The economic progress and sustainable development of our society need to rely on reliable and durable civil engineering infrastructure facilities. Investment in road infrastructure boosts the growth potential of a national economy, which can be fully exploited by efficient utilization of the road infrastructure. The lower bound on the economic benefit of road infrastructure is estimated to be between 4% and 10% of Gross Domestic Product (GDP).
Bridges are critical components of the road infrastructure as they ensure fast and safe passage over otherwise hardly surmountable obstacles. This means that the bridges mustn’t jeopardize the safe and fast travel. India has second largest road network in the world of over 5,472,144 km which includes 100,087 km of National highways consisting of 36,673 bridges. Out of these, over 6551 bridges are structurally inadequate and over 4807 bridge are functionally inadequate. On an average 44% of the bridges are in the age of 24 to 49 years and 14% are above 50 years (Source: IBMS). Considering that there is a significant amount of aged stock in the bridge inventory, it is evident that the longitivity of these old bridges is difficult to define, unless advanced testing and analysis cum prognosis techniques are employed to evaluate their structural safety and residual life.
Bridge mishaps which happened in the past two years in the country has highlighted the need for modern research to be employed for their residual life assessment. The 100 years old bridge connecting from Mumbai to Goa crossing over the Savitri River in Raigad district of Maharashtra collapsed on 2nd August, 2016, see Fig. 1. This accident resulted in 26 casualties and 14 people being missing.
Another collapse of a bridge connecting a village in Araria district of Bihar state on 18th August, 2017 is feared to have drowned 3 people, see Fig. 2.
The latest incident of the Gokhale bridge footover collapse on 3rd July, 2018 in Mumbai, although with no casualties led to loss of connectivity by stalling the suburban railway for 13 hours.
It is evident that bridge failures are unpredictable. This is because of an absence of assessment and prognosis techniques available to owners of bridges to fulfill the task. Availing which the residual life can be predicted and the probability of failure of a bridge during its service life can be computed. In this article, recommendations are made that facilitate the management of existing bridges in the future. Infrastructure Risk Management (IRM) is a company which specializes in ageing management of degrading bridges and possesses necessary skills to provide training to help setup management systems with the bridge asset owners such as Ministry of Road Transport and High¬ways (MoRTH), the Ministry of Railways (MoR), State and Central Public Works Departments (PWD), National Highways Authority of India (NHAI), etc. Infrastructure Risk Management envisions the birth of Indian Bridge Management Industry whose volume of work is estimated to be USD 750 Million.
The current approach and tools available to bridges managers are hardly adequate to cope with predicting bridge failures. In the following sections the the significant challenges are outlined and later the gradual evolution of these tools towards more efficient bridge management is outlined.
Safety and serviceability are the fundamental concerns in bridge design. However, in course of time, the existing bridges are characterized by the several generations of the design codes. The codes change, both regarding traffic actions and bridge resistance models. It is therefore not surprising that the old, otherwise undamaged bridge may not fulfill safety and/or serviceability requirements for the current traffic loading.
Furthermore, safety and serviceability can be jeopardized by deterioration processes. The resistance of deteriorated bridges can in time reach a level, at which there is an immediate danger of structural failure.
The ever increasing traffic volume and loads are probably the most significant challenge regarding the existing bridges. The traffic mix is shifting towards heavy weight vehicles, effectively increasing the occurrence probability of load situations that exceeds current design loads.
In summary, a prudent and proactive bridge owner needs to plan and execute timely intervention in order to cope with following challenges:
- Potentially unsafe bridges designed using bygone codes of practice
- Bridges with reduce resistance due to deterioration or mechanical damage.
- Increasing traffic volume and loads that can render some bridges unsafe.
2.2 Conventional assessment and decision making
Current bridge maintenance practice relies on visual inspections. Based on the visual condition state, the owner often triggers costly in-depth investigation or even maintenance actions. Many a times, once an in-depth investigation is triggered, the maintenance intervention is very likely to follow, even if a bridge can still be used without restrictions. The reason is surely the visual appearance and related perception of safety that entice the owners to remove all visible damages. The state of inspection/repair/retrofit is adhoc, meaning that the owners act only when there is some visual distress or when there is a public uproar or if some sudden shock like an earthquake or tsunami is hit. In most of the cases, it doesn’t give a cost effective solu¬tion given the remaining service life and possibilities to reduce uncertainties on existing bridges.
It is not that bridge owners and operators are not aware of this efficiency, but to remedy it, they need to be provided tools and resources to efficiently store and access the elaborated information on their bridge inventory.
Theoretical deterministic formulas, for example corrosion growth model, fatigue cracking equation, etc. are not very good predictors, when compared with the actual NDT/SHM measurements over time, see Fig. 4. Additionally these measurements do not follow a perfectly smooth trend, but rather show an irregular pattern with a scatter which the model is unable to capture. Clearly available models work only upto a certain extent, because the reality of degradation is very complex to be modelled sufficiently through them.
Since inspection results are instantaneous state they cannot be used independently for forecasting degradation. Additionally uncertainties are introduced through inspections due to:
- Inherent variability of the measured parameter: Degradation is not uniform, for example the amount of corrosion is never the same all over a bridge deck.
- Measurement errors: If the same quantity is measured twice by different engineers or using different NDT instruments, the result is never the same.
- Statistical uncertainty due to a limited number of measurements: Sufficient inspections are seldom possible due to constraint of time and money.
Bridge Management comprises of bridge assessment following the inspections and maintenance planning. The short term maintenance planning is aimed at interventions that are to be taken shortly after in-depth investigations and structural analysis. The mid to long term maintenance planning is a process, in which different interventions scenarios are developed. The goal is to estimate financial and other needs well in advance and avoid unpleasant surprises. Furthermore, proactive planning allows choosing the optimum time for interventions and reducing long term costs without compromising safety and serviceability requirement. For this purpose, safety and serviceability forecasts over time are necessary to define the time instance at which, at the latest, an intervention is inevitable.
3.1 Inspections and impact of damages
The current inspection procedures need not change significantly. However, some additional information to account for the effect of deterioration and damages is to be appropriately considered in assessment of bridges:
- Based on the design documentation, relevant failure modes need to be defined. These failure modes corresponds to the critical load situations used in design and
- For each failure mode, vulnerable zones are to be defined, in which damages have the largest impact on safety and serviceability.
3.2 Deterioration Models
The key ingredient of mid to long term maintenance planning is a deterioration model. The deterioration model allows condition state forecasts of assessment units such as whole bridges, bridge components (super-, substructure and equipment), bridge elements and determine possible maintenance interventions in the future.
Uncertainty plays a major role in the de¬terioration process and our understanding of it. Available models are not good enough for forecasting and the NDT/ SHM gives instantaneous condition. In order to be able to manage a degrading bridge over its service-life, it will be beneficial to combine the predictive capability of the theoretical degradation model and the data obtained from NDT/SHM. However, because of the uncertain nature of degradation process and the inadequacy of any degradation growth model, it is necessary to setup the formulation in a probabilistic framework using Bayesian technology. Fig. 5 shows a schematic of the proposed fusion.
3.3 Reliability analysis
The modern codes define the safety and serviceability in terms of reliability i.e. the probability that a bridge will be fit for its desired purpose during its service life. For example, in EN 1990, 2002 the target annual reliability index for safety is 4.7 and for serviceability 2.9. The bridge is considered as safe and serviceable or intervention-free service life, if specified reliability indexes are not below these target values. Assessing the reliability of existing bridges is a specialised task as one needs to model all actions and material properties as stochastic variables including the effect of deterioration.
3.4 Predictive maintenance
Predictive maintenance is a philosophy or attitude that, simply stated, uses the actual condition to optimize the total maintenance operation. The common premise of predictive maintenance is that regular monitoring of the actual structural condition, will provide the data required to ensure the maximum interval between repairs and minimize the number and cost of unscheduled outages created by failures. Predictive maintenance is a condition-driven maintenance program, which, instead of relying on bridge average-life statistics (i.e., mean-time-to-failure) to schedule maintenance activities, uses direct monitoring of the condition, and indicators to determine the actual failure probability.
The addition of a comprehensive predictive maintenance program will provide factual data on the actual structural condition of the bridge. This data provides the maintenance manager with information for scheduling maintenance activities. A predictive maintenance program can minimize unscheduled breakdowns of the bridge and ensure that repaired component is in acceptable condition. The program can also identify problems before they become serious and minimized if they are detected and repaired early thereby preventing major repairs.
3.5 Improvements to currents approach
In summary, the proposed improvements allow decision maker to
- estimate remaining intervention-free service life of a bridge in rational manner, using reliability assessment,
- considered the improvement in safety and serviceability and compare with the costs to obtain this improvement, and
- plan reliable financial needs and needs for other resources on mid to long term, considering the forecasts of traffic loads.
Schematic of the approach is shown in Fig. 6.
The need for more economic utilization of transportation infrastructure is a challenge for bridge owners. They need to have readily available, high quality information on their bridge inventory in order to cope with gradual deterioration, growing traffic demands and increased frequency and magnitude of natural hazards. In this article, the methods are proposed that use data that is already available and allow their gradual refinement. This allow better exploitation of inspection results to assess safety and serviceability and improved decision making regarding in-depth investigations and maintenance interventions.
- Faroz S. A. and Ghosh S. (2017). Falling Short But Not Falling Down: Challenges and Solution for the Service-Life Estimation of Gradually Degrading Bridges. The Masterbuilder 19(7), 144–148.
- Faroz S. A. (2018). Service-Life Assessment of Gradually Corroding Bridges- A Study. Coatings and Anti Corrosion Engineering Review 9(1), 34–38.
- Faroz S. A. and Ghosh S. (2018). Bayesian Integration of NDT with Corrosion Model for Service-Life Predictions, In 9th International Conference on Bridge Maintenance, Safety and Management (IABMAS 2018), Melbourne, Australia