Justify a Duration
Last reviewed 9 September 20262,614 words12 min read
What to show when someone challenges a number
🟩 In one line — A justified duration fits on one line: quantity, rate and its source, crew and hours per day, and the conditions assumed — plus a comparison to something real, such as a previous floor, a benchmark range or the subcontractor's own signed programme. If you can't produce those, the number is an opinion, and opinions lose review meetings and claims.
👤 Who this is for — Juniors about to sit in their first schedule review, mid-level planners defending a baseline against a client's engineer, and seniors who will one day be asked, under cross-examination, why the baseline said what it said. You should know the duration formula (crew-and-manhour-formulas) and what a schedule basis memorandum is.
First, let's be honest about why this page exists
Reviewers challenge durations for two opposite reasons. The client's engineer says 18 days is padded. The site manager says 18 days is impossible. The planner, caught between them, says "it's based on experience," and both sides hear "I made it up." Then one of two things happens: the planner shortens it in the meeting to make the noise stop, or the client rejects the baseline and three months pass without one.
The deeper cost comes later. A baseline accepted with fiction in it becomes the yardstick for every delay analysis on the project. If the duration was never real, the delay measured against it isn't real either, and the other side's analyst will find that out. What you write down today as the basis of a duration is the most valuable document you'll produce for the eventual claim, and almost nobody produces it.
🟨 The standard — what "good" looks like
AACE RP 32R-04 (Determining Activity Durations) expects durations to come from quantities, productivity and resources, with the basis documented so it can be reviewed.
GAO Schedule Assessment Guide, Best Practice 4 asks whether durations are realistic, based on the planned resources, and supported by a documented basis; it also flags durations that are unusually long as needing explanation. Best Practice 3 ties durations to assigned resources.
DCMA 14-Point, Metric 8 (High Duration) flags incomplete activities over 44 working days, with a 5% threshold. A long duration isn't wrong, but it needs a reason on file.
PMI Practice Standard for Scheduling places duration estimating within the schedule model's documented assumptions and expects those assumptions to travel with the model.
NEC3/NEC4 Cl 31.2 requires the programme to show resources against operations and to identify time risk allowances — so padding is meant to be visible, not hidden in durations. Cl 31.3 limits the Project Manager's reasons for not accepting a programme to four, including that it is not practicable and that it does not represent the Contractor's plans realistically. A duration you can justify clears both.
FIDIC 1999 Cl 8.3 allows the Engineer to give notice that the programme does not comply with the Contract; FIDIC 2017 Cl 8.3 does the same with a wider list of required content and a supporting report. Neither edition gives the Engineer the right to set the Contractor's durations — means and methods remain the Contractor's.
SCL Protocol 2nd ed., Core Principle 1 calls for a properly prepared programme, accepted early, with method and resource information — the recommended companion is a method statement and resource narrative, which is exactly what makes durations defensible.
🟢 Rule to remember: a duration is justified when someone who has never met you could recompute it from what you wrote — and when it matches something that actually happened.
How it actually works
The duration basis line. For every activity that matters, one row:
| Field | Example |
|---|---|
| Activity | BLK-3B-020 Blockwork L3 Zone B |
| Quantity | 1,150 m² |
| Rate and source | 1.3 mh/m², site as-built L1–L2 (1.25–1.40 measured) |
| Crew × hours | 8 masons + 4 helpers, 9 h/day |
| Calendar | 6-day site, summer hours |
| Factors | None (enclosed, low level) |
| Computed | 13.8 days |
| Adopted | 18 days |
| Reason for difference | +2 lintels cast and cured in sequence, +2 MEP conduit chasing hold |
The last row is the one people skip and the one reviewers care about. The gap between computed and adopted is where judgement lives; make it visible.
Proportionate effort. You can't do this for 2,400 activities, and you shouldn't. Three tiers:
| Tier | Activities | Justification |
|---|---|---|
| Full basis line | Critical and near-critical (float under ~10 days), anything over 30 days, anything the reviewer has already asked about | All fields, plus a comparison |
| Rate reference | Everything else with a quantity | Rate and source in a UDF; crew in the resource assignment |
| None | Milestones, LOE, submittals with contractual review periods | The contract or the procedure is the basis |
Triangulate, don't just calculate. One method is an argument; two that agree are evidence.
- Bottom-up: the formula. Quantity × rate ÷ crew.
- Analogous: what did this trade do last time? Same subcontractor on the previous floor, the previous phase, the previous project. Pull it from the as-built P6 file with Schedule Comparison or from the daily reports.
- Benchmark: typical ranges for the type of work. A Gulf flat-slab floor cycle with table forms runs 5–8 working days; a full-height 200 mm block wall lays at 1.2–1.5 m lifts per day. Benchmarks bracket the answer; they don't give it.
- Commitment: the subcontractor's own programme, with their name on it. Useful, but check it against the other three before adopting it — see verifying-subcontractor-progress for why their numbers need reading.
Physical constraints beat manhours. When the challenge is "too long," the strongest justifications are usually not arithmetic. Curing time. Lift height per day in masonry. Pump output per hour. Hoist cycles per day. Inspection hold points with a 24-hour notice. Crane hook time shared across four trades. These don't shrink when you add men, and a reviewer with site experience will accept them immediately.
When the challenge is "too short," the same logic works in reverse. Show that the crew is available, the workface has room, the material is ordered, and the sequence has been done at that pace before. If any of those is missing, the reviewer is right and you should thank them.
Where the basis lives in P6. Three options, use at least one:
- User Defined Fields (Enterprise → User Defined Fields → Activities): "Quantity" (number), "Unit Rate" (number), "Rate Source" (text), "Duration Basis" (text). Exportable to Excel for the review meeting in one filter.
- Notebook topics (Activity Details → Notebook, topic "Duration Basis"): free text, prints on reports, travels with the XER.
- Steps (Activity Details → Steps) for the sub-tasks that explain a longer-than-formula duration.
Do not keep the basis only in a spreadsheet on your own laptop. When you leave, it leaves.
📜 Who owns the duration
The Contractor. Under FIDIC (both editions) the Engineer may notify that a programme doesn't comply with the Contract, and under NEC the Project Manager may decline acceptance for the reasons in Cl 31.3 — but neither gets to write your durations. If a reviewer says "make it 10 days," the correct response is to show the basis and ask which input they disagree with. If they can't say, the duration stands.
Under NEC4, the Accepted Programme is used to assess compensation events (Cl 63.5), so a duration you shortened under pressure becomes the yardstick against which your own delay is measured — you've pre-agreed a standard you can't meet. If you do agree to change a duration in a review, record why in the meeting minutes and in the basis line: "Reduced from 18 to 16 at client request; dedicated hoist window agreed." That sentence is worth a great deal later.
🟥 Where people go wrong
- "Based on experience." It may be true. It's still not a justification anyone else can check. Convert the experience into a rate and a source — "the last two floors took 17 and 19 days" is experience with numbers attached.
- Justifying only the activity that was challenged. The reviewer picked one; they suspect the pattern. If you produce a beautiful basis for one activity and nothing for its forty siblings, you've confirmed the suspicion. Bring the basis for the whole trade.
- Shortening it in the meeting. The pressure is real, the room is senior, and cutting three days makes the noise stop. It also makes the baseline fiction and puts your name on it. Say you'll check the inputs and come back tomorrow. Then do.
- Taking the subcontractor's number at face value. Their 10 days was written by an estimator in a tender to win the job. Compare it to what their crews actually did last month before adopting it.
- Justifying each duration and not the chain. Thirty activities, each perfectly reasonable at 5 days, in a sequence that assumes the same 20 men are in three places at once. Duration justification and resource histogram have to be read together.
- Hiding contingency inside durations. Reviewers can smell it, and under NEC it's explicitly meant to be shown as time risk allowance (Cl 31.2). A visible 3-day allowance on a critical activity is defensible; a 13.8-day calculation adopted as 18 with no explanation is not.
- Keeping the basis in your head or your laptop. The next planner, the claims consultant, and the tribunal all need it. If it isn't in the P6 file or the basis memorandum, it doesn't exist.
⚖️ When you're challenged
"Why 18 days? The last project did this in 14." "The last project's zone was 800 m² with no lintels in the wall. This one is 1,150 m² with cast-in lintels that hold the top course for two days each. Laying alone is 14 days here too — the extra four are lintels and the MEP chase hold. Here's the line."
"Your subcontractor says 10 days." "Their tender says 10. Their crews on Levels 1 and 2 did 19 and 17. I've used what they achieved, not what they bid. If they hit 10 on Level 3, the float goes up and I'll happily report it."
"The client's engineer thinks it's padded." "Then let's look at which input they'd change. Quantity's from the drawings, rate's from this site's own as-built, crew is what's mobilised. The four days above the formula are lintels and the chase hold — both are sequence, not allowance. If they want it shorter, the lever is a dedicated hoist window, and I'll show what that buys."
"If this goes to a claim, can you prove this duration was reasonable?" "Yes. The basis is in the UDFs on every critical activity and in section 4 of the basis memorandum, with the as-built rates it came from. That's what makes the delay measurable — the baseline was real."
📄 Related pages
- Crew and Manhour Formulas — the calculation that produces the basis line
- Productivity Adjustment Factors — when the rate itself needs adjusting, and how to show it
- Activity Duration Limits — how long is too long, and when a long duration is fine
- Schedule Basis Memorandum — where the full set of basis lines lives
- Schedule review comment sheet — the format the reviewer's challenge arrives in, and how to answer it
- Verifying subcontractor progress — why the sub's 10 days needs checking against their 17
- Planner ethics and pressure — holding a number when the room wants it changed
- NEC Accepted Programme — why durations in the Accepted Programme matter twice
✏️ Worked example
Hospital project, Doha. The client's reviewer has commented on BLK-3B-020 Blockwork Level 3 Zone B — 18 working days, stating that "blockwork of this extent should not exceed 10 days" and asking for it to be reduced.
Step 1 — The basis line.
| Field | Value |
|---|---|
| Quantity | 1,150 m² of 200 mm hollow block, walls 4.0 m high, approx. 290 m run |
| Rate | 1.3 mh/m² laying only; site as-built Levels 1 and 2 measured 1.25 and 1.40 |
| Crew | 8 masons + 4 helpers = 12 × 9 h = 108 mh/day |
| Manhours | 1,150 × 1.3 = 1,495 |
| Laying duration | 1,495 ÷ 108 = 13.8 → 14 working days |
| Add: lintels | 38 cast-in-situ lintels, 2-day set before the course above continues; staggered, net +2 days |
| Add: MEP chase hold | Electrical first-fix chasing follows each wall at 1.5 m; agreed sequence hold +2 days |
| Adopted | 18 working days |
Step 2 — Triangulation.
| Method | Result |
|---|---|
| Bottom-up formula | 14 laying + 4 sequence = 18 |
| Analogous — same trade, this site | Level 1 Zone B: 19 days actual. Level 2 Zone B: 17 days actual |
| Benchmark | 4.0 m wall in 1.3 m lifts = 3 lifts minimum; each wall needs 3 laying days regardless of headcount |
| Subcontractor programme | 10 days (tender); their own revised look-ahead says 16 |
Step 3 — Testing the reviewer's 10 days. To lay 1,495 mh in 10 days needs 149.5 mh/day — roughly 17 masons on 290 m of wall. Workface at 1 mason per 12–15 m of wall face is about 20–24, so headcount is physically possible. The constraints that aren't are:
- Lift height. 4.0 m walls, three lifts, mortar set between lifts: three days minimum per wall segment. With 290 m in roughly six segments and scaffold moving between them, the sequence doesn't compress below about 12 days however many masons stand there.
- Hoist. 1,150 m² at 12.5 blocks/m² is about 14,400 blocks — at 60–90 blocks to a pallet, roughly 160–240 pallets — plus mortar and lintel rebar. Zone B's share of the one hoist serving Levels 1–8 is around 8 pallet lifts a day, which is 20 days or more of hoisting on its own, and those lifts have to precede laying by at least a day.
- Chasing. The electrical subcontractor has one chasing crew for the floor; halving the blockwork duration doesn't halve theirs.
Step 4 — The reply and the outcome. The comment response quoted the basis line, the two as-built floors, and the hoist arithmetic, and offered 16 days on condition of a dedicated hoist window 06:00–09:00 daily for Zone B during the activity. The client accepted 16, the site logistics manager confirmed the hoist window in writing, and the basis line was amended to read: "Reduced 18→16 at review R2, on agreed dedicated hoist window; without window revert to 18."
Fourteen months later, when the hoist was reassigned to the façade contractor for three weeks during exactly this activity, that sentence was the entire basis of a four-day extension on Zone B — agreed without argument, because both sides had already written down what the duration depended on.
📖 References
- AACE International RP 32R-04, Determining Activity Durations
- GAO Schedule Assessment Guide (GAO-16-89G), Best Practices 3 and 4
- DCMA 14-Point Assessment, Metric 8 (High Duration)
- PMI, Practice Standard for Scheduling — duration estimating and documented assumptions
- NEC3 ECC Cl 31.2, 31.3; NEC4 ECC Cl 31.2, 31.3, 63.5 (check the edition in your contract)
- FIDIC 1999 Cl 8.3; FIDIC 2017 Cl 8.3 (check the edition in your contract)
- Society of Construction Law, Delay and Disruption Protocol, 2nd ed. (2017), Core Principle 1 and Part B guidance on programme preparation
- Oracle Primavera P6 Professional User Guide — User Defined Fields, Notebook Topics, Activity Steps, Schedule Comparison
Illustrative ranges, for sanity-checking only. Not a substitute for your own as-built data, and not for use as evidence in a contractual claim.
From the field
Experience from working planners. Unreviewed — read it as experience, not guidance.
Add what you know about justify a duration. 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