20-Year Performance of Bridge Maintenance Systems

J. Peter Ault, P.E. – Elzly Technology Corporation (609) 374-7355
Christopher L. Farschon, P.E. – Corrosion Control Consultants & Labs (609) 536-801

In 1986-1987, New Jersey DOT applied forty-seven (47) different coating systems to various individual spans of the Mathis Bridge. The eastbound Mathis Bridge carries Route 37 over the Barnegat Bay from Toms River to Seaside Heights, New Jersey. Each experimental system was applied to a complete span comprising approximately 4,000 square feet of steel. Experimental coating systems included metallizing, various zincbased systems, various levels of surface preparation, and several overcoating strategies (e.g., alkyd over a hand-tool cleaned surface).

This paper will present the results of an inspection conducted in 2007, nominally 20 years after the initial coating application. The inspection showed varied service lives associated with the different coating systems. Some of the systems were in excellent condition after 20 years while others had completely broken down. In addition to the present condition of the test spans, the paper will review the historical performance of the various coating systems as well as the applied cost. Finally, several important implications for maintenance planners will be presented. These will include cost-benefit calculations and risk-reduction strategies.

Background

The New Jersey DOT has an ongoing evaluation of various bridge coatings on the Thomas Mathis Bridge which carries State Route 37 over Barnegat Bay in central New Jersey. 1 The Mathis Bridge consists of 66 spans plus a lift span. Each span is approximately 73 feet long and contains five rolled I-beam stringers of A-36 steel spaced 8 feet apart. Each span contains approximately 4,000 square feet of painted surface area. The bridge is situated over the salt water of Barnegat Bay with vertical clearances from 5 feet at the abutments to 33 feet at the lift span.

Upon construction in 1950, the structure was painted with 3 coats of an oil-based paint containing red lead pigment. The bridge was painted 3 times at various intervals over the next 28 years. The painting just prior to the experimental evaluation occurred in 1978. At that time, a basic lead-silico chromate, oil alkyd system was used with a pigmented fascia coating and “black graphite” on the interior steel.

In 1984, an inspection of the bridge noted that rust and corrosion was extremely heavy on the bearing assemblies, some stringer webs, and bottom flange of the stringers. Corrosion was especially concentrated on stringer ends located at the bridge piers (i.e., steel in the path of run-off water form the bridge deck expansion joints). Rust scale on the steel was as thick as ½-inch. The existing paint was 15 to 25 mils (380 to 635 1 A. Chmiel, V. Mottloa, and J. Kauffman, Structural Coating Evaluation in New Jersey, Research News, Journal of Protective Coatings and Linings, January 1989, pp 23-26.

microns). Concentrated salt deposits were visible on the steel directly beneath the deck joints. The severe marine environment and road salt usage create a severely corrosive environment for the evaluation of different maintenance painting methods. Subsequent to a laboratory evaluation of available maintenance coatings, NJDOT let contract 85-2, Painting of the Mathis Bridge. The bid documents contained specifications for each experimental paint system. Given the project’s timeframe (1986 to 1987) full containment of the blast abrasive and debris was required to comply with recent environmental regulations.

Coating Systems

Eighteen manufacturers donated coatings to be used on 47 of the 66 spans. The experimental systems consisted of inorganic and organic zinc coatings, epoxies, aluminum epoxy urethanes, vinyls, urethanes, oil-alkyds, zinc metallizing, aluminum metallizing, rust converters, and others. These systems represented most feasible options for maintenance overcoating or coating replacement painting of a bridge. Table 1 provides a list of the coating systems along with surface preparation, application date, and span number. The remaining spans were coated with the standard NJDOT Zone 3B system which consisted of a phenoxy organic zinc primer and vinyl intermediate and finish coats.

The surface preparations ranged from SSPC SP-2, Hand Tool Cleaning to SSPC SP-5, White Metal Blast depending on the coating manufacturer’s recommendation. For systems requiring spot cleaning, only loose rust and peeling paint was removed. Containment was not erected during hand tool cleaning. Sand used for blasting was collected on corrugated steel containment floors so that it could be removed for proper disposal.

Seventeen of the eighteen coating manufacturers had a representative on site to approve surface preparation, give mixing instructions, and provide guidance regarding any potential problems. State inspectors worked closely with the paint contractor and manufacturers representative to assure compliance with the manufacturer’s and NJDOT’s minimum specification requirements. Painting began on October 11, 1986. Seven systems requiring spot cleaning were applied before mid November when weather conditions were no longer suitable for any of the systems (some of the systems were designed for application as low as 40 F). Painting resumed in April, 1987 and was completed in October, 1987.

Table 1. Summary of Test Coating Systems

Span Coating System Surface Preparation Application Date 1986 Cost ($/ft2)

Alkyd Systems (6)

7E ALKYD OIL BASE/Si ALKYD SP-2 NOV. 86 $0.71
11E ALKYD/EPOXY/URETHANE SP-2 OCT. 86 $1.04
21W ALKYD/EPOXY/URETHANE SP-6 AUG. 87 $1.56
43W OIL - ALKYD SP-6 OCT. 87 $1.11
13W OIL ALKYD - 3 Cts SP-2 JUNE 87 $0.73
31W OIL-ALKYD SP-6 OCT. 87 $1.37

Aluminum Systems (8)

41W ALUM. URETHANE/ACRYL. SP-6 SEPT. 87 $1.58
12E ALUM. EPOXY/URETHANE SP-7 OCT. 86 $1.00
9E ALUM. EPOXY/URETHANE SP-2/3 NOV. 86 $0.63
8E ALUM. EPOXY/URETHANE SP-2 NOV. 86 $1.07
6E ALUM. EPOXY/URETHANE SP-2 APRIL 87 $0.60
5W ALUM. EPOXY/URETHANE SP-2 MAY 87 $0.70
45W ALUM. EPOXY/URETHANE SP-6 OCT. 87 $0.82
24W ALUM. URETH/URETHANE SP-6 SEPT. 87 $1.28

Epoxy Systems (6)

9W EPOXY MASTIC/EPOXY MAST. SP-6 JUNE 87 $1.00
17W EPOXY MASTIC/URETHANE SP-6 JULY 87 $1.25
18W EPOXY/URETHANE SP-6 JULY 87 $1.29
32W EPOXY/URETHANE SP-6 OCT. 87 $1.12
27W ONE COAT EPOXY SP-6 SEPT. 87 $0.69
29W ONE COAT EPOXY SP-6 OCT. 87 $0.99

Inorganic Zinc Systems (8)

34W H20 INORG. PRIME/SILICONE SP-6 OCT. 87 $1.67
30W H20 INORG. ZINC/ACRYL SP-10 OCT. 87 $1.99
42W INORG. ZINC/VINYL SP-10 OCT. 87 $1.56
46W INORG. ZINC/VINYL SP-10 OCT. 87 $1.26
14W INORG. ZINC/EPOXY/UR. SP-6 JUNE 87 $1.85
35W INORG. ZINC/EPOXY/URE. SP-10 OCT. 87 $1.94
39W INORG. ZINC/URETHANE SP-6 OCT. 87 $1.07
12W INORG. ZINC/VINYL SP-10 JUNE 87 $1.75

Metallizing Systems (2)

37W 100% METALLIZING ZINC SP-5 SEPT. 87 $4.72
38W 85% ZN - 15% AL METALLIZE SP-5 SEPT. 87 $4.85

Miscellaneous Systems (5)

4E CALCIUM BORO-SILICATE - 3Cts SP-2 MAY 87 $0.90
16W CALCIUM BORO-SILICATE - 3Cts SP-6 JULY 87 $1.42
10W LATEX - 3 Cts SP-10 JUNE 87 $1.85
26W THERMOPLASTIC RUBBER SP-10 SEPT. 87 $2.45
40W VINYL/ACRYLIC SP-6 OCT. 87 $1.20

Organic Zinc Systems (7)

7W ORG. ZINC/EPOXY/URET. SP-10 MAY 87 $1.75
28W ORG. ZINC/URETHANE SP-6 OCT. 87 $1.33
20W ORG. ZINC/EPOXY/URETHANE SP-6 AUG. 87 $1.50
23W ORG. ZINC/URETHANE SP-6 SEPT. 87 $1.48
25W ORG. ZINC/URETHANE SP-10 SEPT. 87 $2.09
11W ORG. ZINC/VINYL SP-6 JUNE 87 $1.75
15W ORG. ZINC/VINYL/VINYL SP-10 JULY 87 $1.50

Urethane Systems (5)

33W URETHANE 3-COAT SP-6 OCT. 87 $1.71
44W URETHANE/EPOXY SP-6 OCT. 87 $1.19
10E URETHANE/EPOXY/URETHANE SP-2 NOV. 86 $1.01
5E URETHANE/EPOXY/URETHANE SP-2 NOV. 86 $1.55
19W URETHANE/EPOXY/URETHANE SP-6 AUG. 87 $1.55

Inspections

The data presented in this paper is the result of visual inspections conducted by the authors in 1995, and 2007 in addition to the data presented in the original NJDOT report. The original NJDOT report included one year performance evaluations conducted from a snooper tuck.2 Visual ratings were given to each span based on the percent rusting of the bottom flange. This was deemed to be the harshest exposure and thus the best basis to rank the systems after a relatively short exposure period. The ratings were made in accordance with ASTM D-610, Standard Method of Evaluating Degree of Rusting on Painted Steel Surfaces.

As part of a FHWA project, three inspectors performed a follow-up inspection of the structure in 1995.3 The inspections consisted of assigning a 1-10 rating to the entire span in accordance with ASTM D-610 based on visual assessment from a boat. The inspectors were 0 to 30 feet from the structure, depending on the span. Extensive photographs were taken during the inspections. The ASTM D 610 ratings provided by three individual inspectors were averaged to provide a composite rating. In most cases the inspectors’ ratings were within one unit of each other. For the purposes of this paper, the authors again rated the structures in 2007 using similar procedures to the 1995 inspection.

Discussion

The results of the NJDOT test program after 1 year of exposure indicated mixed performance of overcoating systems.4 Those systems applied over an SP-2 (hand-tool cleaned) surface included alkyds, epoxies, and urethanes. The epoxy mastic systems covered a wide range of performance. Several different manufacturers’ versions of this popular maintenance painting system were applied over SP-2 surfaces. Some of these systems had already failed at the 1 year inspection while others were among the best performers over “surface tolerant” conditions. Other systems performing well over SP-2 surfaces were a calcium borosilicate pigmented alkyd system and an oil-alkyd system. The 1 year results for systems applied over abrasive blasting were consistently good, showing little differences between systems.

2 A. Chmiel, V. Mottloa, and J. Kauffman, Research on Structural Coatings Performance by New Jersey Department of Transportation, presented at the 6th International Bridge Conference and Exhibition, Pittsburgh, Pennsylvania, 1989.
3 Guidelines for Repair and Maintenance of Bridge Coatings: Overcoating, C.L. Farschon, R.A. Kogler, and J.P. Ault, August 1997, FHWA-RD-97-092, 95 p
4 Ibid.

Figure 1 presents the 2007 inspection data on the Y-axis (ASTM D610 – 10 = best, 0 = worst) versus the cost of the coating system ($/ft2 in 1986/87 dollars) on the X-axis. The data suggests a trend toward increased performance with increasing cost, but the relationship has considerable scatter. Cost alone would not be a good basis to assess the overall value of a coating system simply because there are so many other criteria that play into the success of a coating system.

0
1
2
3
4
5
6
7
8
9
10
$0.00 $1.00 $2.00 $3.00 $4.00 $5.00 $6.00
1986 Cost
2007 Rating (10 = perfect)

Figure 1. Correlation between cost and condition after 20 years of service. Each of the tested coating systems was a unique combination of coating and surface preparation. To make generalized conclusions, we grouped the 47 experimental systems into eight generic categories as shown in Table 1. Plots of the condition versus time were generated for each of the individual test span, and then the number of systems in each category were counted that exhibited performance below certain threshold ratings. Systems with an ASTM D610 rating of “4” or below (more than 10% rusting) were classified as candidates for coating removal and replacement. Systems with an ASTM D610 rating of “4” to “7” (between 0.3% and 10% rusting) were classified as candidates for coating maintenance. Systems with an ASTM D610 rating better than “7” (less than
0.3% rusting) were classified as being in good condition. Table 2 shows the data. Table 2. Distribution of Condition Ratings for Coating within Each Category

D610 Rating at 8 Years D610 Rating at 20 Years > 7 7-4 <4 > 7 7-4 <4
Metallizing Systems (2) 2 0 0 2 0 0
Inorganic Zinc Systems (8) 7 0 1 2 5 1
OZ Systems (7) 5 2 0 2 4 1
Miscellaneous Systems (5) 3 2 0 1 3 1
Alkyd Systems (6) 4 2 0 1 3 2
Urethane Systems (5) 2 3 0 1 2 2
Aluminum Systems (8) 1 4 3 0 4 4
Epoxy Systems (6) 0 4 2 0 0 6

Because of the inherent variability in any coating system, the overall performance of a coating system is not reliably quantified with a single life expectancy. Quantifying a coating system life is better suited to a probabilistic or risk-based analysis. To illustrate this analysis the risk associated with maintenance painting for each group of coatings was determined in two ways.

Figure 2 shows the likelihood of reaching each of the above defined end-states (Good – less than 0.3% rusting, maintenance candidate – 0.3% to 10% rusting, remove and replace candidate – greater than 10% rusting) for each coating system group. Notice how this figure ranks the groups of coating systems by performance, with the better performing groups to the left and the poorer groups to the right.

Condition of Systems at 20 Years

0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%

Metalli zing Systems (2)
Inorganic Zinc Systems (8)
OZ Systems (7)
Misc ellaneous Systems (5)
Alky d Systems (6)
Urethane Systems (5)
Aluminum Systems (8)
Epoxy Systems (6)

Percentage of Experimental Systems

Good Maintenance Candidate Removal/Recoat Candidate

Figure 2. Coating systems by category, showing their overall success condition at 20 years.
Figure 3 builds on the data in figure 2 by defining “success” as a system which is “good”
at 8 years and only a “maintenance candidate” at 20 years. Using this definition of
“success” we can determine a “probability of success” for each group. Figure 3 shows
the probability of success, the likelihood of becoming a candidate for removal and
replacement at 8 years, and average applied cost for each of the coating system groups.
Risk Assessment for Bridge Coating Maintenance

0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
70.00%
80.00%
90.00%
100.00%

Metallizing Systems (2)
Inorganic Zinc Systems (8)
OZ Systems (7)
Miscellaneous Systems (5)
Alkyd Systems (6)
Urethane Systems (5)
Aluminum Systems (8)
Epoxy Systems (6)

Percentage of Experimental Systems

$0.00
$1.00
$2.00
$3.00
$4.00
$5.00
$6.00

Probability of Success Remove/Replace Candidate at 8 Years Average Cost, 1986

Figure 3. Risk assessment evaluation for each group of coating systems. Obviously there are nuances in each of the broad categories. Certainly the high cost and high probability of success associated with the metallizing is expected. However, there is also a high probability of success with the inorganic and organic zinc based systems. The aluminum and epoxy systems show a low probability of success and are most likely to be in poor condition after 8 years. The following paragraphs will explore some of the coating groups in more detail.

Metallizing Systems (2)

The metallizing systems are performing extremely well, even after 20 years. At the 20- year inspection, the first signs of rusting were noted on both the zinc and 85 Zn-15 Al metallized spans. For both systems the rusting was at the crevice between the bearings and the stringer flange, and on isolated lower flange spots (likely to be containment hanger locations). It appeared that any place the surface preparation and metallizing thickness were attainable; the bridge was in perfect condition.

Photo 1. Close-up of bearings on metallized systems.

Inorganic Zinc Systems (8)

The inorganic zinc systems performed quite well as a class. Only one system performed unacceptably as defined by the authors. This system was a waterbased inorganic zinc with a silicone topcoat applied over an SP-6 (Commercial Blast) surface. The performance of the inorganic zinc systems is quite interesting because of the variety of systems evaluated. Figure 4 shows the ratings over time for each of the individual systems. The dark blue lines correspond to systems applied over an SP-10 surface and the pink lines correspond to systems applied over an SP-6 surface. Comparable coating systems have similar symbols. It is interesting to note that the waterborne inorganic zinc performed poorly over the SP-6 surface while the solvent based systems performed as well or better over the SP-6 as the SP-10 surfaces. This is in contrast to the standard industry requirement than an inorganic zinc coating should be applied over an SP-10 surface.

Performance of Inorganic Zinc Systems

0
1
2
3
4
5
6
7
8
9
10

0 2 4 6 8 10 12 14 16 18 20

Age, Years
Rating (10=new)
SP-6/WBIOZ/Si
SP-10/WBIOZ/Acry
SP-10/IOZ/VY
SP-10/IOZ/VY
SP-6/IOZ/EP/Ure
SP-10/IOZ/EP/Ure
SP-6/IOZ/Ure
SP-10/IOZ/VY

Figure 4. 20 year performance of inorganic zinc systems.

Photo 2. Example IOZ system with good condition of web, but poorer condition of bottom flanges.

OZ Systems (7)

The organic zinc systems performed quite well as a class. The only system which did not perform well was one of the organic zinc systems with a urethane topcoat over an SP-6 prepared surface. Figure 5 shows the performance versus time of the individual organic zinc systems. Again, the dark blue lines represent systems over an SP-10 surface while the pink lines represent systems over an SP-6 surface. Excepting the organic zinc/urethane system, the data suggest that equivalent performance can be achieved over an SP-6 and SP-10 surface.

Performance of Organic Zinc Based Systems

0
1
2
3
4
5
6
7
8
9
10

0 2 4 6 8 10 12 14 16 18 20

Age, Years
Rating (10=new)
SP-10/OZ/VY/VY
SP-10/OZ/E/U
SP-10/OZ/U
SP-6/OZ/U
SP-6/OZ/U
SP-6/OZ/E/U
SP-6/OZ/VY

Figure 5. 20 year performance of organic zinc systems.

Miscellaneous Systems (5)

Figure 6 presents the performance over time of the miscellaneous systems. All of the systems were candidates for maintenance after 20 years. Worth noting is the performance of the Calcium boro-silicate over the SP-2 surface. This system was the second best performing system over an SP-2 surface. At an applied cost of $0.90 per square foot in 1986, it was the best performing of the low-cost (less than $1 per square foot) systems.

Performance of Miscellaneous Systems

0
1
2
3
4
5
6
7
8
9
10

0 2 4 6 8 10 12 14 16 18 20

Age, Years
Rating (10=new)
SP-2/Calcium Boro-Silicate
SP-6/Calcium Boro-Silicate
SP-10/3 ct Latex
SP-10/Rubber
SP-6/Vy/Acry

Figure 6. 20 year performance of miscellaneous (not categorized) systems.

Alkyd Systems (6)

Figure 7 shows the performance of the alkyd systems over time. As a class, the alkyd systems generally performed well over the first 8 years. One of the systems over SP-2 had an unacceptable level of failure on the flange during NJDOT’s 1-year inspection. However, considering all of the alkyd systems there seems to be relatively little benefit to an SP-6 surface preparation versus an SP-2 surface preparation. Performance of Alkyd Systems

0
1
2
3
4
5
6
7
8
9
10

0 2 4 6 8 10 12 14 16 18 20

Age, Years
Rating (10=new)
SP-2/Alk Oil/ Si Alk
SP-2/Alk/Ep/Ure
SP-6/Alk/Ep/Ure
SP-6/oil-Alk
SP-2/Oil/Alk
SP-6/oil-Alk

Figure 7. 20 year performance of alkyd systems.

Urethane Systems (5)

The urethane systems performed adequately during the first 8 years as a group. Of particular note, the SP-2 surface preparation performed as well as the SP-6 surface preparation. This trend, where performance had little to do with SP-2 versus SP-6 surface preparation, was also seen with the alkyd systems. Another observation is that the 2-coat system was one of the poorest performers. The three better performing urethane systems were all 3 coats with and epoxy intermediate coat. While consistent data on applied thickness was not available for this study, the authors have found through other overcoating research that when surface preparation is minimal, more coating thickness over the “bare” spots equated to better performance.5

Performance of Urethane Systems

0
1
2
3
4
5
6
7
8
9
10

0 2 4 6 8 10 12 14 16 18 20

Age, Years
Rating (10=new)
SP-6/3ct Ure
SP-6/Ure/EP
SP-2/Ure/EP/Ure
SP-2/Ure/EP/Ure
SP-6/Ure/EP/Ure

Figure 8. 20 year performance of urethane systems.

5 “Field Testing Maintenance Overcoating Systems for Bridges,” C.L. Farschon and R.A. Kogler, Journal of Protective Coatings and Linings, January 1997, Volume 14, Number 1, pages 56-67.

Aluminum Systems (8)

Figure 9 shows the performance of the individual aluminum systems. As a class, these systems did not perform well. Of note, the SP-7 surface preparation seemed to perform better than the SP-6 and SP-2. Also notice that the abrasive blasting surface preparations
tended to perform better to the 8 year mark, and then performance across all surface preparations tends to even out. This emphasizes the rather difficult to predict situation where the replacement coating system may not have as good a long-term performance as
a “maintained” original coating system. While this is an interesting observation, note also that all of these systems are D610 of 5 or less, very close to the D610 rating of 4 selected as the “coating system replacement” level of performance.

Performance of Aluminum Systems

0
1
2
3
4
5
6
7
8
9
10

0 2 4 6 8 10 12 14 16 18 20

Age, Years
Rating (10=new)
SP-6/Al Ure/Acry
SP-7/Al Ep/Ure
SP-2-3/Al Ep/Ure
SP-2/Al Ep/Ure
SP-2/Al Ep/Ure
SP-2/Al Ep/Ure
SP-6/Al Ep/Ure
SP-6/Al Ure/Ure

Figure 9. 20 year performance of aluminum based systems.

Epoxy Systems (6)

Figure 10 shows the performance over time for the various epoxy systems. These systems were among the worst performers at the 8 and 20 year inspections. Notice that all of these systems were applied to an SP-6 surface preparation; where most of the prior
lead-based coating would have been removed and where visible amounts of corrosion may have remained prior to painting. Also notice that all of these systems are only 2 coats.

Performance of Epoxy Systems

0
1
2
3
4
5
6
7
8
9
10

0 2 4 6 8 10 12 14 16 18 20

Age, Years
Rating (10=new)
SP-6/Ep/EP
SP-6/Ep/Ure
SP-6/Ep/Ure
SP-6/Ep/Ure
SP-6/Epoxy
SP-6/Epoxy

Figure 10. 20 year performance of epoxy based systems.

Photo 3. Single coat epoxy system.

Barrier Coatings

This study included 30 barrier type coatings and 17 coatings with some kind of zinc metal in the primer. Barrier coatings essentially protect the substrate by separating the environment from the surface. Although some of the barrier systems contained inhibitive pigments, we grouped all barrier coatings together for this analysis. The zinc containing coatings arguably impart some sacrificial protection to a steel substrate and were not considered in this analysis.

Figure 11 shows averaged data for the number of coats in a barrier coating system versus 20 year performance. The trend indicates that more coats equals better performance. Although this trend seems obvious, it is important to consider the nature of the
troublesome areas on a bridge (i.e., those spots that routinely cause low performance ratings). These areas/spots, when maintenance painted, are typically rusted and have no prior coating, so they become “bare spots” after surface preparation. If we look at this
data with coverage of “bare spots” in mind, it is clear that the number of coats applied increased the longevity of the coating system. This data re-affirms the maintenance painting practice of applying spot primers to areas of a prepared bridge with missing coating. It even suggests that more than one spot primer may be appropriate for a longer lasting maintenance overcoating system.

Performance by Number of Coats

0
1
2
3
4
5
6
7
8
9
10

0 2 4 6 8 10 12 14 16 18 20

Age, Years
Rating (10=new)
1-coat systems (3)
2-coat systems (15)
3-coat systems (12)

Figure 11. Performance by number of coats (only non-zinc, non-metal systems).

Conclusions

The original project provides an excellent comparative study of various maintenance painting strategies. While coatings technologies have changed over the 20 years since the test coatings were applied, inspections provide excellent data to form the basis for riskbased decisions regarding maintenance of bridge coatings. The following specific conclusions can also be made:

In many of the instances, surface preparation had less impact on the coating system life than might be expected. Given that surface preparation is a primary cost driver, the opportunity may exist to reduce cost with acceptable (perhaps even negligible) changes in performance.
By far the best performing systems were the metallizing systems. These systems are only just beginning to show rusting after 20 years. Of course, these systems were considerably expensive to apply. Currently, the cost disparity between metallizing and liquid coatings is less than it was in 1987, however the metallizing systems still carry a cost premium.
Of the liquid-applied coating systems, those containing an inorganic zinc or organic zinc primer performed best. The epoxy systems and aluminum-mastic systems performed worst.
The coating systems that are considered traditional overcoating materials (i.e., non-zinc barrier type coatings) had better performance when multiple coats were applied.

Acknowledgements

The authors would like to acknowledge the excellent work and innovative project conducted by NJDOT. The authors would also like to acknowledge Fred Lovett of NJDOT and Bob Kogler of Rampart, LLC for their help during different phases of this project.