Concurrent Delay
Last reviewed 9 September 20262,252 words10 min read
The argument that decides most disputes
🟩 In one line — Concurrent delay is two delays — one the Employer's risk, one the Contractor's — each of which would on its own have delayed completion over the same period; the SCL Protocol's position is time but not money (the Contractor gets the EOT, not the prolongation cost for the overlap), FIDIC 2017 leaves it to the Particular Conditions, NEC handles it through the Accepted Programme, and the planner's job is not to argue the law but to prove which delay was actually driving the critical path, day by day.
👤 Who this is for — Mid-level and senior planners, and anyone about to sit in a claim meeting where the word "concurrent" is going to be used as a weapon. You should understand critical path, float, and how a TIA works.
First, let's be honest about why this page exists
Concurrency is the Employer's favourite word because it's cheap. Say "you were late too" and the burden shifts. Most of the time the Contractor's delay wasn't critical, wasn't in the same period, or wasn't a delay to completion at all — but if the Contractor's planner can't demonstrate that with the programme, the word sticks and the claim is halved.
The reverse also happens. A Contractor months behind on its own account gets an Employer delay in month 15 and claims the lot, hoping nobody notices the critical path ran through the Contractor's own late procurement. Both sides misuse the concept, and the lawyers argue definitions. The planner's contribution is factual: at each point in time, what was driving completion, and what was the float on the other path? Everything else follows from that.
🟨 The standard — what "good" looks like
| Reference | Position |
|---|---|
| SCL Protocol 2nd ed., Core Principle 10 | Where an Employer delay and a Contractor delay are concurrent, the Contractor's delay should not reduce the EOT. |
| SCL Core Principle 14 | The Contractor recovers prolongation cost only for the period in which the Employer delay alone caused the delay — so for the concurrent overlap, no money. |
| SCL Part B §10 | True concurrency (both events at the same time, each independently critical) is rare; more common is sequential or overlapping delays where one is already driving. The Protocol says the assessment should be of the delays' effects on the critical path as they occurred. |
| FIDIC 1999 Cl 8.4 | Silent on concurrency. |
| FIDIC 2017 Cl 8.5, final paragraph | If a delay caused by the Employer is concurrent with a Contractor delay, entitlement is assessed per the Special Provisions; if none, "as appropriate taking due regard of all relevant circumstances." The clause is a pointer, not an answer — look at the Particular Conditions. |
| NEC3 Cl 63.3 / NEC4 Cl 63.5 | No express concurrency clause. Delay is the delay to planned Completion on the Accepted Programme at the time of the event. If the Contractor's own delay had already pushed planned Completion, the CE is measured from that later position — the Contractor's delay is already in the base. Concurrency is largely absorbed by the mechanism. |
| English law position (check , cite carefully) | The prevailing approach (Henry Boot v Malmaison, adopted by SCL) is time-not-money. The Scottish apportionment approach (City Inn) has not been followed in England. North Midland v Cyden (2018) confirmed parties can allocate concurrency risk to the Contractor by an express clause. |
| Gulf jurisdictions | Civil-law systems; tribunals have discretion and may apportion. Local counsel needed — the planner's factual analysis is what they'll apportion from. |
| AACE RP 29R-03, §4.2 | Defines literal versus functional concurrency and lists the criteria: both delays must be critical, independent, and in the same analysis period. |
🟢 Rule to remember: two delays are only concurrent if each was critical on its own in the same window — prove criticality first, then let the lawyers argue about money.
How it actually works
The three tests, in order. Before anything is "concurrent", a delay has to pass all three. Most fail the first.
| Test | Question | How the planner answers it |
|---|---|---|
| Critical | Would this delay, on its own, have moved completion? | Float on the path at the time. Float > 0 and not consumed = not critical = not concurrent, whatever the Employer says. |
| Same period | Did the two delays overlap in time (true concurrency), or did one start after the other was already driving (sequential)? | Delay log dates against the update data dates. |
| Independent | Was one delay actually caused by the other? | If the Contractor's crew left because the Employer's access was late, that's consequence, not concurrency. |
Literal versus functional. AACE distinguishes literal concurrency (both delays happening on the same days) from functional concurrency (both delays in the same analysis window, both critical, even if not on identical days). Most practical analyses — and the SCL Protocol — work in windows, so functional is what you'll usually be assessing. Say which you're using.
The dominant cause fallacy. "The Employer's delay was bigger, so it's the dominant cause and all the delay is theirs." Tribunals have generally not accepted dominant-cause as a way around genuine concurrency. Don't build a claim on it. Build it on float.
Where the Contractor is already late. Employer delay arrives in month 15; the Contractor's own critical path was already 30 days behind. Under NEC the Accepted Programme already shows the later planned Completion, and the CE is measured from there — if the Employer event doesn't push that date, no delay. Under FIDIC the same logic applies through a properly statused base: the TIA base already carries the Contractor's 30 days; the Employer event adds what it adds beyond that. The Contractor doesn't get the 30 back, and shouldn't expect to.
Where the Employer is already driving. Contractor delay arrives on a path the Employer's delay had already made critical. That path now has float relative to the Employer-delayed completion; the Contractor's delay consumes float, not the completion date. Not concurrent.
The P6 tools that answer it. Multiple float paths (Tools → Schedule → Options → Advanced → Calculate multiple float paths; set the end activity as the completion milestone) show the first, second, third most critical chains and their float. Run it on the update before the event and the update after. If the Contractor's problem is on path 3 with 12 days of float, print that. See p6-multiple-float-paths.
The contemporaneous record beats the retrospective argument. If each month's narrative said "critical path runs through X; near-critical path Y at 8 days float", then when the Employer says "you were late on Y", the answer is already written and dated.
📜 Time and money — why they split
The logic of time-not-money: an EOT exists to stop the Employer taking liquidated damages for delay the Employer caused (SCL Core Principle 2). If the Employer caused delay, LDs for that period would be unjust, regardless of whether the Contractor was also late. But prolongation cost requires the Contractor to show the Employer's delay caused the cost — and if the Contractor would have been on site anyway because of its own delay, it hasn't. Hence EOT yes, money no, for the concurrent overlap. Where the contract has an express concurrency clause (increasingly common in Gulf Particular Conditions), it may remove even the time. Read it.
🟥 Where people go wrong
- Calling any Contractor slippage "concurrent". A late subcontractor on a path with 20 days of float is not concurrent with anything. The float printout ends the conversation; the planner who doesn't have it loses the point.
- Claiming from the baseline when already late. The Contractor was 30 days behind; the Employer event adds 10; the claim is for 40. The updates show the 30 was self-inflicted, and the claim's credibility is gone before the 10 is discussed.
- Treating sequential as concurrent. Employer delay drove completion in months 4–6; Contractor delay drove it in months 9–11. Two separate periods, two separate responsibilities, no overlap. Windows analysis sorts it out; a single end-to-end comparison muddles it.
- Ignoring the concurrency clause. Gulf Particular Conditions often say the Contractor gets no EOT for concurrent delay. If that clause exists, the entire approach changes, and the planner needs to know before the analysis starts.
- Arguing law instead of showing facts. The planner's report says "under the Malmaison approach we are entitled to…" and the Engineer's lawyer takes it apart. The planner's report should say "on 14 June the critical path ran through X (TF 0); Y had TF 11" and leave the approach to counsel.
- No monthly critical path record. The whole question turns on what was driving at the time. If the monthly narratives never said, it has to be reconstructed later — and reconstructions are disputed.
⚖️ When you're challenged
"You were late on the façade at the same time. That's concurrent." "The façade was 6 days behind at the May data date, on a path with 14 days of float — it's the third float path in the May update, printed here. It couldn't have delayed completion. The steel delivery you're responsible for was on the critical path with zero float. One was critical; one wasn't. That's not concurrent."
"Your programme was already 30 days late before our variation." "Agreed, and the analysis doesn't claim those 30 days. The variation was impacted into the March update, which already shows the later completion. It moved that date by a further 12 days. We're claiming 12."
"The contract says no EOT for concurrent delay, so this is zero." "That clause applies where both delays were critical over the same period. Our position is that the Contractor delay wasn't critical — here's the float. If the Engineer reads the facts differently, the clause bites; but the facts have to be established first, and they're in the updates."
📄 Related pages
- P6 Multiple Float Paths — the setting that produces the evidence
- Float ownership and the contract — whose float the Contractor's delay consumed
- Near-Critical Paths — the chains that become critical next month, and why you track them
- Time Impact Analysis — the concurrency check inside every TIA
- Windows analysis — separating sequential delays into their periods
- Prolongation costs basics — why the money splits from the time
- Writing a progress narrative — recording the critical path monthly so the record exists
- Scl protocol summary — Core Principles 10 and 14 in context
✏️ Worked example
Hotel, Ras Al Khaimah, FIDIC 1999 with Particular Conditions silent on concurrency. Two delays in the same quarter.
| Delay | Owner | Period | Affected path |
|---|---|---|---|
| D-07: Employer's late approval of kitchen equipment layout | Employer | 1 Feb – 21 Mar (7 weeks) | Kitchen MEP second fix → kitchen fit-out → F&B commissioning |
| C-03: Contractor's late mobilisation of joinery subcontractor | Contractor | 15 Feb – 28 Mar (6 weeks) | Guestroom joinery → guestroom finishes → snagging |
Both look like six-to-seven-week delays in the same window. Everyone in the meeting says "concurrent." The planner ran multiple float paths on the January and March updates:
| Update | Path 1 (TF) | Path 2 (TF) | Path 3 (TF) |
|---|---|---|---|
| DD 31 Jan | Kitchen equipment approval → kitchen MEP → F&B commissioning (0) | Guestroom joinery → finishes → snagging (+18) | Façade → external works (+26) |
| DD 31 Mar | Kitchen path (0, completion +25 days vs Jan) | Guestroom path (+3) | Façade (+9) |
Reading it:
- At the end of January the kitchen path was critical and the guestroom path had 18 working days of float.
- The Employer's approval delay pushed completion 25 working days over the quarter.
- The Contractor's joinery delay consumed 15 of the guestroom path's 18 days of float. It never reached zero. It did not delay completion. Fails the first test — not concurrent.
Had the joinery been three weeks later still, the guestroom path would have gone to −3 while the kitchen path sat at 0: at that point both would be critical, and the last three weeks of the quarter would be genuinely concurrent — time yes, money no for that overlap, under the Protocol approach.
The submission claimed 25 days EOT and prolongation for 25 days. The Engineer's determination: 25 days EOT, prolongation agreed for 25 days after the float paths were reviewed. The float printouts were four pages; the argument they replaced would have run for months.
📖 References
- Society of Construction Law, Delay and Disruption Protocol, 2nd ed. (2017), Core Principles 2, 10, 14; Part B §10
- AACE International RP 29R-03, Forensic Schedule Analysis, §4.2 (concurrency)
- FIDIC 1999 Cl 8.4; FIDIC 2017 Cl 8.5 final paragraph (check the edition in your contract)
- NEC3 ECC Cl 63.3; NEC4 ECC Cl 63.5 (check the edition in your contract)
- Oracle Primavera P6 Professional User Guide — Schedule Options, Multiple Float Paths
General guidance, not legal advice. Contract wording, editions and amendments differ on every project. Read your own contract and take professional advice before relying on anything here in a claim or a dispute.
From the field
Experience from working planners. Unreviewed — read it as experience, not guidance.
Add what you know about concurrent delay. What worked, what the consultant pushed back on, what you would do differently next time. A paragraph is plenty.
Contributors get their name and one link on the site — your own templates, course or consultancy. We take nothing and hold nothing.
Add your experience