When 242 Pipes Stand Between a County and Its Emergency Room
EMBARGOED UNTIL August 18th, 2026
Pipes under rural roads are barely tracked. New research shows what happens when they fail.
AUGUST 18, 2026 (Herndon, VA) – People rarely think about culverts until they are gone. These plain pipes and boxes carry roads over streams and span less than 20 feet — the federal cutoff for what counts as a bridge. That means road-stream crossings stay out of the National Bridge Inventory and therefore, off the inspection schedule. Yet, when Tropical Storm Irene tore through New England in 2011, it destroyed more than 960 of them in Vermont’s Green Mountains alone and left 13 communities entirely cut off for days.
A forthcoming study in the journal Risk Analysis is the first to ask what happens to a whole community when many of them fail at once. Using public data, a team made of researchers from the University of Massachusetts Amherst and an engineering firm modeled the impact of culvert failures in the Deerfield watershed, which is home to 783 road-stream crossings.
Key findings:
- Culverts fall through a regulatory crack. Anything spanning more than 20 feet counts as a bridge and gets regular federal inspection and a public rating. Culverts under that threshold get neither, and they are most common exactly where they matter most: rural roads that are narrow, lightly traveled and short on alternate routes. Several states have improvised their own rating systems, for example Utah uses a 5-point scale, and Massachusetts uses a 9-point scale.
- Nearly a third of the culverts studied are single points of failure. Of the Deerfield watershed network of 783 culverts, 242 turned out to be articulation points, meaning one failure cuts off at least one neighborhood from the hospital. No cluster of failures required; just the one. In Vermont, a single collapsed culvert added as much as 70 miles to some residents’ drives.
- Disconnection leaves people stranded. Across the simulations, many scenarios left more than 30% of the population without a route to the hospital. In a western Massachusetts watershed of 42,498 people served by a single hospital, roughly a third of residents can lose their route to the emergency room.
- Which culverts you call critical depends on how you ask. A method based on network efficiency clustered its picks along a single busy corridor and missed most single points of failure, because severing a sparsely populated area barely moves an efficiency score. However, weighting disconnection and travel time equally spread its picks across the map and caught far more of them. For agencies deciding what to fix first, the choice of metric is what makes the decision.
- Counting broken culverts tells you almost nothing about the risk to a community. The same number of failed culverts produced wildly different community outcomes depending on which ones failed. The authors flag this as a problem for how damage gets reported after disasters, since the standard public tally is the number of components damaged, and that data is a poor proxy for how many people lose access to care.
- Standard vulnerability indexes would have missed the at-risk neighborhoods. Researchers tested whether disconnection tracked income, age or race across the study area’s 36 census block groups. The correlations came back weak at every hazard intensity, meaning there was no systematic bias to find. That is itself the finding. Composite tools like the CDC Social Vulnerability Index can obscure the specific places that need help. Mapping single factors against disconnection instead surfaced in particular pockets, including areas that skewed elderly and are likely to be cut off. In a watershed where nearly 26% of residents are over 65 and median household income runs as low as $21,830 those are the addresses an emergency manager needs to know.
Authors Available for Interview:
- Jessica Boakye, corresponding author, department of civil and environmental engineering, University of Massachusetts Amherst
- Joshua Govina, Lochner, Tewksbury, MA
- Egemen Okte, department of civil and environmental engineering, University of Massachusetts Amherst
“Modeling Culvert Failure and Community Accessibility Impacts During Hazards,” Risk Analysis, August 18,2026. DOI: 10.1002/ISBN.stat00999.pub9
###
About Risk Analysis
Risk Analysis is a peer-reviewed journal publishing original research on the assessment and management of risks across disciplines including public health, engineering, environmental science, social science and policy. Risk Analysis, founded in 1980, is published by Wiley on behalf of the Society for Risk Analysis.
