Disruption and the Measured Mile: Proving Lost Productivity

Disruption is the claim everyone suffers and almost no one proves. It is not delay — the project may even finish on time — it is the loss of productivity within the work: crews achieving 60 units a week where they should achieve 100, because of congested work fronts, out-of-sequence releases, design churn, stop-start access. The money is real and often enormous. The proof discipline is the hardest in construction quantum, and the gap between the two is where disruption claims go to die.

Disruption is not delay

The SCL Delay and Disruption Protocol defines disruption as a disturbance, hindrance or interruption to normal working methods resulting in lower efficiency (Part B, para 18.1). Keep the distinction operational: delay claims compensate the time-related cost of a longer project; disruption claims compensate doing the same work less efficiently. They can coexist, but their costs must be segregated — the overlap is the first thing a respondent's expert attacks, and the double-recovery audit from our prolongation guide applies in full.

The legal frame is unchanged from every other claim: causation on the common-sense approach, proved from contemporaneous records. Disruption just raises the evidentiary bar, because productivity loss is invisible in a programme — it lives in the resource and output data.

The measured mile, properly done

The Protocol's stated preference, where records allow, is the measured mile: compare productivity actually achieved in an un-impacted (or least-impacted) period or area against productivity in the impacted one, on identical or like work. The difference, costed, is the disruption loss. Its strength is that it sidesteps the tender entirely — the contractor's own demonstrated performance is the baseline, so arguments about optimistic pricing fall away.

Its requirements are exactly where claims fail:

• Like-for-like work. The comparison must hold trade, work type and conditions constant, and exclude learning-curve periods at the start of an activity (Part B, para 18.16).
• A reliable baseline sample. Long enough to be representative, genuinely clean of impacts.
• Honest selection. The fatal temptation is cherry-picking. In Civil Mining & Construction Pty Ltd v Wiggins Island Coal Export Terminal Pty Ltd [2017] QSC 85, the contractor's commercial manager had priced an earlier measured-mile claim from "the best week on the project" rather than production over a period. The former project manager agreed in evidence that this inflated the claim ([53]). At trial the contractor's expert said he had abandoned a measured-mile analysis because the records did not support it ([244]). Both appeals were later dismissed ([2019] QCA 12). Choose a representative clean period, disclose the selection criteria, and test the result against other candidate baselines.

Where no genuinely un-impacted stretch exists — common on badly disrupted jobs — the method adapts (least-impacted baseline, adjusted) or yields to alternatives. Pretending a contaminated baseline is clean is worse than using a different method.

The alternatives hierarchy

The Protocol ranks the options roughly by robustness. Below the measured mile sit other productivity-based methods — earned value analysis (planned versus actual labour against quantified output), programme-based analyses, and work sampling. Below those, industry studies (published factors for trade stacking, overtime fatigue, out-of-sequence work), usable as cross-checks or where project data is thin, but explicitly to be treated with caution as theoretical (para 18.20). At the bottom, cost-based methods (estimated-versus-incurred and the total cost family), which provide the least robust support (para 18.21) and inherit every weakness of the global claims we cover separately: the unproven baseline and the buried self-inflicted costs.

The practical hierarchy for a claimant: measured mile if the records permit; earned value as the workhorse where output and hours were tracked; industry factors as corroboration, never foundation; total cost only as a last resort, pleaded honestly as such.

The records that make it possible

Every method above consumes the same inputs, and they cannot be reconstructed later: output by period and area (quantities installed, by week, by work front), hours by period and area (timesheets coded to activities and locations, not just to the project), and the impact record (the site diary capturing congestion, re-sequencing, stop-starts and their causes on the day they happened — with the reason, as our records guide specifies). A contractor that tracks units-per-crew-week by zone has a measured mile waiting to be run; one that tracks only project-level cost has a global claim waiting to fail.

What this means in practice

1. Set up productivity capture on day one of any at-risk trade. Output and hours by work front, weekly. The cost is minutes of engineering time; the absence is unrecoverable.
2. Flag disruption contemporaneously, not at claim time. Diary entries and notices identifying the disrupting events and the affected crews, as they occur — the causation half of the claim, and frequently a contractual notice obligation with a time bar attached.
3. Choose and disclose the baseline like an expert will be cross-examined on it — because one will. Representative, not best-case; criteria stated; sensitivity-tested against alternative baselines (CMC v WICET is the cautionary tale).
4. Cost the productivity gap transparently. Lost hours × actual labour and plant rates for the impacted crews, net of any gains, reconciled against the prolongation and variation claims so nothing is recovered twice.
5. Respondents: attack the baseline and the segregation. Was the clean period really clean? Is like compared with like? Are the claimant's own inefficiencies — rework, supervision gaps, late procurement — excluded? Disruption claims are rarely beaten on the existence of disruption; they are beaten on attribution.

Key takeaways

• Disruption is productivity loss within the work, distinct from delay — segregate the costs or lose both claims.
• The measured mile is the most accepted method where records allow: like-for-like work, a representative clean baseline, honest selection (CMC v WICET's "best week" concession is the standing warning).
• The methods hierarchy runs measured mile → earned value → industry factors (caution) → cost-based (last resort).
• The inputs — output, hours and impacts by period and work front — must be captured during delivery; no method can outrun missing records.

This article is general information only and is not legal advice. For advice on a specific contract or dispute, seek legal counsel or contact Sumit Consulting for commercial and claims advisory support.