How do I prove a construction delay?

The site's own week-by-week account already exists, dated and measured, before anyone asks for it. Proving the delay becomes reading the record across the window the claim covers, rather than reconstructing it afterwards out of photographs, diaries and memory.

What documents do I need for an extension-of-time (EOT) claim?

Entitlement rests on what the site looked like as each week passed — the condition of the works, the access available, the hours plant stood, the shifts weather stopped and the work actually placed against the programme. All of it is captured while it is that week, so the claim is assembled rather than built from memory.

What counts as evidence-grade construction delay claim documentation?

A dated, measured account of the site in which every figure names its source, its method and its uncertainty, and in which the day the work was in that condition is recorded separately from the day the file came to know it.

How is standing time proven?

By two independent measurements read together: the hours plant was on site and not working, and the measured state of the front it was waiting on over the same dates. Stopped work and standing plant are on the record with their dates and their hours, so the line is counted rather than asserted.

Can the record produce an as-built programme?

Yes, as a by-product. A weekly measured record already holds the dates each front actually started and finished, so the as-built programme is read out of the record instead of being commissioned as a reconstruction once the dispute is live.

Does this work under FIDIC or NEC?

The record does not care what the contract calls the event. It supplies dated, measured facts for the window a FIDIC clause 20 notice, an NEC compensation event or a JCT extension-of-time claim is made about — and it supplies the same facts, in the same form, to whoever is assessing it.

Is a progress claim the same thing?

No, and the word collides. A progress claim is a payment application for work done this period. This page is about claims for time and money under the contract. Both read from one record; each is a different question asked of it.

What if the record shows less than the claim does?

Then it shows less, in units, with its uncertainty and its sources named. The record states what was measured and stops there. What that means for entitlement is a matter for the parties, their contract and whoever assesses it — never for us.

Record · Claims

The claim is already written.

A delay claim is decided on what the site looked like as each week passed: the condition of the works, the access, the hours plant stood, the shifts weather stopped, the work that could not start. That account is written while it is still that week — dated, measured, on the same control every time. So when entitlement is argued two years later, the claim is assembled out of the record rather than rebuilt from memory, and every figure in it still carries the day it was measured on.

Record header · sheet R-06

Record: SPECIMEN SITE 04
Works: VIADUCT 2 · PIERS 4–9
Period: 29 JUN – 28 SEP 2026
Epochs: 14 · WEEKLY
Events: 3 NOTIFIED
Written: ON THE DAY, EVERY WEEK
Read: WHEN ENTITLEMENT IS ARGUED

First epoch · 29 Jun 202614 epochs on the recordLatest · 28 Sep 2026

The claim bundle

Pick an event. The bundle is already in the record.

A claim is a window of dates and what is true inside it. Choose one of the three notified events below and the exhibits assemble out of the weeks it spans: the ground state each week, the hours plant stood, the shifts the weather stopped against the allowance the contract gives, and the work placed against the programme. Nothing on the plate was written for the claim — it is the same record the site was running on all along, cut to the window the claim is about.

Bundle bench · specimen record14 EPOCHS · 3 EVENTS NOTIFIED

EV-04 · 4 weeks

Above the rule · plant hours, working then standing · below the rule · shifts stopped for weather

The record, week by week: plant hours worked and stood, shifts stopped for weather, and concrete placed. EV-04, Access withheld, Zone B, spans W-30 to W-33.
Week	Date	Hours worked	Hours standing	Shifts stopped	Placed, m³
W-27	29 JUN 2026	486	34	0	352
W-28	06 JUL 2026	502	28	0	368
W-29	13 JUL 2026	494	31	0	361
W-30	20 JUL 2026	318	196	0	214
W-31	27 JUL 2026	246	268	0	148
W-32	03 AUG 2026	231	284	0	132
W-33	10 AUG 2026	302	208	0	196
W-34	17 AUG 2026	268	142	3	176
W-35	24 AUG 2026	194	176	5	112
W-36	31 AUG 2026	236	158	4	149
W-37	07 SEP 2026	388	118	1	268
W-38	14 SEP 2026	344	162	0	231
W-39	21 SEP 2026	358	151	0	244
W-40	28 SEP 2026	412	96	0	297

Bundle EV-04 · Access withheld, Zone B

5 EXHIBITS · 4 EPOCHS · WINDOW 20 JUL → 10 AUG

Ground state, week by week: 4 epochs · 20 JUL → 10 AUG; MEASURED ON THE SAME CONTROL EACH WEEK · STILL RE-MEASURABLE
Plant standing: 956 h standing · 1,097 h working; STANDING IS 47% OF PLANT HOURS ON SITE IN THE WINDOW
Shifts stopped for weather: 0 of 24 · allowance 2 · net 0; 10 MM RECORDED ON SITE · ALLOWANCE IS THE CONTRACT'S OWN FIGURE
Placed against the programme: 690 m³ ± 17 · programme 1,400 m³; SHORTFALL 710 m³ OVER 4 WEEKS AT 350 m³/WEEK
As-built programme rows: 4 weeks, each dated; READ OUT OF THE RECORD · NOT RECONSTRUCTED FOR THE CLAIM

Every figure on this plate is a sum, a count or a difference over the weeks the window spans. None of it was written for the claim — it is the record the site was already running on, cut to the dates the claim is about, with the sources of each figure named beside it.

Why contemporaneous wins

Made at the time, or made later.

Contemporaneous records are records made while the work was happening rather than written up once a claim is in prospect. Records made as it happened, not reconstructed afterwards — that is what survives cross-examination, and the reason is plain: a note made on the day carries the conditions, the hours and the measurement of that day, while a note made eighteen months later carries what somebody remembers. The two files below hold the same six facts about the same eleven weeks. Only the dates differ.

Two files · the same six factsBAR = DAYS BETWEEN THE DAY AND THE ENTRY

Written as it happened

SIX ENTRIES · SIX DATES · EACH MADE ON THE DAY

Zone B gate closed at 06:40; the piling gang stood down
20 JulD+0
Entered 20 Jul 2026
Two rigs and the crawler crane stood beside a front that could not start
28 JulD+0
Entered 28 Jul 2026
41 mm of rain; the deck pour was called off and the shift lost
25 AugD+0
Entered 25 Aug 2026
112 m³ placed against a programmed 350 m³ for the week
31 AugD+0
Entered 31 Aug 2026
Rock met at Pier 7 above the level the ground investigation showed
08 SepD+0
Entered 08 Sep 2026
Revised setting-out issued; the pier cage was re-detailed
21 SepD+0
Entered 21 Sep 2026

Assembled afterwards

SIX ENTRIES · ONE DATE · ALL MADE 06 SEP 2027

Zone B gate closed at 06:40; the piling gang stood down
20 JulD+413
Entered 06 Sep 2027
Two rigs and the crawler crane stood beside a front that could not start
28 JulD+405
Entered 06 Sep 2027
41 mm of rain; the deck pour was called off and the shift lost
25 AugD+377
Entered 06 Sep 2027
112 m³ placed against a programmed 350 m³ for the week
31 AugD+371
Entered 06 Sep 2027
Rock met at Pier 7 above the level the ground investigation showed
08 SepD+363
Entered 06 Sep 2027
Revised setting-out issued; the pier cage was re-detailed
21 SepD+350
Entered 06 Sep 2027

The facts are identical and both files are specimen. What differs is the gap between the day and the entry — nothing in the left-hand column, up to 413 days in the right. A record made on the day carries the conditions, the hours and the measurement of that day. A record made a year later carries what somebody remembers of it, and every party in the room knows which is which. That is the whole of why contemporaneous records survive cross-examination and reconstructed ones argue for their lives.

A record written after the question is asked can only carry what is remembered.

Every entry carries the date the work was in that condition, not the date the file was made.

The record is written as the site is captured, so nothing depends on anybody remembering to keep it in a bad week.

A correction is added and dated beside the original; the original stays where it was, which is why last year's answer can still be recovered.

What is in it

What an evidence-grade record contains.

Construction delay claim documentation is usually a folder of photographs and a diary, assembled once the argument starts. Evidence-grade is six things carried on every entry from the day it was made — and they are the six a delay expert, an engineer or a tribunal asks for in the first hour of looking at a claim.

The date the site was that way

Every entry carries the day the work was in that condition, kept separately from the day the record came to know it. A claim argued in 2029 about a Tuesday in 2026 gets the Tuesday's answer, not today's opinion of it.

A measurement, not an impression

Quantities arrive with the method that produced them and the tolerance that method earns. “Substantially complete” is not a figure. 1,180 m³ ± 28 is.

Hours, not adjectives

Plant on site is on the record hour by hour, working and standing, at the same weekly cadence as everything else. Disruption gets counted rather than described.

The conditions of the day

What stopped a shift sits on the same sheet as the shift, and is read against the allowance the contract already gives for it — so weather is normalised against the contract instead of argued from a weather report bought afterwards.

The programme the site actually ran

A weekly measured record already holds when each front started and stopped, so the as-built programme is read out of the record rather than commissioned as a reconstruction. Those dates were dates before they were evidence.

A source on every figure

Each number names where it came from, so the expert on the other side can check it instead of taking it. Anything with no independent source is carried as unmeasured and says so on its face.

Standing time

How standing plant gets proven.

Standing time is the hardest line in most claims, because it is usually asserted from a sheet written by the party the hours belong to. Here it is two independent measurements read against each other, and both of them were made before anyone knew there would be an argument.

01The plant is on the record
Every machine on site carries its hours — working, and on site but not working — week by week, on the same cadence the site is captured on. Nobody reconstructs them from a day-work sheet at the end of the job.
02The front it was waiting on is measured the same week
The work the machine was standing beside has a measured state on the same dates: quantity placed, access available, the condition of the ground it had to sit on. Two independent accounts of one week.
03The two are read together, with the dates attached
Standing hours beside a front that did not move is a fact with a date range and two sources behind it. That is what a claim can be built on, and what an assessor can check without taking anybody's word for it.

A note on the word

A progress claim is a payment application: how much of the work is certified this month, answered against measured work on capture-verified billing. This page is about claims in the contract sense — time and money for events that were not the contractor’s to control, notified under whatever the contract calls it: a FIDIC clause 20 notice, an NEC compensation event, a JCT extension of time. One record underneath, two different questions asked of it.

Questions of record

The questions a claim actually starts with.

Including the one about what happens when the record shows less than the claim does.

01How do I prove a construction delay?: The site's own week-by-week account already exists, dated and measured, before anyone asks for it. Proving the delay becomes reading the record across the window the claim covers, rather than reconstructing it afterwards out of photographs, diaries and memory.
02What documents do I need for an extension-of-time (EOT) claim?: Entitlement rests on what the site looked like as each week passed — the condition of the works, the access available, the hours plant stood, the shifts weather stopped and the work actually placed against the programme. All of it is captured while it is that week, so the claim is assembled rather than built from memory.
03What are contemporaneous records, and why do they matter?: Records made at the time the work happened rather than written up afterwards. They are the kind that survives cross-examination, because a note made on the day carries the conditions, the hours and the measurement of that day. Here the record is written as the site moves, so it is contemporaneous without anybody having to remember to make it so.
04What counts as evidence-grade construction delay claim documentation?: A dated, measured account of the site in which every figure names its source, its method and its uncertainty, and in which the day the work was in that condition is recorded separately from the day the file came to know it.
05How is standing time proven?: By two independent measurements read together: the hours plant was on site and not working, and the measured state of the front it was waiting on over the same dates. Stopped work and standing plant are on the record with their dates and their hours, so the line is counted rather than asserted.
06Can the record produce an as-built programme?: Yes, as a by-product. A weekly measured record already holds the dates each front actually started and finished, so the as-built programme is read out of the record instead of being commissioned as a reconstruction once the dispute is live.
07Does this work under FIDIC or NEC?: The record does not care what the contract calls the event. It supplies dated, measured facts for the window a FIDIC clause 20 notice, an NEC compensation event or a JCT extension-of-time claim is made about — and it supplies the same facts, in the same form, to whoever is assessing it.
08Is a progress claim the same thing?: No, and the word collides. A progress claim is a payment application for work done this period. This page is about claims for time and money under the contract. Both read from one record; each is a different question asked of it.
09What if the record shows less than the claim does?: Then it shows less, in units, with its uncertainty and its sources named. The record states what was measured and stops there. What that means for entitlement is a matter for the parties, their contract and whoever assesses it — never for us.

The site's timeline is already written, dated and measured before anyone asks for it.

Written while each week was the current weekStanding hours, stopped shifts and ground state, all datedEvery figure names its source, its method and its ±