Cigarette ignition: the gasoline tests and the limits of a fire reconstruction
Cigarette ignition of spilled gasoline has repeatedly failed in controlled testing. A burning tobacco cigarette and gasoline vapor did not produce ignition in the tested conditions. That finding concerns a particular fuel and ignition source. Federal research on cigarettes in furnishings and guidance on reconstructing fires address different hazards and different evidence.
Howard Marcus and Justin Geiman's research appeared online in 2013 and in Fire Technology's 2014 volume. Both authors were affiliated with the Bureau of Alcohol, Tobacco, Firearms and Explosives. Their study covered five major cigarette brands, with cigarettes both resting and under draw, and several configurations involving gasoline. Across 70 distinct tests, they used 723 cigarettes and recorded more than 4,500 exposures to ignitable vapor concentrations. They reported no ignition of gasoline vapor by the cigarettes. Those numbers describe the experimental record; they are not a measurement of the probability of every conceivable accident.
The investigators counted 70 distinct tests and more than 4,500 exposures within them. The published abstract identifies the vapor concentrations as ignitable, so the failures did not simply reflect a mixture incapable of burning. The conclusion concerns cigarette ignition under the tested conditions. It does not identify the cause of any particular historical fire or establish who was present when one began.
NIST Technical Note 1627, issued in 2009, demonstrates why cigarette fire risk still requires serious measurement. Richard Gann and Edward Hnetkovsky studied cigarettes used as standard ignition sources for testing mattresses and upholstered furniture. A conventional test cigarette was no longer being manufactured, so reproducible replacements were needed. Their work concerned how consistently a cigarette could supply heat during a furnishings test. Gasoline-vapor ignition was not its target. The existence of that testing programme makes the material-specific distinction concrete: cigarettes can be consequential ignition sources even when cigarette ignition of gasoline repeatedly fails.
The NIST report describes a standard measure based on whether a cigarette continues burning its full length despite heat loss to a supporting material. For each cigarette type, the measure uses 40 determinations and records the fraction that burn their full length. Earlier testing linked that performance to ignition in furniture mockups and upholstered chairs. It also reports measurable differences between conventional commercial cigarettes and a reduced-ignition-propensity cigarette. Those findings concern heat transfer, cigarette design and the surrounding material. They cannot be substituted for the gasoline experiment, any more than the gasoline experiment can be used to dismiss the risk of a smoldering fire in furnishings.
That distinction became personally significant to John Galvan. In interviews published by the Innocence Project in October 2022, Galvan recalled watching a television programme whose presenters repeatedly failed to ignite a gasoline pool with a cigarette. His attorney Tara Thompson had seen the same rerun. Both recognized a connection with the ignition mechanism in the statement used against him. Thompson described the programme as a reason to investigate the fire science further. The broadcast was a lead for legal research; the interviews do not make it a substitute for expert evidence or a court decision.
The organization's July 21, 2022, statement records the eventual legal outcome: Cook County prosecutors dismissed the cases against Galvan, Arthur Almendarez and Francisco Nanez. It identifies the September 1986 Chicago apartment fire in which Julio and Guadalupe Martinez died, and describes years of litigation over forensic testimony and alleged coerced confessions. The statement came from organizations representing the men and should be understood in that role. It describes multiple strands of the challenge, including alibi testimony, rather than attributing the outcome solely to a television demonstration.
The National Institute of Justice's 2000 Fire and Arson Scene Evidence guide explains what a fuller investigation requires. Its working group defined a fire's cause through the ignition source, the first fuel ignited and the circumstances bringing them together. It called for collecting and analyzing data, developing and testing hypotheses, and considering alternatives. Investigators were also told to recognize the limits of their expertise and seek specialist help, particularly where deaths or major losses were involved. Cigarette ignition is therefore a proposed mechanism to evaluate against evidence, rather than an explanation established by naming a familiar object.
The same guide emphasizes preserving the scene and documenting evidence before it is disturbed. A fire can consume information about its own origin, and rescue or suppression work can alter what remains. That is why a laboratory result and a historical reconstruction have different evidential jobs. The gasoline study tests a mechanism; scene evidence and witness material address the particular event. Treating those jobs separately preserves the force of the failed ignition results while retaining the wider, well-documented fire hazards of burning cigarettes.
- Howard A. Marcus and Justin A. Geiman, The Propensity of Lit Cigarettes to Ignite Gasoline Vapors — Fire Technology 50 (2014), primary research abstract — link.springer.com
- Richard G. Gann and Edward J. Hnetkovsky, NIST Technical Note 1627 (2009), Modification of ASTM E 2187 — full federal research report — tsapps.nist.gov
- National Institute of Justice, Fire and Arson Scene Evidence: A Guide for Public Safety Personnel (2000), especially pp. 3–4 and 25–26 — www.ojp.gov
- Innocence Project, October 5, 2022 interviews with John Galvan and attorney Tara Thompson about the television experiments — innocenceproject.org
- Innocence Project, July 21, 2022 statement on dismissal of the Galvan, Almendarez and Nanez cases — innocenceproject.org