U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000


Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

Report
This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-HRT-07-045
Date: June 2007

Applications of ISAT, Interchange Safety Analysis Tool (ISAT): User Manual

PDF Version (777 KB)

PDF files can be viewed with the Acrobat® Reader®

Default Data

This section presents the default data that are provided with ISAT. These data include:

  • Safety performance functions (SPFs).
  • Calibration coefficients.
  • Crash distributions by severity and type.

In the case of default SPFs, it is anticipated that most ISAT users will not modify the default coefficients of the SPFs. In other words, it is expected that in most instances ISAT will be run utilizing the default SPFs. In this respect, users will very seldom change the values associated with the default SPFs. Concerning the default calibration coefficients and crash distributions by severity and type, the user should modify these default values to more accurately reflect the safety experience of their local/State agency prior to using ISAT to perform actual safety assessments. Then, it is recommended that the user modify these default values on an annual basis. The user can modify the default data by clicking on the hyperlinks located on the Input-General worksheet. Five hyperlinks are provided to access the default SPFs, one hyperlink is provided to access the default calibration coefficients, and four hyperlinks are provided to access the default crash distributions. The following sections present more detailed information on each type of default data provided within ISAT and further explain the need to modify these default data.

Safety Performance Functions (SPFs)

For many years, statistical techniques have been applied to predict the crash experience along roadway segments and at intersections. These statistical models are developed using crash and inventory databases, selecting an appropriate functional form for the model, and using regression techniques to estimate the values of the coefficients and parameters in the model. Historically, these models were developed with multiple regression techniques. More recently, Poisson and negative binomial (NB) analyses have been used because theoretically they are better suited to crash data with low-frequency observations. These statistical models are often referred to as safety performance functions (SPFs).

ISAT makes use of SPFs for predicting and/or estimating crash frequencies for individual components of an interchange and the surrounding roadway network. Safety estimates are calculated for the individual components, and these safety estimates are summed to obtain safety performance estimates for the interchange as a whole. Within ISAT default SPFs are provided for the following primary interchange elements that can be included in an analysis area:

  • Mainline freeway segments.
  • Interchange ramps (and acceleration lanes).
  • Crossroad ramp terminals and intersections.
  • Crossroad roadway segments.

ISAT was developed using existing knowledge of SPFs or knowledge acquired from other ongoing safety research projects. During the development of ISAT, a review of the technical literature on safety assessment of freeway interchanges was conducted. This review focused on quantitative information that was potentially applicable for use in ISAT. Specifically, interchange-safety-related literature was reviewed to identify quantitative SPFs which could potentially be incorporated into ISAT. The review included both engineering and statistical considerations. The SPFs selected for incorporation into ISAT were obtained from the following primary sources:

Development of Safety Performance Functions for SafetyAnalyst Interim Tools4

SafetyAnalyst is a set of computerized analytical tools being developed for FHWA to aid State and local highway agencies in highway safety management. The main purpose of SafetyAnalyst is to improve a highway agency's systemwide programming of site-specific safety improvements. SafetyAnalyst incorporates state-of-the-art safety management approaches for guiding the decision-making process to identify safety improvement needs and has a strong basis in cost-effectiveness analysis. SafetyAnalyst will help highway agencies get the greatest possible safety benefit from each dollar spent in the name of safety.

Statistical Models for Interchange Ramps and Speed-Change Lanes3

The objective of this research was to develop statistical models for defining the relationship between traffic crashes and highway geometric design elements and traffic volumes for interchange ramps and speed-change lanes. The models were developed using Highway Safety Information System (HSIS) data for interchange ramps and speed-change lanes in the State of Washington. Additional geometric design features were obtained from a review of interchange diagrams, aerial photographs, and other existing highway agency files.

It is important to recognize that the accuracy of ISAT is only as good as the safety knowledge on which it is based. The default SPFs provided within ISAT were selected for incorporation in the tool based primarily for the following reasons: (1) statistical validity of the models, (2) criticality of the components based upon engineering judgment, and (3) methodological consistency.

Due to the comprehensive process for selecting SPFs for incorporation within ISAT, it is anticipated that most users will run the program utilizing the default SPFs; however, the program does not prohibit users from modifying the coefficients of the default SPFs. If users have models that better reflect local safety experience, users can replace the default coefficients with new values. This should only be done if the models were developed with the same functional form as the default SPFs. The following sections provide information on the sources of the default SPFs, their functional form, and select coefficients and parameters of the models.

Mainline Freeway Segment SPFs

The default SPFs for mainline freeway segments provided within ISAT are located on the SPFs Mainline Freeways worksheet and are accessible only through the Go To SPFs for Mainline Freeways hyperlink on the Input-General worksheet. The default SPFs for mainline freeway segments were developed for use within SafetyAnalyst. These SPFs predict the number of crashes that may occur on mainline freeway segments. SPFs are provided for two unique types of mainline freeway segments:

  • Mainline freeway segments within an interchange area.
  • Mainline freeway segments outside an interchange area.

As described in the Mainline Freeway Segment Inputs section, the SPFs for mainline freeway segments within interchange areas attempt to account for the increased level of weaving, lane changing, and acceleration/deceleration that takes place immediately upstream, downstream, and between interchange ramps. The SPFs for mainline freeway segments outside an interchange area model the safety experience of basic mainline freeway segments having homogenous characteristics. In these homogeneous segments, the number of through lanes remains the same, the traffic volume remains the same throughout the segment because there are no ramps associated with the segment, and the primary movements include lane changes between through lanes.

The limits of mainline freeway segments within interchange areas are defined in general terms to extend approximately 0.48 km (0.3 mi) upstream from the gore of the first ramp of a particular interchange to approximately 0.48 km (0.3 mi) downstream from the gore of the last ramp of the given interchange. All mainline freeway segments beyond these boundaries are by definition mainline freeway segments outside an interchange area.

The SPFs for mainline freeway segments within interchange areas include crashes that occur on the speed-change lanes (i.e., deceleration and acceleration lanes) adjacent to the mainline freeway segment through lanes, but the SPFs do not predict crashes that occur on the ramp proper (i.e., downstream from the gore point for off-ramps and upstream from the gore point for on-ramps). Because the SPFs for mainline freeway segments within interchange areas include crashes that occur on speed-change lanes, certain methodological assumptions are made within ISAT so that crashes are not double counted on mainline freeway segments and acceleration lanes.

Table 15 presents the coefficients and parameters of the default SPFs for mainline freeway segments provided within ISAT. Table 15 also shows the functional form of the models. In general, all models included in ISAT were developed using NB regression. Table 16 defines the coefficients and parameters of the defaults SPFs. Twenty SPFs for mainline freeway segments are provided within ISAT and are a function of the following:

  • Area type.
  • Interchange area (i.e., within or outside interchange area).
  • Number of lanes.
  • Severity level.

Portions of the SPFs Mainline Freeways worksheet are protected. In other words, if the user tries to modify the value of a protected cell, a message will appear indicating the cell is protected, and the user is restricted from modifying the default value in the given cell. On the SPFs Mainline Freeways worksheet, the cells with column headings for SPF No., Area type, Interchange area, Number of through lanes, and Severity level are protected. The user may only modify the default values for the Log intercept (a), Coeff of log ADT (b), dispersion parameter, and Max ADT.

Ramp SPFs

Two sets of default SPFs associated with ramps are provided within ISAT. The default SPFs used to predict the safety performance for the ramp proper are located on the SPFs Ramp worksheet and are accessible only through the Go To SPFs for Ramps hyperlink on the Input-General worksheet. The default SPFs used to predict the safety performance for acceleration lanes are located on the SPFs Accel Lanes worksheet and are accessible only through the Go To SPFs for Accel Lanes hyperlink provided on the Input-General worksheet. The default SPFs for the ramp proper were developed for use within SafetyAnalyst. The default SPFs for acceleration lanes were developed by Bauer and Harwood.3

Table 17 presents the coefficients and parameters of the default SPFs for ramps (i.e., ramp proper) provided within ISAT. Table 17 also shows the functional form of the models. Table 18 defines the coefficients and parameters of the defaults SPFs. Twenty-eight SPFs for ramps (i.e., ramp proper) are provided within ISAT and are a function of the following:

  • Area type.
  • Ramp type.
  • Ramp configuration.
  • Severity level.

The ramp SPFs within SafetyAnalyst were developed by combining the data for rural and urban areas; thus, separate ramp SPFs are not necessarily available for rural and urban areas. For consistency purposes the SPFs are presented by area type, but the coefficients and parameters are the same for both area types.

Portions of the SPFs Ramps worksheet are protected. The user may only modify the default values for the Log intercept (a), Coeff of ADTramp (b), Coeff of ramp length (e), dispersion parameter, and Max ADT.

Table 15. SPF Coefficients and parameters for mainline freeway segments.
SPF No. Area Type Interchange Area Number of Through Lanes (Directional) Severity Level Log Intercept (a) Coeff of Log ADT(b) Dispersion Parameter Max ADT
numeric (U, R) (Y, N) (2, 3, 4) (TOT, FI) numeric numeric numeric (veh/day)
SafetyAnalyst Roadway Segment SPF
N = ea x ADTb x SL
DMF1 DMF2 DMF3 DMF4 DMF5 DMF6 DMF7 DMF8 DMF9
1 R Y 2 TOT -7.28 0.92 0.45 60,621
2 R Y 3 TOT -10.05 1.14 0.42 197,798
3 U Y 2 TOT -11.23 1.30 0.81 241,255
4 U Y 3 TOT -11.25 1.28 0.60 255,154
5 U Y 4 TOT -26.76 2.58 0.52 233,323
6 R Y 2 FI -8.68 0.94 0.58 60,621
7 R Y 3 FI -12.07 1.22 0.39 197,798
8 U Y 2 FI -12.89 1.38 0.79 241,255
9 U Y 3 FI -13.62 1.42 0.55 255,154
10 U Y 4 FI -25.63 2.42 0.53 233,323
11 R N 2 TOT -6.46 0.79 0.17 60,621
12 R N 3 TOT -9.67 1.07 0.24 190,403
13 U N 2 TOT -7.85 1.00 0.99 151,038
14 U N 3 TOT -5.96 0.78 0.48 241,255
15 U N 4 TOT -16.24 1.67 0.45 223,088
16 R N 2 FI -8.86 0.9 0.10 60,621
17 R N 3 FI -11.67 1.17 0.21 190,403
18 U N 2 FI -8.82 1.02 1.15 151,038
19 U N 3 FI -7.60 0.85 0.54 241,255
20 U N 4 FI -19.16 1.85 0.52 223,088
Table 16. Definitions of coefficients and parameters of SPFs for mainline freeway segments.
VariableName Variable No. Format Definition
SPF No. DMF1 Numeric Each SPF is identified by a sequential integer, starting with 1; this variable is not used in calculations but is useful for organization.
Area type DMF2 Character(U, R) General character of land use surrounding the interchange, preferably based on FHWA urban area boundaries:
U = Urban
R = Rural
Interchange area DMF3 Character(Y, N) A code identifying whether the segment is located within the interchange area:
Y = yes, segment is located within the interchange area
N = no, segment is not located within the interchange area
Number of through lanes (directional) DMF4 Numeric(2, 3, 4) This field includes all lanes on the segment in a given direction that are used by through traffic; it does not include auxiliary lanes.
Severity level DMF5 Character(TOT, FI) A code identifying the crash severity level:
TOT = total crashes
FI = fatal and injury crashes
Log intercept (a) DMF6 Numeric Intercept of SPF.
Coefficient of log ADT (b) DMF7 Numeric Coefficient of log ADT parameter (this ADT is bi-directional).
Dispersion parameter DMF8 Numeric Dispersion parameter associated with negative binomial regression.
Max ADT DMF9 Numeric(veh/day) The maximum traffic volume for which the coefficients of the SPF were calibrated/calculated.
Table 17. SPF Coefficients and parameters for ramps.
SPF No. Area Type Ramp Type Ramp Configuration Severity Level Log Intercept (a) Coeff of ADTRamp (b) Coeff of Ramp Length (e) Dispersion Parameter Max ADT
numeric (R, U) (ON, OFF, FWY) (D, PL, FFL, DIR) (TOT, FI) numeric numeric numeric numeric (veh/day)
SafetyAnalyst Ramp Proper SPF
N = ea x ADTRampb x RLe
DR1 DR2 DR3 DR4 DR5 DR6 DR7 DR8 DR9 DR10
1 R OFF D TOT -3.17 0.45 1.0 1.49 22,566
2 R ON D TOT -8.28 1.03 1.0 2.57 24,966
3 R OFF PL TOT -4.50 0.73 1.0 1.17 22,538
4 R ON PL TOT -2.11 0.43 1.0 1.77 20,403
5 R OFF FFL TOT -4.50 0.73 1.0 1.17 22,538
6 R ON FFL TOT -2.11 0.43 1.0 1.77 20,403
7 R FWY DIR TOT -1.80 0.45 1.0 1.67 37,474
8 U OFF D TOT -3.17 0.45 1.0 1.49 22,566
9 U ON D TOT -8.28 1.03 1.0 2.57 24,966
10 U OFF PL TOT -4.50 0.73 1.0 1.17 22,538
11 U ON PL TOT -2.11 0.43 1.0 1.77 20,403
12 U OFF FFL TOT -4.50 0.73 1.0 1.17 22,538
13 U ON FFL TOT -2.11 0.43 1.0 1.77 20,403
14 U FWY DIR TOT -1.80 0.45 1.0 1.67 37,474
15 R OFF D FI -6.88 0.78 1.0 2.21 22,566
16 R ON D FI -14.40 1.61 1.0 3.44 24,966
17 R OFF PL FI -3.63 0.53 1.0 1.71 22,538
18 R ON PL FI -3.37 0.44 1.0 0.82 20,403
19 R OFF FFL FI -3.63 0.53 1.0 1.71 22,538
20 R ON FFL FI -3.37 0.44 1.0 0.82 20,403
21 R FWY DIR FI -2.80 0.46 1.0 1.89 37,474
22 U OFF D FI -6.88 0.78 1.0 2.21 22,566
23 U ON D FI -14.40 1.61 1.0 3.44 24,966
24 U OFF PL FI -3.63 0.53 1.0 1.71 22,538
25 U ON PL FI -3.37 0.44 1.0 0.82 20,403
26 U OFF FFL FI -3.63 0.53 1.0 1.71 22,538
27 U ON FFL FI -3.37 0.44 1.0 0.82 20,403
28 U FWY DIR FI -2.80 0.46 1.0 1.89 37,474
Table 18. Definitions of coefficients and parameters of SPFs for ramps.
Variable Name Variable No. Format Definition
SPF No. DR1 Numeric Each SPF included in the analysis must be identified by a sequential integer, starting with 1; this variable is not used in calculations but is useful for organization.
Area type DR2 Character(U, R) General character of land use surrounding the interchange, preferably based on FHWA urban area boundaries:
U = Urban
R = Rural
Ramp type DR3 Character(ON, OFF, FWY) The type of ramps are distinguished by the following codes:
ON = on-ramp
OFF = off-ramp
FWY = freeway-to-freeway ramp
Ramp configuration DR4 Character(D, PL, FFL, DIR) This variable defines the basic geometric design of the ramp; the abbreviated codes correspond as follows:
D = diamond ramp
PL = parclo loop ramp
FFL = free-flow loop ramp
DIR = directional ramp
Severity level DR5 Character(TOT, FI) This variable identifies the crash severity level:
TOT = total crashes
FI = fatal and injury crashes
Log intercept (a) DR6 Numeric Intercept of SPF.
Coefficient of log ADT (b) DR7 Numeric Coefficient of log ADT (ramp) parameter.
Coefficient of ramp length (e) DR8 Numeric Coefficient of ramp length parameter.
Dispersion parameter DR9 Numeric Dispersion parameter associated with negative binomial regression.
Max ADT DR10 Numeric(veh/day) The maximum traffic volume of the ramp for which the coefficients of the SPF apply.

Table 19 presents the coefficients and parameters of the default SPFs for acceleration lanes provided within ISAT. Table 19 also shows the functional form of the models. Table 20 defines the coefficients and parameters of the defaults SPFs. Four SPFs for acceleration lanes are provided within ISAT and are a function of the following:

  • Area type.
  • Severity level.

Portions of the SPFs Accel Lanes worksheet are protected. The user may only modify the default values for the Constant (C), Log intercept (a), Coeff of ADTRamp (b), Coeff of length (c), Coeff of ADTFreeway (d), dispersion parameter, and mean length.

Table 19. SPF Coefficients and parameters for acceleration lanes.
SPF No. Area Type Severity Level Constant (C) Log Intercept (a) CoeffofADTRamp (b) Coeffof Length (c) CoeffofADTFreeway (d) Dispersion Parameter Mean Length
numeric (R, U) (TOT, FI) numeric numeric numeric numeric numeric numeric km (mi)
Bauer and Harwood Acceleration Lane SPF
N = C x ea x ADTRampb x ec x ALL x ADTFreewayd
DAL1 DAL2 DAL3 DAL4 DAL5 DAL6 DAL7 DAL8 DAL9 DAL10
1 R TOT 0.44 -7.19 0.78 -2.59 0.13 0.66 0.16 (0.1)
2 U TOT 0.44 -6.82 0.78 -2.59 0.13 0.66 0.16 (0.1)
3 R FI 0.55 -10.68 0.91 -4.55 0.29 0.52 0.16 (0.1)
4 U FI 0.55 -10.68 0.91 -4.55 0.29 0.52 0.16 (0.1)
Table 20. Definitions of coefficients and parameters of SPFs for acceleration Lanes.
Variable Name Variable No. Format Definition
SPF No. DAL1 Numeric Each SPF in the analysis must be identified by a sequential integer, starting with 1; this variable is not used in calculations but is useful for organization.
Area type DAL2 Character(U, R) General character of land use surrounding the interchange, preferably based on FHWA urban area boundaries:
U = Urban
R = Rural
Severity level DAL3 Character(TOT, FI) This variable identifies the crash severity level:
TOT = total crashes
FI = fatal and injury crashes
Constant (C) DAL4 Numeric This constant is provided because the SPFs were developed to predict crashes for a 3-year period. The constant scales the prediction to an annual basis. The constant also accounts for a ramp length term (i.e., the length of the ramp proper) in the original form of the SPF.
Log intercept (a) DAL5 Numeric Intercept of SPF.
Coefficient of log ADTRamp (b) DAL6 Numeric Coefficient of log ADT (ramp) parameter.
Coefficient of length (c) DAL7 Numeric Coefficient of acceleration lane length (ALL) parameter.
Coefficient of log ADTFreeway (d) DAL8 Numeric Coefficient of log ADT (freeway) parameter (this ADT is directional).
Dispersion parameter DAL9 Numeric Dispersion parameter associated with negative binomial regression.
Mean length DAL10 Numeric(mi) Mean length of acceleration lanes used to develop SPFs.

Crossroad Ramp Terminal and Intersection SPFs

The default SPFs for crossroad ramp terminals and intersections provided within ISAT are located on the SPFs Ramp Terminals worksheet and are accessible only through the Go To SPFs for Ramp Terminals hyperlink on the Input-General worksheet. The default SPFs for crossroad ramp terminals and intersections were developed for use within SafetyAnalyst. Technically, these SPFs were developed using data for conventional at-grade intersections, but due to the lack of any suitable safety prediction models developed specifically for crossroad ramp terminals, these models are used to predict crashes at both crossroad ramp terminals and conventional intersections defined as part of an analysis area.

Table 21 presents the coefficients and parameters of the default SPFs for crossroad ramp terminals and intersections provided within ISAT. Table 21 also shows the functional form of the models. Table 22 defines the coefficients and parameters of the defaults SPFs. Sixteen SPFs for crossroad ramp terminals and intersections are provided within ISAT and are a function of the following:

  • Area type.
  • Type of traffic control.
  • Number of legs.
  • Severity level.

Portions of the SPFs Ramp Terminals worksheet are protected. The cells with column headings for SPF No., Area type, Type of traffic control, Number of legs, and Severity level are protected. The user may only modify the default values for the Log intercept (a), Coeff of ADTmajor rd (b), Coeff of ADToff-ramp (c), dispersion parameter, Max ADTmajor rd, and Max ADToff-ramp.

Crossroad Segment SPFs

The default SPFs for arterial crossroad segments provided within ISAT are located on the SPFs Crossroad Segments worksheet and are accessible only through the Go To SPFs for Crossroad Segments hyperlink on the Input-General worksheet. The default SPFs for arterial crossroad segments were developed for use within SafetyAnalyst. These SPFs predict the number of crashes that may occur on urban and rural arterial streets. These predictions include crashes that occur at intersections located on the crossroad roadway segments and are not related to the operation of the intersections; however the predictions do not include crashes that occur at intersections located on the crossroad roadway segments and are related to the operation of the intersections (i.e., these intersection related crashes are included in the predictions for crossroad ramp terminals and intersections).

Table 21. SPF Coefficients and parameters for crossroad ramp terminals and intersections.
SPF No. Area Type Type of Traffic Control Number of Legs Severity Level Log Intercept (a) Coeff of ADTmajor rd (b) Coeff ofADToff-ramp (c) Dispersion Parameter Max ADTmajor rd MaxADToff-ramp
numeric (U, R) (SG, ST) (3, 4) (TOT, FI) numeric numeric numeric numeric (veh/day) (veh/day)
SafetyAnalyst Ramp Terminal SPF
N = e a x ADTmajor rdb x ADToff-rampc
DRT1 DRT2 DRT3 DRT4 DRT5 DRT6 DRT7 DRT8 DRT9 DRT10 DRT11
1 R ST 3 TOT -8.78 0.71 0.24 1.07 28,500 27,000
2 R SG 3 TOT -6.57 0.66 0.20 0.33 36400 11500
3 R ST 4 TOT -8.96 0.65 0.47 0.70 35,500 26,700
4 R SG 4 TOT -6.57 0.66 0.20 0.33 36,400 11,500
5 U ST 3 TOT -5.35 0.34 0.28 1.28 68,000 18,900
6 U SG 3 TOT -9.85 0.97 0.18 0.23 50,000 25,807
7 U ST 4 TOT -3.12 0.27 0.16 0.86 58,870 81,000
8 U SG 4 TOT -3.47 0.42 0.14 0.32 75,000 81,000
9 R ST 3 FI -9.35 0.71 0.21 1.23 28,500 27,000
10 R SG 3 FI -7.83 0.75 0.14 0.50 36400 11500
11 R ST 4 FI -9.36 0.66 0.40 0.00 35,500 26,700
12 R SG 4 FI -7.83 0.75 0.14 0.50 36,400 11,500
13 U ST 3 FI -8.45 0.49 0.39 1.23 68,000 18,900
14 U SG 3 FI -10.22 0.91 0.21 0.27 50,000 25,807
15 U ST 4 FI -4.35 0.29 0.19 0.99 58,870 81,000
16 U SG 4 FI -5.11 0.49 0.16 0.30 75,000 81,000
Table 22. Definitions of coefficients and parameters of SPFs for crossroad ramp terminals and intersections.
Variable Name Variable No. Format Definition
SPF No. DRT1 Numeric Each SPF included in the analysis must be identified by a sequential integer, starting with 1; this variable is not used in calculations but is useful for organization.
Area type DRT2 Character(U, R) General character of land use surrounding the interchange, preferably based on FHWA urban area boundaries:
U = Urban
R = Rural
Type of traffic control DRT3 Character(SG, ST) A code identifying the type of traffic control for the crossroad ramp terminal or intersection:
SG = signalized intersection
ST = STOP-control on the ramp or minor roadway; no control on the major crossroad
Number of legs DRT4 Numeric(3,4) A code identifying the number of legs of the crossroad ramp terminal or intersection. In determining the number of legs, the user should consider whether it is most appropriate to treat each ramp that is served by the terminal as an individual leg.
3 = three-legs
4 = four-legs
Severity level DRT5 Character(TOT, FI) A code identifying the crash severity level:
TOT = total crashes
FI = fatal and injury crashes
Log intercept (a) DRT6 Numeric Intercept of SPF.
Coefficient of log ADTmajor rd (b) DRT7 Numeric Coefficient of log ADT (major rd) (bi-directional) (ADTmajor rd is calculated 2 x major rd ADT [RT5]).
Coefficient of log ADToff-ramp (c) DRT8 Numeric Coefficient of log ADT (off-ramp) (ADToff-ramp is minor-road or ramp ADT [RT8] when terminal type [RT11] is RT. ADToff-ramp is 2 x minor-road or ramp ADT [RT8] when terminal type [RT11] is CI.)
Dispersion parameter DRT9 Numeric Dispersion parameter associated with negative binomial regression.
Max ADTmajor rd DRT10 Numeric(veh/day) The maximum traffic volume on the crossroad for which the coefficients of the SPF apply.
Max ADToff-ramp DRT11 Numeric(veh/day) The maximum traffic volume on the ramp for which the coefficients of the SPF apply.

Table 23 presents the coefficients and parameters of the default SPFs for arterial crossroad segments provided within ISAT. Table 23 also shows the functional form of the models. Table 24 defines the coefficients and parameters of the defaults SPFs. Twenty SPFs for crossroad roadway segments are provided within ISAT and are a function of the following:

  • Area type.
  • Number of through lanes.
  • Median type.
  • Severity level.

Portions of the SPFs Crossroad Segments worksheet are protected. The cells with column heads for SPF No., Area type, Number of through lanes, Median, and Severity level are protected. The user may only modify the default values for the Log intercept (a), Coeff of log ADT (b), dispersion parameter, and Max ADT.

Calibration Coefficients

ISAT makes use of SPFs from previous and ongoing safety research. SPFs developed for SafetyAnalyst, and as part of research conducted by Bauer and Harwood3, are incorporated into the program. The SPFs for SafetyAnalyst are based upon data from California, Minnesota, Ohio, and Washington. Bauer and Harwood3 used data from Washington to formulate their statistical models. As such, the SPFs in ISAT are valid only for application to the States and time periods for which the models were developed. However, ISAT includes calibration coefficients that allow the SPFs developed for one particular State and one particular time period to be applied to other areas and time periods. By adjusting the calibration coefficients within ISAT based upon local crash data, the calculations within ISAT are more directly applicable to the user's own agency and more useful safety predictions are obtained, better reflecting the local safety experience.

It is recommended that one of the first steps a user should take when beginning to work with ISAT, and definitely before using ISAT to perform actual safety assessments, is modify the calibration coefficients based upon an agency's own crash data. Subsequently, these calibration coefficients should be updated on an annual basis.

Table 23. SPF coefficients and parameters for crossroad segments.
SPF No. Area Type Number of Through Lanes (Directional) Median Severity Level Log Intercept (a) Coeff of Log ADT (b) Dispersion Parameter Max ADT
numeric (U, R) (1, 2, 3) (D, U) (TOT, FI) numeric numeric numeric (veh/day)
SafetyAnalyst Crossroad Segment SPF
N = ea x ADTb x SL
DRS1 DRS2 DRS3 DRS4 DRS5 DRS6 DRS7 DRS8 DRS9
1 R 1 U TOT -3.56 0.55 0.45 30,025
2 R 2 U TOT -2.58 0.44 1.28 42,638
3 R 3 U TOT -2.58 0.44 1.28 42,638
4 R 2 D TOT -8.21 0.93 0.47 31,188
5 R 3 D TOT -8.21 0.93 0.47 31,188
6 U 1 U TOT -7.16 0.84 4.40 29,850
7 U 2 U TOT -10.24 1.29 0.85 57,901
8 U 3 U TOT -10.24 1.29 0.85 57,901
9 U 2 D TOT -11.85 1.34 5.91 77,735
10 U 3 D TOT -11.85 1.34 5.91 77,735
11 R 1 U FI -4.89 0.53 0.45 30,025
12 R 2 U FI -4.96 0.60 1.06 42,638
13 R 3 U FI -4.96 0.60 1.06 42,638
14 R 2 D FI -10.31 1.03 0.33 31,188
15 R 3 D FI -10.31 1.03 0.33 31,188
16 U 1 U FI -8.84 0.89 4.54 29,850
17 U 2 U FI -12.07 1.39 0.81 57,901
18 U 3 U FI -12.07 1.39 0.81 57,901
19 U 2 D FI -14.87 1.52 5.81 77,735
20 U 3 D FI -14.87 1.52 5.81 77,735
Table 24. Definitions of coefficients and parameters of SPFs for crossroad segments.
Variable Name Variable No. Format Definition
SPF No. DRS1 Numeric Each SPF in the analysis must be identified by a sequential integer, starting with 1; this variable is not used in calculations but is useful for organization.
Area type DRS2 Character(U, R) General character of land use surrounding the interchange, preferably based on FHWA urban area boundaries:
U = Urban
R = Rural
Number of through lanes DRS3 Numeric(1, 2, 3) A code identifying the number of through lanes in a given direction on the crossroad segment.
Median DRS4 Character(U, D) A code identifying whether a median is present on the crossroad segment:
U = the segment is undivided
D = the segment is divided
Severity level DRS5 Character(TOT, FI) A code identifying the crash severity level:TOT = total crashesFI = fatal and injury crashes
Log intercept (a) DRS6 Numeric Intercept of SPF.
Coefficient of log ADT (b) DRS7 Numeric Coefficient of log ADT parameter (this ADT is bi-directional).
Dispersion parameter DRS8 Numeric Dispersion parameter with negative binomial regression.
Max ADT DRS9 Numeric(veh/day) The maximum traffic volume on the crossroad for which the coefficients of the SPF apply.

The calibration coefficients adjust the predicted number of total and fatal and injury crashes, calculated within ISAT, to account for differences in crash patterns in different geographical areas that are not directly addressed by the SPFs. These differences may be related to differences in:

  • Driver population and trip purposes.
  • Climate.
  • Animal populations.
  • Crash reporting thresholds.
  • Crash investigation practices.

The calibration coefficients are intended to account for these differences and provide crash predictions that are comparable to the estimates that a highway agency would obtain had the SPFs in ISAT been developed using their own crash records system.

Tables 25 through 28 show the specific SPFs for mainline freeway segments, ramps, crossroad ramp terminals and intersections, and crossroad roadway segments that need calibration coefficients. The nominal or default value of the calibration coefficients is shown in the tables as 1.00 for each of the SPFs. This nominal value for each SPF needs to be replaced with a calculated value appropriate for the highway agency applying the model. In general terms, calibration coefficients greater than 1.00 apply to agencies that experience more crashes than predicted by the default SPFs. Calibration coefficients less than 1.00 apply to agencies that experience fewer crashes than predicted by the default SPFs. A procedure for calculating calibration coefficients is presented below.

Calibration Procedure

This calibration procedure is applicable to each SPF (i.e., each row) in tables 25 through 28. Thus, in total this procedure should be carried out 20 times to calibrate the mainline freeway segment SPFs, 28 times to calibrate the ramp SPFs, 20 times to calibrate the crossroad ramp terminal and intersection SPFs, and 20 times to calibrate the crossroad roadway segment SPFs. The calibration procedure described below makes use of ISAT, but the procedure could be performed externally without making use of ISAT. The nomenclature used for describing the calibration procedure is defined in table 29.

Table 25. Calibration coefficients for mainline freeway segment SPFs.
CCMF1 CCMF2 CCMF3 CCMF4 CCMF5 CCMF6
SPF No. Area Type Interchange Area Number of Through Lanes (Directional) Severity Level Calibration Coefficients
1 R Y 2 TOT 1.000
2 R Y 3 TOT 1.000
3 U Y 2 TOT 1.000
4 U Y 3 TOT 1.000
5 U Y 4 TOT 1.000
6 R Y 2 FI 1.000
7 R Y 3 FI 1.000
8 U Y 2 FI 1.000
9 U Y 3 FI 1.000
10 U Y 4 FI 1.000
11 R N 2 TOT 1.000
12 R N 3 TOT 1.000
13 U N 2 TOT 1.000
14 U N 3 TOT 1.000
15 U N 4 TOT 1.000

16

R

N

2

FI

1.000

17

R

N

3

FI

1.000

18

U

N

2

FI

1.000

19

U

N

3

FI

1.000

20

U

N

4

FI

1.000

NOTE: The default values of 1.00 for the calibration coefficients should be replaced with calculated values appropriate for specific highway agencies.

Table 26. Calibration coefficients for ramp SPFs.
CCR1 CCR2 CCR3 CCR4 CCR5 CCR6
SPF No. Area Type Type of Ramp Ramp Configuration Severity Level Calibration Coefficients
1 R OFF D TOT 1.000
2 R ON D TOT 1.000
3 R OFF PL TOT 1.000
4 R ON PL TOT 1.000
5 R OFF FFL TOT 1.000
6 R ON FFL TOT 1.000
7 R FWY DIR TOT 1.000
8 U OFF D TOT 1.000
9 U ON D TOT 1.000
10 U OFF PL TOT 1.000
11 U ON PL TOT 1.000
12 U OFF FFL TOT 1.000
13 U ON FFL TOT 1.000
14 U FWY DIR TOT 1.000
15 R OFF D FI 1.000
16 R ON D FI 1.000
17 R OFF PL FI 1.000
18 R ON PL FI 1.000
19 R OFF FFL FI 1.000
20 R ON FFL FI 1.000
21 R FWY DIR FI 1.000
22 U OFF D FI 1.000
23 U ON D FI 1.000
24 U OFF PL FI 1.000
25 U ON PL FI 1.000
26 U OFF FFL FI 1.000
27 U ON FFL FI 1.000
28 U FWY DIR FI 1.000
NOTE: The default values of 1.00 for the calibration coefficients should be replaced with calculated values appropriate for specific highway agencies.
Table 27. Calibration coefficients for crossroad ramp terminal and intersection SPFs.
CCRT1 CCRT2 CCRT3 CCRT4 CCRT5 CCRT6
SPF No. Area Type Type of Traffic Control Number of Legs Severity Level Calibration Coefficients
1 R ST 3 TOT 1.000
2 R SG 3 TOT 1.000
3 R ST 4 TOT 1.000
4 R SG 4 TOT 1.000
5 U ST 3 TOT 1.000
6 U SG 3 TOT 1.000
7 U ST 4 TOT 1.000
8 U SG 4 TOT 1.000
9 R ST 3 FI 1.000
10 R SG 3 FI 1.000
11 R ST 4 FI 1.000
12 R SG 4 FI 1.000
13 U ST 3 FI 1.000
14 U SG 3 FI 1.000
15 U ST 4 FI 1.000
16 U SG 4 FI 1.000
Note: The default values of 1.00 for the calibration coefficients should be replaced with calculated values appropriate for specific highway agencies.
Table 28. Calibration coefficients for crossroad segment SPFs.
CCRS1 CCRS2 CCRS3 CCRS4 CCRS5 CCRS6
SPF No. Area Type Number of Through Lanes (Directional) Median Severity Level Calibration Coefficients
1 R 1 U TOT 1.000
2 R 2 U TOT 1.000
3 R 3 U TOT 1.000
4 R 2 D TOT 1.000
5 R 3 D TOT 1.000
6 U 1 U TOT 1.000
7 U 2 U TOT 1.000
8 U 3 U TOT 1.000
9 U 2 D TOT 1.000
10 U 3 D TOT 1.000
11 R 1 U FI 1.000
12 R 2 U FI 1.000
13 R 3 U FI 1.000
14 R 2 D FI 1.000
15 R 3 D FI 1.000
16 U 1 U FI 1.000
17 U 2 U FI 1.000
18 U 3 U FI 1.000
19 U 2 D FI 1.000
20 U 3 D FI 1.000
NOTE: The default values of 1.00 for the calibration coefficients should be replaced with calculated values appropriate for specific highway agencies.
Table 29. Nomenclature for calibration procedures.
Term Explanation
i Subscript to represent site i
y Subscript to represent the year y
The first year of the analysis period is considered Year 1 (i.e., y=1)
The last year of the analysis period is considered Year Y (i.e., y=Y)
TOT Subscript to denote total crashes
FI Subscript to denote fatal and injury crashes
Oiy Observed number of crashes at site i during year y
O Total number of observed crashes for the entire analysis period across all individual sites of the interchange element of interest
Niy Predicted number of crashes at a site i during year y
NA Total number of predicted crashes for the entire analysis period across all individual sites of the interchange element of interest
C Calibration coefficient

Step 1. Select sites for use in applying the calibration procedures.

Select up to a maximum of 50 sites of the same type for use in applying the calibration procedures. There is no specific number of sites that must be selected in this step, in part because the number of sites that should be included is dependent upon the next step (i.e., select analysis years). As a general rule of thumb, at least 20 sites are desirable, and selecting more sites should increase the accuracy of the predictions output by ISAT.

Step 2. Select the analysis years for applying the calibration procedures.

Two approaches could be considered when performing this step. One approach would be to limit the analysis period to the most recent year for which a full calendar year of crash data are available. If this approach is taken, the calibration coefficients and calculations will specifically reflect the crash trends of the selected calendar year.

The second approach is to select multiple years for the analysis period. It is recommended that up to the 10 most recent years for which crash data are available for the selected sites be included in the analysis. If this approach is taken, the calibration coefficients and calculations will not specifically reflect crash trends for the most recent year of crash data, but rather reflect the average crash trend over the collective analysis years. This second approach (i.e., multiple years) is preferred over calibrating the SPFs based upon the most recent year crash data.

A couple of issues should be considered collectively when selecting sites and analysis years on which to base the calibration coefficients:

  • The sites should not have undergone reconstruction at any point during this time period.
  • The analysis period has to be same for all sites.

Step 3. Predict the total number of crashes across the selected sites and analysis years.

On the Input-General worksheet, the user will input data for variables G1 through G4. The beginning and ending year of analysis coincide with the analysis years as selected in Step 2. For variable G5 (i.e., crash data available), a value of "N" should be provided for each primary element (i.e., mainline freeway segments, ramps, crossroad ramp terminals and intersections, and crossroad segments).

On the respective input worksheet either for mainline freeway segments, ramps, crossroad ramp terminals and intersections, or crossroad segments, the user should input data for the sites selected in Step 1.

As part of Step 3, the calibration coefficient for the respective site subtype should be modified to the default value of 1.00.

Having entered all of the input data, the user can click the Perform Calculations button. The table for number of predicted crashes by interchange element type will show the number of predicted crashes for the respective interchange element type for both total and fatal and injury crash levels. Because data were input for only one subtype of the given interchange element, the predicted values are applicable to the given subtype, and these values are truly predicted values because no crash data were considered in the analysis. This calculation is represented as:

. Capital N subscript- A equals summation, lowercase i summation, lowercase y, capital N subscript i-y. Explanation: Total number of predicted crashes across selected sites and analysis years bracket C-A-P parenthesis N close parenthesis subscript- A close bracket equals double summation across select sites, i, and analysis years, y, of predicted number of crashes bracket C-A-P-parenthesis N close parenthesis, close bracket at site i during year y. (Eq. 1)

Step 4. Determine the total number of observed crashes across the selected sites and analysis years.

This step is performed externally to ISAT and is represented as:

Capital O equals summation, lowercase i, summation, lowercase y, capital O subscript i-y. Explanation: Total number of observed crashes across selected sites and analysis years bracket C-A-P parenthesis O close parenthesis, close bracket equals double summation across select sites, i, and analysis years, y, of observed number of crashes bracket C-A-P-parenthesis O close parenthesis, close bracket at site i during year y. (Eq. 2)

When determining the number of observed crashes across selected sites, the same rules for attributing crashes to particular interchange element types should be followed as explained in the General Interchange Inputs section of this manual. The rules are summarized here as follows:

  • Crashes that occur along or within mainline freeway segments, deceleration lanes, and acceleration lanes should be attributed to mainline freeway segments.
  • All crashes that occur within 76.2 m (250 ft) of a crossroad ramp terminal or intersection and are classified as intersection-related should be attributed to crossroad ramp terminals and intersections. All crashes that occur within 76.2 m (250 ft) of a crossroad ramp terminal or intersection but are not classified as intersection-related should be attributed either to ramps or crossroad roadway segments.
  • All crashes that occur along the ramp proper portion of an interchange ramp should be attributed to ramps. For crashes that occur on the ramp proper and are within 76.2 m (250 ft) of the crossroad ramp terminal, if the crash is related to the operation of the ramp terminal (i.e., intersection-related), then the crash should be attributed to the crossroad ramp terminal, but if the crash is not related to the operation of the ramp terminal, then the crash should be attributed to the ramp.
  • Crashes that occur along or within arterial crossroad roadway segments should be attributed as such, except those crashes that occur within 76.2 m (250 ft) of a ramp terminal or intersection and are intersection-related, in which case the crashes should be attributed to crossroad ramp terminals and intersections.

Step 5. Compute the calibration coefficient such that:

Capital C equals capital O divided by capital N subscript A equals summation parenthesis i close parenthesis summation parenthesis y close parenthesis capital N subscript i-y, that amount divided by summation parenthesis ii close parenthesis summation parenthesis y close parenthesis capital N subscript i-y. Explanation: Calibration coefficient bracket Cap parenthesis C close parenthesis, close bracket equals total number of observed crashes across selected sites and analysis years bracket C-A-P parenthesis O close parenthesis, close bracket divided by total number of predicted crashes across selected sites and analysis years bracket C-A-P-parenthesis N close parenthesis) subscript A, close bracket, which equals double summation across select sites, i, and analysis years, y, of observed number of crashes bracket CAP parenthesis O close parenthesis, close bracket at site i during year y divided by double summation across select sites, i, and analysis years, y, of predicted number of crashes bracket C-A-P-parenthesis N close parenthesis, close bracket at site i during year y. (Eq. 3)

This is the value of the calibration coefficient that would be entered into the calibration table for the respective SPF.

Calibration Example

This example illustrates the general calibration procedure for calculating the calibration coefficient for the mainline freeway segment SPF No. 1. Table 15 shows that SPF No. 1 for mainline freeway segments is used to predict crash frequencies for the following situation:

Interchange element:

Mainline freeway segment

Area type:

Rural

Interchange area:

Yes

Number of through lanes (directional):

2

Severity level:

TOT

Step 1. Select sites for use in applying the calibration procedures.

In this hypothetical example, 20 sites were selected (i.e., 10 sites on I-80 and 10 sites on
I-79).

Step 2. Select the analysis years for applying the calibration procedures.

Five years of crash data were used, from 2001 through 2005.

Step 3. Predict the total number of crashes across the selected sites and analysis years.

Table 30 shows the type of data input on the Input-General worksheet. In this case, descriptive data are input for variables G1 through G4, and when asked about the availability of crash data, the input is "N" even though crash data are actually available. If crash data were entered on the Input-General worksheet, then the output would be an EB-adjusted expected value representing a weighted average of the predicted values from the SPF and the observed crash totals. By not including the crash data in these calculations, the output results represent predicted values using mainline freeway segment SPF No. 1.

Table 31 shows the type of data to be entered on the Input Mainline Freeways worksheet. Data for 20 sites are entered into this worksheet. To reiterate, these sites should all be of the same type. In this example, these sites are rural mainline freeway segments within interchange area that have 2 through lanes in the designated direction.

Having entered all of the input data on the Input-General and Input Mainline Freeways worksheets, the calculations are performed by clicking the Perform Calculations button on the Input-General worksheet. Table 32 shows a portion of the output report generated for this example. In this case the table of interest is the number of predicted crashes by interchange element type which shows the number of predicted crashes for the four interchange element types. In this example data were input only for mainline freeway segments. Table 32 indicates that 55.1 total crashes were predicted to occur on these 20 sites during the 5 year period from 2001 to 2005.

Capital N subscript A equals 55.1 total crashes. Explanation: Total number of predicted crashes bracket C-A-P-parenthesis N close parenthesis subscript A close bracket across selected sites and analysis years equals 55.1 total crashes. (Eq. 4)

Step 4. Determine the total number of observed crashes across the selected sites and analysis years.

Assume that 60 total crashes were observed at these 20 sites during the period from 2001 through 2005.

Capital O equals 60 total crashes. Explanation: Total number of observed crashes bracket C-A-P-parenthesis O close parenthesis, close bracket across selected sites and analysis years equals 60 crashes. (Eq. 5)

Step 5. Compute the calibration coefficient.

Plugging the predicted and observed values into equation 3 yields:

Capital C equals capital O divided by capital N subscript A, which equals 60 divided by 55.1, which equals 1.089. Explanation: Calibration coefficient bracket C close bracket equals total number of observed crashes bracket C-A-P parenthesis O close parenthesis, close bracket across selected sites and analysis years divided by total number of predicted crashes bracket C-A-P parenthesis N close parenthesis subscript A close bracket across selected sites and analysis years, which equals 60 divided by 55.1, which equals 1.089. (Eq. 6)

This value would then be entered into the calibration table for mainline freeway segments SPF No. 1 as illustrated in table 33. This procedure would be repeated for each SPF for mainline freeway segments, and the resulting calculated value would be entered as appropriate into the calibration table. Similarly, this procedure would be repeated for each SPF for ramps, crossroad ramp terminals and intersections, and crossroad roadway segments.

Table 30. Example data for calibration procedure (Input-General).
GI1 Project description character MF (calibration)      
GI2 Analyst   MRI      
GI3 Date   2/20/2007      
GI4 Area type (U,R) R      
GI5 Beginning year of analysis numeric 2001      
GI6 Ending year of analysis numeric 2005      
             
             
      Mainline Freeways Ramps Crossroad Ramp Terminals and Intersections Crossroad Segments
GI7 Crash data available? (Y, N) N N N N
GI8 Beginning year of crash data numeric        
GI9 Ending year of crash data numeric        
GI10 Observed number of crashes numeric        
Table 31. Example data for calibration procedure (Input Mainline Freeways).
MF1 MF2 MF3 MF4 MF5 MF6 MF7 MF8 MF9 MF10 MF11
Segment Number Segment Description Direction of Travel Beginning MP Ending MP LengthofSegment Number of Through Lanes (Directional) Mainline ADT (Directional) Mainline ADT Year ADT Growth Rate Within Interchange Area?
numeric character (NB, SB,EB, WB) numeric numeric km (mi) (2, 3, 4) (veh/day) numeric (percent/ year) (Y, N)
1 EB I-80 EB 1.000 1.350 0.350 2 4,000 2004 2.0 Y
2 EB I-80 EB 10.000 10.400 0.400 2 4,000 2004 2.0 Y
3 EB I-80 EB 50.000 50.500 0.500 2 3,500 2004 2.0 Y
4 EB I-80 EB 100.000 100.400 0.400 2 4,500 2004 2.0 Y
5 EB I-80 EB 150.000 150.500 0.500 2 4,500 2004 2.0 Y
6 WB I-80 WB 25.000 25.500 0.500 2 4,500 2004 2.0 Y
7 WB I-80 WB 75.000 75.400 0.400 2 4,500 2004 2.0 Y
8 WB I-80 WB 125.000 125.400 0.400 2 3,500 2004 2.0 Y
9 WB I-80 WB 175.000 175.500 0.500 2 4,000 2004 2.0 Y
10 WB I-80 WB 225.000 225.350 0.350 2 4,000 2004 2.0 Y
11 NB I-79 NB 5.000 5.500 0.500 2 3,000 2005 2.0 Y
12 NB I-79 NB 40.000 40.400 0.400 2 3,250 2005 2.0 Y
13 NB I-79 NB 80.000 80.500 0.500 2 4,000 2005 2.0 Y
14 NB I-79 NB 120.000 120.500 0.500 2 3,700 2005 2.0 Y
15 NB I-79 NB 170.000 170.400 0.400 2 3,000 2005 2.0 Y
16 SB I-79 SB 10.000 10.400 0.400 2 4,000 2005 2.0 Y
17 SB I-79 SB 60.000 60.350 0.350 2 4,250 2005 2.0 Y
18 SB I-79 SB 130.000 130.350 0.350 2 4,500 2005 2.0 Y
19 SB I-79 SB 180.000 180.400 0.400 2 4,000 2005 2.0 Y
20 SB I-79 SB 200.000 200.500 0.500 2 3,800 2005 2.0 Y
Table 32. Example output report for calibration procedure.
General Interchange Information      
Project description: MF (calibration)      
Analyst: MRI      
Date: 2/20/2007      
Area type: Rural      
Beginning year of analysis period: 2001      
Ending year of analysis period: 2005      
    Mainline Freeway Segments Ramps Crossroad Ramp Terminals and Intersections Crossroad Segments      
Crash data available: N N N N      
Beginning year of crash data:              
Ending year of crash data:              
                 
Number of Predicted Crashes for Entire Interchange Area      
Number of Predicted Crashes During Analysis Period Average Number of Predicted Crashes per Year During Analysis Period      
     
Total FI PDO Total FI PDO      
55.1 16.3 38.9 11.0 3.3 7.8      
                 
Number of Predicted Crashes by Interchange Element Type  
Interchange Element Type Number of Sites Number of Predicted Crashes During Analysis Period MVMT MEV Crash Rate
Total FI PDO (per MVMT or MEV)
Mainline freeway segments 20 55.1 16.3 38.9 59.691   0.923
Ramps  0            
Crossroad ramp terminals & ints 0            
Crossroad segments 0            
Total  20 55.1 16.3 38.9 59.691   0.923
Table 33. Example calibration procedure (Input Calibration Worksheet).
CCMF1 CCMF2 CCMF3 CCMF4 CCMF5 CCMF6
SPF No. Area Type Interchange Area Number of Through Lanes (Directional) Severity Level Calibration Coefficients
1 R Y 2 TOT 1.089
2 R Y 3 TOT  
3 U Y 2 TOT  
4 U Y 3 TOT  
5 U Y 4 TOT  
6 R Y 2 FI  
7 R Y 3 FI  
8 U Y 2 FI  
9 U Y 3 FI  
10 U Y 4 FI  
11 R N 2 TOT  
12 R N 3 TOT  
13 U N 2 TOT  
14 U N 3 TOT  
15 U N 4 TOT  
16 R N 2 FI  
17 R N 3 FI  
18 U N 2 FI  
19 U N 3 FI  
20 U N 4 FI  

Three comments concerning this procedure are worth noting:

  • SPFs for a particular site subtype are provided for two severity levels, total and fatal and injuries. The output from Step 3 provides predicted values for both severity levels. In the example above, the predicted number of fatal and injury crashes for the 20 sites is 16.3 crashes for the 5 year period (Table 32). Thus, the output for a particular subtype of interchange element could be used in calculating the calibration coefficients for both SPFs (i.e., TOT and FI).
  • In terms of efficiency, it is possible that when using ISAT to calculate the predicted values for calibration purposes, data could be entered for 4 subtypes of interchange elements as long as they were for different types of elements (i.e., mainline freeway segments, ramps, crossroad ramp terminals and intersections, and crossroad roadway segments). In this case, the output report could be used for calculating the calibration coefficients for up to 8 SPFs (i.e., 2 for each interchange element) in one iteration.
  • In Step 3 it is noted that when performing the calibration procedures, the calibration coefficient for the respective site subtype should be the default value of 1.00. This is important so that the predicted values from the SPFs (i.e., NA) are not scaled/adjusted prior to calculating the calibration coefficient.

Crash Distributions by Severity and Type

This section provides the default distributions of crash severity and crash type for specific types of interchange elements. The default distributions are based upon crash data from Washington State. The crash distributions are used to predict crash frequencies for 3 severity levels (i.e., TOT, FI, and PDO) and 13 crash types. The crash types are classified as either single-vehicle crashes (SV) or multiple-vehicle crashes (MV). In general terms these crash proportions are multiplied by the predicted or estimated values for total and fatal and injury crashes to obtain the number of crashes of a specified collision type and severity. The number of PDO crashes is estimated by subtracting fatal and injury crashes from total crashes.

It is recommended that one of the first steps a user should do when beginning to work with ISAT, and definitely before using ISAT to perform actual safety assessments, is modify the crash distributions based upon an agency's own crash data. Subsequently, these crash distributions should be updated on an annual basis.

The crash proportions are calculated as follows. The calculations apply to a given type of site and severity level. The classifications of subtypes for which crash distributions are necessary are somewhat different than the subtypes of SPFs for the respective interchange elements. In some cases, multiple subtypes are aggregated together for calculating crash proportions so the crash proportions are applied to multiple subtypes of interchange elements. In calculating the crash type proportions, it is assumed that crash data are available for Y years at I sites of a given subtype in a given jurisdiction (e.g., State). The terms in the calculations are defined in table 34. The proportions, P(CT/TOT) and P(CT/FI), are computed across all sites I of a given subtype as follows:

Table 34. Nomenclature for crash distribution calculations.
Term Explanation
i Subscript to represent site i
y Subscript to represent the year y
The first year of the analysis period is considered Year 1 (i.e., y=1)
The last year of the analysis period is considered Year Y (i.e., y=Y)
TOT Subscript to denote total crashes
FI Subscript to denote fatal and injury crashes
CT Subscript to denote crash type
Oiy Observed number of crashes at site i during year y
P(CT/TOT) Calculated proportion of crashes of the crash type of interest for all TOT crashes for all years of available data
P(CT/FI) Calculated proportion of crashes of the crash type of interest for all FI crashes for all years of available data

Capital P subscript capital C-T forward slash T-O-T equals summation parenthesis i close parenthesis summation parenthesis y close parenthesis of capital O subscript i-y capital C-T forward slash T-O-T, that amount divided by summation parenthesis i close parenthesis, summation parenthesis y close parenthesis of capital O subscript i-y T-O-T. Explanation: Calculated proportion of crashes of the crash type of interest for all T-O-T crashes bracket C-A-P parenthesis P close parenthesis subscript C-T forward slash T-O-T close bracket for all years of available data equals double summation across select sites, i, and analysis years, y, of observed number of T-O-T crashes bracket C-A-P-parenthesis O close parenthesis, close bracket of the given type of interest at site i during year y divided by double summation across select sites, i, and analysis years, y, of observed number of T-O-T crashes bracket C-A-P-parenthesis O close parenthesis, close bracket at site i during year y. (Eq 7)

Capital P subscript capital C-T forward slash F-I equals summation parenthesis i close parenthesis, summation parenthesis y close parenthesis of capital O subscript i-y capital C-T forward slash F-I divided by summation parenthesis i close parenthesis summation parenthesis y close parenthesis of capital O subscript i-y F-I. Explanation: Calculated proportion of crashes of the crash type of interest for all F-I crashes bracket C-A-P parenthesis P close parenthesis subscript C-T forward slash F-I close bracket for all years of available data equals double summation across select sites, i, and analysis years, y, of observed number of F-I crashes bracket C-A-P parenthesis O close parenthesis, close bracket of the given type of interest at site i during year y divided by double summation across select sites, i, and analysis years, y, of observed number of F-I crashes bracket C-A-P parenthesis O close parenthesis, close bracket at site i during year y. (Eq. 8)

The default distributions provided within ISAT for total and fatal and injury crashes are the same. This is not a requirement of the program. It is up to the user to determine whether or not providing the actual proportions for the two severity levels is desirable.

Crash Distributions for Mainline Freeways

The default crash distributions for mainline freeway segments provided within ISAT are located on the Distributions Mainline Freeways worksheet and are accessible only through the Go To Distributions for Mainline Freeways hyperlink on the Input-General worksheet. Table 35 illustrates the crash distributions for mainline freeways. For a given group of crash types (i.e., 13 crash types) and subtype, the sum of the proportions should equal 1.00.

Portions of the Distributions Mainline Freeways worksheet are protected. The user may only modify the default values for the mainline freeway segment subtypes (i.e., mainline freeway outside interchange area and mainline freeway within interchange area).

Crash Distributions for Ramps

The default crash distributions for ramps provided within ISAT are located on the Distributions Ramps worksheet and are accessible only through the Go To Distributions for Ramps hyperlink on the Input-General worksheet. Table 36 illustrates the crash distributions for ramps. For a given group of crash types (i.e., 13 crash types) and subtype, the sum of the proportions should equal 1.00.

Portions of the Distributions Ramps worksheet are protected. The user may only modify the default values for the ramp subtypes (i.e., diamond off-ramp, diamond on-ramp, parclo off-ramp, parlco on-ramp, free-flow off-ramp, free-flow on-ramp, and directional ramp).

Table 35. Proportions of crashes by severity and type for mainline freeways.
Crash No. Crash Type Number of Vehicles Involved Area Type Severity Level Mainline Freeway Outside Interchange Area Mainline Freeway Within Interchange Area
    (SV, MV) (R, U) (TOT, FI) numeric numeric
CPF1 CPF2 CPF3 CPF4 CPF5 CPF6 CPF7
1 Fixed object SV R TOT 0.176 0.266
2 Animal SV R TOT 0.206 0.018
3 Pedestrian SV R TOT 0.001 0.000
4 Bicyclist SV R TOT 0.000 0.000
5 Parked car SV R TOT 0.013 0.000
6 Overturn SV R TOT 0.295 0.108
7 Other single-vehicle SV R TOT 0.086 0.149
8 Rear-End MV R TOT 0.115 0.225
9 Head-On MV R TOT 0.006 0.006
10 Angle MV R TOT 0.012 0.020
11 Sideswipe, same direction MV R TOT 0.052 0.113
12 Sideswipe, opposite direction MV R TOT 0.004 0.004
13 Other multiple-vehicle MV R TOT 0.034 0.091
1 Fixed object SV U TOT 0.223 0.237
2 Animal SV U TOT 0.013 0.009
3 Pedestrian SV U TOT 0.009 0.003
4 Bicyclist SV U TOT 0.003 0.001
5 Parked car SV U TOT 0.018 0.012
6 Overturn SV U TOT 0.052 0.047
7 Other single-vehicle SV U TOT 0.016 0.021
8 Rear-End MV U TOT 0.410 0.462
9 Head-On MV U TOT 0.003 0.002
10 Angle MV U TOT 0.083 0.017
11 Sideswipe, same direction MV U TOT 0.113 0.138
12 Sideswipe, opposite direction MV U TOT 0.007 0.003
13 Other multiple-vehicle MV U TOT 0.050 0.048
1 Fixed object SV R FI 0.176 0.266
2 Animal SV R FI 0.206 0.018
3 Pedestrian SV R FI 0.001 0.000
4 Bicyclist SV R FI 0.000 0.000
5 Parked car SV R FI 0.013 0.000
6 Overturn SV R FI 0.295 0.108
7 Other single-vehicle SV R FI 0.086 0.149
8 Rear-End MV R FI 0.115 0.225
9 Head-On MV R FI 0.006 0.006
10 Angle MV R FI 0.012 0.020
11 Sideswipe, same direction MV R FI 0.052 0.113
12 Sideswipe, opposite direction MV R FI 0.004 0.004
13 Other multiple-vehicle MV R FI 0.034 0.091
Table 35. Proportions of crashes by severity and type for mainline freeways. (Continued)
Crash No. Crash Type Number of Vehicles Involved Area Type Severity Level Mainline Freeway Outside Interchange Area Mainline Freeway Within Interchange Area
   

(SV, MV)

(R, U)

(TOT, FI)

numeric

numeric

CPF1 CPF2 CPF3 CPF4 CPF5 CPF6 CPF7

1

Fixed object

SV

U

FI

0.223

0.237

2

Animal

SV

U

FI

0.013

0.009

3

Pedestrian

SV

U

FI

0.009

0.003

4

Bicyclist

SV

U

FI

0.003

0.001

5

Parked car

SV

U

FI

0.018

0.012

6

Overturn

SV

U

FI

0.052

0.047

7

Other single-vehicle

SV

U

FI

0.016

0.021

8 Rear-End MV U FI 0.410 0.462
9 Head-On MV U FI 0.003 0.002
10 Angle MV U FI 0.083 0.017
11 Sideswipe, same direction MV U FI 0.113 0.138
12 Sideswipe, opposite direction MV U FI 0.007 0.003
13 Other multiple-vehicle MV U FI 0.050 0.048

Crash Distributions for Crossroad Ramp Terminals and Intersections

The default crash distributions for crossroad ramp terminals and intersections provided within ISAT are located on the Distributions Ramp Terminals worksheet and are accessible only through the Go To Distributions for Ramp Terminals hyperlink on the Input-General worksheet. Table 37 illustrates the crash distributions for crossroad ramp terminals and intersections. For a given group of crash types (i.e., 13 crash types) and subtype, the sum of the proportions should equal 1.00.

Portions of the Distributions Ramp Terminals worksheet are protected. The user may only modify the default values for the crossroad ramp terminal and intersection subtypes (i.e., 3-leg STOP-controlled [3ST], 4-leg STOP-controlled [4ST], 3-leg signalized [3SG], and 4-leg signalized [4SG]).

Crash Distributions for Crossroad Segments

The default crash distributions for crossroad segments provided within ISAT are located on the Distributions Crossroad Segment worksheet and are accessible only through the Go To Distributions for Crossroad Segments hyperlink on the Input-General worksheet. Table 38 illustrates the crash distributions for crossroad segments. For a given group of crash types (i.e., 13 crash types) and subtype, the sum of the proportions should equal 1.00.

Portions of the Distributions Crossroad Segment worksheet are protected. The user may only modify the default values for the arterial crossroad roadway segment subtypes (i.e., 1-lane undivided, 2-lane undivided, 3-lane undivided, 2-lane divided, and 3-lane divided).

Table 36. Proportions of crashes by severity and type for ramps.
Crash No. Crash Type Number of Vehicles Involved Area Type Severity Level Diamond Off-Ramp Diamond On-Ramp Parclo Off-Ramp Parclo On-Ramp Free-Flow Off-Ramp Free-Flow On-Ramp Directional Ramp
   

(SV, MV)

(R, U)

(TOT, FI)

numeric

numeric

numeric

numeric

numeric

numeric

numeric

CPR1

CPR2

CPR3

CPR4

CPR5

CPR6

CPR7

CPR8

CPR9

CPR10

CPR11

CPR12

1 Fixed object SV R TOT 0.215 0.383 0.292 0.250 0.286 0.355 0.397
2 Animal SV R TOT 0.005 0.005 0.000 0.000 0.000 0.000 0.005
3 Pedestrian SV R TOT 0.000 0.000 0.000 0.000 0.000 0.000 0.000
4 Bicyclist SV R TOT 0.000 0.000 0.000 0.000 0.000 0.000 0.000
5 Parked car SV R TOT 0.000 0.000 0.000 0.000 0.000 0.000 0.000
6 Overturn SV R TOT 0.055 0.137 0.092 0.295 0.357 0.370 0.107
7 Other single-vehicle SV R TOT 0.103 0.178 0.092 0.136 0.143 0.171 0.184
8 Rear-End MV R TOT 0.476 0.150 0.432 0.091 0.143 0.026 0.179
9 Head-On MV R TOT 0.002 0.000 0.000 0.023 0.000 0.000 0.010
10 Angle MV R TOT 0.017 0.016 0.031 0.045 0.071 0.013 0.013
11 Sideswipe, same direction MV R TOT 0.074 0.068 0.015 0.023 0.000 0.039 0.054
12 Sideswipe, opposite direction MV R TOT 0.006 0.000 0.000 0.023 0.000 0.026 0.005
13 Other multiple-vehicle MV R TOT 0.047 0.063 0.046 0.114 0.000 0.000 0.046
1 Fixed object SV U TOT 0.150 0.206 0.280 0.225 0.250 0.292 0.294
2 Animal SV U TOT 0.001 0.000 0.001 0.000 0.000 0.000 0.000
3 Pedestrian SV U TOT 0.000 0.000 0.000 0.000 0.000 0.000 0.000
4 Bicyclist SV U TOT 0.000 0.000 0.000 0.001 0.000 0.000 0.000
5 Parked car SV U TOT 0.000 0.000 0.000 0.000 0.000 0.000 0.000
6 Overturn SV U TOT 0.016 0.036 0.063 0.057 0.040 0.051 0.033
7 Other single-vehicle SV U TOT 0.056 0.077 0.097 0.080 0.082 0.100 0.108
8 Rear-End MV U TOT 0.614 0.421 0.366 0.347 0.518 0.381 0.362
9 Head-On MV U TOT 0.002 0.003 0.016 0.007 0.004 0.003 0.002
10 Angle MV U TOT 0.009 0.020 0.018 0.039 0.010 0.015 0.014
11 Sideswipe, same direction MV U TOT 0.084 0.176 0.086 0.193 0.061 0.109 0.128
12 Sideswipe, opposite direction MV U TOT 0.003 0.011 0.012 0.008 0.003 0.005 0.004
13 Other multiple-vehicle MV U TOT 0.065 0.050 0.061 0.043 0.032 0.044 0.055
Table 36. Proportions of crashes by severity and type for ramps. (Continued)
Crash No. Crash Type Number of Vehicles Involved Area Type Severity Level Diamond Off-Ramp Diamond On-Ramp Parclo Off-Ramp Parclo On-Ramp Free-Flow Off-Ramp Free-Flow On-Ramp Directional Ramp
   

(SV, MV)

(R, U)

(TOT, FI)

numeric

numeric

numeric

numeric

numeric

numeric

numeric

CPR1 CPR2 CPR3 CPR4 CPR5 CPR6 CPR7 CPR8 CPR9 CPR10 CPR11 CPR12
1 Fixed object SV R FI 0.215 0.383 0.292 0.250 0.286 0.355 0.397
2 Animal SV R FI 0.005 0.005 0.000 0.000 0.000 0.000 0.005
3 Pedestrian SV R FI 0.000 0.000 0.000 0.000 0.000 0.000 0.000
4 Bicyclist SV R FI 0.000 0.000 0.000 0.000 0.000 0.000 0.000
5 Parked car SV R FI 0.000 0.000 0.000 0.000 0.000 0.000 0.000
6 Overturn SV R FI 0.055 0.137 0.092 0.295 0.357 0.370 0.107
7 Other single-vehicle SV R FI 0.103 0.178 0.092 0.136 0.143 0.171 0.184
8 Rear-End MV R FI 0.476 0.150 0.432 0.091 0.143 0.026 0.179
9 Head-On MV R FI 0.002 0.000 0.000 0.023 0.000 0.000 0.010
10 Angle MV R FI 0.017 0.016 0.031 0.045 0.071 0.013 0.013
11 Sideswipe, same direction MV R FI 0.074 0.068 0.015 0.023 0.000 0.039 0.054
12 Sideswipe, opposite direction MV R FI 0.006 0.000 0.000 0.023 0.000 0.026 0.005
13 Other multiple-vehicle MV R FI 0.047 0.063 0.046 0.114 0.000 0.000 0.046
1 Fixed object SV U FI 0.150 0.206 0.280 0.225 0.250 0.292 0.294
2 Animal SV U FI 0.001 0.000 0.001 0.000 0.000 0.000 0.000
3 Pedestrian SV U FI 0.000 0.000 0.000 0.000 0.000 0.000 0.000
4 Bicyclist SV U FI 0.000 0.000 0.000 0.001 0.000 0.000 0.000
5 Parked car SV U FI 0.000 0.000 0.000 0.000 0.000 0.000 0.000
6 Overturn SV U FI 0.016 0.036 0.063 0.057 0.040 0.051 0.033
7 Other single-vehicle SV U FI 0.056 0.077 0.097 0.080 0.082 0.100 0.108
8 Rear-End MV U FI 0.614 0.421 0.366 0.347 0.518 0.381 0.362
9 Head-On MV U FI 0.002 0.003 0.016 0.007 0.004 0.003 0.002
10 Angle MV U FI 0.009 0.020 0.018 0.039 0.010 0.015 0.014
11 Sideswipe, same direction MV U FI 0.084 0.176 0.086 0.193 0.061 0.109 0.128
12 Sideswipe, opposite direction MV U FI 0.003 0.011 0.012 0.008 0.003 0.005 0.004
13 Other multiple-vehicle MV U FI 0.065 0.050 0.061 0.043 0.032 0.044 0.055
Table 37. Proportions of crashes by severity and type for crossroad ramp terminals and intersections.
Crash No. Crash Type Number of Vehicles Involved Area Type Severity Level 3ST 4ST 3SG 4SG
   

(SV, MV)

(R, U)

(TOT, FI)

numeric

numeric

numeric

numeric

CPT1 CPT2 CPT3 CPT4 CPT5 CPT6 CPT7 CPT8 CPT9
1 Fixed object SV R TOT 0.126 0.049 0.095 0.024
2 Animal SV R TOT 0.021 0.009 0.000 0.005
3 Pedestrian SV R TOT 0.001 0.003 0.010 0.011
4 Bicyclist SV R TOT 0.003 0.005 0.010 0.009
5 Parked car SV R TOT 0.006 0.006 0.000 0.005
6 Overturn SV R TOT 0.060 0.027 0.029 0.007
7 Other single-vehicle SV R TOT 0.068 0.100 0.019 0.068
8 Rear-End MV R TOT 0.268 0.169 0.541 0.282
9 Head-On MV R TOT 0.015 0.014 0.000 0.016
10 Angle MV R TOT 0.299 0.472 0.229 0.422
11 Sideswipe, same direction MV R TOT 0.055 0.060 0.019 0.053
12 Sideswipe, opposite direction MV R TOT 0.014 0.013 0.010 0.010
13 Other multiple-vehicle MV R TOT 0.064 0.073 0.038 0.088
1 Fixed object SV U TOT 0.048 0.033 0.050 0.026
2 Animal SV U TOT 0.017 0.008 0.012 0.002
3 Pedestrian SV U TOT 0.011 0.009 0.008 0.013
4 Bicyclist SV U TOT 0.010 0.010 0.014 0.013
5 Parked car SV U TOT 0.003 0.002 0.001 0.002
6 Overturn SV U TOT 0.012 0.008 0.011 0.005
7 Other single-vehicle SV U TOT 0.043 0.074 0.021 0.043
8 Rear-End MV U TOT 0.355 0.234 0.557 0.471
9 Head-On MV U TOT 0.003 0.007 0.006 0.011
10 Angle MV U TOT 0.375 0.483 0.241 0.300
11 Sideswipe, same direction MV U TOT 0.039 0.041 0.026 0.041
12 Sideswipe, opposite direction MV U TOT 0.006 0.007 0.003 0.005
13 Other multiple-vehicle MV U TOT 0.078 0.084 0.050 0.068
1 Fixed object SV R FI 0.126 0.049 0.095 0.024
2 Animal SV R FI 0.021 0.009 0.000 0.005
3 Pedestrian SV R FI 0.001 0.003 0.010 0.011
4 Bicyclist SV R FI 0.003 0.005 0.010 0.009
5 Parked car SV R FI 0.006 0.006 0.000 0.005
6 Overturn SV R FI 0.060 0.027 0.029 0.007
7 Other single-vehicle SV R FI 0.068 0.100 0.019 0.068
8 Rear-End MV R FI 0.268 0.169 0.541 0.282
9 Head-On MV R FI 0.015 0.014 0.000 0.016
10 Angle MV R FI 0.299 0.472 0.229 0.422
11 Sideswipe, same direction MV R FI 0.055 0.060 0.019 0.053
12 Sideswipe, opposite direction MV R FI 0.014 0.013 0.010 0.010
13 Other multiple-vehicle MV R FI 0.064 0.073 0.038 0.088
Table 37. Proportions of crashes by severity and type for crossroad ramp terminals and intersections. (Continued)
Crash No. Crash Type Number of Vehicles Involved Area Type Severity Level 3ST 4ST 3SG 4SG
    (SV, MV) (R, U) (TOT, FI) numeric numeric numeric numeric
CPT1 CPT2 CPT3 CPT4 CPT5 CPT6 CPT7 CPT8 CPT9
1 Fixed object SV U FI 0.048 0.033 0.050 0.026
2 Animal SV U FI 0.017 0.008 0.012 0.002
3 Pedestrian SV U FI 0.011 0.009 0.008 0.013
4 Bicyclist SV U FI 0.010 0.010 0.014 0.013
5 Parked car SV U FI 0.003 0.002 0.001 0.002
6 Overturn SV U FI 0.012 0.008 0.011 0.005
7 Other single-vehicle SV U FI 0.043 0.074 0.021 0.043
8 Rear-End MV U FI 0.355 0.234 0.557 0.471
9 Head-On MV U FI 0.003 0.007 0.006 0.011
10 Angle MV U FI 0.375 0.483 0.241 0.300
11 Sideswipe, same direction MV U FI 0.039 0.041 0.026 0.041
12 Sideswipe, opposite direction MV U FI 0.006 0.007 0.003 0.005
13 Other multiple-vehicle MV U FI 0.078 0.084 0.050 0.068
Table 38. Proportions of crashes by severity and type for arterial crossroad roadway segments.
Crash No. Crash Type Number of Vehicles Involved Area Type Severity Level 1-Lane Undivided 2-Lane Undivided 3-Lane Undivided 2-Lane Divided 3-Lane Divided
    (SV, MV) (R, U) (TOT, FI) numeric numeric numeric numeric numeric
CPC1 CPC2 CPC3 CPC4 CPC5 CPC6 CPC7 CPC8 CPC9 CPC10
1 Fixed object SV R TOT 0.330 0.005 0.005 0.005 0.005
2 Animal SV R TOT 0.377 0.061 0.061 0.120 0.120
3 Pedestrian SV R TOT 0.001 0.004 0.004 0.004 0.004
4 Bicyclist SV R TOT 0.000 0.000 0.000 0.001 0.001
5 Parked car SV R TOT 0.000 0.000 0.000 0.001 0.001
6 Overturn SV R TOT 0.023 0.004 0.004 0.005 0.005
7 Other single-vehicle SV R TOT 0.043 0.161 0.161 0.296 0.296
8 Rear-End MV R TOT 0.096 0.343 0.343 0.271 0.271
9 Head-On MV R TOT 0.021 0.009 0.009 0.006 0.006
10 Angle MV R TOT 0.021 0.185 0.185 0.089 0.089
11 Sideswipe, same direction MV R TOT 0.014 0.039 0.039 0.043 0.043
12 Sideswipe, opposite direction MV R TOT 0.026 0.011 0.011 0.005 0.005
13 Other multiple-vehicle MV R TOT 0.048 0.178 0.178 0.154 0.154
1 Fixed object SV U TOT 0.185 0.075 0.075 0.155 0.155
2 Animal SV U TOT 0.173 0.003 0.003 0.129 0.129
3 Pedestrian SV U TOT 0.002 0.019 0.019 0.001 0.001
4 Bicyclist SV U TOT 0.000 0.013 0.013 0.000 0.000
5 Parked car SV U TOT 0.000 0.013 0.013 0.000 0.000
6 Overturn SV U TOT 0.007 0.010 0.010 0.013 0.013
7 Other single-vehicle SV U TOT 0.031 0.004 0.004 0.043 0.043
8 Rear-End MV U TOT 0.437 0.358 0.358 0.463 0.463
9 Head-On MV U TOT 0.022 0.007 0.007 0.007 0.007
10 Angle MV U TOT 0.046 0.359 0.359 0.056 0.056
11 Sideswipe, same direction MV U TOT 0.027 0.102 0.102 0.075 0.075
12 Sideswipe, opposite direction MV U TOT 0.023 0.012 0.012 0.015 0.015
13 Other multiple-vehicle MV U TOT 0.047 0.025 0.025 0.043 0.043
1 Fixed object SV R FI 0.330 0.005 0.005 0.005 0.005
2 Animal SV R FI 0.377 0.061 0.061 0.120 0.120
3 Pedestrian SV R FI 0.001 0.004 0.004 0.004 0.004
4 Bicyclist SV R FI 0.000 0.000 0.000 0.001 0.001
5 Parked car SV R FI 0.000 0.000 0.000 0.001 0.001
6 Overturn SV R FI 0.023 0.004 0.004 0.005 0.005
7 Other single-vehicle SV R FI 0.043 0.161 0.161 0.296 0.296
8 Rear-End MV R FI 0.096 0.343 0.343 0.271 0.271
9 Head-On MV R FI 0.021 0.009 0.009 0.006 0.006
10 Angle MV R FI 0.021 0.185 0.185 0.089 0.089
11 Sideswipe, same direction MV R FI 0.014 0.039 0.039 0.043 0.043
12 Sideswipe, opposite direction MV R FI 0.026 0.011 0.011 0.005 0.005
13 Other multiple-vehicle MV R FI 0.048 0.178 0.178 0.154 0.154
Table 38. Proportions of crashes by severity and type for arterial crossroad roadway segments. (Continued)
Crash No. Crash Type Number of Vehicles Involved Area Type Severity Level 1-Lane Undivided 2-Lane Undivided 3-Lane Undivided 2-Lane Divided 3-Lane Divided
    (SV, MV) (R, U) (TOT, FI) numeric numeric numeric numeric numeric
CPC1 CPC2 CPC3 CPC4 CPC5 CPC6 CPC7 CPC8 CPC9 CPC10
1 Fixed object SV U FI 0.185 0.075 0.075 0.155 0.155
2 Animal SV U FI 0.173 0.003 0.003 0.129 0.129
3 Pedestrian SV U FI 0.002 0.019 0.019 0.001 0.001
4 Bicyclist SV U FI 0.000 0.013 0.013 0.000 0.000
5 Parked car SV U FI 0.000 0.013 0.013 0.000 0.000
6 Overturn SV U FI 0.007 0.010 0.010 0.013 0.013
7 Other single-vehicle SV U FI 0.031 0.004 0.004 0.043 0.043
8 Rear-End MV U FI 0.437 0.358 0.358 0.463 0.463
9 Head-On MV U FI 0.022 0.007 0.007 0.007 0.007
10 Angle MV U FI 0.046 0.359 0.359 0.056 0.056
11 Sideswipe, same direction MV U FI 0.027 0.102 0.102 0.075 0.075
12 Sideswipe, opposite direction MV U FI 0.023 0.012 0.012 0.015 0.015
13 Other multiple-vehicle MV U FI 0.047 0.025 0.025 0.043 0.043

Previous | Table of Contents | Next

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101