HIGHWAY RANKING MODEL

HIGHWAYRANKING

MODEL

HIGHWAY RANKING MODEL

Bill Mann, P.E.Transportation Planning SectionNorthern Virginia DistrictJuly 25, 2006

– Based on Future Congestion and Its Relief Per Cost –(Congestion Relief is only one of many criteria

for selecting projects for construction)

2


• No standardized model exists to objectively rank highway improvements.

• VDOT has developed a method to do this ranking.

• The method piggybacks off MPO traffic forecasting models to rank highway projects.

• This tool is just one objective tool in the final ranking process that addresses gridlock.

PURPOSE

3


There is a funding shortfall of $15 billion to meet

the needs of the 2030 Transportation Plan for the

Northern Virginia District.

[TransAction 2030, January, 2006]

WHY RANK IMPROVEMENTS?

4

INPUTS (Nothing New is Required)• Area’s Future Highway Networks (Transit networks may be added

in later model versions.)• Area’s Future Highway Trips • Improvement Costs

OUTPUTS (Two Ways of Measuring Benefit/Cost Ratios)• Option 1 - Ranking of Improvements by Vehicle Hours of Delay

(VHD) Reductions (above LOS _) Per Cost• Option 2 - Ranking of Improvements by Vehicle Miles of Travel

(VMT) (above LOS _) Reductions Per CostNote: Thresholds for both options are user defined (LOS C, D, E, F, G)


5

CODING REQUIRED FOR RANKING(Took about two weeks for Northern Virginia)

• For each network link affected, code:-- ID Number-- New Lanes (if needed)-- New Functional Class Code (if needed)

• Provide table of construction costs for each improvement ID______________________________________________Note: Once all CLRP elements are coded, one could rank selected elements for any forecast year.


6

MODELING PROCESS(Assumes 70 Improvements as Identified in the CLRP)


Phase I (First Iteration; 70 Individual Assignments)

1. Calculate VHD for Base Network (with no improvements).

2. Add Improvement No. 1 to Base and calculate VHD Reductions with this Improvement.

3. Delete Improvement No. 1 and add Improvement No. 2 to Base; Calculate VHD Reductions with this Improvement.

4. Repeat Step 3 until all 70 improvements are tested.5. Rank all 70 improvements based on their VHD

Reductions/Cost.

7

MODELING PROCESS (Cont’d.)


Phase II (Second Iteration; 69 Individual Assignments)

6. Add the highest-ranked Improvement from Phase I to the Base Network to get a new Permanent Base.

7. Calculate VHD for the new Permanent Base.8. Repeat Steps 2 through 5.

Remaining 68 Phases

• Do 68 Individual Assignments, then 67, 66, etc.• Repeat Steps 6 through 8 until all 70 improvements

are ranked.• Summarize results.

8

MODEL OUTPUTS

Phase I Output shows for each improvement when compared to theBase the following (tested individually):

-- UserBenefits1) Vehicle hours of delay reductions (VHDR) or2) VMT at LOS F reductions (VMTR)

-- Construction Costs-- User Benefit/Cost Ratio expressed as VHDR/C or VMTR/C

________________________________________________________________________

Phase II Output shows for each improvement as it is incrementally added to Base (synergistic effect) what its contribution is relative to the entire plan:

-- User Benefits (VHDR or VMTR)-- Construction Costs-- User Benefit/Cost Ratio


9

ADVANTAGE OF MODELING IMPROVEMENTS

• Improvement A and Improvement B might both be excellent improvements by themselves based on gridlock relief per cost.

• If A is implemented first, the regional benefits of B might diminish and may not reduce regional congestion any further on the regional system until post-2010/2030.

• This process most likely spreads the ranking of improvements around the region, irrespective of geography or political boundaries.


10

• It is cost prohibitive without new software.

• One would have to code thousands of computer networks and run thousands of computer traffic assignments to rank 70 improvement projects in CLRP (e.g., 70 + 69 + 68 + 67 + 66 + 65 + etc.) = 2485 runs.

• Coding and running 3 networks and 3 traffic assignments per week in sequential order is all one could do with traditional modeling techniques.

• This approach would take years to accomplish manually.

WHY HASN’T THIS BEEN DONE BEFORE?


11

EXECUTION TIME WITH NEW VDOT MODEL

-- 6 days of computer time to rank 70 improvement projects in Northern Virginia CLRP using a PC with Intel 3.06 GHz processor.


12

THE NEXT SLIDES SHOW REAL SAMPLE RESULTS

• VDOT’s Central Office selected 52 Northern Virginia District arterial candidates for possible inclusion in the next Six-Year Plan.

• These were run through the Highway Ranking Model using 2005 trips from MWCOG.* (We could have used 2010 trips.)

*MWCOG=Metropolitan Washington Council of Governments


13


14


RANKING RESULTS

EXAMPLE: Project 24 reduces 347 vehicle hours of delay (daily) per construction cost ($ millions). Final Output showed that after adding the top 16 improvements, the bottom 36 did little to reduce VHD.

ID# IMPROVEMENT VHD Red

Cost ($M)

VHDRed/Cost($M)

Existing Future

ADT LOS ADT LOS

24 US 50, I-66 to Jermantown Rd., U8D, 0.81 mi. 3122 9 347

2 VA 27, Rt. 244 S. to Rt. 110, U6D, 1.60 mi. 3813 12 318

7 VA 120, Henderson Rd. to Rt. 50, U6D, 0.64 mi. 1944 7 278

20 VA 7, Patrick Henry Dr. to Rt. 244, U6D, 1.26 mi. 3337 14 238

21 VA 28, Bull Run to Rt. 29, U6D, 2.60 mi. 6042 29 208

28 VA 123, Old Courthouse to Rt. 7, U6D, 0.85 mi. 1028 9 114

29 VA 123, Rt. 7 to I-495, U8D, 0.90 mi. 988 9 110

17 VA 7, Dulles Toll Rd. to I-495, U8D, 2.36 mi. 2723 27 101

1 VA 27, Rt. 50 to Rt. 244 S., U6D, 0.72 mi. 905 9 101

33 VA 236, Pickett Rd. to Chambliss St., U6D, 7.26 mi. 6033 67 90

18 VA 7, I-495 to Birch St., U6D, 1.91 mi. 1437 17 85

25 US 50, Nutley St. to Graham Rd., U6D, 3.57 mi. 1585 34 47

36 VA 244, Sleepy Hol. Rd. to Carlin Spr. Rd., U6D, 2.23 mi. 763 21 36

41 VA 28, Fauquier Dr. to Vint Hill Rd., R4D, 5.14 mi. 506 17 30

16 VA 7, Dranesville Rd. to Dulles Toll Rd., U6D, 8.77 mi. 2806 93 30

19 VA 7, Rt. 50 to Patrick Henry Dr., U8D, 0.80 mi. 137 5 27

15


COST EFFECTIVENESS OF IMPROVEMENTS

NOTES: 1. This is total cost ($) divided by the vehicle hours of delay saved assuming an interest rate of 5% per year for annualizing capital investments. 2. Delay is measured if LOS D or worse. LOS is user defined.

ID# IMPROVEMENT Construction Costs

$ per VHDRed

24 US 50, I-66 to Jermantown Rd., U8D, 0.81 mi. 0.39

2 VA 27, Rt. 244 S. to Rt. 110, U6D, 1.60 mi. 0.43

7 VA 120, Henderson Rd. to Rt. 50, U6D, 0.64 mi. 0.49

20 VA 7, Patrick Henry Dr. to Rt. 244, U6D, 1.26 mi. 0.86

21 VA 28, Bull Run to Rt. 29, U6D, 2.60 mi. 1.02

28 VA 123, Old Courthouse to Rt. 7, U6D, 0.85 mi. 1.20

29 VA 123, Rt. 7 to I-495, U8D, 0.90 mi. 1.25

17 VA 7, Dulles Toll Rd. to I-495, U8D, 2.36 mi. 1.36

1 VA 27, Rt. 50 to Rt. 244 S., U6D, 0.72 mi. 1.36

33 VA 236, Pickett Rd. to Chambliss St., U6D, 7.26 mi. 1.52

18 VA 7, I-495 to Birch St., U6D, 1.91 mi. 1.61

25 US 50, Nutley St. to Graham Rd., U6D, 3.57 mi. 2.91

36 VA 244, Sleepy Hol. Rd. to Carlin Spr. Rd., U6D, 2.23 mi. 3.81

41 VA 28, Fauquier Dr. to Vint Hill Rd., R4D, 5.14 mi. 4.57

16 VA 7, Dranesville Rd. to Dulles Toll Rd., U6D, 8.77 mi. 4.57

19 VA 7, Rt. 50 to Patrick Henry Dr., U8D, 0.80 mi. 5.07

16


WHO IS AFFECTED BY IMPROVEMENTS?

• COLLECTOR ROADWAYS. Users living in the immediate area and those using the road and parallel roads.

• FREEWAYS. This can affect users from parallel roads as far away as 10 miles.

17


WHERE DOES ALL THE NEW TRAFFIC COME FROM?

• In modeling (and in the real world) a roadway widening could result in a LOS F after the widening with great benefits to the road and on parallel roads. • Someone may ask: Where did all this “new” traffic come from? It came from the three “S”s:

-- mode shift, -- time shift, and -- diversion shift.

18


MODEL OUTPUT CORRELATES WITH TTI DATAFOR 2003 NETWORK

• Texas Transportation Institute (TTI) estimates 69 hours of delay per year, per peak period traveler in 2003. [Source: Washington Post, May 10, 2005]

• VDOT/COG model estimates 108 hours of delay per year, per peak period traveler in Northern Virginia.________________________________________________

Explanation of Differences:• TTI measures delay on Freeways and Major Arterials only.

• VDOT measures delays on all roadways, including Minor Arterials and Collectors.

19


RECOMMENDED APPROACH FOR IMPROVED TRANSPORTATION PLANNING PROCESS

• Rank all plan elements (Comp, CLRP, SYIP) using future land uses, trips and networks by forecast year. • For those at bottom half of priority list, investigate to see why they are low in the ranking. What could be changed to improve their ranking? • Take results to decision makers for plan updates.

• This is an ongoing, continuous, iterative process.

20


Project: Widen Route Z from 4 lanes to 6 lanes; grade separate 3 intersections.Ranking Model Results: LowAnalysis: Delete interchanges.Results: Less VHDR, less cost, but benefit/cost ratio improves.Plan Update: Widen roadway now but postpone building interchanges until

after 2020/2030 when LOS drops to F/G based on forecast 2020/2030 land uses.

EXAMPLE OF IMPROVED TRANSPORTATION PLANNING PROCESS

SUBURBS

ACTIVITY CENTER

A

B

C

ROUTE Z

(Post Processing)

21

OUTPUT OPTIONS

• Rank all improvements or only selected improvements.

• Rank selected Functional Classification improvements (Freeways, Major Arterials, and/or Collectors). Example: Rank only primaries assuming:

-- All planned secondary and freeway improvements are built, or-- All planned secondary and freeway improvements are not built.


22

OUTPUT OPTIONS (Cont’d.)

• Select threshold for LOS delays.

• Rank improvements based on: -- VHD Reductions

-- VHD Reduction/Cost -- VMT at LOS F Reductions -- VMT at LOS F Reductions/Cost

• Select small or large area or specific jurisdictions for comparing VHD Reductions.


23


CONCLUSION

• Maximize area-wide traveler benefits at minimum cost

• Develop more cost effective transportation plans

With new and better computer models, we can now help decision makers:

24


Thank You!Questions?

Comments ?

25


CONTACT INFORMATION

Bill [email protected]

Mazen [email protected]


Date post:	08-Jan-2016
Category:	Documents
Upload:	jaeger
View:	30 times
Download:	0 times

HIGHWAY RANKING MODEL

Documents