When a blade breaks over water, we can tell you where the sea takes it.
A blade failure puts a floating debris field on the move under wind, waves and current. Within the hour, the questions arrive: which shorelines are in play, how soon, and who needs to be told first. Pear models the answer, before anything happens and live while it does.
A moving target, on a public clock.
Floating blade debris does not sit where it fell. It spreads and drifts for days, and the difference between a controlled response and a difficult headline is knowing early which coastlines are credibly in play and when the first pieces could arrive. Guessing is not a plan, and a single line drawn on a map is a guess. The honest answer is a probability envelope: the set of places the debris can plausibly reach, ranked by likelihood, with timing.
What we deliver.
Where the floating debris field can credibly go, from the first hours to weeks out. Envelopes with likelihoods, never a single guessed line.
Which coastlines sit inside the plausible fan and the earliest credible arrival for each, so notifications go out in the right order.
Scenario packs prepared for a farm's own turbines and seasons, sitting ready before anything happens. The answer exists before the question is asked.
During a real incident the picture re-runs on current weather and refreshes as the forecast updates, feeding the response as it unfolds.
The envelope renders on the same common operating picture the response runs on, beside the incident log and the federal forms, in Pear EM.
Every output states what it can and cannot see. A forecast is presented as a forecast, with its limits on the face of it.
Tested against the one event with a known answer.
In July 2024 a blade failed at an operating US offshore wind farm and its floating debris reached shore days later. It remains the one real-world case where the outcome is public and known. When we replayed that event against the weather of that week, the model reproduced the observed landfall. We would rather show you one honest reconstruction than ten impressive simulations of events that never happened.
The results are public. The machinery is not.
We publish what the model produces and the limits it carries, every time it runs. We do not publish how it works. If you need to evaluate it, we will gladly walk through a live scenario with you on your own farm's geography.
If you operate turbines over water, this question has your name on it.
One conversation is enough to see it run against your own lease area. Bring a turbine, a season, and a what-if.
Request a briefing part of Pear EM, the emergency-management platform