Hello,
This is more theoretical than technical question. Sorry for the lenght of this post but I didn’t know how to explain it in few words.
I am on early stage of design of office buildings complex. The shape and orientation of each building is known, I would like to decide about glazing and insulation of the facades. As I am using iterative algorithms and the area is huge there is no way I would be able to analyze all individual zones (rooms) for daylight and energy use (loads).
The rooms in the building are quite repetetive. To make things simple I wanted to divide “facade modules” (corresponding to rooms) into (let’s say) 5 groups that would be simmilar. I thought about using radiation analysis for that. Let’s say that throughout the building facade radiation falls in range between 0 and 1000. I could divide it into 5 groups (0-200;200-400 etc.). After that I would pick one representative room for each of these groups and optimize it (f.ex. DA must be above x%, minimize heating/cooling loads). When that is done i would place these “facade modules”, according to their group all around the building.
I wanted to know if that workflow is ok, considering this is early stage and there is no need to be clinical when it comes to analysis? As I understand that radiation is related to gains (so loads as well) I am not sure if this is fine to make this assumption for daylight as well. For example (picture below) - southern facades close to the ground would fall in the same group as eastern/western facades of high rise, and southern ground floor recieves simmilar radiation to northern facade etc.
As I am inexperienced user, I’d be grateful for your thoughts on that method.
Just tackling the energy part here, I think there are a couple of different assumptions that need to be clarified, and refined for this to work, such as what are your assumptions about program type, schedules, ventilation/infiltration, or what are your zone geometries are to get heating/cooling loads.
That being said, I think a version of the radiation analysis your suggesting, to “bin” thermal zones, and use them to simulate heating/cooling loads, can be step one of an early stage design breakdown. Specifically, it reminds me of Zone Multipliers in EnergyPlus or Timur Dogan’s Shoeboxer algorithm, which is also meant for early stage, or large scale energy analysis.
The shoeboxer algorithm works by numerically identifying a sample set of thermal zone shoe-boxes within a larger building, based on a unique configuration of: solar insolation (using Radiance), orientation, height, shading, perimeter and core zone shapes. The energy simulation is then run on these samples. Finally using an area-weighted average the energy demand is multiplied out to the full scale of the building. The key thing here is that the shoebox samples he’s generating can be binned or clustered because it shares similar external and internal loads.
Once you have that figured out you can figure out some schematic optimization for each sample (or in your case a “facade module”). However, without more detailed HVAC and space-type information there’s a limit too how much optimizing you can do…
So check out Chapter 4 of Timur’s thesis, and Zone Multipliers. I also recommend building a proper energy model, and use that to run multiple simulations against your early stage workflow. That way you can iterate your workflow to see how well your “modularization” resembles actual energy loads in that area.