Question 2 of 6: Litigation — delay claims, their contractual prevention, and the delay taxonomy

Question 2: Litigation — delay claims, their contractual prevention, and the delay taxonomy (20 marks)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

(a) Why delay claims arise, how contracts can suppress them, and how they are analysed

Delay claims dominate construction litigation because time is the one commodity a construction contract cannot replace once it is spent, and because almost every project distributes the control of time across parties who do not share an interest in it. The recurring causes fall into a small number of families, and it is worth being precise about them, because the family a delay belongs to usually determines who pays for it.

The largest single family is owner-driven change: design changes and extras issued after the work has been planned, late or defective drawings, late owner decisions on submittals and shop drawings, and late release of the site or of owner-supplied equipment and materials. These are damaging out of proportion to their nominal value because they arrive after the contractor has committed a sequence, a crew size and a supply chain to the original scope. A closely related family is design incompleteness — drawings and specifications that are internally inconsistent, that conflict between disciplines, or that are simply not developed enough to build from — which manifests as a stream of requests for information whose answers arrive too slowly to keep the work in front of the crews.

A second family is differing site conditions: rock where soil was indicated, unrecorded utilities, contamination, groundwater, unexpectedly poor bearing capacity, or archaeological and environmental finds. These are the classic subsurface risk, and their allocation is precisely what a differing-site-conditions clause exists to settle in advance. A third is external and force-majeure events — abnormal weather, labour disputes, permit and approval delays at municipal or provincial authorities, utility company scheduling, and supply-chain disruption — where neither party is at fault but someone must nevertheless bear the consequence.

A fourth family is contractor performance: under-resourcing, low productivity, defective work requiring rework, subcontractor default, poor coordination between trades, and unrealistic tendering in which the bid programme was never achievable. A fifth, and the one that generates the most acrimonious disputes, is acceleration and disruption: the contractor is directed, or believes it has been constructively directed, to recover time it did not lose, and the resulting overtime, shift work, crew stacking, trade congestion and out-of-sequence working destroy the productivity on which the bid was priced. Finally, a great many claims arise not from the delay itself but from administrative failure — no baseline schedule was ever accepted, progress was never updated contemporaneously, notice was not given, and by the time the parties disagree there is no factual record from which anyone can reconstruct what happened.

The contractual modifications that reduce these claims work in three ways: by allocating risk deliberately rather than by default, by forcing information to surface early, and by pricing the outcome in advance so there is nothing left to litigate. The most effective, in rough order of leverage, are as follows.

First, a clear and complete risk-allocation set: a differing-site-conditions clause that entitles the contractor to relief for genuinely unforeseeable subsurface conditions (which is cheaper for the owner than receiving bids padded with subsurface contingency), a force-majeure clause that names the events it covers, and a weather clause that defines “abnormal” against a stated historical baseline — typically the ten-year Environment and Climate Change Canada record for the site — rather than leaving it to argument. Second, a scheduling specification with teeth: a contractually accepted baseline CPM schedule submitted and agreed within a fixed period of award, mandatory monthly updates in native electronic format, a defined float-ownership regime (float belongs to the project, not to either party, so the first user consumes it), a limit on total-float suppression devices such as artificial constraints and preferential logic, and a requirement to submit a time-impact analysis with any claim for extension. Third, strict notice and record-keeping provisions: written notice of a delaying event within a short defined period, with the sanction for failure stated explicitly, plus a contemporaneous requirement for daily reports, manpower returns and photographic records. Notice provisions are the single cheapest claim-reduction device in a construction contract, because they force the parties to address a problem while it can still be mitigated.

Fourth, pre-agreed pricing of time: liquidated damages set as a genuine pre-estimate of the owner's loss (not a penalty, which is unenforceable in Canadian law), and — symmetrically — a stated daily rate for the contractor's extended overhead, which pre-empts the entire Hudson/Emden/Eichleay formula argument about head-office overhead recovery. Fifth, a tiered dispute-resolution ladder that resolves matters while the project is running: project-level negotiation, then a dispute review board or a standing project neutral, then adjudication, then arbitration. Statutory prompt-payment and adjudication regimes now do much of this work by force of law — Ontario's Construction Act and the federal Prompt Payment for Construction Work Act, with comparable provincial regimes following — by imposing interim binding determinations on a statutory timetable so that cash-flow disputes cannot be warehoused until completion. Sixth, and often overlooked, early contractor involvement through a construction-management or design-build delivery model removes a large share of the design-completeness family of claims by putting constructability advice into the design before the price is fixed.

The analyses required to validate and judge a claim divide into three questions that must be answered in order: was there an entitlement, what was the effect on time, and what was the effect on money.

Entitlement is a contractual and factual enquiry, not a scheduling one: does the contract allocate this risk to the owner, was notice given as required, and did the contractor mitigate? It is settled from the contract documents, the correspondence and the daily records, and no amount of schedule analysis rescues a claim that fails here.

Causation and time is where forensic schedule analysis enters. AACE International Recommended Practice 29R-03 classifies the methods along two axes — observational versus modelled, and retrospective versus prospective — and the practical choices are four. As-planned versus as-built simply compares the baseline with what happened; it is cheap and intuitive, but it demonstrates correlation rather than causation and carries little weight where delays overlap. Impacted as-planned inserts the alleged delay events into the baseline and re-runs the CPM; it is a purely theoretical exercise that ignores what actually happened on site and is generally the weakest method. Collapsed as-built (“but-for”) starts from the as-built schedule and extracts the events, asking what the duration would have been without them; it is evidentially grounded but requires a defensible as-built logic, which rarely exists. Time-impact analysis, performed in windows, is the method generally given most weight by courts and boards, and the approach the Society of Construction Law Delay and Disruption Protocol (2nd ed.) recommends for assessing extensions of time while the project is running: the project is divided into windows, the schedule is updated to the start of each window using contemporaneous progress, the delay events of that window are inserted, and the incremental effect on the completion date is measured. It respects the fact that the critical path moves during a project — the single most important thing about delay analysis and the thing the simpler methods get wrong.

Quantum then follows, and is a separate exercise. Prolongation costs are the time-related site overheads — supervision, site accommodation, plant standing, insurance and bonding — recovered for the period of compensable extension at cost, not at tendered rates. Head-office overhead and profit are recovered either at a contractual daily rate or, failing that, by formula. Disruption is quantified by a measured-mile analysis, comparing productivity in an unimpacted period of the same work with productivity in the impacted period, which is far more persuasive than the industry productivity-loss curves precisely because it uses the contractor's own performance as the benchmark. Acceleration costs cover overtime premium, additional crews and shift differentials, and must be supported by evidence of a direction, express or constructive. All of it must be traced to cost-coded records; a claim quantified top-down from the difference between tender and final cost — a total-cost claim — is treated with deep suspicion by Canadian courts because it silently transfers the contractor's own tendering and performance errors to the owner.

(b) The three delay dichotomies

Excusable versus non-excusable. The question is whether the contractor gets more time. An excusable delay is one caused by an event outside the contractor's control and not attributable to its fault — owner changes, differing site conditions, abnormal weather, force majeure — and it entitles the contractor to an extension of the contract time, which relieves it of liquidated damages for that period. A non-excusable delay is one within the contractor's control or arising from its own or its subcontractors' fault: under-manning, rework, late procurement, subcontractor default. It carries no extension, so the completion date stands and liquidated damages run. Two points are frequently missed. The delay must be shown to affect the critical path to be excusable at all — a delay absorbed by float delays nothing and earns nothing. And subcontractor delay is non-excusable as against the owner however blameless the general contractor may feel, because the general contractor bears its subcontractors' performance under the head contract; its remedy lies down the chain, not up it.

Compensable versus non-compensable. This dichotomy is a subdivision of the excusable one and asks whether the contractor also gets money. Every compensable delay is excusable, but not every excusable delay is compensable. A compensable delay is an excusable delay caused by the owner or by someone for whom the owner is responsible — changes, late drawings, late site access, owner-supplied equipment, interference by the owner's other contractors — and it carries both an extension of time and recovery of the associated prolongation and disruption costs. A non-compensable excusable delay is one caused by a neutral event that is nobody's fault: abnormal weather, strikes, epidemics, embargoes. Here the risk is shared — the contractor gets time but bears its own costs for the period, while the owner bears the loss of the use of the facility. This is the deliberate bargain in the standard forms, and CCDC 2 GC 6.5 draws exactly this line, granting an extension of time for the neutral events while reserving cost recovery for delays caused by the owner, the consultant or anyone employed by them (GC 6.5.1), and for a stop-work order issued by a court or public authority through no fault of the contractor (GC 6.5.2). The practical consequence for a candidate is that entitlement must always be tested twice: once for time, once for money.

Concurrent versus non-concurrent. A non-concurrent delay is a single delaying event operating alone on the critical path in a given period, and its consequences follow straightforwardly from the two tests above. Concurrent delay arises where two or more delays — typically one compensable owner delay and one non-excusable contractor delay — operate in the same period and each, independently, would have delayed completion. True concurrency in the strict sense (two delays starting and ending together, each critical) is rare; what is usually argued is the looser sense of two delays overlapping within the same analysis window. Concurrency matters because it decides who pays for a period of time that both parties caused. The prevailing approach in Canadian and English practice, and the one recommended by the SCL Protocol, is that concurrency defeats the money but not the time: the contractor is granted an extension of time (so liquidated damages do not run, on the principle that an owner cannot levy damages for a period it caused in part) but is denied prolongation costs (because it would have incurred those costs anyway on account of its own delay). This is often expressed as “time but no money”. Apportionment between the parties is possible in principle — and Scottish authority has allowed it — but is uncommon in Canada because it demands a factual precision that the records seldom support. Establishing or defeating concurrency is therefore a central objective of the windowed time-impact analysis described in part (a): only a window-by-window reconstruction of the moving critical path can show whether two delays were genuinely both driving the completion date at the same time, or whether one of them was sitting on float and only appeared to be concurrent.