Why does it matter to model the real terrain and real neighboring buildings?
◦ Which aspects of a building design are most affected by surrounding context?
◦ What kinds of mistakes only become visible once the context is present?
Answer: The context changed my numbers more than my own massing did. The empty lot got 3+ hours of winter sun on 87% of the ground, but once my buildings sat next to the real neighbors that dropped to 22-24%. Sun, daylight and wind are the parts of a design most affected by context, because they depend on what is blocking or channeling them. Some mistakes only showed up with the neighbors in. In Proposal B the tower face pressed against an existing building, and 9% of the facade scored below VSC 5, basically no daylight. Without the neighbor that face would have looked fine. I also found that two neighbor buildings overlap my site limit, which would have inflated my GFA if I had trusted the site totals. Context can also be missing. Forma had no traffic or rail data, so noise showed 100% quiet next to Caltrain, which is wrong. A model is only as honest as the context you put in it.
Should the building adapt to the site, or the site to the building?
◦ From a developer's perspective?
◦ From a designer's perspective?
◦ From the perspective of the neighbors who already live there?
Sun hours and wind frequently want opposite things from the same massing.
◦ How should a team break the tie when two analyses disagree?
Answer: On my site they mostly agreed. A's courtyard got the most winter ground sun (24%) and was also calm enough for sitting (41%). But in general I think the team should agree on priorities before running analyses, based on the program and climate. San Mateo is mild, so for housing I would rank daylight and sun above wind, as long as no area becomes uncomfortable. Setting minimums (like 0% uncomfortable wind, and under 10% of facade below VSC 15) helps, because then you only trade off between options that pass. The designer should make the call in the concept stage since they are balancing everything, but the client and future residents should weigh in before the massing is locked, because they live with the result.
◦ Who should have the final say, and at what point in the process?
Forma's rapid analyses are fast because they are approximate.
◦ What is a rapid operational energy analysis leaving out?
◦ At what stage does that approximation stop being good enough?
◦ What would you need before putting a performance number in front of a client?
You generated three proposals and recommended one.
◦ Did the analysis choose it, or did you choose it and then find numbers that agreed?
Answer: Honestly my first favorite was C, the twin bars, because the gap between them felt open and bright. The numbers did not back that up. C had the lowest share of well-lit facade (78%) and the highest carbon per sf (200). A came out ahead on floor area, winter sun and carbon per sf, so the analysis moved me off my instinct. I would have abandoned A if the courtyard had come out windy or dark, for example more than 10% of the facade under VSC 15 or any uncomfortable wind. If the lot next door gets built out in three years, my recommendation might not hold. A taller neighbor could shade the courtyard, which is A's main advantage. Before committing, I would add a placeholder mass on that lot and rerun sun hours and wind to see if A still wins.
◦ What would you have had to see to abandon your favorite option?
◦ If the lot next door is built out in three years, does your recommendation hold?