Coast live oak (Quercus agrifolia) in the utility right-of-way
Coast live oak fails low. Of 619 California failure reports, 40 percent are root failures and 37 percent trunk failures, with decay in 80 percent of trunk failures and 39 percent of all failures under 5 mph. When one took a PG&E 12 kV circuit down near Napa, the CPUC found two GO 95 violations.
Clearance on a coast live oak is set by the voltage above it, exactly as it would be for any other species. No California rule names this one. What Quercus agrifolia adds to the file is not in the crown. It is in the bottom six feet of the trunk and in the ground around it.
The California failure record for this species now runs to 619 reports. Forty percent of them are root failures and 37 percent are trunk failures. When a coast live oak reaches a conductor, it is usually not by dropping a limb into the span. It is by coming down.
The record is 619 reports, and it has held its shape
The California Tree Failure Report Program collects standardized report forms from tree workers, arborists and consultants after a structural failure. Coast live oak has been published out of that database twice: once as a peer-reviewed failure profile in 1999, and once as a running summary the University of California keeps posted. The summary states plainly what the species is in the file: "Coast live oak is the most commonly reported species in the CTFRP database."
1995 | 1996 | 2010 | 2014 | |
All species reports | 1,467 | 2,023 | 4,886 | 5,589 |
Q. agrifolia reports | 122 | 161 | 526 | 619 |
Share of total | 9% | 8% | 11% | 11% |
Failed at the trunk | 29% | 33% | 37% | 37% |
Failed at a branch | 26% | 28% | 24% | 23% |
Failed at the root | 45% | 39% | 39% | 40% |
The sample grew five-fold across 19 years and the proportions barely moved. Root and trunk together carry roughly three quarters of the reports in every column. That stability is what makes the profile usable on a circuit: it is not a reading of one bad winter.
The 2014 column also gives the tree. Mean diameter at breast height of 32 inches, mean height of 48 feet, mean age of 89 years. Edberg and Berry's 1999 species description says the same thing from the other direction, an evergreen oak "growing up to 21 m (70 ft) in height, with a spreading, round-headed, often multi-stemmed crown." A 48-foot tree with a spreading crown reaches a distribution conductor, and it covers the span when it lies down.
Fifty-four percent of the failures were at ground level or below
The 1999 profile grouped the same data a second way, by height rather than by part. Trunk failures at ground level were pooled with root failures, and trunk failures above ground level were pooled with branch failures. For coast live oak the result was 54.0 percent at ground level and below, against 46.0 percent above ground.
That split was the paper's finding, not a rounding artifact. The authors report it as significantly different from other hardwoods and from valley oak, and they say what it means:
This places the major failure problem area for coast live oak at the soil level and below, suggesting that human site and soil modification can have a negative impact on the health of these remnant trees of native origin.
The operational problem is that the deciding part of the tree sits outside what a drive-by look is designed to catch. The Partrick Fire report quotes the standard: "GO 165 defines a patrol inspection as a 'simple visual inspection' meant to identify 'obvious' problems and hazards and may be carried out in the course of other company business."
A basal cavity behind a skirt of suckers on the far side of a 32-inch trunk is not obvious from a truck, and the distinction between the two inspection types is where the question of who was supposed to see it gets settled.
Edberg and Berry's management section names the inspection the species actually needs. It calls for "monitoring decay throughout the tree, especially in below-ground portions." That is a root collar examination, performed on foot, with the duff pulled back.
Decay is present in four trunk failures out of five
Decay is not an aggravating factor on this species. It is the normal condition of a coast live oak failure.
Decay present at | 1995 | 1996 | 2010 | 2014 |
Trunk failures | 97% | 87% | 79% | 80% |
Branch failures | 73% | 69% | 70% | 74% |
Root failures | 78% | 73% | 71% | 73% |
The 1999 paper put decay in 83 percent of coast live oak failures at all locations. The comparison groups in the same database ran well below that: 55 percent for other hardwoods, 39 percent for eucalyptus, 24 percent for conifers. The authors draw the conclusion that matters to anyone writing a prescription:
Our finding that these trees frequently fail at lower wind speeds, together with decay and structural defect data described above, suggests that a substantial decline in structural strength had already occurred in coast live oaks before the trees failed.
A record that notes decay only once it is advanced will flag almost every coast live oak after the failure and almost none before it. The reportable thing on this species is the wound, the cavity, the missing bark at the base, recorded with a measurement, on the date somebody stood next to the tree.
October 8, 2017, 2348 hours: 1721 Partrick Road
On the night of October 8, 2017, a 50-foot coast live oak fell into PG&E's Pueblo 2103 12 kV circuit at the Fontanella Family Winery, 1721 Partrick Road, in the city of Napa. One of the 12 kV conductors fell to the ground.
The Partrick Fire burned 8,283 acres and affected 283 parcels before merging into the Nuns Fire, which burned 56,556 acres, destroyed 1,355 buildings and damaged 172 more, with three fatalities. The CPUC's Safety and Enforcement Division issued Incident Investigation Report E20171020-02 on May 10, 2019. The Violation line on its cover page reads "Yes."
The measurements are worth reading together, because each of them is unremarkable on its own.
Tree: coast live oak, 50 feet tall, approximately 17 inches diameter at breast height, native to the area as part of an oak woodland.
Failure: at the trunk.
Conductors: size #4AR, installed 1981, span approximately 145 feet.
Weather at station EW6860, 4.5 miles southeast: peak wind 9.0 mph, gust 22.0 mph, about 71 degrees Fahrenheit, 14 percent relative humidity.
Seventeen inches is below the mean failure diameter for the species. Fifty feet is about the mean height. Nine miles per hour is a still night. The tree was inside the population the failure profile describes, at the ordinary size, on an ordinary evening.
CAL FIRE contracted an ISA Certified Arborist, Mark Porter, to evaluate the tree. His description of the trunk is the whole case:
A margin of callus (woundwood) surrounds a pre-existing trunk injury exposing the wood without protective bark…The wound is at the ground level extending to approximately 6-foot-high on the trunk, 11 inches wide, and at 3-foot decay is more advanced spreading into the north side of the trunk. On the north side of the trunk there are two open cavities.
Eleven inches wide on a seventeen-inch stem, running six feet up from the ground, with two open cavities. Porter's conclusion was that "the pre-existing conditions of the subject oak was critical to this failure."
The violation turned on a defect nobody had written down
SED found two violations. The first was of GO 95 Rule 31.1, which is the general maintenance rule:
Electrical supply and communication systems shall be designed, constructed, and maintained for their intended use, regard being given to the conditions under which they are to be operated, to enable the furnishing of safe, proper, and adequate service.
The report's violation table captions it "Hazardous tree not identified and abated." What that rule has cost elsewhere, and how a citation under it is dated, is worked through on a different species. The second violation was of GO 95 Rule 35, captioned "Vegetation clearance not maintained," for failing to maintain the clearances required for the hazardous tree that fell into the conductors.
That second finding is the one to hold up against the other October 2017 oak. In the La Porte Fire, a valley oak limb failed with internal decay that had no outward sign, and SED found no Rule 35 violation because PG&E had no actual knowledge of the defect. At Partrick the defect was eleven inches wide and started at the ground. The rule text did not change between the two files. The visibility of the decay did.
SED's conclusion says what should have happened:
PG&E failed to identify a hazardous tree condition despite the tree having visible defects, decay, and rot. PG&E did not take the appropriate steps to prevent the subject tree from falling into the overhead conductors. PG&E did not document the subject tree for trim or removal.
Two facts in the report make that finding hard to argue with. PG&E's own Vegetation Management Distribution Patrol Standard already carried the trigger: "Trees that are dead, show signs of disease, decay or ground or root disturbance that may fall into or otherwise impact the primary conductor shall be removed or made facility safe." And PG&E's Vegetation Management Hazard Tree Rating and Scoring Procedure, VEG-1015P, dated October 13, 2014, "indicates a 'High' failure potential for the Coast Live Oak species."
The criterion existed, the species rating existed, and the defect was visible. SED wrote the gap in one sentence: "As the conditions (decay and wounds) were noted by Mr. Porter to be pre-existing, qualified tree contractors should have identified the Coast Live Oak for removal prior to the incident occurring."
The number that should make a program uncomfortable is the inspection history. SED reviewed PG&E's vegetation management records for the incident area from 2013 to 2017 and found that PG&E performed vegetation management activities in 2013, 2014, 2015, 2016 and 2017. Five consecutive years of work in that area, plus a detailed inspection in 2017 that generated 27 work orders with seven of them in proximity to the incident location, and none of those records carried the oak.
Nine miles per hour is a typical failure wind
The Partrick weather reading sits in the middle of the record rather than at its edge.
Wind at failure, 2014 summary | Share of reports |
Under 5 mph | 39% |
5 to 25 mph | 31% |
Above 25 mph | 29% |
The 1995 profile read the low band higher still, at 49 percent of coast live oak failures at all locations, and its authors noted that only valley oak came close among the groups compared.
The 2014 summary splits precipitation almost evenly as well: 47 percent of failures with some, 53 percent with none. Root failures are the exception inside that, and the 1999 paper puts 73 percent of coast live oak root failures in association with precipitation against 18 percent of trunk failures and 10 percent of branch failures.
Two things follow for scheduling. A post-event patrol after a wind or rain event will catch the root failures and miss most of the trunk failures, because the trunk failures are not waiting for weather. And a basal defect found in July is not less urgent than the same defect found in January, because roughly two failures in five happen in conditions no forecast would flag.
The grade change is what the corridor contributes
Of the six tree groups Edberg and Berry compared, coast live oak had the highest association between a site grade change and failure, at 18 percent fill and 12 percent cut.
Tree group | Trunk | Branch | Root | All locations |
Q. agrifolia | 7.6 | 7.6 | 14.4 | 29.6 |
Q. lobata | 5.8 | 8.7 | 11.6 | 26.1 |
Other oaks | 5.5 | 5.5 | 9.6 | 20.6 |
Conifers | 1.1 | 2.5 | 2.3 | 5.9 |
Eucalyptus | 1.8 | 1.0 | 1.8 | 4.6 |
Hardwoods | 1.6 | 1.0 | 1.6 | 4.2 |
The gap between 29.6 percent and the 4.2 percent recorded for other hardwoods is seven-fold, in a single field on the same report form. Saturated soil runs the same way: 30 percent of coast live oak failures overall, with the species showing a significantly higher association of saturated soil with root failure than every other group except hardwoods as a category.
A pole hole, an anchor trench, a pad, a turnaround for a bucket truck and a graded access road are all grade changes. The general shape of a corridor-caused root failure is worked through on valley oak, and the mechanism does not change here. What changes is the weighting. On coast live oak, grade work is the environmental factor most strongly associated with failure in the published comparison, and PG&E's own patrol standard already names "ground or root disturbance" as a condition that triggers removal or facility-safe work.
The practical consequence is a date. Construction cuts the grade in a dry September and the tree reads normally for several seasons. The excavation date, recorded against the tree rather than against the job, is the only thing that connects the two events when the tree finally goes over, and that is a system of record question before it is an arboriculture one.
The oak was there before the conductor
Two thirds of the coast live oaks in the 2014 summary were in natural stands. In 1995 the figure was 80 percent, against roughly 90 percent planted for other hardwoods, eucalyptus and conifers. The Partrick tree fits: Porter recorded it as "native to the area as part of an oak woodland."
This is the structural difference between coast live oak and a planted corridor species. A eucalyptus or a planted street tree grew into the line after the line was built, so the growth rate and the trim cycle govern it. A remnant oak was standing when the poles were set. Its root plate predates the access road, the pole hole and the conductor.
The tree was never on a growth trajectory that brings it into the clearance radius at all. Half of the 2014 reports came from residential sites, 14 percent from parks and 8 percent from street tree sites, so the population is mostly trees on somebody's property with a line running past them.
The consequence for a pre-inspection is that the clearance question and the hazard question come apart completely. A mature coast live oak can sit at full PRC 4293 clearance, show no growth toward the conductor in a ten-year record, and still be the tree that takes the span out. The only assessment that reaches it happens at the base.
Berkeley's ordinance names the species
The prescription this failure profile points to is removal, and removal is where local law shows up. At least one California city has written an ordinance around this species by name.
Title 6.52 of the Berkeley Municipal Code restricts the removal or excessive pruning of coast live oak (Quercus agrifolia) in Berkeley. Excessive pruning is defined as removal of more than 25% of the functioning leaf, stem, or root system of a tree in any 24 month period. This applies to single stem trees with a circumference of 18" (5.7" diameter) or multi-stem trees with a combined circumference of 26 inches (8.3" combined diameter) measured at 4 feet from grade.
The city's own exception form goes further, recording that Ordinance No. 6,905-N.S. "places a moratorium on the removal of the coast live oak." The threshold catches almost every tree a line clearance crew would be looking at, because the 18-inch circumference is the 5.7-inch diameter the city prints beside it. Twenty-five percent of the functioning leaf, stem or root system in 24 months is a limit a full clearance prune on a low, wide oak can pass without anyone intending to.
Berkeley is one city, and ordinances elsewhere are written differently. What the reader can act on is the shape of the question rather than the answer: whether the parcel sits in a jurisdiction with a species-specific ordinance, what its threshold is, and who signs the exception. Berkeley's form answers the last part for Berkeley. Exceptions "must be approved by the City Manager or their designee." That is the name that has to be in the file before the saw starts, alongside the refusal and notification record for the property owner.
What a coast live oak record has to carry
A general vegetation entry on this species records the wrong end of the tree. Four fields make a coast live oak record defensible, and each of them comes from something above.
A basal observation, made on foot, with the duff pulled back and the root collar examined, dated and signed.
A measurement of any wound, cavity or area of missing bark, in inches, with its height on the trunk and its aspect.
The date and nature of any grade work, excavation or fill within the root zone, recorded against the tree.
The species at species level, so the hazard tree rating the utility already publishes attaches to the row.
The clearest way to see the shape is to write the entry the Partrick file did not contain. Pole and circuit identified, 12 kV, span approximately 145 feet. Quercus agrifolia, approximately 17 inches diameter at breast height, 50 feet tall, standing in a native oak woodland beside the span.
Root collar examined on foot by a named Consulting Utility Forester, with the date of the visit. Open wound at ground level, 11 inches wide, running to approximately 6 feet up the trunk on the north aspect, with two open cavities. Removal prescribed on the basis of the basal defect. Owner notified, and any local tree permit requested, with both dates recorded.
Nothing in that entry is a clearance measurement, and that is the point. The only moment a clearance number described the Partrick oak at all was after it was lying across the conductors, which is where SED's Rule 35 finding was written.
Where this leaves the reader
The species adds nothing to the distance in Table 1 or in PRC 4293. It changes three other things:
where to look, which is the first six feet of trunk and the soil around it
when to look, which is not keyed to weather
what to write down, which is a measured defect rather than a measured gap
Two calls sit outside the rules entirely, and both still get made. On whether a decayed coast live oak is a hazard tree, PG&E's own procedure and rating already answer, and the Safety and Enforcement Division reviews that answer afterward. On whether the tree can come down, the answer is local, and in Berkeley it is the City Manager or a designee.
Further reading
Valley oak: limb failure and line-strike risk: The other California oak on the same circuits, and the opposite failure mode: a live limb, high in the tree, that produced no vegetation violation.
Gray pine and codominant stems near conductors: What GO 95 Rule 31.1 says and what it has cost, worked through a citation rather than an investigation report.
Field hazard tree assessment: The inspection this species needs performed at the root collar rather than from the road.
Frequently asked questions
Does any California rule set a different clearance for coast live oak?
No rule names the species. GO 95 Table 1 is written in voltages and distances, and PRC 4293 in voltages and radial feet. What attaches a species to either number is the utility's own written specification. PG&E's specification does name it: the CPUC's Partrick Fire report records that PG&E's Vegetation Management Hazard Tree Rating and Scoring Procedure, VEG-1015P, dated October 13, 2014, "indicates a 'High' failure potential for the Coast Live Oak species." On the tree that started the Partrick Fire, that rating produced no prescription.
Where does a coast live oak usually fail?
At the bottom. The California Tree Failure Report Program's 2014 summary of 619 coast live oak reports puts 40 percent at the root, 37 percent at the trunk and 23 percent at a branch. The 1999 published profile grouped the same data by height and found 54.0 percent of failures at ground level or below, against 46.0 percent above ground, a distribution the authors reported as significantly different from other hardwoods and from valley oak.
Is a coast live oak with no dead wood in the crown a Rule 35 finding?
It can be, and one was. The Partrick Fire tree was 17 inches in diameter at breast height with a wound running from ground level to about six feet and two open cavities on the north side. The Safety and Enforcement Division found PG&E in violation of GO 95 Rule 31.1 for a hazardous tree not identified and abated, and of GO 95 Rule 35 for vegetation clearance not maintained. The defect was at the base, not in the crown.
Does this species fail in wind?
Less than the other groups in the same database. The 2014 summary puts 39 percent of coast live oak failures in wind under 5 mph, 31 percent in 5 to 25 mph and 29 percent above 25 mph, and splits precipitation almost evenly at 47 percent with some and 53 percent with none. The weather station nearest the Partrick Fire recorded a peak wind of 9.0 mph and a gust of 22.0 mph. A patrol triggered by a storm will not find these trees.
Can a crew prune a coast live oak to clearance without a local permit?
That depends on the jurisdiction, and at least one names the species. Title 6.52 of the Berkeley Municipal Code restricts the removal or excessive pruning of coast live oak, with excessive pruning defined as removal of more than 25 percent of the functioning leaf, stem or root system in any 24 month period, on single stem trees of 18 inches circumference or multi-stem trees of 26 inches combined, measured at 4 feet from grade. Berkeley's exception form states that exceptions must be approved by the City Manager or a designee.
