Skip to contentU.S. Department of Transportation/Federal Highway Administration
Asset Management | Bridge Technology | Operations | Pavement

Asset Management

FHWA > Asset Management > HIF-10-015 > 4.0 Analysis and Evaluation of Alternative Condition Indicators
<< Previous Contents Next >>

Evaluation of Highway Performance Measures for a Multi-Study Corridor - A Pilot Study

4.0 Analysis and Evaluation of Alternative Condition Indicators

This section contains the analysis of condition measure data, both bridge and pavement, received from the participating states. The analysis begins with background on the traditional measures used to evaluate these assets before presenting a statistical breakdown of the specific data captured for this project. Condition is one measure of performance. Other performance measures may relate to functional adequacy or overall importance of a given asset.

Following the analysis, CS presents information obtained from the participating states on the processes used to capture bridge and pavement condition information, including methods employed to validate the quality of the data. This section also includes feedback on the state-specific criteria used to categorize asset condition (e.g., as "good," "fair," or "poor").

Finally, CS presents conclusions on the adequacy of the existing measures for bridges and pavement, highlighting some of the results of the data analysis. CS also makes recommendations on alternative indicators and discusses the issues related to adopting different measures.

4.1 Data Analysis

Section 3 discusses the bridge and pavement data that were collected from the states participating in this study. Due to time and budget constraints as well as the fact that Maryland provided data only for Interstate 95, CS generally restricted its analysis to assets present on I-95.

Review of Statistical Functions

This section briefly reviews some of the statistical functions used in the data analysis.

Correlation coefficient indicates the correlation or linear dependence between two variables. The correlation coefficient is defined as the covariance of the variables divided by the product of their standard deviations. It yields a value between -1 and +1. If the correlation coefficient is equal to 0, it means that there is no linear correlation between the variables. A value of +1 or -1 means that there is a linear equation that describes the relationship between the variables. The correlation coefficient is positive if the variables simultaneously are greater than or less than their respective means. Conversely, the correlation coefficient is negative if the variables lie on opposite sides of their respective means.

For example, the correlation coefficient for health index and sufficiency rating describes the linear dependence between these two variables. If the variables are correlated (i.e., their correlation coefficient is high), a predictive relationship between them can be inferred. On the other hand, if the correlation coefficient is small, no linear relationship exists.

Probability distribution describes the range of possible values that a variable can attain and the probability that the value of the variable is within any subset of that range. Probability distributions are computed using the average and standard deviation for a series of data points. The distributions in this section show possible values for measures such as sufficiency rating, health index, and IRI as well as the probability of certain values occurring. It is important to note that the probability distribution function assumes a normal (i.e., "bell shaped") distribution and does not take into account the fact that sufficiency rating and health index cannot exceed 100.

Cumulative distribution is another method of graphically representing the range of values that a variable can attain. Cumulative distributions also are computed using the average and the standard deviation. However, cumulative distributions run from 0 to 1 and, for any given value, represent the percentage of the total population having that value or less. This calculation also assumes a normal distribution of values and does not reflect the maximum value of 100 for sufficiency rating and health index.

Bridge Analysis

For the bridge data analysis, CS reviewed:

  • Delaware data - 62 bridges with element data obtained from Pontis;
  • Maryland data - 118 bridges with element data obtained from the Maryland bridge management system and 118 bridges without element data obtained from the Maryland National Bridge Inventory (NBI) file; and
  • Virginia data - 385 bridges with element data obtained from Pontis.

CS did review a recent Pontis database for Washington, D.C. However, this database did not contain any structures on I-95. Therefore, bridges for the District of Columbia were not included in this analysis.

The bridge analysis focused on the following key data items, which are available from the NBI or may be calculated using Pontis:

  • Deck Rating (NBI Element 58);
  • Superstructure Rating (NBI Element 59);
  • Substructure Rating (NBI Element 60);
  • Sufficiency Rating (calculated);
  • Health Index for deck elements only (calculated);
  • Health Index for superstructure elements only (calculated);
  • Health Index for substructure elements only (calculated); and
  • Health Index for all elements.

Historically, the sufficiency rating and structurally deficient/functionally obsolete (SD/FO) values have been used as the measure of bridge condition. The development in recent years of the health index in Pontis provides a different measure of the structural condition of a bridge. It is important to note that health index does not attempt to measure the functional adequacy of a bridge.

Discussion of Current Measures

Although states are able to calculate sufficiency rating for themselves, the official calculation is performed by FHWA using NBI data submitted annually by states. The sufficiency rating uses four separate factors to obtain a numeric value that indicates whether a bridge is sufficient to remain in service. The result is a percentage in which 100 percent represents an entirely sufficient bridge and zero percent represents an entirely insufficient or deficient bridge. The sufficiency rating is never less than 0 or more than 100.

A bridge's sufficiency rating affects its eligibility for Federal funding for maintenance, rehabilitation, or replacement activities. For bridges to qualify for Federal replacement funds, they must have a rating of 50 or less. To qualify for Federal rehabilitation funding, a bridge must have a sufficiency rating of 80 or less.

The sufficiency rating factors are:

  1. S1, the structural adequacy and safety factor;
  2. S2, the serviceability and functional obsolescence factor;
  3. S3, the essentiality for public use factor; and
  4. S4, the special reductions factor.

S1 is a function of the lowest rating code of Item 59 (Superstructure Rating), Item 60 (Substructure Rating) or Item 62 (Culvert Rating), and Item 66 (Inventory Rating).

S2 is a function of Item 58 (Deck Rating), Item 67 (Structural Evaluation), Item 68 (Deck Geometry), Item 69 (Underclearances), Item 71 (Waterway Adequacy), Item 72 (Approach and Road Alignment), Item 29 (ADT), Item 51 (Bridge Roadway Width), Item 28 (Lanes on the Structure), Item 100 (Defense Highway), Item 32 (Approach Roadway Width), Item 43 (Structure Type), and Item 53 (Vertical Clearance Over Deck).

S3 is a function of S1, S2, Item 29 (ADT), Item 100 (Defense Highway), and Item 19 (Detour Length).

To obtain the sufficiency rating, S4 is subtracted from the sum of S1, S2, and S3. S4 is only used when the sum of S1, S2, and S3 is greater or equal to 50. S4 is a function of Item 19 (Detour Length), Item 36 (Traffic Safety Features), and Item 43 (Structure Type).

Similar to sufficiency rating, health index provides a single numeric indicator of the structural health of the bridge. This indicator is expressed as a percentage from zero, which corresponds to the worst possible condition, to 100, which is the best condition. Health index is a function of the fractional distribution of the bridge element quantities across the range of their applicable condition states. Concretely, the health index value of an entire bridge is calculated as a weighted average of the health indexes of its elements, where elements are weighted by their total quantity and relative importance (i.e., failure cost). Consequently, the bridge health index is a function of the failure cost of each element of the bridge, quantity of each element on the bridge, and condition state of each element. Likewise, the health index of each element is a function of its unit failure cost, its quantity on the bridge, and its condition state on the bridge.

Though the sufficiency rating provides a consistent standard for the evaluation of sufficiency to remain in service, it is not comprehensive enough to provide performance-based information on each element of the bridge. For example:

  • Sufficiency rating focuses on the overall condition of the bridge, making it irrelevant for maintenance decision-making.
  • Sufficiency rating emphasizes the functionality and geometric characteristics of the bridge rather than an element-based view of the bridge. (Factor S2)
  • Sufficiency rating assumes that the worst distress between the superstructure, substructure, or culvert will represent the overall structural adequacy and safety of the bridge (Factor S1). This means sufficiency rating cannot account for localized problems or safety issues.
  • The superstructure, substructure, and culvert ratings are on a 0-9 scale by severity of deterioration. These ratings do not give a picture of the deterioration process at work or the extent of deterioration.
  • Sufficiency rating focuses only on the major parts of the bridge: superstructure, substructure, deck, and culverts. The lack of detail makes it impossible to estimate the cost of rehabilitation or maintenance.

On the other hand, the health index indicates the condition of each element at a given time. The bridge health index aggregates element-level health indexes where weights, which embody the economic consequences of failure, are assigned to each element. The main advantage of the health index is that it links the condition of bridge to the resources allocated. Consequently, decision-makers can evaluate the impact of several resource allocation scenarios on the future condition of a bridge network. Also, maintenance and rehabilitation decision-making is facilitated since the measures are detailed enough to represent localized problems.

Analysis of Current Measures

The first part of the analysis of bridge data involved the calculation of basic statistical information by state and across all states. These basic statistics, not weighted by bridge deck area, are presented in Table 4.1.

Table 4.1 Basic Bridge Statistics - Not Weighted
State Statistic Deck Rating Superstructure Rating Substructure Rating Sufficiency Rating Health Index - Deck Health Index - Superstructure Health Index - Substructure Health Index - Overall
DE Count 45 45 45 61 44 45 61 61
DE Min 5 5 4 53 50 61 67.5 67.5
DE Max 8 8 8 98 100 100 100 100
DE Average 6.56 6.40 6.33 84.47 89.14 88.63 94.87 92.44
DE Median 7 6 6 85 99.85 92.5 97.4 93.8
DE Std Dev 0.66 0.65 0.77 10.46 14.65 10.70 7.46 6.44
DE Variance 0.43 0.43 0.59 109.36 214.64 114.44 55.62 41.44
MD Count 107 108 108 140 98 108 118 118
MD Min 5 5 5 50.2 25 67 46.2 49.3
MD Max 8 9 9 100 100 100 100 100
MD Average 6.64 6.36 6.19 81.20 84.44 96.63 90.18 88.96
MD Median 7 6 6 83.60 75 99.7 94.35 93.05
MD Std Dev 0.62 0.81 0.78 10.10 17.03 6.56 11.39 11.51
MD Variance 0.38 0.66 0.61 102.02 290.15 43.01 129.78 132.41
VA Count 223 223 223 373 220 222 370 372
VA Min 4 4 4 25.3 25 48.3 12.8 24.4
VA Max 9 9 9 100 100 100 100 100
VA Average 6.57 6.57 6.28 81.97 87.54 92.39 89.77 89.69
VA Median 7 7 6 83 100 97.8 95.3 94.2
VA Std Dev 0.94 1.08 1.07 12.52 15.13 11.38 12.75 11.56
VA Variance 0.88 1.16 1.14 156.78 228.79 129.59 162.65 133.53
ALL Count 375 376 376 574 362 375 549 551
ALL Min 4 4 4 25.3 25 48.3 12.8 24.4
ALL Max 9 9 9 100 100 100 100 100
ALL Average 6.59 6.49 6.26 82.05 86.89 93.16 90.42 89.84
ALL Median 7 6 6 83.00 100 98 95.5 94
ALL Std Dev 0.83 0.97 0.96 11.78 15.65 10.42 12.08 11.12
ALL Variance 0.68 0.93 0.92 138.79 244.83 108.64 145.82 123.69

The following are some observations regarding the basic statistics:

  • Counts only include assets where the measure is not missing. Differences in the counts can be attributed primarily to the absence of certain types of elements. For example, culverts generally do not have deck elements and, therefore, would not have either a deck rating or deck health index.
  • The low standard deviation and variance for the NBI ratings compared to the equivalent health index measures can be attributed to the small number of values the NBI rating can hold. For these statistics, only ratings from zero to nine were considered.
  • The combined statistics for all states do not vary dramatically from the statistics for any individual state. The ratings and the component-based health indexes for Virginia vary more than in the other states. There is no reason to attribute this to anything other than actual variability in the condition of the corresponding structures.

Table 4.2 presents some of the same basic statistics, except that values have been weighted by bridge deck area. Deck area is one of the primary metrics by which bridges generally are categorized. Only the statistics necessary to support calculations of probably and cumulative distributions were calculated.

Table 4.2 Basic Bridge Statistics - Weighted
State Statistic Deck Rating Superstructure Rating Substructure Rating Sufficiency Rating Health Index - Deck Health Index - Superstructure Health Index - Substructure Health Index - Overall
DE Count 45 45 45 61 44 45 61 61
DE Average 6.20 6.03 6.63 78.45 78.87 94.91 95.31 88.96
DE Std Dev 0.62 0.68 0.75 7.81 11.93 8.36 8.14 5.05
DE Variance 0.38 0.46 0.57 61.07 142.31 69.89 66.32 25.55
MD Count 107 108 108 113 98 103 103 103
MD Average 6.61 6.40 6.14 79.66 82.36 95.70 92.23 90.55
MD Std Dev 0.53 0.65 0.69 8.55 17.99 7.56 9.53 9.76
MD Variance 0.28 0.42 0.48 73.17 323.57 57.14 90.80 95.19
VA Count 213 213 213 215 211 211 212 212
VA Average 6.81 6.73 6.62 81.47 88.42 93.79 89.94 90.58
VA Std Dev 1.12 1.38 1.30 12.82 14.63 10.26 12.14 9.89
VA Variance 1.25 1.91 1.70 164.34 213.93 105.29 147.42 97.78
ALL Count 365 366 366 389 353 359 376 376
ALL Average 6.61 6.46 6.39 80.21 84.75 94.48 91.81 90.16
ALL Std Dev 0.80 0.98 0.96 10.37 14.76 9.25 10.68 8.85
ALL Variance 0.64 0.96 0.93 107.49 217.85 85.57 114.03 78.35

The following are some observations regarding the weighted statistics:

  • The count of assets goes down in many cases because bridges without a deck area (e.g., culverts) are removed from the results.
  • Some of the other statistics (e.g., average, standard deviation, and variance) are reduced when weighted by deck area. This reflects both the smaller set of assets being considered as well as a general trend observed in other states that health index weighted by deck area is lower than the average health index.(1)
  • Despite the differences observed, the statistics generally are very close for bridges that have and have not been weighted by deck area. For the data covered by this analysis, this indicates that the values are well distributed across the spectrum of possible results.

For the second phase of the analysis, CS measured the correlation between different measures. The closer the correlation coefficient is to 1 or -1, the greater the statistical relationship between two sets of values. It is important to note that the correlation coefficient does not make any determination regarding accuracy of the values nor does it infer any cause/effect relationship between the values. Table 4.3 presents the correlation coefficients for bridge measures within individual states and across all states.

Table 4.3 Bridge Correlation Coefficients
State Statistic Superstructure Rating Substructure Rating Sufficiency Rating Health Index - Deck Health Index - Superstructure Health Index - Substructure Health Index - Overall
DE Deck Rating 0.05 0.39 0.47 0.62 0.08 0.23 0.60
DE Superstructure Rating Not Applicable 0.09 0.13 0.05 0.43 -0.03 -0.03
DE Substructure Rating Not Applicable Not Applicable 0.51 0.20 -0.12 0.67 0.37
DE Sufficiency Rating Not Applicable Not Applicable Not Applicable 0.37 0.32 0.38 0.25
DE Health Index - Deck Not Applicable Not Applicable Not Applicable Not Applicable 0.12 0.20 0.60
DE Health Index - Superstructure Not Applicable Not Applicable Not Applicable Not Applicable Not Applicable -0.11 0.63
DE Health Index - Substructure Not Applicable Not Applicable Not Applicable Not Applicable Not Applicable Not Applicable 0.67
MD Deck Rating 0.51 0.38 0.16 0.48 0.18 -0.10 0.09
MD Superstructure Rating Not Applicable 0.58 0.32 -0.03 -0.03 0.00 0.04
MD Substructure Rating Not Applicable Not Applicable 0.52 0.17 -0.14 0.02 0.45
MD Sufficiency Rating Not Applicable Not Applicable Not Applicable 0.13 -0.19 0.04 0.43
MD Health Index - Deck Not Applicable Not Applicable Not Applicable Not Applicable 0.26 0.31 0.36
MD Health Index - Superstructure Not Applicable Not Applicable Not Applicable Not Applicable Not Applicable 0.27 0.33
MD Health Index - Substructure Not Applicable Not Applicable Not Applicable Not Applicable Not Applicable Not Applicable 0.99
VA Deck Rating 0.74 0.70 0.40 0.59 0.54 0.49 0.35
VA Superstructure Rating Not Applicable 0.69 0.47 0.47 0.52 0.36 0.25
VA Substructure Rating Not Applicable Not Applicable 0.41 0.47 0.47 0.62 0.28
VA Sufficiency Rating Not Applicable Not Applicable Not Applicable 0.67 0.68 0.64 0.29
VA Health Index - Deck Not Applicable Not Applicable Not Applicable Not Applicable 0.38 0.32 0.74
VA Health Index - Superstructure Not Applicable Not Applicable Not Applicable Not Applicable Not Applicable 0.39 0.77
VA Health Index - Substructure Not Applicable Not Applicable Not Applicable Not Applicable Not Applicable Not Applicable 0.84
ALL Deck Rating 0.65 0.62 0.36 0.58 0.45 0.43 0.36
ALL Superstructure Rating Not Applicable 0.63 0.41 0.38 0.46 0.29 0.18
ALL Substructure Rating Not Applicable Not Applicable 0.44 0.43 0.38 0.60 0.31
ALL Sufficiency Rating Not Applicable Not Applicable Not Applicable 0.62 0.60 0.60 0.30
ALL Health Index - Deck Not Applicable Not Applicable Not Applicable Not Applicable 0.28 0.30 0.59
ALL Health Index - Superstructure Not Applicable Not Applicable Not Applicable Not Applicable Not Applicable 0.30 0.62
ALL Health Index - Substructure Not Applicable Not Applicable Not Applicable Not Applicable Not Applicable Not Applicable 0.86

The following are some observations regarding the correlation coefficients:

  • There are no firm rules regarding how close a coefficient must be to 1 or 1 in order to be considered significant. For purposes of this analysis, CS has chosen 0.70 as a reasonable threshold to identify values that are well correlated.
  • Theoretically, both sufficiency rating and overall health index represent the structural condition of a bridge. If these measures were equally good for this purpose, we would expect a relatively high correlation between these values. However, as shown in Table 4.3, the correlation coefficients are DE=0.25, MD=0.43, VA=0.29, and ALL=0.30. At least part of this difference can be explained by the fact that the sufficiency rating is a combination of structural adequacy (55 percent), serviceability (30 percent), and essentiality (15 percent).(2)
  • The same reasoning applies to Deck Rating versus Health Index - Deck, Superstructure Rating versus Health Index - Superstructure, and Substructure Rating versus Health Index - Substructure. Some of these component values are significantly better correlated than the overall measures. For example, in Delaware the correlations of the deck and substructure measures are 0.62 and 0.67, respectively. These relatively high correlations indicate that the NBI ratings can be reasonable component-based measures of condition once the nonstructural aspects of the sufficiency rating calculation are removed.
  • However, some component-based correlations are much worse than the overall measures. For example, in Maryland, the superstructure and substructure measures are -0.03 and 0.02, respectively. This may indicate a systemic problem with the way in which inspectors capture either element-level data or NBI ratings. CS is aware of similar issues in other states, who report that the FHWA NBI Translator built into Pontis, which calculates NBI ratings from element data, is giving significantly different values than the NBI ratings entered by inspectors.
  • The highlighted values in Table 4.3 are associated with correlations of the Health Index to subcomponents of the Health Index or one NBI rating with another. They are not associated with correlations of Health Index and NBI values. The highest correlations between Health Index and NBI are found for the substructure ratings in Delaware and Virginia.

In addition to computing the correlation coefficients, CS also prepared probability and cumulative distributions and associated graphs. A large number of graphs were produced and representative samples are shown below. Figure 4.1 shows probability distributions across all states for Health Index - Overall, Health Index - Deck, Sufficiency Rating, and Deck Rating. The graphs for other measures and individual states generally are not significantly different.

The similarity between Health Index - Overall and Sufficiency Rating graphs reflects the modest differences in the mean and variance for these measures. The exception is Health Index - Overall for Delaware. A significantly lower variance resulted in a narrower distribution. This trend is demonstrated by the extremely sharp curve for Deck Rating. This graph, which repeats for the other NBI ratings, is driven by the small number of values that an individual rating can assume and the correspondingly lower variance for this measure.

Figure 4.1 Bridge Probability Distributions - All States
This figure contains four graphs that show probability distributions for bridge data. These graphs summarize non-weighted data for all states in this study. The graph in the upper left corner shows the distribution for Health Index for deck elements only. The distribution is a normal curve centered on the average of 86.89 with a standard deviation of 15.65. The X-axis runs from 0 to 200 and the Y-axis from 0 to 0.03. The maximum value of the curve is approximately 0.026. The graph in the upper right corner shows the distribution for Deck Rating. The distribution is a normal curve centered on the average of 6.59 with a standard deviation of 0.83. The X-axis runs from 0 to 200 and the Y-axis from 0 to 0.45. The maximum value of the curve is approximately 0.425. The graph in the lower left corner shows the distribution for Health Index for all elements. The distribution is a normal curve centered on the average of 89.84 with a standard deviation of 11.12. The X-axis runs from 0 to 200 and the Y-axis from 0 to 0.04. The maximum value of the curve is approximately 0.036. The graph in the lower right corner shows the distribution for Sufficiency Rating. The distribution is a normal curve centered on the average of 82.05 with a standard deviation of 11.78. The X-axis runs from 0 to 200 and the Y-axis from 0 to 0.04. The maximum value of the curve is approximately 0.034.

Figure 4.2 shows cumulative distributions across all states for Health Index - Overall, Health Index - Deck, Sufficiency Rating, and Deck Rating. The graphs for other measures and individual states generally are not significantly different. Again, the sharp curve for Deck Rating reflects the small number of values and correspondingly small variance for this measure.

Figure 4.2 Bridge Cumulative Distributions - All States
his figure contains four graphs that show cumulative distributions for bridge data. These graphs summarize non-weighted data for all states in this study. The graph in the upper left corner shows the distribution for Health Index for deck elements only. The distribution is an S-curve with an inflection point on the average of 86.89. The X-axis runs from 0 to 200 and the Y-axis from 0 to 1.2. The minimum value of the curve is 0 and maximum value of the curve is 1. The graph in the upper right corner shows the distribution for Deck Rating. The distribution is an S-curve with an inflection point on the average of 6.59. The X-axis runs from 0 to 200 and the Y-axis from 0 to 1.2. The minimum value of the curve is 0 and maximum value of the curve is 1. The graph in the lower left corner shows the distribution for Health Index for all elements. The distribution is an S-curve with an inflection point on the average of 89.84. The X-axis runs from 0 to 200 and the Y-axis from 0 to 1.2. The minimum value of the curve is 0 and maximum value of the curve is 1. The graph in the lower right corner shows the distribution for Sufficiency Rating. The distribution is an S-curve with an inflection point on the average of 82.05. The X-axis runs from 0 to 200 and the Y-axis from 0 to 1.2. The minimum value of the curve is 0 and maximum value of the curve is 1.

Figures 4.1 and 4.2 show values not weighted by deck area. Figure 4.3 shows Health Index - Overall and Sufficiency Rating comparing values that are weighted and not weighted.

Figure 4.3 Weighted Bridge Distributions - All States
his figure contains four graphs that show probability and cumulative distributions for bridge data. These graphs compare non-weighted and weighted data for all states in this study. The graph in the upper left corner shows the probability distribution for Sufficiency Rating. The distributions are normal curves. The weighted distribution is centered on the average of 80.21 and the non-weighted distribution is centered on the average 82.05. The X-axis runs from 0 to 200 and the Y-axis from 0 to 0.045. The maximum value of the weighted curve is approximately 0.039 and the maximum value of the non-weighted curve is approximately 0.034. The graph in the upper right corner shows the cumulative distribution for Sufficiency Rating. The distributions are S-curves. The weighted distribution has an inflection point on the average of 80.21 and the non-weighted distribution has an inflection point on the average of 82.05. The X-axis runs from 0 to 200 and the Y-axis from 0 to 1.2. The minimum value of both curves is 0 and maximum value of both curves is 1. The graph in the lower left corner shows the probability distribution for Health Index for all elements. The distributions are normal curves. The weighted distribution is centered on the average of 90.16 and the non-weighted distribution is centered on the average of 89.84. The X-axis runs from 0 to 200 and the Y-axis from 0 to 0.50. The maximum value of the weighted curve is approximately 0.045 and the maximum value of the non-weighted curve is approximately 0.036. The graph in the lower right corner shows the cumulative distribution for Health Index for all elements. The distributions are S-curves. The weighted distribution has an inflection point on the average of 90.16 and the non-weighted distribution has an inflection point on the average of 89.84. The X-axis runs from 0 to 200 and the Y-axis from 0 to 1.2. The minimum value of both curves is 0 and maximum value of both curves is 1.

Generally, the weighted measures are sharper (i.e., the probability distributions are narrower and the peaks higher). This reflects the fact that fewer assets are included in the calculations because bridges without deck areas are not included. Otherwise, the curves are very similar, which reinforces conclusions associated with Table 4.2.

Pavement Analysis

For the pavement data analysis, CS reviewed:

  • Delaware data - 193 segments on I-95 totaling approximately 37 miles;
  • Maryland data - 2,179 segments on I-95 totaling approximately 217 miles; and
  • Virginia data - 192 segments on I-95 totaling approximately 371 miles.

Note that these mileage totals generally do not match the official number of I-95 miles by state.(3) The totals reported above were calculated by summing the difference of beginning and ending mile points for each segment. For Maryland and Virginia, the data include separate records for northbound and southbound roadways. The data for Delaware appear to cover one direction only. Also, although CS originally believed that Maryland did not provide data for the section of I-95 that lies with the city limits of Baltimore, a more thorough review of the data determined that these records were present but there was a mismatch between county name and county code. CS corrected these data and matched the records to the correct road section.

The pavement analysis focused on the following key data elements:

  • International Roughness Index (IRI), which was provided by all states;
  • Overall Pavement Condition (OPC), which was reported by Delaware and calculated for Virginia by CS;
  • Critical Condition Index (CCI), which was reported by Virginia; and
  • Other distress index data (e.g., cracking, rutting, etc.) as appropriate.

Historically, IRI has been used as the measure of pavement condition while Present Serviceability Rating (PSR) has measured the ability of the pavement to service expected traffic. These values are provided by states as part of their annual Highway Performance Monitoring System (HPMS) report. Changes to the HPMS reporting process are part of FHWA's HPMS Reassessment 2010+ initiative. These changes include improving the consistency IRI measurement and reporting as well as submitting more data elements (e.g., rutting, faulting, cracking, overlay information).

Discussion of Current Measures

Pavement roughness is defined in accordance with the American Society for Testing and Materials (ASTM) standard E867 as "the deviation of a surface from a true planar surface with characteristic dimensions that affect vehicle dynamics and ride quality." IRI was chosen by the World Bank in the 1980s to quantify roughness. After a detailed study of various methodologies and road profiling statistics, IRI was chosen as the HPMS standard reference roughness index. The HPMS data reporting unit for IRI is meters/kilometer (inches/mile).

IRI is the amount of roughness in a measured longitudinal profile. Lower values for IRI indicate smoother pavement. IRI is based on average rectified slope (ARS), which is a filtered ratio of a standard vehicle's accumulated suspension motion (e.g., millimeters or inches) divided by the distance traveled by the vehicle during the measurement. IRI is equal to ARS multiplied by 1,000. IRI is computed from a single longitudinal profile measured with a road profiler in both the inside and outside wheel paths of the pavement. The average of these two IRI measurements is reported as the roughness of the pavement section.

However, IRI only captures road smoothness. Some states use other indicators, such as OPC or CCI, to describe the general health of the pavement. Indeed, the pavement may be very smooth and yet have deep rutting in the wheel path or cracking that allows water to enter and cause deterioration. OPC and CCI both are composite values that combine several distress ratings to produce an overall pavement condition measure.

Delaware uses the following process to calculate OPC for asphalt pavement:

  • Convert five distress measures into numeric indexes using the tables shown in Figure 4.4;

Figure 4.4 Delaware Pavement Distress Conversation Tables
TThis figure contains five Delaware distress conversion tables. These tables convert severity and extent information for five types of distress into numeric values, which are used by Delaware in the calculation of Overall Pavement Condition (OPC) for a road segment. The five types of distress are Asphalt Patching, Surface Defects, Fatigue Cracking, Block Cracking and Transverse Cracking. Each conversion table categorizes the distress severity (on the Y-Axis) and distress extent (on the X-Axis). Both severity and extent may be high, medium or low. Each cell equates to a number between 0 and 99. The value of 100 is used when the distress is not present for a given road segment.

  • Calculate the average (avg) and the standard deviation (stdev) of the five numeric indexes; and
  • Calculate OPC using the formula OPC = avg - (1.25 * stdev).

Generally, OPC is a number between 0 and 100. In some cases, OPC is divided by 20 and reported as a number between 0 and 5.

Virginia uses the following process(4) to calculate CCI for asphalt pavement:

  • Calculate a load distress index (LDR) to describe distresses related to wheel loads (e.g., alligator cracking, delaminations, patching, potholes and rutting);
  • Calculate a nonload distress index (NDR) to describe distresses related to weathering (e.g., bleeding, block cracking, linear cracking and reflection cracking); and
  • Define CCI as the lower of the LDR and NDR index values.

Both LDR and NDR start at a base value of 100. Points are deducted based on the severity and frequency of occurrence of each distress. Some distresses are classified as more detrimental to pavement and are weighted more heavily. The deductions are based on the deduct curves in the PAVER pavement management system. The specifics of these calculations are beyond the scope of this document but are available from the Virginia Department of Transportation. Like OPC, CCI is a number between 0 and 100. Note that IRI is not one of the inputs into the CCI calculation.

Analysis of Current Measures

As with bridges, the first part of the pavement analysis involved the calculation of basic statistical information by state and across all states. These basic statistics, not weighted by pavement segment length, are presented in Tables 4.4 and 4.5.

Table 4.4 Basic Pavement Statistics Part 1 - Not Weighted
State Statistic IRI - Left IRI - Right IRI - Average Transverse Cracks - Severity 1 (linear feet) Transverse Cracks - Severity 2 (linear feet) Longitudinal Cracks - Severity 1 (linear feet) Longitudinal Cracks - Severity 2 (linear feet) Alligator Cracks - Severity 1 (square feet) Alligator Cracks - Severity 2 (square feet) Alligator Cracks - Severity 3 (square feet)
DE Count 91 91 91 N/A N/A N/A N/A N/A N/A N/A
DE Min 49 49 49 N/A N/A N/A N/A N/A N/A N/A
DE Max 368 354 358 N/A N/A N/A N/A N/A N/A N/A
DE Average 144 156 151 N/A N/A N/A N/A N/A N/A N/A
DE Median 122 135 138 N/A N/A N/A N/A N/A N/A N/A
DE Std Dev 76 80 74 N/A N/A N/A N/A N/A N/A N/A
DE Variance 5731 6334 5458 N/A N/A N/A N/A N/A N/A N/A
MD Count N/A N/A 2179 N/A N/A N/A N/A N/A N/A N/A
MD Min N/A N/A 31 N/A N/A N/A N/A N/A N/A N/A
MD Max N/A N/A 482 N/A N/A N/A N/A N/A N/A N/A
MD Average N/A N/A 83 N/A N/A N/A N/A N/A N/A N/A
MD Median N/A N/A 67 N/A N/A N/A N/A N/A N/A N/A
MD Std Dev N/A N/A 47 N/A N/A N/A N/A N/A N/A N/A
MD Variance N/A N/A 2182 N/A N/A N/A N/A N/A N/A N/A
VA Count 192 192 192 192 192 192 192 192 192 192
VA Min 44 41 45 0 0 0 0 0 0 0
VA Max 160 181 160 12589 9767 5918 9010 15866 31659 10407
VA Average 86 90 88 746 299 180 222 1284 1793 368
VA Median 86 88 88 12 1 0 0 416 414 15
VA Std Dev 21 23 21 N/A N/A N/A N/A N/A N/A N/A
VA Variance 436 523 446 N/A N/A N/A N/A N/A N/A N/A
ALL Count N/A N/A 2462 N/A N/A N/A N/A N/A N/A N/A
ALL Min N/A N/A 31 N/A N/A N/A N/A N/A N/A N/A
ALL Max N/A N/A 482 N/A N/A N/A N/A N/A N/A N/A
ALL Average N/A N/A 86 N/A N/A N/A N/A N/A N/A N/A
ALL Median N/A N/A 70 N/A N/A N/A N/A N/A N/A N/A
ALL Std Dev N/A N/A 48 N/A N/A N/A N/A N/A N/A N/A
ALL Variance N/A N/A 2330 N/A N/A N/A N/A N/A N/A N/A

Table 4.5 Basic Pavement Statistics Part 2 - Not Weighted
State Statistic Patching - Wheel Path (square feet) Patching - Non-wheel Path (square feet) Number of Potholes Rut Depth CCI OPC IRI Condition Index Rut Count Rut Condition Index Friction Number Friction Condition Index Cracking Index Cracking Condition Index
DE Count N/A N/A N/A N/A N/A 193 N/A N/A N/A N/A N/A N/A N/A
DE Min N/A N/A N/A N/A N/A 28 NN/A N/A N/A N/A N/A N/A N/A
DE Max N/A N/A N/A N/A N/A 100 N/A N/A N/A N/A N/A N/A N/A
DE Average N/A N/A N/A N/A N/A 71 N/A N/A N/A N/A N/A N/A N/A
DE Median N/A N/A N/A N/A N/A 71 N/A N/A N/A N/A N/A N/A N/A
DE Std Dev N/A N/A N/A N/A N/A 15 N/A N/A N/A N/A N/A N/A N/A
DE Variance N/A N/A N/A N/A N/A 225 N/A N/A N/A N/A N/A N/A N/A
MD Count N/A N/A N/A 2179 N/A N/A 2179 2179 2179 398 398 1956 1956
MD Min N/A N/A N/A 0.05 N/A N/A 1 0 1 10 1 55 1
MD Max N/A N/A N/A 0.37 N/A N/A 5 48 3 63 3 100 4
MD Average N/A N/A N/A 0.16 N/A N/A 1.95 5 1.28 45 2.81 95 1.18
MD Median N/A N/A N/A 0.14 N/A N/A 2 0 1 46 3 98 1
MD Std Dev N/A N/A N/A 0.07 N/A N/A 0.96 11 0.66 6 0.47 7 0.51
MD Variance N/A N/A N/A 0.00 N/A N/A 0.92 124 0.44 35 0.22 43 0.26
VA Count 192 192 192 192 190 N/A N/A N/A N/A N/A N/A N/A N/A
VA Min 0 0 0 0.1 16 N/A N/A N/A N/A N/A N/A N/A N/A
VA Max 9190 11858 3 0.48 100 N/A N/A N/A N/A N/A N/A N/A N/A
VA Average 411 393 0.08 0.19 73 N/A N/A N/A N/A N/A N/A N/A N/A
VA Median 17 16 0 0.19 77 N/A N/A N/A N/A N/A N/Ae N/A N/A
VA Std Dev N/A N/A 0.41 0.06 20 N/A N/A N/A N/A N/A N/A N/A N/A
VA Variance N/A N/A 0.17 0.00 384 N/A N/A N/A N/A N/A N/A N/A N/A
ALL Count N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
ALL Min N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
ALL Max N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
ALL Average N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
ALL Median N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
ALL Std Dev N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
ALL Variance N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A

1. 2008-2011 Draft Statewide Transportation Improvement Program: Evaluation of the State Bridge Program, Oregon Department of Transportation, Bridge Engineering Section, April 2007.

2. Recording and Coding Guide for the Structure Inventory and Appraisal of the Nation's Bridges, FHWA-PD-96-001, December 1995.

3. http://en.wikipedia.org/wiki/Interstate_95.

4. McGhee, K. H., Development and Implementation of Pavement Condition Indices for the Virginia Department of Transportation, Phase 1: Flexible Pavement, Virginia Department of Transportation, Maintenance Division, September 2002.

<< Previous Contents Next >>

Printable Version

PDF (4 mb)

Events

More Information

Contact

PDF files can be viewed with the Acrobat® Reader®
Updated: 06/18/2012