Grow-in vs fall-in: how to classify vegetation risk on a distribution circuit
Two trees can sit the same distance from a conductor and carry completely different risk. One is a fast-growing species inside the right-of-way, headed for the wire on a predictable schedule. The other is a hundred-foot oak on the neighbor's property, outside your clearance zone entirely, that only matters if it comes down in a windstorm. Lump those together on a worklist and you'll trim the wrong one first.
Most cyclical trim programs were built around grow-in because it's the easier problem: measure clearance, cut back to spec, move to the next span on the schedule. Fall-in runs on weather, root health, and species that fail without much warning, with no schedule behind it at all. Treating both as the same risk category is how crews end up re-trimming spans that were fine while a danger tree three spans over keeps aging toward failure.
Grow-in: predictable, schedule-driven, inside the ROW
Grow-in is vegetation encroaching on conductors from below or beside, inside or near the edge of your right-of-way. Think fast-growing brush, volunteer trees, or species like poplar and willow that can add several feet of height in a season. It's the risk your cyclical trim schedule was designed to catch, and in most cases it does, as long as the cycle length matches the actual growth rate on that span.
The classification question for grow-in isn't "is there vegetation near the line." It's "how many growing seasons until this canopy reaches minimum clearance." A span with slow-growing hardwoods ten feet below the conductor might not need attention for years. A span with fast-growing brush at the same clearance could close the gap in one season. Blanket cyclical trimming misses this distinction because it treats every span on the circuit the same, regardless of species mix or current canopy height relative to the wire.
Fall-in: the tree that's already tall enough, outside the clearance zone
Fall-in risk is a different kind of problem than grow-in entirely. It's a tree, often well outside the right-of-way, that's already tall enough to reach the conductor if it falls or fails. No amount of routine trimming addresses it. The tree already stands next to the line at full height, and the risk is structural: a weak root system, storm damage, disease, lean, or simple height relative to span distance.
This is the category that causes outage events disproportionate to how often it shows up on a worklist. A single compromised tree outside the ROW rarely gets flagged by a crew driving a cyclical route, because it's not encroaching in the way crews are trained to look for. It's just tall and close enough to matter if conditions turn.
Classifying fall-in risk means answering a different question for every tree near the corridor: given its height and its distance from the conductor, would a fall reach the line? It's a geometry problem, checked against actual tree height and span-by-span conductor position rather than a generic buffer distance applied across the whole circuit.
Why one risk score doesn't work for both
A circuit-wide score that averages grow-in exposure and fall-in exposure into a single number tells a trim crew almost nothing useful. A span can score "moderate" overall while hiding a severe fall-in exposure from one danger tree and negligible grow-in risk everywhere else. The crew shows up, trims the brush that was already fine, and the actual threat stays standing.
The fix treats grow-in and fall-in as separate measurements, combined afterward into one ranked worklist. Canopy height segmented against conductor position gives you grow-in timing. Proximity and height of vegetation outside the clearance zone gives you fall-in exposure. Scoring both, span by span, across the whole circuit is what turns an annual imagery pass into a worklist a GF can hand to a crew instead of a map they have to interpret themselves.
If your circuit is still running one trim cycle for both risk types, it's worth seeing what a span-level score actually looks like for your own span IDs.