Create Two New Evaluator Nodes
For the first evaluator node, I decided to do an energy comparison from the solar insolation of the building against average energy consumption usages, with options to flex the model for different seasons and operating cases. Basically, the node takes in a season and a use case and then determines an estimate of the additional energy requirement. This means that a negative value outputted would indicate the building is producing more energy than it is consuming. The average energy usage is based on data provided by Energy STAR for the median usage of US buildings as I could not find reliable data for Dubai. https://portfoliomanager.energystar.gov/pdf/reference/US National Median Table.pdf
The second node calculated the total construction cost of the project based on the square footage at various heights along the project using the equation outlined in the module. I originally thought to calculate the variation due to height from the story height, but to allow for more adaption in the design, I used the level elevations to determine the height of the various mass floors.
Point to Ponder:
Both the construction costs and energy requirements depict meaningful differences between the cases, showing how some are more energy efficient as compared to other cases as well as relative cost between alternatives. Other metrics that could be helpful include constructability / build time and a measure of the sightline throughout the building. For constructability, optimized designs can look ideal on paper, but if the cost to build it is exponentially higher than similar alternatives, this makes the option infeasible. For the sightline, it is often important to residents and office spaces that a nice view is presented by the building, so it may be ideal to the owner to have a building that maximizes the views within it.
Develop a Single-Objective Optimization Scheme
Step 1 Point to Ponder:
I believe a weighted sum combination is the optimal strategy to capture the relationship between evaluation metrics. I was thinking that I would take the maximum of each value over the considered cases to normalize each specific case to a factor of 1, and then I would apply a scaling factor of relative importance to these values. Yet, I then started to consider how some values would need to be maximize and minimized. This then led to a consideration of dividing the maximum by the respective values, and then to normalize these values, they would be divided by the maximum in this segment. With this normalization in mind, I then turned to determining the weights I would apply to the various metrics, including whether it was desirable to minimize or maximize them. The gross volume would sought to be minimized to reduce the impact on the building, but this is a relatively low consideration leading to a weighting of just 1. For the gross surface area, the building performance objective seeks to minimize this, but I believe this to be a minor concern as compared to others, so I will apply a scaling of 1.5. For the gross floor area, it is likely desired to maximize this characteristic, and as this determines the usability of the space, I will give it a scale factor of 2. This in then related to the construction cost which is desired to be minimized. As construction cost governs the viability of the project, I will apply a scaling factor of 2.5. Lastly, the energy requirements tie in with the surface area of the building, so I will once again apply a scaling factor of 1.5. The solar insolation is desired to be maximized, but as the energy requirement subtracts the solar insolation, this value should be minimized in this optimization.
The top three design alternatives are: a base rotation of 10 degrees, 20 degrees, and 0 degrees.
Step 2 Point to Ponder:
The recommended alternative was driven to the top of the list by the construction cost as this had the highest scaling factor. While the third alternative successfully minimized gross volume, gross surface area, and energy requirements, these were all relatively low priority factors in the analysis. The construction costs of this case were relatively high with the smallest gross floor area, causing it to be the third best choice even though it performed the best in 3 categories. The second alternative was slightly less optimized in the gross volume, gross surface area, and energy requirements as compared to the #1 alternative, but it did have slightly more floor area. Ultimately, the better optimization and more efficient construction costs of the #1 option caused it to win in this optimization. This analysis is heavily dependent on the relative scaling, so if other metrics or priorities were incorporated, these alternatives could easily change in their ordering. This single evaluation fails to capture the nuance of the relation between the various factors. It also isn’t able to capture ranges of which a value are consider of equal desirability, instead leading to a linear scaling instead of target ranges that could be equally wanted. It also isn’t able to depict the exact tradeoffs that occur between variables and instead just outputs a ranking result. A more detailed evaluation could provide the client with ways to optimize the building through a combination of the desired traits tested.