Detroit Diesel settles EPA claims over 16 unreported opacity exceedances

The number that crossed the ceiling

A monitor on Detroit Diesel’s M-11 engine-production stack recorded a six-minute average of 51.55% opacity on the evening of December 15, 2023. Opacity measures how much light a plume blocks. It does not directly state how many pounds of pollution entered the air, but it is a regulated indicator of visible emissions.

Michigan’s rule generally limits visible emissions to 20% opacity over a six-minute average. It permits one six-minute average per hour above 20%, but that exception cannot exceed 27%. EPA’s case therefore focused on measurements above 27%, where the hourly exception would not protect the reading.

A visible plume became a continuous test

The enforcement trail began before the monitor produced its December readings. EPA inspected the Detroit facility on October 24, 2022. According to the order, inspectors saw visible emissions from the M-11 stack and used an optical gas-imaging camera to observe emissions trailing and sinking from it.

Facility representatives told EPA that “visible emissions are typically observed from the M-11 Stack.” EPA consent agreement, paragraph 30. The same paragraph says representatives did not believe the routinely observed emissions exceeded 20% opacity.

EPA then issued an information request in July 2023 requiring Detroit Diesel to install a continuous opacity monitoring system and record the stack for a continuous 30-day test. The order says the company supplied evidence that the monitor was installed and certified under the applicable federal performance specification, including calibration and audit records.

What the monitor recorded

The readings were not evenly distributed. Four consecutive exceedances appeared late on December 10. Three more appeared on December 18. The highest result came during one of two listed events on December 15.

Source: the measurement table in the EPA consent agreement. On narrow screens, the table can be scrolled horizontally.

The second alleged failure happened on paper

Detroit Diesel’s 2022 air permit required certified semiannual reports covering required monitoring and every deviation from permit conditions. The company submitted its report for July through December 2023 to the Michigan Department of Environment, Great Lakes, and Energy on March 12, 2024, then shared it with EPA on March 20.

EPA alleged that the report did not identify any of the 16 opacity events as deviations. That omission doubled the structure of the case. Each measurement supported an alleged visible-emissions count, and each unreported event supported a corresponding permit-reporting count.

Reporting rules matter because regulators cannot continuously stand beside every industrial stack. Permit systems depend partly on companies disclosing their own departures from operating limits. When a monitored event is left out, the problem is not only the underlying reading. The official compliance record also becomes incomplete.

What the settlement resolved, and what it did not adjudicate

Detroit Diesel consented to the $125,397 civil penalty, agreed to comply with the order, waived its right to contest the allegations, and waived its right to appeal. The agreement also says the action can be considered in the company’s compliance history in later enforcement proceedings.

The distinction is important. This was not a criminal conviction, and the order does not say a court independently found every allegation true. It is a final administrative settlement in which the company accepted the penalty and surrendered its right to challenge the allegations.

The record includes one contemporaneous company position: facility representatives said in 2022 that they did not believe routinely visible emissions exceeded 20% opacity. The supplied document contains no separate public statement from Detroit Diesel addressing the later monitor results.

What the record says about harm, and what it cannot say

The allegation section describes the engine-production line as emitting air contaminants that include particulate matter, nitrogen oxides, and hydrocarbons. The order does not, however, translate the opacity readings into pollutant mass. It also does not provide air-monitoring results outside the facility, identify individual exposures, or attribute any illness to the cited events.

That means the supplied record supports a clear compliance story, but not a quantified health-damage claim. Opacity can signal a plume dense enough to violate a visible-emissions standard. It cannot, by itself, establish which pollutants were present in a particular six-minute period, their concentrations beyond the stack, or who encountered them.

The documented public consequence is institutional: EPA says an industrial stack exceeded an enforceable visible-emissions ceiling repeatedly, while the permit report that regulators relied on omitted those events. The available source does not establish broader personal harm.

The penalty is public. The corrective plan is not in this record.

The settlement resolves federal civil-penalty liability only for the violations specifically alleged in the order. It does not excuse compliance with other federal, state, or local requirements.

The document also refers to a separate administrative compliance order issued concurrently. That order was not included in the supplied material. As a result, this record does not show what operational changes, monitoring duties, engineering controls, or deadlines may have been imposed outside the penalty agreement.

The 51.55% reading is the most visually striking number in the order. The deeper accountability failure alleged by EPA is quieter: the monitor created a record, but the company’s required compliance report did not carry that record forward. The settlement puts a price on both parts of the case. The missing compliance order will determine how much of the underlying operation is required to change.