The National Academies Press

Currently Skimming:

Pages 89-123

The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.

From page 89... ... 89 Many of the objectives of life expectancy analysis go beyond the simple calculation of life span. Agencies that gather the necessary data and perform the analysis can benefit in many more ways by constructing useful applications that go farther, to help in developing and selecting policies, planning future work programs, and developing cost-effective designs and projects. Read the entire page →
From page 90... ... 90 estimating Life expectancies of highway assets Deterministic models were popular before 1960 when they were developed by the AASHO Road Test (before AASHO became AASHTO) for pavements (Patterson 1987) Read the entire page →
From page 91... ... Develop applications: how to apply Life expectancy Models 91 Subsequent research efforts have developed various forms of non-linear deterioration curves that resembled the AASHO curve and could include explanatory variables such as traffic and climate, but which still can be estimated using simpler linear regression techniques. Life expectancy could be read off the curve where it intersected the minimum tolerable condition level or by inverting the equation to make age a function of condition. Read the entire page →
From page 92... ... 92 estimating Life expectancies of highway assets The model's adjusted R2 was found to be 0.30 with significant autocorrelation problems, suggesting that linear regression is not an appropriate life expectancy method for the given data in the example application. 5.1.2 Markov Models Most common asset management processes use categorical data, which classify condition into a relatively small number of categories. Read the entire page →
From page 93... ... Develop applications: how to apply Life expectancy Models 93 A transition probability matrix can have as few as two condition states, such as pass/fail. It commonly has four or five states for most types of assets. Read the entire page →
From page 94... ... 94 estimating Life expectancies of highway assets Another way to detect possible repair activity is to look for improvements in condition. The following formula can be calculated for each inspection pair: IC y xj k k j k k j = −     = = ∑ ∑max 1 1 where IC is the improvement in condition for the inspection pair; j and k are condition states defined for the asset that was inspected (assuming that k = 1 is the best possible condition state) Read the entire page →
From page 95... ... Develop applications: how to apply Life expectancy Models 95 are less than 0.01, they are changed to 0.01 (or some other small positive value) Read the entire page →
From page 96... ... 96 estimating Life expectancies of highway assets Inspection pairs Condition - start of year Condition - end of year Improvement in condition Road Insp Condition state Condition state Condition state segment Year 1 2 3 4 1 2 3 4 1 2 3 4 Screening RS0028 2004 92 8 0 0 82 17 1 0 -10 -1 0 0 RS0028 2005 82 17 1 0 68 27 4 1 -14 -4 -1 0 RS0028 2006 68 27 4 1 58 32 9 1 -10 -5 0 0 RS0028 2007 58 32 9 1 48 37 11 4 -10 -5 -3 0 RS0028 2008 48 37 11 4 46 35 12 7 -2 -4 -3 0 RS0028 2009 46 35 12 7 37 39 14 10 -9 -5 -3 0 RS0028 2010 37 39 14 10 32 37 19 12 -5 -7 -2 0 RS0061 2005 100 0 0 0 84 16 0 0 -16 0 0 0 RS0061 2006 84 16 0 0 78 19 3 0 -6 -3 0 0 RS0061 2007 78 19 3 0 67 27 5 1 -11 -3 -1 0 RS0061 2008 67 27 5 1 65 23 10 2 -2 -6 -1 0 RS0061 2009 65 23 10 2 55 28 12 5 -10 -5 -3 0 RS0061 2010 55 28 12 5 53 24 15 8 -2 -6 -3 0 RS0035 2004 83 10 5 2 75 17 5 3 -8 -1 -1 0 RS0035 2005 75 17 5 3 68 20 7 5 -7 -4 -2 0 RS0035 2006 68 20 7 5 63 24 7 6 -5 -1 -1 0 RS0035 2007 63 24 7 6 52 32 7 9 -11 -3 -3 0 RS0035 2008 52 32 7 9 43 36 10 11 -9 -5 -2 0 RS0035 2009 43 36 10 11 37 39 11 13 -6 -3 -2 0 RS0035 2010 37 39 11 13 33 34 18 15 -4 -9 -2 0 RS0011 2005 29 21 18 32 25 22 17 36 -4 -3 -4 0 RS0011 2006 25 22 17 36 24 18 18 40 -1 -5 -4 0 RS0011 2007 24 18 18 40 100 0 0 0 76 58 40 RS0011 2008 100 0 0 0 83 17 0 0 -17 0 0 0 RS0011 2009 83 17 0 0 77 22 1 0 -6 -1 0 0 RS0011 2010 77 22 1 0 73 23 3 1 -4 -3 -1 0 RS0001 2003 100 0 0 0 86 14 0 0 -14 0 0 0 RS0001 2004 86 14 0 0 75 22 3 0 -11 -3 0 0 RS0001 2005 75 22 3 0 63 31 5 1 -12 -3 -1 0 RS0001 2006 63 31 5 1 62 26 10 2 -1 -6 -1 0 RS0001 2007 62 26 10 2 51 33 12 4 -11 -4 -2 0 RS0001 2008 51 33 12 4 49 33 10 8 -2 -2 -4 0 RS0001 2009 49 33 10 8 42 36 11 11 -7 -4 -3 0 RS0001 2010 42 36 11 11 35 36 15 14 -7 -7 -3 0 RS0004 2006 24 18 15 43 21 18 15 46 -3 -3 -3 0 RS0004 2007 21 18 15 46 18 17 15 50 -3 -4 -4 0 RS0004 2008 18 17 15 50 16 15 15 54 -2 -4 -4 0 RS0004 2009 16 15 15 54 90 10 0 0 74 69 54 0 Delete 0 Delete RS0004 2010 90 10 0 0 79 19 2 0 -11 -2 0 0 RS0016 2006 81 14 4 1 76 18 4 2 -5 -1 -1 0 RS0016 2007 76 18 4 2 62 30 5 3 -14 -2 -1 0 RS0016 2008 62 30 5 3 59 29 7 5 -3 -4 -2 0 RS0016 2009 59 29 7 5 55 31 8 6 -4 -2 -1 0 RS0016 2010 55 31 8 6 51 28 13 8 -4 -7 -2 0 Change in condition for segments where no work done Condition at start Condition at end 1 2 3 4 1 2 3 4 Avg by state 62.6 22.6 7.0 7.9 55.4 26.2 8.8 9.6 Computed transition probabilities using One-Step Method Condition state probabilities 1 2 3 4 Stay in same state 88.5 84.2 74.7 100 Deteriorate one step 11.5 15.8 25.3 0.0 Table 5-2. Example of the one-step method of estimating Markov models. Read the entire page →
From page 97... ... Develop applications: how to apply Life expectancy Models 97 The second section contains the [X] Read the entire page →
From page 98... ... 98 estimating Life expectancies of highway assets Agrawal and Kawaguchi 2009) Read the entire page →
From page 99... ... Develop applications: how to apply Life expectancy Models 99 µ = threshold parameters, which in comparison to parameter estimates and variable values, indicate the likelihood of being in a given condition state; [µ - S bx] = X value that can be used to calculate normal distribution test statistic via Z X Mean S dard Deviation = − tan N(0, 1) Read the entire page →
From page 100... ... 100 estimating Life expectancies of highway assets In other words, the culvert in this specific example is considered to have a 0.73% chance of being in condition state 0, an 8.48% chance of being in condition state 1, a 31.98% chance of being in condition state 2, and a 58.81% chance of being in condition state 3. Therefore, the most likely condition state for this asset is condition state 3. Read the entire page →
From page 101... ... Develop applications: how to apply Life expectancy Models 101 (i.e., prior means) Read the entire page →
From page 102... ... 102 estimating Life expectancies of highway assets One way to minimize the complication is to use the Markov prediction formula iteratively until the 50% failure criterion is reached. As long as the asset has not already reached its life expectancy, the remaining life can be determined in this way and then subtracted from the life expectancy to compute equivalent age. Read the entire page →
From page 103... ... Develop applications: how to apply Life expectancy Models 103 2. Convert each row of the matrix to median transition time, using the familiar formula t p j jj = ( ) Read the entire page →
From page 104... ... 104 estimating Life expectancies of highway assets 3. Each condition state will be allocated a portion of the asset's life, in proportion to its transition time. Read the entire page →
From page 105... ... Develop applications: how to apply Life expectancy Models 105 5.2.3 Remaining Life One of the most obvious ways to compute remaining asset life is to subtract the actual age of an asset from its life expectancy. This method is valid if no life extension actions have been performed. Read the entire page →
From page 106... ... 106 estimating Life expectancies of highway assets 5.2.4 Lifecycle Cost Models Adding lifecycle cost to life expectancy and deterioration models opens the door to a wealth of useful applications to support transportation asset management decision-making. Among the possible applications are the comparison of design and life extension alternatives, optimizing replacement intervals, optimizing preventive maintenance, evaluating new maintenance materials and techniques, optimizing corridor planning, and responding effectively to funding constraints. Read the entire page →
From page 107... ... Develop applications: how to apply Life expectancy Models 107 Figure 5-13. Example of bridge lifecycle cost alternatives. Read the entire page →
From page 108... ... 108 estimating Life expectancies of highway assets The blanket replacement policy saves one million dollars by reducing the mobilization and traffic control costs of unplanned traffic signal failures. The optimal time for replacement depends on the width of the probability distribution, which is the level of uncertainty in the median failure time. Read the entire page →
From page 109... ... Develop applications: how to apply Life expectancy Models 109 Here FV is again the amount of the future recurring payment, starting 1 year from the present. If the stream of cash flows corresponds to the life span of an asset, then n is typically the median life expectancy of the asset. Read the entire page →
From page 110... ... 110 estimating Life expectancies of highway assets 5.2.4.4 Comparing Alternatives Using Equivalent Uniform Annual Cost For certain purposes and certain audiences, it is useful to compare alternatives by converting net present value to equivalent uniform annual cost (EUAC) Read the entire page →
From page 111... ... Develop applications: how to apply Life expectancy Models 111 costs. In any event, it has a lower cost than do-something. Read the entire page →
From page 112... ... 112 estimating Life expectancies of highway assets 5.3.1 Routine Preventive Maintenance A common maintenance planning issue is the question of whether to start routine programs of crew activities that might have life extension benefits. Common examples are sealing of pavement cracks; washing of bridges, signs, pavements, and guiderails; spot painting; concrete patching; and cleaning of equipment enclosures. Read the entire page →
From page 113... ... Develop applications: how to apply Life expectancy Models 113 With these assumptions, the agency could reduce annual costs by $172 per lane-mile if routine preventive maintenance is completed. 5.3.2 Optimal Replacement Interval Certain types of assets may have various asset life alternatives, depending on different strategies for maintenance and life extension. Read the entire page →
From page 114... ... 114 estimating Life expectancies of highway assets 5.3.3 Comparing and Optimizing Design Alternatives It is a very common need to compare two products or methods that have different costs, different life expectancies, and different life extension possibilities. Here is an example, considering the case of deciding to apply a coating to a pipe culvert. Read the entire page →
From page 115... ... Develop applications: how to apply Life expectancy Models 115 EUAC of Deck Overlay = $15k 1 1+0.04 ∗ ( ) Read the entire page →
From page 116... ... 116 estimating Life expectancies of highway assets From this array of activity options, the improvement strategy that minimizes cost under these assumptions is annual deck overlay and joint replacement. It can also be seen that the life extensions from patching, joint replacement, and rehabilitation under these assumptions are not cost-effective. Read the entire page →
From page 117... ... Develop applications: how to apply Life expectancy Models 117 Consider a small system of assets located along the same roadway (Table 5-7) Read the entire page →
From page 118... ... 118 estimating Life expectancies of highway assets This optimization problem can be solved using a solver software package, although it is simple enough to solve by inspection, recognizing that • Ideally an agency would like to coordinate replacements so as to minimize cost. • The new construction asset life estimates have a common multiple of 5 years. Read the entire page →
From page 119... ... Develop applications: how to apply Life expectancy Models 119 5.3.8 Value of Life Expectancy Information For some of the asset types described in this guide, an agency might not have any data collection processes at all and no way to implement a condition-responsive replacement or life extension program to optimize life expectancy. Usually the cost of data collection is a major barrier to improvement. Read the entire page →
From page 120... ... 120 estimating Life expectancies of highway assets Once the lifecycle costs of different statistical models, as well as the expert opinion, were converted into present worth at perpetuity, those results were plotted against the number of variables in Figure 5-17. The plot suggests that the total cost declines with an increasing number of variables used in the performance model to predict an asset's life, provided that the added variables enabled an extension of asset life for selected assets as shown in Table 5-9. Read the entire page →
From page 121... ... Develop applications: how to apply Life expectancy Models 121 example involves a box culvert built in 2002 and having a culvert condition rating of 8. The culvert is 8 years of age. Read the entire page →
From page 122... ... 122 estimating Life expectancies of highway assets • Become familiar with available methods and tools as described in Chapters 4 and 5 of this guide. • Evaluate possible additional applications and recruit users who may want to see such applications developed. Read the entire page →
From page 123... ... Develop applications: how to apply Life expectancy Models 123 Production • Oversee and attend training classes, for new users and applications, and refresher courses for existing users. • Provide constructive input on new functionality that may be needed. Read the entire page →

From page 89...

... 89 Many of the objectives of life expectancy analysis go beyond the simple calculation of life span. Agencies that gather the necessary data and perform the analysis can benefit in many more ways by constructing useful applications that go farther, to help in developing and selecting policies, planning future work programs, and developing cost-effective designs and projects.