Delay I: types of delay, float and concurrency

Delay claims ask two questions: was the project delayed, and who is responsible? Answering them requires a vocabulary for classifying delays, an understanding of float and the critical path, and a way of dealing with delays that happen at the same time. CON-210 taught the critical path method; CON-220 introduced the types of delay and the law on concurrency; CON-350 treated delay as a risk. This lesson builds the analytical foundation that lesson 5's schedule analysis methods depend on, using the Ashford project's schedule.

Classifying delay

Delays are classified on two axes. The first is whether the contractor is entitled to more time: an excusable delay is one for which the contract grants an extension, and a non-excusable delay is one the contractor must absorb. The second is whether the contractor is also entitled to money: a compensable delay entitles it to delay costs as well as time, and a non-compensable delay entitles it to time only.

The classification of a given event depends on the contract. Weather is usually excusable but non-compensable, and only if it is abnormal for the season; some contracts exclude it entirely. A concealed site condition is usually excusable and compensable under the site conditions clause. A design error is usually owner risk as between owner and contractor. Equipment breakdown is the contractor's risk. On Ashford, the four events classify as follows: contaminated fill, excusable and compensable under the concealed conditions clause; truss redesign, excusable and compensable; crane breakdown, non-excusable; and abnormal rain in the second autumn, excusable but non-compensable.

Delay must affect completion

An event that delays an activity is not necessarily a delay to the project. A delay matters for an extension of time only if it delays the date the contractor could complete the work, which in practice means it affects the critical path, the longest chain of dependent activities through the schedule. An owner's late approval of paint colors, three weeks before painting starts on an activity with six weeks of float, delays nothing. The same late approval on an activity with no float delays completion day for day.

That makes the schedule the central evidence in every delay claim. The baseline schedule shows the plan, the updates show how the plan changed, and the as-built record shows what actually happened. A contractor that keeps a sound, logically linked baseline and updates it monthly with accurate progress has the raw material for a convincing delay analysis. A contractor whose schedule is a bar chart without logic, or whose updates were not done, is starting a delay claim with its strongest evidence missing.

Float: what it is and who owns it

Total float is the amount of time an activity can slip without delaying project completion. Free float is the amount it can slip without delaying any successor. Critical activities have zero total float in a schedule driven by the contract completion date. Float is calculated by the forward and backward passes of the critical path method.

# Forward and backward pass for a simplified Ashford schedule (working days, illustrative)
acts = {  # activity: (duration, predecessors)
    "Mobilize":        (10, []),
    "Excavation":      (25, ["Mobilize"]),
    "Foundations":     (35, ["Excavation"]),
    "Pool tanks":      (60, ["Foundations"]),
    "Change room blk": (70, ["Foundations"]),
    "Steel and roof":  (75, ["Foundations"]),
    "Envelope":        (55, ["Steel and roof"]),
    "Mechanical":      (90, ["Change room blk", "Envelope"]),
    "Pool finishes":   (45, ["Pool tanks", "Envelope"]),
    "Interior fit-out":(60, ["Mechanical"]),
    "Commission":      (25, ["Interior fit-out", "Pool finishes"]),
}
ES, EF = {}, {}
for a in acts:                      # activities are listed in logical order
    d, preds = acts[a]
    ES[a] = max((EF[p] for p in preds), default=0)
    EF[a] = ES[a] + d
finish = max(EF.values())
LS, LF = {}, {}
for a in reversed(list(acts)):
    succs = [s for s in acts if a in acts[s][1]]
    LF[a] = min((LS[s] for s in succs), default=finish)
    LS[a] = LF[a] - acts[a][0]
print(f"{'Activity':<17}{'ES':>4}{'EF':>5}{'LS':>5}{'LF':>5}{'TF':>4}")
for a in acts:
    tf = LS[a] - ES[a]
    print(f"{a:<17}{ES[a]:>4}{EF[a]:>5}{LS[a]:>5}{LF[a]:>5}{tf:>4}{'  critical' if tf == 0 else ''}")
print("Project duration:", finish, "working days")

Activity           ES   EF   LS   LF  TF
Mobilize            0   10    0   10   0  critical
Excavation         10   35   10   35   0  critical
Foundations        35   70   35   70   0  critical
Pool tanks         70  130  245  305 175
Change room blk    70  140  130  200  60
Steel and roof     70  145   70  145   0  critical
Envelope          145  200  145  200   0  critical
Mechanical        200  290  200  290   0  critical
Pool finishes     200  245  305  350 105
Interior fit-out  290  350  290  350   0  critical
Commission        350  375  350  375   0  critical
Project duration: 375 working days

In this simplified schedule the critical path runs through mobilization, excavation, foundations, steel and roof, envelope, mechanical, interior fit-out and commissioning. The pool tanks and pool finishes have float, as does the change room block, because the mechanical work cannot start until the envelope closes in. A delay to the steel and roof flows straight to completion; a delay to the pool tanks of less than their float does not.

Who owns float matters when both parties use it. If the owner delays an activity that had float, and the contractor later delays the same path and pushes it critical, which party's delay caused the late completion? Contracts take three approaches.

Many specifications for public projects say that float is a shared project resource and that the contractor will receive an extension only if a delay pushes completion beyond the contract date. Contractors sometimes respond by building time contingency into activity durations rather than showing it as float, which hides the contingency and makes later analysis harder. CON-210 discussed schedule specifications; for claims, the point is to know what the contract says about float before arguing about it.

Pacing and the 'real' critical path

Not every slowdown is a delay. When one party causes a delay to the critical path, the other party may deliberately slow its own non-critical work to match, rather than hurrying to finish work that will only wait. This is called pacing. A contractor whose mechanical subcontractor slows its rough-in while the roof is redesigned is pacing, not delaying, if the rough-in could not proceed productively beyond a certain point anyway. Pacing is a legitimate response, but it must be shown to be a deliberate choice in reaction to the other party's delay, ideally documented at the time, or it will look like a concurrent contractor delay.

Schedules can also mislead. A baseline may show a critical path that was never realistic, because durations were optimistic or logic was missing. Analysts therefore look at both the planned critical path and the path that actually controlled progress, as shown by the updates and the as-built record. Lesson 5 discusses how different analysis methods handle this.

Concurrent delay

Concurrent delay occurs when two or more delays, for which different parties are responsible, affect the critical path over the same period, so that either one alone would have delayed completion. It is the most argued topic in delay analysis, because each party says the other's delay was the real cause. On Ashford, the truss redesign (owner risk) and the crane breakdown (contractor risk) overlap for two weeks.

The law and the standard forms deal with concurrency in several ways, and the approach in a given contract and jurisdiction should be checked. A common position, reflected in many standard forms and in industry guidance such as the Society of Construction Law Delay and Disruption Protocol, is that where an excusable and a non-excusable delay are truly concurrent, the contractor is entitled to an extension of time for the period of concurrency, so that it is not liable for liquidated damages, but is not entitled to delay costs for that period, because it would have incurred them anyway. Courts in both countries have also applied apportionment, dividing time or money according to the relative causative effect of each delay.

# Apportioning the Ashford truss and crane delays (illustrative working days)
truss = (0, 30)    # owner-risk delay window on the critical path
crane = (12, 22)   # contractor-risk delay window on the same path
overlap = max(0, min(truss[1], crane[1]) - max(truss[0], crane[0]))
truss_only = (truss[1] - truss[0]) - overlap
crane_only = (crane[1] - crane[0]) - overlap
print(f"Concurrent days {overlap}, owner-only days {truss_only}, contractor-only days {crane_only}")
daily_cost = 3_700  # contractor's time-related site overhead per working day
# Approach A: time for all owner and concurrent days, money only for owner-only days
print(f"A  time but no money: extension {truss_only + overlap} days, costs ${truss_only * daily_cost:,.0f}")
# Approach B: 50/50 apportionment of the concurrent days
print(f"B  apportionment:     extension {truss_only + overlap / 2:.0f} days, costs ${(truss_only + overlap / 2) * daily_cost:,.0f}")

Concurrent days 10, owner-only days 20, contractor-only days 0
A  time but no money: extension 30 days, costs $74,000
B  apportionment:     extension 25 days, costs $92,500

The difference is real money. Under the time-but-no-money approach, Kestrel Ridge gets 30 days of extension and is protected from liquidated damages for the whole period, but recovers site overhead only for the 20 days when the owner's delay alone was driving completion. Under 50/50 apportionment, it gets 25 days and 25 days of overhead. Which applies depends on the contract, the jurisdiction and the evidence, and a well-drafted contract states the approach expressly.

True concurrency is rarer than parties claim. Often, on close examination, one delay was on the critical path and the other was not, or one started first and the other was a pacing response. The crane breakdown on Ashford may turn out not to be concurrent at all if, during the redesign, there were no trusses to lift: in that case the crane was not needed on the critical path, and its breakdown delayed nothing. Testing that question with the schedule and the records is what lesson 5's methods are for.

Liquidated damages and the extension of time

Most construction contracts provide for liquidated damages: a fixed daily amount the contractor pays the owner for each day of late completion, agreed in advance as a genuine pre-estimate of the owner's loss. An extension of time moves the completion date and so reduces the period for which liquidated damages apply. That is why time claims matter even when the contractor recovers no money: each day of extension can save a day of liquidated damages. On Ashford, the contract provides for liquidated damages of $6,500 per calendar day (illustrative), so a 30-working-day extension, roughly 42 calendar days, is worth about $273,000 in avoided damages before any delay costs are considered.

If the owner causes a delay and the contract has no mechanism to extend the time for it, the owner may lose the right to liquidated damages altogether under the 'prevention principle', leaving it to prove its actual losses. Standard forms avoid this by allowing extensions for owner-caused delay. CON-220 covered the law on liquidated damages and penalties; the point for claims is that the time and money halves of a delay claim interact, and both must be analyzed.