The concepts of reliability and resilience are often treated synonymously or conflated, leading to painstaking efforts to distinguish between them. While both are complementary and mutually reinforcing, these concepts can produce competing behaviors when it comes to making investment decisions in a resource-constrained environment.  Resource allocation decisions need to strike a balance between achieving the two objectives. 

In general, reliability focuses on many, short-term service interruptions while resilience focuses on infrequent but large-scale service interruptions such as major outage events (MOE) or Black Sky events (BSE).  In a seminal work defining the differences, Dr. Paul Stockton identifies three levels of outages and points out that while reliability metrics are complementary to and very useful for engendering resilience, they nevertheless fall short in attracting regulators’ willingness to allow rate changes for resilience. 

He further points out that some states exclude outages of greater than “X” days when calculating performance measures, such as the System Average Interruption Duration Index (SAIDI), “because including them would distort assessments of utility performance during normal operating days.” In other words, the main focus is placed on maintaining a steady level of service rather than on restoring service during a major disruption—like on rearranging the deck chairs on the Titanic instead of addressing the possibility of an unimagined sinking!

We come to trust in and take for granted the services provided by our supporting infrastructures, especially as they become increasingly reliable through well-meaning investment. We have become comfortable with their performance but are increasingly over-reliant on their reliability!  Myopically, we invest in increasing their reliability from 2- to 3- to 6-sigma deriving only ever-smaller incremental benefits. We may approve rates focused on increasing reliability and efficiency—which address only the ‘now’—while ignoring or under-resourcing efforts to encourage and improve resilience to the level required to contend with a 2-, 3-, or 6-sigma event. 

Let’s broaden our thinking about disaster and move from ‘single-loop’ learning to ‘double-loop’ learning and beyond. We must revisit our goals and decision-making rules and allocate investment across a broader spectrum of risk probabilities—addressing risk across the full spectrum of likelihood and focusing on improving both reliability and efficiency. As Dr. Stockton points out, measures of effectiveness that drive regulator-approved rates should be broadened in scope to stimulate thinking, planning, and contending with Black Sky events.

References:
Dr. Paul Stockton, Resilience for Black Sky Days, Supplementing Reliability Metrics for Extraordinary and Hazardous Events, National Association of Regulatory Utility Commissioners, 2014 https://pubs.naruc.org/pub.cfm?id=536F42EE-2354-D714-518F-EC79033665CD

By: John Organek