Despite being broadly banned in warfare under the terms of the 1925 Geneva Protocol, tear gas as an agent for so-called ‘riot control’ has become the preferred means for police, in the US and around the world, to clear dissenting voices from public spaces. But the toxic chemicals contained in tear gas and other widely-used chemical munitions can cause serious short- and long-term side effects, from asthma and chemical burns to lung injury and neurodegeneration.
In 2020, as protests swept across the US in response to the police killing of George Floyd and Breonna Taylor—just two in a long trajectory of other Black Americans killed with impunity by American law enforcement—police forces in cities from Philadelphia to Minneapolis responded with excessive force to suppress these demonstrations.
In Portland, Oregon, the demonstrations went on for months, and were consistently met with repeated and excessive use of tear gas by the city’s police force, the Portland Police Bureau, or PPB.
On 2 June 2020, the PPB deployed large amounts of tear gas in an attempt to force protestors to withdraw from the streets and squares of downtown Portland, unleashing a highly toxic cloud of airborne chemicals onto the same citizens it was tasked with protecting. That day would become known locally as ‘Tear Gas Tuesday’.
Forensic Architecture worked with the Department of Mechanical Engineering at Imperial College London to analyse the PPB’s use of chemical munitions that day using computer fluid dynamics (CFD) simulation.
Drawing on publicly available information, we counted, located and identified the different munitions used that night by the PPB, and created a simulation of the toxic tear gas cloud in order to measure the exposure, spread, and concentration of tear gas throughout the city.
Our analysis shows how airborne chemical concentrations were in serious excess of recognised safe thresholds, raising important questions about the proportional nature of the PPB’s response, their understanding of the potential harms of the chemical munitions that they use, and the risks inherent in how they use them. The findings of this investigation reinforce concerns raised and legal action taken by local activists and civil rights groups to restrict and ban the use of tear gas.
Our investigation shows how the use of tear gas by the PPB on 2 June exposed civilians exercising their democratic right to protest and assembly to dangerous and toxic levels of CS—or o-chlorobenzylidene malononitrile—one of the most commonly used chemicals found in tear gas.
Within a timeframe of only three hours, between 21:00 and 0:00 that night, the PPB used at least 148 munitions (of which 138 were likely chemical munitions). With our model, we are able to show that these munitions caused airborne chemical concentration levels to exceed the 2mg/m³ threshold beyond which exposure is ‘immediately dangerous to life and health’ (IDLH), as defined by federal US agencies (specifically the US Department of Labor’s Operational Safety and Health Administration). According to their own internal documents, the PPB is aware of these established IDLH thresholds and the ‘severe risks’ caused by exposing civilians to ‘high concentrations’ of such chemical agents—yet these regulations appear to have been actively ignored by the PPB officers on duty.
We evaluated the airborne concentrations of CS in fifteen different locations across downtown Portland, some in immediate proximity to protest sites and others in locations further away. At all fifteen measurement points the IDLH threshold of 2mg/m³ was exceeded for at least thirty seconds.
In nineteen separate instances across the sampling locations, the 2mg/m³ threshold was exceeded continuously for more than a minute.
Throughout the night, across all locations, concentrations of 2mg/m³ were exceeded for a total of more than 73 minutes.
At sampling location A (highlighted in image above), which simulates conditions at the intersection of SW 5th Avenue and SW Yamhill Street, a key location in the 2 June protests, the IDLH value was continuously exceeded for a period of six minutes and twenty seconds. The maximum airborne concentration recorded at sampling location A was 443.9 mg/m³.
At sampling location H (highlighted in image above), just one block north of sampling location A, at the intersection of SW 5th Avenue and SW Morrison Street, our simulation recorded the highest airborne concentration value across the sampled locations: 4,568 mg/m³, more than 2,000 times the IDLH threshold.
In addition to airborne chemical concentrations, our analysis also reveals the quantities of CS particles deposited on the ground after their discharge and dispersion in the air. Our simulation shows that teargas particles travelled up to 809 metres, accumulating in areas close to heavily frequented commercial areas and offices, as well as in the nearby Willamette River—home to a variety of aquatic species for whom CS is highly toxic—where up to 2kg of CS particles were likely deposited.
Our simulation focused on the dispersion of CS, as its health impacts have been studied in more detail than other similarly weaponised chemical compounds. It is worth noting, however, that the impact of exposure to combinations of high concentrations of CS and OC (oleoresin capsicum, the active ingredient in pepper spray)—to which Portland residents were likely subjected on 2 June and regularly throughout the following months—have not been extensively studied.
Using hundreds of videos collected from social media, news outlets, and activists on the ground, we created a synchronised video sequence covering the entire period during which protests took place throughout the night of 2 June.
Based on this video synchronisation file, we counted and geolocated the munitions deployed by the PPB and marked each instance with the relevant time code in a map of downtown Portland (a public version is accessible here), using landscape markers, buildings, and moments in which munitions are visibly discharged as visual indicators.
Next, we developed a multi-layered analysis and verification technique to identify the different types of munitions used by the PPB.
The PPB did not produce an inventory of what types and quantities of chemical munitions it used; however, it did admit to using at least eight different types between June and September 2020. As further evidence of their ongoing use of chemical weapons, we were able to obtain through public record requests invoices of munition procurement by the PPB ranging from 2016 through summer 2020.
In conjunction with these documents, we also sourced technical datasheets from weapons manufacturers corresponding to the purchased and identified munitions, which we cross-checked against the video footage to identify the different explosion and discharge patterns characteristic to each munition. We analysed these characteristics in order to conceive six different categories of emission types, which we then assigned to all the varieties of visible munitions used by the PPB that night.
For each emission type, we ‘weighted’ the quantities of CS and OC according to the variations that the PPB admit to using, and according to the number of those munitions purchased by PPB between 2016 and 2020.
The results of this process of identification, corroboration, and weighting is a table of minimum estimated quantities of CS and OC for each ‘emission type’, a public version of which can be found here. These quantities formed the basis of our simulation.
By applying this methodological approach, we were able to arrive at an estimate of the likely quantities of OC or CS emitted by each of the emission types used that night, and tie this to the locations and times of their deployment.
We created a true-to-scale digital model of downtown Portland and simulated the weather conditions of the night of June 2nd, using the available meteorological data for the relevant hours.
This model served as a digital environment in which we could simulate the dispersion of CS particles, using computational fluid dynamics simulations (for more information, see our methodology report).
Our investigation showed that the use of chemical munitions was concentrated within three peak periods across the night. The first peak lasted eighteen minutes (from 21:16 – 21:34), peak two lasted twenty-four minutes (22:04 – 22:28) and peak three twenty-eight minutes (23:32 – 00:00). Within each of these peak periods, an array of different emission type was used (marked in different colours on the timeline), which in turn produced significant clouds of airborne toxic chemicals and depositions in varying concentrations on the ground – visible and navigable in the interface below.
Using this model, we were able to estimate both the airborne chemical concentrations and the ground depositions resulting from the munition use by recording the amount of particles that pass through a defined 6m x 6m x 3m sample volume and measuring the quantity of CS particles that settle on the ground.
We placed fifteen identically sized ‘volumes’ at different points within our digital model of downtown Portland. Using the aforementioned procedure, we took measurements close to protests areas with a high density of emission points, and at points further away in the direction of the wind, to see if high concentrations also occurred in parts of the city that were a significant distance from the protest sites.
Our investigation makes measurable what was known to many of those present on the protest frontlines on 2 June 2020—that civilian protesters were subjected to extremely high concentrations of dangerous chemicals, at levels considered life-threatening. Our findings corroborate the results of initial studies mapping the range of symptoms reported by individuals present at the protests in Portland in the summer of 2020. Furthermore, airborne CS concentrations and deposits were likely present in areas of the city well beyond the parameters of the demonstration, with the long-term environmental impact and health consequences for both protesters and other residents as yet unknown. These findings lend further weight to calls for the outright ban on the use of tear gas in Portland, Oregon, the United States, and around the world.