I’m working on comparing the effect of tinting speed on energy use and visual comfort for different dynamic glass facades. The main problem i’m facing is that electrochromic windows usually switch states over the span of about 20-30 minutes, while another dynamic glass material - “PD-LC” switches almost instantaneously.
Is there a way to implement this? The way i understand EnergyPlus and the HB Window Construction Shade component, the shading or tinted state will activate instantly when the setpoint for the control strategy is reached(in this case OnIfHighZoneAirTemperature).
Is there a way to ‘‘delay’’ the activation of the shading state to a timestep about 30 minutes after the setpoint condition is reached?
You’re touching on one of the principal difficulties in simulating dynamic glazing systems - most energy simulation software packages assume immediate changing of the shading state, but actual electrochromic windows change over 20-30 minutes. For your comparison of electrochromic to PD-LC glass, I would suggest using a custom EnergyPlus EMS script that adds a time lag. In effect, you’d set the shading control to only engage 30 minutes after reaching temperature setpoint, which better relates to electrochromic response. For PD-LC glass (which changes immediately), you’d use regular HB controls but introduce a 1-timestep delay in order to prevent numerical instability within the simulation. This method is especially relevant in climates such as Ponte Vedra, FL where high rates of changing sun exposure render transition times crucial to precise energy modeling. The EMS technique remains EnergyPlus compatible with providing you with realistic side-by-side comparisons of how each technology operates with their true switching characteristics.
I got around the issue of tinting rate by running a test simulation with the HB Window Construction Shade component, and afterwards running the actual ‘electrochromic’ simulation by constructing a transmittance schedule for the HB Shade component from the zone air-temperature outputs of the test simulation run.
This worked well for emulating linear tinting over 30 minutes, but i have run into 2 additional issues:
ChatGPT claims that the HB Shade component(or at least similar code blocks in different energy simulation softwares, i think it was DesignBuilder) blocks all IR-radiation, regardless of the transmittance state of the HB Shade, even if the HB Shade is supposed to have 100% transmittance. Do you know if this is true in Honeybee as well?
Another issue that has since come up is the fact that electrochromic windows and PDLCs can have different transmittance wavelength spectrums/fingerprints in the clear and dark states.
Without actually having attempted it yet, i assume it is possible to create 30 window constructions with different T_sol and T_vis values(one for each minute during transition between fully clear and tinted states) and apply them based on a schedule with the ‘HB Window Construction Dynamic’ component, but i think this approach could be a simplification of the situation shown in the picture below:
