Sage Journals: Discover world-class research

Abstract

Commercial buildings in Canada consume 1057 PJ of energy annually, with space heating and cooling responsible for 61% of this energy use. Windows play a crucial role in building energy efficiency, but often contribute to significant thermal losses. Automated shading attachments offer a promising retrofit strategy for enhancing building energy efficiency in commercial buildings. The current study uses EnergyPlus and WINDOW to assess the annual energy performance of four internal shading attachments in a US Department of Energy-defined small office building in Canadian cold-climate zones. Six distinct control strategies are analyzed to optimize heating and cooling reductions across five Canadian cities spanning the five climate zones of Canada. The window-shade systems are modelled as Complex Fenestration Systems on WINDOW and integrated into EnergyPlus. The Energy Management System Feature is used to implement custom control strategies. Intermediate shading positions are studied, demonstrating a more refined approach towards automated shading. Findings indicate that a maximum of 22% reduction in the annual heating loads is observed with Shade 2 (Cellular Shade) for the case of Vancouver (ASHRAE 4C) with Control Strategy 3 (CS3), and a maximum of 47% reduction in cooling loads is observed for the case of Yellowknife (ASHRAE 8) with Shade 1 (BO Roller Shade) and CS4 (T_ave). CS4 emerges as an effective control strategy for optimizing heating and cooling load reductions, with heating energy demand reductions in the range of 7.04%–19.69%, compared to 4.14%–17.20% using CS4 (T_ave). However, no improvement in energy performance was observed with the use of intermediate shading positions. These findings support the use of season-specific control strategies for automated shading as an effective means of reducing heating and cooling loads in cold-climate zones.

Keywords

automated shading building energy efficiency cold climate commercial buildings load reduction

Introduction

According to Natural Resources Canada (2024a), commercial buildings in Canada consume 1057 PJ of energy annually. Recent reports (Natural Resources Canada O of EE, 2020) indicate that space heating and cooling account for 61% of this energy use. With the increasing shift toward glazing-dominated architecture in commercial buildings (Shum and Zhong, 2023b), this energy consumption is likely to rise, as windows are typically the weakest components of a building envelope (Tan et al., 2020) in terms of thermal performance. Despite their energy implications, windows offer numerous benefits, especially in commercial office buildings. A study by Boubekri et al. (2014) found that workers in offices with windows reported higher quality sleep and increased physical activity compared to those in offices with no windows, factors which can improve overall productivity (Haynes, 2008). These advantages make windows inevitable components of the building envelope and justify the design choice of increasing the glazing area in office buildings.

However, this architectural preference poses a significant challenge for balancing energy efficiency with occupant comfort and well-being, particularly in cold-climate zones across much of Canada. Buildings with high glazing areas are prone to overheating due to solar exposure (Shum and Zhong, 2023b) and excessive nighttime heat losses, both of which increase energy demands for cooling and heating. This dichotomy underscores the need for innovative solutions that can maintain the benefits of expansive glazing while mitigating its energy costs. Upgrading to windows with better thermal performance is a good solution to this problem, but the disruptive nature and associated costs mean that it is not always feasible (Kunwar and Bhandari, 2020).

A more feasible alternative lies in the strategic use of window attachments as a less disruptive retrofit strategy. While primarily employed for privacy and visual comfort in both commercial and residential settings, these devices can also play a crucial role in reducing energy consumption. By decreasing thermal transmittance and the solar heat gain coefficient (SHGC), attachments such as roller shades, cellular shades and venetian blinds can effectively lower both heating and cooling energy requirements. In an analytical study, researchers (Oleskowicz-Popiel and Sobczak, 2014) concluded that external and internal roller shades can achieve heating energy reductions of 45% and 33%, respectively. However, these findings are limited to nighttime energy savings and do not provide comprehensive insights into daytime energy performance.

Whole-year energy simulations, on the other hand, usually performed on EnergyPlus (U.S. Department of Energy, 2023), give a more comprehensive picture of the energy benefits of shading attachments. In their study, Kunwar et al. (2022) analyzed the energy performance of cellular shades using both experimental testing and EnergyPlus simulations, quantifying a 17%–36% improvement in the heating performance of cellular shades. On the basis of whole-year simulations performed on EnergyPlus, Tan et al. (2020) proposed an Annual Energy Performance Index to rate the energy-saving potential of window attachments. These provide a benchmark for understanding the energy performance of shading attachments.

While this rating highlights the potential of shading attachments to save energy in heating and cooling, it does not account for their dynamic applications. Research suggests that static shade operation does not necessarily reduce energy consumption. In fact, Firlag et al. (2015) found that manually operated shades have a negative impact on energy consumption, resulting in higher heating energy use than unshaded windows. Also, a study on occupant behavior (Nezamdoost et al., 2018) found that the shade positions were not adjusted 50% of the time over an observation period of around 2 weeks, possibly weakening the energy performance. This points to the need for automated control of shading attachments to maximize energy efficiency.

Effective use of automated shading is influenced by the type of control strategy used to trigger the operation of the shade. In terms of control strategies for shading attachments, priority is given to visual comfort in office spaces, as strategies focused on energy efficiency mostly target residential buildings (Firląg et al., 2015). da Silva et al. (2012) studied 11 different control strategies for an office, prioritizing occupant visual comfort in the hot climate of Porto and found that the control strategy based on Daylight Glare Index would be most suitable for office spaces. Researchers (Tzempelikos and Shen, 2013) identified transmitted illuminance as a good metric to control shading operation and suggested that different strategies are suitable for windows in different orientations.

Much research has been conducted on automated shading systems for warm-climate zones, with the primary objective of reducing cooling energy consumption. However, a research gap exists regarding the application of automated shading systems and strategies in cold-climate zones (Shum and Zhong, 2023b). In their study, some authors (Grynning et al., 2014) recognized overheating as a major issue in office buildings, supported by their finding that, even in cold climates, cooling demand dominates total energy demand. Researchers (Shum and Zhong, 2023b) studied the effects of different control strategies on the energy consumption of an office space with south- and west-facing windows across five Canadian cities, with a primary focus on addressing overheating in office buildings. They attained a maximum reduction of 81% in cooling energy. However, they identified the need for strategies that maximize heating-load reduction while not compromising cooling-energy reduction. This, along with the fact that energy consumption in Canadian cold climate zones is predominantly influenced by heating energy demand (Shum and Zhong, 2023a) points toward the need for shading control strategies that optimize both heating and cooling energy reduction. Additionally, a limited number of studies have investigated the effect of automated shading control attachments on whole office spaces with windows in all orientations.

Most studies conducted in the past have focused on energy and daylight performance at extreme shading positions (fully unshaded or fully shaded). In actual scenarios, intermediate shading positions are frequently observed. Tzempelikos and Shen (2013) studied a simple shading control strategy that adjusts the shade position to avoid direct sunlight and found significant improvements in cooling and lighting energy. A study by Kunwar and Bhandari (2020)identified the need to study partially shaded positions and aimed to model intermediate shading conditions with roller shades to better reflect practical scenarios. While they achieved maximum reductions in cooling and lighting energy of 40% and 35%, respectively, they concluded that their control strategies were better suited to cooling-dominated climates. Developing control strategies to enhance heating energy performance in cold-climate zones remains a relatively unexplored yet very significant topic in Canada, where heating loads dominate.

While previous studies have made significant contributions to understanding automated shading systems in cold climates, there remains a critical gap in research on whole-building energy performance and the potential of intermediate shading positions to optimize heating performance, particularly in cold-climate zones in Canada. This study aims to fill this gap by:

Examining the impact of automated shading strategies on both heating and cooling energy demands in cold climate zones of Canada, addressing the need for a balanced approach identified by Shum and Zhong (2023b).

Evaluating four different commercially available shading attachments in conjunction with six distinct control strategies, providing a comprehensive analysis of their performance in a US Department of Energy (US DOE)-defined small office building (DOE & PNNL, 2023).

Exploring the energy-saving potential of intermediate shading positions for enhancing heating energy performance, an aspect that, while studied in literature, has been largely overlooked in cold climate applications. This approach allows for more nuanced control and potentially greater energy savings.

Offering insights into selecting the most energy-efficient control strategies for specific climate zones and shading attachments, considering the whole-building energy performance rather than focusing solely on individual spaces or facades.

By addressing these aspects, this study extends the previous work (Chhetri and Zhong, 2024; Shum and Zhong, 2023b), providing a more holistic understanding of automated shading systems in cold climates. The inclusion of intermediate shading positions to reduce heating loads and the extension of whole-building analysis to cold-climate zones in Canada represent significant advancements in the field, offering practical implications for building designers and energy managers in cold-climate regions.

Methodology

Setup

To study the influence of various control strategies and shading attachments on the yearly heating and cooling energy demands, year-round whole-building simulations were performed using EnergyPlus (version 23.2) on a prototype US DOE small office building based on IECC 2021 (DOE & PNNL, 2023). Figure 1 shows the schematic of the small office building and its top view, demonstrating the five different thermal zones. It consists of four perimeter zones, facing each of South, East, North and West. The north and south perimeter zones have six windows each, while the east and west perimeter zones have four windows each. Each of the 20 windows has dimensions of 1520 mm × 1830 mm. The building has a floor area of 511 m² and a window-to-wall ratio of 21%, which is consistent for all climate zones and is compliant with the maximum allowable WWR of 40% across Canada according to the National Energy Code for Buildings (National Research Council of Canada, 2020).

Figure 1.

Schematic of US DOE small office building.

The prototype building model was obtained from the US DOE website and imported to EnergyPlus (see Table 1). For this study, an ideal air-load system was used for all simulation cases. The ideal air loads system provides the heating and cooling energy required for maintaining a particular thermal zone at the defined setpoint temperatures (EnergyPlus Development Team, 2023) Thus, it effectively represents the demand for heating and cooling energy. The terms cooling and heating loads and cooling and heating demands have been used interchangeably throughout this manuscript. The study of energy demands provides a consistent basis for comparing the different shading attachments, control strategies, and climate zones.

Table 1.

Parameters for EnergyPlus simulation.

Parameters	Values
Building dimensions	27.69 m × 18.46 m × 3.05 m, Gross WWR = 21.9%
Window dimensions	1.52 m × 1.83 m
Baseline window	Double Pane Clear Glass (modelled on Window) U = 2.644 W/m²K SHGC = 0.704, T_vis = 0.786
Envelope	Climate Zone Specific according to IECC 2021 (International Code Council, 2021)
Internal loads	According to IECC 2021 (International Code Council, 2021))
Occupancy	Occupancy Schedule according to IECC 2021 (International Code Council, 2021)
Weather data	All typical meteorological years (TMY3) for Vancouver, Toronto, Quebec City, Edmonton, and Yellowknife
HVAC system	Ideal Air Loads Cooling Setpoint: 22°C (occupied), 26°C (unoccupied) Heating Setpoint: 20°C (occupied), 18°C (unoccupied) according to NECB (National Research Council of Canada, 2020)
Calculation tool	EnergyPlus version 23.2
Building code	IECC 2021 (International Code Council, 2021)

The simulations were performed in five different Canadian cities: Vancouver, Toronto, Quebec City, Edmonton, and Yellowknife, encompassing all the climate zones in Canada. All these cities fall under cold climate zones, defined by ASHRAE as heating-dominated zones with a higher number of Heating Degree Days than Cooling Degree Days (ASHRAE, 2021). Table 2 lists the climate zones of all the cities.

Table 2.

Climate properties for selected cities.

City	Climate Zone, ASHRAE (ASHRAE, 2021)	HDD (18.3°C) (ASHRAE, 2021)	CDD (18.3°C) (ASHRAE, 2021)
Vancouver	Zone 4C (Mixed Marine)	2873	49
Toronto	Zone 5A (Cool-Humid)	3779	339
Quebec City	Zone 6A (Cold-Humid)	4810	162
Edmonton	Zone 7A (Very Cold)	5209	77
Yellowknife	Zone 8 (Subarctic)	8051	38

Window model

This study used a clear-glass double-pane window (5.7 mm thick and 13 mm air gap) as the baseline window. While this window exceeds the maximum allowable U-value according to the latest version of the National Energy Code for Canada (2020), approximately 40% of commercial and institutional buildings in Canada were built more than 50 years ago (Natural Resources Canada, 2024b), during a period of more lenient building code requirements. Thus, it is a good representative of windows used in existing buildings in Canada.

Window models were created in WINDOW 7.8 (Lawrence Berkeley National Laboratory, 2023b), developed by Lawrence Berkeley National Laboratory to model windows and shading attachments. WINDOW creates a Bidirectional Scattering Distribution Function (BSDF) that can be imported into EnergyPlus as an Input Data File (IDF). BSDF files characterize the optical performance of a system through a defined set of incident and outgoing angles. The method represents each layer and the entire system using a matrix that correlates incident and outgoing angles, effectively capturing the system’s light-scattering properties. As a result, the window modeled in WINDOW is imported into EnergyPlus as a Complex Fenestration State, incorporating the BSDF matrix of incident and outgoing angles to ensure an accurate representation of optical interactions.

Table 3 lists the properties of the window shade system modeled in WINDOW 7.8. The Baseline window has a U-value of 2.644 W/m²K and SHGC of 0.704, and the presence of internal shading attachments alters these values. The blackout roller shade does not allow light to pass through it and has a visible transmittance value of 0. Among the other shades modeled, the roller shade with 10% openness has the highest visible transmittance and a higher SHGC, indicating that it allows more light to pass through. These shading materials are obtained from the Complex Glazing Database (Lawrence Berkeley National Laboratory, 2023a), which is built into the software.

Table 3.

Shade properties.

Window system	Shade	Thickness (mm)	Keff (W/mK)	U-value (W/m²K)	SHGC	T_vis
Double-pane window, clear glass, window-shade distance: 35 mm	Shade 1 (Blackout Roller Shade)	0.6	0.152	1.718	0.22	0
	Shade 2 (Cellular Shade)	20	0.134	1.292	0.29	0.21
	Shade 3 (Roller Shade 3% Openness)	0.5	0.114	1.436	0.257	0.042
	Shade 4 (Roller Shade 10%)	1	0.166	1.986	0.335	0.234

Description of control strategies

This study examined four distinct control strategies, along with two baseline cases: Unshaded and Fully Shaded. In EnergyPlus, the distinction between shaded and unshaded positions is made by switching between two different constructions of the Complex Fenestration State: one with shade and the other without shade. These constructions were created using the IDF file imported from WINDOW, which contains all relevant information, including the optical matrices and thermal properties of the window glass and the shading material. The control strategies described below are implemented through the Energy Management System feature on EnergyPlus, which allows users to develop custom control algorithms.

Baseline case

There are two baseline scenarios considered in this study: Unshaded and Always Shaded. In the first scenario, the simulations are performed with all the windows in unshaded positions. In the second case, all the windows are always shaded with no control strategy in place.

Control strategy 1 (CS1)

In this case, the shades are closed when solar radiation falling on the window exceeds 150 W/m²; otherwise, the windows remain unshaded. The World Meteorological Organization defines 120 W/m² as the solar irradiance lower limit to define the start and end times for calculating sunlight duration (World Meteorological Organization, 2021). Thus, 150 W/m² serves as an acceptable lower limit for activating shade operation. This strategy was previously studied by Shum and Zhong (2023b).

Control strategy 2 (CS2)

This strategy is similar to CS 1 in that the shades are closed when the solar radiation falling on the window exceeds 150 W/m². However, the shades are closed overnight, reducing heat losses due to the lower U-value in the unshaded position. This has also been previously studied by Shum and Zhong (2023b).

Control strategy 3 (CS3)

CS1 and CS2 primarily aim to reduce cooling energy consumption, as the shade operation is triggered when the solar radiation falling on the windows exceeds a certain threshold. In Canada’s cold-climate zones, to reduce heating loads, a more season-specific control strategy is needed. As such, a novel seasonal adaptive control strategy was proposed:

Summer (May to September): From May to September, the shades are closed during the daytime if solar radiation exceeds 150 W/m² and are kept in a shaded position overnight. In cold-climate zones in Canada, nighttime temperatures are typically lower than the indoor setpoint temperatures (22°C for cooling and 20°C for heating) used in this study. Due to the absence of solar radiation at night, the primary heat transfer mechanism is conductive and convective heat loss from the system. Thus, keeping the windows in a shaded position is beneficial because the U-value is lower than in the unshaded position. This reduces heat losses and, consequently, heating loads.

Winter (October to April): From October to April, the shades are retracted during the day only if solar radiation exceeds 150 W/m²; otherwise, they are kept shaded, and they are kept in the shaded position overnight. Retracting the shades when there is high solar radiation allows for passive solar gain, which can reduce the heating loads.

This strategy focuses on maximizing heating-load reduction in the winter and cooling-load reduction in the summer. The solar radiation thresholds were kept the same for all the cities to ensure consistency in comparison among different shading attachments and control strategies.

Control strategy 4 (CS4)

This control strategy is based on energy balance equations. The steady-state energy flow through a fenestration, neglecting the air leakage term, is:

\begin{matrix} q = U (T_{out} - T_{in}) + SHGC \times I \end{matrix}

(1)

where q = instantaneous energy flow per unit area (W/m²), U = thermal transmittance (U-value; W/m²K), T_in = indoor air temperature (°C), T_out = outdoor air temperature (°C), SHGC = solar heat gain coefficient, I = incident total irradiance (W/m²).

U and SHGC vary by shade position. The Lawrence Berkeley National Laboratory recommends WINDOW (Lawrence Berkeley National Laboratory, 2023b) as the standard calculation method for determining U-value. While the U-value of a window can vary according to boundary conditions, the calculations in WINDOW are made for a standard set of boundary conditions defined by the National Fenestration Rating Council (NFRC, 2017), and the U-values described in Tables 1 and 2 are calculated under these standard conditions.

For the two different positions, open (up) and closed (down), two different equations can be written:

\begin{matrix} q_{up} = U_{up} (T_{out} - T_{in}) + SHG C_{up} \times I \end{matrix}

(2)

and

\begin{matrix} q_{down} = U_{down} (T_{out} - T_{in}) + SHG C_{down} \times I \end{matrix}

(3)

Equations (2) and (3) describe the heat transfer through the window. In the heating season, the first term on the right-hand side of the above equations is typically negative, as the indoor temperature T_in is generally higher than the outdoor temperature T_out. Conversely, the second term remains consistently positive. On days with high solar irradiance on the windows, radiative heat gains may exceed convective and conductive losses, allowing the shades to be opened to enable passive solar heating.

This control strategy consists of three subcategories, each differing in how the shade position is adjusted when Tin falls between the heating and cooling setpoints.

Primary method

Figure 2 represents the control logic for the primary method. The input parameters required for this logic include U_up, U_down, T_in, T_out, U_up, U_down, T_heat, T_cool, SHGC_up, SHGC_down, and I. The Energy Management System was used to develop and implement this control strategy. The primary goal is to minimize heating and cooling loads by optimizing net heat transfer through the window.

Figure 2.

Control logic for CS4.

In the primary method, when the indoor temperature T_in < the heating setpoint T_heat (20°C), the system enters heating mode. In such a case, the goal would be to minimize heating demand by adjusting the shade position to maximize heat gain into the room or minimize heat loss through the window.

When T_in > T_heat (22°C), the system enters cooling mode. Then, the goal would be to minimize cooling loads by adjusting the shade position to reduce heat gain into the room.

When T_heat < T_in < T_cool, the system is neither in heating nor cooling mode. Under these conditions, the shade position remains unchanged, maintaining its previous state regardless of whether it is open or closed.

Average method

The fundamental control logic remains unchanged for the average method (see Figure 3); however, in this approach, an average temperature T_ave is calculated, which is 21°C in this case. If T_in ≤ T_ave, the room is considered to be in or approaching heating mode. Consequently, the shade is set to a position that maximizes net heat flow into the room, thereby reducing heating loads. If T_in > T_ave, the shade is positioned to minimize the net heat flow into the room, thereby reducing the cooling load. A study (Firląg et al., 2015) explored this control algorithm and found that it effectively optimizes energy consumption for heating and cooling in residential buildings (Figure 3).

Figure 3.

Control logic for CS4 (T_ave).

Partial shading method

The partial shading method builds upon the primary method by incorporating intermediate shading positions, enabling partial shading of the windows. Three intermediate shading levels, 25% shaded, 50% shaded, and 75% shaded, are considered in this approach. A simplified calculation method is proposed for determining the U-value and SHGC at these intermediate positions.

The U-value and SHGC at intermediate shading positions can be calculated using equations (4) and (5), respectively:

U_{partially Shaded} = \frac{U_{shaded} A_{shaded} + U_{unshaded} A_{unshaded}}{A_{shaded} + A_{unshaded}}

(4)

SHG C_{partially Shaded} = \frac{SHG C_{shaded} A_{shaded} + SHG C_{unshaded} A_{unshaded}}{A_{shaded} + A_{unshaded}}

(5)

Since EnergyPlus does not inherently support partial shading of windows using shading attachments, intermediate shading positions cannot be modeled directly. To address this limitation, a workaround is implemented by dividing each window into four equal sections, keeping the total window area unchanged.

For instance, a baseline window with dimensions 1.52 m × 1.83 m is subdivided into four equal sections, each measuring 0.38 m × 1.83 m. The U-value and SHGC of the total window are then estimated as the area-weighted averages of the unshaded and shaded portions.

As an example, under the 25% shaded condition, two separate windows were modeled in WINDOW 7.8: a 0.38 m × 1.83 m window with an interior shade and a 1.14 m × 1.83 m window with no shade. A similar methodology was applied to determine the U-value and SHGC for other intermediate shading positions. The overall U-value and SHGC of the window-shade system were then approximated as the area-weighted average of these two configurations. Table 4 represents the U-values and SHGC for the intermediate shading positions.

Table 4.

U-value and SHGC at intermediate shading positions.

Position	Blackout roller shades		Cellular shade		Roller shade (3%)		Roller shade (10%)
Position	U-value (W/m²K)	SHGC	U-value (W/m²K)	SHGC	U-value (W/m²K)	SHGC	U-value (W/m²K)	SHGC
Open	2.64	0.70	2.64	0.70	2.64	0.70	2.64	0.70
25% shaded	2.45	0.58	2.33	0.60	2.37	0.59	2.50	0.61
50% shaded	2.23	0.46	2.01	0.50	2.08	0.48	2.34	0.52
75% shaded	2.00	0.34	1.67	0.39	1.77	0.37	2.18	0.43
Fully shaded	1.72	0.22	1.29	0.29	1.44	0.26	1.99	0.34

Table 4 provides a framework for analyzing intermediate shading positions. At any given timestep during the simulation, the optimal shading position is determined based on the indoor temperature. The shading is adjusted to maximize net heat flow into the room when T_in ≤ T_heat and to minimize net heat flow into the room when T_in ≥ T_cool. In real-world applications, intermediate shading positions are frequently observed, making this strategy a more accurate representation of actual scenarios. Table 5 summarizes all control strategies used in this study.

Table 5.

Summary of all control strategies.

Control Strategy	Description
Baseline-1	Always unshaded
Baseline-2	Always shaded
CS1	Shaded if solar radiation > 150 W/m², unshaded otherwise
CS2	Shaded during day if solar radiation > 150 W/m², shaded overnight
CS3	Seasonal adaptive strategy: Summer (May to September): Be shaded during daytime if solar radiation > 150 W/m², shaded overnight Winter (October to April): Be unshaded during daytime if solar radiation > 150 W/m², shaded otherwise
CS4	Heat flow-based control strategy, with T_in and T_out (see Figure 2)
CS4 (Tave)	Heat flow-based control strategy, with T_ave (see Figure 3)
CS4 (Partial Shading)	Heat flow-based control strategy, with intermediate shading positions (see Figure 4)

Figure 4.

Control logic for CS4 (partial).

Results

The primary objective of this study is to analyze various control strategies and shading attachments to minimize heating and cooling demand across different climate zones in Canada.

Figures 5 –8 show the annual heating and cooling loads for four shading types: Shade 1 (Blackout Roller Shades), Shade 2 (Cellular Shades), Shade 3 (3% Roller Shades), and Shade 4 (10% Roller Shades). These results are evaluated using different control strategies across five Canadian cities: Vancouver, Toronto, Quebec City, Edmonton, and Yellowknife, which span the five different climate zones in Canada.

Figure 5.

Heating and cooling load distribution for Shade 1 (Blackout Roller Shade).

Figure 6.

Heating and cooling load distribution for Shade 2 (Cellular Shade).

Figure 7.

Heating and cooling load distribution for Shade 3: Roller Shade (3% Openness).

Figure 8.

Heating and cooling load distribution for Shade 4 (Roller Shade with 10% openness).

Under the baseline scenario, heating loads are highest in Yellowknife (climate zone 8) and lowest in Vancouver (zone 4C). In Vancouver, the annual cooling loads are higher than the annual heating loads, which aligns with expected behavior and corroborates the findings of Grynning et al. (2014). The significantly higher heating demand in Yellowknife, compared to the combined heating and cooling demand in Vancouver, indicates the extreme weather conditions and extended heating season characteristic of Climate Zone 8.

Regardless of the climate zone, operating shading attachments in an always-shaded position reduces cooling loads but can negatively affect overall energy efficiency by increasing heating loads. Shade 1 (Blackout roller shades) and Shade 4 (roller shades with 10% openness) incur an energy penalty when kept in an always-shaded condition, reinforcing the importance of automated shading control. However, Shades 2 and 3 contribute to reducing heating demand, even in the absence of automated control. The low U-value and high visible transmittance (T_vis = 0.21) of Shade 2 contribute to its impact on reducing heating loads. Similarly, Shade 3, despite a lower T_vis, has a lower U-value, further improving heating energy performance.

CS1 activates shading only when solar radiation exceeds 150 W/m², primarily to mitigate overheating. While this approach effectively reduces cooling demand, it should be avoided in heating-dominated climates, as it increases heating loads. For instance, in Quebec City, CS1 applied to Shade 1 resulted in a 14% increase in heating demand. The results for CS2 indicate that a combination of nighttime shading and daytime shading activation under high solar radiation yields optimized reductions in heating and cooling loads in office buildings. At night, in the absence of solar radiation, heat losses by conduction and convection are lower when the shade is deployed, reducing heating demand. However, this approach also slightly increases cooling demand compared to CS1.

While CS2 reduces heating demand, its impact is limited to a maximum of 12% for Cellular Shades in Vancouver. It follows the same basic principle throughout the year; it lacks the adaptability needed to fully optimize heating and cooling load reduction.

Season-specific control strategies, such as CS3, offer greater potential to reduce heating demand. In fact, CS3 achieves the highest reduction in heating demand among all the control strategies studied. During the heating season, retracting the shade when sufficient solar radiation is present increases solar gains (SHGC_unshaded > SHGC_shaded), thereby further reducing heating loads. However, this improvement in heating energy efficiency comes at the expense of higher cooling demand.

CS4, which operates based on heat balance at the fenestration, strikes a balance between heating and cooling energy efficiency and achieves the greatest reduction in overall energy demand. For instance, in Edmonton, using Shade 3 with CS4 results in a 17% reduction in heating loads and a 29% reduction in cooling loads, highlighting the effectiveness of this strategy in optimizing energy efficiency.

Overall, all shading attachments analyzed in this study exhibit similar performance trends across different shading strategies and climate zones, differing primarily in their degree of effectiveness.

Performance of CS3 and CS4

A detailed analysis of Figures 5 –8 suggests that CS3 and CS4 are the most effective strategies for reducing heating and cooling loads, yielding the greatest overall energy savings. Figures 9 –12 further compare the absolute energy demand reductions for CS3 and the different variants of CS4.

Figure 9.

CS3 performance across different cities and different shading attachments.

Figure 10.

CS4 performance across different shading attachments and climate zones.

Figure 11.

CS4 (T_ave) performance across different shading attachments and climate zones.

Figure 12.

CS4 (partial shading) performance across different shading attachments and climate zones.

The absolute cooling load reductions achieved with CS4 are significantly higher than those obtained with CS3, though at the cost of slightly reduced heating load savings. For example, in Quebec City with Shade 1, the annual cooling load reduction almost doubles from 20 MJ/m²yr under CS3 to 32 MJ/m²yr under CS4 (T_ave), while the heating load reduction values remain comparable at 19.81 and 19 MJ/m²yr, respectively. This highlights the benefits of considering both indoor and outdoor conditions for effective shading control.

In terms of absolute values of load reductions, the subarctic climate zone of Yellowknife exhibits the highest reductions, which is expected given that the number of Heating Degree Days (HDD) is more than 7000 (ASHRAE, 2021; National Research Council of Canada, 2020). The values for Edmonton and Quebec City are comparable, as both cities have similar HDD values. Additionally, the increase in absolute load reductions from climate zone 4C to climate zone 8 follows an almost linear trend.

A comparison of different shading attachments shows that Shade 2 (Cellular Shades) outperforms other shading types in reducing heating loads. From Figures 10 –12, the average temperature method (T_ave) is more effective for cooling load reduction, whereas CS4 (Primary Method) performs better for heating load reduction. CS4 (partial) exhibits a performance trend similar to CS4, but with slightly lower cooling and heating load reductions.

City-specific comparisons

A closer examination of Figures 13 –17 suggests that maintaining an always-shaded position results in the highest reduction in cooling loads, reaching up to 55% in Yellowknife (see Figure 17(b)) when using blackout roller shades. However, the lack of automated shading control can lead to an energy penalty as high as 9% (see Figure 15(a)), as observed in Quebec City with Shade 4 (10% Roller Shades), making this approach unfeasible for heating season applications.

Figure 13.

Comparison of control strategy in Vancouver for (a) heating and (b) cooling loads.

Figure 14.

Comparison of control strategy in Toronto for (a) heating and (b) cooling loads.

Figure 15.

Comparison of control strategy in Quebec City for (a) heating and (b) cooling loads.

Figure 16.

Comparison of control strategy in Edmonton for (a) heating and (b) cooling loads.

Figure 17.

Comparison of control strategy in Yellowknife for (a) heating and (b) cooling loads.

Cellular Shades prove to be the most effective shading attachment for reducing heating loads, with a maximum reduction of 22% in Vancouver using CS3. This is primarily due to their low U-value and the presence of honeycomb structures, which create air pockets that enhance insulation. However, this insulating effect becomes counterproductive during the cooling season, as the air pockets trap heat, leading to higher cooling loads than with roller shades.

For cooling load reduction, priority should be given to shading attachments with low SHGC and low visible transmittance. Shade 1 (Blackout Roller Shades), with a visible transmittance of 0 and an SHGC of only 0.22, is highly effective in reducing cooling loads. However, due to their relatively high U-value (1.718 W/m²K) compared to Cellular shades, their reduction in heating load is less significant.

Shade 4 (roller shade with 10% openness) performs worst in terms of heating-load reduction, primarily due to its higher U-value compared with the other shading attachments studied. As demonstrated by Shade 4’s poor performance, a combination of high SHGC and high U-value should be avoided in cold climate zones.

Higher cooling-load reductions are observed in Yellowknife and Edmonton, which can be attributed to longer daylight hours, higher latitudes, and lower sun angles, despite their shorter cooling seasons. This makes the impact of shading attachments more pronounced. In contrast, Toronto exhibits the lowest percentage reduction in cooling loads, primarily due to its moderate summer temperatures, with fewer extreme peaks, and higher cloud cover, both of which make shading attachments less effective. However, in terms of heating load reduction, the percentage reductions are comparable across all climate zones.

Discussion

A detailed analysis of Figures 9 and 13 –17 suggests that while CS3 is the most effective strategy for reducing heating loads, its impact on cooling loads is less pronounced. CS3 performs particularly well in colder-climate zones, as evidenced by the higher percentage reduction observed in Yellowknife. Several factors may explain this trend. Keeping the shades open for extended periods when solar radiation is high but below 150 W/m² can lead to overheating, potentially increasing cooling loads. This effect is particularly evident in Toronto, where prolonged warm weather with moderate solar radiation leaves the window unshaded for extended periods, allowing significant heat to enter the room and directly increasing cooling energy demand.

The implementation of climate-specific solar radiation thresholds appears to be a viable alternative. Further, the same months were used to define the heating and cooling seasons for all cities, which may have contributed to lower percentage reductions in cooling demand for Vancouver and Toronto. During transitional months, using a heating-specific strategy when a cooling-specific strategy would have been more beneficial could have inadvertently increased cooling demand. Another influencing factor is the variation in daylight hours across latitudes. Cities like Vancouver and Toronto, situated at lower latitudes, experience shorter daylight hours, reducing the duration of solar exposure, which in turn affects the performance of shading control strategies.

Comparison of CS4 (primary method) and CS4 (T_ave) algorithms

The two different approaches to CS4 provide valuable insights into their performance trade-offs. The average temperature method, which has been previously examined in the literature (Firląg et al., 2015), is more effective for cooling load reduction, while the primary method yields better heating energy performance. For instance, in Edmonton, with Blackout Roller Shades, the primary method results in a 4% greater improvement in heating energy performance. While this 4% improvement may seem modest, it is significant in heating-dominated climates where natural gas is the primary heating source, as it translates to a substantial absolute load reduction.

In Edmonton, using the primary method with Blackout roller Shades, the absolute heating load reduction is 9713 MJ/yr, compared to 5874 MJ/yr with the average method, making a 65% greater heating load reduction. However, this improvement in heating efficiency comes at the cost of increased cooling energy savings. The absolute cooling load reduction using the primary method is 16,686 MJ/yr, lower than the 21,255 MJ/yr achieved with the average method.

Figure 18 demonstrates the shade positions for one specific day (January 2 from TMY3) using CS4 (primary method) and CS4 (T_ave). On this TMY day, the solar radiation was sufficient to drive net heat flow into the room through the window. While both algorithms are based on net heat flow through the window, they differ in how they determine heating and cooling modes. The average temperature method relies on a fixed 21°C threshold, whereas the primary method uses the heating and cooling setpoints to make this distinction.

Figure 18.

Shade position for CS4 and CS4 (T_ave) on January 2 (“0” represents open, and “1” represents closed).

Thus, in Figure 18, it is evident that between 11:00 am and 12:30 pm, CS4 (T_ave) closes the shade when the indoor temperature exceeds 21°C, whereas CS4 (primary method) keeps the shade open. This results in greater solar heat gain, reducing heating loads. However, this increased solar exposure also leads to higher annual cooling loads than the T_ave method.

Heating typically requires more source energy than cooling (Climate Institute C, 2023), especially in Canadian climates. With its greater reduction in annual heating demand, CS4 addresses a larger portion of a building’s overall energy consumption (Leila and Sager, 2019). In addition, the greater reduction in heating energy demand has environmental advantages, as heating energy in some Canadian cities comes from fossil fuels, such as natural gas, which emit higher levels of carbon. In contrast, cooling energy demand, for example, in Vancouver, is largely met through hydroelectric power, a renewable energy source. Using CS4 rather than CS4 (T_ave) can reduce carbon emissions.

CS4 partial shading

A deeper analysis of the partial shading algorithm reveals that shade position is determined by minimizing or maximizing net heat flow at each timestep. However, tracking the simulated shade positions reveals that the intermediate shading positions are never observed in practice for a typical meteorological year. Instead, the system predominantly selects either a fully open or a fully closed position. This outcome is due to the fact that intermediate shading positions are most energy-efficient only under very specific combinations of solar radiation and the indoor-outdoor temperature difference (△T).

For example, considering blackout roller shades under partial shading, equations (6)–(10) describe the heat flow for all five shading positions:

\begin{matrix} q_{up} = 2.64 Δ T + 0.70 I \end{matrix}

(6)

\begin{matrix} q_{\frac{1}{4}} = 2.45 Δ T + 0.58 I \end{matrix}

(7)

\begin{matrix} q_{\frac{1}{2}} = 2.23 Δ T + 0.46 I \end{matrix}

(8)

\begin{matrix} q_{\frac{3}{4}} = 2.00 Δ T + 0.34 I \end{matrix}

(9)

\begin{matrix} q_{down} = 1.72 Δ T + 0.22 I \end{matrix}

(10)

For the 1/4-shaded position to be the most energy-efficient (q_1/4 maximum), q_1/4 > q_1/2, q_1/4 > q_3/4, q_1/4 > q_up, and q_1/4 > qdown, for which ΔT and I must satisfy the following constraint: –0.63I < ΔT < −0.55I.

This condition requires a very specific relationship between outdoor temperature (T_out) and solar radiation (I), which was never observed in the typical meteorological year dataset. Similarly, 50% and 75% shaded positions have their own unique constraints, but the necessary climatic conditions never or very seldom occurred in the dataset, preventing their selection in the simulation.

The observed difference in heating and cooling loads between the CS4 primary and CS4 partial shading methods arises solely from the edge effects introduced when dividing the windows into four equal parts, a workaround used to model partial shading in EnergyPlus. From an energy efficiency perspective, intermediate shading positions are rarely optimal. However, in real-world applications, the practical implementation of sensor-based control strategies could result in intermediate shading positions being utilized. This is because real-world temperature sensors capture temperature gradients within a room, which may influence shade-positioning decisions differently from those in simulations.

Limitations and future work

The simulation results demonstrate that effective control strategies, when combined with appropriate shading attachments for a given climate zone, can significantly reduce heating and cooling demand in commercial buildings in Canadian cold-climate zones. However, this study has a few limitations to consider.

First, the U-values and SHGC values used in this study were calculated under standard NFRC boundary conditions (NFRC, 2017). In practice, U-values can vary based on real-world boundary conditions, as demonstrated in a previous study (Chhetri and Zhong, 2025). Future studies could incorporate U-value variations with outdoor temperature to provide a more realistic estimate of energy demand in real-world settings. Similarly, future studies can combine experimental evaluation of detailed solar heat gain through windows, rather than using average reported values.

Second, this study focused primarily on the effect of shading control strategies on heating and cooling demand, without modeling detailed HVAC systems. Future studies could integrate climate-specific HVAC systems to quantify the environmental benefits of shading control strategies, particularly in terms of greenhouse gas (GHG) emissions reductions.

Third, while retracting shades during high solar radiation periods effectively reduces heating demand by capturing passive solar gains, it may introduce additional challenges, such as glare and visual discomfort (da Silva et al., 2012; Grynning et al., 2014; Tzempelikos and Shen, 2013). These factors are important for occupant well-being but were not explored in this study. Additionally, occupant preferences for shade positioning may differ from energy-optimal positions, leading to variations in actual energy savings compared to simulation results. Therefore, future research should balance energy efficiency with occupant thermal and visual comfort, ensuring practical feasibility in real-world applications.

Another limitation of this study is that it does not account for the influence of adjacent buildings or topographic shading. In reality, the incident solar radiation at the surface of windows can be impacted by the presence of exterior buildings, which may be taller and block direct sunlight, especially in urban environments. This can influence the optimal shading position when implementing a particular control strategy. Future studies should incorporate adjacent buildings and topographic shading, or use urban weather files, to simulate this effect.

Finally, the heating and cooling load reductions observed in this study will vary with shading attachment type and window-to-wall ratio. To enhance the generalizability of the findings, future work should explore a broader range of shading attachments, including external and between-pane shading systems, and different window configurations.

Conclusions

This study explored four shading attachments and six automated shading control strategies across five climate zones in Canada. The simulations were conducted on a representative double-pane window within a DOE small office building, with a total of 160 EnergyPlus simulations, including two baseline scenarios (unshaded and fully shaded). The heating and cooling demands were analyzed across different shading strategies, shading attachments, and climate zones. Based on the results, the following conclusions can be made:

Shading attachments with a low U-value and a finite visible transmittance (T_vis) can positively impact both cooling and heating loads, even in the absence of control strategies. Static shading (always shaded) has the highest cooling-load reduction potential, achieving a 55% reduction for Shade 1 in Yellowknife.

Of the shading attachments studied, Shade 2 (Cellular Shade) demonstrated the best performance for heating energy reduction, while Shade 1 (Blackout Roller Shade) was most effective for cooling energy reduction.

Among the control strategies studied, a maximum heating load reduction of 22% was obtained in Vancouver with CS3 and Shade 2 (Cellular Shade), while the maximum cooling load reduction of 47% was obtained in Yellowknife with CS4 (T_ave) and Shade 1. Parameters such as solar radiation thresholds and heating/cooling season periods must be adjusted according to a specific climate zone to maximize heating and cooling energy savings.

CS4 was identified as the most effective control strategy for combining heating and cooling load reduction. Compared to CS4 (T_ave), CS4 achieved greater heating demand reduction, with percentage improvements ranging from 7.04% to 19.69%, compared to 4.14%–17.20% for CS4 (T_ave). These improvements are significant in terms of GHG emissions.

From an energy efficiency standpoint, partial shading does not provide additional benefits over fully open or fully closed positions for CS4. This is due to the restrictive conditions required for intermediate shading positions to be energy efficient. However, in practical applications, real-world sensor-based implementations may yield better results.

Footnotes

Acknowledgements

The authors would like to express their gratitude to Zack Zhang, Founder of AI Shading, and Caitlyn Shum, for providing valuable industrial and technical insights throughout the duration of the project.

ORCID iD

Lexuan Zhong

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The project was funded by MITACS (IT 33038).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The datasets generated and analyzed during the study are available from the corresponding author upon reasonable request.

References

American Society of Heating, Refrigerating, and Air Conditioning Engineers (ASHRAE) (2021) ASHRAE Handbook—Fundamentals, 2021. ASHRAE.

ASHRAE (2021) ANSI/ASHRAE Addendum a to ANSI/ASHRAE standard 169-2020: Climatic Data for Building Design Standards. Atlanta, GA: ASHRAE. Epub ahead of print 2021.

Boubekri

Cheung

Reid

, et al. (2014) Impact of windows and daylight exposure on overall health and sleep quality of office workers: A case-control pilot study. Journal of Clinical Sleep Medicine 10(06): 603–611.

Chhetri

Zhong

(2024) Load reduction potential by employing internal shading devices: A comparative analysis in Canadian cold climate zones. In: Proceedings of eSim 2024: 13th Conference of IBPSA-Canada, Edmonton, Canada, 5–7 June 2024.

Chhetri

Zhong

(2025) In-situ testing and numerical study of thermal performance of blind systems in a cold climate zone. Journal of Building Engineering 102: 111934.

Canadian Climate Institute (2023) Heat pumps pay off: Unlocking lower-cost heating and cooling in Canada. Executive summary. Ottawa, ON: Canadian Climate Institute.

da Silva

Leal

Andersen

(2012) Influence of shading control patterns on the energy assessment of office spaces. Energy and Buildings 50: 35–48.

DOE and PNNL (2023) Prototype building models and building energy codes program. Available at: https://www.energycodes.gov/prototype-building-models#Commercial (accessed 27 February 2025).

EnergyPlus Development Team (2023) EnergyPlus documentation: Engineering reference. Version 23.1. U.S. Department of Energy, Washington, DC, USA, 14 April.

10.

Firląg

Yazdanian

Curcija

, et al. (2015) Control algorithms for dynamic windows for residential buildings. Energy and Buildings 109: 157–173.

11.

Grynning

Time

Matusiak

(2014) Solar shading control strategies in cold climates – Heating, cooling demand and daylight availability in office spaces. Solar Energy 107: 182–194.

12.

Haynes

(2008) The impact of office comfort on productivity. Journal of Facilities Management 6(1): 37–51.

13.

International Code Council (2021) International energy conservation code. Available at: https://codes.iccsafe.org/content/IECC2021P1 (accessed 27 February 2025).

14.

Kunwar

Bhandari

(2020) A comprehensive analysis of energy and daylighting impact of window shading systems and control strategies on commercial buildings in the United States. Energies 13(9): 2401.

15.

Kunwar

Bhandari

Curcija

(2022) National energy savings potential of cellular shades: A measurement and simulation study. Building and Environment 225: 109593.

16.

Lawrence Berkeley National Laboratory (2023a) Complex glazing database. Available at: https://windows.lbl.gov/complex-glazing-database (accessed 27 February 2025).

17.

Lawrence Berkeley National Laboratory (2023b) WINDOW. 7.8. U.S. Department of Energy.

18.

Leila

Sager

(2019) The future of home heating: Hybrid home heating systems offer energy savings and reduce GHG emissions - Natural Resources Canada. Available at: https://natural-resources.canada.ca/stories/simply-science/future-home-heating-hybrid-home-heating-systems-offer-energy-savings-reduce-ghg-emissions (accessed 1 March 2025).

19.

National Research Council of Canada (2020) National Energy Code of Canada for Buildings 2020. Ottawa.

20.

Natural Resources Canada (2024a) Energy efficiency in existing buildings. Available at: https://natural-resources.canada.ca/energy-efficiency/building-energy-efficiency/energy-efficiency-existing-buildings (accessed 27 February 2025).

21.

Natural Resources Canada (2024b) The Canada green buildings strategy: Transforming Canada’s buildings sector for a net-zero and resilient future. Available at: https://natural-resources.canada.ca/energy-efficiency/building-energy-efficiency/canada-green-buildings-strategy-transforming-canada-s-buildings-sector-net-zero-resilient-future (accessed 27 February 2025).

22.

Natural Resources Canada O of EE (2020) Energy use in the commercial/institutional sector. Available at: https://oee.nrcan.gc.ca/publications/statistics/trends/2020/commercial.cfm (accessed 27 February 2025).

23.

Nezamdoost

Mahic

Van Den Wymelenberg

(2018) A human factors study to update a recently proposed manual blind use algorithm for energy and daylight simulations. In: Proceedingsof the44th annual conference of the IEEE industrial electronics society(IECON 2018), 21–23 October 2018, Washington, DC, USA, pp. 789–794. Institute of Electrical and Electronics Engineers Inc.

24.

National Fenestration Rating Council (NFRC) (2017) NFRC 100: Procedure for determining fenestration product U-factors. Greenbelt, MD: National Fenestration Rating Council.

25.

Oleskowicz-Popiel

Sobczak

(2014) Effect of the roller blinds on heat losses through a double-glazing window during heating season in Central Europe. Energy and Buildings 73: 48–58.

26.

Shum

Zhong

(2023a) A review of smart solar shading systems and their applications: Opportunities in cold climate zones. Journal of Building Engineering 64: 105583.

27.

Shum

Zhong

(2023b) Optimizing automated shading systems for enhanced energy performance in cold climate zones: Strategies, savings, and comfort. Energy and Buildings 300: 113638.

28.

Tan

Peng

Curcija

, et al. (2020) Parametric study of the impact of window attachments on air conditioning energy consumption. Solar Energy 202: 136–143.

29.

Tzempelikos

Shen

(2013) Comparative control strategies for roller shades with respect to daylighting and energy performance. Building and Environment 67: 179–192.

30.

U.S. Department of Energy (2023) EnergyPlus. 23.2.0. Available at: https://energyplus.net/ (accessed 27 February 2025).

31.

World Meteorological Organization (2021) Measurements of meteorological variables. In: Guide to Instruments and Methods of Observation. Geneva, Switzerland: World Meteorological Organization, pp. 308–323.

Energy performance of internal shading attachments: Comparative analysis of control strategies in Canadian cold climates

Abstract

Keywords

Introduction

Methodology

Setup

Window model

Description of control strategies

Baseline case

Control strategy 1 (CS1)

Control strategy 2 (CS2)

Control strategy 3 (CS3)

Control strategy 4 (CS4)

Primary method

Average method

Partial shading method

Results

Performance of CS3 and CS4

City-specific comparisons

Discussion

Comparison of CS4 (primary method) and CS4 (Tave) algorithms

CS4 partial shading

Limitations and future work

Conclusions

Footnotes

Acknowledgements

ORCID iD

Funding

Declaration of conflicting interests

Data availability statement

References

Comparison of CS4 (primary method) and CS4 (T_ave) algorithms