Σ
Anatomy of a Collapse — Ongoing Series
Aerial view of the Brumadinho dam collapse aftermath

Case

01

Anatomy of a Collapse

Brumadinho gave warning.
The contract wasn't listening.

Vale · Córrego do Feijão dam · January 2019

In the 47 days before Vale's Córrego do Feijão tailings dam collapsed, the dam's own monitoring system recorded anomalous readings. The data was logged. It was reviewed. It was assessed against a stability model that had been certified four months earlier by TÜV SÜD. The model said the dam was safe. On January 25, 2019, the dam collapsed. 270 people died.

12 min read Brazil Mining Tailings Law

Date

25 Jan 2019

Deaths

270

Location

Brumadinho
Minas Gerais, BR

Days cert → collapse

47 days

Read case file →

Aerial view of a Brazilian dam collapse area

Case

02

Anatomy of a Collapse

The contract was working.
The dam was not listening.

Samarco · Fundão dam · November 2015

In 2015, the Fundão tailings dam at Samarco collapsed — 19 dead, the Rio Doce contaminated for 663 kilometers, the largest environmental disaster in Brazilian history. Months of anomalous pore pressure readings had been logged and reviewed. The dam had been certified stable in September. The pattern is not accidental. It is architectural. Four years later, Brumadinho.

10 min read Brazil Mining Rio Doce Infrastructure

Date

5 Nov 2015

Deaths

19

Location

Mariana
Minas Gerais, BR

Rio Doce contaminated

663 km

Read case file →

Forest and river landscape in Para state, Brazil

Case

03

Anatomy of a Collapse

The regulation was adopted.
The trigger was suspended.

EUDR · Deforestation Regulation · October 2024

With sixty days remaining before the EU Deforestation Regulation was to take effect, the European Commission announced a one-year delay. The satellite data showing ongoing deforestation had not changed. What changed was the political cost of enforcement.

11 min read EUDR Deforestation EU Law Trigger

Regulation

EU 2023/1115

Delay

1 year

Original trigger

30 Dec 2024

Monitoring

Deforestation continued

Read case file →

The pattern

Both dams had monitoring systems that worked. Both had legal frameworks. Both had recent certifications. Neither had a non-discretionary trigger.

The doctrine that explains this gap is published at hasse.foundation. The analysis applying it to these cases is at Analysis →