- Asset Management
- Solutions or Hardware?
- Key Project Summaries
- Bridge Preservation
- White Papers
- Contact Us
U.S. – Taiwan Bridge Engineering Workshop
Sensor Technology for Assessing Bridge Performance
Peter J. Vanderzee
President and CEO
In recent years, we have learned that visual inspection to determine bridge condition is not necessarily the most effective method to determine if rehabilitation or replacement actions are warranted. Visual inspection of U.S. bridges has been conducted for over thirty-five years under a program called the National Bridge Inspection Standards (NBIS). Nearly ten years ago, the U.S. Federal Highway Administration (FHWA) published results of a study that concluded visual inspection was subjective, highly variable and not sufficiently reliable to optimize bridge capital expenditure programs. This conclusion came as no surprise to practitioners who accept that visual inspection was intended to be overly conservative to assure public safety. However, visual inspection alone has resulted in an overstatement of long-term funding need for bridge rehabilitation and replacement programs.
Given the certainty of insufficient future funding availability (both local and Federal), bridge owners must implement better management paradigms to safely extend the life span of bridges they own. At the owner level, the objective should be to lower life cycle costs by deferring or re-programming (rehab vs. replace) capital expenditure projects. Fortunately, bridge owners now have a variety of advanced condition assessment technologies that can be deployed to gain a much more objective, precise evaluation of actual condition. Given timelier, precise condition information; owners can make more informed, optimal decisions regarding capital expenditures for their bridges, while assuring essential safety margins are maintained.
Importantly, the author is not suggesting that visual inspections should be abandoned. Rather, this paper provides guidelines for when visual inspections should be augmented with advanced condition assessment technologies. This paper also provides experience-based suggestions for scoping project details, such as sensor types and data capture frequency. The business of advanced condition assessment is not research any more – it’s about the highest and best use of taxpayer’s limited funds.
Finally, the author stresses that deployment of advanced condition assessment technologies should provide the owner a return on investment for additional costs incurred. A focus on achieving a return on investment will negate using technology that is more suited for research as owners seek solutions to reduce long-term funding demand for bridge rehabilitation and replacement.
The condition of highway bridges across the United States, as reported by the Federal Highway Administration (FHWA) in their Conditions and Performance Report to Congress, shows limited progress has been made over the past two decades in reducing the number of structurally deficient bridges. While the classification of structurally deficient is not predictive of imminent collapse, there is no question that this number remains stubbornly high (approximately 12%, or 72,000 bridges). Today, the U.S. Congress has few options to reduce this number without a significant and sustained application of billions of dollars in new funding – for decades. Given the current Federal budget issues in the U.S., significantly increased funding is not likely.
After the Ohio River Silver Bridge collapse in 1967, the Federal Highway Administration developed and mandated a State Department of Transportation (DOT) managed periodic bridge inspection program to enhance public safety. This FHWA program, developed in cooperation with the Association of American State Highway and Transportation Officials (AASHTO), is called the National Bridge Inspection Standards (NBIS) program. While a few high-profile bridges have failed and caused loss of life since then, most practitioners agree that bridge inspections were necessary and have resulted in a higher degree of public safety than if they were never conducted.
With few exceptions, a biennial visual inspection by State DOT staff technicians has been the accepted protocol for over 35 years. The author believes this protocol, well intended to be highly conservative, allowing for the subjective and highly variable nature of visual inspection, is not adequate for optimizing the expenditure of billions of dollars per year (both Federal and local owner) in the Highway Bridge Rehabilitation and Replacement Program (HBRRP). New methods must be employed to reduce the number of structurally deficient bridges, while maintaining acceptable levels of safety for transportation users. In essence, we need a better, more cost effective way to achieve a state of good repair for this nation’s bridges.
Using 2008 FHWA statistics, there are approximately 600,000 bridges in the United States exceeding twenty feet in length. Of that total, approximately 72,000 are considered structurally deficient. Another 80,000 are considered functionally obsolete (1). The American Society of Civil Engineers (ASCE) says it will take roughly $140 Billion to correct bridge deficiencies across the U.S. (2). While there are technical options for correcting structural deficiencies, wholesale bridge replacement is the only viable option for correcting functional obsolescence. Structural deficiency is reported to be a $48 Billion dollar problem (3) in the U.S.
Maintaining bridges in a state of good repair is the responsibility of a bridge owner. State DOTs typically own the National Highway System bridges and major State arterial bridges; cities and counties own the remainder and that population includes most of the short-span bridges in a state (<200 feet in length). Cities and counties rely on State DOTs for compliance with the NBIS program, but as owners, they have responsibility for correcting structural deficiencies, erecting load postings, and enforcing weight limits. Bridge inspections are estimated to cost the taxpayers approximately $300-400 Million annually, or about $1,200 per bridge, assuming a standard two year cycle (4). Since most U.S. bridges are short-span, long-span bridges (>200 feet in length) cost far more to inspect. A significant number of bridge inspections are subcontracted by State DOTs to private engineering firms as lack of State DOT staff and specific inspection expertise dictate. Subcontracted bridge inspections generally cost much more than DOT staff inspections.
Bridge inspectors are State DOT employees and can generally be classified as technicians. Inspection team leaders (for a two man team) must have a minimum of five years bridge inspection experience. Program managers and those who determine load ratings must be registered professional engineers (5). Bridge inspection programs are taught by engineering firms and typically require three weeks of initial classroom training, with annual refresher courses. State DOT bridge inspectors (or DOT subcontractors) conduct visual surveillance, take photographs, look for evidence of deterioration and document their observations on a form. It is rare that bridge inspection data is challenged or subjected to a second look, as the cost for that level of quality assurance is very high. In essence, first level, technician-trained bridge inspectors provide the raw data to determine need and priority for billions of dollars in Federal and local funding. It begs the question whether bridge owners and the Federal government (as the primary funding source) would benefit by enhanced information to support and optimize expenditures in the billions of dollars annually?
As bridge inspection data was collected, stored, categorized and analyzed over years by the State DOTs and FHWA, it was used to gain a better understanding of the overall condition of our nation’s bridge inventory. Results from analysis were used to report findings to Congress (6) and support a funding program for Federal matching funds, the major source of funding for bridge repair or replacement. This program is called the Highway Bridge Replacement and Rehabilitation Program (HBRRP).
The prioritization process for the HBRRP has a key metric termed sufficiency rating (0-100 scale) which is based upon visual observation of structural adequacy, safety, traffic volumes, and other germane factors (7). A sufficiency rating below eighty qualifies a bridge for Federal rehabilitation funding; a sufficiency rating below fifty qualifies a bridge for Federal replacement funding. A local owner contribution of twenty percent of the total project cost is normally required.
Nine years ago, the FHWA conducted a project to assess the efficacy of NBIS visual bridge inspection. An article explaining the project and explanation of results was published in Public Roads Magazine in 2001 (8). The analysis showed that NBIS visual inspection was subjective and highly variable (+/- 2 rating grades on a 0-9 scale). A subsequent chapter in a book by an FHWA research engineer and his bridge professional engineer colleague substantiates the basis of this author’s premise: The subjective and qualitative nature of the NBIS fueled concerns that optimal funding decisions were not possible using this data (9). While these findings came as no surprise to bridge practitioners, the NBIS visual inspection process and its subjective, highly variable data are still the determining factor for bridge funding allocations of billions of dollars annually, both Federal and local.
For over ten years, a number of U.S. and foreign companies have offered commercial, technology driven, advanced condition assessment services to bridge owners, from ground penetrating radar for assessing bridge decks to acoustic monitoring of steel or concrete cracking to strain/stress monitoring of major structural components. Monitoring services typically integrate highly accurate sensors, wireless communication and the Internet, allowing bridge owners to ascertain condition remotely and on a near continuous basis. The most comprehensive of these services has been termed structural monitoring and is intended to augment (not replace) NBIS visual inspection when and as warranted. A number of U.S. bridge owners have conducted demonstration projects using this advanced technology (Caltrans, New York State DOT, Pennsylvania Turnpike, South Carolina DOT, and Pennsylvania DOT). However, full scale adoption of advanced condition assessment technologies continues to be delayed for a variety of reasons, most often budgetary issues and resistance to change.
As advanced condition assessment technology demonstration project results have been reported at technical meetings of the Transportation Research Board (TRB), International Bridge Conference (IBC), International Association of Bridge Maintenance and Safety (IABMAS), smaller regional conferences, and technical articles in trade magazines including Roads & Bridges and Bridges Online; technology practitioners are estimating that 20-40% of bridges evaluated using advanced condition assessment technologies are in better to much better condition than NBIS visual inspection initially indicated (10). Again, this is directionally consistent, given that NBIS visual inspection protocols are intentionally conservative, considering their subjectivity and variability.
In 2005, AASHTO convened a bridge practitioner group to develop and publish A Strategic Plan for Bridge Engineering (11). This document, organized to present seven Grand Challenges was also prioritized. The number one priority, or Grand Challenge, was: Extending Service Life in recognition that: A significant portion of the nation’s inventory of 590,000 bridges is rapidly approaching the end of its intended design life (12). In addition, the ASCE, in its annual report card for infrastructure condition, recommends implementation of a risk-based prioritization scheme for the repair or rehabilitation of the nation’s bridges (13).
As bridge owners consider the appropriate use of advanced condition assessment technologies, the author suggests four important questions be asked at the conclusion of this section:
- Is the reported number of structurally deficient bridges sufficiently accurate, or is the NBIS protocol supporting an inflated, overly conservative number of structurally deficient bridges and, consequently, higher overall cost to attain a state of good repair?
- Can owners effectively identify (risk-based) and evaluate (using advanced condition assessment technology) bridges to lower life cycle costs by safely extending life span or re-scoping actions, such as rehabilitation vs. replacement?
- Can bridge owners upgrade daily management of their bridges to enhance safety, minimize cost, and objectively prioritize projects in an era of insufficient funding?
- Can all the above be accomplished while providing a robust return on investment for the taxpayers?
Structural Monitoring Overview
Structural monitoring is an advanced condition assessment technology that can be used to gain an objective understanding of a structure’s health, monitor the progression of deterioration, or expand the safe operating envelope of a bridge. Structural monitoring is enabled by an innovative combination of sensing devices, wireless communication, and the Internet. While not intended to replace visual bridge inspection, structural monitoring should be considered best practice for objectively determining the actual condition of a bridge, after visual inspection rates a bridge (especially the superstructure) as structurally deficient and/or determines that certain bridge elements are suspect (such as visible cracking, out-of-plane bending, heavy corrosion, or fracture critical details).
Using long-term monitoring data (more than one thermal cycle), engineering consultants can build a calibrated finite element model (FEM). Using this mathematical model, structural engineers can accurately and conclusively diagnose structural health, determine the location of hot-spots of concentrated stress, see the unanticipated effects of thermal loading (frozen bearings), determine safe load carrying capacity (posting and permitting decisions), and objectively analyze other relevant bridge management issues. In essence, the bridge owner now has an in-depth understanding about how a bridge performs under actual service conditions. Better financial decisions follow this in-depth understanding.
Examples of Value Delivered Using Structural Monitoring
Example #1: A U.S. Tollway received a third party (outsourced) visual inspection (NBIS) report that concluded corrosion of steel members was sufficiently severe to recommend a steel refurbishment project, strengthening the affected steel to safely carry anticipated service loading. When faced with a report from competent professional engineers, the owner had little choice but to follow the recommendation, or face increased liability exposure and, in a worst case scenario, possible failure. The owner evaluated several contractors’ proposals for the steel repair. The lowest price was $875,000. Ready to award the contract, the owner decided to first install a structural monitoring solution to assess the efficacy of the repair program by monitoring the bridge for six months prior to repair and six months after the repair had been completed.
After six months of monitoring, but before the repair started, the owner’s third party engineer had enough information to calibrate a finite element model of the bridge. Upon completion of the model, the owner’s third party engineer fully analyzed the structure and determined that the steel repair program was not necessary. In addition, the third party engineer concluded that the structure had another serious deficiency that had not been reported on the original visual inspection report. Faced with mounting pressure, the owner instead decided to take the safest course and implement the steel repair program. After six months of monitoring after the repair, at a total cost of approximately $125,000, including subsequent detailed structural assessment, the owner’s engineer concluded that his initial recommendation was correct; there had been no measurable improvement in bridge performance as a result of the $875,000 steel repair program. This case illustrates the potential for improving the cost-effectiveness of bridge maintenance and rehabilitation or replacement actions.
The result: the owner could have saved over $800,000 had the structural health monitoring results been considered in the final action taken on the bridge. These resources could have been invested in other areas of greater need across the owner’s bridge inventory.
Example #2: A U.S. State DOT performed expensive repair actions to relieve expected high stress and observed fatigue cracking in localized areas (crack stopper holes and steel retrofits) as a result of fracture critical structural construction details. Monitoring of key areas on the structure for seven months allowed the owner’s third party engineer to conclude that most of the retrofits were successful (good news), but that a severe hot-spot remained, which apparently was made worse from the retrofit (bad news). In addition, thermal loading caused significant stress excursions, indicative of bridge bearings that were not performing as designed. The owner is considering having the monitoring firm reposition a number of sensing devices and continue the investigation with the third party engineer to resolve the hot spot and better understand structural behavior under service conditions. The objective for continued monitoring and analysis is to safely extend the life span of this bridge, which would seriously disrupt Interstate traffic and cost well in excess of $100 million dollars to replace if it failed.
The result: total expenses for this monitoring/analysis project were less than one month’s interest on capital if the structure had to be replaced too early.
Example #3: A U.S. State DOT recognized they did not have sufficient funding to replace two long-span bridges that had severe corrosion on steel members and bracing. However, due to the extensive nature of the corrosion and unknown effects on bridge performance, the owner opted to install a structural monitoring system on each bridge, consisting of both displacement and temperature sensors, automatic data capture and reporting, and display of results over the Internet. Both bridges are fracture critical.
Initial results show that temperature differentials are the major driving force for observed strains in structural members, while worst-case (peak to peak) observed strains typically range from 300-500 microstrain. These initial results are similar to other bridges in that, despite structural concerns, the bridge members are performing within a tolerable range. More data will be collected over the next two years to verify or upgrade the bridge condition classification.
The result: Despite visual inspection classification of structural deficiency, both bridges have reasonable reserve capacity for safe extension of operating life. Costs for the solution implemented suggest a robust return on investment.
Guidelines for Structural Monitoring Solutions
The previous project examples demonstrate the value of deploying structural monitoring, a more precise, objective asset assessment technology available to bridge owners. Importantly, like most specialized analytical tools, structural monitoring is not necessary for every structure, but rather for structures that, when properly and objectively assessed, have a high probability of producing significant cost-savings for the owner, which generates a return on investment.
Suggested guidelines for consideration of structural monitoring condition assessment solutions include the following:
- When a repair program is being considered for superstructure corrosion (loss of section), typically exceeding $300,000 dollars (USD).
- When a major replacement project is being considered for superstructure structural deficiencies that exceeds $2 million dollars (USD).
- When a concrete or steel member has significant visual cracking that can propagate rapidly, without warning, such as fatigue of steel.
- When unexplained member movement or out-of-plane bending is revealed by visual inspection and is apparently the result of service loading.
- On key, short span bridges having load postings that severely restrict commercial traffic, causing significant detours and added costs.
- On bridges that are frequently permitted for heavy loads, to objectively determine if the heavy load movement has damaged the bridge (insurance recovery).
- On bridges that are structurally deficient, but simply cannot be replaced due to lack of funding – immediate warning alerts should be sent to responsible parties 24/7 in the event of high observed strains or other deleterious events.
- When a vital bridge in a seismic region must be evaluated as quickly as possible after a major seismic event.
Sensor Technology Fundamentals
Like most solutions, one size does not fit all. Owners should be judicious in their use of structural monitoring solutions, as outlined in the previous section. However, once a structural monitoring solution is considered appropriate, the author offers the following technology and solution specification guidance, based upon his experience:
Data intensity is over-rated. Experience suggests that capturing and returning data six to ten times per day is sufficient for detailed analytics, including building a calibrated finite element model (FEM). Typically, we suggest data collection times at peak and non-peak traffic times to capture best and worst case loading events. Despite that confirming experience, we continue to question why certain owners, academics, and other solution suppliers suggest capturing data at high frequencies, only to be overwhelmed by data and consequently, result in added costs for the bridge owner.
Sensor accuracy is over-rated. Bridge owners can specify and purchase structural monitoring solutions that can provide accuracy to one microstrain. While that appears useful, owners should ask if they need to spend more to make certain the bridge member is experiencing 399 or 400 microstrain. Will the owner make a different decision based upon one microstrain – more or less? We believe this level of accuracy not worth the added investment, especially since there is no actionable value in knowing such small variations. When capturing data, we are finding that accuracies of 25-40 microstrain are more than adequate to conduct sophisticated analytics, including building a calibrated finite element model.
Only use sensors that fit the need. There are a substantial number of sensor types that have been developed over the past twenty years. Tilt meters, accelerometers, and even corrosion sensors are available that can provide data regarding bridge or member condition. The question all owners should ask is this: Will this sensor provide data necessary for subsequent analytics to determine the actual condition of my bridge? If a particular sensor isn’t essential for that purpose, adding such sensors only have research value and detract from the owner’s return on investment. We believe that displacement (strain) and temperature sensors are essential for structural monitoring. Other sensors should pass the test of necessity before deployment.
Monitoring periods are under-rated. Short term monitoring is only useful if you are conducting a static load test to determine a bridge’s ability to safely carry a known load. In service, a bridge sees loading from a variety of sources over time, most importantly a range of applied live load and thermal cycles. This author’s experience suggests that structural monitoring is most effective if the owner captures data over a full thermal cycle, not just for one to two months. We have seen, conclusively, that thermal cycles can account for more than 80% of the observed strain in a member. Having strain and thermal data over a longer period also supports development of a finite element model.
Minimize sensors to start; allow for progressive diagnostics. The fewer sensors to start, the better. The author suggests using a minimum number of sensors to capture essential information for subsequent analytics. But allowance should be made in the system controller for inexpensive expansion or movement of sensor locations as more information about structure condition becomes known. Fewer sensors avoid the data overload others have reported as unwieldy and expensive. Also, fewer sensors generally result in a more reliable overall system. Hardware and software installed in outdoor environments with potentially harsh environmental conditions tend to experience lower overall reliability as system complexity increases. As more diagnostic information leads to a better understanding of structure hot spots, sensors can be moved to new locations for enhanced monitoring effectiveness.
Professional installation is essential. Climbing over and around bridges, especially large, long-span structures, is not a job for amateurs. There are substantial safety risks and only committed professionals with adequate safety training and gear should be used for installation of structural monitoring solutions. The author specifically warns against using graduate students, who may not have the requisite insurance coverage or safety training. Also, if the monitoring period exceeds three months, we strongly advocate use of conduit for cable runs. Bridges are frequently visited by birds and small animals who like to chew insulation; conduit prevents this from happening. Of course conduit is more expensive, but this is clearly a scope item that supports overall system reliability, consistent data capture and minimizes potentially expensive rework.
A professionally managed Network Operations Center (NOC) is crucial. Consider the impact of a bridge owner not receiving an alert message when a bridge member experiences strain or crack propagation that is above an established threshold. The NOC must be owned and managed by professionals who can provide believable reliability statistics on the order of at least four nines, e.g. .9999; the equivalent of five minutes of downtime per year. A situation where the owner’s data is stored in a server, located in a closet, with minimal daily management, no back-up power supply or adequate back-up servers strike this author as extremely risky. Why take that chance when the added cost for superb reliability is relatively low?
Confirm a return on investment. If a bridge owner wants a structural monitoring solution, solution providers should be able to support a return on investment. Various non-quantifiable value drivers can be discussed, such as added safety margin, but this author recommends use of hard-number financial analysis based on extending asset life as the primary metric. By simply accounting for deferrals of capital expenditures (cost of rehabilitation or replacement project) at reasonable interest rates (3-7%), and factoring in the probability of successfully achieving that result (25-40%), owners should be cognizant of a robust financial return as the main reason for spending money to get enhanced bridge condition information.
Advanced condition assessment technologies are ready for deployment and use to help bridge owners safely defer expensive rehabilitation and replacement projects. In particular, structural monitoring solutions have been repeatedly shown to provide an objective basis for safely extending asset life while providing bridge owners with a robust return on investment. While bridge owners can choose among a few technology platforms for structural monitoring, we strongly recommend that owners conduct due diligence before specifying system needs and committing to spend significant sums of money on these solutions.
As is true for most financial decisions, it pays to do your homework.
4. Private communication with Tom Everett, FHWA, August 7, 2009
5. Code of Federal Regulations; 23 CFR 650, §650.309
6. FHWA‟s biennial Conditions and Performance Report to Congress
7. Code of Federal Regulations; 23 CFR 650, §650.409
9. Transportation Research Circular, Number E-C104; 50 Years of Interstate Structures; September 2006; Walther and Chase.
10. LifeSpan Technologies; White Paper #2; www.lifespantechnologies.com