Practices for Establishing Contract Completion Dates for Highway Projects (2017)

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.

22 Following initial analysis of the survey results, several states were identified for potential follow-up discussions with the purpose of gleaning additional details and depth regarding their practices in devel- oping estimates of contract time. The criteria used to select these STAs involved their willingness to participate in follow-up interviews and how they responded to specific questions of the survey. The goal was to query a cross section of states using a variety of tools, whether they be custom developed or off-the-shelf, as well as having varied responses regarding revision procedures, accuracy, and the specific business unit responsible for preparing the contract time estimates. The states selected for case examples are listed in Table 3. The designated STAs New Jersey, Oklahoma, Ohio, and Kentucky were contacted and interviews scheduled. The findings of these interviews and the examination of their procedures and tools are provided in the following sections. NEW JERSEY CONTRACT TIME ESTIMATION PROCEDURE The New Jersey DOT develops individual critical path schedules for projects using Primavera templates. These templates require the designer to develop a work breakdown structure (WBS) with common codes for activities and responsible parties and production rates selected from project type tables, and adjusted as needed. This type of contract time estimation method is indicative of a non-automated procedure in use at many STAs and relying on the estimator to create a project-specific construction schedule with no default project logic. In New Jersey procedures, schedules are developed by the designer according to the standards presented in the 2013 Construction Scheduling Manual. The manual provides guidance on the standards for developing the schedule and WBS, default production rates for roadbuilding activi- ties, and approval and review procedures for the construction duration. The project components are defined using a WBS, which is coded to ensure uniformity across projects. The agencies use a Construction Scheduling Template (Figure 21) in Primavera that provides formatting and default coding for different activity types, but does not include generic project logic. The manual contains guidance on many considerations for designers as they are preparing the contract time estimate, including seasonal letting accommodations, weather delays, accelerated construction schedules, accounting for incentives and disincentives, and A+B bidding. There are also some provisions provided for design-build projects, although the interviewee noted that New Jersey does not currently use design-build projects. The area for improvement suggested in the inter- view was to examine the production rates within the procedure for their accuracy to existing construction methods. The designer uses guidance from the manual to account for the duration from final design to award and then is able to make an estimate of a completion date. The designer’s prepared sched- ule and estimate of contract time are provided to the construction manager and other STA staff for feedback. Once the project is awarded, the contractor uses the Contract Scheduling Manual to develop the project schedule akin to the process used by the designer. The contractor’s initial schedule becomes the baseline schedule for the project and is updated biweekly throughout the project. chapter four CONTRACT DURATION CASE EXAMPLES

23 OKLAHOMA CONTRACT TIME DETERMINATION PROCEDURE The procedure in use at the Oklahoma DOT is an example of the type of integrated automated sched- uling procedure in place at many STAs. This procedure uses a custom developed user interface and a Microsoft Access database (Figure 22) of default project type templates and production rates that is then linked with Microsoft Project to create a project schedule (Figure 23). To generate a schedule estimate the user selects a project type template, has the ability to adjust the simple finish-to-start logic of the activities in the template, and inputs bid item quantities for different activities. The default production rates and bid item quantities are used to calculate durations for the project activi- ties, although these may also be user adjusted. This information is then used by Microsoft Project to generate a construction schedule using a predetermine project template. The benefit of this procedure is that the workflow is automated for the user such that predefined criteria and logic are provided, but there are opportunities for adjustments. In addition, the automated use of Microsoft Project easily creates a CPM schedule from the users completed inputs. The developed schedule is them refined by Design and Construction staff to set the project’s contract time. OHIO DEPARTMENT OF TRANSPORTATION CONSTRUCTION DURATION TOOL The Ohio DOT implemented an agency-specific regression-based tool to determine preliminary con- tract durations. Development of this tool was outsourced to the Parsons Brinckerhoff consulting firm and a final product was delivered to the Ohio DOT in 2013. This tool predicts contract construction durations using a series of regression models inclusive of multiple project variables such as project cost, project type, project location, and starting season. The Ohio DOT provided the consultant with 8 years of project data and included the following parameters: • Project Construction Duration (in calendar days) • Project Construction Cost (in dollars) • Type of Roadway—Interstates, U.S. Routes, State Routes, or Other • Location—Urban or Rural • Project Starting Season—Fall, Winter, Spring, or Summer • Project Type (19 categories as follows) – Bridge painting – Bridge repair – Bridge replacement – Culverts State 1 State 2 State 3 State 4 State 5 State 6 State 7 Have Documented Procedure Yes Yes Yes Yes Yes Yes Yes Have Official Revision Procedure No No Yes Yes No Yes Yes Evaluate the Effectiveness of Their Procedure Unsure No Unsure Yes No Unsure Yes Accuracy of Procedure Unsure ±50–100% ±0–25% ±0–25% Unsure Unsure Unsure Division That Handles Contract Time Determination (C—Construction, D—Design) D C C and D C D C and D D How Their Tools Are Developed (CD—Custom Developed, P—Purchased Off the Shelf, C—Combined) CD P CD C C C CD Same or Individual Procedure for Different Delivery Methods Same Same Same Same Same Same Individual TABLE 3 CRITERIA FOR SELECTING STATES FOR INTERVIEW

FIGURE 21 Screenshot for Primavera scheduling template.

FIGURE 22 Screenshot for Microsoft Access quantity input.

FIGURE 23 Screenshot for Microsoft Project schedule output.

27 – Guardrail – Intersections – Landslide/drainage repair – Lighting – Maintenance – Major construction – Minor rehabilitation – Minor rehab—General systems – Noise walls – Other – Pavement markings – Resurfacing – Signals – Signing – Widening. A unique regression equation was developed for each of the 19 project types. The results present a mean predicted duration and a 90% and 95% confidence interval. The regression equations were assembled into a computer program for ease of use. More information on this specific tool can be found in Appendix D. The Ohio DOT uses the previously mentioned tool as a preliminary estimate of construction dura- tion early in the design phase. It relies on a production rate-based tool for the final setting of contract time. The production rate tool is also completed by designers with the recommendation that they will solicit feedback from construction forces within the department. The production rate tool was also recently developed (2012) through reference of other STA approaches such as those of the Indiana and Virginia DOTs. KENTUCKY CONTRACT TIME DETERMINATION PROCEDURE Within the Kentucky Transportation Cabinet (KYTC) contract time estimates are prepared by the design division, but typically finalized with feedback from the construction personnel. Since 2000, the KYTC implemented a custom scheduling procedure that utilized six generic project templates with generic Microsoft Project activity logic and user adjustable default production rates (Hancher and Werkmeister 2000). In 2011, KYTC revised the existing contract time determination procedure based on user feed- back and internal questions regarding the setting of contract time. The revised procedure is a multivariate regression model-based procedure that is operated within a Microsoft Excel Macro-enabled template. This Macro-enabled template assists users in the workflow of the procedure and the development of the time estimate. Projects to be analyzed are categorized-based and project sized by dollar volume, with projects of greater than $1,000,000 and less than $1,000,000 being analyzed separately. Projects under $1,000,000 use a regression model developed from the historic data of small proj- ects based solely on the construction estimate and selection of ten different project types. An analy- sis of KYTC projects of less than $1,000,000 found that 90% of these projects were completed in less than 6 months and that project duration was not significantly impacted by controlling bid item quantities (Zhai et al. 2016). Interviews with KYTC personnel showed that these projects were typi- cally let with the intention to be completed within the construction season and contractors executed these projects based on project location and time-dependent resource availability. For example, a small resurfacing project might be executed when a contractor is moving equipment from one large project to another large project. A screen capture from the less than $1,000,000 project tool is shown in Figure 24. The small project tool requires the user to input the project identification number, the con- struction estimate, the letting date, and the selected design project type. The procedure returns an estimated lower, mean, and upper range of completion dates and working days based on the 95% confidence interval.

28 Contract times for projects greater than $1,000,000 are estimated using the large project tool. The large project duration estimate tool is based on a multivariate regression model for five different project types (Table 4). Each project type has a unique multivariate regression equation. The inputs to the model include the construction estimate and key bid item quantities that have been found to be predictors of project duration (e.g., cubic yards of roadway excavation and linear feet of storm sewer). Figure 25 shows the opening macro of the Excel template and Figure 26 an example of an open access template input FIGURE 24 Input screen for projects less than $1,000,000 (KYTC). TABLE 4 LARGE PROJECT TYPES (>$1,000,000) WITHIN THE KENTUCKY CONTRACT TIME DETERMINATION PROCEDURE

29 FIGURE 25 Screenshot of KTC title page. FIGURE 26 Screenshot for open access calculation.

30 and calculation sheet (other project types are similar). The tool returns a lower, mean, and upper contract duration estimate based on the 95% confidence interval. One key point of emphasis with the revised Kentucky contract time determination procedure is that the tool provides an estimate that is to be used as a starting point for the establishment of contract time. The user’s guide for the tool has the following statement regarding its use: Once a contract time estimate is determined, careful consideration should be given to its significance. It should be also noted whether it is essential for the project be completed in the scheduled time calculated or if there is flexibility to build in float. For example, if a project would realistically take 15 days to complete, but KYTC did not need the job complete for 90 days, then the project should not be scheduled for 15 days. (KYTC 2014) The tool and the users’ guide are available for download from the KYTC Highway Design Soft- ware and Support webpage under “working days” at the following link: http://transportation.ky.gov/ Highway-Design/Pages/Software-and-Support.aspx. An online training webinar on the use of the tool is available at https://www.youtube.com/watch?v=pXxtYYc-rGA&feature=youtube.