Case Study 2 — The Five-Day Cycle That Ran Eight

Every person, company, and project in this book is a Tier-3 illustrative composite. The numbers are internally consistent and realistic; they are not a real project.

Setup

Project: Wexford Flats — a six-story, 128,800 SF cast-in-place flat-plate apartment building in Rivermont. Lump-sum contract, $24,600,000. Built by Kestrel Construction Group three years before Northgate started. Ray Alvarez was not on it; the project manager was a Kestrel PM named Theo Brandt, and Jamal Foster ran the self-perform concrete, which is why this story circulates inside the company.

The structure: six elevated floors at 18,400 SF each — 110,400 SF of elevated deck — plus a slab on grade. Flat plate, conventionally reinforced, 8-inch slab, specified 28-day compressive strength (f'c) of 4,500 psi. The structural drawings required 75 percent of f'c — 3,375 psi — before shores could be removed.

The plan: a 5 work-day cycle per floor. One full-floor set of deck forms and shoring, two levels of reshores, a 9-person carpenter crew plus a placement crew, and a tower crane shared with nothing much else, because the façade did not start until the frame topped out.

The schedule that plan produced: 6 floors × 5 WD = 30 work days, or about 42 calendar days. Frame start in early October, top out in mid-November, enclosure through the winter.

What actually happened: the cycle ran 8 work days. 6 × 8 = 48 work days, about 67 calendar days. Top out came just before Christmas.

What Happens

Nobody noticed for three floors.

That is the first thing worth sitting with. Floor 2 took seven days and everybody called it a start-up floor, which is a real phenomenon and a real excuse. Floor 3 took eight, and the explanation was rain. Floor 4 took eight, and by then the schedule update showed the frame nine work days behind, and Theo asked Jamal for two more carpenters.

Jamal said no, and then he did the thing that saved the last two floors: he timed the cycle, step by step, for floor 5.

The cycle, planned versus actual

Two steps account for 2.5 of the 3 lost days. Adding carpenters would have improved step 1, which was not the problem.

Why step 5 ran 2.8 days instead of 1.0

The concrete mix had been bought in the buyout. Two suppliers quoted; the low quote came in $6.00 per cubic yard cheaper on a 4,500 psi mix. The frame took roughly 2,100 CY, so the saving was $12,600, and it was taken without a conversation about early strength.

The cheaper mix reached 4,500 psi at 28 days perfectly well. What it did not do was reach 3,375 psi in three days at 44°F, which is what a five-day cycle in October and November in Rivermont actually requires. Nobody had asked the supplier that question, because the specification is written in terms of 28-day strength and the buyout was evaluated in terms of dollars per cubic yard.

Then the testing agreement made it worse. The laboratory's contract covered weekday breaks. A cylinder broken Friday afternoon that came back at 3,100 psi meant nothing could be stripped until a Monday break — a two-day wait for a slab that had probably reached the criterion Saturday morning.

Why step 3 ran 1.2 days instead of 0.5

Cardinal Mechanical and the electrical subcontractor were producing sleeve and hanger layouts floor by floor, and their layouts required Kestrel's coordination review plus, twice, an RFI to the structural engineer about penetrations near columns. The layouts consistently arrived while the deck below was being formed rather than a floor ahead. The carpenters and the rebar crew waited on paper, on the deck, at seven in the morning.

What it cost

And the costs above are only the ones Kestrel could trace. Every trade behind the frame started 25 calendar days later than planned, into worse weather, and the enclosure sequence that had been designed for autumn ran in January.

The conversation on floor 5

Theo: "I can get you two more carpenters by Thursday."

Jamal: "I don't need carpenters. I need concrete that makes thirty-four hundred pounds in three days at forty-four degrees, and I need the sleeve layouts a floor ahead, and I need somebody at the lab on Saturday morning. Two more carpenters just means more people standing on a deck we can't strip."

Theo: "What does the mix change cost?"

Jamal: "Six dollars a yard on what's left. Call it eight thousand. And the lab wants a change order for weekend breaks — maybe four hundred a visit."

Theo: "Do it."

They did it for floors 5 and 6. The cycle came back to 6.5 days, not 5 — the sleeve-layout problem was only half solved and there was no second form set to buy back the rest — but it saved about three work days on the last two floors and it told everybody exactly where the money had gone.

Analysis

The failure was in preconstruction, and it showed up in the field. This is theme 3 — the project is built twice — with a receipt attached. The decision that cost $222,050 was made at a buyout table, in about ninety seconds, by people comparing two numbers that were correctly calculated and answered the wrong question. A concrete mix on a cycle-driven structure is not a commodity purchased by the cubic yard; it is a schedule component whose specification is "reaches the stripping criterion by day X at the temperature we will actually have." Nobody wrote that requirement down, so nobody bought it.

The cycle is a measurement problem before it is a management problem. Nobody knew where the three days went until floor 5, because nobody had timed the steps. A cycle reported as a single number — "eight days" — is unmanageable. A cycle reported as eight numbers is a list of specific problems with specific owners. Time your cycle from floor 1, not from the floor where somebody panics.

Adding labor to a cycle problem is the default wrong answer. Theo's instinct — more carpenters — is universal and almost always wrong, because cycles are dominated by hold points rather than by production. The four questions from §22.2.3 of the chapter exist for exactly this moment: which step is long, is the crew waiting or working, what does more labor do to the crane, and what is the marginal productivity of the added people. On Wexford Flats, all four questions pointed away from carpenters.

The two decisions that would have prevented it. Not five, not a culture change — two, and both were available before the first floor was formed.

1. Specify and buy the mix against the cycle. Write the early-strength requirement into the concrete scope sheet at buyout: "Mix shall achieve 3,375 psi field-cured at 72 hours at an ambient temperature of 40°F, verified by trial batch." Then evaluate quotes against that requirement instead of against dollars per cubic yard. Cost: roughly $12,600 of the "saving" given back, plus a trial batch. Value: most of $222,050.

2. Resource the cycle completely before floor 1 — the second half-floor form set and the third level of reshores if the cycle needs them, a testing agreement that includes weekend breaks, and a rule that MEP sleeve and hanger layouts are approved one floor ahead with a named due date per floor. Cost: on the order of $60,000 to $75,000, most of it recoverable as residual value. Value: the rest of $222,050.

And the one that is free. Publish the cycle-step table every floor, on the trailer wall, next to the schedule. Eight numbers, six floors, updated the morning after each strip. The team that can see which step is drifting fixes it two floors earlier than the team that cannot.

Discussion Questions

1. The mix decision saved $12,600 and cost $222,050. Where, structurally, does an organization catch that kind of decision? Name the specific document, meeting, or approval step you would add — and say who would have to attend.

2. Theo's instinct was to add carpenters. Write the four questions you would ask before approving added labor on a cycle that has slipped, and then say what you would do if the answers showed the crew was genuinely undersized.

3. The laboratory did not break cylinders on weekends. That is a $400-per-visit problem that cost multiple days. Why do constraints like this survive? What does it suggest about how you should read a testing-and-inspection agreement before you sign it?

4. Floors 5 and 6 improved from 8 days to 6.5, not to 5. Explain, using the cycle-step table, exactly where the remaining 1.5 days lived and what it would have taken to remove them. Then decide whether it would have been worth doing at that point in the job, and show your arithmetic at $5,150/CD.

5. Compare this case with Case Study 1. Both are failures of a temporary structure — one a scaffold, one a form cycle — and both trace back to a management decision made weeks earlier by competent people. What is the common mechanism, and what single practice would have caught both?

Your Turn

Take the Willow Street Community Center from this chapter's Project Checkpoint and write the concrete scope sheet language you would use at buyout to prevent the Wexford Flats failure on your job. Three to five sentences. It must state the early-strength requirement in terms of a strength, an age, a curing condition, and a verification method — and it must say what the contractor owes you if the mix does not perform.

Then price the trade-off honestly. Find out what an early-strength mix typically costs per cubic yard as a premium, apply it to Willow Street's concrete quantity, and compare it against $1,200 per calendar day of liquidated damages plus your own extended general conditions. Write the two-sentence recommendation you would give your own project executive. If the answer on a small building turns out to be "do not pay the premium," say so and show why — the point of the arithmetic is that it decides the question, not that it always produces the same answer.