Genevieve Gandara

A very brief description of the design decisions from Step 1 following the Generative Design Framework.

• Design Decision 1: Sustainability
• Design Variables
• Floor height
• Floor area per story
• daylight
• structural materials
• Evaluators
• embodied carbon
• operational carbon
• wellbeing
• cost
• Most Important Tradeoffs to Consider
• embodied carbon vs. operational carbon
• state-of-the-art design vs. budget
• Design Decision 2: Energy Efficiency
• Design Variables
• base radius
• story height
• insulation thickness
• top rotation
• Evaluators
• Total Capital Cost
• PV electricity cost savings
• Daylit floor area
• Thermal efficiency
• Most Important Tradeoffs to Consider
• Upfront costs vs. long-term savings
• Design Decision: Daylighting
• Design Variables
• Window-to-wall ratio
• Glazing type
• Shading depth
• Building orientation
• Evaluators
• Daylit floor area
• Glare risk
• occupant comfort
• operational cost
• Most Important Tradeoffs to Consider
• Increasing daylight access can improve wellbeing, but it may also increase glare and energy loads.
• Reducing floor plate depth can improve daylighting, but it may reduce usable or rentable floor area.

• I chose to focus my optimization scheme on energy efficiency and daylighting. The main tradeoffs are upfront costs vs. long-term savings. For example, paying more upfront for PV panels and thick insulation is expensive, but leads to operational energy savings over time. Thus, my variables non-trivially interact with each other by incentivizing savings in the upfront capital cost, while also incentivizing state-of-the-art technology and operational savings.
• The scheme maximizes PV electricity cost savings, but the tradeoff is the added cost to the total cost. The scheme also maximizes daylit floor area, but this leads to a narrower building and less roof space for the PV panels. The scheme maximizes the insulation score, which costs more upfront but will lead to operational energy savings, as the building uses less heating and cooling. The total cost is the sum of the capital cost, PV panels cost, and insulation material cost. The total cost is minimized.
• A genetic algorithm with a population of 20 and 10 generations was used to evaluate the optimization scheme.

The optimization results show that increasing daylit area generally increases PV electricity cost savings, but the relationship is clustered rather than continuous. This suggests that both evaluators are partially driven by building size: options with larger base radii tend to have more roof area for PV and more floor area within the 30 ft daylight zone. However, the horizontal bands show that PV savings are also being affected by another discrete factor, likely tower story height, which was only evaluated at discrete values (10, 11 ,12 ,13 ,14, and 15 ft), which causes groups of designs to have similar PV performance even when their daylit area differs. The best-performing options appear in the upper-right region of the graph because they combine high daylit area with high PV savings.

In the set of optimal solutions, as total cost increases, so does daylit area. Both are functions of the building’s radius, so they likely both increase as radius increases.

In the set of optimal solutions, there is virtually no correlation between insulation score and total cost, and insulation score and daylit area. This suggests that insulation thickness is acting as an independent design variable rather than being strongly controlled by the massing variables, such as base radius, story height, or top rotation. In other words, a design can have a high insulation score without necessarily being more expensive overall or having more/less daylit area. Although insulation thickness contributes to total cost, its cost impact is small compared to other cost drivers, especially building floor area. Insulation score and daylit area are measuring two very different aspects of the design. Daylit area is mostly controlled by geometry, especially the base radius and number of floors, while insulation score is controlled by insulation thickness. Because these variables do not strongly affect each other in the model, there is no clear tradeoff between improving insulation and improving daylight access.

The 12-generation run appears to have found more high performing solutions than the 10-generation run solution set. In the 12 generation run, there are stronger patterns for optimal variables compared to the 10 generation run. For example, many of the optimal solutions prefer a shorter story height, a top rotation greater than 40 degrees, and an insulation thickness greater than 3 inches.

A sample of geometries for optimal solutions.

• An image of your Dynamo Study Graph.