Bridges for Service Life Beyond 100 Years: Service Limit State Design (2014)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

56 C h a p t e r 4 4.1 Introduction Deterioration and a degradation of serviceability indicated by a reduction in usable capacity (either by a change in rating or a change in reliability index) are not interchangeable terms, although they may be related for some elements or systems. For example, a steel girder, especially a rolled beam, may have visible corrosion suggesting that the element is deteriorating and in need of maintenance, but it may have little or no per- ceptible change in deformations, stresses, or rider comfort. Sometimes a corrosion hole may exist in a place that does not control an evaluation. Similar observations can be made about the early stages of damage to prestressed beams resulting from poor drainage control at expansion joints. The resulting spall- ing and possible rusting of rebar and strands may be unsightly but have relatively little structural effect until the damage is well advanced. For these types of elements or systems, deterio- ration models or databases geared to predicting maintenance budget needs are not especially useful to the calibration pro- cess needed for Project R19B. Degradation first leads to loss of service, and if left untreated, can lead to loss of load-carrying capacity resulting in failure. However, for some elements and subsystems, a high cor- relation may exist between loss of serviceability and deterio- ration as indicated by a change in National Bridge Inspection Standards (NBIS) sufficiency ratings or the results of a dete- rioration model. Such might be the case for decks and bear- ings, for which a deteriorated state could be considered in the calibration by owner adjustment of nominal resistance based on bridge-specific knowledge or deterioration modeling. Several researchers have proposed algorithms predicting the change in condition number, either the National Bridge Inventory (NBI) condition number or a variation, over time for bridge details or complete structures. Five of these pro- posals are reviewed in this chapter. Bridge owners could use one or more of these proposals, or others that may be found to more accurately reflect local conditions, as a basis for including an estimate of deterioration in recalibrating the service limit states by using the framework described in Chapter 3 of this report. One simple way to do this would be to accept the premise that until more usable data on the change of resistance with time are available, it is reasonable to treat the percentage change in condition number as a surro- gate for change in resistance and adjust the resistance in the calibration spreadsheets accordingly. In some cases, the use of the equations included in this chapter may be beyond the range for which they were developed. 4.2 Bolukbasi et al. (2004) Bolukbasi et al. (2004) used historic NBI rating data for 2,601 bridges from Illinois to determine regression equations relating the bridge age to the condition rating of the deck, superstruc- ture, and substructure. No distinction between cast-in-place (CIP) decks and precast panels was indicated in the reference. The data were adjusted such that bridges with a sudden rating increase are excluded from the study (a sudden increase in rating indicates performance of maintenance). The resulting equations suggest the rating and a corresponding service life if no maintenance occurs. Equations are provided for the follow- ing categories: all bridges; steel bridges; reinforced-concrete (RC) bridges; prestressed concrete bridges; Interstate bridges; non-Interstate bridges; bridges with annual average daily traffic (AADT) <5,000; bridges with 5,000 < AADT < 10,000; and bridges with AADT >10,000. Within each bridge category, equations are provided to estimate the rating for the deck, superstructure, and substructure. Table 4.1 shows the rating prediction equations for the nine categories, as well as a graph showing the condition rating versus time. The end of service life is typically defined as when a rating of 3 is achieved, and maintenance would be required to continue using the structure. The equations were also plotted by component (deck, superstructure, substructure), allowing an investigation into Deterioration

57 Table 4.1. Rating Prediction Equations and Graphs for Nine Categories of Bridges 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) All Bridges Deck Superstructure Substructure Deck: R = 8.960814 - 0.20144T + 0.006719T2 - 9.67 × 10-5T3 Superstructure: R = 8.854089461 - 0.144890772T + 0.003122716T2 - 2.91 × 10-5T3 Substructure: R = 8.767383274 - 0.127816817T + 0.002736488T 2 - 2.57 × 10-5T3 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Steel Bridges Deck Superstructure Substructure Deck: R = 8.922947 - 0.19861T + 0.006735T 2 - 9.77 × 10-5T3 Superstructure: R = 8.895666888 - 0.160854616T + 0.004406448T2 - 5.36 × 10-5T3 Substructure: R = 8.822326892 - 0.148338077T + 0.004166181T 2 - 4.83 × 10-5T3 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Reinforced Concrete Bridges Deck Superstructure Substructure Deck: R = 8.605268604 - 0.1277358696T + 0.0023501188T 2 - 3.643 × 10-5T3 Superstructure: R = 8.662249581 - 0.145660594T + 0.003299188T2 - 3.09 × 10-5T3 Substructure: R = 8.624414481 - 0.123890228T + 0.002486843T 2 - 2.21 × 10-5T3 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Prestressed Concrete Bridges Deck Superstructure Substructure Deck: R = 9.243165 - 0.25857T + 0.01004T 2 - 1.5 × 10-4T3 Superstructure: R = 9.134415141 - 0.213185033T + 0.006920265T2 - 8.77 × 10-5T3 Substructure: R = 9.075226897 - 0.19604399T + 0.006203563T 2 - 7.49 × 10-5T3 (continued on next page)

58 Table 4.1. Rating Prediction Equations and Graphs for Nine Categories of Bridges 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Interstate Bridges Deck Superstructure Substructure Deck: R = 8.920346 - 0.21323T + 0.007687T 2 - 1.5 × 10-4T3 Superstructure: R = 8.974079168 - 0.193652056T + 0.006771473T2 - 1.2065 × 10-4T3 Substructure: R = 8.956002854 - 0.205796117T + 0.008041095T 2 - 1.31981 × 10-4T3 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Non-Interstate Bridges Deck Superstructure Substructure Deck: R = 8.981204 - 0.20173T + 0.007319T 2 - 1.1 × 10-4T3 Superstructure: R = 8.823963724 - 0.134551029T + 0.002855493T2 - 2.73 × 10-5T3 Substructure: R = 8.745705917 - 0.113435369T + 0.002153535T 2 - 2.01 × 10-5T3 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bridges with AADT < 5,000 Deck Superstructure Substructure Deck: R = 8.974903 - 0.20009T + 0.007589T 2 - 1.1 × 10-4T3 Superstructure: R = 8.793293844 - 0.128307613T + 0.002753594T2 - 2.73 × 10-5T3 Substructure: R = 8.688714213 - 0.10308987T + 0.001890448T 2 - 1.87 × 10-5T3 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bridges with 5,000 < AADT < 10,000 Deck Superstructure Substructure Deck: R = 8.887719688 - 0.1873850501T + 0.0047333447T2 - 7.2279 × 10-5T3 Superstructure: R = 8.812137936 - 0.144747111T + 0.002574894T2 - 2.06 × 10-5T3 Substructure: R = 8.791762968 - 0.139058986T + 0.0026443T2 - 2.02 × 10-5T3 (continued) (continued on next page)

59 Table 4.1. Rating Prediction Equations and Graphs for Nine Categories of Bridges 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bridges with AADT > 10,000 Deck Superstructure Substructure Deck: R = 9.047654 - 0.24391T + 0.009357T 2 - 1.7 × 10-4T3 Superstructure: R = 9.056469638 - 0.195142224T + 0.00607703T 2 - 9.69 × 10-5T3 Substructure: R = 9.045892953 - 0.218679597T + 0.008359897T 2 - 1.31336 × 10-4T3 Source: Bolukbasi et al. (2004). how other criteria affect the service life of the bridge. Figure 4.1, Figure 4.2, and Figure 4.3 show the predicted deck, super- structure, and substructure condition rating, respectively, versus time. The graphs in Table 4.1 show that the service life of the deck is typically shorter than the service life of either the superstructure or the substructure. All categories have a deck service life less than 55 years. For Interstate bridges and high-traffic-volume bridges, Figure 4.1 shows the deck ser- vice life is much closer to 40 years. High-traffic-volume bridges have an estimated service life of 41 years, medium- traffic-volume bridges have an estimated service life of 47 years, and low-traffic-volume bridges have an estimated ser- vice life of 54 years. The medium- and high-traffic bridge deck condition ratings decrease at a faster rate than the low-volume bridges. For the first 30 years, the medium and high traffic have nearly identical deck condition ratings. After the first 30 years they split, with the high-traffic bridges decreasing faster. Figure 4.2 shows the superstructure rating versus time for the nine bridge categories. Similar to the deck condition rating, Interstate bridges and bridges with AADT >10,000 have the shortest service life (45 to 50 years). Steel and prestressed con- crete bridges also have shorter service lives (55 to 65 years), but this is likely due to the fact that many of these are located on the Interstate and are subject to high traffic counts. The basis for this difference, 55 years and 65 years, could not be found in the reference. Non-Interstate, RC, and low-traffic bridges have an estimated service life of approximately 75 years. Bridges with AADT between 5,000 and 10,000 are shown to have the longest service life, greater than 80 years. The AADT between 5,000 and 10,000 category may have a longer esti- mated life due to less traffic than the Interstate and AADT >10,000 categories, combined with routine maintenance and repair, resulting in a longer service life. Figure 4.3 shows the substructure condition rating versus time for the nine bridge categories. As with the other condi- tion ratings, bridges falling into the high-traffic and Interstate categories have the shortest service life (an estimated 50 years). The substructures of prestressed concrete bridges and steel girder bridges have estimated service lives of approximately 62 and 67 years, respectively. The basis of this difference could 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 R ati ng Time (years) Deck Ratings All Bridges Steel Bridges RC Bridges Prestressed Concrete Bridges Interstate Bridges Non-Interstate Bridges AADT < 5,000 5,000 < AADT < 10,000 AADT > 10,000 Figure 4.1. Deck condition rating versus time. 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 140 Ra tin g Time (years) Superstructure Ratings All Bridges Steel Bridges RC Bridges PS Concrete Bridges Interstate Bridges Non-Interstate Bridges AADT < 5,000 5,000 < AADT < 10,000 AADT > 10,000 Figure 4.2. Superstructure condition rating versus time. (continued)

60 not be found in the reference. Non-Interstate, RC, and low- traffic bridges all have an estimated service life of approxi- mately 80 years, and bridges in the medium-traffic category have the longest estimated service life (approximately 90 years). Similar to the superstructure ratings, the substructure service life for the medium-traffic category is the longest and may be due to better maintenance. 4.3 Jiang and Sinha (1989) In their 1989 report on bridge performance and optimization, Jiang and Sinha discussed the results of regression analy sis and Markov chain analysis to estimate the average rating of a group of bridges. They considered Interstate and non-Interstate bridges, as well as steel and concrete bridges; no distinction was made between reinforced or prestressed concrete construction. Geographic location and traffic volume were initially consid- ered, but because they did not appear to influence the regres- sion analysis, they were not considered as separate categories. A relatively small sample (several hundred bridges) was used in the regression analysis, and at the time of the analysis, biennial NBI inspections had only been occurring for approximately 10 years. Thus, the results may have been influenced by the limited amount of data available and used. The results of the regression analysis were coefficients for a third-order polynomial describing the NBI condition rating as a function of bridge age. Coefficients were determined for the different bridge types and for the deck, superstructure, and substructure. The equations and a graph of the equations showing the NBI condition rating as a function of time are shown in Table 4.2. Unlike the equations by Bolukbasi et al. 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 140 Ra tin g Time (years) Substructure Ratings All Bridges Steel Bridges RC Bridges PS Concrete Bridges Interstate Bridges Non-Interstate Bridges AADT < 5,000 5,000 < AADT < 10,000 AADT > 10,000 Figure 4.3. Substructure condition rating versus time. Table 4.2. Rating Prediction Equations and Graphs for Four Categories of Bridges 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Steel Interstate Bridges Deck Superstructure Substructure Deck: R = 9 - 0.41141790T + 0.02116563T 2 - 4.0387 × 10-4T3 Superstructure: R = 9 - 0.45572206T + 0.02399958T 2 - 4.4201 × 10-4T3 Substructure: R = 9 - 0.44818105T + 0.02555900T 2 - 4.9875 × 10-4T3 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Concrete Interstate Bridges Deck Superstructure Substructure Deck: R = 9 - 0.36622617T + 0.01659520T 2 - 2.7162 × 10-4T3 Superstructure: R = 9 - 0.34704791T + 0.01598966T 2 - 2.7160 × 10-4T3 Substructure: R = 9 - 0.34508455T + 0.01575857T 2 - 2.6681 × 10-4T3 (continued on next page)

61 (2004), the constant term in the prediction equation is always 9; this assumes that the bridge component was in perfect condition when new. The equations were also plotted by component (deck, superstructure, substructure), allowing an investigation into how other criteria affect the service life of the bridge. Decks are believed to be CIP concrete decks. Figure 4.4, Figure 4.5, and Figure 4.6 show the predicted deck, superstructure, and substructure condition ratings, respectively, versus time. It is typically assumed that the end of service life occurs when the condition rating reaches a value of 3. Figure 4.4 shows that the Interstate bridges typically have a shorter deck service life than the non-Interstate bridges. All bridge types have a similar deterioration rate until approximately 10 years. Table 4.2. Rating Prediction Equations and Graphs for Four Categories of Bridges 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Steel Non-Interstate Bridges Deck Superstructure Substructure Deck: R = 9 - 0.34979283T + 0.01036093T 2 - 1.1009 × 10-4T3 Superstructure: R = 9 - 0.34616183T + 0.01088174T 2 - 1.1870 × 10-4T3 Substructure: R = 9 - 0.34059831T + 0.01093574T 2 - 1.1953 × 10-4T3 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Concrete Non-Interstate Bridges Deck Superstructure Substructure Deck: R = 9 - 0.30199933T + 0.00915111T2 - 9.409 × 10-5T3 Superstructure: R = 9 - 0.29095931T + 0.00860726T 2 - 8.815 × 10-5T3 Substructure: R = 9 - 0.31267496T + 0.00961677T 2 - 9.876 × 10-5T3 Source: Jiang and Sinha (1989). 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Deck Ratings Steel Interstate Concrete Interstate Steel Non-Interstate Concrete Non-Interstate Figure 4.4. Deck predicted condition ratings for different bridge types. 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Superstructure Ratings Steel Interstate Concrete Interstate Steel Non-Interstate Concrete Non-Interstate Figure 4.5. Superstructure predicted condition ratings for different bridge types. (continued)

62 Non-Interstate bridges have a longer period during which the rating does not change significantly; the concrete bridge rating is higher than the steel bridge rating during this pla- teau period and throughout most of the service life. Con- crete bridges are shown to have longer deck service lives than steel bridges. The predicted service life for Interstate bridge decks is approximately 36 years for steel bridges and 42 years for concrete bridges. For non-Interstate bridges, the service life increases to approximately 59 years for steel bridges and 62 years for concrete bridges. The reference did not provide any information as to why the deck service life varies between steel and concrete bridges for both Interstate and non-Interstate conditions. Figure 4.5 is very similar to Figure 4.4 with regard to mate- rial and highway type. Concrete superstructures have longer service lives than steel bridges subjected to the same volume of traffic. Interstate bridges have a shorter service life than non- Interstate bridges. The superstructure service lives predicted are very similar to those predicted for the deck. The difference in service life between concrete and steel bridges is not dis- cussed in the reference. Figure 4.6 shows the predicted substructure condition ratings versus times. This figure is similar to the two previous figures for deck and superstructure service lives. The predicted substructure service lives are very similar to those predicted for the deck and also for the superstructure. This is a surprising result as it is typically expected that the substructure will last longer than either the deck or the superstructure. The report did not indicate any specific reasons for the substructure having predicted service lives similar to the deck and superstructure. 4.4 hatami and Morcous (2011) A 2011 report by Hatami and Morcous, Developing Deteriora- tion Models for Nebraska Bridges, presented the results of a project performed for the Nebraska Department of Roads in which deterioration models were developed specifically for Nebraska bridges. The deterioration models were based on NBI condition ratings for bridge decks, superstructures, and substructures by using data from 1998 to 2010. Factors such as structure type, deck type, wearing surface, deck protection, average daily traffic (ADT), average daily truck traffic (ADTT), and location were considered in the development of the dete- rioration models, which were determined using deterministic and stochastic methods. NBI data were obtained for all bridges in Nebraska from 1998 to 2010; only data for state bridges were used in the analysis as the authors believed that inspections performed by state inspectors have stricter requirements. The determin- istic deterioration models developed for state bridges in Nebraska are shown in Table 4.3. In the second figure in the table, in which deterioration is related to ADTT, decks sub- jected to more truck traffic appear to have a longer expected life than those subjected to fewer trucks, which is contrary to what would be expected. In the third figure, the rating starts to increase in District 2 after approximately 60 years, which is likely a sign that more data were needed to more accurately develop the deterioration model. The last figure shown in Table 4.3 indicates that the service life of the deck exceeds that of either the superstructure or substructure, which is also contrary to what would be expected. 4.5 Comparison of equations from Bolukbasi et al. (2004), Jiang and Sinha (1989), and hatami and Morcous (2011) 4.5.1 Introduction The results from Bolukbasi et al. (2004), Jiang and Sinha (1989), and Hatami and Morcous (2011) are generally simi- lar. The equations are plotted together to provide a comparison between resulting equations. Various comparisons are pro- vided below. Comparisons are based on material type, as well as highway type and ADTT. Typically the Bolukbasi et al. equations have a slower deterioration rate over the service life of the structure. This may be due to a larger number of struc- tures being considered and the availability of more inspection data. The equations from the Nebraska study are only included in the superstructure ratings for steel bridges; the results for other bridge and component types were not spe- cific enough to include elsewhere. 4.5.2 Concrete Superstructure Bridges Plots of the prediction equations for deck, superstructure, and substructure condition ratings for concrete bridges are 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Substructure Ratings Steel Interstate Concrete Interstate Steel Non-Interstate Concrete Non-Interstate Figure 4.6. Substructure predicted condition ratings for different bridge types.

63 Table 4.3. Nebraska Deterioration Models 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Deck and Overlay Ratings Original Deck Replacement Deck Overlay Low Slump Concrete Overlay Original Deck: R = 10.2915 - 0.2531T + 0.0093T 2 - 0.0001T3 Replacement Deck: R = 8.48681 + 0.34139T - 0.05392T 2 + 0.00222T 3 - 0.00003T 4 Overlay: R = 9.6499 - 0.0829T + 0.0009T 2 - 0.0002T3 Low Slump Concrete Overlay: R = 10.094 - 0.1902T + 0.0087T 2 - 0.0004T 3 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 R at in g Time (years) Deck Ratings based on ADTT Deck w/ ADTT<100 Deck w/ 100<ADTT<500 Deck w/ ADTT>500 ADTT < 100: R = 10.189 - 0.233T + 0.0092T 2 - 0.0002T3 100 < ADTT < 500: R = 10.754 - 0.342T + 0.0127T 2 - 0.0002T 3 ADTT > 500: R = 10.372 - 0.2311T + 0.0039T 2 - 0.00004T3 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Deck Ratings based on Location Deck Dis. 1, 3, 4 Deck Dis. 2 Deck Dis. 5-8 Districts 1, 3, and 4: R = 9.9984 - 0.1948T + 0.0052T 2 - 0.00008T 3 District 2: R = 9.9374 - 0.1267T - 0.0015T 2 + 0.00003T3 Districts 5–8: R = 10.252 - 0.2214T + 0.0067T 2 - 0.0001T3 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Deck and Overlay Ratings Original Deck Steel Superstructure Substructure Original Deck: R = 10.2915 - 0.2531T + 0.0093T 2 - 0.0001T3 Steel Superstructure: R = 10.2731 - 0.1727T + 0.0046T 2 - 0.0001T 3 Substructure: R = 9.6098 - 0.0657T - 0.0017T 2 + 0.00001T3 Source: Hatami and Morcous (2011).

64 shown in Figure 4.7. For the deck condition ratings, the Bolukbasi et al. (2004) equation indicates the highest condi- tion rating until an age of approximately 35 years, after which the non-Interstate equation for concrete bridges pro- posed by Jiang and Sinha (1989) indicates the highest con- dition rating. For the superstructure and substructure, the Bolukbasi et al. equations always indicate the highest con- dition rating. The prediction equations provide an estimated service life for the deck, superstructure, and substructure. The estimated service lives, or the predicted times until a condition rating of 3 is achieved, are provided in Table 4.4. The Jiang and Sinha equations generally predict a similar service life for all major components of a bridge. The Bolukbasi et al. equation suggests that the deck has the shortest service life. The super- structure and substructure service lives are significantly longer. 4.5.3 Steel Superstructure Bridges Plots of the prediction equations for deck, superstructure, and substructure condition ratings for steel bridges are shown in Figure 4.8. For the deck condition ratings, the Bolukbasi et al. (2004) equation indicates a higher condition rating until an age of approximately 50 years, after which the non-Interstate equation for steel bridges proposed by Jiang and Sinha (1989) indicates a higher condition rating. The superstructure and substructure condition ratings are always higher when using the Bolukbasi et al. equations versus either Jiang and Sinha equation. The Hatami and Morcous (2011) equation Table 4.4. Service Life Comparison: Reinforced Concrete (RC) Bridges Equation Service Life (years) Deck Superstructure Substructure Bolukbasi et al., RC 54 77 82 Jiang and Sinha, RC Interstate 42 41 41 Jiang and Sinha, RC non-Interstate 63 63 63 Figure 4.7. Comparisons of concrete bridge predicted condition ratings for (a) decks, (b) superstructures, and (c) substructures. (a) 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Deck Ratings Bolukbasi - RC Bridges Jiang and Sinha - Concrete Interstate Jiang and Sinha - Concrete Non-Interstate (b) 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Superstructure Ratings Bolukbasi - RC Bridges Jiang and Sinha - Concrete Interstate Jiang and Sinha - Concrete Non-Interstate (c) 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Substructure Ratings Bolukbasi - RC Bridges Jiang and Sinha - Concrete Interstate Jiang and Sinha - Concrete Non-Interstate

65 [identified as NEDOR (Nebraska Department of Roads) in Figure 4.8] predicts higher condition ratings than both Bolukbasi et al. and Jiang and Sinha until approximately 30 years. Unlike the other equations, the Hatami and Morcous equation does not indicate a period of time when the condition rating plateaus. The prediction equations pro- vide an estimated service life for the deck, superstructure, and substructure. The estimated service lives, or the predicted times until a condition rating of 3 is achieved, are provided in Table 4.5. The Jiang and Sinha equations generally predict a similar service life for the major components of a bridge. The Bolukbasi et al. equation suggests that the deck has the short- est service life. The superstructure and substructure service lives are somewhat longer. 4.5.4 Interstate Bridges Plots of the prediction equations for deck, superstructure, and substructure condition ratings for Interstate bridges are shown in Figure 4.9. For the deck condition ratings, the Bolukbasi et al. (2004) equation indicates a higher condition rating until an age of approximately 25 years. After 25 years, the Bolukbasi et al. equation and the Jiang and Sinha (1989) concrete equations are very similar. The superstructure and substructure condition ratings are always higher when using the Bolukbasi et al. equations versus either Jiang and Sinha equation, but overall the three indicate similar estimated Table 4.5. Service Life Comparison: Steel Bridges Equation Service Life (years) Deck Superstructure Substructure Bolukbasi et al., steel 53 63 68 Jiang and Sinha, steel Interstate 36 37 36 Jiang and Sinha, steel non-Interstate 54 56 57 Hatami and Morcous, steel NA 47 NA Note: NA = not available. Figure 4.8. Comparisons of steel bridge predicted condition ratings for (a) decks, (b) superstructures, and (c) substructures. (a) Deck Ratings 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - Steel Bridges Jiang and Sinha - Steel Interstate Jiang and Sinha - Steel Non-Interstate (b) Superstructure Ratings 2 3 4 5 6 7 8 9 10 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - Steel Bridges Jiang and Sinha - Steel Interstate Jiang and Sinha - Steel Non-Interstate NE DOR - Steel Bridges (c) Substructure Ratings 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - Steel Bridges Jiang and Sinha - Steel Interstate Jiang and Sinha - Steel Non-Interstate

66 service lives. The prediction equations provide an estimated service life for the deck, superstructure, and substructure. The estimated service lives, or the predicted times until a con- dition rating of 3 is achieved, are provided in Table 4.6. The estimated service lives are similar for all components and equations. 4.5.5 Non-Interstate Bridges The prediction equations for deck, superstructure, and substructure condition ratings for non-Interstate bridges are shown in Figure 4.10. For the deck condition ratings, the Bolukbasi et al. (2004) equation indicates a higher rat- ing until an age of approximately 40 years. After 40 years, the Jiang and Sinha (1989) concrete equation indicates the highest condition rating, and after approximately 45 years, the Jiang and Sinha steel equation provides an estima- ted service life greater than the Bolukbasi et al. equation. The super structure and substructure condition ratings are always higher when using the Bolukbasi et al. equations versus either Jiang and Sinha equation. The Bolukbasi et al. equations indicate an overall slower deterioration rate. The prediction equations provide an estimated service life for the deck, superstructure, and substruc- ture. The estimated service lives, or the predicted times until a condition rating of 3 is achieved, are provided in Table 4.7. The table shows that the Bolukbasi et al. Table 4.6. Service Life Comparison: Interstate Bridges Equation Service Life (years) Deck Superstructure Substructure Bolukbasi et al., Interstate 41 45 48 Jiang and Sinha, steel Interstate 36 37 36 Jiang and Sinha, concrete Interstate 42 41 41 Figure 4.9. Comparisons of Interstate bridge predicted condition ratings for (a) decks, (b) superstructures, and (c) substructures. (a) (b) (c) Deck Ratings 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - Interstate Bridges Jiang and Sinha - Steel Interstate Jiang and Sinha - Concrete Interstate Superstructure Ratings 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - Interstate Bridges Jiang and Sinha - Steel Interstate Jiang and Sinha - Concrete Interstate Substructure Ratings 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - Interstate Bridges Jiang and Sinha - Steel Interstate Jiang and Sinha - Concrete Interstate

67 Table 4.7. Service Life Comparison: Non-Interstate Bridges Equation Service Life (years) Deck Superstructure Substructure Bolukbasi et al., non-Interstate 52 77 81 Jiang and Sinha, steel non-Interstate 54 56 57 Jiang and Sinha, con- crete non-Interstate 62 62 63 equations predict the shortest deck service life, but they also predict the longest superstructure and substructure service lives. 4.5.6 Low- and Medium-AADT Bridges The prediction equations for deck, superstructure, and substructure condition ratings for bridges with AADT <10,000 for Bolukbasi et al. (2004) and for non-Interstate bridges for Jiang and Sinha (1989) are shown in Fig- ure 4.11. For the deck condition ratings, the Bolukbasi et al. equation for AADT <5,000 indicates the highest condition rating until an age of approximately 42 years. After this time, the Jiang and Sinha concrete non-Interstate equation indicates the highest condition rating, and after approximately 50 years, the Jiang and Sinha steel non- Interstate equation provides an estimated service life similar to the Bolukbasi et al. equation for AADT <5,000. The Bolukbasi et al. equation for AADT between 5,000 and 10,000 is greater than the Jiang and Sinha equations until approximately 27 years. The superstructure and substructure condition ratings are always higher when using the Bolukbasi et al. equations versus the Jiang and Sinha equations. The Bolukbasi et al. equations indicate an overall slower deterioration rate. The equation for AADT <5,000 is greater until approximately 63 years for the superstructure and approximately 50 years for the substructure. The prediction equations provide an estimated service life for the deck, superstructure, and Figure 4.10. Comparisons of non-Interstate bridge predicted condition ratings for (a) decks, (b) superstructures, and (c) substructures. (a) Deck Ratings 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - Non- Interstate Bridges Jiang and Sinha - Steel Non-Interstate Jiang and Sinha - Concrete Non-Interstate (b) Superstructure Ratings 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - Non-Interstate Bridges Jiang and Sinha - Steel Non-Interstate Jiang and Sinha - Concrete Non-Interstate (c) Substructure Ratings 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - Non- Interstate Bridges Jiang and Sinha - Steel Non-Interstate Jiang and Sinha - Concrete Non-Interstate

68 substructure. The estimated service lives, or the predicted times until a condition rating of 3 is achieved, are provided in Table 4.8. The table shows that the Bolukbasi et al. equa- tions predict the shortest deck service life, but they also predict the longest superstructure and substructure ser- vice lives. 4.5.7 High-AADT Bridges The prediction equations for deck, superstructure, and sub- structure condition ratings for bridges with AADT >10,000 for Bolukbasi et al. (2004) and for Interstate bridges for Jiang and Sinha (1989) are shown in Figure 4.12. The Bolukbasi et al. deck condition rating equation is greater for approxi- mately 30 years, after which it is very similar to the Jiang and Sinha concrete Interstate prediction equation. The superstructure and substructure condition ratings are always higher when using the Bolukbasi et al. equation versus the Jiang and Sinha equations. The Bolukbasi et al. equations indicate an overall slower deterioration rate. The prediction equations provide an estimated service life for the deck, superstructure, and substructure. The estimated service lives, or the predicted times until a condition rating of 3 is achieved, are provided in Table 4.9. The table shows that the Bolukbasi et al. equations predict a deck service life approximately equal to the service life predicted by the Jiang and Sinha concrete Interstate bridge equation, and they also predict the longest superstructure and substructure service lives. Table 4.8. Service Life Comparison: Low- to Medium-AADT Bridges Equation Service Life (years) Deck Superstructure Substructure Bolukbasi et al., AADT <5,000 54 76 80 Bolukbasi et al., 5,000 < AADT < 10,000 47 81 90 Jiang and Sinha, steel non-Interstate 54 56 57 Jiang and Sinha, concrete non-Interstate 62 62 63 Figure 4.11. Comparisons of low- to medium-AADT bridge predicted condition ratings for (a) decks, (b) superstructures, and (c) substructures. (a) Deck Ratings 2 3 4 5 6 7 8 9 10 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - AADT < 5,000 Bolukbasi - 5,000 < AADT < 10,000 Jiang and Sinha - Steel Non-Interstate Jiang and Sinha - Concrete Non-Interstate (b) Superstructure Ratings 2 3 4 5 6 7 8 9 10 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - AADT < 5,000 Bolukbasi - 5,000 < AADT < 10,000 Jiang and Sinha - Steel Non-Interstate Jiang and Sinha - Concrete Non-Interstate (c) Substructure Ratings 2 3 4 5 6 7 8 9 10 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - AADT < 5,000 Bolukbasi - 5,000 < AADT < 10,000 Jiang and Sinha - Steel Non-Interstate Jiang and Sinha - Concrete Non-Interstate

69 4.6 agrawal and Kawaguchi (2009) A 2009 report by Agrawal and Kawaguchi provides regression equations relating condition rating (CR) to age for common bridge components in New York State. The following list indi- cates the number of options and examples of each component: • Abutment backwall (all grouped together); • Abutment stem (all grouped together); • Abutment wingwall (four options: none, other, wingwall exists, and reinforced earth wingwall); • Abutment bearing (six options: none, steel, polytetra- fluoroethylene [PTFE], multirotational, elastomeric, and others); • Abutment pedestal (all grouped together); • Abutment joint (12 options: none, open, finger, sliding plate, filled elastic material, preformed elastomeric seals, strip seal, sawed and filled, compression, modular, armored, and other or unknown); • Pier bearing (six options: none, steel, PTFE, multirota- tional, elastomeric, and other or unknown); • Pier pedestal (five options: none, concrete, masonry, steel, and timber); • Pier cap top (five options: none, concrete, masonry, steel, and timber); • Pier cap (five options: none, concrete, masonry, steel, and timber); • Pier stem (all grouped together); • Pier column (five options: none, concrete, masonry, steel, and timber); • Pier footing (all grouped together); Table 4.9. Service Life Comparison: High-AADT Bridges Equation Service Life (years) Deck Superstructure Substructure Bolukbasi et al., AADT >10,000 41 48 49 Jiang and Sinha, steel Interstate 36 37 36 Jiang and Sinha, concrete Interstate 42 41 41 Figure 4.12. Comparisons of high-AADT bridge predicted condition ratings for (a) decks, (b) superstructures, and (c) substructures. (a) Deck Ratings 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - AADT > 10,000 Jiang and Sinha - Steel Interstate Jiang and Sinha - Concrete Interstate (b) Superstructure Ratings 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - AADT > 10,000 Jiang and Sinha - Steel Interstate Jiang and Sinha - Concrete Interstate (c) Substructure Ratings 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Bolukbasi - AADT > 10,000 Jiang and Sinha - Steel Interstate Jiang and Sinha - Concrete Interstate

70 10,175 and 25,457 tons), high (between 16,969 and 40,195 tons), and very high (between 46,375 and 94,739 tons); • Snow accumulation—There are three snow accumulation categories: low (<171 in.), medium (between 171 and 278 in.), and high (between 278 and 458 in.); • Climate groups—The 10 groups are based on climate data provided by the National Oceanic and Atmospheric Administration; • Functional class—There are five functional classes ranging from Interstate to none; and • Feature under—This factor has three categories: Interstate under, highway under, and water under. The factors listed above were used to create a class of bridges that have similar characteristics. The number of characteris- tics selected allows the deterioration rate to be calculated for a very narrow or a very broad group of bridges. Within a specific component, multiple equations may be provided for different materials or types of components. As an example, for abutment bearings, four equations (one each for steel bearings, elastomeric bearings, multirota- tional bearings, and PTFE sliding bearings) were provided. The equations and graphs are shown in Table 4.10. The ratings in New York vary from 1 to 7, with 7 indicating perfect condition; 5 indicating minor deterioration but still functioning as designed; 3 indicating serious deterioration or not functioning as designed; and 1 indicating a failed condition. Even-numbered ratings (2, 4, and 6) are used to provide a middle ground between the odd numbered, defined ratings (1, 3, 5, and 7). If failure is defined as a condition rating of 3 and the component no longer functioning as intended, then the • Pier recommendation (five options: none, concrete, masonry, steel, and timber); • Pier joint (12 options: none, open, finger, sliding plate, filled elastic material, preformed elastomeric seals, strip seal, sawed and filled, compression, modular, armored, and other or unknown); and • Primary member design type (19 options, such as rolled beam, truss, and deck arch). The work by Agrawal and Kawaguchi resulted in a com- puter program based on synthesized Pontis data that calcu- lates the deterioration rates of bridge components using Pontis data. The program contains a cascading algorithm to classify bridges based on several factors. These factors are • Element design type—For bearings, for example, the type can be one of six choices (none, steel, PTFE, multirota- tional, elastomeric, and others); • New York State Department of Transportation (NYSDOT) Region—There are 11 regions in New York State; • Bridge ownership—Various organizations own bridges within New York State, including NYSDOT, park authori- ties or commissions, nonpark authorities or commissions, and the New York State Thruway Authority. Bridges are also owned locally, privately, by railroads, and by other entities; • Superstructure design type—These include girder and floorbeam system, truss, and suspension; • Superstructure material type—These include weathering steel, timber, and prestressed concrete; • AADT—AADT is divided into five groups ranging from no trucks to >5,000 trucks per day; • Salt usage—Salt usage is divided into four categories: low (between 6,893 and 13,492 tons), medium (between Table 4.10. Regression Equations and Graphs Based on New York State Bridge Data 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Abutment Backwall Abutment Backwall CR = 7 - 0.0564703T + 0.0000667T2 (continued on next page) (text continues on page 78)

71 Table 4.10. Regression Equations and Graphs Based on New York State Bridge Data 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Abutment Bearings Steel PTFE Multirotational Elastomeric Steel: CR = 7 - 0.0773187T + 0.0002408T 2 PTFE: CR = 7 - 0.1369652T + 0.0023073T 2 - 2.25 × 10-5T3 Multirotational: CR = 7 - 0.1276043T + 0.0020318T 2 - 1.8 × 10-5T3 Elastomeric: CR = 7 - 0.0633160T + 0.0002109T 2 - 1 × 10-7T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Abutment Joints Open Compression Modular Armor Sliding Plate Filled Elastic Preformed Open: CR = 7 - 0.1544542T + 0.0019093T 2 - 1.01 × 10-5T3 Compression: CR = 7 - 0.1546255T + 0.0019008T 2 - 9.3 × 10-6T3 Modular: CR = 7 - 0.1944402T + 0.0054188T 2 - 7.4 × 10-5T3 Armor: CR = 7 - 0.1667466T + 0.0022536T 2 - 1.29 × 10-5T3 Sliding Plate: CR = 7 - 0.1955859T + 0.0043095T 2 - 3.42 × 10-5T3 Filled Elastic: CR = 7 - 0.1416458T + 0.0016176T 2 - 6.3 × 10-6T3 Preformed: CR = 7 - 0.1563427T + 0.0014834T 2 - 5.0 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Abutment Pedestal Abutment Pedestal CR = 7 - 0.0484691T - 0.0000925T 2 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Abutment Stem Abutment Stem CR = 7 - 0.0562065T - 0.0000832T 2 (continued) (continued on next page)

72 (continued)Table 4.10. Regression Equations and Graphs Based on New York State Bridge Data 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Abutment Wingwall Abutment Wingwall CR = 7 - 0.0500728T - 0.0000546T 2 - 6.0 × 10-7T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Deck Curb Granite/Stone Steel Plate Timber Concrete Granite/Stone: CR = 7 - 0.0605424T + 0.0001089T 2 - 1.0 × 10-7T3 Steel Plate: CR = 7 - 0.0577393T - 0.0001956T 2 - 1.7 × 10-6T3 Timber: CR = 7 - 0.0584921T - 0.0003144T 2 - 2.4 × 10-6T3 Concrete: CR = 7 - 0.0507576T - 0.0002625T 2 - 1.9 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Pier Bearings Steel Multirotational Elastomeric Steel: CR = 7 - 0.0681319T - 0.0001597T 2 + 3.4 × 10-6T3 Multirotational: CR = 7 - 0.0833154T + 0.0008055T2 - 3.8 × 10-6T3 Elastomeric: CR = 7 - 0.0845871T + 0.0008876T 2 - 7.3 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Pier Cap Concrete Masonry Steel Timber Concrete: CR = 7 - 0.0575767T + 0.0000583T 2 + 1.0 × 10-7T3 Masonry: CR = 7 - 0.0347071T - 0.0002426T2 + 1.1 × 10-6T3 Steel: CR = 7 - 0.0172139T - 0.0008876T 2 + 3.8 × 10-6T3 Timber: CR = 7 - 0.0674187T + 0.0001438T 2 + 1.0 × 10-6T3 (continued on next page)

73 Table 4.10. Regression Equations and Graphs Based on New York State Bridge Data 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Pier Cap Top Concrete Masonry Steel Timber Concrete: CR = 7 - 0.0475800T - 0.0001091T2 + 1.2 × 10-6T3 Masonry: CR = 7 - 0.0094394T - 0.0007153T2 + 3.8 × 10-6T3 Steel: CR = 7 - 0.0131302T - 0.0007820T 2 + 4.9 × 10-6T3 Timber: CR = 7 - 0.0467232T + 0.0001051T 2 - 1.3 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Pier Column Concrete Masonry Steel Timber Concrete: CR = 7 - 0.0486218T - 0.0001326T 2 + 1.2 × 10-6T3 Masonry: CR = 7 - 0.1461181T + 0.0028522T 2 - 2.66 × 10-5T3 Steel: CR = 7 - 0.0594952T + 0.0002300T 2 - 4.0 × 10-7T3 Timber: CR = 7 - 0.1077933T + 0.0012051T 2 - 7.9 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Pier Footing Pier Footing CR = 7 - 0.0361181T - 0.0001836T 2 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 R at in g Time (years) Pier Joint Open Strip Seal Compression Modular Armor Sliding Plate Filled Elastic Preformed Open: CR = 7 - 0.1746867T + 0.0029733T 2 - 2.24 × 10-5T3 Strip Seal: CR = 7 - 0.2222855T + 0.0043429T 2 - 3.68 × 10-5T3 Compression: CR = 7 - 0.2047452T + 0.0034777T 2 - 2.09 × 10-5T3 Modular: CR = 7 - 0.1178004T + 0.0000691T2 + 1.37 × 10-5T3 Armor: CR = 7 - 0.1623125T + 0.0012891T 2 - 1.0 × 10-7T3 Sliding Plate: CR = 7 - 0.1581306T + 0.0016926T 2 - 8.9 × 10-6T3 Filled Elastic: CR = 7 - 0.1937046T + 0.0028916T 2 - 1.3 × 10-5T3 Preformed: CR = 7 - 0.1725949T + 0.0020362T 2 - 9.6 × 10-6T3 (continued) (continued on next page)

74 Table 4.10. Regression Equations and Graphs Based on New York State Bridge Data 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 R at in g Time (years) Pier Pedestal Concrete Masonry Steel Concrete: CR = 7 - 0.0427029T - 0.0003432T 2 + 2.8 × 10-6T3 Masonry: CR = 7 - 0.0214166T - 0.0007708T2 + 5.0 × 10-6T3 Steel: CR = 7 - 0.0294246T - 0.0002940T 2 + 1.5 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Pier Design Concrete Masonry Steel Timber Concrete: CR = 7 - 0.0616063T + 0.0001235T2 - 1.0 × 10-7T3 Masonry: CR = 7 + 0.0189981T - 0.0013498T2 + 7.5 × 10-6T3 Steel: CR = 7 - 0.0335030T - 0.0004089T2 + 2.4 × 10-6T3 Timber: CR = 7 - 0.1156794T + 0.0014818T2 - 9.8 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Pier Stem Pier Stem CR = 7 - 0.0445180T - 0.0001482T2 + 1.1 × 10-6 T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Primary Members Slab, Box, Box Channel Tee or I-Beam Rolled Beam Plate Girder Slab, Box, Box Channel: CR = 7 - 0.0724412T + 0.0002255T2 - 4.0 × 10-7T3 Tee or I-Beam: CR = 7 - 0.0509168T - 0.0001729T2 + 2.1 × 10-6T3 Rolled Beam: CR = 7 - 0.0573849T + 0.0000603T2 + 1.0 × 10-7T3 Plate Girder: CR = 7 - 0.0533815T + 0.0000618T2 (continued) (continued on next page)

75 Table 4.10. Regression Equations and Graphs Based on New York State Bridge Data 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Primary Members Truss Deck Arch Truss: CR = 7 - 0.0962120T + 0.0005460T2 - 1.6 × 10-6T3 Deck Arch: CR = 7 - 0.0608540T - 0.0001644T2 + 2.5 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Primary Members Metal Pipe Arch Frame Box Culvert Pipe Culvert Metal Pipe Arch: CR = 7 - 0.0917752T + 0.0006315T2 - 1.2 × 10-6T3 Frame: CR = 7 - 0.0586090T - 0.0000153T2 + 9.0 × 10-7T3 Box Culvert: CR = 7 - 0.0662312T + 0.0002877T2 - 1.1 × 10-6T3 Pipe Culvert: CR = 7 - 0.0918358T + 0.0005486T2 - 1.9 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Secondary Members Slab, Box, or Box Channel Tee/I-Beam Rolled Beam Plate Girder Slab, Box, or Box Channel: CR = 7 - 0.0705115T + 0.0002846T2 - 2.0 × 10-7T3 Tee or I-Beam: CR = 7 - 0.0371296T - 0.0004970T2 - 4.1 × 10-6T3 Rolled Beam: CR = 7 - 0.0536963T + 0.0002090T2 - 1.6 × 10-6T3 Plate Girder: CR = 7 - 0.0403950T - 0.0002383T2 + 1.6 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Secondary Members Truss Deck Arch Truss: CR = 7 - 0.0600905T - 0.0001653T2 + 2.1 × 10-6T3 Deck Arch: CR = 7 + 0.0225284T - 0.0019546T2 + 1.38 × 10-5T3 (continued) (continued on next page)

76 Table 4.10. Regression Equations and Graphs Based on New York State Bridge Data 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Secondary Members Frame Frame: CR = 7 - 0.0031620T - 0.0007666T2 + 3.9 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Sidewalk/Fascia Concrete Steel Plate Asphalt Concrete Steel Concrete: CR = 7 - 0.0697598T + 0.0001899T2 - 4.0 × 10-7T3 Steel Plate: CR = 7 - 0.0636279T + 0.0001742T2 + 1.0 × 10-7T3 Asphalt Concrete: CR = 7 - 0.1145251T + 0.0015822T2 - 1.34 × 10-5T3 Steela: CR = 7 - 0.0055077T - 0.0034812T2 + 4.78 × 10-5T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Structural Deck CIP Concrete Black Rebar Precast Plank CIP w/ Epoxy Rebar CIP w/ Other Prot. or Coating CIP Concrete Black Rebar: CR = 7 - 0.0675608T + 0.0001411T2 + 1.0 × 10-7T3 Precast Plank: CR = 7 - 0.1188157T + 0.0018646T2 - 2.04 × 10-5T3 CIP with Epoxy Rebar: CR = 7 - 0.0767927T + 0.0007988T2 - 5.1 × 10-6T3 CIP with Other Protection or Coating: CR = 7 - 0.0793700T + 0.0005157T2 - 2.3 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Structural Deck Timber Steel Grating Steel Plate Timber: CR = 7 - 0.1015141T + 0.0010366T2 - 7.3 × 10-6T3 Steel Grating: CR = 7 - 0.0971087T + 0.0005147T2 - 7 × 10-7T3 Steel Plate: CR = 7 - 0.1387853T + 0.0032377T2 - 2.53 × 10-5T3 (continued) (continued on next page)

77 Table 4.10. Regression Equations and Graphs Based on New York State Bridge Data 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Superstructure Type Slab Stringer/Multi-beam/Girder Girder/Floorbeam System Tee Beam Box Beam or Box Girder Slab: CR = 7 - 0.0550966T - 0.0000107T2 + 5.0 × 10-7T3 Stringer/Multi-beam/Girder: CR = 7 - 0.0608104T + 0.0001228T2 - 2 × 10-7T3 Girder/Floorbeam System: CR = 7 - 0.0375553T - 0.0003374T2 + 1.9 × 10-6T3 Tee Beam: CR = 7 - 0.0334694T - 0.0005675T2 + 4.1 × 10-6T3 Box Beam or Box Girder: CR = 7 - 0.0671339T + 0.0000287T2 + 1.5 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Superstructure Type Frame Culvert Frame: CR = 7 - 0.0374148T - 0.0004245T2 + 3.1 × 10-6T3 Culvert: CR = 7 - 0.0683836T + 0.0002159T2 - 1.0 × 10-7T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Superstructure Type Thru Truss Deck Arch Thru Truss: CR = 7 - 0.0719036T + 0.0001651T2 Deck Arch: CR = 7 - 0.0209106T + 0.0007879T2 + 5.1 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Wearing Surface Integral/Monolithic PC Concrete w/ Membrane Class H Concrete Integral/Monolithic Portland Cement (PC): CR = 7 - 0.1178904T + 0.0012462T2 + 7.0 × 10-6T3 Concrete with Membrane: CR = 7 - 0.3488945T + 0.021168T2 - 5.196 × 10-4T3 Class H Concretea: CR = 7 - 0.0417046T + 0.0022971T2 (continued) (continued on next page)

78 Table 4.10. Regression Equations and Graphs Based on New York State Bridge Data 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Wearing Surface PC Overlay Asphalt PC Overlay: CR = 7 - 0.1517338T + 0.0019529T 2 - 9.7 × 10-6T3 Asphalt: CR = 7 - 0.1215795T + 0.0008883T2 - 1.9 × 10-6T3 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 0 20 40 60 80 100 120 Ra tin g Time (years) Wearing Surface Wood/Wood Block Open Steel Grate Wood/Wood Block: CR = 7 - 0.168718T + 0.0020122T 2 - 9.2 × 10-6T3 Open Steel Grate: CR = 7 - 0.0726944T + 0.0004775T 2 - 3.9 × 10-6T3 Note: CR = condition rating. a Indicates equations that, when plotted, do not appear correct based on graphs provided in the report. Source: Agrawal and Kawaguchi (2009). service life of each component can be estimated. The esti- mated service lives are shown in Table 4.11 to Table 4.13. The reported service lives were determined by extending the graph until a condition rating of 3 was reached. Doing so may have resulted in some equations being used outside their intended range of applicability. In addition, the bridges used in this analysis were combined into one large group. This grouping may result in service lives for one material being greater than that for another material that might be expected to last longer. As an example, in Table 4.12 timber pier caps are predicted to last longer than concrete pier caps. Performing the analysis on a smaller group of bridges may result in the concrete service life being greater than that of the timber. 4.7 Stukhart et al. (1991) In a report to the Texas Department of Transportation (DOT), Stukhart et al. (1991) presented numerous equa- tions predicting the condition rating for bridge decks, superstructures, and substructures. No distinction between CIP decks and precast panels was noted in the reference. Several of the equations are from work completed by others, but most of the equations either use NBI data for Texas bridges or the expert opinion of Texas bridge engineers. The equations are shown and plotted in Table 4.14. The first set of equations was determined using regression analysis by the Transportation Systems Center and is a function of both age and ADT. NBI data for Texas bridges were used to determine addi- tional equations relating age and ADT to condition ratings. Linear, piecewise linear, and nonlinear equations were pro- posed; in addition, through a survey, equations based on expert opinion were determined considering the worst- case scenario, the most likely scenario, and the best-case scenario. Table 4.15 shows the prediction equations for coastal bridge substructures for different functional clas- sifications, and Table 4.16 shows the prediction equations for substructures in all regions not considering functional classification. The graphs in Table 4.16 and Table 4.17 show that the linear equations suggest service lives (time to reach (continued) (continued from page 70)

79 Table 4.11. Abutment Component Estimated Service Lives Component Service Life (years) Abutment Backwall 78.0 Abutment Bearings Steel 65.0 PTFE 51.4 Multirotational 57.0 Elastomeric 88.0 Abutment Joints Open 45.0 Compression 46.0 Modular 41.3 Armor 42.4 Sliding plate 66.4 Filled elastic 58.0 Preformed 37.0 Abutment Pedestal 72.5 Abutment Stem 81.0 Abutment Wingwall 79.0 Table 4.12. Pier Component Estimated Service Lives Component Service Life (years) Pier Bearings Steel 61.4 Multirotational 98.0 Elastomeric 69.0 Pier Cap Concrete 76.0 Masonry 84.5 Steel 83.4 Timber 81.6 Pier Cap Top Concrete 82.6 Masonry 93.0 Steel 91.7 Timber 84.8 Pier Column Concrete 77.4 Masonry 57.1 Steel 98.0 Timber 63.3 Pier Footing 79.0 Pier Joints Open 47.7 Strip seal 34.3 Compression 41.5 Modular 49.0 Armor 33.7 Sliding plate 37.2 Filled elastic 42.0 Preformed 35.6 Pier Pedestal Concrete 76.0 Masonry 78.3 Steel 91.4 Overall Pier Concrete 75.7 Masonry 96.6 Steel 78.7 Timber 66.0 Pier Stem 81.0 a condition rating of 3) significantly longer than 100 years, which is possible, but unlikely. The nonlinear equations are terminated at the minimum value; beyond this point the condition rating would appear to increase, which is not possible without maintenance. As only bridges without maintenance, repair, or rehabilitation were used in the analysis, the rating should not increase with increasing age. A new structure would have a condition rating of 9; all the prediction equations indicate the condition rating to be near 8 when new. Piecewise linear equations were determined for differ- ent functional classifications for the deck, superstructure, and substructure condition ratings. The coefficients B0, B1, B2, and B3 used in the piecewise linear equations are presented in Table 4.17. The condition rating is described by three linear equations that are applicable during cer- tain times of the bridge life; these equations are shown as Equation 4.1. In the study by Stukhart et al., t1 and t2 are defined as 25 and 45 years, respectively. Several of the graphs are terminated at 45 years as the results of the regression analysis indicate that the condition rating would increase, which cannot be true without maintenance being performed.

80 Component Service Life (years) Deck Curb Granite or stone 75.6 Steel plate 75.9 Timber 58.3 Concrete 66.9 Primary Member Slab, box, or box channel 67.8 Tee or I-beam 77.3 Rolled beam 76.7 Plate girder 82.9 Truss 56.9 Deck arch 65.7 Metal pipe arch 87.0 Frame 72.8 Box culvert 79.5 Pipe culvert 61.0 Overall Superstructure Slab 75.4 Multistringer or beam 76.0 Girder or floorbeam 76.6 Tee beam 75.6 Box beam or girder 68.9 Frame 77.4 Culvert 76.0 Through truss 65.5 Deck arch 77.9 Secondary Member Slab, box, or box channel 82.8 Tee or I-beam 53.2 Note: NA = not available. Component Service Life (years) Rolled beam 84.3 Plate girder 81.3 Truss 64.5 Deck arch NA Frame 98.0 Structural Deck CIP with black rebar 69.9 Precast plank 51.9 CIP with epoxy rebar 87.0 CIP with other coating 74.5 Timber 61.0 Steel grating 57.0 Steel plate 85.4 Wearing Surface Integral or monolithic Portland cement 57.7 Concrete with membrane 26.3 Class H concrete NA Portland cement overlay 53.0 Asphalt 48.0 Wood or wood block 37.8 Open steel grate 68.6 Sidewalk or Fascia Concrete 68.0 Steel plate 82.0 Asphalt concrete 59.0 Steel NA Table 4.13. Superstructure and Deck Estimated Service Lives if if if (4.1) 0 1 1 0 1 1 2 1 1 2 0 1 1 2 2 1 3 2 2 CR B B t t t B B t B t t t t t B B t B t t t t t t ( ) ( ) ( ) = + ≤ + + − < ≤ + + − + β − >      A nonlinear regression analysis was performed to deter- mine the parameters for the best-fit exponential decay curve. Parameters were determined for bridge decks and superstructures based on functional classification using the multiyear data set. The best-fit parameters and equations were used to estimate the service life of bridge decks and superstructures; for most cases, the estimated service lives were in excess of 150 years. Although this would seem like a good thing, it is known that most bridge decks and super- structures will not have a service life of this length. In fact, it is more likely that the service life of a bridge deck is closer to 40 or 50 years than 150 years. Although the estimated service lives seem extreme, the results are shown to indicate the avail- able data. The basic equation used to estimate the service life is shown as Equation 4.2; the graphs and parameters are shown in Table 4.18. The estimated service life is approxi- mately equal to the absolute value of b2 for this set of data. Looking at the values of b2, the only reasonable values are for (text continues on page 86)

81 Table 4.14. Transportation Systems Center Prediction Equations 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 R at in g Time (years) Deck Ratings Deck (ADT = 1000) Deck (ADT = 5000) Deck (ADT = 10000) Deck (ADT = 15000) CR = 9 - 0.119t - 2.158 × 10-6 (ADTAGE) 10 ADTAGE ADT AGE( )( ) = 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 R at in g Time (years) Substructure Ratings Substructure (ADT = 1000) Substructure (ADT = 5000) Substructure (ADT = 10000) Substructure (ADT = 15000) CR = 9 - 0.105t - 2.105 × 10-6 ADT 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Superstructure Ratings Superstructure (ADT = 1000) Superstructure (ADT = 5000) Superstructure (ADT = 10000) Superstructure (ADT = 15000) CR = 9 - 0.103t - 1.982 × 10-6 ADT Source: Stukhart et al. (1991).

82 Table 4.15. Coastal Substructure Condition Rating Prediction Equations 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 R at in g Time (years) Coastal Substructure Rating-Interstate Highways IH 1 IH 2 IH1 = 7.80 - 0.022t IH2 = 7.98 - 0.036t + 3.89 × 10-4t2 - 8.00 × 10-8 × t × ADT ADT assumed as 25,000 vehicles 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Coastal Substructure Ratings-US Highways US 1 US 2 US1 = 7.81 - 0.017t US2 = 7.94 - 0.028t + 2.02 × 10-4t2 - 6.00 × 10-8 × t × ADT ADT assumed as 15,000 vehicles 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Coastal Substructure Ratings-State Highways SH 1 SH 2 SH1 = 8.12 - 0.025t SH2 = 8.47 - 0.064t + 7.45 × 10-4 t2 - 2.20 × 10-7 × t × ADT ADT assumed as 10,000 vehicles (continued on next page)

83 Table 4.15. Coastal Substructure Condition Rating Prediction Equations 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 R at in g Time (years) Coastal Substructure Ratings-Farm-Market Highways FM 1 FM 2 Other FM1 = 8.11 - 0.028t FM2 = 8.18 - 0.032t + 1.84 × 10-6t3 Other = 7.93 - 0.034t Note: IH = Interstate highways or principal arterials; US = U.S. highways (non-Interstate) or minor arterials; SH = state highways or minor arterials; FM = farm-to-market roads or collectors. Source: Stukhart et al. (1991). Table 4.16. Substructure Condition Rating Prediction Equations by Region 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 R at in g Time (years) Coastal Substructure Ratings C1 C2 C3 C1 = 7.97 - 0.024t C2 = 8.14 - 0.040t + 3.31 × 10-4t2 - 9.00 × 10-8 × t × ADT C3 = 8.22 - 0.057t + 1.05 × 10-3t2 - 8.25 × 10-6t3 - 9.00 × 10-8 × t × ADT 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) East Texas Substructure Ratings ET1 ET2 ET1 = 8.20 - 0.037t ET2 = 8.20 - 0.012t - 1.58 × 10-3t2 + 2.24 × 10-5t3 (continued on next page) (continued) (continued)

84 Table 4.16. Substructure Condition Rating Prediction Equations by Region 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Inland Texas Substructure Ratings IT1 IT2 IT1 = 7.93 - 0.015t IT2 = 8.05 - 0.028t - 4.40 × 10-4t2 + 3.76 × 10-6t3 - 6.00 × 10-8 × t × ADT 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) West Texas Substructure Ratings WT1 WT2 WT1 = 7.85 - 0.015t WT2 = 8.27 - 0.059t - 1.22 × 10-3 t2 + 9.00 × 10-6 t3 - 4.40 × 10-7 × t × ADT 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 0 20 40 60 80 100 120 R at in g Time (years) Panhandle Region Substructure Ratings PH1 PH2 PH1 = 7.72 - 0.015t PH2 = 8.56 - 0.109t - 2.52 × 10-3 t2 + 1.62 × 10-5t3 - 3.60 × 10-7 × t × ADT (continued on next page) (continued)

85 Table 4.16. Substructure Condition Rating Prediction Equations by Region 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 R at in g Time (years) All Region Substructure Ratings AR1 AR2 AR1 = 7.98 - 0.023t AR2 = 8.21 - 0.043t - 3.68 × 10-4t2 - 5.00 × 10-8 × t × ADT Note: Equations are named according to region (C for coastline, ET for east Texas, and so forth) and are numbered in the order in which they are presented in Stukhart et al. (1991). Source: Stukhart et al. (1991). Mat. B0 B1 B2 B3 IH RC 8.17 -0.051 0.003 -0.046 IH Other 8.18 -0.025 0.004 -0.063 SFM RC 8.04 -0.029 -0.016 -0.034 SFM Other 8.06 -0.012 -0.004 -0.032 US RC 8.25 -0.056 0.018 -0.134 US ST 8.08 -0.016 0.004 -0.040 Table 4.17. Piecewise Linear Condition Rating Equations and Coefficients 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Deck Ratings IH Reinf. Conc IH Other ST/FM Reinf. Conc ST/FM Other US Reinf. Conc US Other 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 R at in g Time (years) Superstructure Ratings IH P/S Conc. IH Reinf. Conc IH Steel St/FM P/S Conc. ST/FM Reinf. Conc ST/FM Steel US P/S Conc. US Reinf. Conc US Steel Mat. B0 B1 B2 B3 IH PS 8.23 -0.029 -0.132 -0.184 IH RC 8.27 -0.036 -0.036 -0.036 IH ST 8.16 -0.056 0.002 -0.050 SFM PS 8.33 -0.033 -0.027 0.084 SFM RC 8.08 -0.016 -0.026 -0.013 SFM ST 8.08 -0.035 -0.016 -0.056 US PS 8.34 -0.038 -0.038 0.004 US RC 8.25 -0.034 -0.002 -0.125 US ST 8.17 -0.053 -0.016 -0.147 (continued on next page) (continued)

86 Table 4.17. Piecewise Linear Condition Rating Equations and Coefficients 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 R at in g Time (years) Substructure Ratings IH Reinf. Conc. IH Steel IH Other St/FM Reinf. Conc. ST/FM Steel ST/FM Other US Reinf Conc. US Steel US Other Note: IH = Interstate highways or principal arterials; RC = reinforced concrete; SFM = state farm-to-market road; US = U.S. highways (non-Interstate) or minor arterials; ST = state highways; PS = prestressed; SH = state highways or minor arterials; FM = farm-to-market roads or collectors; ST/FM = state highways and farm-to-market combined. Source: Stukhart et al. (1991). Mat. B0 B1 B2 B3 IH RC 8.23 -0.039 -0.018 -0.004 IH ST 8.46 -0.066 -0.066 -0.066 IH Other 8.30 -0.029 -0.029 -0.029 SFM RC 8.32 -0.039 -0.039 0.020 SFM ST 7.87 -0.028 -0.048 -0.094 SFM Other 8.23 -0.032 -0.032 -0.032 US RC 8.34 -0.046 0.007 -0.118 US ST 8.08 -0.033 -0.063 -0.063 US Other 8.43 -0.040 0.009 -0.109 Table 4.18. Exponential Best-Fit Graphs and Parameters 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Deck Ratings IH R/F Conc. IH Other US R/F Conc. US P/S. Conc. US Other State R/F Conc. State P/S Conc. State Other FM R/F Conc. FM P/S Conc. FM Timber FM Other Material b1 b2 IH RC 7.997 -198.662 IH Other 8.046 -718.254 US RC 7.882 -266.081 US PS 8.229 -255.085 US Other 7.980 -1295.727 SH RC 7.901 -316.403 SH PS 8.876 -66.876 SH Other 7.980 -937.262 FM RC 7.991 -330.937 FM PS 8.998 -59.482 FM Timber 7.032 -363.467 FM Other 8.046 -718.254 decks of prestressed concrete bridge on the state and farm-to- market highway systems. (4.2)1 2CR e t = β β The final method used to develop equations to predict con- dition ratings for the deck, superstructure, and substructure was a survey of Texas bridge engineers, who were asked to pro- vide estimates of the worst-case, the most likely, and the best-case expected remaining service life based on expert opin- ion. The expected remaining service life was based on a given condition rating: new (9), good (7), fair (5), and poor (3). From these responses, an estimated condition rating deterio- ration rate was determined. As would be expected with any opinion-based survey, there was significant variation in the responses; in several cases, the standard deviation was greater than the mean. The equations and graphs are shown in Table 4.19 for the deck, superstructure, and substructure condition ratings. (continued on next page) (continued) (continued from page 80)

87 Table 4.18. Exponential Best-Fit Graphs and Parameters 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Superstructure Ratings IH R/F Conc. IH P/S Conc. IH Steel US R/F Conc. US P/S Conc. US Steel State R/F Conc. State P/S Conc. State Steel FM R/F Conc FM P/S Conc. FM Steel Source: Stukhart et al. (1991). Material b1 b2 IH RC 8.177 -256.504 IH PS 8.269 -247.523 IH Steel 7.938 -184.626 US RC 8.155 -302.925 US PS 8.356 -204.955 US Steel 7.630 -371.819 SH RC 8.186 -332.936 SH PS 8.373 -217.760 SH Steel 7.929 -291.797 FM RC 8.194 -352.926 FM PS 8.187 -410.558 FM Steel 7.957 -289.826 (continued) Table 4.19. Expert Opinion Condition Rating Prediction Equations 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Deck Ratings Min Mean Max 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Superstructure Ratings RC Min RC Mean RC Max P/S Min P/S Mean P/S Max Steel Min Steel Mean Steel Max Min = 7.560 - 0.145t Mean = 7.758 - 0.115t Max = 7.655 - 0.083t Reinforced Concrete: Min = 7.642 - 0.136t Mean = 7.775 - 0.107t Max = 7.698 - 0.076t Prestressed Concrete: Min = 7.737 - 0.138t Mean = 7.707 - 0.099t Max = 7.633 - 0.076t Steel: Min = 7.752 - 0.146t Mean = 7.864 - 0.117t Max = 7.803 - 0.089t (continued on next page)

88 4.8 Massachusetts DOt The Massachusetts DOT conducted a study of its bridges to gain a better understanding of the dynamics of how bridges age and deteriorate. This knowledge is intended to be used to plan strategies for bridge work and to determine required lev- els of funding. The main aspects of this study were to determine the following: • The makeup of the bridge population by age and material of construction; • The average age of bridges for a given average condition rating by material; • The probability that a bridge in a given average condition rating will transition to a structurally deficient condition in the following year based on the age and current condi- tion of the bridge; • The percentage of bridges in each age group that are in one of the following categories: structurally deficient, fair, or satisfactory; and • Equations to predict the growth of bridges in the fair or satisfactory condition categories. In undertaking this analysis, the Massachusetts DOT defined the bridge condition categories as described here. A structurally deficient bridge was defined as a bridge with any one of the NBI Items 58, 59, or 60 (deck, superstructure, or substructure, respectively) condition ratings less than or equal to 4. A fair bridge was defined as a bridge with an aver- age condition rating of Items 58, 59, and 60 greater than 4 but less than or equal to 5, but with none of the individual condi- tion ratings being 4 or lower. Similarly, a satisfactory bridge was defined as a bridge with an average condition rating of Items 58, 59, and 60 greater than 5 but less than or equal to 6, but with none of the individual condition ratings being 4 or lower. Bridges with a condition rating average greater than 6 were considered excellent. Figure 4.13 shows the average age of a Massachusetts bridge in a given average condition rating. The average age for all bridges to reach an average condition rating of 1.0 is greater than that for either steel or concrete and can be attrib- uted to the age effect from old masonry bridges, the oldest of which in Massachusetts is 250 years old. By knowing the time it takes a bridge to deteriorate into the next lower average condition rating, the additional service life that could be obtained by increasing the average condi- tion rating can be estimated for a given preservation strategy. A regression analysis could be used to develop equations relating age to condition rating, as done in the previously 0 1 2 3 4 5 6 7 8 9 0 20 40 60 80 100 Br id ge C on di tio n Bridge Age Concrete Steel All Structurally Deficient Fair Satisfactory Figure 4.13. Massachusetts bridge conditions by age. Table 4.19. Expert Opinion Condition Rating Prediction Equations 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 20 40 60 80 100 120 Ra tin g Time (years) Substructure Ratings RC Min RC Mean RC Max P/S Min P/S Mean P/S Max Steel Min Steel Mean Steel Max Timber Min Timber Mean Timber Max Source: Stukhart et al. (1991). Reinforced Concrete: Min = 7.654 - 0.144t Mean = 7.740 - 0.107t Max = 7.701 - 0.081t Prestressed Concrete: Min = 7.710 - 0.131t Mean = 7.739 - 0.097t Max = 7.723 - 0.073t Steel: Min = 7.881 - 0.177t Mean = 7.866 - 0.138t Max = 7.883 - 0.105t Timber: Min = 7.535 - 0.202t Mean = 7.846 - 0.174t Max = 7.992 - 0.140t (continued)

89 presented studies. The same process could be used to develop additional equations based on ADTT, location, or owner if the required data were available. The average condition rating versus age for Massachusetts bridges is similar to those pre- sented above for other states. However, the purpose of the Massachusetts DOT study was not to develop equations to predict the condition rating as a function of time (or age) as the studies presented previously, though if desired, equations could be developed using the available data. Instead, using the data acquired for the third and fifth bullet points above, the number of bridges that become structurally deficient in any given year can be esti- mated from the number of bridges predicted to be in a given condition (i.e., satisfactory or fair). The analysis results indicated that, for the bridge population as a whole, approximately 4.25% of bridges in fair condition transition to structurally deficient the following year (see Figure 4.14). Similarly, Figure 4.15 indicates that approxi- mately 1.11% of bridges in satisfactory condition transition to structurally deficient the following year. Figure 4.14 and Fig- ure 4.15 also show the transition probabilities for steel bridges and concrete bridges; these probabilities could be used if the analyst wished to look only at steel or concrete bridges. The graphing of these transition probabilities indicates that, except for concrete bridges in the fair category, which show some age- related influence, age is not as much of a factor as the current condition category of the bridge in determining the transition probability. Similar transition probabilities could be devel- oped for different geographic regions or different levels of ADT or ADTT. To predict the growth of bridges in the fair and satisfactory categories, best-fit equations were developed from a regression 0.00% 2.00% 4.00% 6.00% 8.00% 10.00% 12.00% 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 10 0 10 5+ % o f F ai r Br id ge s Be co m in g St ru ct ur al ly De fic ie nt Age Group Fair Fair-Conc Fair-Steel Figure 4.14. Probability of all bridges in fair condition becoming structurally deficient in the following year. 0.00% 0.50% 1.00% 1.50% 2.00% 2.50% 3.00% 3.50% 4.00% 4.50% 5.00% 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 10 0 10 5+ % o f S ati sf ac to ry B rid ge s Be co m in g St ru ct ur al ly D efi cie nt Age Group Satisfactory Satisfactory-Conc Satisfactory-Steel Figure 4.15. Probability of all bridges in satisfactory condition becoming structurally deficient in the following year.

90 analysis of the number of bridges that were in those two categories for each year from 2002 through 2009. These equa- tions were used to predict the number of fair and satisfac- tory bridges in future years. These equations are graphed in Figure 4.16 and Figure 4.17 and show the predicted numbers compared with the actual numbers from 2002 to 2009. Due to concerns that the regressed exponential equation for the growth of fair bridges was too aggressive, the Massa- chusetts DOT decided to use the number of fair bridges that would be obtained by averaging the number of fair bridges predicted by the best-fit exponential equation and the best-fit straight line equation for further analysis. After applying the transition probability for fair bridges, the growth in structurally deficient bridges from this category is shown in Figure 4.18 for each of the regression equations, as well as from the average. A final needs analysis spreadsheet was developed that combined the structurally deficient growth predictions and the predictions of the number of projects that could be undertaken for a given amount of funding. The number of 1540 1560 1580 1600 1620 1640 1660 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 1588 1603 1615 1640 1627 1653 1635 1639 1588 1605 1615 1622 1628 1632 1636 1639 1642 1645 1647 1649 Co un t Satisfactory Bridges Growth Actual "Predicted" Figure 4.16. Growth of bridges in satisfactory category. 0 200 400 600 800 1000 1200 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 496 521 535 567 642 704 716 757 479 511 546 583 622 664 709 757 808 863 921 983 476 516 556 597 637 677 717 758 798 838 879 919 Co un t Fair Bridge Growth "Actual" "Predicted Exp" "Predicted Lin" Figure 4.17. Growth of bridges in fair category.

91 structurally deficient bridges for a given year was estimated by multiplying the transitional probability by the predicted number of fair or satisfactory bridges for that year. The cost model was calibrated with actual project costs and considered the costs for a full replacement versus a preserva- tion project. Replacement projects assumed the replacement of an already structurally deficient bridge and hence resulted in a reduction in the number of structurally deficient bridges estimated for the following year. It was assumed that preser- vation projects were to be performed on fair bridges and that only a percentage of bridges, based on the transitional prob- ability times the number of preservation projects undertaken, would be prevented from becoming structurally deficient in the following year. However, the total number of bridges that had preservation work done were removed from the fair con- dition rating population in calculating structurally deficient bridges for subsequent years. Figure 4.19 and Figure 4.20, respectively, show the effect of different funding levels on the number of bridges that will be structurally deficient or in fair condition. The funding levels, shown in the far-left column of Figure 4.19 and Figure 4.20, are normalized to the lowest funding level shown in the top row. Figure 4.19 shows the overall bridge program funding level and includes both replacement and preservation projects. 0 10 20 30 40 50 60 70 80 90 100 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 57 59 61 63 64 66 68 70 71 73 75 57 60 62 64 67 70 73 76 80 83 87 57 59 61 64 66 68 70 73 75 78 81 Co un t Growth in SD Bridges from Fair Bridges Linear Exponential Average Figure 4.18. Number of fair bridges becoming structurally deficient (SD). 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 $1.00 541 545 559 553 521 494 471 467 473 484 500 526 553 585 620 660 $1.59 541 545 559 553 521 494 463 445 434 424 414 412 407 404 402 402 $1.93 541 545 559 553 521 494 452 421 397 373 348 329 307 287 266 248 0 100 200 300 400 500 600 700 C o u n t YEAR SD TRENDS BASED ON CAPITAL AND PRESERVATION SPENDING, 4% INFLATION Figure 4.19. Number of structurally deficient bridges based on various spending levels.

92 The second and third rows of Figure 4.19 assume that 1.59 and 1.93 times as much money is available for structurally deficient bridges. The 1.53 funding level will trend to a steady state number of structurally deficient bridges. The 1.93 fund- ing level will reduce the number of structurally deficient bridges at a rate that will result in zero structurally deficient bridges in 20 years. In Figure 4.20, the spending levels corre- late to the preservation spending for each of the total funding levels in Figure 4.19. As expected, spending more money leads to fewer bridges that are structurally deficient or in fair condition and spend- ing less money leads to more bridges that are either structur- ally deficient or in fair condition. This graph also indicates the level of funding that would be needed to achieve a given desired outcome. For example, to achieve a net annual reduc- tion in the number of structurally deficient bridges for the long term, program funding above the 1.59 level is needed. In addition to the preceding analysis, the Massachusetts DOT also developed a utility to rank all the bridges in the state to prioritize which bridges should be worked on first. This ranking methodology is being used to develop the bridge State Transportation Improvement Program lists. The rank- ing is a function of three values: the condition loss value, the change in health index, and the highway evaluation factor. Condition loss is simply the difference between a perfect con- dition rating (9.0) and the current average condition rating divided by nine and multiplied by 100 to achieve a percentage value. Health index is the change in the bridge’s health index that AASHTOWare Bridge Management (formerly Pontis) predicts will occur over 15 years, expressed as a percentage value. The health index is calculated using the current CoRe Element condition state for an existing bridge versus that of a new bridge, as given in Equation 4.3 (Thompson and Shepard 2000). Current element value is calculated using Equation 4.4, and total element value is calculated using Equation 4.5. The CoRe elements and associated condition states can be found in the AASHTO Guide Manual for Bridge Element Inspection (2011). Health Index HI CEV TEV 100 (4.3) ∑ ∑( ) = × where CEV is current element value and TEV is total element value. CEV Quantity in Condition State FC (4.4) i WF i∑( )[ ]( )= × × where WF(i) is the condition state weighting factor given in Table 4.20 and FC is the failure cost of the element. TEV Total Element Quantity FC (4.5)= × Figure 4.20. Number of fair bridges based on various spending levels. 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 $1.00 803 811 813 822 829 858 894 936 979 1028 1081 1138 $3.64 803 811 754 707 660 638 623 618 614 619 630 646 $4.48 803 811 735 670 606 567 536 515 497 488 484 487 0 200 400 600 800 1000 1200 C o u n t YEAR FAIR BRIDGE GROWTH TRENDS BY PRESERVATION SPENDING, 4% INFLATION Table 4.20. Condition State Weighting Factors Number of Condition States State 1 State 2 State 3 State 4 State 5 3 1.00 0.50 0.00 4 1.00 0.67 0.33 0.00 5 1.00 0.75 0.50 0.25 0.00

93 The highway evaluation factor is a measure of the func- tionality of the bridge and considers the ADT, detour length, functional classification, load-carrying restrictions, and deck geometry deficiencies. The categories within each variable are given a value between 1 and 5; the average value for the five variables is determined and then divided by five and multi- plied by 100 to achieve a percentage value. The values for con- dition loss (CL), health index (HI), and highway evaluation factor (HEF) are then combined using Equation 4.6 to deter- mine the final ranking factor for each bridge: Ranking Factor 0.3CL 0.4HI 0.3HEF (4.6)= + + The ranking factor is then used to sort the bridges to determine each bridge’s overall rank within the Massachu- setts bridge population and, hence, its priority for work; bridges with the highest ranking factor values are those that require repair or maintenance in the future. The ranking is not a set order (Bridge 2 can go before Bridge 1) but, in gen- eral, higher-ranked bridges should be improved before lower-ranked bridges. The ranking factor, if calculated over a number of years, may lead to a reasonable estimate of the amount of deterioration and possible loss of serviceability for a bridge.