How do investigators start? What do they look for at the scene?
The first step is always to find the fire's point of origin. Investigators look for classic burn patterns—V-shaped marks on walls, deeper charring, and the lowest point of damage—to narrow down where the fire started. This is the foundation of every investigation, as described in a 2023 review of fire chemistry [7]. Once the origin is identified, the investigator examines potential ignition sources in that area: electrical wiring, appliances, smoking materials, or chemical reactions.
But the scene alone can't tell the whole story. A 2026 conceptual paper argues that traditional methods lack the ability to establish precise timing or reconstruct pre-incident events, which is why they must be combined with other evidence [2]. For example, in a 2022 analysis of a deadly fire in Jilin Province, China, investigators determined the direct cause was an arc fault in an electrical circuit, but they also identified indirect causes like poor installation and lax supervision [5]. This shows that the physical scene provides the 'what,' but not always the 'why' or 'how.'
How do lab tests and electrical equations confirm or rule out a cause?
Once investigators have a hypothesis, they turn to lab analysis to test it. Fire debris samples are collected from the origin area and analyzed for ignitable liquid residues using gas chromatography, the most sensitive method for detecting accelerants like gasoline or lighter fluid [7]. This can prove whether a fire was intentionally set.
For electrical fires, investigators use physics. A 2022 paper introduced the 'quarter-power equation,' which calculates whether a resistive heating fault (like a loose connection or overloaded wire) could have delivered enough heat to ignite nearby materials. The author showed how this equation helped eliminate a suspected cause in a real case, proving that not every electrical fault is powerful enough to start a fire [6]. This is a key tool because electrical faults are a primary consideration in almost every fire investigation [6].
In more complex industrial fires, investigators recreate the fire using computer models. A 2025 study of a factory fire in Taiwan used a Thermogravimetric Analyzer (TGA) and Differential Scanning Calorimetry (DSC) to measure the heat released by stored materials—paper lunch boxes, tissue paper, and corrugated boxes—finding they released 848, 468, and 301 Joules per gram, respectively. These numbers were fed into a Fire Dynamics Simulator (FDS) to reconstruct the fire's spread and confirm the official cause [1]. This combination of lab chemistry and computer modeling is a powerful way to test hypotheses.
What role do digital records and virtual reality play in modern investigations?
Digital forensics is increasingly critical, especially for fires involving complex systems. In South Korea, a government committee investigated a series of Energy Storage System (ESS) fires by analyzing the battery management system (BMS) log data stored in fire-resistant safes. This data revealed that all fires occurred when the battery state of charge was above 95% and during the initiation of thermal runaway in specific cells—a pattern invisible to physical scene examination alone [3]. However, even with this data, the committee could not determine the root cause, showing that digital evidence has limits [3].
Virtual reality (VR) is also emerging as a tool for training and scene reconstruction. A 2024 paper from a cross-border partnership argues that VR can provide immersive, reusable, and contamination-free training for fire investigators, overcoming the high cost and single-use limitation of full-scale simulated fire scenes [4]. While still being evaluated for admissibility in court, VR offers a way to revisit and re-analyze a scene without returning to the physical location [4].
The key insight across all these studies is that no single method is sufficient. The 2026 paper on integrating digital forensics makes this explicit: traditional fire scene investigation and digital forensics complement each other, with digital data filling in the gaps in timing and sequence that the physical scene cannot provide [2]. The strongest investigations combine scene patterns, lab chemistry, electrical modeling, and digital logs to build a complete, evidence-based picture.
About These Sources
This answer is built on 7 peer-reviewed studies — published from 2022 to 2026, 3 from 2024 or later, 1 in Q1–Q2 journals — selected as the most relevant from 8 studies that passed quality screening, drawn from 68 papers retrieved from a database of over 500 million.
Sources used in this answer
Utilization of Calorimetric Analysis and Fire Dynamics Simulator (FDS) to Determine the Cause of Plant Fire in Taiwan: Thermogravimetric Analyzer (TGA), Differential Scanning Calorimetry (DSC), and FDS Reconstruction
In a 2025 study of a Taiwan factory fire, calorimetric analysis of stored materials (paper lunch boxes, tissue, corrugated boxes) measured their heat release (848, 468, and 301 J/g, respectively) and mass loss, which were then used in a Fire Dynamics Simulator (FDS) to reconstruct the fire's spread and confirm the official cause [1].
Integrating digital forensics into fire scene investigation: Enhancing reconstruction and analysis.
A 2026 conceptual paper argues that traditional fire scene investigation cannot establish precise timing or pre-incident events, and that integrating digital forensics fills these gaps, making the combination essential for a complete investigation [2].
Unraveling the Characteristics of ESS Fires in South Korea: An In-Depth Analysis of ESS Fire Investigation Outcomes
In an analysis of South Korean ESS fires, battery management system (BMS) log data showed fires occurred when state of charge was above 95% and during thermal runaway initiation, but the root cause remained undetermined [3].
Virtual reality aiding fire scene investigation: A thankless endeavor or catalyst for change?
A 2024 paper from a cross-border partnership explores using virtual reality (VR) for fire investigator training and scene reconstruction, arguing it is a reusable, contamination-free alternative to costly full-scale simulated scenes [4].
Analysis of Characteristics of Fire Incident on 24 July 2021 in Jilin Province, China
In a 2022 analysis of a deadly fire in Jilin Province, China, the direct cause was an arc fault, with indirect causes including poor circuit installation and lax supervision; the paper also identified electrical failure, human factors, and aging wires as common causes in public buildings [5].
The ‘quarter-power equation’ for resistive heating faults
A 2022 paper introduces the 'quarter-power equation' for fire investigation, which calculates whether a resistive heating fault could supply enough energy to start a fire, and demonstrates its use in eliminating a suspected cause in a real case [6].
Fire Chemistry and Forensic Analysis of Fire Debris
A 2023 review of fire chemistry explains that the first step is identifying the fire's origin, then collecting debris for analysis; gas chromatography is the most sensitive method for detecting ignitable liquid residues in fire debris [7].