The 2020 wildfires along the West Coast caused unprecedented damage, leaving many without power for days. Deciding when, or if, to shut off power can be a precarious decision for utilities.
When facing extreme weather conditions like this, utilities turn to their Public Safety Power Shutoff (PSPS) processes to help determine if shutting off power is necessary. Although most data collection and dissemination related to PSPS is automated, the emergency management personnel ultimately decides whether to turn off a circuit. Emergency management personnel walk a fine line between shutting off circuits too early and depriving critical infrastructure of power or shutting off circuits too late and risking sparking new wildfires1.
Making timely and accurate high-stake decisions during fast moving weather events requires utilities to quickly comprehend and relate complex data sources. How well are the nuanced cognitive requirements of time-pressured decision making supported? How well are ‘human factors’ supported during a PSPS?
With a Small Business Innovative Research (SBIR) Phase I grant from the Office of Electricity, Pacific Science & Engineering Group, Inc. (PSE) of San Diego, CA, has been researching the human factors of PSPS. Human Factors is the science and practice of supporting the needs of humans in complex systems2. Those needs may be overlooked in system design because systems are typically developed by engineers who are neither familiar with, nor know how to specifically support, the task or cognitive information requirements of the system’s end users. This could lead to delayed or flawed decision-making.
PSE partnered with a utility to identify the roles, tasks and decisions made by those involved in a PSPS and analyzed how well display systems and tools supported end users. They focused on the Incident Commander (IC), the person who interprets grid and weather data on their displays in real time to decide if preemptive power shutoffs were warranted. PSE found specific gaps in support for ICs, specifically in supporting them to prioritize their limited available attention to data systems and relating information to make informed judgements and decisions. For example, when the information required to make a de-energization decision was examined, PSE discovered that only a subset of risk factors related to current weather conditions were easily available to ICs, forcing them into a time-consuming and error-prone scavenger hunt for the additional information needed.
PSE worked with ICs and other PSPS utility stakeholders to design an integrated suite of tools that provides the necessary support to the identified PSPs workflow and addresses cognitive issues. By relating each challenge to existing human factors and reviewing cognitive and perceptual science literature, PSE was able to provide a scientific basis for understanding and addressing the challenges through display redesign. To demonstrate the benefits of the redesigned PSPS displays, in addition to collecting subjective user feedback, computational clutter and perceptual salience modeling were applied as a quick and cost-effective means to evaluate the effectiveness of the proposed changes to the utility’s system.
This project showed that a human factors and applied cognitive science approach to a problem in emergency grid management could lead to important changes in PSPS processes.
Applying this approach to training could also offer on-the-job experiences with PSPS processes, preparing the workforce to make these important decisions in the real-world. Learning from experiences rather than through traditional methods will make training more engaging and effective. Training also helps identify additional data sources that may be beneficial in the decision making process. Thus, including a human-in-the-loop (through training) can serve as means of validation, helping ensure that the system design is more robust.
Shutting down a power system during an extreme weather event continues to be a difficult decision, but now decision-makers are armed with clearer information to make the best choice for the utility and the community it serves.