The collapse of a tailings dam at the Samarco Mine in the Brazilian state of Minas Gerais on November 5th 2015 suddenly released around 62 million cubic meters of semi-liquid mining waste. The subsequent "wave of mud" (Creado and Helmereich 2018) travelled 879 km until reaching the Atlantic ocean, leaving a landscape of massive destruction, pollution, and death in its wake. The magnitude of the disaster immediately turned the spill into the worst environmental disaster in Brazil's history, according to most commentators. Then, on January 25th 2019, the collapse of another tailings dam at the Brumadinho Mine, located in the same Brazilian state and owned by the same mining corporation, killed at least 248 people, mostly mine workers, becoming the deadliest industrial disaster in Brazil's history. In just three years, a single region and company experienced two of the most polluting and deadliest disasters, not only in Brazil's history, but in Latin America as a whole. At the center of both disasters was the sudden collapse of a humble piece of mining infrastructure: a tailings dam. Simply defined in technical mining literature as "purposebuilt sedimentation lagoons" (Lottermoser 2007, 157), tailings dams are the place where all the unwanted leftovers of the mining extraction process come to rest, forever, materializing the "ultimate sink" (Tarr 1996) principle behind most mining enterprises worldwide. In most aspects, tailings dams appear as the ultimate example of a "boring" infrastructure, or complex pieces of technology that are "designed to become invisible as [they are] stabilized" (Lampland and Star 2009, 207). But then something happened. The ever-precarious balance between water and minerals that holds tailings dams together shifted in favor of water, and the dam collapsed. We will likely never know the exact moment or place where the collapse of the dams began, but in both cases the result was the same: destruction, death, and pollution. Everywhere. Brazil's twin mining disasters dramatically show us how infrastructures never become completely stabilized, or even invisible, in the first place. As a large body of STS literature has explored, instead of being perfectly immutable devices, infrastructures are highly paradoxical thingslarge and global, yet local and intimate; malleable, yet