This section describes the methodology of the paper including neighborhood selection, energy modeling, simulation details, and solar strategies. Firstly, a selection of five neighborhoods representative of common urban layouts in Canada (and North America) are analyzed. The assessed neighborhoods are all residential. The archetypes that compose each neighborhood are detached houses, attached houses (duplexes and townhouses), and multi-family apartment buildings. A 3D model of each of these neighborhoods is developed to assess the solar potential and have a base to develop an energy model. Energy Plus in conjunction with SketchUp (using a plugin-Euclid) is employed in development.
To develop an energy model, a baseline is developed using the Natural Resources Canada energy use database for validation purposes30, which has the average energy use based on building types, building use, and heating and cooling systems, divided by each province. Energy simulations are performed to estimate the demand of each of the selected neighborhoods and validated against the baselines. These models are then systematically modified to reduce the overall energy consumption and to implement various solar strategies. The study focuses on the reduction of heating loads by altering the materials of the building envelope. Energy consumption of the neighborhoods is compared to energy generation to determine whether a net-zero energy status can be achieved. Figure 1 illustrates the methodology workflow.
Overall workflow of proposed methodology.
A matrix is used to classify, evaluate, and understand the neighborhood characteristics in a study developed by the International Energy Agency (IEA) Task 63: Solar Neighborhood Planning (International Energy Agency, 2022). This matrix exhibits five different types of street layouts that represent North American street networks: conventional grid, conventional grid with tilted orientation, curvilinear loop, cul-de-sac, and radial. Street layouts have an immediate impact on the solar generation potential of a neighborhood since they affect the design and set of buildings.
Matrix for neighborhood characterization
Cities are composed of districts, which all have distinct characteristics, but most of them morph into urban patterns that can be visually identified. The International Energy Agency (IEA) Task 63: Solar Neighborhood Planning developed a matrix to catalog different patterns of neighborhoods, classifying them by neighborhood types and uses, street layouts, building designs, and several other neighborhood parameters to recognize such patterns. The matrix is based on an extensive spreadsheet that allows inputting data related to the features of neighborhoods, separated into six categories:
-
A)
Neighborhood type: this section displays general data including total area, demographics such as population, number of dwellings, as well as street layout characteristics, type of neighborhood (e.g., residential, mixed-use), and green areas.
-
B)
Neighborhood building structure and passive design: related to the details of building block design as well as individual building designs, number of stories, units, and area. This section tackles as well as roof details, construction materials, the orientation of the façades, and window design.
-
C)
Solar energy generation: this part includes various parameters that affect of the design of solar technologies (photovoltaic and thermal collector systems), like orientation, tilt angle, efficiency, and surface location.
-
D)
Energy systems: this section displays the energy systems employed in the buildings of the analyzed neighborhood, divided by heating, cooling, and water heating systems.
-
E)
Other information: this section shows other features that might be relevant to the energy model, such as lighting, systems set points, and appliances.
-
F)
Simulation outputs: this part presents a summary of the results simulated in the energy model.
Overall, using the IEA matrix as a template, this study aims to understand how the different aspects of a neighborhood impact energy performance, especially in the application of solar strategies (a detailed matrix is provided in Appendix A).
Neighborhood selection
An analysis of urban patterns is performed to identify representative layouts across multiple cities in Canada. Various uncontrollable parameters such as climate, topography, and usage patterns are similar in all the selected neighborhoods. Based on this extensive analysis using satellite images, five different street layouts are assumed as representative of urban planning in Canada, as shown in Table 1, below. This study includes only existing residential neighborhoods of single-family residential units including different types of units. Residential units are divided into three categories such as single detached houses, single attached houses, and apartment buildings.
Five neighborhoods in different provinces are selected as a case study for each layout representative of urban patterns in Canada. The neighborhoods analyzed are Brighouse located in Richmond-BC, Parkdale, located in Calgary-Alberta (AB), Glendale, located in Winnipeg-Manitoba (MB), East York, located in Toronto-Ontario (ON), and Mount Royal, located in Montreal-Quebec (QC). Figure 1 shows the selected area of each neighborhood.
A small sample of each neighborhood is analyzed to ease the simulation process, ranging from 1.7 to 4.8 hectares (17,000–48,000 m2) due to the form of the block, which is affected by the street layout and type of construction. Density and demographic indicators, as well as other data from each community profile, are extracted from the Canada Census 201632 and Natural Resources Canada 2019 database30. Characteristics of the selected neighborhoods are presented in Table 2.
Energy modeling
This section presents the assumptions employed in the energy modeling process of the selected neighborhoods, such as materials used, the development of a comparable baseline, and some rules followed to develop the 3D model in the energy simulation (Fig. 2).
Aerial pictures of selected neighborhoods (taken from Google Earth): (A) Brighouse, (B) Parkdale (C) Mount Royal Brighouse, (D) Glendale, and (E) East York.
Energy modeling assumptions
The energy modeling process consists of first inputting all the obtained data into an energy modeling program to generate a comprehensive energy model of the neighborhood. EnergyPlus33, an energy analysis and thermal load simulation program, is utilized for this investigation. SketchUp and the plugin Legacy Open Studio34 are also used in conjunction with EnergyPlus to simplify the input of geometric data of all the buildings within a neighborhood. Euclid plugin uses SketchUp’s modeling platform to transform geometric data (created in SketchUp) data from a 3D model into an EnergyPlus-compatible energy model. Satellite imagery is used to compute measurements for 3D modeling of neighborhood buildings. Furthermore, Google Street View function and 3D models from the Google Earth35 are used to represent the structures’ rooftops as exactly as possible. The neighborhood energy models are formulated for this study are simulated as whole building stock to account the mutual effects of buildings. In the energy simulations, hourly weather data provided by Canadian energy and engineering data sets (CEEDS) is used for this work. In the development of the neighborhood models, the following assumptions were considered and can be also correlated from Fig. 3.
-
a)
Intending to ease the neighborhood modeling, the roof shape determines each building envelope forms, assuming external walls are 1 m offset from the roof edge.
-
b)
Each storey is assumed to have a ceiling height of 2.5 m, 0.40 m higher than the minimum required by the Ontario Building Code for room ceiling heights36. Building interior walls and internal divisions were not considered.
-
c)
If there are no street views available from Google Earth, fenestrations are based on the patterns of surrounding buildings.
-
d)
Each dwelling in a single-family structure is a zone, whereas multi-family buildings include a zone for each unit as well as a common space.
-
e)
Due to its widespread use in North American homes30, the heating system used in all energy simulations considered is the natural gas furnace with 80% efficiency considering ideal loads in order to standardize the simulation process. Heating setpoint is taken as 20 °C for all buildings.
-
f)
Because cooling systems are less commonly used in Canada in residential buildings due cold climates, space cooling systems are not taken into consideration.
-
g)
The building envelope is the sole upgrade researched and implemented for buildings retrofit.
Visual representation of various solar strategies modeling assumptions.
Further, a single neighborhood building energy model (for each of the considered neighborhood) is constructed with different buildings and simulated in this work. It is essential to capture the mutual impact of buildings as well as the effect of the configuration of the neighborhoods on energy simulations.
Building envelope materials
To estimate the energy demand of the neighborhoods, two sets of simulations are proposed for the study. The first simulation set, referred to as “typical simulation,” uses standard materials applied in the North American construction sector. Those materials consist of insulated wood frame walls, plywood floors, asphalt shingle roofs coated with oriented strand board (OSB) boards, and double pane glazing. The second simulation set is conducted based on high-performance materials that present a superior R-value, referred as “high-performance simulation. The fenestration materials use the typical fenestration characteristics chart available in the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) Handbook Fundamentals as a reference37. For the energy simulations, two glass layers with operable aluminum frames are used for the typical simulation, while triple-layered low-e, low-solar glazing with operable aluminum frames is assumed for high-performance simulation. This study considers only the heating load reduction when proposing the implementation of new building envelope materials. Table 3 illustrates the characteristics of the materials assumed in the simulations.
Model validation
Following the neighborhood simulations, the database created by Natural Resources Canada for energy use statistics30 is employed to develop baselines to compare and validate the simulations. The comparison is conducted based on energy consumption per area and end-use (kWh/m2/year).
The database stated above has a great deal of information regarding energy use, broken down by province and building type. Three baseline models are built for each province for this study, based on the following building types: single detached houses, single attached houses, and apartment buildings Space heating, water heating, appliances, and lighting are among the end-uses evaluated. Whilst all these end-uses contribute to the overall energy demand of buildings; this research focuses on measures to minimize space heating demand since it is the most significant in Canada and is directly tied to building envelope design. By combining data from the database with the Canada Census 201632, it is possible to estimate the energy demand per area for each baseline model. Table 4, based on the NRCan database set30 shows the assumptions for each baseline model.
Solar strategies application
Solar strategies applications are simulated employing PVsyst. Prior to running these simulations, a shading analysis is carried out employing SketchUp in order to ensure that solar systems are not shaded, for at least 7 h per day during the whole year. This analysis is crucial to understand the surroundings before designing solar strategies for each neighborhood, to optimize the energy generation of the photovoltaic modules. The shading analysis is used to recognize some of the environmental elements, like trees, other buildings, or different parts of the roof, that may shade the PV system.
Simulating large-scale PV systems is challenging since PVsyst can only compute 8 distinct solar systems within the same simulation model. the models with minimal shading explored in SketchUp, are then applied in different settings, such as region, azimuth, tilt, and albedo, using PVsyst. A line horizon shading cut-off is applied to replicate the shade caused by the horizon, using the horizon shading function in PVsyst. In this study, the sun is assumed dispersed from 6 a.m. to 8:30 p.m. in the summer and 8 a.m. to 4:30 p.m. in the winter, if not behind the horizon plane. The horizon line, for Calgary, is drawn with the assumptions indicated in Fig. 4.
Horizon line shading, based on the simulations conducted in PVsyt.
For this study, monocrystalline, which are displayed in rooftop and east–west mounts, and bifacial panels, paired with ground mounts on flat rooftops, as well as solar carports and building integrated photovoltaic (BIPV) systems, and solar trees, will be employed to fulfill energy production requirements, depending on the energy demand and qualities of the surface to be placed. Table 5 shows the pros and cons of each selected solar strategy used in the study that would be further useful in the decision-making process for urban planners.
An albedo of 0.15 is considered for solar carports, which have an asphalt reflecting surface, 0.25 for solar trees (considering grass as the reflecting surface), and 0.9 for flat rooftops assumed to be painted white38. For the solar generation simulation, twelve baseline systems are simulated to make solar simulation easier and to create a standard system that can be replicated and adapted to varied urban layouts and applied to each neighborhood according to its suitability. The assumptions used in these baselines are shown in Table 6, considering the region where each neighborhood is located as well.
The yield of each system is used as a foundation to estimate the energy output of each system. This system configuration is simulated in three distinct azimuths: 0°, 45°, and 90° angles, as well as a 25° plane tilt. A ground cover ratio of 57.1% is chosen for systems put on a level area. This ratio allows a 3.5 m space between modules of 2 m in length, reducing the mutual shade between the modules. The power output of all solar panels is considered as 420 W, with an efficiency of 20.4%. The nominal power per inverter ratio is 1.16, having no overload loss. The reason behind overloading is standard in solar systems design since it allows the inverter to harvest more solar energy during most parts of the day, compensating for the loss due to the inverter clipping when it peaks the energy production output39.
Solar strategies implementation rationale
To evaluate the benefits and drawbacks of each solar strategy when applied in various situations, this study provides a decision-making process that relies on Table 7, in which green signifies a favorable outcome, yellow is an average outcome, and red is a negative outcome.
The outcome is assessed based on comparative performance among the various strategies. Table 7 summarizes the decision-making process considering various factors. The criteria employed in evaluating these strategies, namely implementation cost, assembly difficulty, efficiency, power-to-area ratio and power output are briefly explained below.
Strategies implementation cost
The decision-making process for implementing solar strategies considers multiple factors. The cost of implementing such technology is the first one to be analyzed. Considering tilted rooftops, the cost of installing solar panels is low: $1.00 to $1.50 per watt for monocrystalline modules, $0.90 to $1.00 per watt for polycrystalline, and $2.41 to $3.42 per watt for bifacial solar panels40, 41. Since the structures to install solar systems in tilted rooftops are cheap, monocrystalline and polycrystalline are rated as “low” in the cost category, while bifacial is rated as “medium”.
When considering the installation of solar systems in flat roofs, the cost may vary depending on the design due to the elevated ground mount structures needed to space and tilt the modules. Assuming that there is an extra cost for such structures, systems that require a ground mount are also rated “medium” in the cost factor.
Other than flat and tilted rooftops, there are options to implement solar systems, such as solar carports in parking lots, solar trees, and BIPV using solar panels as building elements. All the strategies cited above require custom designs and structure manufacturing, as well as specialized labor to install, and thus are rated as “high” in the cost factor of the implementation rationale.
Assembly difficulty
Following the same approach as the cost factor, assembly is considered when choosing the best solar strategy for each situation. Installing solar panels on tilted rooftops is the most common type of installation, having standard clamps and an anchoring system that makes the process very fast and easy, therefore is rated as “easy”. Ground mounts require assembly of the ground structures, anchoring into the slab, and waterproofing it. The installation process requires more material and more attention in the assembly, consequently being rated as “medium”.
Again, custom designs like carports, solar trees, and BIPV takes usually more time to install, assemble and design due to the complexity of the structures to install. Also, most of it requires specialized labor. Consequently, this type of strategy is rated as “hard”.
Efficiency
On tilted rooftops, monocrystalline has a better performance due to the pureness of the silicon used in the manufacturing process, therefore is rated as “high”42. Polycrystalline has a lower efficiency than monocrystalline, rated as “medium”. Bifacial modules are not considered in tilted rooftops because their rear surface does not face a reflecting surface, so it makes no use of their advantage over similar technologies.
On flat surfaces, ground-mounted systems have a better performance than east–west mount because of the orientation of the arrays and optimized tilting, ranked as “high”. East and west mount have poor orientation, therefore having a “low” efficiency.
Since custom designs such as carports, solar trees, and BIPV are sometimes the only option to be implemented in some environments, it is assumed that there is the freedom to choose whatever technology fits best under the determined circumstance, despite the efficiency of the systems. Solar carports efficiency is usually “high” but might be affected if there are nearby buildings shading the parking area. BIPV performance depends heavily on the method of implementation including the tilt angle of the surface used for the integration of this PV system, therefore each BIPV system proposed in the study will be evaluated and simulated separately and not included in this solar strategy implementation rationale. Solar trees are assumed “high” since they can be placed in no shading areas, but the albedo from grass is low.
Power to area ratio
This category factors in the amount of solar power that can be installed in each area. Since all the modules are the same size, tilted rooftops are not applicable to this category. Ground mounts used in flat rooftops have better efficiency, however, it needs more spacing due to mutual shading, having a lower power per area ratio than east–west mount. Figure 5 illustrates how many solar panels are suitable in a 100 m2 area. After the comparison, the east–west mount is rated as “high”, while the ground mount is rated as “low”.
Power to area ratio comparison: (a) east–west mount and (b) ground mount on the right.
Power output
The efficiency and power output of each technology is simulated and compared using PVSyst. Thus, the power-to-area ratio is determined by designing each strategy for the rooftops available in each neighborhood using SketchUp as a modeling tool. After simulating all the technologies in different environments, the following conclusions are achieved: for tilted rooftops, monocrystalline is the technology that has the better energy generation, therefore rated as “high”. For flat rooftops, the east–west mount has a slightly better power output, even though its efficiency is lower.
When the flat rooftop is not oriented south, ground mount power output is better than east–west mount. Also, this study does not consider systems that are oriented north, northeast, and northwest, so an east–west mount is not used on flat rooftops that are not south-oriented because half of the system would not be in an acceptable orientation. Solar carports, solar trees, and BIPV are used as suitable disregarding the efficiency or power output of the system since it is considered as an alternative system to be employed in alternate locations.
Moreover, East–West mount PV employs monocrystalline modules due to the low lighting incidence of its back surface, whereas ground mount PV uses bifacial due to the considerable exposure to light reflection from its surface. Flat south-facing rooftops will benefit more from East–West mount PV since the power output and area ratio is higher than Ground Mount PV by being the same effect on cost and installation. When the building is not south-faced, east–west mounts have a large loss in power production and therefore do not meet with the premise of this study of not having solar systems facing northwest or northeast. When analyzing the available rooftop systems, monocrystalline came out ahead because it presents the biggest benefits when compared to similar technologies.
