COURT OF APPEAL FOR ONTARIO

CITATION: Green Rise Foods Inc. v. N.V. Hagelunie, 2026 ONCA 334

DATE: 20260511

DOCKET: COA-24-CV-1361

Tulloch C.J.O., Roberts and George JJ.A.

BETWEEN

Green Rise Foods Inc.

Plaintiff/Applicant (Appellant)

and

N.V. Hagelunie

Defendant/Respondent (Respondent)

Keith Marlowe, K.C., for the appellant

Lindsay Lorimer and Rachel Cooper, for the respondent

Heard: October 1, 2025

On appeal from the order of Justice J. Ross Macfarlane of the Superior Court of Justice, dated November 27, 2024, with reasons at 2024 ONSC 6622.

L.B. Roberts J.A.:

A. Overview

[1] This appeal concerns the correct analytical approach to be followed regarding the interpretation and application of the standard form greenhouse insurance policy issued by the respondent to the appellant (“the Policy”) in the context of a claimed loss caused by a series of events.

[2] The appellant’s greenhouse tomatoes were destroyed by the faulty emission of excessive carbon monoxide from a malfunctioning boiler that went undetected by a malfunctioning monitor. The appellant’s loss for the destroyed tomato plants was denied by the respondent. The motion judge upheld the respondent’s denial of coverage and granted it boomerang summary judgment.

[3] The appellant appeals from the summary dismissal of its coverage claim for loss under the Policy and raises several grounds of appeal. They can be more compendiously expressed in the following overarching grounds: 1) did the motion judge fail to carry out the requisite legal analysis of the standard form policy that applies in the context of a loss caused by a series of events; and 2) if so, should partial summary judgment be granted to the appellant or are there genuine issues requiring a trial?

[4] The correctness standard of appellate review applies to the motion judge’s interpretation of this standard form insurance policy: Ledcor Construction Ltd. v. Northbridge Indemnity Insurance Co., 2016 SCC 37,[2016] 2 S.C.R. 23, at para. 4; Emond v. Trillium Mutual Insurance Co., 2026 SCC 3, 509 D.L.R. (4th) 583, at para. 67.

[5] I accept that the motion judge’s analysis of the Policy was legally flawed.

[6] First, in relation to the determination of the issue of causation for the claimed loss, the motion judge erred by failing to determine the effective cause or causes of the claimed loss in the context of this case where there were a series of potentially causal events. Specifically, the motion judge erred by only considering the immediate-in-time effect of the carbon monoxide emissions as the cause of the claimed loss for the purposes of determining coverage, without regard to the boiler and monitor malfunctions and other potential causes, as the potential effective cause or causes of the claimed loss.

[7] Second, the motion judge erred in his consideration of the machinery breakdown exception to coverage under the initial grant of coverage under the Policy because he failed: 1) to address who had the onus to demonstrate whether the machinery breakdown exception did or did not apply; and 2) to make the requisite factual findings as to the causes of the boiler and monitor malfunctions.

[8] I would allow the appeal. As I explain in more detail below, I am of the view that summary judgment should not have been granted as there are genuine issues requiring a trial. I would remit the entire action for trial.

B. Factual background

i. The loss claimed under the policy

a. The greenhouse operations

[9] The appellant grows tomatoes in a large greenhouse operation, which it purchased on February 1, 2021.

[10] Boilers are used in the greenhouse operation: 1) to provide heat for the greenhouses; and 2) to generate carbon dioxide, which is vented into the greenhouses to accelerate photosynthesis and promote growth of the tomato crop.

[11] By burning natural gas, boilers produce “flue gas” consisting of the carbon dioxide required for distribution into the greenhouses, as well as carbon monoxide. Specifically, carbon monoxide is produced by incomplete combustion when there is not enough oxygen to react with all the carbon in the fuel. A portion of the produced flue gas at a temperature ranging from 200°F to 325°F is extracted and introduced to a condenser unit.

[12] The condenser unit is designed to introduce safe flue gas into the greenhouses by reducing the temperature and moisture of the flue gas. Additionally, the condenser unit is equipped with a Sercom Monitor (the “monitor”) to monitor the flue gas’s carbon monoxide level. The monitor is triggered if the level of carbon monoxide exceeds 25 ppm/30 ppm.

[13] The system supplies carbon dioxide-enriched flue gas to the greenhouses for crop growth. When the boilers are functioning safely and efficiently, the flue gas contains only safe levels of carbon monoxide. If carbon monoxide emissions exceed safe levels, the monitor is intended to alarm and vent unsafe carbon monoxide gas out of the greenhouse.

[14] Since 2016, the inspection, maintenance and repair of the boilers had been carried out by a third-party company, Powerhouse Boiler & Combustion Ltd. (“Powerhouse”). The motion judge found that days prior to the purchase of the greenhouse operation, Powerhouse carried out an inspection of the boilers and determined they were efficiently and safely functioning. There is no evidence that Powerhouse examined the monitor at that time.

b. The Greenhouse Policy

[15] When it purchased the greenhouse operation, the appellant entered into a greenhouse horticultural policy of insurance issued by the respondent. The Policy provides standard and extended property and crop and business interruption coverage. Under the standard crop and business interruption coverage, a loss is covered only if the event causing damage to the property is covered under the standard property coverage. The nature of the standard property coverage is described in section I of the terms and conditions and is subject to an exception for loss caused by machinery breakdown, which is included in the initial grant of coverage, as follows:

Standard property coverage

You are insured for accidental physical loss of or damage to:

• Greenhouses,

• Commercial buildings,

• Inventory, agricultural produce and items in stock,

• functional furnishings, fixtures and fittings inside commercial buildings and greenhouses.

The standard property coverage applies:

• Only if the item is noted on your "Declaration Page".

• While the item is at the risk location.

• Unless the loss or damage was caused by machinery breakdown. [Emphasis in original.]

[16] The definition of machinery breakdown under the Policy reads as follows:

Machinery breakdown means the accidental loss as a result of inherent vice and/or any other internal cause to any machine (including computers), instrument, appliance, equipment (including all electronic equipment attachments, accessories, fittings, cables, pipes and such associated property but excluding foundation and brickwork) during the period that this item is in use or installed and ready for use. [Emphasis in original.]

[17] In addition to the machinery breakdown exception included in the standard property coverage provisions, there are certain exclusions applicable to all coverage under the Policy, including exhaust gas, machinery breakdown, lack of maintenance, nutrient water contamination and setting error. The relevant exclusions in issue in this appeal are exhaust gas, machinery breakdown and pollution. They read as follows:

Excluded perils and cover Limiting conditions

These excluded perils and cover limiting conditions apply to all coverages under this policy. You are never insured for any loss or damage, actual or alleged legal liability or business interruption caused by, arising from, or in connection with any of the following:

11. Exhaust gas

Any exhaust gas.

21. Machinery breakdown

• Machinery breakdown on crop

• when causing standstill of crop handling systems and therefore a delay in delivery of your output, or climate damage, water and nutrition damage or light damage to your crop.

• Unless the extended coverage is noted on your "Declaration Page" under coverage and limits insured.

• Machinery breakdown to moving work equipment.

• Machinery breakdown due to standstill or maintenance.

33. Pollution and contamination

All pollution and contamination

• Unless caused by an event that is covered under this policy.

• Unless caused by smoke and/or soot from an insured event at the risk location. [Emphasis in original.]

[18] Exhaust gas is not a defined term under the Policy.

[19] “Pollution” is defined under the Policy, as follows:

damage or loss arising from the discharge, dispersal, release or escape of smoke, vapours, soot, fumes, acids, alkalis, toxic chemicals, liquids or gases, waste materials or other irritants, contaminants or pollutants into or upon any property, land, atmosphere or any water course or body of water. This includes any effects or by-products of above ground or subterranean mining.

c. The claimed loss

[20] On February 23, 2021, the appellant began observing signs of damage to its tomato plant crop. Investigation into the cause of the damage ensued. It was not until March 17, 2021, that the appellant discovered excessive amounts of carbon monoxide were being emitted by an improperly functioning boiler and that the monitor was not working and had failed to detect the high levels of carbon monoxide flowing into the greenhouses through the carbon dioxide flue-gas distribution system. In consequence, the appellant suffered a loss of 23 acres of its tomato plants.

[21] It was agreed that the appellant’s loss arose from the following series of events. One of the appellant’s boilers malfunctioned because of a displaced internal burner diffuser, although the cause of the displacement was unexplained. This displacement caused the boiler to emit excess levels of carbon monoxide, which entered into the greenhouses through the flue-gas distribution system. The excess levels of carbon monoxide poisoned the tomato plants. The boiler’s malfunction was not detected by the monitor. The monitor did not trigger an alarm because it, too, malfunctioned. It was common ground that if the monitor had not failed, the boiler’s excess emission of carbon monoxide would have been detected before the tomato plants were poisoned and lost.

[22] On March 19, 2021, the appellant made a claim for the loss of its tomato plants under the Policy.

[23] On March 22, 2021, a crop expert hired by the respondent attended the greenhouse operation and concluded that the appellant’s loss was caused by excessive levels of toxic gases, including carbon monoxide. He recommended that the entire crop be removed. He did not inspect the malfunctioning boiler or monitor.

[24] In its letter dated March 24, 2021, the respondent informed the appellant that the crop damage was not covered under the Policy. It based its denial of the appellant’s claim on its crop expert’s opinion that the crop damage was likely incurred due to incomplete combustion of the boiler causing emissions of different toxic exhaust gases, including carbon monoxide. The respondent stated that the loss was caused by machinery breakdown and relied in part on the Policy exclusion for losses caused by the emission of exhaust gas, as follows:

The loss is caused by a machinery breakdown. A malfunctioning / defect of the burner and [carbon monoxide] monitor. Damage to your crop due to machinery breakdown is only covered if causing climate damage, water and nutrition damage, light damage or operational delay and needs to be registered and alarmed. This machine breakage coverage does not include damage due to incorrect flue gasses. ...

Besides, damage due to exhaust gasses in general is specifically excluded under exclusion 11. [Emphasis in original.]

[25] By its letter dated June 3, 2021, the appellant challenged the denial of its loss, asserting that the loss was caused by an “accidental event”, which is covered under the Policy.

[26] In its August 17, 2021 response, the respondent repeated and expanded on its denial of the appellant’s claim because of the machinery breakdown exception as well as the machinery breakdown, exhaust gas, and other exclusions. It added that the machinery breakdown of the boiler and monitor was caused by lack of maintenance, despite not inspecting them or requesting maintenance records. With respect to the monitor, the respondent advised that “any loss or damage from incorrect settings, programming, operation, composition, dosage, installation and alarm threshold of equipment is also excluded, and these are also potential causes of the failure of the [monitor].”

[27] The appellant commenced an action for damages for breach of contract arising from the respondent’s failure to indemnify it under the Policy for its claimed loss.

ii. The decision below denying coverage and granting boomerang summary judgment

[28] The appellant brought a motion for partial summary judgment, seeking a declaration that it was entitled to coverage for its claimed loss and that the respondent was liable to indemnify it under the Policy.

[29] The motion judge commenced his reasons with an accurate review of the applicable rules and governing principles regarding summary judgment. He noted the parties’ well-established obligation to put “their best foot forward” on the motion and the moving party’s onus to establish that there is no genuine issue requiring a trial: Chernet v. RBC General Insurance Company, 2017 ONCA 337, 11 M.V.R. (7th) 1, at para. 12; Chao v. Chao, 2017 ONCA 701, 99 R.F.L. (7th) 281, at para. 16. He acknowledged the circumstances when partial summary judgment is appropriate and when boomerang summary judgment may be ordered: Rules of Civil Procedure, R.R.O. 1990, Reg. 194, r. 20.01(1); Graham v. Toronto (City), 2022 ONCA 149, 23 M.P.L.R. (6th) 179, at paras. 9-13. There is no issue taken with the motion judge’s recitation of the rules and governing principles on summary judgment.

[30] The motion judge moved on to consider the general principles guiding the interpretation of standard form insurance policies set out in Ledcor, at paras. 49-52. In addition to reviewing the governing principles concerning the treatment of unambiguous and ambiguous policy language, the motion judge instructed himself on some of the shifting onuses: the insured has the initial onus of proving that the damage or loss claimed falls within the initial grant of coverage; the onus then shifts to the insurer to establish that none of the exclusions to coverage applies; and, if successful, the onus then shifts back to the insured to prove that an exception to the exclusion applies. The motion judge did not expressly address who bore the onus of proof with respect to the exception in the initial grant of coverage for machinery breakdown.

[31] In determining the cause of the loss, the motion judge accepted and neatly summarized the agreed series of events that culminated in the appellant’s loss in para. 27 of his reasons:

I find that the direct cause of the Loss was excessive levels of [carbon monoxide], which were created by the malfunctioning Boiler and went undetected by the malfunctioning Monitor.

[32] The motion judge dismissed the appellant’s motion. He found that the language of the Policy – including the language of the exhaust gas exclusion and machinery breakdown exception – was unambiguous. He concluded that the policy exclusion for exhaust applied; alternatively, if he were in error, he applied the machinery breakdown exception in the initial grant of coverage.

[33] For the purpose of considering the application of the exhaust gas exclusion, the motion judge tacitly assumed that the claimed loss fell within the initial grant of coverage. He accepted the respondent’s position that the policy exclusion for exhaust gas applied. Specifically, he found that:

[T]he “exhaust gas” exclusion in the Policy is a complete answer to [the appellant’s] claim. It provides, in clear and unambiguous terms without any limits or exceptions, that “you are never insured for any loss or damage, actual or alleged legal liability or business interruption caused, by, arising from or in connection with ... any exhaust gas.” I find that [carbon monoxide] is an exhaust gas within the plain and ordinary meaning of the term. All of [the appellant’s] witnesses understood this.

[34] In concluding that the exhaust gas exclusion applied, the motion judge rejected on two grounds the appellant’s argument that the discharge of carbon monoxide into the greenhouse was also “pollution” as defined under the Policy.

[35] First, he distinguished two decisions in which this court interpreted a pollution exclusion in the context of an insurer’s duty to defend third party claims under commercial general liability policies, relied upon by the appellant: Zurich Insurance Co. v. 686234 Ontario Ltd. (2002), 62 O.R. (3d) 447 (C.A.), leave to appeal refused, [2003] S.C.C.A. No. 33; ING Insurance Company of Canada v. Miracle (Mohawk Imperial Sales and Mohawk Liquidate), 2011 ONCA 321, 105 O.R. (3d) 241. He concluded that “[t]he test in a ‘duty to defend’ case is much broader than in a ‘coverage case’ such as this.”

[36] Second, he determined that the exceptions stipulated under the pollution exclusion did not exclude “exhaust gas”:

When reading the Policy as a whole, the “pollution” exclusion, which includes two exceptions, is easily read together with the “exhaust gas” exclusion, which has no exceptions: if the cause of the loss is both pollution and an exhaust gas, there is no coverage because of the clear, broad wording of the exhaust gas exclusion. Again, there is no ambiguity or need to resort to other tools of interpretation.

[37] The motion judge provided an alternate basis for denying coverage. If he were incorrect about the exhaust gas exclusion, he found, in the alternative, “that the malfunctions of the Boiler and the Monitor were ‘machinery breakdowns’ as defined in the Policy.” This put the loss outside the scope of coverage under the Policy because of the exception for machinery breakdown under the initial grant of coverage. He relied on the definition of “machinery breakdown” in the Policy, which I referenced above, and explained his reasoning as follows:

The Loss would only be covered if, under the standard crop and business interruption coverage, the “event” causing the damage to the property (i.e., the Boiler and the Monitor) is covered under the standard property coverage. That standard property coverage under the Policy applies “unless the loss or damages was caused by machinery breakdown.” As noted above, the malfunctions of the Boiler and the Monitor were caused by machinery breakdown. As a result, there is no coverage under the standard property coverage, and therefore no coverage under the standard crop and business interruption coverage under the Policy.

[38] The motion judge granted boomerang summary judgment to the respondent and dismissed the appellant’s action. Costs of the motion and action of $157,377.37 were awarded to the respondent.

C. Analysis

i. What is the effective cause or causes of the loss?

a. General principles: the required analytical approach to causation in insurance policy interpretation in series of events cases

[39] In insurance coverage disputes, the cause of the loss typically drives the coverage analysis. As Professor Erik S. Knutsen observed, this is because “[m]ost insurance policy coverage and exclusion clauses contain some causation element whereby a certain class of event must cause the loss in order for the clause to be triggered”: Halsbury’s Laws of Canada, “Insurance” (Toronto: LexisNexis Canada, 2023 Reissue), at HIN-13.

[40] The causation analysis becomes more difficult when multiple factors potentially operate at the same time to bring about a loss. In Craig Brown et al., Insurance Law in Canada (Toronto: Thomson Reuters Canada, 2025) (loose-leaf 2025-Rel. 5), the authors identify two types of scenarios in cases where multiple factors function to bring about the claimed loss. First, there are chain of causation cases, in which “the occurrence of a factor sets off a series of incidents eventually resulting in the loss which is the subject of the claim”: § 8:12. This triggering factor or another event in the chain is the effective cause of the loss. Second, there are concurrent cause cases in which “two or more independent causes contribute to the loss and where the loss cannot, as a practical matter, be apportioned among the various contributing factors”: § 8:12. In other words, in concurrent cause cases, there is more than one effective cause of the loss.

[41] As I shall explain in more detail below, this case involved a series of factors that potentially operated together to bring about the claimed loss. While the motion judge found that the “direct cause” of the loss was excessive levels of carbon monoxide, he also determined that the excessive levels of carbon monoxide “were created by the malfunctioning Boiler and went undetected by the malfunctioning Monitor.” In light of this determination, the motion judge was required to determine what was the effective cause or causes of the loss by employing the chain of causation or concurrent cause analytical framework.

[42] Causation in chain of causation and concurrent cause cases, in the context of the interpretation of insurance policies, is “to be decided on common sense principles and to be ascertained by determining what is in substance the cause” of an event: Boiler Inspection & Insurance Co. of Canada v. Sherwin‑Williams Co. of Canada Ltd., [1951] 3 D.L.R. 1 (J.C.P.C.), at p. 15 (emphasis added), aff’g 1949 60 (SCC), [1950] S.C.R. 187. In chain of causation or concurrent cause cases, the immediate cause of the loss is not necessarily the effective cause of the loss.

[43] Although courts have often adopted the shorthand of “proximate” cause to aid in the analysis, it is important to remember that “proximate”, in this context, does not mean “closest in time” to the loss; rather, “proximate” refers to the “effective cause of the loss”: 942325 Ontario Inc. v. Commonwealth Insurance Co. (2006), 81 O.R. (3d) 399 (C.A.), at para. 3, aff’g (2005), 75 O.R. (3d) 653 (S.C.). See also Sher-Bett Construction (Manitoba) Inc. v. The Co-Operators General Insurance Company, 2021 MBCA 10, 457 D.L.R. (4th) 111, at paras. 46-51, 75; Co-operative Fire & Cas Co. v. Saindon, [1976] 1 S.C.R. 735, at pp. 747-48; and Milashenko v. Co-operative Fire and Casualty Co. (1968), 1 D.L.R. (3d) 89 (Sask. C.A.), at p. 100, per Culliton C.J.S. (dissenting), rev’d (1970), 72 W.W.R. 228 (S.C.C.), for the reasons of Culliton C.J.S.

[44] Most helpful is the explanation of the effective cause of the loss, per Lord Shaw of Dunfermline, from Leyland Shipping Company v. Norwich Union Fire Insurance Society, [1918] A.C. 350 (H.L. (Eng.)), at p. 369, as follows:

What does "proximate" here mean? To treat proximate cause as if it was the cause which is proximate in time is, as I have said, out of the question. The cause which is truly proximate is that which is proximate in efficiency. That efficiency may have been preserved although other causes may meantime have sprung up which have yet not destroyed it, or truly impaired it, and it may culminate in a result of which it still remains the real efficient cause to which the event can be ascribed. [Emphasis added.]

This decision was cited with approval by the Privy Council in Canada Rice Mills Ltd. v. Union Marine & General Insurance. Co., [1941] 1 D.L.R. 1 (J.C.P.C.), at p. 11. As the Privy Council noted, at p. 11, citing to the reasons of Lords Finlay and Dunedin in Leyland Shipping Company, “causa proxima in insurance law does not necessarily mean the cause last in time but what is ‘in substance’ the cause, per Lord Finlay at p. 355, or the cause ‘to be determined by commonsense principles,’ per Lord Dunedin at p. 362.”

[45] Consistent with this approach, the following cases are useful in illustrating the coverage analytical principles that the motion judge should have considered in this case to determine the effective cause or causes of the loss.

[46] In Boiler Inspection & Insurance Co. of Canada, a chain of causation case, a tank exploded due to the mixing of certain substances, causing vapours to escape, which ignited and caused a fire. Although the policy specifically excluded losses from fire, the Privy Council held the insured was entitled to coverage, emphasizing that proximate cause is determined by common-sense principles and what is, in substance, the cause of the loss: at pp. 9, 15. Lord Porter concluded, at p. 15: “the cause, whether it be described as dominant or proximate or by any other of the numerous epithets which have been used, was the bursting open of the door of the tank.” In other words, the effective cause of the loss was not the most immediate event, namely, the fire, nor the event that originated the chain leading to the loss, namely, the mixing of vapours. Rather, the effective cause of the loss was the bursting open of the door of the tank.

[47] Shea v. Halifax Insurance Co., [1958] O.R. 458 (C.A.), is another chain of causation case. There, a company’s policy insured against damage “caused solely by fire”. In the course of delivering fuel oil, the company’s trailer was badly damaged by an explosion. The explosion resulted from the oil level in the tank dropping too low, which exposed a heating tube that caused the oil to distill and form a highly combustible mixture of oil vapour and oxygen. This in turn led to a flash fire inside the tank, which generated intense heat and pressure resulting in the explosion. Schroeder J.A. held that the cause of the loss was the fire, not the explosion. He reasoned that the cause of a loss need not be “the actual instrument of destruction” and concluded that “the loss [in that case] was directly attributable to fire in its popular and well understood sense”: at p. 470.

[48] In Derksen v. 539938 Ontario Ltd., 2001 SCC 72, [2001] 3 S.C.R. 398, the Supreme Court considered causation in situations of concurrent causes. In that case, one child was killed and three were seriously injured when a base plate, left unsecured on a tow bar attached to a supply truck during a worksite cleanup, flew off the tow bar and hit a school bus. Major J. agreed with the motion judge that the accident was the result of two concurrent causes: the failure to safely clean up the work site and the failure to ensure that the truck could be operated safely: at para. 32. He rejected the argument that the single dominant cause of the loss was driving the truck with an insecure load: at paras. 30-31.

[49] In 942325 Ontario Inc., a chain of causation case, a grocery store owner sustained significant losses of its perishable food inventory when a blackout led to a loss of refrigeration at its stores. At para. 3 of its reasons, this court held that the effective cause of the loss was the blackout, not the immediate cause of the lack of refrigeration resulting from the blackout.

[50] This court cited favourably to 942325 Ontario Inc. in Caneast Foods Limited v. Lombard General Insurance Company of Canada, 2008 ONCA 368, 91 O.R. (3d) 438, aff’g (2007), 86 O.R. (3d) 385 (S.C.). In that chain of causation case, an insured similarly sustained substantial spoilage of a large quantity of produce following a blackout. This court again concluded that the effective cause of the loss was not the immediate cause of the lack of refrigeration but the interruption to the power supply because of a regional blackout. Borins J.A. for the court reasoned that the insured’s “refrigeration … stopped because the power to it was cut off”: at para. 24.

[51] Finally, in O’Byrne v. Farmers’ Mutual Insurance Company (Lindsay), 2014 ONCA 543, 121 O.R. (3d) 387, aff’g 2012 ONSC 468, 10 C.C.L.I. (5th) 103, a chain of causation case, this court’s coverage analysis again centred on the effective cause of the loss. In that case, a tenant had inserted a piece of cardboard into the primary control of a furnace, presumably to keep the furnace in constant hot operation while she was away. Forcing the furnace to continuously run caused the ignition component to fail, resulting in oil being pumped but not burned. This led to a significant spill causing damage to the building. One of the reasons the insurer argued that coverage did not apply was the mechanical breakdown exclusion in the policy, which provided in part that there was no coverage for “loss or damage directly or indirectly caused by ... mechanical or electrical breakdown or derangement”. The insurer argued the failure of the ignition to reignite was one of the factors that caused or contributed to the loss and, relying on Derksen for the proposition that an exclusion can be worded so as to apply to a loss with multiple causes if only one of the causes is contemplated by the exclusion, asserted the mechanical breakdown exclusion ought to apply. Justice van Rensburg disagreed for the following reasons at para. 36:

In my view, the error in the appellant’s argument is in seeking to characterize the oil spill damage as a multi-causal loss. What occurred here was a chain of events set in motion by the tenant’s insertion of a piece of cardboard, which in turn bypassed the thermostat, which forced the furnace to run at an excessively high temperature, causing the ignition component to fail and oil to be pumped continuously without burning.

Stated otherwise, the effective cause of the loss in O’Byrne was the tenant’s insertion of a piece of cardboard, not the mechanical failure of the ignition component.

[52] In short, as these cases demonstrate, it is critical to determine the effective cause or causes of a loss in insurance coverage disputes. In chain of causation cases, a loss arises from a series of events where there is one effective cause, sometimes, though not necessarily, the first event in the chain. In concurrent cause cases, multiple effective causes operate to bring about the loss.

[53] With these governing principles in mind, I turn to the particulars of this case.

b. Principles applied

[54] The motion judge’s analytical error arose in his causation analysis. The motion judge found that the loss resulted from a series of events: the malfunctioning boiler that went undetected by the malfunctioning monitor that resulted in the excessive emissions of carbon monoxide that entered into the greenhouses through the flue-gas distribution system and, ultimately, the loss of the tomato plants. Yet he ignored his factual finding of a series of potentially causal events and erroneously focused on the most immediate-in-time cause of the loss, namely, the carbon monoxide poisoning of the tomato plants, rather than determining what in substance was the effective cause or causes of the loss. This approach represented an error in principle.

[55] In light of the motion judge’s finding of a series of potentially causal events, he should have ascertained whether the claimed loss was caused by one effective cause in the chain of causation culminating in the loss (such as in Boiler Inspection & Insurance Co. of Canada, Shea, 942325 Ontario Inc., Caneast Foods and O’Byrne), or by the operation of concurrent causes (such as in Derksen). This required him to analyze the malfunctions of the boiler and the monitor, as well as the reasons for these malfunctions, to determine whether any of these events, by themselves or in combination with the carbon monoxide poisoning, were the effective cause or causes of the claimed loss. The effective cause or causes of the loss will determine coverage.

[56] It is not appropriate for this court to determine the effective cause or causes of the loss and whether the loss was caused by one effective cause in a chain of causation or by concurrent causes operating together. There was competing evidence about the factual cause or causes of the claimed loss that must be considered with the correct analytical lens. Moreover, the fresh determination of the effective cause or causes of the loss will require a new consideration of whether the exception or exclusion clauses apply in this case.

[57] These constitute genuine issues requiring a trial that must be considered afresh on a complete evidentiary record.

[58] For the purposes of completing the correct analytical framework for the presiding judge and the parties at trial, I turn now to that analysis, indicate where the motion judge erred, and set out the genuine issues requiring a trial.

ii. How are exception and exclusion clauses in an insurance policy to be interpreted?

a. General principles

[59] I start this part of the analysis with the same interpretative principles referenced by the motion judge.

[60] As recently reaffirmed by the Supreme Court in Emond, per Rowe J., at para. 37: “Where the language of the insurance contract is unambiguous, effect should be given to that clear language, reading the contract as a whole”. If, however, the policy’s language is ambiguous, the court must employ other rules of contractual interpretation to resolve that ambiguity: Emond, at para. 48.

[61] The framework for determining whether there is coverage under an insurance policy generally requires the court to consider:

(i) Whether the insured has met the onus of establishing that the loss claimed falls within the initial grant of coverage;

(ii) If the insured meets that onus, whether the insurer can establish that one of the exclusions to coverage applies; and

(iii) If the insurer meets that onus, whether the insured can establish that an exception to the exclusion applies.

See Progressive Homes Ltd. v. Lombard General Insurance Co. of Canada, 2010 SCC 33, [2010] 2 S.C.R. 245, at paras. 28-29, 51; Ledcor, at para. 52; and Emond, at para. 33.

[62] In Sam’s Auto Wrecking Co. Ltd. (Wentworth Metal) v. Lombard General Insurance Company of Canada, 2013 ONCA 186, 114 O.R. (3d) 730, at para. 37, Laskin J.A. summarized the principles for interpreting insurance policies and exclusion clauses in particular:

The principles for interpreting insurance policies, and in particular exclusion clauses, are well-established. The policy should be interpreted to promote a reasonable commercial result; provisions granting coverage ought to be construed broadly; provisions excluding coverage ought to be construed narrowly; in the case of ambiguity, the interpretation most favourable to the insured should be adopted; and even a clear and unambiguous clause should not be given effect if to do so would nullify the coverage provided by the policy.

See also Ledcor, at para. 51; Emond, at paras. 51-66.

[63] This last principle – that “even a clear and unambiguous clause should not be given effect if to do so would nullify the coverage provided by the policy” – is referred to as the nullification of coverage doctrine and was also recently reaffirmed in Emond. As Rowe J. explained, even when the language of a provision is unambiguous, it should not be applied to the extent it would render coverage nugatory: Emond, at para. 66.

[64] These general principles do not expressly explain how to interpret exceptions in the initial grant of coverage, such as the machinery breakdown exception in the present case. In Munro, Brice & Co. v. War Risks Association, [1918] 2 K.B. 78 (Eng.), at p. 88, rev’d on other grounds, [1920] 3 K.B. 94 (Eng. C.A.), the English Court of King’s Bench provided the following guidance about the onus of proof for exceptions:

The rules now applicable for determining the burden of proof … may, I think, be stated thus: –

1. The plaintiff must prove such facts as bring him prima facie within the terms of the promise.

2. When the promise is qualified by exceptions, the question whether the plaintiff need prove facts which negative their application does not depend upon whether the exceptions are to be found in a separate clause or not. The question depends upon an entirely different consideration, namely, whether the exception is as wide as the promise, and thus qualifies the whole of the promise, or whether it merely excludes from the operation of the promise particular classes of cases which but for the exception would fall within it, leaving some part of the general scope of the promise unqualified. If so, it is sufficient for the plaintiff to bring himself prima facie within the terms of the promise, leaving it to the defendant to prove that, although prima facie within the terms, the plaintiff's case is in fact within the excluded exceptional class.

[65] This passage was cited with approval by the court in Luciani v. British America Ass’ce Co., [1931] 1 D.L.R. 166 (Ont. S.C.(A.D.)), at pp. 173-74, and more recently by Gordon G. Hilliker in his text Liability Insurance Law in Canada, 7th ed. (Toronto: LexisNexis Canada, 2020), at § 2.93. It has also been recently applied by the Court of Appeal for British Columbia in Honeywell International Inc. v. XL Insurance Company Ltd., 2024 BCCA 375, 502 D.L.R. (4th) 365, at paras. 23-25. This court, however, has not recently referenced this passage from Munro, nor considered the distinction drawn in that passage between a qualified promise of coverage where “the exception is as wide as the promise, and thus qualifies the whole promise”, and a promise of coverage containing exceptions that leave “some part of the general scope of the promise unqualified”, including in light of the current principles for interpreting insurance policies, such as the nullification of coverage doctrine referenced above in para. 63 of these reasons.

[66] With these principles in mind, I turn to the particulars of this case.

b. Principles applied

[67] After determining the cause of the loss, the motion judge started his coverage analysis by tacitly assuming the loss was covered under the initial grant of coverage before concluding that the exhaust gas exclusion applied, and the loss was therefore not covered. Alternatively, he determined that the loss was not covered under the initial grant of coverage because the machinery breakdown exception applied. This approach did not accord with the approach recommended by Progressive Homes and led to analytical error.

[68] The motion judge should have undertaken the following analysis, considering, at each stage, who bore the onus of proof. Rather than starting his initial analysis with the question of exclusions, he should have started his analysis of coverage by considering, as he did in his alternative analysis relating to the machinery breakdown exception, whether the claimed loss is covered under the initial grant of coverage. According to the policy, this question should have included consideration of the effective cause or causes of the claimed loss. He should then have turned to determine whether the exception under the initial grant of coverage and/or the exclusions under the Policy applied. Specifically, depending on the effective cause or causes of that loss, he should have considered whether the machinery breakdown exception in the initial grant of coverage applied. If it did not, he should have gone on to consider whether the exhaust gas, pollution or machinery breakdown exclusions from coverage, and any exceptions to those exclusions, applied.

i. The initial grant of coverage and the machinery breakdown exception

[69] Given the parties’ positions, the loss of the tomato plants would have been covered under the initial grant of coverage unless the machinery breakdown exception applied. The standard crop and business interruption coverage (which required that the damage be covered under standard property coverage) applied unless the loss or damages were caused by machinery breakdown.

[70] The motion judge adverted to the definition of “machinery breakdown” in the Policy, which reads as follows:

the accidental loss as a result of inherent vice and/or any other internal cause to any machine (including computers), instrument, appliance, equipment (including all electronic equipment attachments, accessories, fittings, cables, pipes and such associated property but excluding foundation and brickwork) during the period that this item is in use or installed and ready for use. [Emphasis in original.]

[71] He applied the term “inherent vice”, which is in turn defined to mean “hidden defect (or the very nature) of the insured property which in itself is the cause of (or contributes to) its deterioration, damage, or wastage”. The Policy does not define “other internal cause”.

[72] The motion judge concluded that the malfunctions of the boiler and the monitor were due to machinery breakdowns. He appeared to rely on the evidence that the boiler malfunctioned because of a displaced internal burner diffuser. However, he did not address what displaced the internal burner diffuser nor indicate the cause of the monitor malfunction.

[73] The difficulties with the motion judge’s determination are two-fold.

[74] First, the motion judge failed to address expressly and to apply who had the onus of proving that the machinery breakdown exception operated here. This was a significant analytical omission, especially on a motion for summary judgment where each party was required to put their evidentiary “best foot forward”. For example, if the respondent bore the onus to prove that the machinery exception operated here, it would have been required to put forward evidence demonstrating that the boiler and the monitor malfunctions were the result of machinery breakdown because of an inherent vice or internal cause.

[75] Second, and relatedly, the motion judge did not clearly find what caused the boiler and monitor to malfunction. His conclusion is cryptic: “As noted above, the malfunctions of the Boiler and the Monitor were caused by machinery breakdown. As a result, there is no coverage under the standard property coverage, and therefore no coverage under the standard crop and business interruption coverage under the Policy.” Regardless of who bore the onus, it was incumbent on the motion judge to make the requisite findings of fact as to what caused the boiler and monitor to malfunction. He failed to do so.

[76] There was evidence that the boiler malfunction resulted from a displaced internal burner diffuser. However, the motion judge did not make an explicit finding that the displacement of the internal burner diffuser was the cause of the boiler’s malfunction. His reference to what was “noted above” about both the boiler and the monitor is too obscure. Moreover, he did not examine or make any findings as to what caused the internal burner diffuser to displace and, importantly, whether the cause of the displacement was the result of an internal vice or cause. The latter step was necessary in the circumstances of this case given the undisputed evidence that the malfunctioning boiler was inspected and cleared to be in good working order mere weeks before the malfunction.

[77] With respect to the monitor, the motion judge failed to make any findings as to what caused it to malfunction. There was conflicting evidence on the issue that the motion judge was required to resolve. For example, while an expert witness, Ramtin Movassaghi, agreed that the malfunction may have resulted from the electrical components in the monitor being past their lifespan, he also indicated in his report that it may have occurred because of an electrical surge. This latter explanation was consistent with the testimony of Jeremy Wood, the President of Powerhouse, who inspected the monitor on March 17, 2021, and ruled out end-of-life failure in favour of electrical damage, likely caused by a power surge or spike. As Mr. Wood explained, a monitor typically shuts off and ceases to operate when it reaches the end of its lifespan, but the monitor was still lit up and would not reset. Whether or not Mr. Wood’s evidence was admissible opinion evidence also had to be determined.[^1]

[78] These are not issues that the panel can determine. The parties did not address on appeal the issue of which party bore the onus to prove that the machinery breakdown exception applied or did not apply to foreclose the initial grant of coverage. To be fair to the motion judge, it does not appear that the parties argued this issue before him. Nevertheless, this was a material issue that must be determined if the boiler or monitor is determined to be an effective cause of the loss. As it was not squarely addressed by the parties or the motion judge, it is a genuine issue requiring a trial. Moreover, findings about what factually caused the boiler and the monitor to fail are required for the determination of whether the exception applies. Therefore, the factual causes of the boiler and monitor malfunction are also genuine issues requiring a trial.

ii. The exhaust gas, pollution and machinery breakdown exclusions

[79] For completeness and guidance to the trial judge and the parties at trial, I turn to the exclusions at issue on this appeal. The only exclusions raised by the parties on the appeal are the exhaust gas, pollution and machinery breakdown exclusions.

[80] I earlier explained why the motion judge erred in his determination that carbon monoxide was the sole cause of the claimed loss for the purposes of determining coverage under the Policy. Depending on the result at trial, it may be that the effective cause or causes of the claimed loss are found to be the carbon monoxide, the boiler, and the monitor, some combination thereof, or something else entirely, including the reasons for the boiler and the monitor malfunctions.

[81] If the effective cause of the claimed loss arises in the context of a cascading chain of causation – where the boiler or monitor malfunction, or another previously unconsidered event, such as the cause of the boiler or monitor malfunction, is determined to be the effective cause of the claimed loss – the exhaust gas exclusion becomes immaterial because the carbon monoxide is not the effective cause of the loss.

[82] The exhaust gas exclusion may become relevant under the concurrent causes analysis if the trial judge finds that there are two or more effective causes of the loss that operate concurrently. For example, it may be found that the boiler, the monitor, and, perhaps also, the emitted carbon monoxide gas are independent causes that operated concurrently to produce the loss.

[83] If so, the court will have to determine whether the emitted carbon monoxide is an exhaust gas within the meaning of the exhaust gas exclusion and whether, as the appellant argues, the emitted carbon monoxide is also “pollution”, as defined under the Policy, and caused by an event that is covered under the Policy, such that the exception to the pollution exclusion would apply.

[84] The motion judge’s findings about the exhaust gas and pollution exclusions reveal analytical error and are not owed deference. The motion judge found that the wording of the exhaust gas exclusion was “clear, broad” and contained “no ambiguity” and that the pollution exclusion was “easily read together” with the exhaust gas exclusion. There are several analytical difficulties with the motion judge’s findings.

[85] First, the motion judge’s finding that the wording of the exhaust gas exclusion was unambiguous depended on a literal interpretation of the words – “exhaust gas” – and the evidence given by the appellant’s witnesses that the emitted carbon monoxide was an exhaust gas. If the meaning of exhaust gas was clear, there should have been no need to depend on the parties’ evidence as to its meaning.

[86] Moreover, the witnesses’ agreement that the emitted carbon monoxide gas was, factually, an exhaust gas does not determine the interpretation of “exhaust gas” within the meaning of the Policy. The motion judge failed to interpret the meaning of “exhaust gas” by considering the Policy as a whole and how it applied to the particular carbon monoxide emissions in this case, including with respect to how the carbon monoxide entered the greenhouse system. As earlier described, the carbon monoxide entered the greenhouses through the carbon dioxide flue-gas distribution system and not by ordinary exhaust venting into the greenhouses: what did “exhaust gas” mean in these circumstances? Moreover, not all of the carbon monoxide emissions were toxic: there were safe levels of carbon monoxide that entered into the greenhouses through the flue-gas distribution system. To which carbon monoxide emissions, if any, did “exhaust gas” refer: all of the emissions or only the toxic levels of emissions that destroyed the tomato plants? The motion judge had to determine these questions.

[87] Further, the motion judge failed to meaningfully consider the meaning of “exhaust gas” in the context of the definition of “pollution” and the pollution exclusion. Notably, his rejection of the Zurich and ING cases on the basis that they were duty to defend cases was incorrect. Duty to defend cases involve the construction of insurance policies to determine whether the pleaded claim falls within the grant of coverage under an insurance policy in accordance with the same interpretative principles to which the motion judge made reference: Progressive Homes, at paras. 19, 21-24.

[88] Moreover, the Zurich and ING cases are relevant to the coverage analysis in this case insofar as they survey relevant decisions from this court and other jurisdictions and discuss the historical context of the pollution exclusion. The interpretative value of these cases is doctrinal, not textual. Despite the difference in the policy wording, these cases remain instructive in the present case because they articulate a general judicial approach to interpreting pollution exclusions. This includes the treatment of ambiguity in exclusion clauses, the adoption of a narrow rather than a broad reading of exclusion clauses in favour of the insured and, specifically, the interpretation of pollution exclusions in the context of the overall risk allocation of the policy. As a result, they inform the analysis of the Policy in the present case.

[89] For example, in Zurich, this court found that the historical context of the exclusion suggested “its purpose is to bar coverage for damages arising from environmental pollution”: at para. 39. There was nothing in that case “to suggest that the [insured]’s regular business activities place[d] it in the category of an active industrial polluter of the natural environment”, as the insured “did not discharge or release carbon monoxide from its furnace as a manufacturer discharges effluent, overheated water, spent fuel and the like into the natural environment”: at para. 38. Therefore, a “reasonable policyholder would … have understood the clause to exclude coverage for damage caused by certain forms of industrial pollution, but not damages caused by the leakage of carbon monoxide from a faulty furnace”: at para. 39.

[90] It was a necessary step in the analysis that the motion judge consider the purpose of the pollution exclusion and whether the emitted carbon monoxide was a pollutant in the historical context of the exclusion. He also had to grapple with the purpose of the exhaust gas exclusion and explain why the pollution gas exclusion could be read easily beside it. In other words, the motion judge was required to analyze what the exhaust gas exclusion was meant to exclude, if not a pollutant, and what kind of exhaust gases were meant to be excluded other than pollutants. Finally, it was incumbent on the motion judge to consider whether the exhaust gas exclusion was meant to cover gases vented into the greenhouses that were produced as part of the boiler’s operations that malfunctioned (like the carbon monoxide leakage from a faulty furnace in Zurich), as opposed to exterior, industrial pollutants.

[91] These questions raise genuine issues requiring a trial.

[92] Further, even if the emitted carbon monoxide is determined to be a concurrent cause falling within the exhaust gas exclusion, this exclusion does not necessarily operate to exclude coverage.

[93] In Derksen, Major J. rejected the argument that if a loss is caused by concurrent causes – one covered by the policy and the other excluded by an exclusion clause – there is a presumption of no coverage. He reasoned that “[w]hether a particular exclusion clause actually ousts coverage in a given case is a matter of interpretation” and underscored “the well-established principle in Canadian jurisprudence that exclusion clauses in insurance policies are to be interpreted narrowly and generally in favour of the insured in case of ambiguity in the wording (contra proferentem)”: at paras. 40, 46. An insurer must therefore clearly specify that if a loss is caused by an excluded peril, all coverage is ousted despite the fact the loss is also caused by another covered peril: at para. 47.

[94] In Derksen, there were two exclusion clauses at issue, one which excluded coverage for injury and damage arising out of the use or operation of a motor vehicle and another which excluded coverage for injury and damage with respect to which a motor vehicle liability policy was in effect. Major J. concluded that the exclusion clauses were triggered, but only with respect to that portion of the loss that was attributable to the auto‑related cause: at paras. 55, 62. He underscored that “[w]hether an exclusion clause applies in a particular case of concurrent causes is a matter of interpretation” to be made in accordance with the general principles of interpretation of insurance policies: at para. 49. Ultimately, he held that the policy covered the portion of loss attributable to non‑auto‑related negligence: at para. 63.

[95] As the causation analysis must be considered afresh, if it is determined that this is a case of multi-causal loss, it will be necessary for the trial judge to interpret the Policy language anew with the view of determining whether the exhaust gas exclusion clause applies to bar all coverage. Specifically, in accordance with Derksen, if the respondent intended for all coverage to be ousted if a loss is caused by an excluded peril, namely exhaust gas, despite the fact that the loss is also caused by another covered peril, the respondent needed to clearly specify this in the Policy. The trial judge will have to determine whether the respondent did so. This is a genuine issue requiring a trial.

[96] Finally, the trial judge will have to consider whether the machinery breakdown exclusion has any application here. The machinery breakdown exclusion was not considered by the motion judge because of his conclusion that the machinery breakdown exception or the exhaust gas exclusion applied. As set out above, the machinery breakdown exclusion, which differs from the definition of “machinery breakdown” and the machinery breakdown exception, deploys the definition to three specific situations, each with its own conditions: “machinery breakdown on crop”, “machinery breakdown to moving work equipment” and “machinery breakdown due to standstill or maintenance” (emphasis in original). It is common ground that “machinery breakdown on crop” and “to moving work equipment” do not apply here because neither of those situations relates to the malfunctioning boiler or monitor. Whether “machinery breakdown due to standstill or maintenance” applies is a question of fact that this court is unable to make. As noted above, in its August 17, 2021 letter to the appellant, the respondent relied on the machinery breakdown exclusion due to lack of maintenance as a basis for denying coverage. In its factum on appeal, the appellant denies that the machinery breakdown exclusion due to standstill or maintenance applies. These are genuine issues to be resolved at trial.

D. Disposition

[97] I would allow the appeal and set aside the motion judge’s decision, including the costs order. As I earlier explained, it is not possible for us to determine the issue of coverage because of all the genuine issues requiring a trial which require a complete evidentiary record and factual findings that are absent here. I would remit the entire action for trial.

[98] As agreed, I would order the respondent to pay the appellant its appeal costs in the all-inclusive amount of $70,000.

[99] Given that there will be a trial, I would leave the question of the costs below to be determined by the trial judge.

Released: May 11, 2026 “M.T.”

“L.B. Roberts J.A.”

“I agree. M. Tulloch C.J.O.”

“I agree. J. George J.A.”

[^1]: The respondent submitted that Mr. Wood was not qualified to opine as to the cause of the monitor’s malfunction. Mr. Wood conceded that he was not on site when the monitor was damaged and that he is not an electrical engineer. However, he is an experienced tradesperson called as a fact witness and was qualified to give evidence on what he observed. The admissibility of and weight given to Mr. Wood’s evidence on these issues are genuine issues for trial.