Design development

In week eight, Marcus Oyelaran brings a question from his firm's insurer to the big room. The insurer has looked at the team's tower and wants to know how the building will be protected from rain and fire while it is under construction, before the encapsulation, the roof and the sprinklers are complete. It is the kind of question that arrives in design development, when the design is settled enough for people outside the team to read it closely. Ingrid opens the decision log to see what the team decided about the tower in week four, and why, before anyone answers.

What design development is

Design development is the stage between the schematic design, which fixes the building's form, organization and main systems, and the construction documents, which describe it completely enough to build. In design development every system is defined, sized and coordinated: the structure's members are sized and its connections typed; the envelope's assemblies are chosen and their key details drawn; the mechanical, electrical, plumbing and fire protection systems are laid out with their main equipment, routes and sizes; the code analysis is completed; the interiors are planned with their finishes; and the cost estimate is updated with real quantities. At the end of it, the owners should be able to say yes to the building, knowing what they are getting, and the team should know that nothing fundamental will change in the documents.

Solberg describes design development as the stage where the building stops being an idea and becomes a set of commitments. Each decision now carries more weight, because more work depends on it. The integrated team's advantage is that most of the decisions that cross disciplines were made in the first five weeks, together, so design development is mostly a matter of each discipline working out its own system within decisions everyone has agreed. Its risk is that decisions made together can be unmade separately: a discipline that changes something in its own model, without bringing it back to the big room, can break an agreement that three others depend on.

Change control

For that reason the team adopts a change rule at the start of design development, as most integrated projects do. Any change to a decision in the log, or any change that affects another discipline's model, must be proposed at the big room, with its reasons and its effects on the targets, and agreed before it is made. Small changes inside one discipline's own work need no approval but must be recorded in the model's change notes. Changes requested by the owners go through the same process, so that their effect on cost, schedule and targets is known before they are accepted.

The rule sounds bureaucratic, and Rafael says so. Qureshi answers with an example from her own practice: a hospital in which the electrical engineer moved a main switchboard 5 feet to make room for a door, without telling anyone, and the structural engineer's opening in the floor below, already detailed and priced, ended up in the wrong place. Finding and fixing the error took three weeks. 'The rule is not there because people are careless,' she says. 'It is there because everyone is busy, and in a busy team the only way to know what has changed is to say so.'

The decision log

The decision log is the team's memory. Each entry records the week, the decision, who decided it, the alternatives considered, the reasons, and the targets it affects. It is kept by Ingrid and reviewed at each big room. By the end of design development Ingrid's team has made fourteen major decisions, and three of them have been reopened. The script below summarizes the log.

# The decision log: how many decisions were made, by whom, and how often they were reopened
log = [  # (week, topic, decided by, reopened?)
    (2, "grid aligned through podium", "big room", False), (2, "two cores", "big room", False),
    (3, "transit hall outside the tower footprint", "big room", False), (3, "clinic on levels 2 and 3", "owner", False),
    (4, "tower structure", "big room", True), (5, "heat recovery chiller", "big room", False),
    (5, "window types limited to four", "architect", False), (6, "clinic air handler on level 3", "big room", True),
    (6, "bike hub entrance on the side street", "owner", False), (7, "roof plant enclosure height", "architect", True),
    (7, "air curtains at transit doors", "engineer", False), (8, "stair pressurization", "engineer", False),
    (8, "balconies thermally broken", "big room", False), (9, "pharmacy drive-up dropped", "owner", False),
]
by = {}
for w, t, who, re in log:
    by[who] = by.get(who, 0) + 1
reopened = [t for w, t, who, re in log if re]
print(f"{len(log)} decisions in the log; by: " + ", ".join(f"{k} {v}" for k, v in sorted(by.items())))
print(f"reopened {len(reopened)} ({len(reopened) / len(log):.0%}): " + "; ".join(reopened))
late = [t for w, t, who, re in log if w >= 7]
print(f"decisions in weeks 7 to 9: {len(late)}")

14 decisions in the log; by: architect 2, big room 7, engineer 2, owner 3
reopened 3 (21%): tower structure; clinic air handler on level 3; roof plant enclosure height
decisions in weeks 7 to 9: 5

Half of the decisions were made in the big room, by the whole team, which is what an integrated process intends. The owners made three, all about their own operations: putting the clinic on the second and third floors, placing the bike hub's entrance on the side street, and dropping the pharmacy's drive-up window, which would have put a car lane across the bus stop's sidewalk. The architect and the engineers made the rest within their own disciplines, after consulting the others.

The three reopened decisions are the more interesting part of the log. The tower structure was reopened in week eight by news that the state's rulemaking on the 2024 code had moved forward. The team had chosen a post-tensioned concrete tower in week four, keeping mass timber as an alternative. With the new code possibly in effect before the permit, the team agreed to develop the timber alternative far enough to price it, as an alternate in the builder's bid, without changing the base design. Oyelaran's insurer then asked how a timber tower would be protected from rain and fire during construction, and the team added a construction sequence to the alternate: encapsulation and sprinklers following the timber erection closely, and floor-by-floor temporary roofing. The clinic's air handling unit was reopened in week six, when Mensah pointed out that its fans, on the clinic's upper floor, would be directly under the first floor of homes, and the vibration and noise would reach them through the podium roof. The unit stayed where it was, but on vibration isolators and a floating slab, with the homes above it given a deeper acoustic floor. The roof plant enclosure was reopened in week seven when Haugen confirmed that the zoning counts any roof structure over a certain height toward the building's height; the enclosure was lowered and the heat pumps rearranged.

Solberg points out what the log shows. A reopened decision is not a failure: two of the three were reopened because new information arrived from outside the team, and one because an engineer saw a problem others had missed. What would have been a failure is a decision reopened without anyone knowing, or one that should have been reopened and was not. The log makes both visible.

Developing the parts

Each discipline spends most of design development on its own system, inside the agreed decisions. Tomasz sizes the tower's members and the podium's columns, the cores' walls and the transfer girder, and draws the main connection types. Adaeze lays out the heating, cooling and ventilation systems on every floor, sizes the heat pumps, the tanks and the heat recovery chiller, and draws the plant rooms. Rafael sizes the electrical service, lays out the risers and the panels, and designs the battery, the solar panels and the protected circuits for the outage rooms. Hamid designs the site: the grading, the stormwater storage under the forecourt, the connection to the sewers, the curb at the bus stop and the fire department's access. Ingrid develops the plans, the facades and the sections, the code analysis with Kostyshyn, the accessible routes and the interiors.

Some decisions in design development still cross disciplines and come back to the big room. The facade is one. Ingrid has proposed prefabricated wall panels, on the concrete tower's grid, hung from the edge of each slab, with four window types; Adaeze needs the panels' thermal performance and airtightness; Tomasz needs their weight and the way they are fixed; Oyelaran needs to know how they will be lifted and how many can be delivered in a day down the lane. The balconies are another. Projecting balconies make thermal bridges where the floor passes through the wall, and the team agrees to thermally broken connections, priced by Okafor and modeled by Adaeze, rather than giving up the balconies Lafontaine asked for.

Acoustics are a third. The clinic sits between the transit hall and the homes. The buses pull up outside the hall every few minutes; the hall's doors open constantly; the clinic needs quiet counseling rooms; and the homes need to be quiet at night above a clinic that may have cleaning crews until midnight and a transit hall open until one in the morning. Claire Beaulieu, the acoustician who advised the students in their first studio, reviews the separations and asks for a heavier floor between the clinic and the homes, which the podium roof's concrete slab provides, and for the counseling rooms to be away from the hall and the stair, which Ingrid has already done.

The owners' review

Design development ends with an owners' review, at which the three clients see the developed design, the updated estimate and the updated energy and carbon results, and approve the design for construction documents. Ingrid's team presents the building as it now stands, the decision log, and a short list of open questions with the date each must be closed. Courchesne asks about the clinic's waiting room windows, which face the transit hall's roof; the team shows that the hall's roof is a planted terrace for the clinic's staff, visible from the waiting room. Sadeghi asks whether the hall's doors will keep it warm on the coldest mornings; Adaeze shows the vestibules, the air curtains and the radiant floor, and the model's result. Lafontaine asks whether every home still has light from two sides; Ingrid shows that all but four do, and that the four are studios on the north-east corner with a corner window. All three approve.

Exercise 9: design development

• Adopt a change rule for design development and apply it to every change that affects another discipline or a target.
• Keep a decision log with week, decision, decider, alternatives, reasons and targets, and summarize it at the end of the stage.
• Develop each system within the agreed decisions, and bring any crossing decisions, such as facades, balconies and acoustics, back to the big room.
• Prepare an owners' review: the developed design, the updated estimate and performance results, and a list of open questions with closing dates.

Self-check

• Has every reopened decision been recorded with the new information that caused it?
• Is there any change in your model that another discipline does not know about?
• Could the owners approve your design today knowing what they are getting?