Trim type correct: the audit against the prescription, cut by cut
Trim type correct is the audit that compares completed tree work against the prescription recorded before the crew arrived. It asks whether the prescribed work type was performed on the prescribed tree, and whether the result will hold for the cycle the prescription assumed. A clear span with the wrong trim type fails.
Two documents decide whether a trimmed tree passes. One is the clearance rule. The other is the prescription: the work a pre-inspector assigned to that tree, in the system of record, before the crew arrived. Trim type correct is the audit of the second document. It asks whether the crew performed the work that was prescribed, on the tree that was prescribed, and whether the result will hold for as long as the prescription assumed.
Writing the prescription is a different job, on a different visit, by a different person. This page starts the day after the crew leaves.
The prescription is what the work is measured against
PG&E publishes its distribution inspection procedure, TD-7102P-01, at Revision 2, published April 20, 2023 and effective June 20, 2023. Its section 4, "Prescribing Work," does not leave the inspector a free choice of trim type. It sets an order, and the inspector moves down it only when the rung above is unavailable:
(1) Prescribe removal. … (2) If removal is not practical, THEN prescribe pruning such that the tree will maintain compliance for three annual work cycles. (3) If pruning to maintain compliance for three annual work cycles is not an option, THEN prescribe pruning such that the tree will maintain compliance for one annual work cycle. (4) If pruning to maintain compliance for one annual work cycle is not an option, THEN prescribe Bi-Annual clearance.
Every rung carries two facts about one tree: what to do, and how long the result has to last. That second fact is what makes trim type auditable at all.
Take a valley oak on Circuit 12JY prescribed at rung 2. The prescription bought three annual work cycles of compliance from that tree. A crew that cut it back to the clearance the line needs today, and no further, has worked the tree and has not satisfied the prescription. The auditor's job is to say so, in the record, against the rung that was written.
An auditor who instead grades the tree against the prescription they would have written is reviewing the pre-inspection. That is a separate finding with a separate owner, and it belongs in the record as one.
What the mark on the trunk is worth as evidence
The same procedure tells the inspector to leave a physical signal on the tree, in section 5.1:
Spray the paint near the base of a tree using one of the following shapes: • A dot for pruning. • An X for removal
An X on a trunk that is still standing, with a fresh pruning cut above it, is the most legible trim type mismatch in the field. An auditor can read it from the shoulder of the road without opening a tablet.
It is corroboration, not the record. Four lines later the same section tells the inspector to overwrite the marking history:
Spray new marks over any marks from previous years, but with some of the older mark still showing. Cover incorrect marks with black or brown paint.
So paint carries the current intent, the previous year's intent, and any correction, all on one stem. Where the paint and the prescription in the system of record disagree, the record governs, and the disagreement is itself a finding worth writing down. Some trees carry no mark at all. Section 5.1 provides for that too:
IF the VMI cannot paint or flag a tree, THEN UPDATE the tree record with the code CNP (cannot paint) or CNF (cannot flag) and a description of the tree's location within the span.
A tree identified only by a sentence of prose has to be resolved to a specific trunk before anything about it can be graded. If the auditor cannot make that resolution with confidence, the honest entry is that the tree could not be located, not a pass.
Clearance today, and the cycle the prescription bought
PG&E's quality control standard for routine distribution, SW-0003 at Revision 03, published May 31, 2024 and listed among the documents on the utility's Community Wildfire Safety Program page, writes the acceptance test forward rather than backward. Section 3.1:
Verify all trees at location will maintain MDR until next routine cycle (13 months from VMI date)
Thirteen months, measured from the inspection date rather than the work date. A tape held to the conductor on the day the crew leaves cannot answer that question, which is why trim type and clearance are graded separately. The measurement itself belongs to clearance achieved.
What fills the gap is the method of the cut. The Utility Arborist Association states the endpoint rule plainly:
Trees that are pruned in accordance with ANSI A300 Standards will have branches pruned back to a lateral branch that is at least one third the diameter of the branch being removed.
The same source defines directional pruning as removing "branches growing toward the power lines while leaving those that are growing away." It adds that "cuts should be made close to the parent or main limb so as to not leave a stub or damage the main limb." A cut that ends at a dominant lateral heading away from the conductor is the cut the three-cycle rung was written for.
A cut that ends in a stub is not, whatever the tape reads that afternoon. The Tree Care Industry Association, which publishes the standard, lists topping, lion's tailing and rooster-tailing as unacceptable practices under A300 Part 1.
Whether a given cut conforms to A300 is a question of its own, and it is covered in ANSI standards met. The trim type audit uses the cut endpoints for a narrower purpose: they are the evidence of which prescribed work type was actually performed.
Two tags, one definition, and where the published criteria run out
Section 4.2 of SW-0003 lists the tag types a QC specialist selects from. Two of them print the same definition:
Potential Grow-in (PGI) • Vegetation is currently compliant with PG&E's Minimum Distance Requirements (Distribution voltage) but won't hold 13 months from the VMI inspection date. Non-compliant • Vegetation is currently compliant with PG&E's Minimum Distance Requirements (Distribution voltage) but won't hold 13 months from the VMI inspection date.
Two tags, one sentence, in the revision whose own revision note reads "Minor clarifications in Section 4.2." Earlier in the same section the distinction does appear, carried entirely by a parenthetical:
Tag Type: Non-Compliant (within MDR) & Potential Grow-in (At or approaching MDR)
Read together, a tree inside the minimum distance requirement today is non-compliant, and a tree at or approaching it is a potential grow-in. The bullet definitions lost that difference in print. Here is the document's own amendment history:
Rev | Date | Description, as printed |
0 | `MM/DD/YYYY` | Original Publication |
01 | February 26, 2024 | Formatting updated, new criteria for 2024 added. |
02 | May 10, 2024 | Added guidance to verify the PG&E facilities at the location reside in HFTD otherwise constrain. Added QC Timeframe of Assessment to select Post VMI or Post TC. Updated to clarify differences in reporting veg deficiency findings depending on Post VMI or Post TC. |
03 | May 31, 2024 | Minor clarifications in Section 4.2. |
The published standard does not settle which tag a borderline tree gets. The program has to, in its own work instruction, and then prove that two auditors apply it the same way. That proof is a calibration exercise, and it is covered in what a QA/QC audit actually covers. An audit that leaves the choice to the individual specialist will report a tag mix that moves with staffing rather than with the trees.
Cut by cut: what a removal and a prune each have to show
A prescription at rung 1 and a prescription at rung 2 leave different evidence on the ground, and the audit asks different questions of each.
A prescribed removal is nearly binary. The tree is gone or it is not. The case that hides between those two is the stump, and TD-7102P-01 routes it to its own attachment in section 4.1:
IF prescribing removal of a tree that may resprout, OR IF a stump is currently resprouting, THEN refer to Attachment 4, "Handling Stump Resprouts."
A felled tree with untreated resprouts at the collar is a removal that started and did not finish. The finding belongs against the original removal prescription, with the original tree identifier, because recording it as a new tree resets the clock on work somebody has already been paid for.
A prescribed prune is resolved cut by cut. Two crews can leave the same tree at the same final clearance and have performed different work types: one set of cuts ending at laterals growing away from the line, the other ending at stubs on branches still aimed at it. What separates them is the endpoint of each cut, so the endpoint is what the auditor photographs. A photograph of the whole canopy from thirty feet away proves clearance and proves nothing about trim type.
The verdict is recorded per tree, against the tree's own identifier, even when the sample is drawn by span. The unit definition and the evidence file that carries it are set out in what a QA/QC audit actually covers. Rolling four tree-level verdicts into one span-level pass is how a type mismatch disappears from a report.
Who owns the miss, the pre-inspector or the crew
SW-0003 asks the specialist to attribute the finding before anyone has time to argue about it. Section 4.1:
Which VM work group is likely responsible for the QC vegetation finding? Options: • VMI • Tree Crew • Inconclusive This initial assessment will aid VM execution with their root cause analysis.
Attribution is what makes a trim type finding act on something. A prescription written at the wrong rung and a cut made at the wrong endpoint produce the same unhappy tree and call for different corrective actions, one in pre-inspection training and one in crew supervision. Corrective action and its due date are covered in what a QA/QC audit actually covers.
The attribution only works because the specialist first declares when the visit happened. Section 2:
Select Post VMI when the Distribution location has NONE OR ANY prescription trees PENDING completion by the tree crews. Select Post Tree Crew (TC) when the Distribution location has ALL prescription trees worked by the tree crews.
That declaration changes what counts as a deficiency:
During Post VMI, deficiency findings are reported by QC when vegetation is identified that does NOT meet scope requirements AND is NOT recorded by the VMI as a pending work prescription in our QC Map. During Post TC, deficiency findings are reported by QC when vegetation is identified that does NOT meet scope requirements and is NOT mitigated.
Before the crew has finished, an unworked prescribed tree is pending work. After the crew has finished, the same tree is a finding. Visit timing decides which, so the timeframe field has to be filled from the work status in the system of record, not from the auditor's impression of how fresh the sawdust looks.
Keep "Inconclusive" available and keep using it. An audit that forces a binary attribution moves the number without changing the trees, and the trend it produces will point at whichever group is easier to blame. Those counts roll up into a pass rate: section 4 of SW-0003 has the findings collected "in our QC System of Record for each survey location to determine the QC Pass Rate."
When the prescribed work could not be performed
Some prescriptions are written and never executable, and the audit has to tell that apart from a miss. TD-7102P-01 defines both situations:
Refusal: A situation that occurs when a customer / property owner refuses to allow PG&E to perform pre-inspection work or complete 100% of the work prescribed.
Constraint: A situation that occurs when a customer, property owner, or agency obstructs or delays PG&E pre-inspection work or the completion of the intended tree work.
A tree prescribed for removal and pruned instead, on a property whose owner refused the removal, is a refusal with an attested record or it is a type mismatch. The difference is entirely documentary. What the attestation has to contain is set out in who signs off on a span, and the notice and contact thresholds that a California refusal has to clear are covered in notice and refusal limits.
The procedure names its escalation path for that case in section 4.1: the inspector follows the refusal procedure TD-7102P-04 "AND ESCALATE to the VPM." It names no equivalent path for a crew that decides in the field to perform a different trim type than the one prescribed. That gap is real and it is common.
The program has to close it in its own work instruction, by naming who can authorize a change of prescription and requiring that authorization before the cut rather than after it. An audit finding written weeks later is the wrong place to discover that a crew foreman made a silvicultural decision alone.
When a work order is ready to be audited for trim type
Auditing a work order that is half worked produces a type mismatch rate that is really a completion rate. PG&E's quality control business process document, MID Quality 0002 at Revision 02, published April 10, 2024, sets a numeric gate:
QC should only review these completed WO's when they also have 85% or more of tree work prescriptions completed and documented within the VM System of Record.
Two conditions, joined: completed and documented. A tree worked yesterday and not yet closed in the system is not available to be audited against its prescription, because the prescription record is where the expected work type lives.
SW-0003 gives the specialist the matching exit. Section 1 permits a skipped review for six stated reasons:
Safety Concern
Accessibility issues
All vegetation work at location not complete
Customer concerns or refusal
2nd patrol inspection occurring
Other
The third of those is the trim type audit's own stop condition. A specialist who walks a location where work is still in progress and writes findings anyway has produced numbers that will be argued away, and the argument will be correct. How the lot is defined and how the sample is drawn from it is covered in sampling versus a full census.
Where the trim type record is read by somebody else
The comparison this page describes is not an internal quality habit. It turns up in regulatory filings as a described practice. The Independent Evaluator report filed with California's Office of Energy Infrastructure Safety records PacifiCorp's approach in one line:
Forester signing off work as it is complete. PacifiCorp notes it reviews pre-inspections (prescriptions) at the time of conducting audits of completed work …
The same office's Independent Evaluator filing in the 2026 docket sets out its own purpose in similar terms. Its stated aim is to "verify WMP activities performed by PacifiCorp, control (QA/QC) programs". The prescription sits inside that scope, because the utility named it as part of how work is accepted.
So the file has a second reader who did not commission it and has no stake in its pass rate. That reader will want to see, for one tree on one span, the prescription as written, the work type recorded as performed, the evidence that settles the difference, and the name against the verdict. How that record ties into the inspection obligations it sits beside is covered in GO 165 inspection cycles and in what a QA/QC audit actually covers.
Further reading
What a QA/QC audit actually covers in utility vegetation work: The unit, the evidence file, the defect classes and the CAPA this finding feeds into.
ANSI standards met: A300 in the audit: Whether the individual cut conforms, once the trim type question is settled.
Who signs off on a span, and what that signature means: Who can accept the tree, and how a refusal or exception is attested instead.
Frequently asked questions
If the span measures clear, can a tree still fail on trim type?
Yes. PG&E's published QC standard SW-0003 tells the specialist to verify that trees at a location will maintain the minimum distance requirement until the next routine cycle, 13 months from the inspection date. Clearance on the day the crew leaves does not answer that. A tree cut to today's number by a method that will not hold the cycle is a finding.
Who is responsible when a prescribed removal was pruned instead?
The audit records an attribution before anyone argues about it. SW-0003 asks which work group is likely responsible for the finding and offers three options: the vegetation management inspector, the tree crew, or inconclusive. A pruned tree against a removal prescription usually lands on the crew, but the option to record inconclusive exists and should be used when the record does not settle it.
Does the paint on the trunk decide what trim type was prescribed?
No. PG&E's distribution inspection procedure assigns a dot for pruning and an X for removal, but the same section tells inspectors to spray new marks over old ones and to cover incorrect marks with black or brown paint. The prescription in the system of record governs. Where the paint and the record disagree, that disagreement is itself a finding.
When is a work order ready to be audited for trim type?
PG&E's quality control business process document sets a readiness gate: quality control should review completed work orders when they also have 85 percent or more of tree work prescriptions completed and documented in the system of record. Auditing earlier than that produces a number that reports progress rather than quality.
