Fire pattern analysis is the backbone of origin determination, and flashover is the event that most reliably undermines it. Once a compartment transitions to full room involvement, every exposed surface burns, ventilation drives damage far more than proximity to origin does, and the clean directional indicators that pattern analysis depends on are progressively overwritten. Many of the origin opinions that fail under examination fail here.
What flashover does to the evidence
Before flashover, burning is localised and the damage record is roughly a map of where the fire was and how it moved. After flashover the whole compartment is a fire, and continued burning reflects where fuel and air were available rather than where ignition occurred.
The practical result is that intensity is no longer a proxy for proximity to origin. A deep char at a doorway may record nothing more than that the doorway supplied air. An area of severe damage at the far end of a room may record a heavy fuel package, not a starting point.
Ventilation-generated patterns
The most common misreading in post-flashover compartments is treating a ventilation-generated pattern as an origin indicator. Openings — doors, windows that failed early, HVAC penetrations — introduce air, and burning concentrates where air arrives. These patterns can be sharp, well defined and entirely unrelated to origin.
Distinguishing them requires reconstructing the ventilation history: which openings existed, which failed and roughly when, whether suppression created openings, and how the compartment was ventilated during firefighting. That history is assembled from the structure itself, from suppression records and from imagery, not from the char alone.
What still survives
Several categories of evidence are comparatively robust to full room involvement. Damage below the level of the burning — penetrations through floors, damage to surfaces that were protected by furnishings, and the boundary between burned and protected regions under collapsed materials — records conditions early in the fire.
So does the vertical dimension. Depth-of-char and calcination surveys across a compartment produce a gradient that is more resistant to late-stage overwriting than surface appearance is, particularly where measurements are systematic rather than selective. Arc mapping on the electrical system, where circuits were energised, provides an independent line that does not depend on thermal pattern interpretation at all.
Excavation is where origin work usually succeeds or fails
In a burned-out compartment, the informative material is under the debris. Layer-by-layer excavation — documenting each layer before removing it, recording the position of contents and reconstructing where furnishings stood — recovers the pre-fire arrangement, and the pre-fire arrangement is what makes a pattern interpretable.
This is slow and is frequently curtailed by cost, weather or demolition schedules. An excavation that stops at the first plausible finding tends to produce an origin opinion that cannot be defended, because the alternative areas were never examined to the same standard.
Hypothesis testing rather than pattern reading
The methodology set out in NFPA 921 is explicit that origin and cause are reached by developing and testing hypotheses against all the data, not by reading patterns and reporting the answer. In a post-flashover compartment this matters more, not less, because the pattern evidence alone under-determines the answer.
Testing means asking, for each candidate origin area, whether the fire could have developed as observed from that point given the fuel arrangement, the ventilation history and the timeline. Candidate areas that cannot produce the observed development are excluded on a physical basis rather than by preference.
Where modelling helps and where it does not
Fire dynamics modelling can test whether a proposed origin and fuel package are consistent with the observed development and with the reported timeline. Used that way — as a consistency check on a hypothesis — it is a legitimate and often decisive tool.
Used to generate an origin it is much weaker, because the model's output is governed by inputs that encode the very assumptions in question: fuel load, arrangement, ignition location, ventilation schedule. As with any reconstruction, the meaningful presentation shows which conclusions survive across the plausible input range.
The timeline as an independent constraint
Data generated independently of the investigation is disproportionately valuable in a compartment where the physical record has been degraded. Alarm and detection activations, sprinkler operation, utility interval metering, security and doorbell imagery, 911 call timing and dispatch records together establish when the fire reached particular stages.
A candidate origin that requires a development time inconsistent with that record is constrained regardless of how the patterns read. This data is also perishable — panels are reset and cleared, and retention on video and monitoring systems is often measured in days.
How these opinions are tested
Challenges to post-flashover origin opinions concentrate on a few points: that the area of origin was identified from patterns known to be unreliable under those conditions, that ventilation effects were not accounted for, that excavation was incomplete or that alternative areas were not examined equivalently, and that the opinion was not tested against the independent timeline.
An examination that documents the ventilation reconstruction, the full excavation, and the basis on which each alternative area was excluded is doing the work that survives scrutiny. Reporting a conclusion with photographs of the patterns that supported it is not the same exercise.
This article is general technical orientation, not a failure analysis, an engineering opinion, or advice on any specific matter. Determining the cause of a particular incident requires hands-on examination by a credentialed expert.