Case Study 26-1 — The Same Week, Written Twice

What This Case Is

This is not a story about a project that went wrong. It is a story about a week that went exactly the way weeks go, recorded two different ways, and what each version turned out to be worth eight months later when the delay claim was assembled.

Everything below is a Tier-3 illustrative composite built on the canonical Northgate record.

Setup

The project. Northgate Outpatient Pavilion. Meridian Health System, owner; Halvorsen + Pike, architect; Kestrel Construction Group, construction manager at risk under a $47,500,000 guaranteed maximum price and 565 calendar days. Liquidated damages $5,500 per calendar day; the contractually agreed extended general-conditions rate is $5,150 per calendar day. Total daily exposure to slipping substantial completion: $10,650 per calendar day.

The week. Weeks 45 through 52 of the job are the Level 2 and Level 3 interior build-out — roughly early January through late February of Year 2. Full building dry-in is not until March 28 of Year 2, but the south wing is enclosed and released, so interior work is running there while the enclosure crews work the north and east elevations. We are looking at one week in that stretch: five working days, Monday through Friday, on Level 3 Areas B and C.

The people. Dani Okonkwo, field engineer, writes the daily reports. Margo Deacon, general superintendent, reviews them. Ray Alvarez, senior project manager, reads them on Friday afternoons if the week has been quiet and not at all if it has not. Sightline Interiors has the framing and board. Halcyon Electric and Cardinal Mechanical are in the walls ahead of them. Dale Whitcomb at H+P owes two RFI responses.

The pressure. Kestrel accelerated in the fall to recover seventeen of the twenty-three calendar days lost when steel erection slipped from August 4 to August 27 of Year 1. That acceleration cost $168,000 and it consumed essentially all of the interior float that used to sit between trades. In week 45, an idle crew is no longer a nuisance. It is a schedule event.

What Happened That Week

Here is the week, as it actually occurred.

Nothing dramatic. No injury, no collapse, no catastrophe. A perfectly ordinary week on a job under schedule pressure.

Version A: The Reports That Were Actually Written

These are real in the sense that they are what busy, competent, well-intentioned people write at 4:55 p.m. Dani was not lazy. Dani was twenty-four, had walked eleven floors that day, and had four other things due.

Mon. Weather: cool, dry. Manpower: Sightline 22, Cardinal 16, Halcyon 13, Thermaline 4. Work: framing L3, board L2, MEP rough-in continuing. Remarks: Waiting on RFI.

Tue. Weather: cool. Manpower: Sightline 22, Cardinal 16, Halcyon 13, Thermaline 4. Work: framing L3, board L2, MEP rough-in continuing. Remarks: Still waiting on RFI 112.

Wed. Weather: cool, some sun. Manpower: Sightline 22, Cardinal 15, Halcyon 13, Thermaline 4. Work: framing L3 Area D, board L2, MEP rough-in. Remarks: Emailed architect re RFI 112.

Thu. Weather: mild. Manpower: Sightline 22, Cardinal 16, Halcyon 12, Thermaline 4. Work: framing L3, board L2, MEP. Remarks: RFI 112 answered. Some rework required.

Fri. Weather: mild. Manpower: Sightline 22, Cardinal 16, Halcyon 13, Thermaline 4. Work: framing L3 Area B, board L2, MEP. Remarks: Crews progressing.

Read those five entries as an outsider would in eight months.

They establish that Sightline had twenty-two people on site every day of the week. They establish that RFI 112 existed and was answered on Thursday. They establish nothing else. They do not say who was idle, where, for how long, or why no other work was available. They do not distinguish the six framers from the eight board hangers from the eight in the other trades within Sightline's twenty-two. They do not quantify the rework. And the manpower count — twenty-two, every day, unchanged — actively argues against the claim, because it shows a subcontractor at full strength for five consecutive days.

The word "progressing" appears once. That word will be read aloud.

Version B: The Reports That Should Have Been Written

Same week. Same person. Roughly eight additional minutes a day.

Mon. Problems, delays, and interferences. Sightline Interiors framing, L3 Area B, grids C–F: 6 carpenters idle 0700–1130 = 27 labor hours. Cause: no response to RFI 112 (head-of-wall detail, rated corridor), issued 12 work days ago; response due under the 10-work-day turnaround recorded in the preconstruction meeting minutes of March 12, Year 1. No other released work available in Area B; Area C not released pending in-wall inspection not yet called by Halcyon. Mitigation: crew redeployed 1130 to Area D at reduced efficiency — unfamiliar layout, stock relocated by hand. Kestrel emailed H+P 0740. Photos 0112-L3B-01 to -03.

Tue. Sightline framing, L3 Area B: 6 carpenters idle 0700–1015 = 19.5 labor hours. Same cause. Cardinal Mechanical, L3 Area B: 4 sheet-metal workers idle 0700–1130 = 18 labor hours — duct penetrations at the same head-of-wall condition cannot be set. No other released work in Area B. Both crews redeployed to Area D/L2. Kestrel emailed H+P 0720 requesting a response date. Photos 0113-L3B-01, -02.

Wed. Sightline framing held out of L3 Area B entirely; 6 carpenters worked Area D at estimated 70% efficiency = approximately 14.4 labor hours lost to relocation of material and unfamiliar layout. RFI 112 remains open, 14 work days. Kestrel emailed H+P 0715, copy P. Sethi, requesting a response by end of day Thursday and advising that framing in Area B is the controlling activity for the Area B interior sequence per Schedule Update No. 18.

Thu. RFI 112 answered 1420, 13 work days after issue (agreed turnaround: 10 work days). Response revises the head-of-wall detail. 240 LF of track set in L3 Area B under the prior released detail is now nonconforming and must be removed and reset. NCR-022 opened. 2 carpenters pulled to demolition 1430–1600 = 3 labor hours. Photos 0114-L3B-01 to -06 show the as-installed condition and the revised detail. Kestrel advised H+P and P. Sethi in writing at 1615 that the response requires rework and that Kestrel will submit a change proposal and reserves its rights as to cost and time.

Fri. Framing restarted L3 Area B 0700 with 6 carpenters. Reset of 240 LF of track consumed 2 carpenters for 6.5 hours = 13 labor hours of rework. Cardinal returned to Area B 0700. Week summary for the record: Area B framing lost 4 of 5 planned days against Look-Ahead No. 18; total idle and inefficiency 83.9 labor hours; rework 16 labor hours; NCR-022 open. Photos 0115-L3B-01 to -04.

Same events. Same people. Same week.

Analysis: What Each Version Is Worth

Eight months later, Kestrel is assembling a delay and disruption position covering the interior build-out. A claims consultant reads both versions. Here is the arithmetic.

The labor hours, from Version B:

At an illustrative blended burdened interior-trade rate of $62.00 per labor hour — your market and labor agreement will differ, sometimes substantially:

94.9 MH × $62.00/MH = $5,884

Just under six thousand dollars of craft labor for one ordinary week, in one area, on one unanswered RFI.

Now the part that matters. That week is one of eleven similar weeks across the interior build-out. Eleven weeks at a comparable magnitude is roughly $65,000 of direct craft labor, before any claim for extended general conditions, before any inefficiency factor applied to the surrounding work, and before the time impact.

Version A is not merely weaker. It is affirmatively harmful. The constant manpower count of twenty-two, repeated across five days, is the strongest single piece of evidence against a disruption claim, because it shows a subcontractor fully staffed and, by implication, fully productive. The word "progressing" on Friday reads as an admission that the week went fine. And an analyst who finds five consecutive days of near-identical entries begins to doubt whether anyone was observing anything — which then infects every other report in the file, including the good ones.

The cost of the difference. Version B is about eight additional minutes a day. Five days is forty minutes. Across eleven weeks, roughly seven and a half hours of writing. Against a defensible position worth on the order of $65,000 in direct labor alone, that is roughly $8,700 per hour of writing — and it does not count the time position, which on a job with $10,650 per calendar day of exposure is worth far more than the labor.

And the honest caveat. Version B does not guarantee Kestrel wins. Entitlement, causation, and damages are three separate proofs and you must carry all three (Chapter 33). H+P may have a defensible reason the RFI took thirteen days. Meridian may argue concurrency. Version B does not settle those arguments. What it does is put Kestrel in a position to have the argument. Version A does not.

That is the whole lesson, and it is Theme 5 in its most literal form: documentation is the memory of the project. A project with Version A reports has amnesia, and it will lose arguments it should win, on facts that actually happened, to people who were not there.

Discussion Questions

1. Dani Okonkwo wrote Version A while being genuinely diligent about everything else. What in the system — not in Dani — produced Version A? Name at least three changes to the system that would produce Version B without anyone working harder.

2. Version B's Wednesday entry claims a 30 percent inefficiency for a crew working an unfamiliar area with relocated material. That is an estimate, not a measurement. Is including it stronger or weaker than omitting it? Defend your answer, and say how you would make the estimate more defensible.

3. The Friday entry in Version B includes a week-summary line. No standard daily report form has such a field. What does it add, and what is the risk of a field engineer writing weekly summaries into a daily record?

4. Margo Deacon reviews these reports and signs them. What should she be looking for in ten seconds per report, and what should she send back?

5. Version A's constant manpower count of 22 was accurate — Sightline did have 22 people on site. Explain precisely how an accurate number becomes harmful evidence, and what the fix is.

Your Turn

Take the Tuesday of this week and write the daily report entry three ways:

1. As Version A wrote it (one line).
2. As Version B wrote it (the full impact statement, both trades).
3. As an over-reacher would write it — an entry that overstates, editorializes, and blames H+P by name for incompetence.

Then write one paragraph explaining why version 3 is worse than version 1 in the hands of a claims consultant, even though it contains more information. This is the line between a strong record and an advocacy document, and learning to feel where it sits is most of the skill.