Quebec appeal court restores 1991 start date in Horne foundry class action

The Quebec Court of Appeal has restored a 1991 starting date for a group seeking damages over emissions from the Horne foundry, ruling that a dispute about when residents understood their alleged harm could not be resolved at the class-action authorization stage. In Fortier c. Procureur général du Québec, 2026 QCCA 1239, released September 21, 2026, the court allowed the residents’ appeal and amended the authorized group and proposed compensation plan.

The decision expands the period covered by one subgroup in the proceeding against Glencore Canada Corporation and Quebec’s attorney general. It leaves liability, the existence and value of compensable harm, and the contested limitation-period questions for the merits of the case. The proposed monthly compensation figures are claims, not damages awarded by the court.

A dispute over when residents understood the alleged harm

Julie Fortier and Miguel Charlebois applied for authorization on October 23, 2023. They alleged that emissions of toxic or carcinogenic contaminants from the Glencore-operated foundry caused fear, anxiety, stress, anger, guilt and similar harm. Their proposed action also challenged the Quebec government’s authorization of Glencore’s activities and alleged failure to give residents adequate and timely information about the exposure and its health effects. They sought compensatory and punitive damages.

Superior Court Justice Daniel Dumais authorized the action on September 30, 2025, in Fortier c. Glencore Canada Corporation, 2025 QCCS 3525. However, he applied the general three-year prescription period in article 2925 of the Civil Code of Québec and limited the relevant subgroup to people who had lived within 10 kilometres of the foundry from October 23, 2020 onward.

The judge concluded that the proposed representatives sufficiently knew the elements of their claim by 2019. Among the evidence was a petition both had signed that year concerning arsenic emissions and health risks. He also considered Charlebois’s statements about his concerns and Fortier’s participation in a biomonitoring study and receipt of information about exposure.

The residents’ appeal challenged that determination at the preliminary authorization stage. They argued that information released in 2022 had revealed the full extent of the contamination and its effects, producing the moral harm for which they sought compensation.

Exposure history was part of the damages theory

Writing for a unanimous panel with Justices Simon Ruel and Sophie Lavallée, Justice Stéphane Sansfaçon explained a distinction in the residents’ case. They said they were seeking compensation for harm experienced in 2022 upon learning the extent of past exposure, rather than for distress they had necessarily experienced throughout the years before then.

On that theory, the number of months of exposure would help measure the intensity of the later distress. Someone learning of 30 years of exposure could, they argued, suffer greater harm than someone exposed for only a year or two. The appeal court acknowledged that this theory had been explained much more clearly on appeal than in the authorization application.

The court also accepted that an authorization judge can consider proportionality and the demands a broad class action may place on limited judicial resources. Those considerations did not resolve the disputed factual question in this case.

Prescription issue required a merits hearing

Prescription generally begins when a claimant knows the fault, damage and causal connection supporting the claim. The court explained that a pure question of law can be decided at authorization. A claim that is prescribed on its face can also be screened out at that stage. But determining when a right of action arose is ordinarily a highly factual exercise.

Here, the parties disputed what the residents knew in 2019 and what changed in 2022. The court noted that the earlier biomonitoring reports focused on the Notre-Dame neighbourhood, while information about other areas and exposure scenarios became available in 2022.

The authorization judge could not settle that dispute without exceeding his screening role, the court held at paragraph 24. Assessing the residents’ knowledge and alleged moral harm raised factual, or at least mixed factual and legal, questions for the merits. Although they had been examined out of court, they had answered defence lawyers’ questions and had not yet presented their complete account at a hearing where credibility could be assessed.

What the court changed

The appeal court replaced October 23, 2020 with January 1, 1991 in the relevant subgroup’s description and the corresponding compensatory-damages portions of the proposed plan. It made a matching January 1991 change to the proposed punitive-damages period and standardized one territorial description in the plan. All other conclusions of the authorization judgment remained unchanged. The respondents were ordered to pay the appeal’s legal costs.

The ruling preserves a broader case for trial without deciding whether the residents will recover compensation. Its procedural significance is that a contested account of when harm became sufficiently known cannot be treated as a settled limitation issue simply because the case is at authorization.