← All case studies Case Study · Electric Utilities · Transmission Inspection

Seventeen years without a climb: the Camp Fire transmission tower

Butte County, California · November 8, 2018 · 85 fatalities · CPUC Safety and Enforcement Division investigation
This case study is based on public reporting of the California Public Utilities Commission Safety and Enforcement Division's investigation report on the Camp Fire. Pacific Gas and Electric Company is not a SureSight customer, and SureSight was not involved in the incident or its investigation.

01What happened

Early on November 8, 2018, a worn C-shaped hook holding up part of a high-voltage line on a PG&E transmission tower in Butte County broke. The energized equipment it had been supporting came loose, and the resulting fault ignited dry vegetation below. Driven by strong winds, the Camp Fire burned more than 153,000 acres, destroyed most of the town of Paradise, and killed 85 people. It remains the deadliest and most destructive wildfire in California's history.

The tower stood on the Caribou-Palermo line, which was nearly 100 years old. PG&E accepted that its equipment started the fire.

02What investigators found

The state utility commission's Safety and Enforcement Division cited PG&E for a dozen safety violations. The findings that matter most here are about inspection:

The division's summary was that these gaps were “not isolated, but rather indicative of an overall pattern of inadequate inspection.”

WEAR YOU CAN ONLY SEE FROM THE TOWER WORN C-HOOK NOT DETECTED FROM THE AIR 2001 LAST CLIMB (AT LATEST) NO CLIMBING INSPECTION NOV. 8, 2018 HOOK FAILS Seventeen years of records said "inspected." None showed the hook.
Fig. 1 — Illustrative diagram based on findings of the CPUC Safety and Enforcement Division as publicly reported. Not to scale.

03The inspection gap

The tower was not ignored. It was patrolled, flown over, and logged, year after year. What the records could not show was what each of those inspections had actually looked at. A flyover and a hands-on climb both ended up as an entry saying the tower had been inspected, and nothing in the system distinguished the tower whose load-bearing hardware had been examined up close from the one whose hardware no one had seen in a generation.

Three separate failures hid behind that one word: an inspection method that could not see the wear, a known warning sign that did not trigger the closer look the company's policy required, and a form that did not even ask about the part that broke.

How SureSight closes this gap

SureSight cannot inspect a tower nobody visits. What it does is make the content of every inspection — and the absence of one — plain to the people responsible for the asset.

Inspection records that show what was seen. A SureSight inspection is first-person imagery tied to the specific asset and to each item on the checklist. For a tower, that means the record either contains verified views of the hooks, hanger plates, and insulator attachments, or it shows those items as not covered. “Inspected” stops being a single checkbox and becomes a list of components with evidence behind each one. A tower with no close-up hardware imagery since 2001 is a visible, sortable fact, not something reconstructed by investigators afterward.

Flagged wear that stays open. When an inspector's view shows wear, the finding is bound to the image and to the asset, and it stays open until a follow-up closes it. Visible wear on a tower's arms would stand as an open item calling for the climbing inspection the utility's own policy required — carried forward on every later patrol instead of fading into a prior year's file.

One current checklist, enforced at the workface. The inspection steps live in the system, not on whatever form is in the truck. When the procedure is revised to add a hardware assessment, every inspector's next job includes it, and a hardware item with no supporting imagery is flagged as unverified the same day.

Trend evidence for slow failures. A hook wears through over decades. Comparable imagery of the same component across inspection cycles gives engineers what a pass/fail history cannot: the rate of change, and time to act on it.

Sources & further reading