U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590

Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-HRT-06-139
Date: October 2006

Traffic Detector Handbook:Third Edition—Volume II


A number of studies were conducted during the 1980s to determine the extent and causes of inductive-loop detector failures. The underlying purpose of those investigations was to identify potential solutions to increase reliability and life expectancy of inductive-loop installations. Several of the studies involved testing various loop configurations and selected materials in agency facilities. None of these particular studies, however, tested all possible configurations or all available materials or products.


Four studies were particularly significant in that they contained the results of questionnaire and interview surveys, which documented the experience of numerous agencies. These studies are listed below and their results are summarized in this appendix:


The Oregon study was designed to:

  1. Identify the causes of loop system failures and the methods used to remedy the identified problems.
  2. Assess the accumulated data.
  3. Test and evaluate those materials and procedures that appeared most likely to be effective in the Oregon state system.

The required information was obtained by distributing a lengthy questionnaire to the following State departments of transportation:

Some local agencies also participated in the survey.


The Oregon study reported two major causes of loop failure: loss of adhesion in the sealant and poor installation techniques.

Although different materials and installation methods were used by the agencies surveyed, two major causes of loop failure were consistently reported: loop sealant failure (loss of adhesion) and poor installation techniques (e.g., inadequately cleaned or dried saw slots). Another key problem common to many agencies was wire breakage due to asphalt pavement deterioration. The City of Albany, OR, cited pavement flexing and shoving, which caused the epoxy sealant to pop out of the saw slot, requiring resealing. The Portland, OR, metropolitan region had many loop failures because of road wear down to the loop wire, primarily in wheel rut locations. Many agencies also experienced loop loss resulting from trenching for the installation or maintenance of underground utility lines.

Alaska reported no failures when using preformed loops.

Of particular interest was Alaska’s exclusive use of the preformed loop for all of their roadway-installed vehicle detection. Their preformed loop is composed of #12 AWG cross-linked polyethylene wire encased in 1-inch (2.54- cm), schedule 80 PVC conduit. The loop wire is placed in the conduit, which is then assembled to the specified loop dimensions. Alaska reported no loop failure using the preformed loops. Other States also reported selected use of preformed loops. Oregon uses preformed loops mostly in signalized gravel or dirt detour areas around bridge construction sites. California and Utah have experimented with preformed loops and have achieved satisfactory performance with no loop failures reported.

Table M-1 summarizes the Oregon survey results. With the exception of the State of Utah, most loop installations were performed by contractors.

Table M-1. Oregon inductive-loop failure survey results.
 Percent installed by  
StateStateContractorMajor failuresRemarks
1090No loop failures reportedExclusive use of preformed loops
595Improper sealing and foreign material in saw slotUses preformed loops in poor pavement and dirt detours
1090Improper sealingNo failure for loops made of #20002 cable
1090Improper sealing
595Improper sealing and pavement deterioration
1090Improper sealing
7030Improper sealing and pavement deteriorationUsed some preformed loops with no failures
1090Improper sealing and foreign material in saw slot Need better inspection to improve loop performance


Previous to this study, the Oregon vehicle detection system consisted of a 3- by 3-ft (0.9- by 0.9-m) diamond loop, constructed with four turns of THWN wire. The loop was installed in a 1/4-inch (6.5-mm) wide sawcut and encapsulated in a preapproved sealant. The splice between the loop wires and the lead-in cable was made in proximity to the loop and was soldered, covered with heat-shrinkable tubing, and encapsulated with sealant. The lead-in cable (Belden No. 8720) was installed in the sawcut from the loop to the curb or edge of pavement and then in a conduit underground to the controller.

Based on the research study results, Oregon has amended many of their specifications and practices. They continue to specify a 3- by 3-ft (0.9- by 0.9-m) diamond configuration when the loops are in series in the field. However, any single loops are now required to be 4 ft by 4 ft (1.2 m by 1.2 m). Backer rod to hold down the loop wires is now specified on all State projects. A significant change is that the splice between the loop wires and the lead-in cable is now made exclusively at pull box locations.

Presence loop configuration evolved to its present specification in order to accommodate the detection of smaller vehicles in the traffic stream. Two 3- by 3-ft (0.9- by 0.9-m) diamond loops are connected in series. A single 4- by 4-ft (1.2- by 1.2-m) diamond is placed upstream to provide a long loop effect. Information from this detector is input into a separate electronics unit channel so that the Model 170 controller carry-over feature may be used.

Presence type operation is used on left-turn lanes and most side streets. Loop spacing is now 4 ft, 12 ft, and 60 ft (1.2, 3.6, and 18 m) from the stop line as shown in figure M-1.

Figure M-1 shows two 3-foot by 3-foot (0.9-meter by 0.9-meter) diamond loops connected in series with the first at the stop bar and the second 12 feet (3.6 meters) from the first. A third 4-foot by 4-foot (1.2-meter by 1.2-meter) loop is placed 72 feet (21.6 meters) from the loop in front of the stop bar to provide the effect of a long loop.
Figure M-1. Oregon presence loop configuration.

Because of sealant failure incidents, 16 sealant products were tested on both asphalt and cement pavements. Sealant installation was made in July 1983 and was continually observed until the spring of 1984.

The sealant was subjected to a cross section of weather. July, August, and September were very warm, and in December, snow and freezing were present. Cold and wet conditions continued though the winter months. Ten of the sealants tested were approved and added to the approved product list. It was recommended that testing of various epoxy sealants be continued.

Oregon State maintenance personnel expressed interest in enclosing the loop wire in a 0.25-inch (6-mm) vinyl tube prior to inserting the wire in the saw slot. While this technique was worthy of consideration, the State expressed concern about the need for a wider saw slot, extra sealant, and the requirement to twist the loop wire 4 to 6 turns per foot (13 to 20 turns per meter) from the loop to the pull box.

It was apparent from the Oregon experience and from the gathered data that complete cleanout and drying of the loop saw slot are vital to long lasting, effective loop performance. Oregon became interested in the sawcut cleanout nozzle developed by the State of New York. This nozzle uses the Venturi principle to supply pressurized water for cleaning out the slot after sawing. A nozzle was fabricated by Oregon based on the New York design. Valves were added to control both air and water feed lines so that options existed to use air only, water only, or an air/water mixture. This nozzle is being used in loop installations by three of Oregon’s five regions. They report a better bonding of sealant to the saw slot because of the more efficient cleaning of the saw slot by the nozzle.


The information base for the study consisted of a literature search and a State and national telephone survey of traffic engineers. A questionnaire was developed for the survey that could accommodate either verbal or written responses. The survey participants represented 23 cities and 7 counties within Washington State, 6 districts of the Washington State Department of Transportation (WSDOT), and 9 other State agencies, representing a geographic cross section of the Nation. Many of the State and local agencies did not maintain precise documented records related to inductive-loop detectors; accordingly, respondents were asked to provide estimates as necessary. As a result, much of the information obtained was qualitative.


This study was conducted by the University of Washington Civil Engineering Department and sponsored by WSDOT in cooperation with FHWA. The purpose of this research was to identify the types and frequency of loop detector failure and to identify possible solutions. The specific study focus was on the reasons for loops becoming dislodged from the saw slot, the frequency of failure, the optimum products, and installation techniques.

The number of loops within the responding jurisdictions and the corresponding failure rates are given in table M-2. The cities and counties within the State of Washington had the same median number of loops, while the WSDOT districts had a much greater mean, median, and range for the number of loops. County officials reported the lowest incident of loop failures. The WSDOT districts and cities seemed to have relatively comparable failure rates. Out-of-State respondents reported much higher failure rates than those reported in the State of Washington.

The majority of the responding agencies felt that their loop failures were not excessive. Those indicating excessive failures complained that repair procedures were very costly and that loop performance needs to be improved.

When asked how failures were detected, the survey showed that notification of loop failure by Washington cities was primarily from individuals outside the traffic department (i.e., the general public, police, other agencies, etc.). For the WSDOT districts and county agencies, most failures were discovered by their own personnel.

In WSDOT, failures were reported as a consequence of their 12 inspections per year of each installation. Out-of-State agencies reported that failures were most frequently identified by sources outside their responsible department.

Pavement failure was frequently identified as the cause of loop failure. Cracking was listed as the primary mode of pavement distress, while permanent pavement deformation (rutting and shoving), or a combination of this with cracking was also reported as a primary reason for loop failure.

Table M-2. Washington State inductive-loop failure rate survey results.
JurisdictionNumber of loopsFailure rate (% per year)
Cities (23):
Counties (7):
WSDOT Districts (6):
Out-of-State (9)*:
* AZ, FL, GA, IL, MS, NE, NC, PA, and TX.

The most frequent causes for electronics unit failures, as defined by the respondents, included lightning damage and the need for constantly retuning older models to accommodate temperature changes. Other respondents stated that, although retuning was occasionally necessary, it was not considered to be a failure.

The questionnaire asked what methods were used to monitor detector loops. The answers were categorized as follows: periodic inspection, waiting for complaint about a traffic signal, continuous monitoring, or a combination of the above. Periodic inspections were the predominant method, with periods ranging from 1 to 25 times per year per installation. Continuous monitoring was used by some agencies, particularly in areas of high congestion where optimizing traffic flow was a concern.

When asked who installed the loops, the response was contractors approximately 90 percent of the time. Only one State DOT indicated that in-house personnel installed 98 percent of its loops.

Most of the cities and counties within Washington State cited WSDOT specifications for loop detectors. The preferred loop shape was the square loop or quadrupole. Very few agencies specified rectangular or diamond loops.

Out-of-State agencies specified a variety of loop shapes, including the chevron. The quadrupole was given the highest ratings for sensitivity and reliability. Square and rectangular loops scored high on reliability but somewhat less on sensitivity. Most sensitivity problems were associated with vehicles having a high ground clearance such as logging trucks. Most agencies expressed some concern for the detection of light vehicles such as bicycles and small motorcycles.

The preponderance of agencies used only pressurized air to clean out saw cuts. High-pressure water followed by pressurized air was seldom used. The reasoning cited for this was the need for the cuts to be dry prior to the placement and sealing of the loop wire.

The choice of loop wire was fairly consistent among the agencies surveyed. Although Washington State specifies #14 gauge wire with RHH-RHW insulation, many of the city agencies preferred #12 THHN or TWN wire. Out-of-State agencies were about equally divided between #12 and #14 gauge wire.

The methods and products used for sealing the sawcuts and splices varied considerably. Many of the agencies within the State used the WSDOT specifications for sealants. Some city agencies used either polypropylene rope or cotton rope on the top and bottom of the wire to prevent the wire from floating to the surface. Cotton rope was substituted for polypropylene rope because the polypropylene rope tends to deform when covered with hot asphalt, forcing the rope out of the slot. Most WSDOT districts used roofing asphalt as a sealant, while out-of-State agencies tended to use commercial sealants that were tested and approved prior to their use. For splicing, more out-of-State agencies used sealant kits and/or heat-shrink tubing with silicone sealer for their splices than agencies within Washington State.


Primary cause of loop failure uncovered by the WSDOT study was pavement cracking. Secondary causes included wire breakage, poor inspection, pavement rutting, and sealant failure.

The failure rates reported by Washington State agencies were much lower than reported by out-of-State DOTs and in the literature. The largest single factor contributing to loop failure appeared to be pavement cracking. The next most important factor was breaking of loop wires due to utility repair or construction. Other significant factors included poor inspection procedures, pavement rutting, and sealant failure.

Problems associated with the electronics unit could be attributed to the inability of these units to adjust to temperature changes and to moisture or other changes within the loop wire or lead-in cable. This could be resolved by the purchase of self-tuning amplifiers and by complete encapsulation of the loop wires. The study recommended the elimination of using rope to hold down wire.

WSDOT districts expressed much concern about the poor quality of inspection of loop installations. It was pointed out that the best remedy for poor inspection is training and experience. Training may be accomplished on the job by assigning new inspectors to more experienced personnel or through formal classes. Inspectors must remain in their jobs long enough to be effective in the enforcement of specifications.

It was also suggested that forms and procedures be developed to track loop performance. Such records would provide valuable information on the expected life of loops, failure rates, failure modes, quality and reliability of products, etc.

Finally, it was recommended that self-sealing heat-shrink tubing be considered for splicing wires. It is used quite successfully in the marine industry in deep underwater applications and adds extra strength and abrasion resistance to the area of the splice. Heat guns with fixed temperature heads should be used with the temperature selected ahead of time to avoid damage to the wire coating.


Inductive-loop detector failures reported in the MnDOT study were due to broken wires or deteriorated insulation at pavement joints, pavement cracks running through the loops, poor splice technique, treatment of loop corners, and improper sealant choice or application technique.

The Minnesota Department of Transportation (MnDOT) Office of Traffic Engineering determined that 14 percent of their signal maintenance time was spent repairing loop detectors because of poor installation procedures, pavement deterioration, or environmental effects. Consequently, the State Office of Research and Development was asked to conduct a study to identify the reasons for loop detector failure and where within the installation the failure usually occurred.

A literature search revealed that Minnesota’s problems were not unique. Loops installed by other States in the Snowbelt area were also affected by moisture, loop sealant failure, and pavement cracks and joints. A survey questionnaire was developed and distributed to all the Snowbelt States (all States above 36 degrees latitude were considered to be in the Snowbelt region). The District of Columbia and Provinces of Canada were also included, as were the major city and county agencies within Minnesota. The questionnaire covered two critical areas of interest: (1) reasons why the loop detector failed and (2) the materials and procedures used by each agency when installing loop detectors.

Completed surveys were received from 26 States, two counties, two cities, and three Canadian provinces. In addition to the literature search and the survey, MnDOT personnel were observed while making loop repairs.


An initial review of the completed questionnaire revealed that most States, including Minnesota, did not keep good records concerning reasons for loop failures. Characteristically, if it were determined that a short or open circuit had occurred in the loop wire, no investigation was made as to where or why the short occurred. The loop wire was simply replaced.

Many failures occurred because of broken wires or deteriorated insulation at concrete pavement joints, pavement cracks running through the loop, or at the pavement interface with a curb or shoulder. In addition, failures frequently occurred in pull boxes where the loop wires are spliced with the lead-in cables. Only one agency indicated that they did not make their splices in a pull box.

Splice failures occurred because of corrosion, moisture, or poor connecting methods. A number of problems occurred with the "V" formed between wires held side by side as their ends were twisted together. Soldering was favored over crimping, although some agencies crimp and then solder. Many of the agencies sealed their connections with a waterproofing agent and then wrapped with tape. Other agencies encapsulated the connections in an epoxy or other slot sealant using drug store pill boxes or paper bags to serve as inexpensive molds. Commercial sealing kits were also used.

Another frequently mentioned problem was electronics unit failure caused when older models of electronics units become detuned. Deteriorated wire insulation, poor splices, temperature changes, moisture, vibration, electrical storms, and wire movement all contributed to this type of failure. Many agencies have alleviated this problem by using digital self-tuning electronics units.

As for wire type, 25 agencies used #14 AWG wire for their loops. Stranded wire was always used, and type THHN and THWN were favored for insulation. Eleven agencies enclosed the insulated wire in thin vinyl tubing, which creates a conduit effect and provides added protection. Three agencies were reported experimenting with this concept. This tube-encased wire is commercially available or can be made up in-house.

For sealing the saw slots, 12 commercial brands of sealant were mentioned, as were many types of epoxy and polyester resins and unspecified asphalt compounds. Experience and price seemed to be the criteria used in making a selection.

The question of most concern in this study involved the way the various agencies installed their traffic loops. The response indicated three particular procedures were most important in installing loops: (1) laying and sealing the loop wires, (2) addressing the loop corners, and (3) crossing expansion joints or pavement-shoulder joints.

Initial MnDOT procedures used to lay and seal the loop wires merely specified inserting the turns of insulated wire into the slot and sealing with any type of inexpensive sealant. It was soon learned that these inexpensive, untested sealants were unsuitable for preventing water and foreign materials from entering the slot. The sealant would either loose its bond with the sides of the slot or would deteriorate to the point where it would just disintegrate. Foreign materials entering the slot could pierce the wire insulation and allow moisture to come into contact with the loop wire thus causing detuning or possibly a short.

This led to the testing of various sealants to find one that would maintain adhesion to the side of the slot during expansion, withstand compression when joints were contracting, and yet be of the right viscosity to be poured into the slot but not run out if the slot were on an incline. It also became apparent that sharp edges along the slot edges and at the loop corners could create breaks in the wire insulation. Therefore, procedures were undertaken to smooth sharp edges, flush debris from slots with pressurized water, and blow-dry the slot with compressed air.

Cutting 45-degree diagonal slots reduced sharp bends at the loop corners. However, this method created triangles at each corner that could eventually break down from the wheels of vehicles passing over the area. Several agencies resolved this problem by drilling 1- or 2-inch (2.5- or 5.1-cm) holes at each corner instead of cutting the diagonals. The corners of the hole were chiseled out to remove the sharp edges. Other agencies did not drill holes at the corners; rather, they chiseled out the inside of the corner to create small open triangular areas. Slack in the wire should be left at each corner, whether holes are drilled or chiseled to accommodate any movement of the pavement.

Many ideas were presented on how to fill the slots and corners with sealant. Some agencies simply laid the wire in the bottom of the slot and filled it with sealant, hoping that most of the wire would be encased. To better encapsulate the wire, others partially filled the bottom of the slot with sealant, laid the loop wire, and then added more sealant until it was almost flush with the top of the slot. Still other agencies went to the trouble to put sealant in the bottom of the cut and between each turn of the loop wire before final sealing to achieve total encapsulation.

To create a conduit effect without the need for wide sawcuts, some agencies chose to cut a standard slot, lay the wires, and then wedge a form of rope in the slot over the wire before adding sealant. This procedure also reduced the amount of sealant needed. One agency suggested that this method created problems in that the rope was not as resilient as the sealant and that voids could occur between the wires and rope and/or the slot. A conduit effect could be achieved during asphalt pavement construction by cutting the slots and inserting the wire before adding the final layer of bituminous asphalt.

PVC rigid plastic tubing was used as conduit for preformed loops by a number of agencies. This 0.5- to 1-inch (13- to 25-mm) tubing was often installed during construction before the final courses were laid. However, if it were used in an existing installation, much wider cuts would be made to accommodate the tubing, thus weakening the pavement structure. Experience demonstrated that preformed loops were particularly effective where the pavement was broken or in poor condition.


MnDOT contracts out the installation of new traffic signals, ramp meters, and automatic traffic counters, all of which use loop detectors for activation. MnDOT inspection personnel preform inspection of new installations. MnDOT maintenance staff primarily carry out repairs and replacement of failed loops.

Previous to this study, MnDOT did not require its in-house crews to follow the same procedures and State specifications as those required of contractors. For example, a contractor is required to place a layer of sealant in the bottom of the slot before installing the loop wire, while MnDOT crews simply lay in the wire and cover with sealant. In addition, in-house crews use #14 AWG wire instead of the specified #12 AWG. However, MnDOT personnel use insulated wire encased in a thin vinyl tubing, which is not required of the contractor.

It was found that, on occasions, MnDOT crews created problems by choosing poor locations for the new loops or routing the lead-in wires through areas already showing distress. Moreover, they did not follow specifications regarding expansion joints, cracks, or pavement edges.

The primary recommendation of the MnDOT study was that MnDOT repair crews use the same procedures as those required of contractors installing new loops. The study also recommended that consistency be achieved among all crews making repairs. Using MnDOT standard plates as guidelines, it was suggested that the repairmen meet and decide how future repairs would be made. The major points to be addressed included:

Poor installation procedures by contractors were the cause of a significant portion of loop failures experienced in the State. Faulty inspection procedures contributed to these failures. It was recommended that the MnDOT inspectors become more knowledgeable in the proper installation of traffic loops and conduct more aggressive inspections during installation.

A major conclusion from this study was consistent with other survey findings: that records and documentation of loop failures throughout the United States are wholly inadequate. It is common practice to simply replace a failed loop installation without investigating the cause of failure or the precise mechanism of failure. The study concluded that detailed information on loop failures remains critical to determining the procedures and products most appropriate to enhanced reliability and loop life expectancy.

To resolve this inadequacy, MnDOT has recommended that all new loop installations be documented and stored in electronic form. Initial input would include type and size of wire, model of inductive-loop detector electronics unit, sealant type, identity of installer (contractor or in-house crew), etc. This installation would then be tracked throughout its lifetime with all repairs being fully documented.

A new repair sheet would be developed that would include a checkoff list for use during troubleshooting and details on reasons for failures and on repairs made. All replaced equipment would be identified and input into the computer for future reference.


This study was based on an in-depth survey of 9 New York State DOT regions responsible for more than 15,000 loop installations. Although the study did not include a survey of out-of-State agencies, several other State DOTs were contacted to ascertain the types of sealants they used. The survey was initiated after a 1980 study determined that as many as 25 percent of New York State’s 15,000 loop detectors were not operational at any given time. The 1980 study further determined that loop detectors were maintenance-free for an average of only 2 years. Consequently, this investigation was designed to discover the major causes of loop failures and to identify steps to reduce the frequency of failure.


All current loop failures in New York State were listed and a number of installations were inspected to determine if the actual failure causes were readily discernible. In most instances, failures were attributed to one or more of the following factors:

An evaluation sheet was designed to record failure types during the full-scale investigation. A total of 340 failed loops were fully documented and sorted into various categories of major failure type. In all cases, final failure was due to a broken or grounded wire resulting from installation error, material problems, or other difficulties. Table M-3 lists the number of loops installed in each of three pavement types, AC, composite, or PCC. Table M-4 summarizes the survey results by the types of failures that occurred in each type of pavement.

Table M-3. New York State distribution of inductive loops by pavement type in failure survey.
 Total loopsAC pavementComposite pavementPCC pavement
Number 3402752045
Percent 10081613
Table M-4. Inductive-loop failure percentage by cause of failure and pavement type.
Failure typeTotal loops% Failure AC*% Failure composite*% Failure PCC*
Surface Cracking
None Visible53504078
Pavement Heaving
None Visible9088100100
Pavement Patching
No Patching86868091
Sealant Loss
Loop Wire Visible
Not Visible61616562
Broken Loop Wire
None evident78717571
*AC = asphalt pavement
 Composite = asphalt overlay on concrete pavement
 PCC = concrete pavement

Of the surveyed loops, approximately 40 percent had exposed wires and approximately 20 percent had broken wires. Broken or exposed wires were attributed to three major causes: sealant failure, pavement failure (cracking, potholes, and shoving), and wire float. Almost 60 percent of the failed loops that were examined had partial or complete sealant failure.

When the sealant did not achieve a good bond to the sidewalls of the slot, debris and water could infiltrate the slot. Sealant could then be forced out of the slot, exposing the wire to traffic and inducing loop failure. Moisture could also enter the slot and create an electrical ground. Once moisture entered, expansion and contraction of the water during freeze-thaw cycles would further disrupt the loop and abrade the insulation, exposing the wire to an electrical ground.

The most frequent pavement failure affecting the loop wire was cracking. In general, pavement failure caused the loop wire to be strained, resulting in wire breakage, wire insulation wear, or the infiltration of foreign materials. Of the failed loops, 50 percent of the AC pavement and asphalt-overlaid concrete pavement surfaces displayed some cracking.

Loops in PCC without overlays represented approximately 20 percent of the cracking failures. However, this type of pavement presented other problems, such as gross slab movements that strained the wire at the pavement–shoulder interface or where loop wires cross joints.


Major causes of inductive-loop failure uncovered by the New York study were sealant reliability and wire breakage due to pavement failure.

This extensive examination of failed loops revealed two major areas for further study: sealant reliability and reduction of wire breakage due to pavement failure. The survey also concluded that problems existed in the installation process.

Some failed wires were found very close to the surface even when the sealant had adequate adhesion to the sidewalls of the slot. It was concluded that the wire had floated to the surface either before the sealant could cure, or because the sealant remained plastic.

Inspection of splicing methods and materials used in the pull box found that different techniques were applied depending on the individual doing the splice. To guarantee waterproofing integrity and longevity of the splice, better techniques should be evaluated and the best methods incorporated into a Statewide standard.

After a thorough investigation of various installation elements and procedures, the following changes were recommended:

Previous | Table of Contents | Next

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101