Case Study 1 — The Northgate Crane Decision: Tower, Two Mobiles, or One That Walks

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

The project. The Northgate Outpatient Pavilion. Four stories, 132,000 gross square feet, structural steel frame on spread footings, composite metal deck floors, unitized curtain wall and architectural precast enclosure. Owner: Meridian Health System. Contractor: Kestrel Construction Group, CM at Risk with a $47,500,000 guaranteed maximum price and 565 calendar days of contract time.

What the crane has to do.

The site. 6.2 acres, sloping, urban-edge. The north property line is fourteen feet from the building at the closest point, and beyond it is an active clinic — a tenant of Ardmore Properties — that sees patients five days a week for the entire 565 days Kestrel is on site. Marquand Avenue, a four-lane arterial, is on the south. Selby Street, a two-lane residential street with a noise ordinance, is on the east. Meridian's existing medical office building and its patient parking lot are on the west.

The people. Ray Alvarez, senior project manager. Margo Deacon, general superintendent. Grace Lindqvist, VDC/BIM manager. Tomás Reyes, chief estimator. Hank Duffy, project manager for Ironbridge Steel, the erector. Pri Sethi, Meridian's owner's representative.

The moment. The second Tuesday in December — twelve weeks before Notice to Proceed on March 3 of Year 1, and eight months before the planned steel erection start on August 4.

What Happens

Three options on the table

Margo wants a freestanding hammerhead tower crane. Ray's GMP estimate carries two mobile cranes — a 275-ton crawler on the west and a 165-ton hydraulic truck crane on the east, both working from prepared pads outside the building footprint. Grace has spent six weeks building a 4D erection model against Hank Duffy's preliminary piece list, and she has a third option nobody asked for: one 275-ton crawler working inside the footprint on a timber mat road, setting steel around itself in five sequences and walking out through the last bay.

Grace's model kills option 2 first

She loads the two mobile cranes onto their pads, turns on capacity shading, and walks the sequence. Half the interior column lines light up red.

Grace: "From outside the footprint, with the boom length you need to clear the frame you've already erected, neither machine reaches column line D-5 with capacity to spare. That transfer girder is 21,400 pounds and it's at a ninety-six-foot radius from the nearest legal setup point on the west. The chart says no."

Ray: "So we upsize the crawler."

Grace: "Four hundred ton. Ninety-six thousand in premium and nine calendar days, because a bigger machine at a longer radius makes smaller picks and more of them. And the mat road has to be rebuilt heavier."

That figure — $96,000 and nine calendar days — is the same upsize option Kestrel had priced during preconstruction planning and set aside.

Then the model kills option 1

Grace swaps in the tower crane. Green everywhere. Coverage of the whole footprint from one position, no setup and teardown, a hook available to every trade every hour. Her 4D run shows it twenty work days faster than either mobile option across steel, deck, precast, and curtain wall.

Then she turns on one more layer: a pale grey cone extending well past the north property line.

Grace: "Out-of-service radius. A freestanding hammerhead has to weathervane when it isn't working — the slewing brake comes off so the jib turns with the wind instead of taking load on the side. That's not optional; it's how the machine survives a storm. Which means the jib crosses Ardmore's parking lot every time the wind comes out of the south. Not four picks in Sequence 4. Twenty-four hours a day for eleven months, over a medical practice."

Margo: "Show us the third thing."

The full comparison, worked

Option 3 wins by $112,400 against the tower and by $263,700 against the two-mobile plan — and it wins before you even try to price the thing that actually disqualified the tower.

The constraint that decided it

Ray had assumed the tower crane's problem was the foundation: a base in the west drive aisle takes 34 of Meridian's patient parking spaces out of service for eleven months, and Pri Sethi's hospital operations committee would never approve it. That was true, and it was not the biggest problem.

The biggest problem was the grey cone. A $22,000 license for four picks in one sequence is a negotiation. A permanent, around-the-clock, eleven-month easement over the parking lot of an operating medical practice is a different transaction entirely — and Meridian, who has to be a good neighbor to that practice for the next thirty years, would have had opinions of its own long before Ardmore's counsel did.

Margo, at the end of it: "So the answer is the machine I'd have picked in 1998 and the reason is something none of us knew this morning."

Ray: "The reason is that nobody in this room asked what the crane does when it isn't working."

What actually happened afterward

The decision saved money nobody forecast. In August of Year 1, steel erection slipped 23 calendar days — from August 4 to August 27 — because the anchor-bolt and embed submittal sat in Kestrel's office for 11 days before it went to Caldwell Structural, Caldwell took its full 14-day review, and Ironbridge lost its mill rolling slot.

A crawler crane is a rental with a monthly rate and a mobilization date. Kestrel moved the mobilization with ten days' notice for a $6,400 rescheduling fee.

A tower crane erected in July would have been standing on the site, complete, tied in, insured, with an operator on payroll, for 23 days of nothing. At the all-in rate of $77,602 a month, that is roughly $59,500 of pure standby — plus a re-permitted street closure for the assist crane and a rescheduled aviation obstruction evaluation. Net advantage of the crawler in a delay nobody predicted: about $53,000.

And the decision cost money too. The crawler demobilized in December of Year 1, once the frame and the deck were complete. When four rooftop mechanical units and the two cooling towers needed setting in February of Year 2 — with enclosure well underway — Kestrel had to remobilize a 165-ton hydraulic truck crane for three days.

A tower crane still standing in February would have made those picks at no incremental cost. It also cost Margo a Saturday she had promised her crew.

Net: the December decision was right by roughly $91,600 against the tower and $284,500 against the two-mobile plan — and it was right for a reason that never appeared on anybody's spreadsheet.

Analysis

1. The estimate was not wrong; it was unexamined. Tomás Reyes priced two mobile cranes because that is what a 132,000 SF four-story steel building on a 6.2-acre site usually takes. It is a defensible assumption at the conceptual stage. The failure was that nobody tested it against the actual erection sequence until Grace built the model — and the model is where a $200,000 error costs nothing to find. Theme 3: the project is built twice, and the first build is cheap.

2. Grace won because she asked a different question. Ray was comparing machines on price. Margo was buying availability. Grace asked what the work required — which pieces, at what radius, in what order, against what obstruction — and the machine fell out of the answer. Equipment selection is a schedule decision priced in dollars. The order of the questions is the whole method.

3. The disqualifying constraint was not in the cost column. The tower crane's out-of-service radius is not a line item, does not appear on a rental quote, and does not show up in a schedule comparison. It appears only if somebody asks what the machine does at 2 a.m. on a Sunday in a south wind. Every equipment decision has at least one constraint of this shape, and it is almost never financial.

4. The option you choose changes what future risk costs you. Nobody in December knew steel would slip 23 days. But the crawler was a flexible commitment and the tower was a rigid one, and when uncertainty arrived, flexibility paid $53,000. That is theme 1: risk management is not only about identifying the risk; it is about choosing arrangements whose failure modes you can afford.

5. And the decision was not free. Kestrel gave up crane availability in February and paid $20,800 for it. Honest analysis names what a decision costs as well as what it saves. A case study where the chosen option is perfect is a case study that has been sanded smooth.

Discussion Questions

1. Suppose Meridian had owned the adjacent parcel instead of Ardmore Properties, so the oversail question simply disappeared. Does the tower crane win? Work the numbers, and then name at least two non-financial constraints that would still have to be resolved before you could say yes.

2. Grace's 4D model showed the tower crane 20 work days faster. Ray converted that to a credit at the extended general-conditions rate of $5,150 per calendar day rather than at the total daily exposure of $10,650. Explain why that is the correct rate for a savings calculation, and identify the circumstance in which the higher rate would be right.

3. Option 2 needed a 400-ton machine to reach the transfer girder. An alternative nobody priced: redesign the connection so the transfer girder ships in two pieces and is field-spliced in place. Sketch the arguments for and against, and name the three parties whose agreement you would need.

4. The crawler's mat road cost $61,000 net of resale value. Kestrel bought the mats and sold them at the end of the job. What is the argument for renting mats instead, and what does the answer depend on? Connect it to the break-even analysis in §21.3 of the chapter.

5. The tower crane would have sat in standby for 23 days during the steel delay. Kestrel could have paused the crane rental — or could it? Look at what the $77,602 monthly all-in figure is made of and say which components would have continued regardless.

Your Turn

Take the same three options and re-run the comparison for a different building: a 9-story, 210,000 SF residential tower on a 0.9-acre downtown site with zero laydown area, no on-site parking, a subway line under the north sidewalk, and adjacent buildings on three sides.

Do not compute new dollar figures. Instead, write one page that does three things:

1. States which option wins and why, in two sentences.
2. Identifies the binding constraint — the single thing that decides it — and explains why it is different from Northgate's.
3. Names the one question you would ask first if you walked into that December meeting, and says what answer would change your recommendation.

Then compare your answer to the framework in §21.1. If your first question was about price, go back and try again.