High Risk Rural Roads (HRRR) Program in Virginia

High Risk Rural Roads (HRRR)
Program in Virginia

In-Kyu Lim Ph.D.

High Risk Rural Roads Program Manager
Traffic Engineering Division (TED)

High Risk Rural Roads

ƒ For reducing traffic fatality and incapacitating
injuries on rural roadways

ƒ Eligible Roads per SAFETEA-LU
“..on roadway functionally classified as”

– Rural Major Collector
– Rural Minor Collector
– Rural Local

High Risk Rural Roads (cont)

Statutory Requirements

A.on which the accident rate for fatalities and
incapacitating injuries exceeds the statewide
average for those functional classes of roadway

OR

B. that will likely have increases in traffic volume that are
likely to create an accident rate for fatalities and
incapacitating injuries that exceeds the statewide for
those functional classes for roadway

Purpose

ƒ Develop a method to identify Virginia roadway for
implementing HRRR program

– Must be a systematic approach in Oracle dB and not a
manual approach

– Must satisfy FHWA statutory requirements

ƒ Improve safety adopting High Risk Rural Roads
Program in Virginia

VVAARoadRwayos adways

1 Rural Interstate VA Road Length = 72,676 miles
2 Rural Principal Arterial
3 Rural Minor Arterial Urban (30.8%)
4 Rural Major Collector Rural (69.2%)
5 Rural Minor Collector
6 Rural Local
A Urban Interstate
B Urban Freeway/Expressway
E Urban Principal Arterial
H Urban Minor Arterial
I Urban Collector
J Urban Local

VA Roadways (cont)

1 Rural Interstate VA DVMT = 225VA DVMT = 225 million miles million miles
2 Rural Principal Arterial
3 Rural Minor Arterial Urban (62.4%)
4 Rural Major Collector Rural (37.6%)
5 Rural Minor Collector
6 Rural Local
A Urban Interstate
B Urban Freeway/Expressway
E Urban Principal Arterial
H Urban Minor Arterial
I Urban Collector
J Urban Local

VA Roadway Crash

ƒ Average Crash Rate

Crash Rate 150 113.5

115.6 Urban

120 Based on 2003 – 2007 Crash

90

60

30

0
Rural

VA Roadway Crash (cont)

ƒ Average Severity Rate

Severity Rate 40 33.4 15.0

30 Urban
20
10 Based on 2003 – 2007 Crash

0
Rural

VA Rural Roadways
Severity Rate

80 6677..22

70 5544..66 5588..77
60
Severity Rate 36.1 Avg. Rural
50 33.4
40 26.2
Avg. Urban
30 15.0
20 13.5
10 2 Rural 3 Rural Minor 4 Rural Major 5 Rural Minor 6 Rural Local
0 Principal
Arterial Arterial Collector Collector
1 Rural
Interstate

Functional Class

HRRR Program Funding
and Crash Distribution

ƒ VDOT has 2 millions/year Functional Class Road Length
in HRRR Program Rural Major Collector (Miles)

ƒ Need to focus on hot 9,293
spots to maximize benefit
from the limited funds Rural Minor Collector 2,440

Rural Local 33,062

TOTAL 44,796

Based on 2007 Inventory

Roadways Fatal or Incapacitating
Injuried People

Segement 19,665 (76.1%)

Intersection 6,161 (23.9%)

HRRR Intersection
Analysis Method

Eligible HRRR

Intersection Criteria
2
Functional Class 44

1. Rural Interstate = 42 (Eligible)

2. Rural Principle Arterial 2
3. Rural Minor Arterial 1
4. Rural Major Collector

5. Rural Minor Collector 4 5 = 41 (Eligible)
6. Rural Local

A. Urban Interstate 6
B. Urban Freeway/Expressway

E. Urban Principle Arterial 6

H. Urban Minor Arterial 4 4 = 4J (Not Eligible)
I. Urban Collector

J. Urban Local J

HRRR Intersection

Category

HRRR ID Description ƒ Crash Data Analysis

41 Rural Major Collector - Rural Interstate – 2003 – 2007 Crash
Data
42 Rural Major Collector - Rural Principle Arterial – Intersection Related
Crash Only (Offset <=
43 Rural Major Collector - Rural Minor Arterial 250 ft)

44 Rural Major Collector - Rural Major Collector

45 Rural Major Collector - Rural Minor Collector

46 Rural Major Collector - Rural Local

51 Rural Minor Collector - Rural Interstate

52 Rural Minor Collector - Rural Principle Arterial

53 Rural Minor Collector - Rural Minor Arterial

55 Rural Minor Collector - Rural Minor Collector

56 Rural Minor Collector - Rural Local

61 Rural Local - Rural Interstate

62 Rural Local - Rural Principle Arterial

63 Rural Local - Rural Minor Arterial

66 Rural Local - Rural Local

Network Screening –
Step 1

Intersections
Fatal & incapacitating

Injury Crash = 0
2003 - 2007 (9,360)

Intersections
Fatal & incapacitating

Injury Crash >= 1
2003 - 2007 (5,156)

Network Screening –
Step 2

Applied Threshold Prioritized by

1. Intersections have no Safety 1. Crash Frequency (CF)
Improvement during last 24 2. Crash Rate (CR)
and next 12 months 3. Severity (people)
2. Fatal & Incapacitating Crash
Frequency >= 5 AND Frequency (SF)
3. Fatal & Incapacitated 4. Severity (people) Rate
People >= 5 OR
4. % of Fatal & Incapacitating (SR)
Crash >= 40%
Final Rank by

Identify 71 Severe Crash
Intersections under HRRR

12
10 9.59

Avg. Frequency 8 District # of Intersections %
1 Bristol 6 8.5
6.21 2 Salem 12 16.9
6 3 Lynchburg 14 19.7
4 Richmond 4 5.6
4 2.27 5 Hampton Roads 13 18.3
2 1.57 6 Fredericksburg 12 16.9
7 Culpeper 6 8.5
0 8 Staunton 4 5.6
Total (5,156) Selected 71 Total (5,156) Selected 71 9 NOVA 0 0.0
71
Fatal & Incapacitating Crash Fatal & Incapacitated People TOTAL 100.0

RSA for Improvement
Safety

Detail crash data Conduct analysis Countermeasures
analysis and propose • Short-term
• Intermediate
Perform field countermeasures • Long-term
reviews
Prepare written
report with final

decision

Implementing the
decsion

Intersection HRRR
Program Status

Roadway Safety Assessment (RSA)
were conducted imperatively safety
needed (Most Critical) 71 eligible
HRRRP Intersections in statewide of
Virginia

HRRR Segment
Analysis Method

Network Screening

Developing Eligible No
Procedure HRRR Roadway?

13 Steps Join Crash Data Do Not
TMS Data Consider
2 YES
Join
3
Severe No
Crash Rate

>
SW Average?

YES

Prioritization Screening

Network Screening

ƒ VDOT has interrelated Oracle database system
called Highway Transportation Records
Information System (HTRIS) which is stored as
historical crash, road inventory and traffic
information.

ƒ Network screening was conducted utilizing 3

1 subsystem as crash (ACC), road inventory (RDI)

and traffic monitoring (TMS) in HTRIS system.

ƒ Structured Query Language (SQL) was used
for data manipulation

2 ACC

SQL
RDI TMS

3

Step 1

Network Screening ƒVirginia roadways consist of approximately
72,600 miles and are classified into 12 different
1 Eligible No functional classes, same as FHWA in VDOT
HRRR Roadway? road inventory sub-system (RDI).

Join Crash Data ƒAbout 60% (43,123 miles) of roadways are
TMS Data functionally classified as eligible in HRRR
YES program Statewide.
Join
ƒSummary of eligible HRRR roadways in
Statewide Virginia

Severe FUNCTIONAL CLASS # OF LENGTH
SEGMENT (mile)
2 Crash Rate No 4 Rural Major Collector
> 5 Rural Minor Collector 19,921 9,293.5
6 Rural Local 3,908 2,440.2
SW Average? 31,389.2
TOTAL 63,681
YES 43,122.9
87,510

3 Prioritization Screening

Joining Crash Data

Network Screening ƒ Collect only crashes that occurred on roadway
segments from crash (ACC) sub-system using
1 Eligible No offset information
HRRR Roadway?
Intersection crash (offset <= 250ft)
Join Crash Data Segment crash (offset > 250ft)
TMS Data
YES ƒ Join only the segment crashes on the roadways
Join
identified from Step 1 .

ƒ Summary of joining Crash Data (2005 – 2007)
on HRRR Roadways

FUNCTIONAL CLASS TOTAL SEVERE
CRASH CRASH

Severe

2 Crash Rate No 4 Rural Major Collector 18,782 2,917
> 5 Rural Minor Collector 2,976 492

SW Average?

YES 6 Rural Local 14,422 2,220

Severe Crash : Fatal and Injury Type “A” Crash Only

3 Prioritization Screening

Joining TMS Data

Network Screening ƒ Collect Average Annual Daily Traffic (AADT) on
roadway from traffic monitoring sub-system
(TMS)

ƒ Join the AADT on the roadways identified from

Step 1 .

1 Eligible No ƒ Summary of joining TMS data on HRRR
HRRR Roadway? Roadways

Join Crash Data Average Annual Daily Traffic
TMS Data
YES FUNCTIONAL # OF (AADT)
Join CLASS
SEGMENT MAX. AVG.
MIN.

4 Rural Major Col 19,921 102 25,944 2,339.3

Severe

2 Crash Rate No 5 Rural Minor Col 3,908 102 9,553 833.3
>

SW Average? 6 Rural Local 63,681 101 21,719 452.9

YES

3 Prioritization Screening

Severe Crash Rate Measurement

Severe Crash Rate (SCR): 3 year

Network Screening

1 Eligible No SCR = # Severe Crash ×100 ,000 ,000
HRRR Roadway? Length × AADT × 365

Statewide Average Severe Crash Rate
(SWASCR) was calculated for each HRRR
functional class as below

SWASCR i = ∑ # of Severe Crash ×100 ,000 ,000
∑ Length i × ∑ AADT i × 365
Join Crash Data
TMS Data i = functional class
YES
Join

Severe FUNCTIONAL CLASS Statewide
Average Severe
2 Crash Rate No
> Crash Rate

SW Average? 4 Rural Major Collector 49.17
5 Rural Minor Collector 69.70
YES 6 Rural Local 64.68

3 Prioritization Screening TOTAL 61.37

Severe Crash : Fatal and Injury Type “A” Crash Only

Step 2

Network Screening ƒ Based on Step 1, Identify and collect only the
segments which have severe crash rate greater
than statewide average in same functional class.

1 Eligible No ƒ Approximately, 13% (5,552 miles) of the
HRRR Roadway? segments had severe crash rate greater than
average statewide severe crash rate in each
functional class.

Join Crash Data Summary Results.
TMS Data
YES 1 2
Join

Severe

2 Crash Rate No
>

SW Average?

YES

3 Prioritization Screening

FHWA Statutory for HRRR
Program

ƒ VDOT identified 5,552 miles of roadways segments which

satisfied FHWA required statutory to implement HRRR program
as

1. “..on roadway functionally classified as” √ Satisfied
• Rural Major Collector
• Rural Minor Collector
• Rural Local

2. “..on which the accident rate for fatalities and incapacitating

injuries exceeds the statewide average for those functional

classes of roadway” √ Satisfied

Historical Crash Review

ƒ 14,260 crashes were occurred on 5,552 miles between
2005 – 2007

ƒ If those crashes are uniformly distributed, crash is
occurred ..

5552 × 5280 = 2056 ft / crash
14260

Very dispersed!!

Spot Crash Distribution

Crash Frequency 11 1
10 1
4000 6000 8000 10622
83 # of Spot 10000 12000
75
65
5 20
4 62
3 226
2 1251
1

0 2000

Screening Method

ƒ High Crash Segment Determination

S ft

D1
D2

D3

Screening Method (cont)

ƒ High Crash Segment Consolidation

AL B

If Route (Prefix, Number, Suffix) is same
L <= 0.5 mile Î Consolidate
L > 0.5 mile Î Not Consolidate

Data Analysis

ƒ 5,552 miles HRRR segments were screened

ƒ Modified sliding window technique was utilized

ƒ Threshold

– Segment must have at least one spot had more than 3
crashes

– Total Crash >= 5
– Segment Crash Density >= 20.0

Analysis Results

ƒ Most hazardous 93 HRRR segments were
identified in statewide

– Total 25.06 miles

• Min. : 0.23 miles Max. : 1.2 miles

1 Bristol # of Segment Length (Mile) Total Crash RD Crash
2 Salem 12 3.01 96 66
3 Lynchburg 17 4.76 133 92
4 Richmond 3 0.68 16 10
5 Hampton Roads 10 2.52 76 41
6 Fredricksburg 6 1.76 46 28
7 Culpeper 11 4.36 118 74
8 Staunton 12 2.83 71 54
9 Northern Virginia 14 3.28 93 65
8 1.86 72 30
TOTAL
93 25.06 721 460

Example Site 1

ƒ Route: SR00218
ƒ District:

Fredericksburg
ƒ Length: 1.2 mile
ƒ Total Crash: 30

Density: 25.0/mile
ƒ RD Crash: 28

Density: 23.3/mile

Example Site 2

ƒ Route: 9200609
ƒ District: Bristol
ƒ Length: 0.23 mile
ƒ Total Crash: 12

Density: 52.2/mile
ƒ RD Crash: 12

Density: 52.2/mile

Identified Results

ƒ Most Hazardous 785 miles

– Avg. Crash Density: 7.0 /mile
– Avg. RD Crash Density: 4.8 /mile

ƒ Top 100 miles

– Avg. Crash Density: 11.9 /mile
– Avg. RD Crash Density: 7.5 /mile

ƒ Selected 25.06 miles using Modified Sliding Window
Method

– Avg. Crash Density: 27.7 /mile
– Avg. RD Crash Density: 17.7 /mile