⚡ electrical engineering

Changes in upper and lower carbon monoxide concentrations during butane heater operation in an enclosed tent

This study demonstrates that carbon monoxide concentrations in an enclosed tent with a butane heater vary significantly by height and duration of operation, initially showing higher levels near the ceiling before shifting to higher concentrations near the floor over time, thereby highlighting the critical importance of considering both measurement height and elapsed time in incident reconstructions.

Original authors: Jihyuk Kim

Published 2026-10-13

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Jihyuk Kim

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

When a fire burns inside a small, closed space like a camping tent, it consumes oxygen and releases invisible gases, including carbon monoxide. This gas is dangerous because it binds to the blood more tightly than oxygen does, starving the body of what it needs to breathe. For decades, investigators have known that the amount of this gas in a tent is not the same everywhere; it can be higher near the ceiling or higher near the floor depending on how the air moves and how long the fire has been burning. This variation makes it difficult to reconstruct what happened in a tragic accident, especially when people are sleeping on the ground. If a person is found unconscious near the floor, a simple assumption that the air was worst at the ceiling could lead to the wrong conclusion about their exposure. Understanding how these gases move and change over time is essential for piecing together the story of a fatal incident.

A researcher at the National Forensic Service in South Korea set out to map this movement with precision. They used the exact tent and portable butane heater involved in a real fatal camping incident to conduct three controlled experiments. Their goal was not to guess what happened, but to measure exactly how the air changed from the moment the heater was lit until the experiment ended. They placed sensors near the floor, just a few centimeters above the ground, and others near the top of the tent, about 140 to 150 centimeters up. They let the heater run continuously for forty minutes, recording the levels of carbon monoxide and oxygen every three minutes.

The results revealed a shifting pattern that challenges a simple view of how these gases behave. In the first few minutes after the heater was turned on, the air near the top of the tent contained significantly more carbon monoxide than the air near the floor. At the three-minute mark, the sensors near the roof recorded an average of 40.33 parts per million, while the sensors on the floor read only 8.67 parts per million. This initial difference makes sense because the heat from the fire pushes the warm, gas-filled air upward. However, as the heater continued to run, the situation reversed. By the thirty-nine-minute mark, the air near the floor had become more dangerous, with levels rising to an average of 101.67 parts per million, while the levels near the roof settled at 93.33 parts per million.

The researcher found that the air near the floor accumulated carbon monoxide at a faster rate than the air near the roof. Over the course of the experiment, the concentration near the floor increased by about 2.68 parts per million every minute, compared to an increase of about 1.51 parts per million every minute near the top. This meant that while the top of the tent started with higher levels, the bottom eventually caught up and surpassed it. The moment this switch happened varied slightly between the three trials, occurring sometime between the twenty-seventh and thirty-third minute. Despite this late surge near the floor, the average amount of carbon monoxide a person would have breathed over the entire forty-minute period was still higher near the top of the tent.

The study also tracked oxygen levels, which dropped steadily in both locations as the heater consumed the air. The oxygen near the roof fell from about 20 percent to 16.8 percent, while the oxygen near the floor dropped from 20.9 percent to 17.9 percent. The researcher noted that the air near the roof remained slightly more depleted of oxygen throughout the entire experiment. These findings highlight a critical lesson for forensic investigators: the height at which a measurement is taken and the time at which it is taken both matter immensely. A single reading taken at the end of an incident might show the highest danger near the floor, but that snapshot misses the earlier period when the danger was greater near the ceiling.

This work does not calculate exactly how much poison a specific person inhaled, as that depends on their breathing rate, their position, and how long they slept. Instead, it provides a clear record of how the environment itself changed. The researcher showed that the air in a tent is not static; it evolves. The early dominance of carbon monoxide at the top gives way to a later dominance near the floor. For anyone trying to understand a camping tragedy, this means that relying on a single measurement or a single assumption about where the gas settles is insufficient. The full picture requires knowing how the concentrations shifted from the moment the fire started until the end, acknowledging that the most dangerous air can move from the ceiling to the ground as time passes.

Technical Summary: Changes in Upper and Lower Carbon Monoxide Concentrations During Butane Heater Operation in an Enclosed Tent

Problem Statement
Carbon monoxide (CO) poisoning remains a significant cause of accidental fatalities in temporary shelters, such as camping tents, often resulting from the use of gas appliances or charcoal for heating. While previous experiments have characterized peak CO concentrations and variations based on appliance design and shelter geometry, the temporal evolution of vertical concentration gradients during continuous heater operation is less understood. Specifically, it is unclear whether the ordering of upper versus lower CO concentrations remains consistent throughout an incident. This uncertainty complicates forensic reconstructions, where the interpretation of environmental conditions relies on understanding how CO distributes vertically over time, particularly for occupants sleeping near the tent floor.

Methodology
The study conducted three full-scale reconstruction trials using a dome-style camping tent (260 × 260 × 150 cm) and a portable liquefied-butane heater, both submitted from a fatal incident investigation. The heater was placed centrally on the tent floor and operated continuously for 40 minutes in an unventilated, enclosed environment.

Gas monitoring was performed using two MultiRAE Lite portable multigas monitors equipped with electrochemical sensors for CO and oxygen (O₂). Measurements were taken concurrently at two distinct heights:

Upper position: 140–150 cm above the floor (near the tent roof).
Lower position: 0–10 cm above the floor.

To minimize instrument bias, monitor assignments were alternated between runs. Data were recorded at 3-minute intervals from 3 to 39 minutes post-ignition. The primary analysis focused on the paired difference between upper and lower CO readings ( $\Delta CO$ ), the time-weighted mean concentrations, and the rate of O₂ depletion. Linear regression was used to describe concentration trends, and residuals were analyzed to examine departures from linearity.

Key Results

CO Accumulation and Vertical Reversal: CO concentrations increased at both heights over the observation period. However, the vertical distribution shifted significantly over time.
- Early Phase (3 min): The upper position showed a mean CO concentration of 40.33 ppm, while the lower position was significantly lower at 8.67 ppm.
- Late Phase (39 min): The trend reversed, with the lower position reaching a mean of 101.67 ppm and the upper position at 93.33 ppm.
- Transition: The difference between upper and lower readings ( $\Delta CO$ ) decreased from +31.67 ppm at 3 minutes to −8.33 ppm at 39 minutes. The transition point, where lower readings exceeded upper readings, occurred between 27 and 33 minutes, varying by trial.
Time-Weighted Means: Despite the higher near-floor concentration at the 39-minute endpoint, the time-weighted mean CO concentration from 3 to 39 minutes remained higher at the upper position (62.99 ppm) compared to the lower position (51.10 ppm) in all three trials.
Oxygen Depletion: O₂ levels declined steadily at both positions. The upper mean dropped from 20.03 vol% to 16.80 vol%, and the lower mean decreased from 20.90 vol% to 17.97 vol%. The rate of depletion was consistent between heights (slopes of approximately −0.085 percentage points/min).
Slopes: The linear slope of CO accumulation was steeper near the floor (2.684 ppm/min) than at the upper position (1.509 ppm/min) across all trials.

Key Contributions
The study provides empirical data demonstrating that the vertical ordering of CO concentrations in an enclosed tent is not static during continuous heater operation. It establishes that an initial excess of CO at the upper level, driven by buoyancy, is followed by a period where near-floor concentrations exceed upper concentrations. The paper highlights that relying solely on endpoint measurements or early readings can lead to incomplete reconstructions, as the "higher" concentration location changes over time. Furthermore, it documents that while the endpoint concentration may be higher near the floor, the cumulative exposure (time-weighted mean) over the interval can still be greater at the upper level.

Significance and Claims
The author asserts that these findings are critical for forensic interpretations of tent incidents. The study does not claim to establish individual inhaled doses or fatal exposure durations, noting that posture, movement, and respiration affect actual exposure. Instead, the significance lies in demonstrating that both measurement height and elapsed time are essential variables when reconstructing the environmental conditions of a tent fire or heater incident.

The paper concludes that a forensic assessment must consider the concentration history at specific locations rather than assuming a uniform distribution or a fixed vertical gradient. The observed shift from upper to lower dominance in CO concentration suggests that buoyancy-driven transport is initially dominant but is eventually overtaken by cooling, recirculation, and restricted ventilation effects. The author emphasizes that these results apply specifically to the tested unoccupied tent-heater configuration and that the study establishes repeatability only for this specific setup, without extrapolating to other shelter designs or occupied scenarios.

Technical Summary: Changes in Upper and Lower Carbon Monoxide Concentrations During Butane Heater Operation in an Enclosed Tent

More like this