Skip to contentUnited States Department of Transportation - Federal Highway AdministrationSearch FHWAFeedback
Highways for LIFE

Arrow I-85 Interchange Design-Build Project Using Prefabricated Bridge Elements in West Point, GA

<< Back Content Next >>

Data Acquisition and Analysis

Data on safety, traffic flow, quality, and user satisfaction before, during, and after construction were collected to determine if this project met the HfL performance goals.

The primary objective of acquiring these types of data was to quantify project performance and provide an objective basis from which to determine the feasibility of the project innovations and to demonstrate that the innovations can be used to do the following:

  • Achieve a safer work environment for the traveling public and workers.
  • Reduce construction time and minimize traffic interruptions.
  • Produce a high-quality project and gain user satisfaction.

This section discusses how well the GDOT project met the specific HfL performance goals related to these areas.


The crash data from the I-85 corridor (see table 1) shows many vehicular crashes resulting in 43 injuries and no fatalities within the project limits (mileposts 3 to 8) during the 3-year study period before construction. This is a significant numbers of crashes. To help keep all types of crashes to a minimum, concrete Jersey-type barriers were used instead of cones or barrels to permanently separate the traffic on I-85 from construction workers. For the first time, GDOT required the contractor to set these barriers along the I-85 shoulders within 6 months of the beginning of construction and maintain them until all final roadway features were installed (guardrail, striping, etc.) and construction equipment was removed at the end of the project.

Table 1. Historical 3-year crash data.*
Year Number of Crashes Number of Vehicles Number of Injuries Number of Fatalities
2006 31 42 20 0
2007 21 29 14 0
2008 17 26 9 0
Total 69 97 43 0

This project was the first in which GDOT required the D-B contractor to provide the hardware, methods, and process to monitor speed bands during construction with the primary goal of improving work zone safety. Real-time monitoring enabled GDOT to evaluate daily reporting of traffic volumes and speeds through the construction zone.

As a result of the enhanced safety features included in this project, worker and motorist safety during construction exceeded the HfL performance goals. During construction, no worker injuries were reported, which means GDOT exceeded the HfL goal for worker safety (incident rate of less than 4.0 based on the OSHA 300 rate). Only a single motorist incident occurred in the construction zone on I-85, resulting in minor vehicle damage and no personal injury.

Contruction Congestion


Minimal queue lengths were observed during construction. To keep congestion down, GDOT required the contractor to prioritize work elements under the D-B delivery method, enabling the contractor to schedule the most efficient use of lane closures and lessen congestion. Additional methods used to limit congestion include the following:

  • Restricting lane closures to 180 days in any given direction (NB or SB) and to a total of 270 days when overlapping work is done on both NB and SB lanes. The contractor realized its targets (SB took about 150 days to complete and NB took 120 days to complete).
  • Restricting lane closures to offpeak times, typically 9 a.m. to 4 p.m. during the week and on weekends. The contractor had an option to close lanes after 7 p.m. on any day and up to 6 a.m. the following day on weekdays. However, a review of traffic data for this corridor indicated that starting work at 8 p.m. Sunday and working through Tuesday evening would have much less impact on the motoring public.
  • An additional benefit of the contractor-installed speed band system was allowing GDOT to provide real-time traffic information on delays or incidents to the traveling public through remote dynamic message boards placed along I-85.

Finally, the project also required the contractor to provide methods for noninjury incident clearance time management. This included methods to reach a goal of clearing noninjury incidents from the construction zone travel lanes within 20 minutes. The one minor vehicle incident that did occur was promptly cleared in less than 20 minutes.


Pavement Test Sections

This is a unique HfL project in that the bridge construction did not replace an existing structure. Therefore, preconstruction test sections were chosen from the nearest interstate exit that represents typical in-service pavements. Interstate interchange ramps at Exit 13 and Exit 14 located just 6 mi (9.6 km) north were chosen for comparison with the postconstruction pavements of the new interchange on- and off-ramps for Exit 6. Exit 13 has an aged dense-graded asphalt surface and Exit 14 is transverse-tined concrete (figure 14). The new Kia Boulevard bridge deck and the off-ramp at Exit 6 both have a transverse-tined concrete surface, while the Exit 6 on-ramp is dense-graded asphalt (figure 15).

Figure 14. Comparison pavements (left to right): Exit 13 ramp and Exit 14 ramp.
Figure 14. Comparison pavements (left to right): Exit 13 ramp and Exit 14 ramp.

Figure 15. Newly constructed pavement test sections (left to right): Kia Boulevard bridge deck, Exit 6 off-ramp, and Exit 6 on-ramp.
Figure 15. Newly constructed pavement test sections (left to right): Kia Boulevard bridge deck, Exit 6 off-ramp, and Exit 6 on-ramp.

Sound Intensity Testing

Presently, GDOT does not use the onboard sound intensity (OBSI) test method on any projects. However, this method was used to collect tire-pavement sound intensity (SI) on the newly constructed pavements of this project and U.S. 27, Exit 13, and Exit 14 for comparison.

Sound intensity measurements were made by the National Center for Asphalt Technology personnel and equipment using the OBSI technique AASHTO TP 76-08, which uses dual vertical sound intensity probes and an ASTM-recommended standard reference test tire (SRTT). The sound measurements were recorded and analyzed using an onboard computer and data collection system. A minimum of three runs were made in the right wheelpath with two phase-matched microphone probes simultaneously capturing noise data from the leading and trailing tire-pavement contact areas. Figure 16 shows the dual probe instrumentation and the tread pattern of the SRTT.

Figure 16. OBSI dual probe system and the SRTT.
Figure 16. OBSI dual probe system and the SRTT.

The OBSI measurements were conducted at 45 miles per hour (mi/h) (72.4 kilometers per hour (km/h)). The average of the front and rear SI values was computed over the full length of the pavement sampled to produce SI values. Raw noise data are normalized for the ambient air temperature and barometric pressure at the time of testing. The resulting mean SI levels are A-weighted to produce the sound intensity frequency spectra in one-third octave bands, as shown in figure 17 for the exit ramps and figure 18 for the Kia Boulevard bridge deck and U.S. 27.

Figure 17. Mean A-weighted sound intensity frequency spectra for the exit ramps.
Figure 17. Mean A-weighted sound intensity frequency spectra for the exit ramps.

Figure 18. Mean A-weighted sound intensity frequency spectra for the Kia Boulevard bridge deck and U.S. 27.
Figure 18. Mean A-weighted sound intensity frequency spectra for the Kia Boulevard bridge deck and U.S. 27.

The figures show that transverse-tined concrete pavement of Kia Boulevard has slightly higher decibel levels than the U.S. 27 pavement above about 1,000 hertz (Hz), which is typical of this type of surface texture. The ramps show generally similar spectra except for some variance in the lower frequencies.

Sound levels were calculated using logarithmic addition of the one-third octave band frequencies across the spectra. The noise levels are presented in table 2. Overall, the sound levels among the existing and newly constructed pavements were very similar. However, the sound levels from the newly constructed ramps and bridge deck were slightly higher than the HfL goal (96.0 dB(A) or less) by a range of 1.0 to 2.7 dB(A). Note that the HfL goal of 96.0 dB(A) was intended for pavement surfaces, not elevated bridge decks. It is also noted that this goal is difficult to achieve on any concrete surface using current technology.

Table 2. OBSI summary.
Pavement Feature Baseline* OBSI dB(A) As-built OBSI dB(A)
Ramps (hot-mix asphalt (HMA) Surface) 97.2 98.7
Ramps (PCC surface) 97.4 97.0
Interchange (PCC Bridge Deck) N/A 98.6
*Baseline ramp data taken from Exits 13 and 14.
Smoothness Measurement

Smoothness measurements on the test sections were collected by the Auburn University Automatic Road Analyzer (ARAN) van (figure 19). The ARAN is a high-speed inertial profiler able to perform roughness measurements of the pavement surface in both wheelpaths. Roughness is reported in inches per mile as recommended by the International Roughness Index (IRI) approach and consists of a mathematical assessment of the section profile aimed at quantifying the quality of the ride in a passenger car. The ARAN van performed three runs in each direction at a speed of 45 mi/h (72.4 km/h).

Figure 19. ARAN van.
Figure 19. ARAN van.

The average of the left and right wheelpaths are calculated and presented in table 3.

Table 3. Mean roughness measurements.
Pavement Feature Baseline* IRI (in/mile) As-built IRI (in/mile)
Ramps (HMA Surface) 98 68
Ramps (PCC surface) 115 89
Interchange (PCC Bridge Deck) N/A 77
*Baseline ramp data taken from Exits 13 and 14

Overall, the roughness values are lower for both the newly constructed asphalt and concrete pavements. However, the newly constructed bridge deck and ramp pavements did not reach the HfL goal of 48 inches per mile or less (43.8 inches per mile target value for this specific project). The contractor's testing of all other pavement sections of the project (excluding bridge deck and ramps) concluded that the project goal was indeed satisfied as reported by the project team and is included as following:4

4. Arcadis, HfL Performance Review, West Point 85 Interchange Project.

The IRI for the West Point 85 Interchange ranges from 20.9 to 32.2 inches per mile, never exceeding the goal of a finished pavement smoothness of 43.8 inches per mile.

The IRI statistics above do not include smoothness data for concrete-paved facilities, such as shoulders, bridges, and ramps. Data for these facilities are reported as Profile Index. Georgia DOT uses a Rainhart-type profilograph with specifications for both overall smoothness and localized profile deviations (scallops) to determine initial smoothness. A Profile Index is determined from profilograms of pavements for every 0.25-mi (0.4-km) section of pavement. Vertical deviations exceeding 7 in (177.8 mm) on the mainline and 12 in (304.8 mm) on ramps were corrected. The finished Profile Index for all segments met these requirements.

Attention to pavement smoothness on this project contributed to the overall quality and durability of the West Point 85 Interchange.

User Satisfaction

Under the D-B contract, the contractor was required to involve the community and the traveling public through a public involvement and communications plan. Requiring the contractor to maintain communication is a new business practice for GDOT. The plan kept the public informed of the construction schedule through tactics such as the following:

  • Online information
  • News releases and mailing
  • Local signage
  • Construction information hotline

These communication efforts had a positive impact on the user satisfaction as indicated by the survey results. During construction and upon completion of the project, the contractor conducted four Likert scale user satisfaction surveys (at 25, 50, 75, and 100 percent of project completion). The approval rating goal was set at 80 percent or better (i.e., 80 percent of the surveyed customers approve of the job being done in the construction work zone). The remainder of this subsection is taken from Arcadis, HfL Performance Review.5

The project team successfully measured the level of user satisfaction by completing surveys via telephone interviews. An internal team goal of 300 respondents per round of surveys was established to obtain adequate statistical reliability and allow for breakouts in cross-tabulations. The target audience for the surveys included the following:
  • Half were constituents living in a 3-mi (4.8-km) radius of the construction.
  • The remaining 50 percent were taken from a random sample of constituents purchased by the surveyor who lived in West Point and Troup County, GA, and Lanett and Valley, AL.
  • All respondents were age 18-plus and could drive.
  • There was a 50-50 mix of males and females.
  • GDOT employees were excluded.

At the 100 percent completion point for the project, respondents were asked to indicate their level of satisfaction with the new facility. At the 25, 50, 75, and 100 percent completion points for the project, respondents were asked to indicate their level of satisfaction with the approach used to construct the new facility in terms of minimizing disruption.

The level of user satisfaction was determined for all respondents. However, survey results showed that a large group of neutral respondents emerged for each survey. It was determined from cross-tabulation of the data that these respondents tended to be infrequent travelers into the construction area. Therefore, to get a better idea of actual user satisfaction, the project team also determined the level of satisfaction for non-neutral respondents who expressed an opinion during the survey. This group included those who were very satisfied, somewhat satisfied, somewhat dissatisfied, and not at all satisfied. The survey results for both non-neutral respondents and all respondents are provided in table 4.

Table 4. User satisfaction summary.
  % Very/Somewhat Satisfied With New Facility % Very/Somewhat Satisfied With Construction
(average of all four surveys)
Nonneutral Respondents 91% 75%
All Respondents 78% 49%

As table 4 shows, 91 percent of non-neutral respondents were very to somewhat satisfied with the new I-85 facility overall, and 78 percent of all respondents were very to somewhat satisfied. Only 8 percent of all respondents indicated dissatisfaction (somewhat to very dissatisfied). On average, over the four surveys, 75 percent of non-neutral respondents were very to somewhat satisfied with construction activities in terms of minimizing traffic delays. Forty-nine percent of all respondents were very to somewhat satisfied, and only 16.5 percent of all respondents were somewhat to very dissatisfied.

In addition, using cross-tabulations in the data, the project team determined that satisfaction levels among respondents increased with a respondent's frequency of travel through the construction zone. This is a key indicator of project success.

Strategies implemented by the project team to alleviate traffic congestion and improve work zone safety likely contributed to the level of user satisfaction. Other influencing factors include the following:

  • Other construction activities: During the survey, considerable construction activity was occurring at the Kia Motors plant and on surrounding portions of I-85, which may have been confused with construction activities for the new I-85 interchange. This possibility is clearly seen in the results of the first survey (at the 25 percent project completion point). Almost 18 percent of the respondents expressed dissatisfaction with construction efforts in terms of delays experienced. However, no lane closures were implemented for this project until after the 25 percent complete survey was administered.
  • Survey question clarification: The survey question asked at the 25 percent point did not specify construction activity along I-85. This was corrected for subsequent surveys. Specifically, the question changed from "How satisfied are you to date with the approach used to construct the new I-85 interchange near the Kia Motors plant in Troup County in terms of minimizing disruption?" to "How satisfied are you to date with the approach used to construct the new I-85 interchange near the Kia Motors plant in Troup County in terms of minimizing disruption to you on the new interchange and on I-85?"
  • Communication activities: Communication activities (including news releases, postcard mailers, and the project Web site) likely called attention to construction activity in the area, regardless of the location of these activities. Communication efforts appear to have had a positive influence on user satisfaction and survey results as well. Respondents included in the project database for information dissemination activities (received postcard mailer and factsheet on construction activities), especially at the 50 percent complete point, were more likely to have an opinion and be satisfied with construction efforts.
  • Neutral responses: A large group of respondents for each survey had a neutral opinion, and many respondents may not have encountered construction, particularly those who live outside the immediate area.
<< Back Content Next >>

More Information



Mary Huie
Center for Accelerating Innovation

This page last modified on 04/04/11

United States Department of Transportation - Federal Highway Administration