Your loads were determined by decisions you made weeks ago.
◦ Which single earlier decision had the largest effect?
◦ If you could go back to Week 3 knowing this, what would you change?
◦ Who on a real project team should have raised this, and when?
Occupant comfort is not the same as air temperature.
◦ What else determines whether an occupant feels comfortable?
◦ Which of those can your model predict, and which can it not?
The Bay Area has a mild climate.
◦ Where in your typical floor, and where else in your building, is mechanical cooling genuinely necessary?
◦ What would it take to eliminate it, and what would that cost the occupants?
You computed loads for one typical floor.
◦ Which floors in your building would have a different loads, and why?
The first floor (B01-L01) and the top residential floor (B01-L06) would differ most. L01 sits over the L00 level and its podium roofs, so its floor slab exchanges heat with that space below. L06 sits under the roof (B01-L07 is the green roof), so it gets roof heat gain and loss that the middle floors don’t. Floors L02 to L05 are sandwiched between conditioned floors and should be close to my typical floor. Corner units and units on different facades also change with solar and envelope exposure, wherever they are.
◦ Would the system you modeled for your two units still work on those floors, or what would you change?
Yes on the middle floors, since the same fan coil, trunk and branch layout would carry over. On L06 I’d check each unit’s peak load against the 600 CFM fan coil, and I’d add airflow or upsize equipment if the roof load pushes a unit past it. On L01 I’d check the plenum depth and the longest condenser line (about 63 ft up to the roof), and I’d group the lines in a vertical chase. Upper floors have shorter refrigerant runs, which helps.
You were told not to agonize over duct connections.
◦ What is the difference between a detail that matters now and one that can wait?
A detail matters now if it changes a number or a decision that everything after it depends on, such as airflow, duct sizes, equipment capacity, whether the system fits in the plenum, and where the condenser goes. A detail can wait if it only affects how a connection is drawn and not what it carries, such as exact fitting geometry, a neat tap, or tags and labels.
◦ How would you know you had deferred something you should not have?
I’d ask one test: would fixing it later force me to redo sizing, the loads, or the coordination with structure? In this project it happened with the 1-inch diffuser stubs. I treated them as a drawing detail, but they blocked automatic sizing and I had to raise the branches to fix it. Other warning signs are a system whose flow doesn’t equal the sum of its terminals, errors that appear in Check Duct Systems, and a duct that can’t fit in the plenum.