The Grid Needs the Network. The Network Needs the Grid.

Something breaks in the grid. Lights go out across a cluster of substations, and the first thing anyone wants is data: which breakers tripped, which lines are live, where to send a crew. That data travels over a network. The network runs on power. So the tool that tells you why the power is out needs power to run.

Grid operators have worked this way for decades. Substations report back through SCADA systems and remote terminals, over fiber or microwave links, so a control room can see the territory and flip the right switches in order. Take away the network and the grid does not just get less convenient to run. It turns into a pile of disconnected equipment, invisible as a system.

Telecom has the mirror problem. Every cell tower and exchange runs on grid power, and so does whatever keeps a long fiber run alive in between. When utility power drops, the network holds on: a battery cabinet at most towers, good for a few hours, and a diesel generator at bigger sites, good for longer if the tank is full and a truck can reach it. That is the entire plan for outliving an outage, local and finite, betting it stays short.

The loop closes there. The grid's control systems need the network to coordinate restoration. The network needs the grid to outlast its own batteries. In a short outage this never comes up. In a long one, two systems fail in slow motion, in each other's direction.

Two resilience plans, each betting the other holds

A utility's resilience plan runs to redundant transmission paths, contingency planning, black start procedures, mutual aid agreements. A telecom operator's runs to backup batteries, portable generators, cell-on-wheels units, priority fuel contracts. Both are careful, both tested. Neither treats the other sector as a variable. The utility assumes its comms backbone will be there when needed, because it usually is. The telecom operator sizes batteries assuming power returns inside a window they comfortably cover, because it usually does. Each plan suits the ordinary bad day and stays optimistic about the extraordinary one.

When a whole region has a bad week

A single substation fire or one cut fiber line tests none of this; the surrounding grid and network absorb it without drama. The real test comes with a large regional event, an ice storm or a heat dome sitting over an area for days instead of hours. Every substation is stressed at once, and every cell site drains its battery on the same clock. Crews reaching the sites that need diesel are stuck on roads that are blocked or iced over themselves.

Past the first day, the failure turns worse than either sector's own worst case. The network fails site by site, in whatever order each battery drains, so the coverage map crews rely on for coordination keeps shrinking under them, because the control room's own visibility depends on that same shrinking network. Crews lose contact with dispatch exactly where they need direction most, while dispatch loses sight of exactly the substations that need attention most. Blind spots and outage grow together, each making the other harder to fix.

The gap nobody is paid to close

Ask who owns the scenario where grid and network fail together, and you get a shrug dressed up as an answer. Utility regulators regulate utilities. Telecom regulators regulate telecom. Each sector's continuity plan is audited by its own regulator, against its own worst case, so the assumption that the other sector will be there survives every audit unchallenged. Cross-sector mapping exists on paper in most infrastructure plans, but gets far less drilling than either sector spends on itself.

None of this makes either sector negligent; both are individually good at their jobs, which is the problem. Each has built such a competent plan for its own domain that it is easy to mistake “our system is resilient” for “the system is resilient.” A regional event does not care about the distinction. It finds the seam between the two plans and sits there.

Maybe the fix is smaller than a grand cross-sector initiative: a battery at every substation control point sized for the outage you are actually likely to get, and a comms fallback for grid operators that skips the towers the grid itself powers. Mostly it needs someone admitting, on a slide nobody wants to present, that their resilience plan has a dependency they do not control.

Until then, both systems are betting the other will politely stay up long enough to not become their problem. Call it a pact of mutual denial. Sooner or later, someone will have to explain to a control room lit by a phone flashlight why the backup plan for talking to each other was each other.