When the Towers Are Already Gone

The infrastructure failure sequence during a fast-moving wildfire is well-documented and predictable. Cell towers near the fire zone lose power or are destroyed. Most towers carry between eight and twelve hours of battery backup once grid power fails — after that, any phone call or text message that depends on them fails too. In a rapidly expanding fire scenario, that window can close before many residents know they need it.

A mesh communication network built before fire season operates on a completely different set of dependencies. No cell tower. No internet backhaul. No power grid. Just radio, battery, and the repeaters you placed when conditions let you place them carefully.

What the coverage plan actually needs to cover

The most common mistake in mesh planning for wildfire scenarios is planning for neighborhoods instead of evacuation routes.

When a fire moves fast, people are moving — on roads, not in houses. The communication gaps that hurt most are not in backyards. They are on the mountain road between the subdivision and the highway, in the drainage where the canyon narrows, at the junction where the route decision matters. If someone loses contact with their group while moving along an evacuation corridor, they need to know whether to turn back or continue.

Before fire season, walk or drive your primary evacuation routes with a companion device and note where packets stop arriving. That boundary — where your map transitions from connected to silent — is where coverage planning starts. The question is not whether the map looks good in the neighborhood; it is whether a message from the back of the evacuation column can reach someone in the front.

Repeater placement for fire scenarios

Two constraints are specific to wildfire that do not apply to general mesh planning.

Place outside the likely burn perimeter. A repeater that burns in the first hour of a fire provides no coverage. For your area, identify the most probable fire corridors based on terrain, vegetation, and historical patterns — not the worst-case scenario, but the credible-scenario perimeter for a typical fire. Repeaters that fall inside that perimeter should be treated as conditional: available in ordinary conditions, not available when most needed.

For coverage on evacuation routes that pass through fire-risk terrain, the repeater has to be close enough to cover the route but high enough and far enough from the fuels to have a reasonable chance of surviving. A ridge point above the chaparral, a rocky summit above the timber line, or a metal structure (water tower, communications mast) in a cleared area — these are the sites that are still there when the route is in use.

Power for the duration. A repeater with eight hours of battery provides no more coverage than a cell tower during a multi-day fire event. Remote repeaters should have solar panels sized for cloudy conditions plus enough battery storage to carry through periods when panels produce little. The specific sizing depends on device power consumption and your climate; an ESP32-based repeater drawing 100–200 mA continuous is a different calculation than a nRF52-based repeater at much lower draw. What matters is that the power budget includes a realistic worst-case, not an optimistic best-case.

Fire scenarios amplify every single point of failure. A SPOF in a hurricane deployment is a reliability risk. A SPOF in a wildfire deployment, where the repeater itself may be in the affected zone, is a near-certain failure. For every segment of your evacuation coverage, ask: if this repeater fails, what coverage remains? If the answer is “none,” that is the planning priority before fire season.

Testing before you need it

A coverage map that was never tested against reality is a planning assumption, not a plan.

The verification process is the same as any mesh coverage check: take a companion device on the evacuation routes you plan to cover, note where you can and cannot reach the network, and cross-reference that against what Waev’s Live Map shows from enrolled observers. The areas where the map is silent are either dead zones you need to address with additional repeaters, or observer coverage gaps that mean your health data for that zone is incomplete. Either finding is worth knowing in June.

After placement, watch the Network Stats for your new repeaters for at least two weeks before fire season. A repeater whose SNR trend is degrading from its first week is already showing a problem — marginal power, sub-optimal antenna orientation, interference. Find it while there is still time to fix it. For detailed guidance on reading that health data, see When the Grid Goes Down.

For repeater placement methodology — how to identify coverage gaps in existing mesh data and pick a candidate site — see Where Your Network Ends.

The plan document

A fire-season mesh communication plan should answer, in writing:

• Which nodes cover which evacuation routes, verified by field test
• Power specs for every remote repeater, including worst-case scenario
• Who monitors which channels and what protocols they follow
• What the fallback is if a given repeater fails
• When the plan was last tested and what was found

The document is not the capability. The tested, operational mesh is the capability. The document tells you what you verified, when, and what you need to fix before you can rely on it.

---

The mesh either exists when the towers go dark or it doesn’t. Building it on a quiet day in late spring is the only prep window that exists.