Petroleum is still the most important energy source globally and essential to daily human activities. Its exploration in deep waters and the logistics-related operations present the risk of oil spills, with direct and indirect negative impacts on society, the environment, and the economy. In this paper, we bring a systematic risk-based perspective of oil spills, jointly describing why they are harmful, how to prevent them, how to detect them, the possible mitigating actions, and how to remediate affected areas. Such a comprehensive view relies on information about crude oil's physical and chemical properties, consequences, and existing safety barriers for prevention and mitigation. We also provide an in-depth review of the available oil spill detection methods, as rapid identification is essential to enable mitigation and remediation. Most articles used accuracy as a performance metric and Synthetic Aperture Radar imagery; deep learning has shown the most promising results. The proposed oil spill risk-based view subsidizes efficient risk management and communication, and the identified gaps can guide practitioners and researchers in their investigation efforts.