The Washington Post published “Are induction stoves that much safer than gas? We tested them.” on January 16, 2024. This piece presents an unfair picture of the data from the study commissioned by the Washington Post and conducted by RoundhouseOne. Outlined below are the claims made in the article and responses to each.

Claim: The natural gas stove is representative of that used in the average American kitchen.

Response: The stove in question is a Wedgewood natural gas stove. Wedgewood went out of business by approximately 1970, setting a minimum age for the stove of 54 years. However, based on the appearance and on visual searches of similar Wedgewood vintage/antique stoves (retail price approximately $4,000), it appears more likely to have been manufactured in the 1940’s or 1950’s, setting the likely age at between 70 and 84 years. The fact that this stove theoretically could have belonged to President Roosevelt is relevant for multiple reasons.

This stove uses a pilot light. Pilot lights on natural gas ranges were phased out starting in 1990. As the average lifespan of a natural gas stove is 15 years, and the average electric stove 13 years, there are likely to be very few 80-year-old stoves still in use today. Natural gas stoves have improved dramatically over the past eight decades. A more appropriate comparison would have used a modern stove manufactured less than 15 years ago.

The authors provide no documentation on how the stove was maintained. Had it been serviced in the past one to three years? Had it ever been serviced? The answer to that question is likely to have a major impact on the test results.

Claim: “Based on several safety standards, levels of pollution detected by the sensors regularly rose to hazardous levels.”

Response: NO2 is difficult to measure directly, and NOx and NO2 sensors can be difficult to calibrate and prone to bias and error. Standard practice for scientific studies using them is to provide extensive documentation on how they were calibrated. This information is entirely absent from the article. Where in the kitchen were the sensors placed? We simply don’t know, but the question is as significant to the outcome as the question of how they were calibrated.

The way the authors interpret their air quality measurements appears deeply disingenuous, for multiple reasons. The authors do point out that the test kitchen started off with NO2 levels significantly elevated before starting the test– fully halfway to the final reported level. The authors however, stating that one will not be attempting to account for confounding variables is deeply insufficient if one then attempts to draw a conclusion, as the authors did, while still ignoring the many potential confounding factors.

Even with the unaccounted-for elevated background levels of NO2, the study found that the average 20-minute NO2 exposure was a third less than a third of the EPA national ambient air quality standards for one-hour exposure to NO2. They also mention that the peak level of NO2 briefly exceeds the most stringent one-hour standard, which is not a meaningful comparison. Comparisons of peak NO2 concentrations, which are instantaneous and transient, to the 1-hour time-average NAAQS are inappropriate and can lead to misleading conclusions. Why did the authors not provide the one-hour average NO2 concentration while referencing the one-hour safety threshold?

Claim: “Our findings suggest that some people — particularly those with respiratory or heart conditions, the elderly and children — may have reason to worry [about cooking with natural gas].”

Response: The study found that ventilation is exceptionally effective for removing excess NO2. However, it found that ventilation was less effective for removing PM 2.5 particulate matter that comes from food, meaning the food being cooked has a significant impact on indoor air quality.

The authors present no information on controlling for the pan, cooking procedures, and food products used in the testing. Work from GTI Energy indicates that heat distribution across the pan and inconsistency in meat products utilized in cooking have significant effects on emissions associated with cooking.

The study found that the air quality impact of using an induction stove is worse than that of a natural gas stove, with their burger test producing an air quality index fully 23 points higher (unhealthier) for an induction stove than a natural gas stove.

In other words, the induction stove has a higher risk for sensitive individuals according to the information provided by the authors, with this risk being more difficult to ameliorate.