This afternoon, a power failure at a communications centre brought major disruption to rail services across the Midlands and north-west England. Avanti West Coast, CrossCountry, East Midlands Railway, London Northwestern, Northern, TransPennine Express and Transport for Wales were all affected. Not one operator having a bad day. Seven, at once, because of a single point that sits underneath all of them.
We don’t yet know the full detail of what failed or why, and that’s not really the point of this post. The point is what the incident illustrates about how disruption actually spreads, and it’s worth ten minutes of any risk or continuity manager’s time.
The failure wasn't in the trains
Every one of those operators will have their own business continuity plan. Fleet contingencies, crew rosters, station evacuation procedures, the usual. None of that helped this afternoon, because the fault didn’t sit inside any of their own operations. It sat in shared infrastructure that none of them individually own, control, or can fix.
This is the pattern behind a lot of the disruption that makes the news: not a failure of the organisation’s own plans, but a failure of something upstream that the plans never accounted for because it technically belongs to someone else.
Most dependency maps stop too early
Ask most organisations to map their critical dependencies and you’ll get a list of suppliers, IT systems and key staff. Useful, but it usually stops at the first layer. It rarely asks whether two apparently different suppliers are actually drawing on the same thing further up the chain. I’ve seen this with a manufacturing client who used two separate suppliers for a key raw material, on the reasonable assumption that this gave them genuine redundancy. It didn’t. Both suppliers were sourcing the same chemical from the same producer. The diversity was real on paper and non-existent in practice.
But today's case is a harder problem
That kind of exposure is at least fixable, because a choice was made somewhere along the line and a different choice could have been made instead. Today’s rail disruption isn’t that. The train operators didn’t choose to share a communications and power backbone; it’s simply a condition of running services on that network. There’s no alternative supplier to switch to and no diversification strategy that gets you out of it.
Which means the response has to be different too. Where you have genuine choice over your dependencies, the job is to check whether the diversity you think you have is real. Where you don’t, because the infrastructure is shared by design, the job is knowing exactly where that exposure sits and having a plan for what you do the moment it’s realised, since you can’t design your way out of it.
The question worth asking this week
Worth five minutes with your own resilience lead: split your critical dependencies into the two categories. Where you have a choice of supplier, are you actually getting the redundancy you think you’re paying for. Where you don’t, because you’re sat on shared national infrastructure like everyone else, does your plan assume that infrastructure will simply be there, or does it have a genuine answer for the afternoon it isn’t.
