Estimating Road-Closure Costs with Limited Demand Data

Estimating disruption costs when traffic counts are available but trip origins and destinations are unknown.

Johannes Bleher
Non-peer-reviewed working paper, September 2026

Traffic counts reveal how many vehicles use a road, but the cost of closing it depends on where those journeys begin and end. This working paper develops a method for assessing disruption costs when those origins and destinations are unknown.

The method searches the road network for candidate origin-destination pairs whose routes use the affected link, calculates their additional travel time and distance under closure, and allocates observed traffic across these trips using explicit spatial demand weights. Population, commuting, and other spatial information inform those weights. An optional extension incorporates counts on other roads, conditional on an assumed background traffic pattern.

Two maps showing successive steps of the adaptive car-network search on either side of the B19 Kocher Bridge
The adaptive network search, illustrated for the B19 Kocher Bridge. The animation shows algorithmic search steps, rather than observed traffic movements or estimated costs. Labels are in German.

Experiments on three standard transportation networks and a rural network evaluate cost estimates and mitigation decisions using synthetic demand. The results examine when limited demand information supports useful appraisals: decisions become less reliable near break-even, and finer spatial information does not improve performance on every network.

The B19 Kocher Bridge provides an application of this general workflow. The methodological contribution concerns how demand assumptions affect disruption-cost estimates and mitigation decisions across settings.