Case Study 14-1 — Twenty-Three Days: The Anchor Bolts, the Mill Slot, and the Time Impact Analysis

The projects, companies, and people in this case study are illustrative composites created for teaching. The arithmetic is real; the names are not.

Setup

Project: Northgate Outpatient Pavilion — 132,000 SF, four stories, structural steel frame, outpatient clinics and an ambulatory surgery suite. Owner: Meridian Health System. Owner's rep: Priyanka "Pri" Sethi. Contractor: Kestrel Construction Group, CM at Risk, GMP $47,500,000, contract time 565 calendar days, liquidated damages $5,500 per calendar day. Structural engineer: Caldwell Structural (principal: Ruth Caldwell). Steel subcontractor: Ironbridge Steel (project manager: Hank Duffy). Kestrel team: Ray Alvarez (PM), Wei Chen (project controls), Margo Deacon (general superintendent), Dani Okonkwo (field engineer), Nadia Haddad (VP Operations).

Kestrel's accepted baseline showed steel erection starting August 4, Year 1, and substantial completion on September 18, Year 2. Steel erection was on the critical path.

The chain that fed it was not.

Read that table the way Ray read it in the first week of the job: four days of float in a chain that terminates at a mill rolling slot. A mill slot is not a crew you can double or a shift you can extend. It is a reservation. Ironbridge held a rolling week at the mill, and the release-to-fabrication deadline for that slot was 10 calendar days before rolling.

Ray wrote a note on the back of a meeting agenda: A/B submittal — 4 days float — hard downstream constraint at fabricator. Treat as critical.

Nobody treated it as critical.

What Happened

Step 1 — Eleven days in an office

The anchor-bolt and embed package arrived from Ironbridge and went into the submittal queue. Kestrel's submittal log was a spreadsheet at that point in the job; the package had a "received" date and no "required by" date. Dani Okonkwo, three weeks into their first job, assumed a package that important was somebody else's priority. Ray assumed Wei was tracking it because it was on the schedule. Wei was tracking the activity, which was showing four days of float and therefore not appearing on any critical-path report.

The package sat for 11 calendar days before it was transmitted to Caldwell Structural.

Step 2 — Fourteen days, exactly as contracted

Kestrel's baseline had assumed Caldwell would return embed drawings in about 10 calendar days, because early in preconstruction Ruth Caldwell had said her office "usually turns embeds in about a week." That sentence was never written down and never made it into the contract. The contract allowed 14 calendar days, and Caldwell — busy, entirely within its rights — took all fourteen.

That is an important lesson on its own: a duration based on a verbal expectation is not a duration; it is a hope. Kestrel's schedule carried the risk of the difference and nobody had priced it.

Step 3 — The slot

Ironbridge's mill release was required by April 18. Caldwell Structural returned the approved submittal on April 21 — fourteen calendar days after Kestrel finally transmitted it on April 7, which is exactly the review period Caldwell agreed to. Ironbridge missed the release deadline by 3 days — and a missed deadline on a mill slot is binary. Hank Duffy called Ray on a Wednesday afternoon.

Hank: "We're out. The slot's gone. Next opening they'll give me is five weeks after the one we just missed."

Ray: "Five weeks. That's thirty-five days. Can you do anything with it?"

Hank: "Some. My detailer's already through the shop drawings, so that time isn't lost. I can roll the secondary framing out of stock instead of waiting on the mill for it. And I can deliver by erection zone instead of by piece mark, so you can start the north bays before the last truck comes. That's the best I've got."

Hank's recovery was worth 12 calendar days.

Steel erection start moved from August 4 to August 27, Year 1.

Step 4 — The time impact analysis

Wei Chen ran a time impact analysis (TIA): take the last accepted schedule update, insert a fragnet representing the delay, recalculate, and measure how far the completion milestone moves. That method is the standard prospective approach in the industry, and its virtue is that it isolates this delay from everything else happening on the job.

The inserted fragnet added 23 calendar days to the steel erection start. Because erection was on the critical path and the four days of float in the procurement chain had already been consumed by the eleven-day hold, all 23 days carried straight through to substantial completion.

That second row is not a contract date. Meridian's leased interim clinic space expires October 1, Year 2. It is a business fact with no clause behind it, and it turned out to be the number that decided everything.

Step 5 — The decision

Nadia Haddad convened the meeting. Wei put three options on one page.

The daily exposure of $10,650/CD is $5,150 of extended general conditions plus $5,500 of liquidated damages.

Do the arithmetic: $244,950 − $231,900 = $13,050. On a $47.5 million project, accelerating saved about three one-hundredths of one percent. Financially it is a coin flip.

Nadia: "Thirteen thousand dollars. That's not a decision, that's noise. What's the real number?"

Ray: "October first. Meridian's lease. At twenty-three days late they have clinics scheduled into a building that doesn't exist. At six days late they have a hard week and a lot of apologies."

Nadia: "Then it isn't a money decision. Accelerate."

Kestrel accelerated. Forecast substantial completion: September 24, Year 2 — six calendar days past the contract date, but seven days ahead of the lease expiry.

Analysis

What actually failed. Not the schedule. The schedule was correct; it reported four days of float on the procurement chain, and four days is what there was. What failed was the management response to a small float number attached to a chain with a hard downstream constraint. Every report Kestrel produced filtered on zero float. Four days is not zero. The activity was invisible by design.

The fix is a rule, and Kestrel adopted it after this job: any activity whose downstream chain ends at an external constraint you cannot buy back — a mill slot, a permit, a utility company cut-in, an owner-furnished delivery, a seasonal restriction — is managed as critical regardless of its float. Float measures how much a chain can slip before it moves the project. It says nothing about whether the slip is recoverable once it happens.

Where the 23 days really came from. Notice that the two things everyone argued about — Kestrel's 11 days and Caldwell's 4 extra days — total 15, and the actual damage was 23. The gap is the character of the downstream constraint. Fifteen days of ordinary lateness became 23 days of project delay because it crossed a threshold in a system that does not degrade gracefully. This is the single most useful thing to carry out of this case: delay is not always proportional to lateness. Look for the thresholds.

Why the acceleration math was almost irrelevant. Kestrel's comparison table showed a $13,050 benefit. If you stop there, you conclude that acceleration was barely worth doing, and a purely financial decision maker might have declined. The correct analysis brought in a constraint that does not appear anywhere in the contract documents: the owner's own operations. Any project manager who thinks the schedule exists to protect the contractor's dates will miss this every time. Pri Sethi's problem was never liquidated damages; it was patients.

The bill that was not on the page. Acceleration is not free of non-monetary cost, and Kestrel paid that bill in weeks 34 through 36: trade stacking on the enclosure, a rework event on deck-edge detailing, and a spike in near-misses that culminated in a scaffold event on the north elevation, level 3, where a plank modified overnight by another trade shifted under a mason tender named Milo Serrano. He caught himself on the top rail. No injury. Bea Salgado's investigation found three failures, and the third was "a crew running behind after the steel acceleration, with an unwritten 'make it up' pressure." That finding is a direct line from a submittal that sat in an office for eleven days.

What good documentation would have been worth. Two contemporaneous records would have changed Kestrel's position materially: (1) a submittal log with a required-by date back-scheduled from the mill release deadline, which would have made the eleven-day hold visible on day three instead of day eleven; and (2) a written record of the 10-day review expectation, which would at least have created a conversation with Caldwell before the schedule depended on it. Neither is expensive. Both are boring. That is what protective documentation looks like.

Discussion Questions

1. Kestrel's own delay (11 days) was larger than Caldwell's contribution (4 days), and Ironbridge's recovery (12 days) was larger than either. If Meridian had assessed liquidated damages, how would you apportion responsibility — and what would you need in writing to defend your apportionment?
2. Wei Chen's TIA inserted the delay into the last accepted update and recalculated. What is the alternative — analyzing the delay retrospectively, after the job is over — and what are the advantages and disadvantages of each approach for the contractor?
3. Kestrel's report filtered on total float = 0. Design the filter you would use instead. What float threshold, and what additional attribute, would have surfaced the anchor-bolt submittal?
4. Suppose Meridian's lease had not expired until December. Does Kestrel still accelerate? What changes in the analysis, and who makes the decision?
5. Hank Duffy recovered 12 of the 35 days by pre-completed detailing, stock secondary steel, and zone-sequenced delivery. Which of those three should have been in the baseline plan all along, and what would including it have changed about the float number?

Your Turn

You are Wei Chen, five days after the missed mill slot. Nadia Haddad wants a one-page decision memo by 8:00 a.m.

Produce it. It must contain: (a) a statement of the delay and its cause in three sentences, naming who did what and when; (b) the TIA result, including the forecast substantial completion date and the float against both the contract date and the lease-expiry date; (c) the three-option cost comparison table; (d) a recommendation with the reason stated in one sentence that does not contain a dollar sign; and (e) the notice you are giving, to whom, and under which contract provision — and if you do not know which provision, say what you would look up first.

Keep it to one page. If it does not fit on one page, it will not get read, and a memo that does not get read is the same as a memo that was never written.