Note: Because I’m working with files that are based on the ones from Module 5, there are some discrepancies for what’s loaded into the file or what’s named what after doing renames. I tried making sure everything was linked properly, but I’m not entirely sure it’s all ok.
- For 2 or More Units: Create Two New Evaluator Nodes
- Initial Build Cost: Based on square footage and height for each floor, the cost per square foot increases exponentially (instead of linearly like in the assignment spec). Also added another output to know the average cost per floor. One weird thing is that the sum inside this node doesn’t work for some reason, and therefore I have to sum outside of the node within my testing node that I have.
- HVAC Costs (approximations): These calculations are very much approximated and take lots of liberties. However, considering the location is in Dubai where it has high temperatures and humidity, HVACs will definitely be important to consider. I output both the initial build cost with assumptions of costs (installation labor per ton, equipment cost per ton, etc.) as well as the operational costs per year (electricity needed, maintenance which is estimated at about 8% of the total initial build cost).
I wanted to try to use Ladybug, but ran into issues that a lot of others had with dependency graphs just not lining up.
For my two nodes, I wanted to look into the construction costs as well as a subset of operational costs for the building that I have designed. I think when evaluating a building design, it is incredibly important to the developer to know how much money it will cost to build but also maintain each year. I don’t have exactly good math or estimates, but I have two sets of node logic:
Results:
Points to Ponder #1: I’m honestly not too sure that my metrics are incredibly useful because they are based in a lot of speculation around the costs. While I did try to do general research into what these costs would look like, it’s very much taking a lot of assumptions into account and I don’t know how well we can see differences. The differences in the costs don’t differ a ton in magnitude but there is a little bit of a difference that is significant enough (in my opinion) to let the developers compare between. I think besides these metrics, it’s definitely useful to look into the weather analysis and solar efficiency especially because I took the assumption of hot weather to be the general and didn’t vary my HVAC values based on that. Some other metrics, non-cost based would be possibly the amount of people it can house depending on the space dedicated towards one person, to account for sizes of companies that may be able to occupy each floor or floors or the whole building.
- For 3 or More Units: Develop a Single-Objective Optimization Scheme
My optimization scheme is as follows:
- Boolean for if the floor area falls within the project spec of 25000000 SF and 3000000 SF with a scaling factor of 2
- Optimal Costs for the Building Cost, HVAC Initial Cost, and HVAC Operational Cost (the lower the better) with scaling factors of 1, 0.5, and 0.2 respectively
In retrospect, the factors that I chose to look at are a bit one dimensional as both the building cost and the HVAC costs are a function of the floor surface area, and so the optimal result will likely choose the cheapest option that has the smallest floor surface area within that range. Everything is linearly related and so realistically, the relationship between the scaling factors aren’t the most important except the boolean so that the floor area falls within the acceptable range.
Out of these options, it’s difficult to say what’s the best in my opinion, again because it’s so one-dimensional, but I think I would choose the 5th option (first one highlighted red) because it falls within the middle of the floor area requirement range while also giving costs that are reasonably optimized for.
As for the points to ponder, there are most definitely a lot of nuances lost in this single evaluation. I don’t really have diverse enough metrics to create something isn’t looking a minimum that falls within a range because everything is a function of practically one option.