|Title||Control algorithms for dynamic windows for residential buildings|
|Publication Type||Journal Article|
|Year of Publication||2015|
|Authors||Mehry Yazdanian, D. Charlie Curcija, Christian Kohler, Dragan Vidanovic, Robert Hart, Stephen Czarnecki|
|Journal||Energy and Buildings|
The present study analyzes the influence of control algorithms for dynamic windows on energy consumption, daylight access and shade operations in residential buildings. Five different control algorithms – heating/cooling, simple rules, perfect citizen, heat flow and predictive weather were developed and compared. The proposed algorithms can work with any window, not only dynamic - standard window with dynamic attachment or window with dynamic glazing, in new or renovated buildings. Results of the calculations were compared with base cases – no shade, always shaded, half shade and observed manual shade operations.Evaluation of different control algorithms for dynamic windows was based on whole building energy simulation. The performance of a typical residential building was modelled with EnergyPlus. The program Widow was used to generate a Bi-Directional Distribution Function (BSDF) for two window configurations – double low solar gain with external roller blinds and triple glass high solar gain with between glass cellular shading. The BSDF was exported to EnergyPlus using the IDF file format. The Energy Management System (EMS) feature in EnergyPlus was used to develop custom control algorithms.The calculations were made only for USA but they include four locations with diverse climates. Atlanta has a humid subtropical climate (hot and humid summers and cool but variable winters), Phoenix has a subtropical desert climate (extremely hot summers and warm winters), Minneapolis has a continental climate (winters are cold and snowy, while summers are mild and can be humid) and Washington DC is in the humid subtropical climate (spring and fall are warm, winter is cool, summers are hot and humid).The results showed that: a) manual operation of shade has on average no effect on site (final) energy consumption in comparison to windows without shade; use of automated shading with proposed control algorithms can reduce on average the site energy in the range of 11.6% to 13.0%; in regard to source (primary) energy, manual operation of shade reduces on average the consumption by 8.6%, while automatically controlled in the range of 20.1% to 21.6%, b) automatic shade operation is more effective in cooling dominated climates, c) the differences between algorithms in regard to energy savings are not high, d) use of windows with low U-value and high SHGC is not appropriate in all climates, e) the differences between algorithms in regard to daylight access are visible, f) the control algorithms have a strong influence on shade operation and oscillation of shade can occur, g) additional energy consumption caused by motor, sensors and a small microprocessor in the analyzed case is very small.