Resilience: Adaptation to Climate Change or creating the next normal accident?

The UK Government under the Department for Environment, Food and Rural Affairs (Defra) has a series of programmes running to enable the UK to adapt to the challenge of climate change.  This includes engaging with the engineering communities to consider the issues that are raised and considering how these may be addressed.  One of the interesting challenges highlighted by the engineers is the increasing interconnectedness of the systems that we are building.

Consider the widely accepted need to reduce energy consumption through the introduction of the Smart Grid.  This introduces a dependency on the communications infrastructure, and both are dependent on the electrical infrastructure, with this in turn connecting to water management (water being key to power stations, flooding can lead to problems with underground cables).  Given the current concerns regarding cyber attacks, this leads to dependency on the effectiveness of a security management system too.  It is clear that these changes introduce new vulnerabilities to the critical national infrastructure.  Perrow has written about the closely coupled systems that lead to inevitable "accidents", the complexity becomes such that operators cannot understand the interactions and mistakes become inevitable and these are magnified through the interconnected systems.  Such events describe nuclear power accidents but have also been used to describe many other forms of "accident", including the roots of the current financial crisis (look up Tim Harford on this).

So I have a concern, indeed a growing concern, that the increasing desire to connect everything to the internet will lead to unexpected outcomes, failures in one system become magnified through their interaction with connected systems.  The best forms of security can sometimes be obscurity and isolation; a cyber attack on the power grid in the UK would have had no meaning when we still entirely relied on mechanical switches and engineers.  We may be designing a world in which a failure of security leads to a failure of some internet services which leads to the failure of power systems which causes more internet services to fail, and hence more power systems to fail and then the water pumps stop...and well I guess you get the idea.

More positively, the engineers are right in suggesting that there are strategic opportunities to improve resilience.  So, it may be very expensive to move a railway that is on the coast and may be inundated.  However gradual and planned change over many years can make this affordable and practical.  This is something that all organisations can consider.  Recognising that the limitations of a current facility may not be entirely addressed, for instance it is in a flood plain, one can plan strategically for resilience to gradually evolve and reduce the risk.  So when considering expansion plans, put the new plant or office in a different site outside the flood plain.  Over time gradually migrate to the new location and have this designed to optimise resilience from the outset, much cheaper than trying to retrofit an old facility.

Adaptation, interconnectedness, resilience and inter-disciplinary approaches will all feature greatly in our increasingly complex society.  The Defra work is but one example and I am sure we will be hearing more on this very soon.