ABSTRACT
Bridge maintenance and preservation activities are widely accepted to slow bridge deterioration rates and extend service life. In order to reflect the quantitative impact of bridge maintenance and preservation on bridge performance, State Departments of Transportation (DOTs) typically have utilized deterioration models or service life extension values that are based on expert elicitation in the United States. Research on quantifying the benefits of bridge preservation and maintenance, based on actual bridge data has been limited, if any. Consistently, the variables used for quantifying the impact of preservation in past research studies have either been assigned by expert elicitation, or the service life of treatments, as provided for specific bridge preservation products. Also, performance of bridge preservation treatments and their impact on bridge condition is highly dependent on bridge characteristics and external factors (e.g. traffic volume, climate, bridge condition). Therefore, there is imminent research need to quantify the impact of bridge preservation based on actual data sets for state DOTs. This paper presents preliminary research findings from an ongoing project in Minnesota. To date, the maintenance history data gathered by the Minnesota Department of Transportation (MnDOT) has undergone comprehensive review and evaluation. This data has been integrated with pertinent historical bridge condition information, and a preliminary modeling process has been conducted to forecast the extension of service life for deck components. Minnesota bridge decks were partitioned into four clusters using K-Means clustering analysis technique. Markov transition times for deck component ratings were computed separately for these clusters. Decks with recorded repair and maintenance had significantly higher transition times for ratings 7, 6, and 5 which correspond to good and fair conditions. The findings indicate that deck repair and maintenance extend the service life of bridge decks and significantly delay expensive bridge replacements. Future research steps include computation of Markov transition times for select bridge components and elements, bridge-, component- and element-level life cycle cost analysis and development of decision trees.
