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.
The L07 final report (MRIGlobal 2013b) provides equations for calculating the benefits using the standard deviation on page 88 of an earlier draft version of the report. Essentially, an agency would calculate the difference in the standard deviations estimated for the build and no- build scenarios and multiply the difference by the value of reliability and total vehicle-miles traveled (VMT) along the facility. FREEVAL-RL provides VMT for the volume handled in the model in its comprehensive outputs spreadsheet. A VMT for each segment is found in the âResults Summaryâ tab of the spreadsheet on the row labeled âVMTV Veh-miles (Volume).â These figures can be multiplied by the segment length in miles, summed to estimate the total VMT for the facility, and used in the above calculations. The results of the calculations would be a reliability value equivalent to the method used in the L07 Analysis Tool. Using the standard deviation is the method most commonly used in recent reliability research. Alternatively, the reliability benefits could be estimated using the difference between the 50th and 80th percentiles. The technical documentation for the C11 Reliability Analysis Tool (Cambridge Systematics et al. 2013b) provides formulas for using this inter-percentile difference on pages 16 and 17. These formulas could be modified to remove the recurrent delay portion of the calculation and may have been addressed in recent updates to the tool. The resulting formula basically multiplies the change in the inter-percentile difference by the value of reliability and the VMT. This method would estimate reliability benefits equivalent to those reported in the C11 Reliability Analysis Tool. This value would not be the same as (and likely slightly larger than) the one calculated using the L07 method. 8.2 Procedure for Estimating Reliability Benefits Once the study team settled on using the C11 tool for estimating travel time reliability benefits, the team estimated the impact of including travel time reliability in the following steps: ⢠Start with the C11 tool to estimate the cost of unreliability for each CSMP scenario. ⢠Calculate the travel time reliability benefit by estimating the reduction in the cost of unreliability from one scenario to the next. ⢠Add the reliability benefits to the CSMP benefit-cost analysis. ⢠Analyze the change in the overall benefit-cost ratio and the rank order of projects when reliability is included as a benefit. This procedure required the C11 tool to be recalibrated for the I-210 and I-5 facilities. This is because the CSMP benefit-cost analyses covered entire facilities, while the C11 calibration and scenario testing discussed in Chapter 5 covered only portions of the facilities. The recalibration process was aided by the earlier calibration efforts, but it required additional work to identify appropriate segmentation. For the I-210 facility, the study team focused on the four-lane and five-lane segments since these had proven adequate in earlier testing. Although the eastbound direction was already 176
calibrated, the study team had to download additional PeMS data to calibrate the westbound direction. The study team knew from the earlier calibration efforts that obtaining the necessary hourly VMT and TTI data would be very time-consuming. PeMS allows only two to three weeks of each year to be downloaded at a time. The study team decided to examine the data available for the eastbound direction and select the most representative two-week period (i.e., the first two weeks of November). The study team then downloaded the necessary hourly data for the two westbound segments and calibrated the C11 tool for these segments. Although this shortcut did not save much time for the I-210 facility, it was absolutely necessary for re-calibrating the I-5 facility. As Figure 8.2 shows, the C11 tool calibrated fairly well to real-world conditions (as measured by PeMS) for the westbound I-210 segments. The calibration required slightly higher peak capacities for the westbound direction than the eastbound direction (roughly 1,700 to 1,900 vphpl westbound compared to 1,460 to 1,490 vphpl eastbound). This result makes sense because the westbound direction experiences less congestion (and has a lower throughput or productivity loss) than the eastbound direction. (a) Five-lane segment (b) Four-lane segment Figure 8.2. Calibration of westbound I-210 segments. For the I-5 facility, the study team needed to recalibrate the entire facility, since the previous calibration had focused on a small, 6.5-mile segment. The study team decided to divide the facility into segments according to the bottleneck areas identified in the CSMP. These are shown in Figure 3.5. This resulted in seven segments in the northbound direction and nine segments in the southbound direction. This segmentation is consistent with the microsimulation modeling and aggregation of results used for the benefit-cost analysis found in the CSMP. The study team used the percent demand per hour distribution already calculated for the 6.5-mile segment for the northbound segments. The distribution for the southbound segments was calculated using a representative two-week period. Once the calibration was completed, the study team tested each of the CSMP scenarios in the models. For the I-210 facility, this involved running the model for only the two new 177
calibrated westbound segments. For the I-210 facility, the testing required running the C11 tool 16 times for each scenario. The CSMP benefit-cost analyses covered a 20-year life cycle. The C11 tool does not calculate life-cycle benefits within the tool. Rather, individual year benefits are calculated for a current and future year. Ideally, reliability benefits should be calculated for each year, summed, and discounted. However, the study team decided to estimate benefits for the current year and 20 years later in the C11 tool using an ADT growth factor estimated from the CSMP microsimulation runs and travel demand models for each facility. The study team developed a spreadsheet to interpolate the current and future year delay and reliability benefits over a 20-year life cycle. Figure 8.3 shows a snapshot from this spreadsheet for the I-5 facility. For both facilities, the benefits were interpolated and discounted using the same 4-percent discount rate used in calculating benefits for the I-5 and I-210 CSMPs. 178
Year Base vs. NB S1-S2 Base vs. Total S1-S2 Base vs. NB S3-S4 Base vs. Total S3-S4 Base vs. NB Incident Mgmt Base vs. Total Incident Mgmt Base vs. NB S7 Base vs. Total S7 1 $1,298,209 $2,234,467 $2,134,750 $2,627,168 $4,605,747 $6,002,362 $24,520,895 $35,046,001 21 $6,144,234 $10,578,351 $14,508,886 $22,131,369 $30,437,340 $45,426,526 $24,520,895 $35,046,001 1 $1,298,209 $2,234,467 $2,134,750 $2,627,168 $4,605,747 $6,002,362 $24,520,895 $35,046,001 2 $1,481,260 $2,549,674 $2,647,554 $3,463,825 $5,670,507 $7,666,895 $23,577,784 $33,698,078 3 $1,648,310 $2,837,329 $3,117,754 $4,232,238 $6,646,548 $9,194,507 $22,670,946 $32,401,998 4 $1,800,318 $3,099,085 $3,547,869 $4,936,418 $7,539,121 $10,593,269 $21,798,986 $31,155,767 5 $1,938,195 $3,336,509 $3,940,286 $5,580,169 $8,353,203 $11,870,832 $20,960,564 $29,957,469 6 $2,062,803 $3,551,085 $4,297,268 $6,167,099 $9,093,510 $13,034,451 $20,154,388 $28,805,258 7 $2,174,959 $3,744,219 $4,620,961 $6,700,625 $9,764,514 $14,091,001 $19,379,220 $27,697,364 8 $2,275,435 $3,917,244 $4,913,399 $7,183,988 $10,370,450 $15,046,995 $18,633,865 $26,632,081 9 $2,364,966 $4,071,421 $5,176,505 $7,620,257 $10,915,331 $15,908,607 $17,917,178 $25,607,770 10 $2,444,243 $4,207,943 $5,412,104 $8,012,340 $11,402,957 $16,681,683 $17,228,056 $24,622,856 11 $2,513,924 $4,327,940 $5,621,922 $8,362,990 $11,836,927 $17,371,757 $16,565,438 $23,675,823 12 $2,574,629 $4,432,482 $5,807,594 $8,674,815 $12,220,646 $17,984,072 $15,928,306 $22,765,214 13 $2,626,945 $4,522,580 $5,970,668 $8,950,281 $12,557,338 $18,523,587 $15,315,679 $21,889,629 14 $2,671,429 $4,599,191 $6,112,606 $9,191,726 $12,850,053 $18,994,998 $14,726,614 $21,047,720 15 $2,708,604 $4,663,218 $6,234,793 $9,401,357 $13,101,675 $19,402,745 $14,160,206 $20,238,192 16 $2,738,969 $4,715,517 $6,338,540 $9,581,266 $13,314,933 $19,751,028 $13,615,583 $19,459,800 17 $2,762,990 $4,756,894 $6,425,082 $9,733,428 $13,492,405 $20,043,817 $13,091,906 $18,711,347 18 $2,781,112 $4,788,113 $6,495,591 $9,859,712 $13,636,529 $20,284,866 $12,588,372 $17,991,679 19 $2,793,753 $4,809,894 $6,551,172 $9,961,884 $13,749,607 $20,477,723 $12,104,203 $17,299,692 20 $2,801,308 $4,822,916 $6,592,868 $10,041,611 $13,833,814 $20,625,737 $11,638,657 $16,634,319 TOTAL (Presen t Value) $46,462,362 $79,987,721 $101,959,286 $150,283,197 $214,955,817 $313,550,934 $346,576,845 $495,338,057 Figure 8.3. Example of interpolating benefits on the I-5 facility. (Discount Factor = 4.00%; Average Vehicle Occupancy (AVO) = 1.24) In previous unpublished research, study team members tested the difference in interpolating model inputs compared to benefits for benefit-cost analysis and found that interpolating the benefits does not accurately estimate the results from year-by-year estimation. However, the study team chose to interpolate the benefits given the time and resources available for the L38 project. The C11 tool should be modified to estimate life-cycle benefits, so agencies do not need to interpolate benefits or engage in time-consuming year-by-year estimation of benefits. 179
