Simulation Metric : Point-in-time Sunlight Hours Simulation
Location : Pittsburgh
Performance Objective(s) : Isolate areas >6h of sunlight
Analysis Period : 21st December, 9:00am - 3:00pm
The objective of this exercise was to give a simulation-aided quantitative solution to a landscape designer to avoid spaces in the near vicinity of the given building space that don’t receive adequate hours of incident sunlight during the winter solstice (21st of December). The specific time period under consideration for the decision support was 9:00 am to 3:00 pm.
Landscape Workflow
To execute this workflow, the performance metric used was computationally calculating the sunlight hours for the given date (21st December) and time period (9:00 am to 3:00 pm) and filter regions that receive less than the preferred threshold hours, which was chosen to be 6 hours.
Sunlight-hours simulation is a computer-based calculation of the number of hours of incident sunlight available for a given test site (analysis mesh) under one or several sky conditions. Simulation outputs are given as the number of hours the sun - vectors fall under selected sensor points for a given analysis mesh. The sun-vectors are extracted from the sun-path component based on the different sun positions for the given analysis day and time period.
Initial Simulation
The initial simulation setup was as follows. Using the Ladybug plugin, the Pittsburgh weather file (.EPW file of TMY3) is fed into the sun-path component to extract the sun vectors. Then these vectors are fed into the sunlight hours component and an output false-color mesh containing the sunlight-hours data for the input test points (sensors) is generated. This false color shows the amount of sunlight hours in the vicinity of the given analysis building context.
Site Details
Simulation Interfaces
The initial simulation results yielded 94% of the site area considered to be favorable for landscape design, meaning that these areas received more than 6 hours of sunlight for the given analysis period. The false color representation visualizes these results as the Yellow regions, and the non-favorable spaces as darker colors.
We can notice that the spaces close to the vicinity of the context building show lesser sunlight hours compared to the surrounding areas. This implies that the building acts as an active shader, blocking incoming sunlight throughout the day.
Post this simulation run, the aim was to understand how different orientations of the building context - with respect to the base case orientation - affect the overall area receiving preferable incident sunlight. 8 different orientations were taken into consideration based on passive design principles. The 8 orientations are as follows :
Angles taken for orientation study
Parametric Orientation Study
The same simulation was conducted for the aforementioned orientations and results were recorded.
Some interesting inferences were identified. The areas of the site getting preferable sunlight were greater for angled (Non-Orthogonal) orientations as compared to the Orthogonal orientations. We believe that this is due to the alignment of the building with the sun-vectors for most of the hours of the day, causing minimal intrusion and hence shading of the sun vectors.
This gives us a good idea on the ideal kind of orientation needed for ensuring that a large amount of site area is available with preferable sunlight. It was also identified that in the orthogonal orientations, those with the longer facades facing north-south direction had a slight edge in preferable areas over the other orthogonal orientations. The overall results are summarized below :
Site Area and Orientations
Angles and corresponding results
The next aim was to understand how this simulation result would change or affect design decisions while there are additional context objects in the site. Since it is a landscape design exercise, we chose to go with trees as appropriate context objects for this problem.
To understand the shading effect of context objects together with each other and apart from each other, we chose the following configurations :
The spread out configuration involved placing of context trees in an open fashion in such a way that the apparent shaded areas don’t intersect with each other. The second configuration involved clustering the context trees in such a way that the apparent shading areas intersect with each other.
Sunlight hours simulation was run for both the configurations and the results are recorded as below :
Parametric Context Layout
From the results, it can be inferred that the area of preferred site that receives more than 6h of sunlight for the given analysis period was higher for the clustered context configuration as compared to the spread-out configuration. The reason for this can be attributed to the fact that clustered context intersects the areas shaded by them as compared to spread-out contexts that cover / shade a larger area of the site from incident sunlight.
This can be a great design insight while trying to respond to context landscape as well, while choosing preferable landscaping areas based on sunlight hours
The final analysis that was done on this front is to compare how the simulation results might change when taken for a point in time, vs taken over an annual period.
The results are summarized below :
Point in Time and Annual Study
It can be inferred that the drastic difference in the simulation difference for the two simulation periods is due to averaging out of results over the whole year (through different seasons) for the annual simulation as compared to the point in time simulation. Sun goes through different positions throughout the year and there might be complementary and supplementary positions of the sun that may lead to cancelling out of several shaded regions. This implies that a larger area of the site receives sunlight more than 6 hours as compared to the point in time. This can be evidently seen in the simulation results.