely Increase in walk person trips +0.38% 0.30% Yes Yes +1.46% Unknown Yes Likely Decrease in private auto mode person trips -0.06% 0.04% Yes Yes -0.09% Unknown Yes Unlikely No change in total person trips +0.002% 0.04% n/a No +0.10% Unknown No Unlikely Test 5. Transit Route Presence (bus line deleted) Decrease in transit person trips -1.75% 1.06% Yes Yes +1.76% Unknown No Unlikely Increase in walk person trips +0.29% 0.30% Yes No +0.74% Unknown Yes Unlikely Increase in private auto mode person trips +0.02% 0.04% Yes No +0.01% Unknown Yes Unlikely No change in total person trips +0.02% 0.04% n/a No +0.08% Unknown Yes Unlikely Notes: Shaded areas indicate results which were not the correct sign and the test effect was greater than, or likely to be greater than, the random effect. Random effects are unknown for C10B model. Estimate of random effect for purposes of evaluation was three times the SACSIM random effect. C10B model test effects greater than three times SACSIM random effect were deemed likely to be greater than random.

64 using SACSIM and the C10B integrated model. “Test effect” shows the percentage change from the baseline scenario for each measure. “Random Effects (+/-)” shows the expected range of variation (available only from SACSIM for certain measures). “Correct Sign?” is an indicator of whether the model results changed in the expected direction for the given measure. “Test Effect > Random?” is an indicator of whether the change in results exceeds the expected range of variation. For the C10B integrated model, this last indicator is shown as “likely” if the difference is greater than three times the SACSIM variation and “unlikely” otherwise. No analysis of changes to highway assignment was included in the transit related tests (1, 4, and 5). The results for the individual scenarios are discussed in the following subsections. Test 1 Results This test was performed using only the C10B integrated model. Scenario 1 extended service coverage on one bus route, the #11-Truxel route, which connects the Natomas neighborhoods to Downtown Sacramento, just to the south of Natomas and across the American River. The route currently runs from 6:00 a.m. to 6:00 p.m. on 30-min headways during the com- mute peaks, and hourly headways during the midday. The test involved extending service coverage beyond 6:00 p.m. to mid- night, at 60-min headways. Because SACSIM is limited to representing day-long service in two generic peak and off- peak service periods, the effect of extending service coverage is impossible to explicitly model using SACSIM. Highlights of this test include the following: • Demand effects conformed to the expected direction of the result for three metrics, as shown in Table 4.4. These metrics are 44 Transit person trips (increase); 44 Walk person trips (increase); and 44 Private auto person trips (decrease). Table 4.4 provides details of the trips by mode for the baseline and Scenario 1, and Table 4.5 provides details of the transit trips in the vicinity of the revised route by time period. The test effects were unlikely to exceed random effects for two of the three metrics. Oddly, for Scenario 1 compared with the baseline, the test resulted in slightly fewer total transit person trips made between 6:00 p.m. and 8:00 p.m., which is a significant part of the test. Most of the added transit person trips occurred between 10:00 a.m. and 3:00 p.m., and between 8:00 p.m. and midnight. While this result seems likely to be the result of the stochastic nature of the simulation, it does indicate that some of the “new” riders after 6:00 p.m. are attracted from other transit routes. • Transit assignment effects conform to the expected direc- tion of result. Under Scenario 1, Route 11 generated Table 4.4. Person Trips by Mode—Test 1 Mode Baseline Scenario 1 Difference Percent Difference Transit auto access 10,999 10,915 -84 -0.8% Transit walk access 59,759 60,765 +1,006 +1.7% Transit (total) 70,758 71,680 +922 +1.3% School bus 91,627 92,248 +621 +0.7% Shared ride 3+ 2,264,407 2,261,893 -2,514 -0.1% Shared ride 2 2,044,692 2,042,891 -1,801 -0.1% Drive alone 3,615,848 3,615,022 -826 -0.0% Bike 97,686 99,043 +1,357 +1.4% Walk 521,898 530,005 +8,107 +1.6% Total 8,706,916 8,712,782 +5,866 +0.1% Table 4.5. Transit Trips by Time Period—Test 1 in Localized Test Area Time Period Transit Trips Difference Baseline Test # % 5 a.m. to 10 a.m. 13,411 13,380 -31 -0.2% 10 a.m. to 3 p.m. 5,448 5,592 +144 +2.6% 3 p.m. to 6 p.m. 7,527 7,513 -14 -0.2% 6 p.m. to 8 p.m. 1,269 1,219 -50 -3.9% After 8 p.m. 793 812 +19 +2.4% Total 28,448 28,516 +68 +0.2%

65 about 50 additional passenger boardings compared with the baseline (see Table 4.6). About 80 boardings were after 6:00 p.m. (see Figure 4.1), implying some rescheduling of trips due to the extended service. In summary, in Test 1, the C10B integrated model behaved plausibly in an aggregate sense, shifting trips to the transit and walk modes from the auto modes and showing reasonable sensitivity and magnitude of response while maintaining a relatively constant level of demand. Board- ings on the route for which service was extended increased while boardings on nearby routes declined. A significant part of the added boardings occurred in the extended ser- vice period between 6:00 p.m. and midnight. Even with the level of noise in the C10B model, it seems unlikely that the entirety of the model response is indistinguishable from random noise since the mode shifts and changes in boardings on individual routes are nearly all in the correct direction. In terms of localized effects, however, the C10B integrated model showed only a minor impact on transit trips. The tem- poral shifts are also counterintuitive since trips shifted from the period when the service was extended. Test 2 Results Scenario 2 involved “uncoding” an operations-oriented inter- change design improvement from the baseline scenario. The project location was the Douglas Boulevard interchange of Interstate 80 in the City of Roseville. The improvement that was removed involved three major components: (1) construction of a direct connector ramp from the eastbound Douglas Table 4.6. Summary of Transit Passenger Boardings—Test 1 Test Metric Count SACSIM Baseline C10B Baseline C10B Scenario 1 C10B Percent Increase Daily Boardings on Test Route Route 11 940 631 948 984 +3.8% Other Nearby Routes Rte 13 (near Rte 11) 480 309 168 159 -5.4% Rte 86 (near Rte 11) 2,240 1,797 1,700 1,610 -5.3% Rte 88 (near Rte 11) 1,280 1,586 1,163 1,166 +0.3% Sum of nearby routes 4,000 3,692 3,031 2,935 -3.2% All other RT buses in C10B model 49,100 49,634 48,996 49,591 +1.2% Baseline Boardings Scenario 1 Boardings Baseline Max Load Scenario 1 Max Load Figure 4.1. Route 11 transit boardings and vehicle loads for Scenario 1.

66 Boule vard overcrossing to southbound Sunrise Boulevard, a major north/south arterial just east of the interchange; (2) construction of a tunnel and direct ramp connection from northbound Sunrise Boulevard to eastbound/northbound I-80, with the tunnel running below the Douglas Boulevard overcrossing and other ramps; and (3) widening of the west- bound Douglas Boulevard overcrossing itself, and addition of a second point of access to the westbound Douglas–to– westbound/southbound I-80 loop on-ramp. Highlights of this test include the following: • A major impact on traffic patterns was expected (given that the test removes a significant improvement), resulting in higher congestion in the project area and fewer tour des- tinations in the project area. For both models, however, testing showed that tour destinations increased to most subareas around the project. The summary Table 4.3 dem- onstrates these effects. Table 4.7 presents details for the neighborhoods near the project location. One exception was for Roseville East for the SACSIM model, which showed the expected decrease; the decrease was large enough that the overall number of tour destinations decreased across all subareas. • Another expected result of the demand testing was a decrease in the number of peak-period trips made with destinations in the project area. Both models showed this expected result (see Table 4.3 for the summary and Table 4.8 for details). However, the magnitude of the C10B model test effect— nearly a 10% reduction—is questionable. • The traffic assignment results on the two new direct con- nectors for the two models (see Table 4.9) were fairly consistent: 44 The C10B integrated model showed about 12,000 daily vehicles using the northbound Sunrise to eastbound/ northbound I-80 ramp (which was part of the interchange improvement and therefore appears only in the baseline scenario) compared with 10,500 for the same link for SACSIM. 44 The C10B integrated model showed about 17,500 daily vehicles using the eastbound Douglas to southbound Sunrise connector (which was part of the interchange improvement and therefore appears only in the baseline scenario) compared with about 15,000 for the same link for SACSIM. • Table 4.9 also provides summaries of total volumes on rel- evant highway segments. Overall, the C10B model shows greater sensitivity to the change of segment volumes on ramps and arterials (a decrease of 4.5% in daily volumes compared with a 0.2% decrease for the SACSIM model) and on I-80 as well (2.5% decrease in total daily volumes compared with a 1.1% decrease for the SACSIM model). 44 The C10B model showed significant changes in total volumes on I-80 westbound, north of Douglas (-11%). Changes of this magnitude on I-80 were not expected as test results. 44 Both models showed an increase in traffic on Douglas Boulevard east of I-80, and some level of increase makes sense—some of the traffic which uses the new direct ramps to/from I-80 and Sunrise might shift to Douglas when those facilities were taken out. However, the mag- nitude of the increase in the C10B model (about 30% compared with about 15% for the SACSIM model) seems high. • Figure 4.2 provides segment-by-segment graphical repre- sentations of C10B and SACSIM model results. As men- tioned earlier, the most remarkable difference is in the level of detail in results for the C10B model. While SACSIM can provide total volumes for aggregate periods ranging from 3 to 13 h in length, the C10B model provides results for any time periods that the analyst wishes to define (for these figures, the results were aggregated to individual hours for the entire 24-h day). Table 4.7. Person Tour Primary Destinations Near Project—Test 2 Tour Destination SACSIM C10B Integrated Model Baseline Scenario 2 Difference Percent Difference Baseline Scenario 2 Difference Percent Difference Roseville West 95,600 96,029 +429 +0.4% 105,541 106,045 +504 +0.5% Roseville East 150,651 149,540 -1,111 -0.7% 157,201 158,839 +1,638 +1.0% Granite Bay 20,142 20,164 +22 +0.1% 21,903 22,334 +431 +2.0% Citrus Heights 139,596 139,784 +188 +0.1% 147,325 147,205 -120 -0.1% Combined areas 405,989 405,517 -472 -0.1% 431,970 434,423 +2,453 +0.6% All other areas 3,110,819 3,110,647 -172 -0.0% 3,377,289 3,376,375 -914 -0.0% All destinations 3,516,808 3,516,164 -644 -0.0% 3,809,259 3,810,798 +1,539 +0.0%

67 In summary, in this scenario, the highway system reverted to an earlier state in that a key interchange improvement was removed. Highway capacity was therefore lower, resulting—as expected—in a higher level of congestion in the affected area in both models. The higher level of congestion apparently caused some travelers to shift to transit, as shown in Table 4.3. Overall, the C10B integrated model was more sensitive to con- gestion than SACSIM, shifting a significantly greater number of travelers from peak periods to adjacent time periods. The SACSIM results showed reductions in all time periods (though very small reductions) rather than any noticeable peak spread- ing. Interestingly, on the one hand, the C10B model showed a smaller reduction in trips on work tours than on nonwork tours; this result is consistent with the notion that work tours are more inelastic. SACSIM, on the other hand, showed a greater reduction in trips on work tours. It is unclear whether the magnitude of the sensitivity of the C10B integrated model is reasonable; the SACSIM results seem too inelastic. The C10B model seems very sensitive in terms of shifts in demand from peak periods, but the relative inelasticity of the SACSIM results does not provide a worth- while basis for comparison. The assignment results for both SACSIM and the C10B integrated model for five key road- ways showed changes in the expected direction although the predicted volume levels and the magnitude of the impacts vary among roadways. Test 3 Results Scenario 3 involved coding an additional general purpose freeway lane on a congested segment of the Capital City Free- way (I-80 Business) at its crossing of the American River, Table 4.8. Timing of Person Trips on Tours to Local Test Area—Test 2 Timing SACSIM Model C10B Model Baseline Scenario 2 Difference Percent Difference Baseline Scenario 2 Difference Percent Difference Trips on Work Tours A.M. peak (3 hrs) 43,059 43,171 +112 +0.3% 47,915 38,029 -9,886 -20.6% P.M. peak (3 hrs) 53,196 52,696 -500 -0.9% 61,144 56,562 -4,582 -7.5% Total peak period 96,255 95,867 -388 -0.4% 109,059 94,591 -14,468 -13.3% Midday (5 hrs) 35,332 35,049 -283 -0.8% 43,425 45,758 +2,333 +5.4% Late evening/early A.M. (13 hrs) 50,484 50,209 -275 -0.5% 59,218 64,180 +4,962 +8.4% Total off-peak period 85,816 85,258 -558 -0.7% 102,643 109,938 +7,295 +7.1% Total weekday 182,071 181,125 -946 -0.5% 211,702 204,529 -7,173 -3.4% Trips on All Nonwork Tours A.M. peak (3 hrs) 128,482 128,200 -282 -0.2% 148,263 124,393 -23,870 -16.1% P.M. peak (3 hrs) 160,092 159,459 -633 -0.4% 146,244 146,004 -240 -0.2% Total peak period 288,574 287,659 -915 -0.3% 294,507 270,397 -24,110 -8.2% Midday (5 hrs) 253,768 253,203 -565 -0.2% 248,074 250,861 +2,787 +1.1% Late evening/early A.M. (13 hrs) 232,659 232,581 -78 -0.0% 237,508 219,603 -17,905 -7.5% Total off-peak period 486,427 485,784 -643 -0.1% 485,582 470,464 -15,118 -3.1% Total weekday 775,001 773,443 -1,558 -0.2% 780,089 740,861 -39,228 -5.0% All Tours A.M. peak (3 hrs) 171,541 171,371 -170 -0.1% 196,178 162,422 -33,756 -17.2% P.M. peak (3 hrs) 213,288 212,155 -1,133 -0.5% 207,388 202,566 -4,822 -2.3% Total peak period 384,829 383,526 -1,303 -0.3% 403,566 364,988 -38,578 -9.6% Midday (5 hrs) 289,100 288,252 -848 -0.3% 291,499 296,619 +5,120 +1.8% Late evening/early A.M. (13 hrs) 283,143 282,790 -353 -0.1% 296,726 283,783 -12,943 -4.4% Total off-peak period 572,243 571,042 -1,201 -0.2% 588,225 580,402 -7,823 -1.3% Total weekday 957,072 954,568 -2,504 -0.3% 991,791 945,390 -46,401 -4.7%

68 Table 4.9. Vehicle Volumes on Roads In/Around Local Test Area—Test 2 Roadway Segment Lane Type Direction Lanes Weekday Total Volume NotesBaseline Test Baseline Scenario 2 Difference Percent Difference SACSIM Model I-80 freeway South of Douglas General purpose EB 3 3 94,393 94,189 -204 -0.2% No large changes expected on I-80 North of Douglas General purpose EB 3 3 84,884 84,095 -789 -0.9% South of Douglas General purpose WB 3 3 95,071 96,234 +1,163 +1.2% North of Douglas General purpose WB 3 3 84,373 80,407 -3,966 -4.7% I-80/Douglas Blvd. ramps From I-80 EB to EB Douglas Diagonal n/a 1 1 19,059 17,994 -1,065 -5.6% No large change expected From I-80 EB to WB Douglas Loop n/a 1 1 3,007 2,991 -16 -0.5% No large change expected From WB Douglas to I-80 EB Diagonal n/a 1 1 2,057 10,893 +8,836 +429.6% Change in expected direction From I-80 WB to WB Douglas Diagonal n/a 1 1 11,936 5,598 -6,338 -53.1% Change in expected direction From WB Douglas to I-80 WB Loop n/a 1 1 18,151 15,805 -2,346 -12.9% No large change expected From EB Douglas to I-80 WB Diagonal n/a 1 1 4,482 5,618 +1,136 +25.3% No large change expected Douglas Blvd. West of I-80 Arterial EB 2 2 15,286 13,308 -1,978 -12.9% No large change expected I-80 overcrossing Arterial EB 2 2 6,124 11,046 +4,922 +80.4% Change in expected direction East of I-80 Arterial EB 2 2 24,432 28,297 +3,865 +15.8% No large change expected East of I-80 Arterial WB 3 2 28,432 32,877 +4,445 +15.6% No large change expected I-80 overcrossing Arterial WB 3 2 30,136 25,720 -4,416 -14.7% No large change expected West of I-80 Arterial WB 2 2 13,839 12,152 -1,687 -12.2% No large change expected Sunrise Blvd. South of I-80 interchange/Douglas Arterial NB 2 2 24,700 23,102 -1,598 -6.5% Change in expected direction South of I-80 interchange/Douglas Arterial SB 2 2 26,285 22,027 -4,258 -16.2% Change in expected direction Direct connectors From NB Sunrise Blvd. On-ramp n/a 1 10,500 — -10,500 To SB Sunrise Blvd. Connector n/a 1 14,757 — -14,757 Subtotal of all ramps and arterials without new direct connectors 227,926 227,428 -498 -0.2% Change in expected direction Subtotal of freeway segments 358,721 354,925 -3,796 -1.1% Change in expected direction (continued on next page)

69 Table 4.9. Vehicle Volumes on Roads In/Around Local Test Area—Test 2 (continued) Roadway Segment Lane Type Direction Lanes Weekday Total Volume NotesBaseline Test Baseline Scenario 2 Difference Percent Difference C10B Integrated Model I-80 freeway South of Douglas General purpose EB 4 4 96,492 97,097 +605 +0.6% No large changes expected on I-80 North of Douglas General purpose EB 4 4 89,678 87,600 -2,078 -2.3% South of Douglas General purpose WB 3 3 88,582 89,568 +986 +1.1% North of Douglas General purpose WB 4 4 76,712 68,305 -8,407 -11.0% I-80/Douglas Blvd. ramps From I-80 EB to EB Douglas Diagonal n/a 1 1 18,684 17,006 -1,678 -9.0% No large change expected From I-80 EB to WB Douglas Loop n/a 1 1 4,464 4,553 +89 +2.0% No large change expected From WB Douglas to I-80 EB Diagonal n/a 1 1 4,375 12,062 +7,687 +175.7% Change in expected direction From I-80 WB to WB Douglas Diagonal n/a 3 1 33,053 11,014 -22,039 -66.7% Change in expected direction From WB Douglas to I-80 WB Loop n/a 2 1 5,079 10,357 +5,278 +103.9% Large change not expected From EB Douglas to I-80 WB Diagonal n/a 3 1 39,844 21,920 -17,924 -45.0% Large change not expected Douglas Blvd. West of I-80 Arterial EB 2 2 23,620 22,897 -723 -3.1% I-80 overcrossing Arterial EB 2 2 7,533 16,453 +8,920 +118.4% Change in expected direction East of I-80 Arterial EB 2 2 25,537 33,624 +8,087 +31.7% Large change not expected East of I-80 Arterial WB 2 2 29,954 38,664 +8,710 +29.1% Large change not expected I-80 overcrossing Arterial WB 2 2 30,723 30,990 +267 +0.9% No large change expected West of I-80 Arterial WB 2 2 17,396 16,171 -1,225 -7.0% No large change expected Sunrise Blvd. South of I-80 interchange/Douglas Arterial NB 2 2 19,230 16,606 -2,624 -13.6% Change in expected direction South of I-80 interchange/Douglas Arterial SB 2 2 24,166 18,612 -5,554 -23.0% Change in expected direction Direct connectors From NB Sunrise Blvd. On-ramp n/a 1 11,959 -11,959 -100.0% To SB Sunrise Blvd. Connector n/a 1 17,544 -17,544 -100.0% Subtotal of all ramps and arterials without new direct connectors 283,658 270,929 -12,729 -4.5% Change in expected direction Subtotal of freeway segments 351,464 342,570 -8,894 -2.5% Change in expected direction Notes: EB = eastbound; WB = westbound; NB = northbound; SB = southbound.

70 Segment SACSIM C10B Integrated Model Fig. 4.2A I-80 freeway eastbound, south of Douglas, general purpose lanes Fig. 4.2B I-80 freeway eastbound, north of Douglas, general purpose lanes Fig. 4.2C I-80 freeway westbound, south of Douglas, general purpose lanes Figure 4.2. Vehicle volumes and speeds on key test segments—Test 2. (Continued on next page.)

71 Segment SACSIM C10B Integrated Model Fig. 4.2D I-80 freeway westbound, north of Douglas, general purpose lanes Fig. 4.2E I-80 eastbound off-ramp to eastbound Douglas Fig. 4.2F I-80 eastbound off-ramp to westbound Douglas Figure 4.2. Vehicle volumes and speeds on key test segments—Test 2 (continued).

72 Segment SACSIM C10B Integrated Model Fig. 4.2G I-80 eastbound on-ramp from Douglas Fig. 4.2H I-80 westbound off-ramp to Douglas Fig. 4.2I I-80 westbound on-ramp from westbound Douglas Figure 4.2. Vehicle volumes and speeds on key test segments—Test 2 (continued).

73 Segment SACSIM C10B Integrated Model Fig. 4.2J I-80 westbound on-ramp from eastbound Douglas Fig. 4.2K Douglas Blvd. eastbound, west of I-80 ramps Fig. 4.2L Douglas Blvd. eastbound on overcrossing Figure 4.2. Vehicle volumes and speeds on key test segments—Test 2 (continued).

74 Fig. 4.2M Douglas Blvd. eastbound, east of I-80 ramps Segment SACSIM C10B Integrated Model Fig. 4.2N Douglas Blvd. westbound, east of I-80 ramps Fig. 4.2O Douglas Blvd. westbound on I-80 overcrossing Figure 4.2. Vehicle volumes and speeds on key test segments—Test 2 (continued).

75 Segment SACSIM C10B Integrated Model Fig. 4.2P Douglas Blvd. westbound, west of I-80 ramps Fig. 4.2Q Sunrise Blvd. northbound, south of I-80 interchange Fig. 4.2R Sunrise Blvd. northbound, south of I-80 interchange Figure 4.2. Vehicle volumes and speeds on key test segments—Test 2 (continued).

76 between the North Sacramento and Arden-Arcade areas and downtown Sacramento. This segment consistently comes up as one of the most congested freeway segments in the region. The test scenario added a fourth general purpose lane between the E Street and Exposition Boulevard interchanges. High- lights of this test include the following: • A major impact on demand was expected, given that the test adds significant capacity on a highly congested, high- volume roadway segment; adding capacity was expected to increase the number of tours with destinations in and around the test area. The SACSIM results conformed to this expectation, with a 0.2% increase over the baseline scenario. The biggest effect was on downtown Sacra- mento, with a 0.4% increase. (In addition to the summary provided in Table 4.3, Table 4.10 provides further details.) The C10B model showed the opposite, counterintuitive result—a decrease in tour destinations in and around the test area. • Another expected result of the demand testing was an increase in the number of peak-period trips made with des- tinations in the project area, assuming that the higher con- gestion in the baseline scenario had caused some peak spreading. SACSIM showed a small increase, but the C10B model showed a decrease of about 8%. (In addition to the summary provided in Table 4.3, Table 4.11 provides addi- tional details.) The effect on timing of travel was expected to be greater for nonwork, discretionary trips than for work trips; this result was borne out in the SACSIM results but not in the results for the C10B integrated model. • Traffic assignment results on the two test segments varied significantly between the two models (see Table 4.12): 44 SACSIM showed balanced increases in both directions (+5.4% eastbound, +6.4% westbound). Segment SACSIM C10B Integrated Model Fig. 4.2S I-80 eastbound on-ramp from northbound Sunrise (direct ramp taken away in Test 2) Fig. 4.2T Douglas Blvd. eastbound connector to southbound Sunrise (direct ramp taken away in Test 2) Figure 4.2. Vehicle volumes and speeds on key test segments—Test 2 (continued).

77 Table 4.10. Person Tour Primary Destinations Near Project—Test 3 Tour Destination SACSIM C10B Integrated Model Baseline Scenario 3 Difference Percent Difference Baseline Scenario 3 Difference Percent Difference Downtown 277,046 278,180 +1,134 +0.4% 321,117 316,927 -4,190 -1.3% North Sacramento 131,128 131,162 +34 +0.0% 146,843 145,648 -1,195 -0.8% Arden Arcade 196,925 197,193 +268 +0.1% 216,477 213,634 -2,843 -1.3% East Sacramento 232,428 232,431 +3 +0.0% 260,984 259,205 -1,779 -0.7% Carmichael 67,112 67,272 +160 +0.2% 71,756 71,654 -102 -0.1% Antelope-North Highlands 181,089 181,424 +335 +0.2% 191,791 194,718 +2,927 +1.5% Combined areas 1,085,728 1,087,662 +1,934 +0.2% 1,208,968 1,201,786 -7,182 -0.6% All other areas 2,430,436 2,429,560 -876 -0.0% 2,600,291 2,618,140 +17,849 +0.7% All destinations 3,516,164 3,517,222 +1,058 +0.0% 3,809,259 3,819,926 +10,667 +0.3% Table 4.11. Timing of Person Trips on Tours to Local Test Area—Test 3 Tour Timing SACSIM Model C10B Model Baseline Scenario 3 Difference Percent Difference Baseline Scenario 3 Difference Percent Difference Trips on Work Tours A.M. peak (3 hrs) 153,833 153,545 -288 -0.2% 180,292 151,595 -28,697 -15.9% P.M. peak (3 hrs) 203,781 204,228 +447 +0.2% 242,494 236,692 -5,802 -2.4% Total peak period 357,614 357,773 +159 +0.0% 422,786 388,287 -34,499 -8.2% Midday (5 hrs) 131,855 131,877 +22 +0.0% 162,893 177,244 +14,351 +8.8% Late evening/early A.M. (13 hrs) 209,858 209,560 -298 -0.1% 241,942 261,098 +19,156 +7.9% Total off-peak period 341,713 341,437 -276 -0.1% 404,835 438,342 +33,507 +8.3% Total weekday 699,327 699,210 -117 -0.0% 827,621 826,629 -992 -0.1% Trips on All Nonwork Tours A.M. peak (3 hrs) 298,202 298,985 +783 +0.3% 340,622 282,904 -57,718 -16.9% P.M. peak (3 hrs) 377,627 379,106 +1,479 +0.4% 351,438 359,901 +8,463 +2.4% Total peak period 675,829 678,091 +2,262 +0.3% 692,060 642,805 -49,255 -7.1% Midday (5 hrs) 626,030 628,225 +2,195 +0.4% 653,334 667,418 +14,084 +2.2% Late evening/early A.M. (13 hrs) 538,063 538,188 +125 +0.0% 573,705 591,718 +18,013 +3.1% Total off-peak period 1,164,093 1,166,413 +2,320 +0.2% 1,227,039 1,259,136 +32,097 +2.6% Total weekday 1,839,922 1,844,504 4,582 +0.2% 1,919,099 1,901,941 -17,158 -0.9% All Tours A.M. peak (3 hrs) 452,035 452,530 +495 +0.1% 520,914 434,499 -86,415 -16.6% P.M. peak (3 hrs) 581,408 583,334 +1,926 +0.3% 593,932 596,593 +2,661 +0.4% Total peak period 1,033,443 1,035,864 +2,421 +0.2% 1,114,846 1,031,092 -83,754 -7.5% Midday (5 hrs) 757,885 760,102 +2,217 +0.3% 816,227 844,662 +28,435 +3.5% Late evening/early A.M. (13 hrs) 747,921 747,748 -173 -0.0% 815,647 852,816 +37,169 +4.6% Total off-peak period 1,505,806 1,507,850 +2,044 +0.1% 1,631,874 1,697,478 +65,604 +4.0% Total weekday 2,539,249 2,543,714 4,465 +0.2% 2,746,720 2,728,570 -18,150 -0.7%

78 Table 4.12. Vehicle Volumes on Roads In and Around Local Test Area—Test 3 Roadway Segment Lane Type Direction Lanes Weekday Total Volume NotesBaseline Test Baseline Scenario 3 Difference Percent Difference SACSIM Model Capital City Freeway South of E St. General purpose EB 3 3 81,496 86,095 +4,599 +5.6% South of test segment— expected increase South of Exposition Blvd. General purpose EB 3 4 89,004 93,828 +4,824 +5.4% Test segment—expected increase South of Arden Way General purpose EB 2 2 59,562 62,368 +2,806 +4.7% North of test segment— expected increase North of Arden Way Auxiliary EB 1 1 10,893 10,837 -56 -0.5% No large change expected North of Arden Way General purpose EB 4 4 86,247 87,693 +1,446 +1.7% North of test segment— expected increase North of Arden Way General purpose WB 4 4 101,635 103,731 +2,096 +2.1% North of test segment— expected increase South of Arden Way General purpose WB 3 3 68,761 71,661 +2,900 +4.2% North of test segment— expected increase South of Arden Way Auxiliary WB 1 1 3,648 4,442 +794 +21.8% North of test segment— expected increase South of Exposition Blvd. General purpose WB 3 4 95,150 101,213 +6,063 +6.4% Test segment—expected increase South of E St. General purpose WB 3 3 69,153 73,789 +4,636 +6.7% North of test segment— expected increase South of E St. HOV WB 1 1 19,057 19,751 +694 +3.6% North of test segment— expected increase 16th St. South of N B St. Arterial NB/EB 3 3 35,966 34,618 -1,348 -3.7% Expected decrease 12th St. South of N B St. Arterial SB/WB 4 4 32,412 31,259 -1,153 -3.6% Expected decrease SR 160 West of Royal Oaks General purpose EB 2 2 26,413 25,325 -1,088 -4.1% Expected decrease West of Royal Oaks General purpose WB 2 2 34,794 33,618 -1,176 -3.4% Expected decrease Exposition Blvd. East of Bus 80 Arterial EB 3 3 28,882 29,180 +298 +1.0% No large change expected East of Bus 80 Arterial WB 3 3 17,896 18,697 +801 +4.5% No large change expected Arden Way East of Bus 80 Arterial WB 4 4 15,211 15,007 -204 -1.3% No large change expected East of Bus 80 Arterial EB 4 4 21,138 20,562 -576 -2.7% No large change expected West of Bus 80 Arterial EB 4 4 19,398 19,367 -31 -0.2% No large change expected West of Bus 80 Arterial WB 4 4 26,497 26,085 -412 -1.6% No large change expected Subtotal of freeway segments 684,606 715,408 +30,802 +4.5% Expected increase Subtotal of arterial segments 258,607 253,718 -4,889 -1.9% Expected decrease (continued on next page)

79 Table 4.12. Vehicle Volumes on Roads In and Around Local Test Area—Test 3 (continued) Roadway Segment Lane Type Direction Lanes Weekday Total Volume NotesBaseline Test Baseline Scenario 3 Difference Percent Difference C10B Integrated Model Capital City Freeway South of E St. General purpose EB 3 72,896 78,240 +5,344 +7.3% South of test segment— expected increase South of Exposition Blvd. General purpose EB 3 90,548 93,834 +3,286 +3.6% Test segment—expected increase South of Arden Way General purpose EB 2 63,919 63,267 -652 -1.0% North of test segment— expected increase North of Arden way Auxiliary EB 1 19,565 18,770 -795 -4.1% No large change expected North of Arden Way General purpose EB 4 92,241 92,202 -39 -0.0% North of test segment— expected increase North of Arden Way General purpose WB 4 108,024 109,606 +1,582 +1.5% North of test segment— expected increase South of Arden Way General purpose WB 3 62,094 71,032 +8,938 +14.4% North of test segment— expected increase South of Arden Way Auxiliary WB 1 11,008 11,215 +207 +1.9% North of test segment— expected increase South of Exposition Blvd. General purpose WB 3 94,223 112,221 +17,998 +19.1% Test segment—expected increase South of E St. General purpose WB 3 69,800 81,511 +11,711 +16.8% North of test segment— expected increase South of E St. HOV WB 1 11,387 15,124 +3,737 +32.8% North of test segment— expected increase 16th St. South of N B St. Arterial NB/EB 3 42,217 43,189 +972 +2.3% Expected decrease 12th St. S of N B St. Arterial SB/WB 4 37,157 35,357 -1,800 -4.8% Expected decrease SR 160 West of Royal Oaks General purpose EB 2 40,590 41,764 +1,174 +2.9% Expected decrease West of Royal Oaks General purpose WB 2 33,448 28,964 -4,484 -13.4% Expected decrease Exposition Blvd. East of Bus 80 Arterial EB 3 19,740 21,492 +1,752 +8.9% No large change expected East of Bus 80 Arterial WB 3 17,086 21,044 +3,958 +23.2% No large change expected Arden Way East of Bus 80 Arterial WB 4 30,960 30,431 -529 -1.7% No large change expected East of Bus 80 Arterial EB 4 32,060 33,075 +1,015 +3.2% No large change expected West of Bus 80 Arterial EB 4 17,155 18,791 +1,636 +9.5% No large change expected West of Bus 80 Arterial WB 4 20,959 21,058 +99 +0.5% No large change expected Subtotal of freeway segments 695,705 747,022 +51,317 +7.4% Expected increase Subtotal of arterial segments 291,372 295,165 +3,793 +1.3% Expected decrease Notes: EB = eastbound; WB = westbound; NB = northbound; SB = southbound.

80 Table 4.13. Travel Volume, Time, and Delay Summary and Comparison—Test 3: P.M. Peak Period Eastbound (Peak Period and Direction) Segment/ Scenario Volume (vpd) Speed (mph) Travel Time (minutes) Total Vehicle Delay (hours per day) SACSIM C10B SACSIM C10B SACSIM C10B SACSIM C10B River crossing segment Baseline 19,058 13,857 29.5 15.6 4.30 9.52 261.6 1,225.9 Scenario 3 20,533 15,454 41.25 16.7 3.09 7.63 44.4 1,066.4 Difference +1,475 +1,597 +11.8 +1.1 -1.21 -1.88 -217.3 -159.5 Change +8% +12% +40% +7% -28% -20% -83% -13% Average of downstream segments (Exposition Blvd. to North of Arden) Baseline 15,631 12,579 38.8 36.4 2.59 2.29 103.4 39.3 Scenario 3 16,415 12,645 36.5 28.3 2.94 2.96 182.1 155.2 Difference +783 +66 -2.3 -8.1 +0.35 +0.67 +78.7 +116.0 Change +5% +1% -6% -22% +13% +29% +76% +295% All segments (E Street to North of Arden) Baseline 17,629 13,310 29.7 17.7 6.89 11.81 365.1 1,265.1 Scenario 3 18,800 14,309 34.0 19.7 6.03 10.59 226.5 1,221.6 Difference +1,171 +999 +4.3 +2.0 -0.86 -1.22 -138.6 -43.5 Change +7% +8% +14% +11% -13% -10% -38% -3% Note: vpd = vehicles per day. 44 The C10B integrated model showed asymmetrical increases, with only a +3.6% change eastbound, but a +19.1% change westbound. Such a wide difference by direction does not seem to make sense. • There are differences between the results of the two models in terms of speeds and delays, in terms of the magnitudes, the locations of the delays, and the time of day. (For this study, delay was defined by SACOG using a base speed of 35 mph. The specific measure of delay is the difference between actual travel time and travel time at 35 mph, when the actual travel speed is less than 35 mph. In other words, delay is not accrued if travel speeds exceed 35 mph.) Table 4.13 compares the volumes, speeds, travel times, and total vehicle delay for the river crossing seg- ment (the section that is widened in Scenario 3), the segments immediately downstream from the improve- ment, and the total of all segments of the relevant section of I-80 between the two models, for the p.m. peak period (3:00 to 6:00). The volumes are lower in all cases for the C10B integrated model, as are the speeds. On the wid- ened highway segment, there is more total vehicle delay in the C10B model results for both the baseline scenario and Scenario 3, and a lower percentage decrease in delay due to the improvement. Downstream, however, the C10B model shows a higher percentage increase in delay, nearly offsetting the decrease in delay on the improved segment. • The differences in speed results between the two models by time of day are illustrated in Table 4.14. The generally higher speeds in SACSIM are evident across all time periods, espe- cially the p.m. peak and evening periods. For the river cross- ing segment and upstream links, the C10B integrated model shows greater percentage increases in speed for all periods except the p.m. peak. Downstream from the improvement, the C10B model shows substantial speed reductions while SACSIM shows only minor decreases. Table 4.15 shows the corresponding results for vehicle delay. The delays across the segment for the p.m. peak period are illustrated in Figure 4.3. • Figure 4.4 provides segment-by-segment graphical repre- sentations of C10B and SACSIM model results. As already mentioned, the most remarkable difference is in the level of detail in results for the C10B model. Where SACSIM can provide total volumes for 3- to 13-h demand periods, the C10B model provides results for any time period the ana- lyst wishes to define (for these figures, the results were aggregated to individual hours for the entire 24-h day).

81 Table 4.14. Travel Speeds by Segment, Capital City Freeway Eastbound/Northbound—Test 3 Segment Direction Lanes Distance (mi) Travel Speed (mph) Baseline Scenario 3 Percentage Difference Baseline Scenario 3 A.M. Midday P.M. Evening A.M. Midday P.M. Evening A.M. Midday P.M. Evening SACSIM South of E St. EB 3 3 0.50 47 48 37 53 45 46 30 52 -4% -4% -19% -2% North of E St. EB 3 4 0.26 46 48 27 54 53 54 45 57 15% 13% 67% 6% River Crossing EB 3 4 0.64 46 48 27 54 53 54 45 57 15% 13% 67% 6% South of Exposi- tion Blvd. EB 3 4 0.68 46 48 27 54 53 54 45 57 15% 13% 67% 6% Between Exposi- tion Blvd. ramps EB 3 3 0.16 51 52 45 53 51 51 43 53 0% -2% -4% 0% South of Arden Way EB 3 3 0.14 51 52 45 53 51 51 43 53 0% -2% -4% 0% South of Arden Way EB 2 2 0.69 46 45 25 48 44 42 21 47 -4% -7% -16% -2% North of Arden Way EB 4 4 0.36 50 49 40 50 50 49 39 50 0% 0% -3% 0% Total 3.43 47 48 30 52 49 49 34 53 5% 3% 14% 2% C10B Integrated Model South of E St. EB 3 4 0.50 28 32 7 9 49 49 14 20 72% 54% 92% 128% North of E St. EB 3 4 0.26 28 29 10 10 42 43 16 20 50% 45% 54% 103% River Crossing EB 3 4 0.71 30 32 19 17 40 48 20 21 31% 53% 9% 25% South of Exposi- tion Blvd. EB 3 4 0.68 30 30 26 32 22 22 17 19 -26% -26% -35% -40% Between Exposi- tion Blvd. ramps EB 3 3 0.16 47 46 38 46 33 26 22 26 -31% -43% -43% -45% South of Arden Way EB 3 3 0.14 43 41 26 37 22 16 15 19 -50% -60% -44% -49% South of Arden Way EB 2 2 0.69 41 38 32 39 32 28 27 34 -21% -26% -15% -14% North of Arden Way EB 4 4 0.35 56 55 49 53 55 53 50 54 -2% -3% 1% 2% Total 3.48 34 35 18 19 33 33 20 23 -2% -6% 11% 20% Note: EB = eastbound.

82 Table 4.15. Total Vehicle-Hours of Delay by Segment, Capital City Freeway Eastbound/Northbound—Test 3 Segment Direction Lanes Distance (mi) Vehicle-Hours of Delay Baseline Scenario 3 Percentage Difference Baseline Scenario 3 A.M. Midday P.M. Evening A.M. Midday P.M. Evening A.M. Midday P.M. Evening SACSIM South of E St. EB 3 3 0.50 0 0 0 0 0 0 44 0 — — n/a — North of E St. EB 3 4 0.26 0 0 43 0 0 0 0 0 — — -100% — River Crossing EB 3 4 0.64 0 0 106 0 0 0 0 0 — — -100% — South of Exposition Blvd. EB 3 4 0.68 0 0 113 0 0 0 0 0 — — -100% — Between Exposi- tion Blvd. ramps EB 3 3 0.16 0 0 0 0 0 0 0 0 — — — — South of Arden Way EB 3 3 0.14 0 0 0 0 0 0 0 0 — — — — South of Arden Way EB 2 2 0.69 0 0 103 0 0 0 182 0 — — 77% — North of Arden Way EB 4 4 0.36 0 0 0 0 0 0 0 0 — — — — Total 3.43 0 0 365 0 0 0 227 0 — — -38% — C10B Integrated Model South of E St. EB 3 4 0.50 49 53 611 1,315 0 0 319 636 -100% -100% -48% -52% North of E St. EB 3 4 0.26 25 35 259 699 0 0 148 378 -100% -100% -43% -46% River Crossing EB 3 4 0.71 63 89 259 864 31 0 270 741 -51% -100% 4% -14% South of Exposi- tion Blvd. EB 3 4 0.68 48 78 97 161 203 299 329 782 323% 283% 239% 386% Between Exposi- tion Blvd. ramps EB 3 3 0.16 0 0 3 0 7 29 32 75 — — 967% — South of Arden Way EB 3 3 0.14 0 0 17 16 30 87 64 131 — — 276% 719% South of Arden Way EB 2 2 0.69 0 0 19 26 24 79 59 105 — — 211% 304% North of Arden Way EB 4 4 0.35 0 0 0 0 0 0 0 0 — — — — Total 3.48 185 254 1,265 3,081 295 493 1,222 2,849 59% 94% -3% -8% Note: EB = eastbound.

83 SACSIM C10B Integrated Model Figure 4.3. Travel time and average delay by segment, p.m. peak period—Test 3. Segment SACSIM Model C10B Model Fig. 4.4A Capital City Freeway EB, south of E St. Fig. 4.4B Capital City Freeway EB, south of Exposion Blvd. Figure 4.4. Vehicle volumes and speeds on key test segments—Test 3. (Continued on next page.)

84 Figure 4.4. Vehicle volumes and speeds on key test segments—Test 3 (continued). Segment SACSIM Model C10B Model Fig. 4.4C Capital City Freeway EB, south of Arden Way Fig. 4.4D Capital City Freeway EB, north of Arden Way, auxiliary lanes Fig. 4.4E Capital City Freeway EB, north of Arden Way, general purpose lanes Fig. 4.4F Capital City Freeway WB, north of Arden Way, general purpose lanes

85 Segment SACSIM Model C10B Model Fig. 4.4G Capital City Freeway WB, south of Arden Way, general purpose lanes Fig. 4.4H Capital City Freeway WB, south of Arden Way, auxiliary lanes Fig. 4.4I Capital City Freeway WB, south of Exposion Blvd., general purpose lanes (widened from 3 to 4 lanes in Test 3) Fig. 4.4J Capital City Freeway WB, south of E St., general purpose lanes (widened from 3 to 4 lanes in Test 3) Figure 4.4. Vehicle volumes and speeds on key test segments—Test 3 (continued).

86 Segment SACSIM Model C10B Model Fig. 4.4K Capital City Freeway WB, south of E St., HOV lanes Fig. 4.4L 16th St. NB/EB, south of N B St. Fig. 4.4M 12th St. SB/WB, south of N B St. Fig. 4.4N SR 160 EB, west of Royal Oaks, general purpose lanes Figure 4.4. Vehicle volumes and speeds on key test segments—Test 3 (continued).

87 Segment SACSIM Model C10B Model Fig. 4.4O SR 160 WB, west of Royal Oaks, general purpose lanes Fig. 4.4P Exposion Blvd. EB, east of Capital City Freeway Fig. 4.4Q Exposion Blvd. WB, east of Capital City Freeway Fig. 4.4R Arden Way WB, east of Capital City Freeway Figure 4.4. Vehicle volumes and speeds on key test segments—Test 3 (continued).

88 A general note on why the C10B model delay estimates are so much higher than the SACSIM estimates is needed. The delay measure, as already mentioned, is based on a threshold travel speed of 35 mph. Because SACSIM models aggregate time periods (e.g., 3-h peaks and a 13-h late evening/early morning period), modeled travel speeds are blended across all hours in the aggregate period. The highs and lows are blended out. The C10B model operates at very small time slices, which for this set of estimates are aggregated to hours. The highs and lows which occur within the aggregate time periods are specifically modeled in the C10B model. Stated more simply, to calculate delay relative to the 35 mph threshold, the aggre- gate time period travel speed would have to be below 35 mph for the SACSIM model. For the C10B model, delay can accrue for each time slice within the aggregate time period—any time slice within the aggregate period can generate delay. In summary, in Scenario 3, an additional general purpose lane was incorporated on a congested segment of the Capital City Freeway which is the most congested freeway in the region. Both models showed a small increase in the total Segment SACSIM Model C10B Model Fig. 4.4S Arden Way EB, east of Capital City Freeway Fig. 4.4T Arden Way EB, west of Capital City Freeway Fig. 4.4U Arden Way WB, west of Capital City Freeway Figure 4.4. Vehicle volumes and speeds on key test segments—Test 3 (continued).

89 amount of regional travel, with the C10B integrated model showing a larger increase. However, in the SACSIM model this increase was mainly concentrated near the vicinity of the improved highway; in the C10B model, destinations near the improvement decreased while they increased farther away from the improvement. The C10B integrated model results are different from SACSIM for this segment. Both the baseline and Scenario 3 show slower speeds and higher travel times than SACSIM. It is unknown in which model’s results the speeds and volumes are more accurate. Both SACSIM and the C10B model show higher volumes on the widened highway for the test, and both show added delay in the downstream segments. But the C10B model shows the impact of higher volumes on the downstream seg- ments to be much greater than SACSIM. In other words, by widening the crossing segment, delay is reduced on that seg- ment, but that improvement is offset by much higher delay downstream. Widening the bridge segment alone would be nearly net-zero in delay reduction, according to the C10B model results. In this scenario it was anticipated that the increased capac- ity would result in a higher number of trips to the affected area both spatially and temporally. However, such an impact is seen only in SACSIM, not in the C10B integrated model. For this particular scenario, perhaps less than ideal convergence in the C10B model may have left the model with too many localized sources of instability and congestion which have distorted the final outcome. The study team tested using a higher number of DynusT iterations in the last overall iteration, which improved convergence and reduced excessive congestion, but this did not eliminate the counterintuitive results. Test 4 Results Scenario 4 involved tripling the service frequency on the “23– El Camino” bus route. The route currently runs at 30-min headways from about 5:30 a.m. to 9:00 p.m. on weekdays; Scenario 4 reduced the headways to 10 min for the entire ser- vice period. Highlights of this test include the following: • An increase in person trips by transit modes (+3.1% with SACSIM, +5.1% with C10B), and an increase in walk trips (+0.4% with SACSIM, +1.5% with C10B). These changes conformed to expectations on direction of change. (In addition to the summary provided in Table 4.3, Table 4.16 provides further details.) • As shown in Table 4.17, transit boardings on the test route nearly tripled in SACSIM, from about 2,300 to 8,700 daily boardings. This increase seems unreasonably large. The test effect estimated with C10B was about 66%, from 2,800 to 4,700 daily boardings, which is more reasonable. Boardings on surrounding transit routes decreased in both models. • Figure 4.5 shows the transit boardings on Route 23 by time of day. SACSIM produces summaries only for the aggre- gated peak and off-peak periods while the C10B model estimates volumes by bus run (the volumes are summa- rized by hour in Figure 4.5). The C10B model is also able to estimate maximum loads of passengers, also shown by hour. In the baseline scenario, for 2 h in the morning and 2 h in the afternoon, buses reached the maximum loads on the bus (40 seated, 20 standees, total 60). For the test sce- nario, buses in only 1 h did so. In summary, in this test, SACSIM produced an unexpect- edly large shift in ridership on Route 23 as a result of the Table 4.16. Person Trips by Mode—Test 4 Mode SACSIM C10B Integrated Model Baseline Scenario 4 Difference Percent Difference Baseline Scenario 4 Difference Percent Difference Transit auto access 11,920 11,268 -652 -5.5% 10,999 11,855 +856 +7.8% Transit walk access 85,977 89,709 +3,732 +4.3% 59,759 62,525 +2,766 +4.6% Transit (total) 97,897 100,977 +3,080 +3.1% 70,758 74,380 +3,622 +5.1% School bus 132,096 131,801 -295 -0.2% 91,627 93,813 +2,186 +2.4% Shared ride 3+ 1,580,119 1,578,941 -1,178 -0.1% 2,264,407 2,265,069 +662 +0.0% Shared ride 2 2,180,452 2,178,753 -1,699 -0.1% 2,044,692 2,037,101 -7,591 -0.4% Drive alone 3,568,112 3,566,803 -1,309 -0.0% 3,615,848 3,616,052 +204 +0.0% Bike 142,852 142,430 -422 -0.3% 97,686 99,224 +1,538 +1.6% Walk 524,481 526,456 +1,975 +0.4% 521,898 529,507 +7,609 +1.5% Total 8,226,009 8,226,161 +152 +0.0% 8,706,916 8,715,146 +8,230 +0.1%

90 Table 4.17. Transit Passenger Boardings Summary for Test 4 Test Metric Count Base Model Boardings Scenario 4 Boardings Percent Increase from Base SACSIM C10B SACSIM C10B SACSIM C10B Daily boardings on Test Route 23 2,550 2,289 2,804 8,684 4,666 +279.4% +66.4% Other nearby routes Route 22 (near Route 23) 520 727 494 589 453 -19.0% -8.3% Route 25 (near Route 23) 1,160 1,197 1,427 928 1,192 -22.5% -16.5% Route 82 (near Route 23) 2,130 2,400 2,870 2,457 2,752 +2.4% -4.1% Sum of nearby routes 3,810 4,324 4,791 3,974 4,397 -8.1% -8.2% All other RT buses IN C10B model 49,100 49,634 48,996 49,652 50,071 +0.0% +2.2% RT buses NOT IN C10B model 10,110 14,139 10,811 RT light rail (all lines) 48,300 42,278 20,864 42,581 21,397 +0.7% +2.6% All RT system boardings 117,870 116,356 69,860 119,439 71,468 +2.6% +2.3% Source: SACOG. SACSIM C10B Integrated Model Weekday boardings Change from base in weekday boardings Test 4-Test Route (Route 23) Only Test 4-Change from Base for Test Route & Combined Nearby Route Boardings Test 4-Change from Base for Test Route & Combined Nearby Route Boardings Test 4-Route 23 (Test Route) Only Figure 4.5. Boarding summaries for C10B integrated model—Test 4.

91 Table 4.18. Person Trips by Mode—Test 5 Mode SACSIM C10B Integrated Model Baseline Scenario 5 Difference Percent Difference Baseline Scenario 5 Difference Percent Difference Transit auto access 11,920 11,006 -914 -7.7% 10,999 11,257 +258 +2.3% Transit walk access 85,977 85,179 -798 -0.9% 59,759 60,744 +985 +1.6% Transit (total) 97,897 96,185 -1,712 -1.7% 70,758 72,001 +1,243 +1.8% School bus 132,096 132,067 -29 -0.0% 91,627 92,276 +649 +0.7% Shared ride 3+ 1,580,119 1,581,303 +1,184 +0.1% 2,264,407 2,264,047 -360 -0.0% Shared ride 2 2,180,452 2,179,814 -638 -0.0% 2,044,692 2,042,917 -1,775 -0.1% Drive alone 3,568,112 3,568,654 +542 +0.0% 3,615,848 3,618,680 +2,832 +0.1% Bike 142,852 143,798 +946 +0.7% 97,686 98,389 +703 +0.7% Walk 524,481 525,996 +1,515 +0.3% 521,898 525,769 +3,871 +0.7% Total 8,226,009 8,227,817 +1,808 +0.0% 8,706,916 8,714,079 +7,163 +0.1% decreased headway. It is not clear why this should occur in SACSIM since the mode choice model should not be overly sensitive to headway assumptions, and the same mode choice model is used in the C10B integrated model. Nor should the static transit assignment process be overly sensi- tive to headway assumptions. This result is particularly puzzling given that the C10B integrated model had a larger overall increase in transit demand (5% compared with 3% for SACSIM). Examining the reasons behind the unusual SACSIM results was beyond the scope of this project; for whatever reason, the C10B integrated model results were more reasonable. Both models showed about the same (reasonable) shifts in ridership from nearby routes. Test 5 Results Scenario 5 involved deleting the “23–El Camino” bus route. Highlights of Test 5 include the following: • A decrease in total regional transit person trips and walk person trips was expected. The SACSIM model test effects conformed to this expectation. (In addition to the sum- mary provided in Table 4.3, Table 4.18 provides further details.) The C10B model showed an increase of 1.8% in transit person trips, counter to expectation. • Both models showed an increase in passenger boardings for some surrounding routes (see Table 4.19), but the increases only partly offset the loss of passenger boardings on the test route; so overall transit boardings decreased in both models. Interestingly, both models showed similar increases on nearby routes. However, the C10B integrated model showed increases on other Sacramento Regional Transit District (RT) bus routes as well as the light rail lines. It is interesting that SACSIM showed substantial decreases in ridership on the RT routes not included in the C10B integrated model. These are generally low ridership routes that are mainly not in the vicinity of Route 23. More than half of the decrease in total regional transit boardings in SACSIM occurs on these routes, which is also an anomalous result that skews the direct comparisons of the results from the two models. • Figure 4.6 illustrates the change in boardings on the test route and those on the combined surrounding routes. The SACSIM model produces summaries only for the aggregated peak and off-peak periods while the C10B model estimates volumes by bus run (the volumes are summarized by hour in Figure 4.6). The C10B model is also able to estimate maximum loads of passengers, also shown by hour. In summary, in contrast to the results of Test 4, which used the same transit route as its basis, the results of Test 5 were more reasonable for SACSIM than for the C10B integrated model. The deletion of Route 23 should have resulted in a decrease in overall transit ridership, but in the C10B model, the opposite occurred. Both models did show increases in ridership on nearby routes, as expected. There were some unusual results in SACSIM away from the deleted route, making some direct com- parisons difficult.

92 Table 4.19. Transit Passenger Boardings Summary—Test 5 Test Metric Count Base Model Boardings Scenario 5 Boardings Percent Increase from Base SACSIM C10B SACSIM C10B SACSIM C10B Daily boardings on Test Route 23 2,550 2,289 2,804 0 0 -100.0% -100.0% Other nearby routes Route 22 (near Route 23) 520 727 494 727 606 — +22.7% Route 25 (near Route 23) 1,160 1,197 1,427 1,525 1,879 +27.4% +31.7% Route 82 (near Route 23) 2,130 2,400 2,870 2,391 2,871 -0.4% +0.0% Sum of nearby routes 3,810 4,324 4,791 4,643 5,356 +7.4% +11.8% All other RT buses IN C10B Model 49,100 49,634 48,996 49,590 50,231 -0.1% +2.5% RT buses NOT IN C10B Model 10,110 14,139 11,214 -20.7% RT light rail (all lines) 48,300 42,278 20,864 42,207 21,231 -0.2% +1.8% All RT system boardings 117,870 116,356 69,860 111,243 71,462 -4.4% +2.3% Source: SACOG. SACSIM C10B Integrated Model Weekday boardings Change from base in weekday boardings Test 5-Test Route (Route 23) Only Test 5-Change from Base for Test Route & Combined Nearby Route Boardings Test 5-Change from Base for Test Route & Combined Nearby Route Boardings Test 5-Route 23 (Test Route) Only Figure 4.6. Boarding summaries for SACSIM and C10B integrated model—Test 5.