What Liquidated Damages Actually Require, and How Schedule Evidence Defeats Them

Liquidated damages, usually shortened to LDs, are one of the most common clauses in a Canadian construction contract, and one of the most misunderstood. The owner sets a fixed daily amount the contractor pays for finishing late. On the surface it looks automatic. The job is late, so the money is owed.

It is not automatic. An LD clause has to clear a legal bar to be enforceable at all. Even a valid clause only bites if the owner can show the delay was the contractor's responsibility and actually drove the completion date. Both of those questions are settled in the schedule. This article walks through what an LD clause must satisfy under Canadian law, the defences that defeat or reduce an assessment, and how forensic schedule analysis proves each one. The focus is Ontario and Canada, and amounts are in Canadian dollars.

What a liquidated damages clause must satisfy to be enforceable

Canadian law draws a hard line between liquidated damages and a penalty. The distinction comes from the House of Lords decision in Dunlop Pneumatic Tyre Co. v. New Garage & Motor Co. (1915) and was received into Canadian law the same year in Canadian General Electric Co. v. Canadian Rubber Co. A clause that fixes the sum payable for late completion is enforceable as liquidated damages only if it was a genuine pre-estimate of the loss the owner expected from late completion. If the figure is extravagant and unconscionable next to the greatest loss that could conceivably follow from the breach, it is an unenforceable penalty.

Three points follow from that test. The clause is judged as at the date the contract was made, not with hindsight after the breach, so the question is what was reasonable to estimate when the deal was signed. The fact that the loss was hard to quantify in advance supports using an LD clause, it does not condemn one. And the label is not decisive. Calling a clause "liquidated damages" in the contract does not make it so. Courts look at substance.

Canadian courts have layered an oppression screen on top of the basic test. In H.F. Clarke Ltd. v. Thermidaire Corp. Ltd. (1976) the Supreme Court struck down a damages formula that produced a grossly excessive, punitive result. In Elsley v. J.G. Collins Insurance Agencies Ltd. (1978) the Court framed the power to strike a penalty as relief against oppression, with no role to play where there is no oppression. A stipulated sum is set aside only where it is so extravagant relative to the expected loss as to be oppressive to the party paying it. The genuine pre-estimate remains the yardstick.

This is still the law in Ontario. In Haas v. Viscardi (2019) the Court of Appeal confirmed the genuine pre-estimate rule remains in force and declined an invitation to replace it with a pure unconscionability analysis. An earlier decision, Peachtree II Associates (2005), suggested moving in that direction but expressly left the change for another day. Canada has also not adopted the United Kingdom's newer "legitimate interest" test from Cavendish Square v. Makdessi, so the pre-estimate standard governs here.

How LD clauses actually operate in construction

Construction LDs almost always run as a per-diem amount, a daily figure that accrues from the contractual completion date until completion is reached, rather than as a single lump sum. The daily rate is itself reviewable. If it has no sensible relationship to the owner's anticipated daily loss, things like financing costs, lost revenue, and extended supervision and consultant costs, it is vulnerable.

The milestone that stops the clock matters, and it has moved. Under older CCDC forms, Contract Time ran to Substantial Performance. Under CCDC 2 (2020), the relevant milestone became Ready-for-Takeover, which requires Substantial Performance plus further items such as occupancy compliance, final cleaning, as-built drawings and operating manuals, completed start-up testing, and demonstration and training. Ready-for-Takeover lands later than Substantial Performance, which pushes the LD-stopping date out and increases the contractor's exposure compared with the older form.

The CCDC core forms do not fix the LD rate or any cap. Those are inserted through the Agreement, the Definitions, and the Supplementary Conditions, which is where the real negotiation happens. A total cap on LD exposure, often expressed as a percentage of contract value, is a key risk control. It matters most on small or thin-margin jobs, where a handful of delay days can swallow the entire fee.

A valid LD clause usually does two things at once. It caps the owner's recovery, and it is the exclusive remedy for the delay it covers. The owner recovers the agreed rate and no more, even if its real loss was higher. That is the double edge contractors should remember, because the same clause also fixes their exposure. The cap is a default rather than an iron rule. Where the contract preserves all rights and remedies provided by law and by the agreement, an owner may be able to claim above the LD sum, as in 1252662 Ontario Inc. v. Swisslog Logistics, Inc. (2016). Where the clause is valid, the owner does not have to prove the amount of its actual loss. It recovers the stipulated sum on proof of a qualifying, contractor-caused late completion. Removing the burden of proving quantum is the whole point of an LD clause. It is also exactly what the schedule defence attacks, because the owner must still prove the delay was the contractor's and on the critical path. One more practical point: an owner that takes possession or certifies completion may be found to have waived its right to delay damages, so the timing of takeover or early occupancy can extinguish an LD claim.

The defences that defeat or reduce an LD assessment

Attack the clause. The contractor can argue the figure was not a genuine pre-estimate and, where the rate dwarfs any conceivable loss, that it is oppressive. There is an important caveat. Striking the clause does not erase liability. It pushes the owner back to proving its actual delay damages, which is usually harder and often smaller.

The prevention principle. An owner that causes or contributes to the delay and fails to extend the completion date cannot enforce the LD clause. In Perini Pacific Ltd. v. Greater Vancouver Sewerage and Drainage District (1967) the owner failed to deliver engines in good repair for the works, an item on the contractor's critical path, and did not extend the date. The Supreme Court held the owner could not levy LDs, even though part of the delay was the contractor's own fault. Serious interference is enough. Total prevention is not required.

Time at large. When the prevention principle applies and the contract has no working mechanism to absorb the owner-caused delay, there is no longer a fixed completion date to count days from. Time goes "at large", the contractor is held only to finishing within a reasonable time, and the LD clause loses the anchor it needs. This does not mean the owner walks away with nothing. It can still pursue general delay damages it is able to prove.

The extension-of-time clause is the hinge. A well-drafted EOT clause that does reach owner-caused delay usually preserves both the completion date and the LD entitlement, because it extends the date to absorb the delay. Time goes at large, and LDs fall away, mainly where the clause does not cover the owner's act, or where the owner or its certifier failed to operate it. The contractor's side of this is real too. Notice requirements and conditions precedent have to be met, and a court will check whether the contractor actually triggered the mechanism.

Excusable, compensable, or non-excusable. Delay analysis sorts disputed days into three buckets. Excusable delay earns a time extension and relief from LDs. Excusable and compensable delay earns time and money. Non-excusable delay is the contractor's own risk and is what LDs are for. The contractor's first job in an LD fight is to move disputed days into the excusable bucket. AACE International Recommended Practice 29R-03 sets out these definitions.

Concurrency. Where owner-caused critical delay overlaps contractor-caused critical delay, Canadian and SCL practice generally treats the concurrent period as excusable but not compensable. The contractor gets time relief, which defeats LDs for that window, but no delay money for the overlap. In Schindler Elevator Corporation v. Walsh Construction Company of Canada (2021) the Ontario Superior Court adopted a practical definition of concurrent delay, overlapping delays to co-critical and co-controlling activities, and rejected the rigid idea that the delays must start and end at the same moment. Apportionment starts by identifying the actual critical path, and each party answers only for its own share.

Mitigation and pacing. A contractor resisting LDs is expected to have taken reasonable steps to mitigate the delay. Pacing, deliberately slowing work to match a controlling owner delay so long as it does not affect completion, is a recognized and legitimate response, not culpable delay. The line between the two turns on whether the paced activity was on the critical path at the time.

How forensic schedule evidence proves each defence

Every defence above runs through one question. Was the disputed delay on the critical path. In Walsh Construction v. Toronto Transit Commission (2024) the court defined the critical path as the longest path, the connected sequence that sets the minimum duration for the project, and held the contractor must prove the delay sat on it. Delay that did not drive completion does not count. With that spine established, recognized forensic methods carry the rest. The method numbers below follow the AACE 29R-03 framework.

• As-planned versus as-built (MIP 3.1 and 3.2). Overlay the contract baseline against the as-built record to show where, when, and by whom the schedule diverged. It exposes owner-driven critical slippage and quantifies the real magnitude of the delay. That quantification also feeds the penalty challenge, because if the LD rate vastly exceeds the owner's demonstrable as-built loss, the figure starts to look extravagant.
• Windows analysis (MIP 3.3 and 3.4). Divide the project into periods at the schedule data dates and read the as-built critical path in each window to see whose delay actually controlled completion. Because it uses the contemporaneous schedules, it reveals genuine concurrency, where owner and contractor delay were co-critical, and supports the excusable-versus-compensable split window by window.
• Time Impact Analysis (MIP 3.6 and 3.7). Insert an owner-delay fragnet into the schedule that was current when the event happened, then re-run the forward pass to measure how far the event pushed the completion date. If the modelled completion moves past the LD date because of the owner's event, the contractor has shown the delay was owner-caused and that an extension was due. This is the prospective, cause-first method that proves prevention and a failure to grant an EOT.
• Collapsed as-built (MIP 3.8). Start from the as-built schedule and remove the owner-caused activities to compute a but-for completion date. If taking out the owner events pulls completion back to or before the contractual date, it proves the owner caused the LD exposure. Run as a second, independent method alongside windows or TIA, it gives the dual-method corroboration that answers any warning against leaning on a single contested methodology.
• Apportioning concurrency the right way. Map each party's critical-path delay on one common timeline, flag only the genuinely overlapping co-critical days as concurrent, and assign the rest. This is the defensible alternative to the meaningless exercise of adding up every delay regardless of whether it was on the critical path.

Walsh v. TTC as a concrete anchor

Walsh Construction Company of Canada v. Toronto Transit Commission (2024) is the clearest recent Ontario illustration of how this plays out. On the Pioneer Village subway station contract, originally the Steeles West station, the TTC counterclaimed roughly $22 million in liquidated damages for missed milestones. The counterclaim was dismissed in full, and the contractor recovered a substantial delay award.

What carried it was schedule evidence. The contractor's delay expert ran a critical-path delay analysis in Primavera P6 across roughly 780 owner-related conditions, identified the critical path as the longest path, and tested for concurrency. The decisive contrast was on the other side of the table. The TTC offered a critique of the expert's methodology but did not produce its own independent delay analysis. A party that only attacks the other side's schedule analysis, without performing its own, is in a weak position when the court has to decide what actually drove completion. That is the lesson worth carrying into any LD dispute, from either chair.

Practical takeaways

For owners. Protect the entitlement with a genuine, proportionate, well-documented LD rate, an EOT mechanism that actually works and is actually operated, and clear remedy and cap wording. The strongest LD assessment is one built on a clean schedule record and properly granted extensions, not one imposed over the top of owner-caused delay.

For contractors. The defences are only as good as the records behind them. A credible baseline, contemporaneous schedule updates at every cycle, and timely EOT notices are the raw material every defence here depends on. The time to protect them is during the job, not after the assessment lands.

Keep expectations honest. A functioning EOT clause usually preserves LDs. Unconscionability rarely succeeds between sophisticated commercial parties, because the current test from Uber Technologies Inc. v. Heller (2020) requires both an inequality of bargaining power and a resulting improvident bargain. And losing the LD clause does not end delay liability. It shifts the owner to proving its actual damages.

Whether you are an owner testing an assessment or a contractor facing one, the answer lives in the schedule, and the analysis has to be done with the contemporaneous records and a recognized method. That is where forensic scheduling earns its place.