Some third-party automation tools (e.g., for parametric optimization) fail to release study handles properly, leaving exclusive locks behind.

If a specific study is corrupted: