Publication Cover

International Journal of Architectural Heritage

Conservation, Analysis, and Restoration
Volume 16, 2022 - Issue 1
414
Views
3
CrossRef citations to date
0
Altmetric
Research Article

Balancing Trade-offs between Deep Energy Retrofits and Heritage Conservation: A Methodology and Case Study

, , &
Pages 97-116
Received 28 Oct 2019
Accepted 05 Apr 2020
Published online: 27 Apr 2020

ABSTRACT

Drastic reductions in energy consumption within existing buildings are required to achieve climate change mitigation targets. However, a portion of existing buildings have important historic values that need to be conserved. The goal of this paper is to present a methodology and decision-framework for deep energy retrofit analyses that balances trade-offs between conservation and sustainability. This methodology includes historic recording, documentation, a detailed energy model, and calibration to monthly data. An historic house in Ottawa, Canada is studied to demonstrate the use of the methodology. The energy retrofit analysis suggests 67% energy savings are achievable by increasing envelope thermal resistance to 4.1 m2-K/W, reducing air infiltration by 70% to 4.2 ACH at 50 Pa through air sealing and an air-vapour barrier, rehabilitating windows to be triple-pane low-E assemblies, using an air-source heat pump to supplement the existing gas boiler, daylight sensors and controls, and solar PV panels.

1. Introduction

While we are making our best efforts to reduce our impact on the planet through better performance standards for new buildings and thriving as a society, there is still missed opportunity in the sustainability of conserving heritage buildings. The development of new buildings to replace existing building is around 1.0–3.0% per year globally (Ma et al. 2012), which means that building new net-zero energy buildings is not the way to solve the planet’s climate change crisis. The National Energy Code of Canada (NECB) (NRC 2015) is a national guideline aiming to minimize the impact of new buildings on the environment. The Ontario provincial building energy code has identified that conservation of existing buildings and energy upgrades should be a focus for achieving energy reduction targets (MMAH 2016), but the rate of energy retrofits within existing buildings globally is at a low rate of 2.2% per year (Ma et al. 2012). It also considers that heritage buildings should be upgraded abiding by the Heritage Act, ensuring that changes are not detrimental to the conservation of the cultural values of the building (MMAH 2016).

Reusing buildings is sustainable because it: reduces the depletion of natural resources, alleviates landfill pressures, reduces energy use from demolition and embodied energy to create new buildings, and reduces the environmental, social and economic costs due to suburban expansion and land use intensification (Brand 1994; Rypkema, Cheong, and Mason 2011; Young 2012). The existing building stock in the U.S. consists of over 90% built before the 1990s which are now requiring major retrofits (Spengler, Samet, and McCarthy 2001). These retrofits may be for seismic performance and prevent safety risks due to natural deterioration, to improve thermal comfort and energy performance (Webb 2017), to conserve heritage values through preservation and rehabilitation (Grimmer 1995), and for adaptive reuse by current and future generations (Bullen and Love 2011). Historic buildings1 (HBs) worldwide present a great opportunity to achieve large energy reductions as they comprise between 10% and 40% of the building stock and will continue to exist because of their cultural value (CIBSE 2002; DOE 2011; Economidou et al. 2011; Troi 2011; Webb 2017). There is potential for large-scale energy savings due to energy retrofits in HBs of 75% from a baseline of 240 Mt of CO2 emissions without retrofits (Troi 2011).

Currently, HBs are exempt from energy efficiency codes in many regions (EU-Parliament 2010; NECB 2015), but it is likely that large energy reduction targets will also be required in HBs soon to meet climate change targets. There are similarities between implementing energy retrofits in existing buildings and in historic and traditional buildings (Webb 2017). The main differences between energy retrofits in these buildings are within the physical characteristics of the building (e.g. the construction materials are non-homogenous, HVAC systems are older and inefficient, etc.) and the constraint of conserving the cultural values within these HBs (Webb 2017). These differences provide challenges in implementing deep energy retrofits.

The difficulty in achieving deep energy retrofits in HBs is that a prescriptive compliance pathway is not a viable option to conserve the heritage values of the building. This pathway prescribes low levels of air infiltration or specific R-values (thermal resistance) for envelope components, among other things, which may not be attainable for the type of construction of HBs (Pracchi 2014). A performance compliance pathway allows deviations from prescribed energy requirements through energy analyses that demonstrate the desired building performs better than a prescribed building of the same massing and internal loads. This makes a performance compliance pathway a better option for historic buildings. However, creating an accurate energy model is more difficult for HBs because of their unique constructions (Adhikari et al. 2011) and lack of as-built documentation.

There are a few studies that demonstrate deep energy retrofits can save between 40% and 89% energy in historic buildings in cold climates (Bin and Parker 2012; Jermyn and Richman 2016), exploring combinations of envelope retrofits and HVAC upgrades to achieve great energy savings. See Table 1 for a summary of these deep energy retrofit case studies. Additional case studies in North American cold climates are important, as less than 5% of publications originate in continents other than Europe and Asia (Loli and Bertolin 2018). This is also important as (Blaszak and Richman 2012) noted that energy retrofit strategies will have varying effects for different types, styles, and ages of houses, which varies from region to region.

Table 1. Findings from case studies about energy retrofits in historic buildings in cold climates

Table 1: Findings from case studies about energy retrofits in historic buildings in cold climates

2. Literature review of methodologies

Many papers – Ma et al. (2012), Loli and Bertolin (2018), Ascione, de Rossi, and Vanoli (2011), Adhikari et al. (2011), Alongi et al. (2015) and Grytli et al. (2014), Young (2012) – outline methodologies for conservation and sustainability projects with a commonality being that a high level of understanding of the building is required prior to analyses. Ma et al. (2012) outline a methodology for cost-optimal retrofits of existing buildings. This methodology describes many important things to consider in the methodology for a HB energy retrofit such as an initial survey of the building and definition of project objectives, energy audit and performance assessment, identification of retrofit options through several analysis methods (energy, economic, risk assessment, etc.), site implementation, and validation and verification of energy savings. Loli and Bertolin (2018) outline that the following should be included in a methodology balancing heritage conservation and sustainability: identifying the cultural values of a building and protecting them, determining level of intervention based on condition of structure, reducing costs without compromising occupant comfort, compatibility with existing materials, and life-cycle assessment to maximize reuse of materials and reduce carbon footprint of interventions.

Ascione, de Rossi, and Vanoli (2011) consider energy savings and economic benefit in their methodology. This study collected data from heat flux sensors, calibrated to historical monthly, and simulated individual retrofit options. Then, the most viable retrofits were combined to predict the total savings. A cost analysis was conducted for all the retrofit options. A similar approach is used in this paper, but a simple payback is calculated to consider cost.

Adhikari et al. (2011) proposes a procedure for modelling and simulating energy performance of HBs: documenting the building, calculating thermal transmittance of walls, climate and microclimate data, model calibration using measured indoor air temperature, and then simulation. Alongi et al. (2015) conducted a detailed performance analysis of a historic castle in Italy and described their modelling and simulation methodology using TRNSYS, but there was no calibration method explained. Roberti, Oberegger, and Gasparella (2015) start with a detailed initial model based on energy data of the building, a sensitivity analysis, calibration to indoor air temperatures, and validation of the calibration. The importance of the calibration is to determine values for the most sensitive parameters in the model.

Grytli et al. (2014) outline a methodology and case study that assesses retrofits based on their environmental impacts and their impact on heritage values. Their approach focused on the heritage assessment from archival analysis, to surveying, condition assessment, values assessment, and life-cycle assessment in addition to the energy assessment, which was not seen in many other papers that considered energy retrofits in HBs. For the heritage assessment, a numerical ranking was given based on how an intervention would impact the heritage values of the building. A life-cycle assessment of energy which includes embodied energy was not conducted for this paper but is a consideration for future work.

This paper aims to combine the many critical elements described above in these methodologies to present one cohesive general methodology for conducting deep energy retrofits in historic buildings, while considering other important aspects (occupant comfort, cost, etc.). The application of the methodology is demonstrated for an historic wood-framed house in Ottawa, Canada, where there has not been a study published before for this harsh cold climate and where conservation is considered. The methodology delineates the creation of an accurate energy model for an HB without invasive time-consuming testing and an appropriate level of documentation by leveraging the use of calibration to measured data. A decision-framework is presented to guide decision-making for balancing the trade-offs between heritage conservation and energy savings objectives. While this study is conducted on a single-detached home, the goal is that this methodology and decision-framework can be broadly applicable to any type of building.

3. Methodology and case study

Figure 1 presents an overview of the methodology used for deep energy retrofits in the historic house studied for this paper. The following sections explain the use of the methodology in the case study and demonstrate potential energy savings for the final recommended package of measures.

Figure 1. Proposed methodology for deep energy retrofit analysis in historic buildings. Size of step suggests relative amount of time to be spent. The process of additional measurements and re-calibration may be iterative. Validate the model using different year weather data and utility bill data if possible.

3.1. Historical analysis and documentation

A deep energy retrofit analysis was conducted for an historic house in Ottawa that has a brick-cladded wooden balloon frame construction, semi-rigid cavity insulation (retrofit), a stone foundation, gas boiler for heating, and natural ventilation for cooling. The house is used as a bed and breakfast, a daycare, and home offices. Figure 2 shows the front façade of the house. The back section of the house was added more than 40 years ago, whereas the original main part of the house was built in 1889.

Figure 2. Front view of historic house studied in Ottawa (Ide 2019).

3.1.1. Character-defining elements

An historical analysis was first conducted to determine the character-defining elements of the house following guidelines in Parks Canada (2006) and Van Balen (2008). The Nara grid outlining key values of the place are provided in Table 2. A thorough condition assessment of the house was performed including indoor environmental quality (IEQ). Many features were found to be character-defining of the Dutch Colonial and Queen Anne Revival architectural styles of the house (e.g. large bay windows for daylight, mansard roof, red brick exterior finish), the historical connections (James and William Hopper architects, Glebe community leader Angela Keller-Herzog), and contextual value of having mature trees on site. Documentation was performed earlier in the year using photogrammetry, laser scanning, total station, and hand recording.

Table 2. Nara grid for case building. A rating of low, medium, high, or very high significance (outstanding universal value) has been assigned to each value within the Nara Grid

3.1.2. Measured energy use

As part of the historical analysis of the building, past utility bill data was analysed and annual energy use (kWh/m2) was compared to similarly aged buildings and newer buildings with the same construction type. Knowing the heritage values, condition, and comparative energy use intensity of the house help determine a realistic energy reduction target. For comparison, space heating energy statistics were analysed because this was the most complete data and is dominant in residential buildings in Canada. The oldest and most complete data in Canada for buildings constructed in 1990 show that the space heating energy use is 219 kWh/m2 and 206 kWh/m2 for Canada and Ontario (NRCan 2016). The average space heating for buildings constructed between 2011–2016 is 136 kWh/m2 and 133 kWh/m2 for Canada and Ontario.

The measured gas and electricity data from utility bills for 2015 to 2017 were analysed. The average annual electricity use is 40 kWh/m2 (24% of total) for lighting, appliances and a small portion for pumps used in the hydronic heating system.2 The electricity use does not have much variation throughout the year because of its type of usage (mostly appliances and lighting). The average annual gas use is 128 kWh/m2 (76% of total) for the gas boiler used for heating and a tankless gas heater for the domestic hot water (DHW). The gas use increases for the heating season and decreases for the cooling season when it is only used for DHW.

A heating degree day analysis3 was conducted for the gas energy use since it is used for space heating and DHW heating (Figure 3). A regression was conducted to determine what portion of the gas energy use was not correlated to the required heating for the house, therefore indicating the energy use for the DHW. The coefficient of determination is 0.876, since this building envelope allows more airflow, the performance of the house is more dependent on the outdoor temperature than a newer building that has a more air-tight envelope. This regression analysis also gave the DHW baseload of 1.7 kWh/m2 per month and about 20 kWh/m2 per year. This means that space heating is about 108 kWh/m2 which is 19% better than Ontario’s 2011–2016 space heating energy use for residential buildings, so this house is already performing better than many houses. The house also has photovoltaic panels (PVs) on the roof and shed, which generates an average of 25% of the electricity use for 2015–2017. The energy generated by the PVs is supplied to the grid and is credited on the electricity energy bill. The electricity use increases during the summer as the number of guests in the bed and breakfast increases during this time.

Figure 3. Heating degree day analysis for the house with a base temperature of 15°C. Baseload is 1.7 kWh/m2/month and 20.4 kWh/m2/year of domestic hot water energy use.

3.2. Detailed energy model, sensitivity analysis and calibration

A detailed energy model was created using OpenStudio and EnergyPlus (Energyplus (version 8.9) 2018; Openstudio (version 2.6.0) 2018). To achieve a deep energy retrofit, multiple retrofits need to be combined, and a simple model will not predict the interactions of multiple retrofits. First, a representation of the geometry of the building (see Figure 4) was created using floor plans (see Appendix) and elevation drawings that were recorded in 2018. All sides of the final building model are shown in Appendix.

Figure 4. Geometry of model in SketchUp plug-in and view of back of house from the same perspective (Larissa Ide 2019). Shading objects used to represent trees and adjacent building.

Simplifications were made for the geometry as the geometric resolution does not significantly impact the simulation results for the purposes of this study. This improves reproducibility, reduces model complexity and simulation run time, while keeping the necessary data. The simplifications made were: modelling smaller dormer windows as in plane windows, creating a flat-roofed third floor/attic with an equivalent volume as the mansard roof attic (as the existing roof has a small slope), and using an average floor area to have a consistent footprint for each floor (no small overlaps in floors). Shading by trees (deciduous and evergreen) was represented in the model to account for reduction in solar radiation, daylighting, wind and change in long-wave radiation to environment. Thermal zones were separated by floor (because of stack effect and air leakage) and by the different constructions of the house (main part of house separated from more recent addition), resulting in seven thermal zones in total. Separating the house into these thermal zones allows for adjustments for each individual thermal zone as necessary to match the behaviour of the existing house. The schedules, plug loads, occupancy, and HVAC were defined for the model based on information obtained from an interview with the house owner and quick survey of the house. The average permanent occupancy for the calibration year was four occupants. Additionally, there are 5–10 children and one adult in the house during the day for the daycare that is run in the house. The occupancy varies throughout the year because of the guests that come for the bed and breakfast service in the house. The summer occupancy and DHW use are higher than the rest of the year due to increased number of guests. Unique schedules for occupancy, equipment and DHW were created based on this information instead of using ASHRAE recommended occupancy schedules.

3.2.1. Modelling

The temperature setpoint for heating was supposed to be 21°C, but because of the thermostat proximity to the entrance of the house, the house may be heating more often as if the setpoint were greater. This was accounted for in the calibration by increasing the temperature setpoint. The radiant hydronic heating system was modelled using a baseboard radiant convective water system. A baseboard radiator was assigned to each conditioned space. The radiators were connected to the gas boiler with a water outlet temperature of 82°C indicated by the boiler. The initial air infiltration was modelled using the calculated flow coefficients based on air changes per hour measured in a blower door test in 2017 which indicated 11 ACH at 50 Pa. The method used to calculate the air infiltration for the house in the energy model was the AIM-2 model by Walker and Wilson (1998) which is appropriate for smaller, residential buildings. The flow coefficient was calculated rearranging the equation in the AIM-2 model: (1) C=Q/ΔPn(1)

where Q is the air infiltration flow rate (m3/s), C is the flow coefficient, ∆P is the pressure difference between inside and outside (about 50 Pa for a blower door test), and n is the power law pressure-flow relationship assumed to be 0.67 (Energyplus Version 8.9 Documentation 2018). Air changes per hour (ACH) were converted to flow rate using the volume of the house and then converted to flow coefficients using equation 1. The flow coefficients for each of the seven thermal zones were calculated for each iteration of total ACH for the house and weighted 20% basement, 20% ground floor, 20% level 1, and 40% level 2 based on the patterns observed in thermography. The air infiltration fluctuates with the windspeed and temperature difference between indoors and outdoors throughout the year (Energyplus Version 8.9 Documentation Input Output Reference 2018).

3.2.2. Sensitivity and calibration

A one factor at a time sensitivity analysis was conducted to identify the most sensitive parameters that may require additional measurements to improve accuracy of calibration and to inform most effective retrofit options. The air leakage, heating setpoint temperature, boiler efficiency, wall thermal resistance and window properties were the most sensitive parameters. Parameters that were found to be less sensitive in the model were the ground temperature, stone foundation properties, and others, so reasonable baseline values were chosen for these parameters. Figure 5 shows a summary of the sensitivity analysis results and Appendix for the full sensitivity analysis results.

Figure 5. Sensitivity analysis results for change in annual gas energy use from baseline 117 kWh/m2/year. Most sensitive parameters were determined to be air changes per hour of air leakage, heating setpoint, boiler efficiency, main house insulation, and improved windows.

An actual meteorological year (AMY) file (Urbandale 2017) was used for calibration of the model to utility bill data over a full year following the ASHRAE Guideline 14 criteria for calibration to monthly data: NMBE (Normalized mean bias error) between −5% and 5% and CV(RMSE) (Coefficient of variation of the root-mean-square error) <15% (ASHRAE 2014). Information obtained about the house was input into the initial model to improve estimates and acceptable discrete values of sensitive uncertain parameters were simulated to complete calibration (heuristic, evidence-based calibration as described by Coakley, Raftery, and Keane 2014). The energy model was manipulated interactively using the Python package Eppy. The energy use for plug loads (appliances) and lighting was estimated using the average observed in the actual electricity consumption from 2015–2017 (see Appendix for details of model inputs). The R-values were estimated based on single-family-detached houses in Toronto of similar age and construction in Blaszak and Richman (2012). The simulated gas energy output was compared with the measured gas energy data using the NMBE and CV(RMSE) as indicators, and the most sensitive parameters were adjusted to achieve better NMBE and CV(RMSE) values based on observations in the house during site visits.

Thermography was used to qualitatively enhance understanding of thermal bridging, air infiltration, and other issues in the house. The thermography revealed that significant air infiltration occurs around the baseboards on the ground floor (Figure 6(a)), around windows, and around the header of the second floor. The thermography revealed a large hole in the basement wall that was previously undiscovered. Thermal bridging was considered in the model through the chimney because of its large area (Figure 6(b)). The air infiltration from the blower door test was a good initial value for the energy model, but due to uncertainties and sensitivity of this parameter, it was adjusted for the calibration of the model.

Figure 6. (a): Example of thermography done for house showing patterns of air infiltration along baseboards of north wall of house at ground level and digital photo. (b): Heat transfer around roof (third floor) and chimney.

Digital photo to accompany thermography. Taken during day to see building features.

The effective R-value of the walls in the main house was estimated to be around 2.2 m2-K/W based on an initial value of 3.32 m2-K/W for a double-wythe brick with 38 × 140 mm fibreglass batt insulation from Blaszak and Richman (2012) and then adjusting for calibration. For the addition of the house, 3.0 m2-K/W was the thermal resistance because of the recent insulation improvement which was found in documentation for the retrofit. The schedules for occupancy, equipment use, and DHW were adjusted based on detailed occupancy variation throughout the year (see Appendix for occupancy schedules). Table 3 summarizes important modelled parameters and discrete values tested for calibration of uncertain parameters. The calibration achieved for electricity had a NMBE of 4.1% and CV(RMSE) of 13.7%. The calibration achieved for the gas use had a NMBE of −2.1% and CV(RMSE) of 14.9% (see Figure 7). The annual gas and electricity consumption of this baseline-calibrated model with a TMY (typical meteorological year) are 121 kWh/m2 and 39 kWh/m2.

Table 3. Model and calibration parameters

Figure 7. Measured gas and calibration results. NMBE = −2.1% and CV(RMSE) = 14.9%.

After calibration, the model should be validated with a separate year of measured utility bill data (if possible) to ensure that the model still meets calibration guidelines and check for overfitting. Validation with another year of utility bill data was not possible for this case study because the previous and following years involved additional changes to the building. The changes included additional insulation, air sealing, LED lighting retrofit, and appliance upgrades which would not allow to validate the baseline energy model. Due to the lack of validation with a different set of weather data, there is a possibility of overfitting the model to the calibration year.

3.3. Viable retrofits and ranking of options

3.3.1. Decision-framework for balancing heritage and sustainability objectives

Once the model is fully calibrated and validated, viable retrofit options should be selected based on sensitive parameters and other risk factors in terms of conserving heritage values. For example, if the lighting power is sensitive and can be changed without impacting the heritage character, a lighting retrofit that maximizes daylight should be selected for the analysis. This may also favour inherent energy-efficient features in an HB and help conserve the heritage characteristics. Important risk analyses to consider include heritage impact, hygrothermal analysis to assess change in moisture behaviour (e.g. when adding insulation), cost analyses, and occupant health/comfort.

The balance between heritage and environmental sustainability can be achieved by comparing the energy savings and risk to heritage values, while also considering other aspects of sustainability (e.g. occupant comfort). The following metrics and criteria may be used in the analysis of balancing heritage and sustainability. Environmental Sustainability Metrics and Criteria: Overall wellness of occupants (IEQ, CO2, daylight); GHG emissions including embodied carbon of new materials vs existing materials; Reduction in total EUI (energy use intensity) by 50%4; and Climate change resilience – indoor air temperature in free-floating conditions, change in energy use and renewable energy. Heritage Metrics and Criteria: Minimal intervention (refer to the Standards and Guidelines 2010); Respectful to heritage values; Reversibility (i.e. can not solely use spray foam on entire wall, but may be able to do this on new materials which do not have heritage values); Adaptability to current and future needs; and Proven technology.

Knowing the sustainability and heritage criteria above, the balancing of these two broad objectives becomes clearer, but which has a higher priority in the comparison between the two? Does heritage have higher priority over environmental sustainability, or the other way around? Prioritization will depend on the significance of the element impacted by each retrofit option. That is why the heritage impact assessment is crucial in utilizing the decision-framework to balance heritage and energy reductions. The potential heritage impact assessment should be carried out following Section 5 of ICOMOS (2011). To summarize this evaluation method, a more significant element of a place and a greater severity or scale of impact results in a greater overall potential impact.

Many retrofit options will have low heritage impact and implementing more changes to more HBs would help to mitigate the severity of climate change and conserve the heritage values in a new manner. This does not mean drastically changing the existing building to construct a net-zero energy/carbon building. With too many drastic changes, there is a risk of jeopardizing the authenticity and integrity of the heritage values of the building. It is still best to utilize the existing materials of the building and prioritize minimal intervention. Minimal intervention can be executed in different ways. It may mean repairing the existing historic windows so that they are airtight and adding a storm window because this is the minimal intervention required to improve the thermal performance of the windows.

First, we look at reducing the current operational energy use intensity by a total of 50%. Any measure that contributes to EUI reduction individually should be considered. A measure is determined to be viable based on reduction in EUI, occupant comfort, and respecting heritage values using the guidance provided by the decision-framework shown in Figure 8. Then, it can be evaluated based on climate change resilience, greenhouse gas emissions, and other project objectives not included in this study.

Figure 8. Decision-framework for balancing energy reductions and heritage conservation, while also considering occupant comfort. If not proven technology, perform in-depth research to understand the risk before considering the technology.

3.3.2. Retrofit options

First retrofit measures for the building envelope were considered because this would impact the peak load for the HVAC system. Adding insulation was considered for the interior of the exterior walls because of the character-defining exterior brick cladding. To determine appropriate levels of insulation, the existing wall assembly was examined and only a thickness of insulation that would fit in the existing stud cavity or on the interior of the studs before the new gypsum board was considered (see Figure 9 for wall details). It was important to ensure no increase or minimal increase in wall assembly thickness. PIR (polyisocyanurate) rigid insulation board was considered as the insulation material for its high thermal resistance of 0.022 W/m-K and potential to easily remove this retrofit. Since there is already some insulation on the basement walls but not on the floor, insulation for the basement floor was considered. The roof was recently insulated in 2013 with blow-in cellulose and is performing well according to the thermography. It is unlikely that the owner would invest to re-insulate the roof to a higher thermal resistance, so it was not explored as an option.

Figure 9. Wall details of existing wall in main part of house and three rehabilitation scenarios. Similar rehabilitation scenarios for addition of house except stud space is 152 mm and second scenario includes 12 mm extra of PIR on the interior side of the stud space.

Window retrofits were considered to reduce infiltration and improve thermal resistance, such as adding a storm window to effectively make the assembly a triple-pane clear glass window or a new triple-glazed window with low-E coating and argon gas. If it is possible to modify the existing window to have low-E coating and argon gas filling, this would be a good option to conserve the authenticity of this value of the house (Historic England 2016; Parks Canada 2010). Air sealing the leakage pathways in the house was also explored as an option.

On the mechanical side, an air-source heat pump was also analysed as an option because the house does not have mechanical ventilation for cooling during the summer, and the owner noted that guests are feeling uncomfortable as temperatures increase each summer. This would not have a negative impact on the character of the house if the historic radiators are conserved. The air-source heat pump is assumed to provide supplemental heating to the house to reduce gas energy use of the boiler above outdoor temperatures of −5°C (NRCan 2004). Table 4 provides a summary of the retrofits explored, values simulated for the retrofit analysis, heritage significance of the element impacted by retrofit, scale of the change, and the potential overall heritage impact.

Table 4. Summary of retrofits explored, values simulated for analysis, and heritage impact assessment

3.3.3. Building science considerations

Since there are changes to the thermal resistance and vapour permeability of the wall assembly, the risk of moisture accumulation needs to be considered. To consider the moisture accumulation, vapour diffusion through the wall, RH (relative humidity) at each interface between materials and the rate of condensation were calculated. One way to minimize the vapour diffusion from the interior to the exterior in a cold climate and prevent moisture accumulation in the insulation is to add a vapour barrier to the inner side of the insulation layer. Vapour diffusion calculations were done using the simple Glaser Method. The aluminum foil on the PIR insulation acts as a vapour barrier and provides a vapour resistance of 0.35 Pa-s-m2/ng. The outdoor conditions of −20°C and 75% RH were chosen for Ottawa as a worse case where the temperature is lower than average, and RH is higher than average. Indoor conditions are assumed to be 21°C and 35% RH, and the insulation has a thermal resistance of 3.7 RSI. With these conditions, there is no condensation plane within the wall assembly. The relative humidity within the wall assembly reaches a maximum of 72% in the bricklayer under these conditions.

3.3.4. Ranking of options and recommended package

A parametric analysis aided in the ranking of retrofit options. The results of the parametric analysis are shown in Figure 10(a) which are based on the calibrated model. The modelled air-source heat pump produces the highest individual annual energy savings of about 27% of the total baseline energy use. Reducing air infiltration by 80% resulting in 2.8 ACH at 50 Pa achieves the next highest energy savings of about 12%. Thermal resistance of 5.1 m2-K/W in the main house exterior walls achieves 11% energy savings. Reducing air infiltration by 70% and insulation to achieve 4.1 m2-K/W walls also achieve significant energy savings of 10% and 8% annual energy savings, respectively. New triple-pane low-E argon windows and rehabilitated triple plane low-E windows both achieve 4% energy savings. Insulating the basement floor to a high level of 6.0 RSI only achieves 2% energy savings so it was not be considered for the retrofit packages because of its invasiveness and low energy savings. Daylight controls save about 1% energy through the reduction of electrical lighting use and has no impact on the heritage character of the building, so it was considered for the retrofit packages. Increasing insulation in the addition of the house achieved only 1% energy savings due to the smaller surface area and higher starting level of insulation. It was recently insulated in 2017, so it is also not practical to retrofit the walls of the addition of the house again (time, money, disturbance to occupants). Specific envelope improvements were selected based on the sensitivity of a retrofit on the energy savings; for example, if the walls are to be insulated, it can be insulated to 4.1 m2-K/W instead of 5.1 m2-K/W because of the small change in savings (3%).

Figure 10. (a) Results of parametric analysis of retrofit options. Thermal resistances are indicated for the level of insulation, not the effective thermal resistance of the entire wall assembly. (b) Incremental annual energy savings of package of retrofits that balances heritage conservation and energy reduction target relative to calibrated baseline.

The retrofits that are viable from a sustainability and heritage perspective were combined and simulated to demonstrate the potential energy savings. Two different packages were simulated to compare the energy savings of highest feasible energy savings to a package which balances heritage conservation with energy reduction targets. A third option would be the minimal intervention approach of only air sealing large cracks in the envelope without any other upgrades which is assumed to achieve a 40% reduction in air leakage. This minimal intervention option would only achieve about 6% energy savings, which is not adequate for achieving deep energy savings. The best energy savings package (air-source heat pump, 80% reduction in air leakage to 2.8 ACH at 50 Pa, 5.1 m2-K/W exterior walls, new triple-pane low-E argon windows, and daylighting controls) achieved 48% energy savings from the baseline.

The package that balanced heritage conservation and energy reductions implemented an air-source heat pump, reduced air infiltration by 70% to 4.2 ACH at 50 Pa, insulated to have 4.1 m2-K/W exterior walls, used the existing windows with storm windows to achieve triple-pane low-E windows, and daylight controls. A 70% reduction in air leakage is likely more feasible in this house based on the construction and should not require the removal of the character-defining chimney. The lower level of insulation still achieves a significant amount of savings and only increases the wall thickness by 15.5 mm, which will have a smaller impact on the interior space. The new window frames compared to rehabilitated windows, do not have a significant difference in energy savings and achieve lower than 10% energy savings so it does not make sense to risk impacting the heritage character with new window assemblies. For this reason, storm windows are a good option to improve the window assemblies. This second package achieved 46% energy savings (Figure 10(b)). The incremental savings of each option demonstrates that it is crucial to simulate retrofits combined to get an accurate estimate of total savings and impact of each retrofit when packaged with others. It is also recommended that additional solar PV panels are considered to reduce electricity energy consumption, GHG emissions, and payback period of retrofit package (see Section 2.4 for details). With an additional 26 kWh/m2 PV power generation and adjusting baseline value to consider the current 11.4 kWh/m2 PV power, the total energy savings for this package is 67%.

3.4. Cost analysis

A cost analysis was conducted to compare the operational costs of the baseline and retrofit scenarios and determine the simple payback for the capital costs (CAD$) of the retrofit package. Capital costs were calculated by multiplying the cost per unit by the number of units. The windows had the highest capital cost of all the building retrofits (excluding renewables) at 20,730. USD For this reason, two different window options were analysed to see if capital cost could be reduced. The first option explored was a rehabilitation of the existing windows and adding a storm window with low-E coating. The second option explored was the rehabilitation of the existing windows and adding only a low-E film to the interior side of the glazing. The daylight sensors had the lowest capital cost of 98.50. USD

Electricity costs increase by 543 USD (21%), and gas savings are 1061 USD (41%), resulting in total annual savings of 518 USD (20%). New PVs on available site area were considered for the retrofit package to help offset costs and carbon emissions. Based on the average of 2015–2017 current PV generation, the generation split between winter and summer was estimated to be about 32% during the winter and 68% during the summer. About four times greater annual operational savings of 2151 USD (84%) are achieved when new PVs generating 8608 kWh/year are added.

The simple payback with the window retrofit is around 35 years for just the low-E film and 62 years for a full retrofit when the savings of renewables is not considered. With additional PVs, the simple payback period is shortened significantly to 21 years and 28 years for the two options. The reduction in energy savings for just the low-E coating is about 2%. From a cost perspective, option 2 of rehabilitating the windows and adding only a low-E film is the better option. However, adding the storm window will have a beneficial impact on the occupant comfort, so this should be a consideration in the final decision.

3.5. Discussion

The incremental energy savings of the recommended retrofits shown in Figure 10b demonstrate that upgrading the HVAC system and reducing air infiltration into the house are the most important retrofit measures to implement. Reducing the air infiltration to 4.2 ACH at 50 Pa (70% reduction) was recommended because any lower is unlikely to be feasible in the house due to the air-leakage pathways in the existing construction. Obtaining this level of airtightness would require the addition of an air-vapour barrier within the envelope. The hole in the basement, the chimney, and other problem areas should be sealed to achieve this level of airtightness. When air sealing the chimney, the fireplace should be conserved as a decorative character-defining element. Since the benefit of adding insulation is greater than the capital cost shown in literature (Chapman et al. 2009), it should be implemented along with the addition of an air-vapour barrier to prevent moisture issues in the envelope.

Greater energy savings from the air-source heat pump compared to other retrofits were expected since replacing older HVAC equipment was found to have the greatest energy savings in literature. The air-source heat pump increases electricity use by 21 kWh/m2/year when combined with envelope retrofits, whereas when implemented alone, it increases by 47 kWh/m2/year. This demonstrates the importance of retrofitting the envelope before applying HVAC retrofits, so the systems can be reduced in size and operate efficiently. A ductless air-source heat pump should not be an invasive retrofit for the heritage values of the building, and it improves the comfort of occupants by providing cooling in the summer, making the building more sustainable long term.

The end use breakdown for the recommended retrofits is shown in Figure 11 (excluding renewables). The gas energy used for heating can be reduced to 2.9 kWh/m2/year. After retrofit, the most carbon-intensive source of energy use is the DHW, heated using natural gas,5 which can be reduced with an integrated air-source heat pump system. The appliances are estimated to use about 25.5 kWh/m2 of electricity so this should be the next focus for further energy reductions using kill switches and automation. Further reductions in electricity can be achieved using a smart thermostat, which were not modelled due to uncertainties in occupant behaviour. More PV panels should be added on-site to provide additional electricity generation, especially with the increased electricity consumption due to the air-source heat pump.

Figure 11. End use breakdown for recommended retrofits combined.

A balance between environmental sustainability and heritage conservation objectives was achieved in the energy retrofit recommendation as the two are mostly complimentary to each other. The methodology presented does not aim to achieve an optimized solution in terms of energy efficiency. Due to the qualitative nature of defining the level of significance of heritage elements; the outcome is an appropriate balance between the objectives to protect heritage values and mitigate climate change. The trade-offs between the two objectives are minimal and ensure that society continues to thrive through the connection to a sense of place, as promoted in Sustainable Development Goal 11.4 (United Nations 2019), and through minimal impact on the environment. Following the methodology presented in this paper facilitates the appropriate balance by first understanding the heritage values of a place, assessing the condition of the building and site, carefully selecting options for the retrofit analysis based on the heritage values and risk assessments of the site, and recommending only the retrofits that achieve a significant amount of energy savings and/or require minimal intervention. If a retrofit option involves an invasive modification impacting the character defining elements of the building (e.g. replacing historic windows, removing historic fabric, adding new elements that diminish the historic elements), then the energy savings must be great enough, and this option must not have a detrimental impact on the heritage values of the site.

4. Conclusion

In conclusion, about 67% energy savings can be achieved in an historic single-detached house in a cold climate through deep energy retrofits of the building envelope, the HVAC systems, and addition of on-site renewable energy, while conserving the character defining elements of the house. A balance between energy savings, sustainability, occupant comfort impacts, cost and heritage conservation were achieved using the methodology and decision-framework presented in this paper. The recommended retrofits include reducing air infiltration by 70% to 4.2 ACH at 50 Pa, adding insulation and an air-vapour barrier to achieve 4.1 m2-K/W exterior walls, rehabilitating the windows to triple-pane low-E windows, the addition of an air-source heat pump for supplemental heating, daylighting sensors and controls, and additional on-site solar PVs. The annual operational cost savings are 84% with the recommended retrofit package. The simple payback for the capital costs of this retrofit package maybe 21 years to 28 years depending on the window retrofit chosen. The methodology in this study may be applied to retrofits for all types of existing buildings to provide more holistic sustainability retrofit recommendations. Further research for this study will include comparing retrofit carbon emissions to carbon savings due to energy reductions. The results of this study suggest that the presented methodology and decision-framework facilitate deep energy retrofits in historic residential homes in a cold climate, while conserving the heritage values of an historic house. In most cases, the energy retrofits benefitted heritage conservation values and occupant comfort, demonstrating how heritage conservation can play a key role in sustainable development.

Acknowledgments

This work was funded through the NSERC CREATE Heritage program at Carleton University (NSERC grant number: 465459-2015). The author would also like to acknowledge that the owner of the studied house was very accommodating of the research conducted.

Disclosure statement

The authors declare no conflicts of interest.

Additional information

Funding

This work was supported by the NSERC CREATE Heritage [465459-2015].

Notes

1 An historic building in this paper defines a building that is 40 years or older and has retained most of its original building fabric.

2 Note: The power generated by the existing PVs on site is not included in the baseline EUI of the house since PV calculations are done outside of the detailed energy model to be calibrated. There is no air conditioning or mechanical ventilation in this house.

3 See (Meng and Mourshed 2017) for details on heating degree day analyses.

4 Deep energy retrofits are defined as “a major building renovation project in which site energy use intensity (including plug loads) has been reduced by at least 50% from the pre-renovation baseline with a corresponding improvement in indoor environmental quality and comfort” (IEA Annex 61 Business and Technical Concepts for Deep Energy Retrofits of Public Buildings 2017).

5 The carbon intensity depends on region; in Ontario, Canada, the annual marginal emission factor of electricity is about 0.134 kgCO2/kWh (Sotes 2019) and 0.525 kgCO2/kWh for natural gas (Intrinsik Corp., 2016).

References

Appendix

A: Floor Plans used to create initial model geometry. Only second floor is provided as it gives an idea of massing and spaces in building (other floors are similar).

  (B. Hobson et al. 2018)

B: Model Details

The figure below shows occupancy during holidays/Saturdays/Sundays. The typical occupancy was assumed to be 3.6. The occupancy reduces for a few hours during the morning, assuming that occupants may need/want to leave for errands or other things. The figure below also shows the occupancy schedule for a weekday; it starts at the typical occupancy and increases to six people to account for the daycare provider and children for the daycare. Then at 5 pm, the daycare service ends, and the occupancy returns to the regular level.

  Occupancy schedule for holidays (left). Assumed Jan 1–5 and Dec 25–31 are typical holidays. Occupancy for weekdays Jan 6–Dec 24 (right).

Internal Gains:

People: Fraction radiant = 0.3. Fraction convective = 0.7.

Lighting was 2.56 W/m2 or 2.40 W/m2 depending on the zone. Fraction radiant = 0.5. Fraction convective = 0.5.

Electrical equipment (appliances): 6.2 W/m2 for all zones. Fraction latent = 0.1. Fraction radiant = 0.2. Fraction convective = 0.7.

Total sum of internal loads = 8.6 to 8.7 W/m2

Alternative formats

 

Related research

People also read lists articles that other readers of this article have read.

Recommended articles lists articles that we recommend and is powered by our AI driven recommendation engine.

Cited by lists all citing articles based on Crossref citations.
Articles with the Crossref icon will open in a new tab.