Extension of Time Basics
Last reviewed 9 September 20262,248 words10 min read
How an EOT claim is assembled
🟩 In one line — An EOT claim is four things stapled together: a notice given in time, a delay event the contract puts on the Employer, a schedule analysis showing that event drove the critical path and by how many days, and the records that prove the first three. The planner owns the third and half of the fourth — and if the notice was late, the other three usually don't matter.
👤 Who this is for — Juniors who've been asked to "run the delay through P6" for the first time, and mid-level planners writing their first claim narrative. You should know what the critical path is, what a data date is, and how a monthly update works.
First, let's be honest about why this page exists
Most planners meet an EOT claim eighteen months into a project, when the completion date has drifted three months and the commercial manager says "we need an EOT for all of it." At that point the planner discovers that the baseline was never accepted, half the updates have no narrative, the delay events were never logged with dates, and nobody sent notices. The claim that follows is a single global figure with a long story attached, and the Engineer rejects it in two pages.
The alternative is unglamorous: a delay log kept from day one, a notice sent every time something happens, a fragnet run through the current update while people still remember the facts. An EOT is not written at the end. It's assembled from things that were done at the time. This page is about what those things are.
🟨 The standard — what "good" looks like
| Reference | What it says, in planner's terms |
|---|---|
| FIDIC 1999 Cl 8.4 | Grounds for EOT: Variations, causes listed in the contract, exceptional climate, unforeseeable shortages caused by epidemic or government, delay by the Employer or its people. Entitlement only if completion "is or will be delayed" — the causal link. |
| FIDIC 1999 Cl 20.1 | Notice within 28 days of becoming aware (or should have become aware), or you lose the claim. Fully detailed claim within 42 days. Contemporary records required. Engineer responds within 42 days. |
| FIDIC 2017 Cl 8.5 / Cl 20.2 | Same grounds, with a concurrency sentence added at the end of 8.5. Notice 28 days (20.2.1), fully detailed claim 84 days (20.2.4), Engineer's agreement or determination 42 days under Cl 3.7 (20.2.5). Time bar is express and the Engineer must state it. |
| FIDIC 1999 Cl 3.5 / 2017 Cl 3.7 | The Engineer determines, "fairly" (1999) or after attempting agreement (2017). |
| NEC3/NEC4 ECC Cl 60.1 | The list of compensation events — the Employer-risk events. Cl 61.3: Contractor notifies within 8 weeks or loses the right, unless the Project Manager should have notified it. Cl 62: quotation, including a revised programme. NEC3 Cl 63.3 / NEC4 Cl 63.5: delay assessed as the delay to planned Completion shown on the Accepted Programme. |
| NEC3 Cl 16 / NEC4 Cl 15 | Early warning — not the notice, but failing to give one can reduce the assessment (Cl 61.5, 63.5 NEC3 / 63.7 NEC4). |
| SCL Protocol 2nd ed. | Core Principle 2: the purpose of an EOT is to relieve the Contractor of delay damages, not to give it money. Core Principle 3: follow the procedure. Core Principle 4: don't wait and see — assess contemporaneously. Core Principle 5: Engineer should assess within a reasonable time. Core Principle 6: entitlement follows critical delay only. |
| AACE RP 29R-03 | Taxonomy of analysis methods; the choice of method should suit the records available. |
🟢 Rule to remember: notice, entitlement, critical delay, records — in that order, and the first one has a clock on it.
How it actually works
The four legs of the claim
| Leg | Question it answers | Who owns it | Where it comes from |
|---|---|---|---|
| Notice | Did we tell them, in writing, within the time allowed? | Contracts / commercial | Letter register; see notice-requirements-and-records |
| Entitlement | Is this event the Employer's risk under the contract? | Contracts / commercial | The clause list (FIDIC 8.4/8.5, NEC 60.1, particular conditions) |
| Causation and quantum | Did it delay completion, and by how many days? | Planner | Schedule analysis — fragnet, TIA or windows, against the accepted programme |
| Records | Can we prove the dates, the sequence, the resources? | Everyone; planner for schedule records | Daily reports, updates, correspondence, photos |
The planner's leg, step by step
- Log the event the day it happens. A delay register: event ID, description, date started, date ended (or "ongoing"), affected activities by ID, notice reference, source document. A planner who keeps this log is worth more to the claim than any consultant hired later.
- Identify the affected activities in the current programme. Not the baseline — the last accepted update before the event. That's the programme that shows what the Contractor was actually going to do and where the float was.
- Build the fragnet. The delay modelled as a small network: a milestone for the event, activities for the extra or delayed work, links into the existing activities. See fragnets. Keep it small and honest — a fragnet that adds 40 days for a 3-day access problem is rejected on sight.
- Impact and measure. Insert the fragnet, reschedule, read the movement of the completion milestone. That's the prospective delay. If the event is over and the actual result is known, say so — the Engineer will look at the as-built anyway. See time-impact-analysis.
- Check what else was happening. If the Contractor was also late on that path, the concurrency question arrives. Don't hide it; deal with it. See concurrent-delay.
- Write it up. The narrative reviewers actually read: event, dates, affected activities, programme used, method, result, records referenced. Two pages for a simple event; the tables and the P6 comparison as appendices.
- Submit within the clock. FIDIC 1999: 42 days from awareness for the detailed claim; 2017: 84 days; NEC: quotation within 3 weeks of instruction (Cl 62.3). If the event is ongoing, interim submissions (FIDIC 20.1 / 20.2.6) — monthly, with the update.
What the schedule analysis has to show. Entitlement is to critical delay (SCL Core Principle 6). A ten-day delay to an activity with thirty days of float is not an EOT, however clearly it was the Employer's fault. The analysis must show the event consuming the float, reaching the critical path, and moving completion. If float ownership is contested — see float-ownership-and-contract — say which position you're taking and why.
Prospective versus retrospective. Assessed at the time, an EOT is a forecast: "this event will delay completion by 12 days." Assessed after, it's a measurement: "this event did delay completion by 9 days." FIDIC and the SCL Protocol both prefer the first — decide it while the facts are fresh, adjust later if the contract allows. NEC forces it: the quotation is based on the Accepted Programme at the time, and the assessment isn't revisited if the forecast turns out wrong (Cl 65.2 NEC3 / 66.3 NEC4 — verify).
📜 Time versus money
They are different claims, assessed differently. Time relieves you from liquidated damages (SCL Core Principle 2); prolongation cost is a separate entitlement, usually needing the delay to be both critical and not concurrent (Core Principles 12 and 14). The planner's analysis feeds both, but the money question needs the commercial team and, in the Gulf, often the lawyers. A programme that shows 30 days of critical delay does not automatically produce 30 days of prelims.
🟥 Where people go wrong
- Claiming the gap, not the events. Completion has moved 90 days, so the claim is for 90 days. Nobody has said which event caused which part. That is a global claim, and SCL Core Principle 17 tells you how the Engineer will treat it.
- Analysing against the baseline instead of the update. The delay in month 14 is run through the month-1 programme. Everything has changed since; the analysis is fiction, and the other side will build a better one from your own updates.
- Waiting until the end. SCL Core Principle 4 exists because of this. Analysed contemporaneously, a 12-day access delay is a 12-day EOT in month 5. Analysed at the end, it's one line in a 200-page claim fighting for space with the Contractor's own slippage.
- Fragnets that carry the Contractor's own problems. The Employer's late drawing is modelled with 20 days of "re-mobilisation" that was really the subcontractor's absence. The reviewer compares to daily reports and finds the crew was never there.
- No records for the dates in the fragnet. The claim says the site was handed over on 14 March. The Engineer's diary says 9 March. Without a signed handover note, the Engineer's date wins.
- Treating the EOT and the money as one document. The time analysis gets tangled in cost narrative, the entitlement argument gets buried in P6 screenshots, and neither reader finds what they need.
⚖️ When you're challenged
"The delay was only three days of access. Why are you claiming twelve?" "The access was three days late, but it pushed piling into the two-week rig demobilisation window we'd agreed with the piling contractor. The fragnet shows the three-day event, the rig standing down, the remobilisation — nine days of consequence on top of three of cause. Every date in it is from the piling daily reports, attached."
"You should have mitigated." "We did — the programme used is the one where we'd already re-sequenced Zone B ahead of Zone A to keep the rig busy. That's why the impact is twelve and not nineteen. Mitigation short of acceleration is what the Protocol asks for; we've done it and shown it."
"This event was months ago. Why are we only seeing the analysis now?" "The notice went in within the 28 days and it's referenced. The detailed particulars were held until the event closed out so the analysis could use actual dates — the contract allows interim submissions and we made them. If you'd prefer the prospective version from the month-8 update, I have that too, and it says fourteen days rather than twelve."
📄 Related pages
- Notice Requirements and Records — the clock and the paperwork; where claims are actually lost
- Time Impact Analysis — the method used above, step by step
- Concurrent Delay — what happens when both parties were late on the path
- Fragnets — building the event as a mini-network
- Choosing a Delay Method — when TIA isn't the right tool
- Float ownership and the contract — whose float the event consumed
- Prolongation costs basics — the money side, and why it's separate
- Template delay notice — the letter, ready to fill
✏️ Worked example
Substation and cable route, KSA, FIDIC 1999. In month 7, the Employer's utility hands over the last 400 m of cable trench route 18 calendar days late (handover 3 May instead of 15 April, Cl 2.1). Six-day site calendar.
| Item | Detail |
|---|---|
| Notice | Letter ref 214, dated 20 April — 5 days after the due date, within 28 |
| Entitlement | Cl 8.4(e) — delay by the Employer; Cl 2.1 right of access |
| Programme used | Update at data date 30 April (last accepted before handover) |
| Affected activity | TR-040 Trench excavation Ch 3+200 to 3+600, planned start 16 April, total float 6 working days at DD 30 April |
| Fragnet | Milestone "Access Ch 3+200–3+600 granted" constrained SOoA 3 May → FS to TR-040 |
| Result | TR-040 starts 4 May; float of 6 consumed; completion milestone moves from 22 Nov to 4 Dec — 10 working days |
| Records | Signed access handover 3 May; Engineer's site diary; TR-040 daily reports showing crew idle then redeployed to Ch 2+800 duct bank |
The planner's note on concurrency: cable laying (CB-060) was 4 working days behind at DD 30 April because of the Contractor's late cable delivery, but it sat on a path with 11 days of float and was not critical before or after the impact. Stated in the submission, with the float paths printed. The Engineer's determination under Cl 3.5: 10 days, no deduction.
Twelve days of EOT was the initial prospective figure from an earlier draft using a 20-day handover assumption; the actual handover was 18 days late and the final figure was 10. That correction was made by the Contractor in the detailed claim, which did more for the Engineer's confidence in the analysis than any argument.
📖 References
- FIDIC Conditions of Contract for Construction 1999, Cl 2.1, 3.5, 8.4, 20.1; FIDIC 2017, Cl 2.1, 3.7, 8.5, 20.2 (check the edition in your contract)
- NEC3 ECC Cl 16, 60.1, 61.3, 61.5, 62, 63.3, 65.2; NEC4 ECC Cl 15, 60.1, 61.3, 62, 63.5, 63.7, 66.3 (check the edition in your contract)
- Society of Construction Law, Delay and Disruption Protocol, 2nd ed. (2017), Core Principles 2, 3, 4, 5, 6, 12, 14, 17
- AACE International RP 29R-03, Forensic Schedule Analysis
- AACE International RP 52R-06, Time Impact Analysis — As Applied in Construction
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 extension of time basics. 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