Resilience is not one perfect network. It is layers.

We already know what a single point of failure looks like.

Okanogan County has repeatedly lost essential communications during major fires. This isn't a hypothetical about which technology is safest — it's the county's own record.

The lesson is not that fiber failed and satellite never would, or that towers are safe and cables are weak. Any communications path can fail. The dangerous condition is different:

The danger isn't that a route fails. It's having no independent path left when it does.

And the way we warn people now makes this sharper. A Wireless Emergency Alert — the evacuation push to your phone — only arrives if you're in range of a live cell tower. A tower that's burned, out of power, or in one of the county's dead zones sends nothing. That is exactly why more than one independent path matters.

Fiber: the durable foundation, wherever it can be built affordably.

Fiber is the strongest long-term asset where the construction pencils out. The buried glass can stay useful for decades while the electronics at each end are refreshed — the backbone the whole system grows on.

It belongs in towns, denser corridors, tower backhaul, public facilities, businesses, and every rural pocket where grants or density make it practical. It should keep expanding. But the District's own projects show that reaching the most dispersed homes can cost tens of thousands of dollars per location — which makes universal fiber a long-term direction, not an immediate replacement for wireless. Fiber vs. wireless, in full →

The value isn't that one never fails. It's that they don't fail the same way.

Resilience comes from uncorrelated failure. A second connection is most valuable when it doesn't depend on the same equipment, route, power system, company, or support process as the first.

● stays up   ● partial or conditional   ● that path fails. "May" and "can" are deliberate: outcomes depend on which route, site, or home is hit. No one technology is guaranteed to survive every major fire — which is the whole argument for keeping more than one.

A second path is worth most when it fails for entirely different reasons than the first.

Resilience is physical — and economic.

A communications system can fail its people even while every cable and radio still works: through unaffordable price increases, neglected maintenance, poor service, missing parts, discontinued gear, delayed upgrades — or a company deciding a sparse area isn't worth serving. Competition and local control protect against those failures too.

That's why local control belongs in a resilience argument. Starlink gives rural customers leverage when a local provider performs poorly. Fiber gives businesses and denser communities a high-capacity alternative. And the PUD wireless network gives the rural majority a locally controlled option whose costs, inventory, upgrades, and policies can be questioned in public — and hardened, restored, and answered for by people who live here. No layer should be trusted without alternatives; each behaves better because the others exist. Rural self-reliance →

Competition is another form of backup — it gives people somewhere to go before high prices or neglected infrastructure becomes a crisis.

Every layer needs power. Shared infrastructure lets us plan for it.

Wildfires and storms take the grid down with them, so what keeps you online runs on a battery or a generator. All three technologies need electricity somewhere — the difference is where the backup can be planned.

Fiber and wireless customers need power for a router and the ONT or radio. A Starlink dish draws roughly 50–75 watts continuously — and in winter it heats itself to melt off snow, which can roughly double that draw. Tower sites, fiber nodes, and satellite ground infrastructure all need power too.

The advantage of locally managed infrastructure is that backup power at shared towers and network sites can be budgeted, tested, fueled, and prioritized for the community — one hardened site keeps whole neighborhoods connected. Starlink adds an independent path, but each participating household must provide its own backup power. That's another reason to keep both: a backed-up tower supports many homes efficiently; a battery-backed dish keeps one home or emergency site online when that tower fails.

Community backup and household backup solve different problems. A resilient county needs both.

Sources: independent watt-meter measurements of Starlink Standard dishes (~50–75 W typical; snow-melt heating adds ~50–100 W) — seabits.com and multi-owner metered reports (2024–2025); fixed-wireless tower backup-power practice — WISP operator documentation. Figures are typical ranges, not guarantees.