Dependency map
A single view of the site and everything it relies on, with each dependency positioned against the hazards that can reach it.
The hazard map stopped at the property line.
Why standard flood layers often fail to reveal operational consequences for facilities and supply chains.
01
The situation
An intense rainfall event exposed a dependency nobody had mapped.
After sustained rainfall, a debris flow moved through the valley floor below a production site. The buildings were untouched — the site had been located deliberately above the modelled flood extent, and the hazard map confirmed it.
Production stopped anyway. The single access road was cut, the medium-voltage substation feeding the site was submerged, and the two nearest logistics operators were both on the valley floor. The facility was intact and unreachable at the same time.
At a glance
- Site situated above the modelled flood extent
- Access road, substation and logistics partners all below it
- Zero structural damage, three weeks of lost output
02 — The question
Is this site still viable, or is it one storm away from the next shutdown?
The board had to decide whether to invest further in the location or begin planning an alternative. The available flood mapping could not answer that, because it described where water goes — not what stops working when it gets there.
03
What we did
We assessed the territory the site depends on, not just the ground it stands on.
Three stages, run over six weeks, combining desk analysis with survey work in the valley itself.
-
01
Locate the dependencies
We mapped every element the operation needs in order to run — access routes, power and water supply, telecoms, upstream suppliers and outbound logistics — and placed each one geographically.
-
02
Stress-test the territory
Each dependency was tested against the hazards present in the valley — river flooding, debris flow and slope instability — at several intensities rather than a single design event.
-
03
Translate into downtime
For each scenario we established what fails, in what order, how long restoration realistically takes, and which failures the company can absorb versus which ones stop production outright.
04
What was delivered
A decision basis, not a hazard report.
Three documents the board could act on in a single sitting.
Scenario matrix
Four disruption scenarios with the resulting failure sequence and expected production loss for each, making the cost of inaction explicit.
Priority roadmap
Interventions ranked by how much exposure each removes — separating what to fix now, what to plan for, and what only relocation would solve.
0
Dependencies mapped
0
Disruption scenarios modelled
0wks
Longest modelled shutdown
We had been assessing the wrong thing for years. The site was never the risk — the valley was.Operations director, manufacturing client
Facing a similar situation?