3. AGREEMENT BETWEEN PMIS OUTPUT AND ACTUAL
3.2 Project Agreement Analysis
3.2.1 Project Agreement Analysis for Bryan District
Table 3.1 illustrates the construction from FY 2007 compared with the Needs Estimate from FY 2006.
Table 3.1 indicates that 629 sections received some type of preservation action from a construction project let in FY 2007. Of these sections, approximately 42 percent had Needs Nothing (NN) suggestions generated by the FY 2006 Needs Estimate. The obvious question is why would TxDOT repair or improve pavement sections that its own management system indicates require no repair or improvement? The answer to this is twofold. First, a proactive pavement management program will inevitably result in spending money where a problem has not yet manifested to ensure that the system remains at or above an acceptable level. This is especially true with the seal coat program throughout Texas. Second, TxDOT lets construction projects, not construction sections. PMIS provides information on pavement sections, usually a ½-mile in length, and then aggregates those sections to create information about the network. Category 1 projects often span several miles and therefore multiple sections within PMIS. From the district level standpoint, it is important to work in terms of projects, not just sections.
Bridging the gap between network-level (section) and project-level is where new decision making tools are needed to help pavement managers.
Table 3.1 continues the illustration of the FY 2006 needs estimate preservation suggestions with a more detailed evaluation of FY 2007 projects. The lengths shown in Table 3.2 are approximate and are based on the average section length corresponding to the project length provided in the contract description and the number of sections affected according to PMIS. This approximation results in sections lengths of approximately 0.48-mile, an expected result because most of the sections within PMIS are ½-mile in length. Tables 3.2 and 3.3 provide similar information for FY 2008 and FY 2009 construction.
For FY 2007 construction (Table 3.1), the 65 projects consisted of 56 seal coat projects, 3 restoration projects, and 6 grading/structures/base/surface projects, spread over 629 sections, 262 of which received an NN suggestions from the needs estimate.
Of the 262 NN sections, 257 were part of seal coat projects. Seal coat projects often contain a significant number of NN sections. In FY 2008 there were 319 NN sections receiving a seal coat, there were only 345 NN sections that received any sort of preservation work. The trend continues for FY 2009 construction where 249 NN sections received a preservation action, 219 of these received a seal coat.
Table 3.1. 2007 Construction Project Description and Needs Estimate Suggestions (Bryan)
Project Type
# of
Projects NN PM LRhb MRhb HRhb Total
Length (mi) 120.8 76.6 38.5 18.3 0.5 254.7
# of Sections 257 163 82 39 1 542
Length (mi) 123.2 93.8 47.4 29.2 3.9 297.5
% of Length 41% 32% 16% 10% 1%
Table 3.2. 2008 Construction Project Description and Needs Estimate Suggestions (Bryan)
Project Type
# of
Projects NN PM LRhb MRhb HRhb Total
Length (mi) 149.9 32.0 12.2 5.6 0.5 200.2
# of Sections 319 68 26 12 1 426
Length (mi) 162.1 56.6 13.7 7.1 0.5 240.0
% of Length 68% 24% 6% 3% 0%
Table 3.3. 2009 Construction Project Description and Needs Estimate Suggestions (Bryan)
Project Type
# of
Projects NN PM LRhb MRhb HRhb Total
Length (mi) 102.9 126.0 13.6 8.5 1.9 252.9
# of Sections 219 268 29 18 4 538
Length (mi) 110.6 140.1 32.5 26.4 8.6 318.2
% of Length 35% 44% 10% 8% 3%
Through an interview with Bryan district personnel, it was discovered that Bryan has tried to develop a seal coat cycle by using the previous 10 years’ seal coat plans, thus helping to explain why and how this occurred. Seal coat projects will not be evaluated further for project agreement in the Bryan district; instead the analysis will focus on heavier preservation treatments. The obvious question with the exclusion of seal coat projects becomes, does the needs estimate agree with construction projects of a heavier type? The answer is, not always. Table 3.4 shows preservation types suggested by the
Needs Estimates tool for the pavement sections within the boundaries of three grading/structure/base/surface projects from 2008 construction. The PMIS versus actual projects agreement statistics are shown in Table 3.5.
Table 3.4. Needs Estimate for 2008 Construction Projects FM 39, SH 7, and US 190
FM0039 0388 +01.0 0388 +01.5 A999 NN
FM0039 0388 +01.5 0390 +00.0 A999 NN
FM0039 0390 +00.0 0390 +00.5 A999 NN
FM0039 0390 +00.5 0390 +01.0 A999 NN
FM0039 0390 +01.0 0390 +01.5 A999 NN
FM0039 0390 +01.5 0392 +00.0 A999 NN
FM0039 0392 +00.0 0392 +00.5 A999 NN
FM0039 0392 +00.5 0392 +01.0 A999 NN
FM0039 0392 +01.0 0392 +01.5 A105 MRhb
FM0039 0392 +01.5 0394 +00.0 A999 NN
FM0039 0394 +00.0 0394 +00.5 A999 NN
FM0039 0394 +00.5 0394 +01.0 A999 NN
FM0039 0394 +01.0 0394 +01.5 A105 MRhb
FM0039 0394 +01.5 0396 +00.0 A610 PM
FM0039 0396 +00.0 0396 +00.5 A999 NN
FM0039 0396 +00.5 0396 +01.0 A705 PM
FM0039 0396 +01.0 0396 +01.5 A999 NN
FM0039 0396 +01.5 0396 +02.0 A999 NN
SH0007 0624 +00.0 0624 +00.5 A999 NN SH0007 0624 +00.5 0624 +01.0 A999 NN
SH0007 0624 +01.0 0624 +01.5 A615 PM
US0190 0744 +00.0 0744 +00.5 A999 NN US0190 0744 +00.5 0744 +01.0 A999 NN US0190 0744 +01.0 0744 +01.5 A999 NN
US0190 0744 +01.5 0746 +00.0 A520 PM
US0190 0746 +00.0 0746 +00.5 A999 NN US0190 0746 +00.5 0746 +01.0 A999 NN
US0190 0746 +01.0 0746 +01.5 A520 PM
US0190 0746 +01.5 0748 +00.0 A520 PM
US0190 0748 +00.0 0748 +00.5 A999 NN
US0190 0748 +00.5 0748 +01.0 A520 PM
US0190 0748 +01.0 0748 +01.5 A999 NN
Table 3.5. FM 39, SH 7, & US 190 Construction Summary Project
HWY
Cost ($,
million) NN PM LRhb MRhb HRhb Total
FM 39 3.4 Length (mi) 7 1 0 1 0 9
# of Sections 14 2 0 2 0 18
% in Length 78% 11% 0% 11% 0% 100%
SH 7 1.2 Length (mi) 1 0.5 0 0 0 1.5
# of Sections 2 1 0 0 0 3
% in Length 67% 33% 0% 0% 0% 100%
US 190 0.47 Length (mi) 3.5 2 0 0 0 5.5
# of Sections 7 4 0 0 0 11
% in Length 64% 36% 0% 0% 0% 100%
The question becomes why a $3.4 million project has almost 80% of its sections as Needs Nothing? One answer includes, the Needs Estimate tool does not accurately estimate preservation needs; or maintenance has been performed within the project limits that is represented in the Needs Estimate suggestions, while in reality the root of the problem has not been corrected. Based on the district interviews, it was determined that the Needs Estimate tool in PMIS does not completely account for all factors that the decision makers consider when making pavement preservation decisions. The lack of agreement between actual decisions and the Needs Estimate tool highlights the need to investigate improvements to decision support tools available to TxDOT.
TxDOT districts are broken into area offices and within area offices there are maintenance sections. Each of these sections has employees that perform various maintenance activities throughout their respective section. Some of these maintenance activities include pavement repair to keep the pavement in a condition acceptable to the
general public. This type of maintenance could affect the information in PMIS and cause the needs estimate to change. It is hypothesized that only looking one-year prior to construction could discount routine maintenance and mask a larger problem that needs attention. For that reason, before continuing the analysis, the researcher generated needs estimates for as far back as five years prior to construction. With this information it is assumed that routine maintenance will be accounted for and the true problem spots will rise to the top, so to speak. Table 3.6 summarizes 10 construction projects completed in the Bryan district in FY 2007, 2008, and 2009.
Table 3.6. Bryan District Construction Projects in FY 2007-2009
HWY many policy and funding factors, the percent sections with NN needs estimate for the 10
projects listed above was computed using pavement condition data from one year, three years, and five years prior to the actual construction year (see Figure 3.1). In six of the 10 projects, the percent of NN sections increases when moving from 1-year prior to 3-years prior. For all 10 projects, the percent of NN sections 5-3-years prior to construction was greater than 3-years prior to construction. In fact, 5-years prior to construction have the largest percent of NN sections for nine of the 10 projects, with the only exception occurring on FM 39, let in FY 2008. For SH 7, let in FY 2008, the 1-year and 5-year NN sections are equal. This increasing trend in %NN indicates that, in Bryan district, it takes one or two years to let a project after the pavement is identified as in need of maintenance or rehabilitation (based on condition). For the 1-year prior period, the percent of NN sections within Bryan district projects ranged between zero (perfect agreement between PMIS output and the actual project) to 77.8% (poor agreement between PMIS output and the actual project). On average, the percent of NN sections within Bryan district for all projects, across all years, 1-year prior to construction was 45 percent. For projects analyzed further (Figure 3.1) the average percent of NN sections for 1, 3, and 5 years prior to actual project construction year was 33 percent, 31 percent, and 42 percent, respectively.
Figure 3.1. Percent NN Sections with Sample Construction Projects in Bryan District