I
The admonition to manage ahead latent interconnectivities before they are triggered into manifest ones by a disaster is easier said than done. What are the practices to do so? Mitigations like retrofitting a bridge, installing automatic shut-off valves, 2 week readiness supplies in advance of an earthquake are examples. So too are table-top exercises, increasing one’s contact list for emergencies, and advanced contingency planning.
Each of these comes with no guarantee that they will actually mitigate once the disaster hits. It might be useful, then, to start with what is guaranteed to happen and see if that offers insights in what to do beforehand and afterwards.
II
One guarantee when major disaster hits: Latent interconnections unmanaged beforehand, particularly those that are invisible or dormant, necessitate improvisations in immediate emergency response afterwards. That is, disaster is the only way these vulnerabilities become visible for management, if any.
Obviously, not all latent interconnectivities are invisible beforehand. To bring to light what can be made visible and manage ahead for them is the function of contingency planning, table-tops, joint drills and other mitigations, like retrofitting. The only thing I have to offer here based in our Oregon and Washington State research on a Magnitude 9 earthquake is that a number of interviewees did not have specific response scenarios for their own departments or units.
This lack of granularity is understandable (i.e., the more specific the scenario the more likely it is wrong about actual events unfold), but it makes some M9 discussions, in the words of one state infrastructure coordinator, “theoretical”. To avoid that, increased granularity in what-if scenarios seems necessary in managing for vulnerabilities that are not hidden out of view. Think again of table-tops, but this time around multiple what-if scenarios and interconnections.
III
But what to do beforehand for those cases where latent vulnerabilities are altogether unknown until disaster makes them manifest?
One answer follows from the guarantee that, when it comes to major disaster, prior latencies are joined at the hip with subsequent improvisations. Managing ahead means the latter are to be more doable and effective. I think immediately of cross-desk or cross-position training, e.g., control room dispatchers have also trained on the scheduler’s desk or water department staff can clear a major road even if the roads department staff have priorities elsewhere.
But it must also be recognized that some improvisations would not happen, cross-training or not, without the disaster. One state coordinator involved in communications management during emergencies told us about convening an online group of competing companies and infrastructure providers:
During a winter storm we had a utility or provider say we’ve got fiber cuts in this area, we don’t have the fiber to replace it in that area, our resources are in this other area—that allowed us to look at the group and say now is the time for some teamwork: Can anyone else solve that problem and be a good team member? And we’ve seen a lot of that sort of problem-solving manifest among the agencies with very little input from us. Another example might be a cellular carrier who is a competitor of another carrier going “Hey, we’re going to fill our generator, can we top off your fuel tank while we’re up there?. . .But I don’t think [those kinds of cooperation] would occur if we didn’t coordinate it and get everybody on the same call and provide a platform for them to kind of air those sorts of things.
Disaster shifts the interconnectivity configurations of staff and infrastructures not only in ways that open up opportunities to improvise but also in ways that make any such missed opportunities mitigable errors to be avoided.
IV
So what, practically?
Many interviewees reiterated they have no idea who or how many of their staff will be able to resume work immediately after the M9 earthquake. “The first 72 hours and you’re still trying to figure out who’s alive out there and those who can communicate,” said a state emergency manager. In other words, referring to “the M9 event” is misleading if it’s taken to imply one event and not thousands or more of them unfolding unpredictably.
One major implication is that it’s better to assume infrastructure cascades are part of the unfolding nature of the M9 earthquake, where just-in-time joint improvisations play an important role in addressing those cascades. Far too often “inter-infrastructure cascades” are assumed–and not just by modelers–to be instantaneous and unmanageable when in fact they are delayed and open to human intervention.
For managing latencies ahead, please also see https://mess-and-reliability.blog/2024/06/03/managing-ahead-for-latent-risks-and-latent-interconnectivity-2/
WHEN COMPLEX IS AS SIMPLE AS IT GETS 