Fabric Retrofit of a Hard-to-Treat, Pre-1919 House in Preparation for Heat Pump Use

Abstract

The uptake of low-carbon domestic heating systems is a significant strategy towards global targets of reducing greenhouse emissions and mitigating climate change. Pre-1900 hard-to-treat houses will still be existing in the next 25 years, and they have the greatest potential for improved energy-efficiency. This study investigates the potential of fabric retrofit to prepare an older, hard-to-treat house type for heat pump use. The house type was modelled in DesignBuilder and validated using the Ulster University test house. The wall, loft and floor insulation, as well as glazing upgrades can yield up to 50% reduction in heating demand for a hard-to-treat house type, thereby preparing it for heat pump installation. Additionally, upgrading insulation and glazing in line with the current building standards was cost-effective, with a net present value of approximately GBP 12,000.

1. Introduction

In the United Kingdom (UK), domestic stock retrofit is essential to achieve the government’s targets of reaching net zero emissions by 2050, [1,2]. The majority of the UK’s older (pre-1919) housing stock has the largest scope for improved energy efficiency through retrofit [3,4]. They commonly feature underperforming solid walls, uninsulated roofs and floors, and single glazing; therefore, they are classified as ‘hard-to-treat’ [3,5]. In the UK, the Standard Assessment Procedure (SAP) is a point-based system which rates the energy and environmental performance of houses from bands A to G (with A and G being the most and the least efficient, respectively) [6]. The older UK housing stock has the lowest SAP ratings, with pre-1919 houses more being likely to have an SAP rating of E, F or G [7,8,9,10].

A proportion of 75% of fuel-poor UK households live in a pre-1965 house [11]. Fuel poverty is compounded in hard-to-treat dwellings, as old fabrics have high air permeabilities (or a low airtightness) and tend to allow draughts. Draughts contribute to higher heat demand than what can be afforded by fuel poor households. However, these properties will still be in use by 2050 [3]. Retrofitting older UK homes can be complex as they were originally designed for past lifestyles [12]. The fabric-first approach is well suited to the retrofit of older house types, and low-risk measures such as draught-proofing and loft insulation should be considered before high-risk measures such as solid wall insulation [12].

Despite general government guidance, each UK region is expected to develop tailored energy strategies to meet its unique energy-efficiency needs [13,14,15,16]. In Northern Ireland (NI), the government recently published its energy strategy, which aims for a 56% reduction in energy-related carbon emissions from the 1990 levels by 2030 [17]. It states that improved domestic fabric energy-efficiency will be essential in the transition to low-carbon heating. Energy-efficiency upgrades will be made to the existing housing stock to make them heat-pump ready while new builds are expected to install similar low-carbon heating systems. Therefore, there is a need for research on how the fabric-first approach can facilitate retrofit in preparation for heat pump installations in NI.

1.1. Problem Statement

Heat pump installations will be combined with the uptake of domestic energy-efficiency measures, as the UK works towards net-zero emissions by 2050 [1]. Therefore, improved fabric energy-efficiency is essential to the uptake of domestic heat pumps, which are typically air-source and ground-source heat pumps [18]. Potent measures, such as solid wall insulation and improved glazing should be implemented on fabrics towards heat pump installation [18,19]. In 2020, Lingard investigated the impact of retrofit of a solid-walled dwelling in readiness for heat pump use [20]. He suggests that solid wall insulation and low U-value glazing are the most cost-optimal retrofit measures, which can facilitate demand reductions of approximately 80%. He states that if fabrics are retrofitted in line with current building regulations, heating demand and emissions can be reduced by up to 65%, taking the input electrical demand for the heat pump to under 1 kW.

Fabric retrofit encourages a reduced heat demand and, therefore, lower flow temperatures that supply low-grade heat [19]. Low-grade heat supply can be maximised by the output of heat emitters (such as radiators) and large bore pipework [18,20,21]. There are constraints on increasing the domestic heat emitter sizes or changing the pipework in older dwellings [19]. Consequently, fabric retrofit is a more feasible option than these and is effective in encouraging the feasibility of low flow temperatures [19]. Additionally, the Carbon Trust recommends that all buildings with an SAP rating below D require fabric upgrades for flow temperatures of 55 °C to be possible [19]. Their study presents scenarios where the SAP rating of dwellings improves from band D to bands A, B and C after whole-house/deep retrofit (such as insulation, glazing, ventilation and airtightness upgrades) and the installation of heat pumps.

Accordingly, existing studies have evaluated the performance of heat pumps in various house types. In 2023, Mohammadpourkarbasi, et al. compared the impacts of Passivhaus retrofit standards and conventional retrofit standards on pre-1919 houses built in Manchester [22]. Furthermore, they investigated the differences between deep retrofit and shallow retrofit with a heat pump. Their results indicate that most effective retrofit methods should be considered, materials used in retrofit have a significant role to play in retrofit scenarios, and life cycle analysis should be applied in retrofit scenarios. Finally, they suggest that deep retrofit scenarios with natural materials reduce demand and emissions more compared to shallow retrofit with heat pumps. However, their study does not evaluate how retrofit can prepare a pre-1919, hard-to-treat house for heat pump installation. It does not consider the cost implications of the suggested deep and shallow retrofit scenarios. In 2023, Saffari, et al. examined the effectiveness of an air–water heat pump for various retrofit scenarios for a residential building in Ireland [23]. Their results show that with deep retrofit and heat pump use, heat demand and emissions can reduce by up to 72% and 74%, respectively. Additionally, their results demonstrate that 41–80% of the capital cost for a heat pump can be recovered for all retrofit scenarios. However, their study does not consider the effectiveness of installing a heat pump in an older (pre-1919), hard-to-treat house type. In 2024, Considine, et al. assessed the cost implications of improving loft insulation towards the installation of an air source heat pump in various Irish house types [24]. Their study shows that grants will contribute to positive returns on loft improvements for homeowners. Additionally, significant energy savings can be achieved if the Irish Climate Action Plan 2023 retrofit targets are followed with these improvements. However, their study does not consider the cost and energy implications of other retrofit measures, such as wall insulation, floor insulation, glazing and airtightness.

In Northern Ireland, similar studies have been conducted. In 2018, Le, et al. evaluated the annual performance of a variable-capacity, high-temperature air source heat pump (ASHP) retrofitted in a mid-terrace house using TRNSYS models [25]. They recommend that in milder climates, the efficiency of the ASHP increases, and its running cost reduces. Furthermore, they indicate that while new buildings can reduce the COP, as the ASHP works at lower loads, more energy, CO₂ and cost savings can be achieved due to lower heat losses in new buildings. However, they do not provide any insights on fabric retrofit and its role in preparing the dwelling for heat pump installation. In 2019, Le, et al. conducted a techno–economic assessment of the performance of a cascade air-to-water heat pump (ASHP) retrofitted into a Northern Irish pre-1919 terrace house [26]. They suggest that reductions in the coefficient of performance (COP) of the heat pump are due to ambient temperatures and coupling with thermal storage but improve with direct heating and load-shifting. Their study demonstrates that while the cascade heat pump has higher operation costs compared to gas boilers and high-efficiency boilers, it facilitates CO₂ reductions by an additional 43%. Nevertheless, this study does not indicate the role of fabric retrofit in preparing a pre-1919 house for heat pump installation. In 2021, Abid, et al. compared the performances of a gas boiler and an ASHP installed in a 1900s mid-terraced house [27]. Their results show that the ASHP led to a 59% reduction in carbon emissions at a low heat supply temperature when compared with the performance of the gas boiler with 90% efficiency. The study provides valuable insights into the superior energy efficiency of a domestic heat pump, but the role of fabric retrofit in preparation for heat pump installation is outside of its scope.

1.2. Contributions and Justification for the Study

This study contributes new and valuable knowledge because it investigates the impact of fabric (insulation and glazing) improvements on annual demand and emissions of a typical hard-to-treat house type, using dynamic thermal simulation methods. The house type is modelled with building energy analysis software DesignBuilder version 6.0), and the model is calibrated against a real-life test house built on the Ulster University Jordanstown campus. This study’s recommendations are practical because validation is achieved with a real-life case. It demonstrates how the fabric-first approach can be used to improve the energy-efficiency of a hard-to-treat dwelling, thereby making it heat-pump ready. This study presents cost-benefit analyses, which can also inform on the financial ramifications of making hard-to-treat house types heat-pump ready.

2. Material and Methods

2.1. Three-Dimensional (3D) Model of a Typical NI Hard-to-Treat House Type

DesignBuilder is a high-quality, specialist software that simulates the environmental performance of buildings [28,29,30,31]. A 3D model of a typical hard-to-treat archetype terraced house was created in DesignBuilder. The terraced house is the most popular house type in NI and the second most common house type occupied by fuel-poor households (see Figure 1) [8,32]. The terraced house is the most common house type with uninsulated solid walls, single glazing and a need of loft insulation in NI.

This study’s hard-to-treat model was based on the purpose-built demonstration test houses located on the Jordanstown campus of Ulster University (see Figure 2). The test houses were built to imitate the pre-1919 to 1944 terrace houses located in Belfast, and both have identical construction features and layout (see

Table 1. Construction features of the house type and corresponding U-values.

Component	U-Value (W/m2K)
External solid wall made with 215 mm clay bricks, 15 mm inner plaster	1.81
Pitched roof with resin slate tiles, timber battens, Dupont Tyrek vapor barriers	1.63
Single glazing, 4 mm clear glass panes	5.87
External door	5.87 for glazed portions
Ground floor, 150 mm thick concrete slab, 75 mm screed	0.73
Air tightness (air changes per hour)	0.15

). The house on the left was used as the reference for the building model, which simulated the construction features of an uninsulated hard-to-treat house with an SAP rating of F (see Figure 2).

The external solid walls are made of 215 mm clay bricks with a 15 mm inner plaster. The pitched roof has resin slate tiles fixed on timber battens and Dupont Tyrek vapor barriers. The first-floor ceiling consists of 15 mm plasterboard with 19 mm plywood flooring in the loft. The external windows are single-glazed, with 4 mm clear glass panes. The ground floor is made up of a 150 mm thick concrete slab with a 75 mm screed.

The infiltration rate (airtightness) was assumed as 0.15 air changes per hour, as the air permeability was assumed to be greater than 5 m³/h.m² at 50 Pa [33]. The model has 2 floors, with a total living area of 113 m². On the ground floor, there is a lounge area and a kitchen, while the first floor comprises three bedrooms and a bathroom. The average household size in NI is 2.4 [34]. Therefore, this study simulated an occupancy of 2 people, with an occupancy density of 0.02 people/m². The left-side house was retrofitted with an air source heat pump (ASHP), which was also served as a reference for the DesignBuilder heat pump model. The test house heat pump is a Daikin Altherma high-temperature air-to-water heat pump with a COP of 2.5.

Table 2. Heat pump model specifications.

Model Name	Daikin Altherma HT
Heat pump type	Air to water (cascade unit)
Indoor model number	EKHBRD011ACV1
Indoor refrigerant	R134a
Outdoor model number	ERSQ011AAV1
Outdoor refrigerant	R410-A
Leaving water temperature *	25–80 °C
Manufacturer	Daikin
City	Ostend
Country	Belgium

* [35].

shows the specifications of the heat pump. The base-case leaving water temperature (flow temperature) was set at 80 °C. A thermal energy store (hot water tank) supported the performance of the heat pump.

Table 3. Heat pump specifications and corresponding DesignBuilder parameters.

Heat Pump Feature	DesignBuilder		Value
	Model Component	Parameter
Crankcase heater power consumption	Air-to-water heat pump coil	Crankcase heater capacity	33 W
Heating capacity for combination indoor + outdoor units		Rated heating capacity	11 kW
Indoor water circulation pump consumption		Condenser water pump power	87 W
Coefficient of performance (COP)		Gross rated COP	2.5
Ambient dry bulb temperature		Rated evaporator inlet air dry-bulb temperature	7 °C
Wet bulb temperature		Rated evaporator inlet air wet-bulb temperature	6 °C
Leaving water temperature (upper limit)	Water heater	Maximum temperature limit	80 °C

details the specification of the heat pump and the equivalent parameters in DesignBuilder, which were used to model the heat pump. The ASHP template features an air-to-water heat pump coil, a supply fan and water heater tank components (see Figure 3).

2.2. Heat Pump Modelling, Weather File and Heating Profile Calibration

An Eltek Genll wireless telemetry data logging system (manufactured in Cambridge, UK), was installed in the test house to monitor the performance of the heat pump [36] (Eltek Ltd., 2022). The system was made up of a logger (receiver) with wireless sensor transmitters which could interface with other sensors. A temperature and humidity sensor was installed in each room of the house (see Figure 4). The system transferred data from the datalogger to a computer through an automation application. The automation application installed on the computer downloaded data at midnight daily in CSV format. Data were extracted for the winter period of 12 December 2019 to 13 January 2020. These data were primarily collected towards evaluating heat pump performance. Therefore, it included thermal power delivered by the heat pump to the house, electrical power consumed by the heat pump when delivering heat to the house, indoor temperature and outdoor/ambient/dry bulb temperature. These data were used to calibrate the external climatic conditions, the heating profiles and indoor temperature set points for this study’s model. DesignBuilder provides weather files in .epw format for many urban and rural locations. These .epw files can be imported, exported and edited to suit the nature of the project. There was an .epw file for Belfast, which was the location of the test house. Although DesignBuilder weather files are reliable, the external conditions were simulated more accurately. The Belfast ‘epw file was exported and edited by adding the actual dry bulb temperature obtained during the field work. During the simulated period, multiple daily heating profiles were observed. Therefore, to best replicate the variability of the actual heating profiles, this study modelled a range of heating profiles for different dates during the simulated periods. Due to the limited actual data available, it was assumed that the heating system was switched off or running for a few hours on warmer days during spring, summer and autumn seasons. Figure 5 shows that hourly heating loads reduced during the summer periods.

2.3. Validation

This study’s model was validated by comparing the measured and modelled outdoor and indoor temperatures during the simulated period (see Figure 6). The HVAC was switched on to calibrate the performance of the modelled heat pump. The maximum indoor air temperature for the simulated period was 25 °C, which indicates the maximum heating temperature set by the thermostat in the actual house. The heating set point temperature in DesignBuilder models the setting of the thermostat [37]. During the calibration of the model, it was observed that a heating set point temperature of 25 °C resulted in simulated indoor air temperatures that were higher than actual indoor air temperatures on cooler days. Therefore, the heating set point temperature was set at 24 °C in DesignBuilder. Both modelled and actual data sets had the same mean and standard deviation values; however, the margin of error between the actual and simulated mean indoor temperatures was 0.8% (see

Table 4. Comparison between mean and standard deviation values of actual and simulated data for indoor and outdoor temperatures.

	Indoor Temperature		Outdoor Temperature
	Actual	Modelled	Actual	Modelled
Mean	23.0 °C	22.8 °C	6.6 °C	6.6 °C
Standard Deviation	1.0	0.7	2.7	2.7

).

The actual and simulated heat demand was compared for the simulated period of 12 December 2019 to 13 January 2020 (see Figure 7). The results showed that the simulated heat demand was less than the measured data by 25%, despite the consistencies between the actual and modelled indoor and outdoor temperatures. This can be attributed to the tendency of thermal modelling software such as DesignBuilder to simplify mechanical systems [38]. Therefore, they do not portray the mechanical system’s dynamic behaviours and part-load operations [38]. Estimating occupants’ behaviour is a significant cause of inconsistencies between actual and measured energy consumption [39]. In this study, actual occupancy data could not be obtained. Therefore, the authors created a simple occupancy schedule where the heating was consistently switched on to simulate the daily recorded indoor temperatures during the simulated period. This caused similar daily thermal loads throughout the simulated period and therefore, spikes in the recorded thermal loads could not be simulated. Consequently, the simulated thermal load was less than the measured thermal load. Nevertheless, the model generated an annual heat demand which is consistent with the heat demand associated with the modelled house type, as stated in the next section.

2.4. Fabric Retrofit Scenarios

The base-case fabric was retrofitted at two different standards—current building regulations for new dwellings in NI and Passivhaus standards (see

Table 5. Current construction standards according to the building regulations in NI, England and Passivhaus principles.

Component	Passivhaus Standards U-Value (W/m2K)	Recommended U-Value (England) (W/m2K)	Recommended U-Value (NI) (W/m2K)
Solid wall	0.1	0.26	0.21
Cavity wall	0.1	0.26	0.21
Roof	0.1	0.16	0.16
Glazing	0.8	1.60	1.60
External door	0.8	1.60	1.60
Ground floor	0.1	0.18	0.21
Air tightness (air changes per hour)	0.6	0.15	-

Source: [40,41,42,43,44].

) [40]. The optimal specifications for the retrofit measures were selected based on impact on heat demand and heat losses through the building fabric. As the assumed airtightness standard is within building regulations, this study did not investigate the impacts of airtightness retrofit on space heating demand.

Base-case annual heat demand and emissions for the terraced house

The annual heat demand and emissions for the modelled house was estimated as 14,616 kWh and 5457 kgCO₂. These values are close to typical annual demand and emissions of 13,482 kWh and 4944 kgCO₂e, respectively, for pre-1919 terraced houses [45].

2.5. Estimating the Cost-Effectiveness of the Fabric Retrofit Measures and Scenarios

In the building life cycle, cost-effectiveness can be determined by calculating the net present value indicator (NPV) [46]. The cost-effectiveness of retrofitting the fabric of the F-rated house to higher standards (building regulations and Passivhaus) was assessed using the NPV formula below:

$$\mathrm{NPV}=\sum _{\mathrm{t}=0}^{\mathrm{T}}{\displaystyle \frac{{\mathrm{C}}_{\mathrm{t}}}{{(1+\mathrm{r})}^{\mathrm{t}}}}$$

where

• C_t = cash flow in year t;
• r = discount rate;
• t = analysed year (t = 0,1,2…T);
• T = life cycle in years.

The cash flow was the product of the annual heating bill savings, current and future fuel prices, and the life cycle for each measure, all of which were considered in the calculations. The cost-effectiveness of the entire retrofit package was determined by considering a lifetime of 30 years, as the older housing stock would be existing by 2050. The National Infrastructure Commission suggests that electricity prices for heating will rise, as less energy will be used for heating due to increased energy-efficiency and heat pump use [47]. It was assumed that electricity prices will increase by 1% every year between now and 2050. The current electricity prices were derived from the Consumer Council’s electricity price comparison table [48] . As different supply companies have different rates, an average electricity price of 22.128 p/kWh was used as the current price. It was assumed that the current elevated electricity price will reduce by 2024, as wholesale prices have started falling and energy suppliers are buying gas and electricity in advance [49,50]. Therefore, a discount rate of 5% is typically used for NPV calculations regarding building projects [46]. As the performances of different insulation materials for wall and loft insulation were very similar, the NPVs of the measures with various insulation materials were estimated.

3. Simulation Results

3.1. Impacts on Annual Demand and Emissions with Fabric Retrofit to Meet Building Regulations and Passivhaus Standards

Four retrofit measures were applied: external solid wall insulation (ESWI), loft insulation, solid floor insulation (SFI) and double glazing. These measures yield the highest demand savings among all domestic insulation, glazing, airtightness and ventilation improvements [16]. Seven different types of insulation were modelled during the simulations, and they include mineral wool quilt, glass wool slab, cellulose, heavyweight expanded polystyrene (EPS), polyurethane (PUR), polyisocyanurate (PIR) and phenolic foam.

3.1.1. External Solid Wall Insulation

Internal solid wall insulation (ISWI) is recommended for pre-1919 houses, as their owners would prefer to preserve their historic appearance [51]. ISWI is typically less expensive than ESWI when installed as a DIY project and can be installed room by room or in single flats [52,53]. However, the installation of ISWI can encourage overheating, cause disruption to occupants and cause a loss of indoor space [53]. ESWI possesses more thermal and structural advantages. ESWI causes minimal disturbance to occupants and has better thermal performance compared to ISWI [53]. It can improve the building’s external appearance and presents a lower risk of moisture accumulation and condensation [52]. ESWI is well-suited to houses, as opposed to blocks of flats [52]. When ESWI was applied, all seven types of insulation, in standard thicknesses of 50 mm and 100 mm, were modelled in line with recommendations by Palmer’s et al 2017 study and the Planning Portal [54,55]. A total of 14 iterations of ESWI specifications were simulated (see Figure 8). The best thermally performing specification was 100 mm polyurethane insulation; however, 100 mm expanded polystyrene, and 100 mm glass wool insulation yielded slightly higher losses through the external walls. The 100 mm PUR insulation fostered the lowest amount of heat losses through the exterior walls. The U-value of the base-case exterior wall improved from 2.1 W/m²K to 0.2 W/m²K. The ESWI resulted in a 19% reduction in annual heat demand and 13% of the annual emissions removed. To attain the Passivhaus standard, increased insulation thicknesses of 150 mm, 200 mm, 250 mm, 300 mm and 350 mm were modelled. The 350 mm PUR insulation resulted in the lowest demand and heat losses through the external walls. The U-value of the external wall improved to 0.1 W/m²K. The base-case annual heat demand and emissions reduced by 22% and 16%, respectively.

3.1.2. Loft Insulation

The optimal thickness for loft insulation is 270 mm [56]. However, 270 mm is the recommended thickness for mineral wool insulation, while other materials can be installed with different thickness [57]. Phenolic foam is not commonly used for loft insulation [56]. All previously mentioned insulation types, except phenolic foam, were used to simulate insulation between the rafters only, joists only and both rafters and joists simultaneously. This study modelled 35 iterations of insulation thickness (120 mm, 150 mm, 180 mm, 210 mm, 240 mm and 270 mm) with different insulation types. This resulted in 107 iterations of loft insulation specifications with various thicknesses of different insulation types between joists alone, rafters alone and joists and rafters simultaneously. Incremental variations in the thickness of insulation in the loft yielded very marginal impacts on demand and heat losses through the roof (see Figure 9).

All insulation types facilitated close demand reduction and heat loss values. The greatest reduction in demand and lowest heat losses through the roof were observed with the 270 mm PIR insulation positioned between and above ceiling joists. The optimal performance of the loft insulated with 270 mm PIR insulation confirms the insulation thickness recommendations from Energy Saving Trust [57]. Furthermore, the U-value of the roof ceiling improved to 0.1 W/m²K, with this optimum specification. This U-value specification aligned with the Passivhaus specification for roofs. Loft insulation yielded a 10% reduction in demand and a removal of 7% of annual emissions. Insulation of further thicknesses: 300 mm, 350 mm and 450 mm PIR insulation was placed between the joists alone, rafters alone and joists and rafters simultaneously. The 400 mm PIR placed between ceiling joists resulted in the highest demand savings and lowest heat losses through the roof. Therefore, insulating the roof to the Passivhaus standard resulted in an 11% and 8% reduction in demand and emissions respectively. The U-value of the roof ceiling improved to 0.1 W/m²K.

3.1.3. Glazing

NI building regulations recommend double glazing for the windows in dwellings [40]. Double-glazed windows usually feature glass panes with typical thickness of 4 mm, 6 mm, 8 mm, 10 mm and 12 mm [58]. Common glass panes include clear, low emissivity (low-E), obscure, tinted/solar control and so on; each of these types possesses thermal and functional benefits [59]. The most energy-efficient glass for double glazing is low-E glass [60]. Low-E glass permits light while reducing heat loss [60]. Domestic double glazing is usually characterised by cavity thicknesses of 6 mm–16 mm; sometimes, thicknesses greater than 16 mm are used [61]. Cavity fillings include air and inert gases, such as krypton, argon and xenon; air and argon are popular choices for domestic double-glazed windows [61]. Clear glass with thickness of 4 mm, 6 mm, 10 mm and 12 mm and low-E glass with thickness of 4 mm, 6 mm and 10 mm were modelled. The air gap thicknesses for clear double-glazing options ranged from 6 mm to 30 mm, while those of low-E double-glazing options ranged from 6 mm to 20 mm. Air and argon were selected as cavity fillings. A total of 112 double glazing iterations were simulated based on the selected glass pane, cavity thickness and cavity filling options.

Table 6. Glazing specifications which were used to generate iterations of the double- and triple-glazing options modelled for this study.

Pane Type	Pane Thickness	Filling	Gap Thickness
Clear glass	4 mm	Air	6 mm, 8 mm, 10 mm
	6 mm		12 mm, 13 mm, 16 mm
	10 mm		20 mm, 25 mm, 30 mm
	12 mm	Argon	6 mm, 8 mm, 10 mm
Low-E glass	4 mm
	6 mm		12 mm, 13 mm, 16 mm, 20 mm
	10 mm

details the specifications for double glazing iterations simulated based on these parameters. There were narrow differences among the thermal properties of the different double-glazing specifications (see Figure 10). There were marginal changes to the thermal energy required to heat the house and the heat losses through the windows with each specification. The double-glazing specification, which facilitated the lowest heating demand and heat losses through the exterior windows, was the 4 mm low-E glass panes with 16 mm argon-filled cavity (which is specified as 4-16-4 double glazing). This glazing upgrade resulted in a 11% and 7% reduction in annual heat demand and the removal of annual emissions, respectively. The U-value of the external windows improved from 5.9 to 1.2 W/m²K. Triple-glazed windows typically have a U-value of 0.9 W/m²K and below, as specified by Passivhaus standards [62,63]. Triple-glazing options were modelled based on the glass pane types, glass pane and cavity thicknesses and cavity fillings used for the double-glazing options. A total of 112 triple-glazing iterations were modelled. The results showed that there were minuscule changes to heat demand and heat losses through the external windows due to the lack of disparity among the thermal properties of the triple-glazing iterations (see Figure 10). The 10 mm low-E glass panes with a 20 mm argon-filled cavity (10-20-10-20-10) resulted in the lowest heat demand and heat losses through the exterior windows. With triple glazing, the U-value of the exterior windows improved to 0.7 W/m²K. The heat demand and emissions were reduced by 10% and 6%, respectively. Contrary to existing research, the optimum double-glazing option yielded a slightly greater reduction in annual demand and emissions compared to the optimum triple-glazing option. The net solar heat gains with the triple-glazed external windows are lower than those with the double-glazed external windows [64] (see

Table 7. There were higher net solar gains through double glazing than through triple glazing.

	Triple Glazing (U-Value: 0.7 W/m2K)	Double Glazing (U-Value: 1.2 W/m2K)
Solar gains exterior windows (kWh)	1214	2001
Losses through windows/year (kWh)	105	409
Heat required to heat house/year (kWh)	8834	8750
Net solar gains (kWh)	1316	1589

). More heat was needed to balance out the slight drop in air temperature caused by the lower net solar heat gains.

3.1.4. Solid Floor Insulation

SFI is usually installed above or under the floor [65,66]. When SFI is installed above the slab, the space heats up and cools down quickly [66]. Below-slab SFI has superior structural and thermal properties, as it stores heat during the day, which helps keep the space warm at night [65]. Fitting SFI above the slab requires the adjustment of door lengths, skirting boards, electrical wiring and so on, and some headroom height is used up. Below-slab SFI is more disruptive compared to above-slab SFI, and it is recommended for major renovations. Mineral wool is not usually used for SFI, but some types of mineral wool have been engineered to be installed below the screed [67]. EPS, PIR and PUR insulation are used for below-slab SFI [66]. Insulation thicknesses typically range from 25 mm to 200 mm [68,69]. A total of 28 different iterations for SFI installed below and above the ground floor and with different insulation types and thicknesses were modelled. These 28 specifications showed marginal differences in thermal properties and facilitated similar space heating demand and heat losses (see Figure 11). Among the above-slab options, 200 mm mineral wool fostered the lowest heat losses through the floor slab and improved the U-value of the ground floor from 0.7 to 0.2 W/m²K. Among the below-slab specifications, the 200 mm EPS insulation resulted in the lowest heat demand. There were very minor differences between heat demand and losses facilitated by below-slab 200 mm EPS and those facilitated by above-slab 200 mm mineral wool.

With the 200 mm EPS insulation installed below the slab, the U-value of the ground floor improved from 0.7 to 0.1 W/m²K. This U-value indicated that the floor had been insulated to Passivhaus standards. With this solid floor insulation specification, the annual demand and emissions were reduced by 6% and 3% proportionately. The optimal SFI thickness was increased to 250 mm, 300 mm and 350 mm; however, the U-value of the ground floor remained at 0.1 W/m²K. An SFI thickness of 350 mm fostered the highest demand reductions and emission removals: a 6% reduction in annual demand and a 4% removal of annual emissions.

3.1.5. Total Reductions in Annual Demand and Emissions with Fabric Retrofit to Meet Building Regulations and Passivhaus Standards

When the base-case fabric was retrofitted to building regulations and Passivhaus standards, there was a 46% and 50% reduction in annual demand, respectively. Proportions of 30% and 34% of emissions were removed with fabric retrofit to building regulations and Passivhaus standards, respectively.

Table 8. U-values of the fabric when improved to current NI building regulations and Passivhaus standards.

Component	Passivhaus Standard U-Value (W/m2K)	NI U-Value (W/m2K)	Base-Case Fabric U-Values (W/m2K)	Optimised Fabric U-Values (W/m2K)	Passivhaus Fabric U-Values (W/m2K)
Solid wall	0.1	0.21	1.81	0.2	0.1
Roof	0.1	0.16	1.63	0.2	0.1
Glazing	0.8	1.60	5.87	1.2	0.7
External door (glazed portions)	0.8	1.60	5.87	1.2	0.7
Floor	0.1	0.21	0.70	0.1	0.1

shows improvements to thermal properties of the fabric with each standard. The CCC’s Sixth Carbon Budget (6CB)’s recommendations suggest that on average, the annual demand and emissions of a terraced house can be reduced by 11% and 5%, respectively, due to fabric retrofit to building standards [16]. In contrast, in their 2019 study, Ji, et al. indicate that the annual demand of a hard-to-treat English terrace house can be reduced by up to 50% when its fabric is upgraded to building regulations and up to 77% when it is improved to Passivhaus standards [70]. Similarly, in their 2012 study, Dowson, et al. report that insulating older buildings to more recent standards can result in a 50–80% reduction in demand [3]. This study’s results corroborate other studies [3,70] by indicating that the unique, outdated features of hard-to-treat homes give them a significant potential for demand reduction through fabric-first retrofit.

3.1.6. Potential for Reduced Demand on the Grid Due to Lower Domestic Heat Demand and Emissions

The CCC 6CB’s recommendations indicate that the most potent measures for a terraced house are ESWI and triple glazing (see

Table 9. Comparison between the reduction in demand per measure for a terraced house according to the CCC, and according to this study.

Measure	CCC Data—Typical Terraced House Fabric	Improved Fabric	Passivhaus Fabric
External solid wall insulation	12%	19%	22%
Loft insulation	4%	10%	11%
Single to double glazing	5%	11%	-
Single to triple glazing	7%	-	10%
Solid floor insulation	8%	6%	6%

Source of CCC data: [16].

) [16]. This study shows that the most potent measures to be installed in an uninsulated, pre-1919 terrace house, according to current building regulations, are ESWI and pre-2006 double glazing. For this study, insulation and glazing measures can yield greater savings than those prescribed by the CCC. This demonstrates the potential of retrofit measures to achieve substantial reductions in demand and emissions and significantly improve the energy-efficiency of E-G rated pre-1919 houses in NI. In NI, pre-1919 houses account for an estimated heating demand and emissions of 938 GWh and 0.4 MtCO₂e, respectively. These properties contribute 13% of existing domestic demand and 13% of existing emissions [45]. If all pre-1919 houses had insulation and glazing upgrades in line with building regulations, the average demand for pre-1919 houses would fall from 11,742 kWh to 6355 kWh. The average emissions for a pre-1919 house would reduce from 4753 KgCO₂e to 3317 KgCO₂e. There would be a 6% reduction in the existing domestic demand, and 4% of emissions would be removed from existing domestic emissions. There is significant potential for pre-1919 domestic fabric improvements to reduce heat demand in NI and contribute towards the region’s aim of 56% reduction in energy-related carbon emissions from the 1990 levels by 2030.

4. Cost and SAP Results

4.1. SAP Rating Improvements and Retrofit Costs

SAP improvements with fabric retrofit were only assessed when the base-case fabric was improved to building regulation standards. The current SAP methodology is outdated and does not account for improvements with the application of Passivhaus standards [71]. Figure 12 shows that fabric upgrades (insulation and glazing) alone can enhance the SAP rating of a house from band F to band D. The heat pump adds 29 points. An F-rated, hard-to-treat house can be potentially enhanced to band B with wall, loft and floor insulation, glazing and airtightness improvements and the installation of a heat pump.

The costs of implementing these fabric measures were calculated by considering retrofit material and labour cost data from the Building Cost Information Construction Data and other studies [54,57,72,73]. These costs cover material and labour costs associated with installing each retrofit measure, and they are typical to Northern Ireland. The unit cost of installation, in pounds per square meter, was estimated and applied it to the floor/wall/fenestration areas of the house. Measures such as loft and floor insulation can be applied by homeowners themselves in some cases [57,65]. A medium cost of upgrading the fabric of an F-rated hard-to-treat house to building regulation standards was GBP 12,494, including material costs and professional installation of all measures. This cost was appraised using mineral wool quilt or glass wool for loft insulation and glazing upgrades to double glazing with aluminium or aluminium-cladded timber frames. With non-professional or DIY projects for loft and floor insulation, labour costs decreased, bringing the medium cost down to GBP 10,744. A high cost of GBP 13,114 was calculated with the material costs and professional installation of all measures, the use of PIR insulation in the loft and installation of double glazing with a uPVC frame. The high retrofit cost was reduced to GBP 11,143 with DIY projects for loft and floor insulation. GBP 12,280 was the low cost for the materials and professional installation of all measures if cellulose or EPS insulation was used to insulate the loft insulation, and the glazing was improved without the window frame. With the DIY installation of loft and floor insulation, the lowest retrofit cost falls to GBP 10,551. Appendix A 

Table A1. Installation costs of measures according to current building regulations.

Costs of Measures Installed According to Current Building Regulations
Cost Level	Mode of Installation	Measure	Installation Cost	Total Installation Cost
Low	Professional	External solid wall insulation	GBP 6928	GBP 12,280
		Loft Insulation (cellulose/EPS)	GBP 705
		Double glazing (no frame)	GBP 2385
		Solid floor insulation (EPS)	GBP 2263
	DIY	External solid wall insulation *	GBP 6928	GBP 10,551
		Loft Insulation (cellulose/EPS)	GBP 452
		Double glazing (no frame) *	GBP 2385
		Solid floor insulation	GBP 787
Medium	Professional	External solid wall insulation	GBP 6928	GBP 12,494
		Loft Insulation (mineral wool/glass wool)	GBP 761
		Double glazing (aluminium/aluminium-cladded timber frame)	GBP 2542
		Solid floor insulation (EPS)	GBP 2263
	DIY	External solid wall insulation *	GBP 6928	GBP 10,744
		Loft Insulation (mineral wool/glass wool)	GBP 487
		Double glazing (aluminium/aluminium-cladded timber frame) *	GBP 2542
		Solid floor insulation (EPS)	GBP 787
High	Professional	External solid wall insulation	GBP 6928	GBP 13,114
		Loft Insulation (PIR/PUR)	GBP 1377
		Double glazing (uPVC frame)	GBP 2547
		Solid floor insulation	GBP 2263
	DIY	External solid wall insulation *	GBP 6928	GBP 11,143
		Loft Insulation (mineral wool/glass wool)	GBP 881
		Double glazing (uPVC frame) *	GBP 2547
		Solid floor insulation (EPS)	GBP 787

* These measures are typically recommended for professional installation.

shows a breakdown of the installation costs. The Carbon Trust suggests that the installation of energy-efficiency measures (without deep retrofit measures) in pre-1919 properties cost, on average, GBP 10,783 [19]. Nevertheless, cost variations will depend on the unique energy-efficiency requirements of properties. This study’s fabric retrofit costs are slightly higher than the Carbon Trust’s recommendations, considering retrofit to building regulations standards and professional installation of loft and floor insulation.

With Passivhaus retrofit, the medium retrofit cost rose to GBP 43,461 with professional installation, the use of mineral wool quilt or glass wool insulation in the loft and the fitting of triple-glazed windows with uPVc, aluminium or timber framing. The cost decreased to GBP 36,201 with DIY installation for loft and floor insulation. The high cost for Passivhaus retrofit was calculated as GBP 44,087 with professional installations, the use of PIR insulation for the loft and the installation of triple-glazed windows with aluminium-cladded timber framing. With DIY installations for loft and floor insulation, this cost was reduced to GBP 36,605. The lowest cost with Passivhaus retrofit was estimated as GBP 43,296 due to the installation of triple glazing without replacing existing window frames, the use of cellulose/EPS insulation in the loft and professional labour costs. Double-glazing panes can be replaced with triple-glazing frames if the latter are thin enough [74]. The lowest cost was reduced to GBP 36,056 with DIY installations of loft and floor insulations. Appendix A 

Table A2. Installation costs of measures installed according to Passivhaus standards.

Costs of Measures Installed According to Passivhaus Standards
Cost Level	Mode of Installation	Measure	Installation Cost	Total Installation Cost
Low	Professional	External solid wall insulation	GBP 25,739	GBP 43,296
		Loft Insulation (cellulose/EPS)	GBP 1064
		Triple glazing (no frame)	GBP 7583
		Solid floor insulation (EPS)	GBP 8910
	DIY	External solid wall insulation *	GBP 25,739	GBP 36,056
		Loft Insulation (cellulose/EPS)	GBP 681
		Triple glazing (no frame) *	GBP 7583
		Solid floor insulation	GBP 2053
Medium	Professional	External solid wall insulation	GBP 25,739	GBP 43,461
		Loft Insulation (mineral wool/glass wool)	GBP 1120
		Triple glazing (uPVC/aluminium/timber frame)	GBP 7692
		Solid floor insulation (EPS)	GBP 8910
	DIY	External solid wall insulation *	GBP 25,739	GBP 36,201
		Loft Insulation (mineral wool/glass wool)	GBP 717
		Triple glazing (uPVC/aluminium/timber frame) *	GBP 7692
		Solid floor insulation (EPS)	GBP 2053
High	Professional	External solid wall insulation	GBP 25,739	GBP 44,087
		Loft Insulation (PIR/PUR)	GBP 1736
		Triple glazing (aluminium-cladded timber frame)	GBP 7702
		Solid floor insulation (EPS)	GBP 8910
	DIY	External solid wall insulation *	GBP 25,739	GBP 36,605
		Loft Insulation (mineral wool/glass wool)	GBP 1111
		Triple glazing (aluminium-cladded timber frame) *	GBP 7702
		Solid floor insulation (EPS)	GBP 2053

* These measures are typically recommended for professional installation.

shows a breakdown of these costs. The Passivhaus Trust associates Passivhaus retrofit with deep retrofit, which typically reduces space heat demand by 75% or more [75]. Deep retrofit packages typically feature insulation and glazing improvements, as well as passive ventilation (heat recovery and airtightness) upgrades [75]. A deep retrofit package has an average cost of GBP 48,190 [19]. This study’s cost estimates are about 10% lower than this average, as the estimated costs cover only insulation and glazing upgrades. Figure 13 shows a comparison between installation costs of retrofit, in line with building regulations and PassivHaus standards, with the accompanying reductions in demand and emissions for each measure.

4.2. Potential Consumer Energy Cost Savings

Appendix A 

Table A3. Annual heating bills with different fabrics.

Electricity Supplier	Base-Case Fabric	Optimised Fabric	Passivhaus Fabric	% Reduction in Bills with Optimised Fabric	% Reduction in Bills with Passivhaus Fabric
Power NI	GBP 3868	GBP 2093	GBP 1947	46%	50%
SSE Airtricity	GBP 2426	GBP 1313	GBP 1222
Budget Energy	GBP 4816	GBP 2607	GBP 2425
Electric Ireland	GBP 2533	GBP 1371	GBP 1275
Click Energy	GBP 2528	GBP 1368	GBP 1273

Source of electricity prices: [48] .

shows annual heating bills with different levels of fabric retrofit. These figures were calculated using the electricity prices per unit presented by the Consumer Council’s Electricity Price Comparison Table from 4th November 2022 [48]. If homeowners of a house with an F-rated SAP rating improved the fabric to building regulation standards, they would achieve a 46% reduction in annual heating bills. If the fabric is insulated and the glazing is upgraded to Passivhaus standards, they can acquire a 50% reduction in annual heating bills.

A household is fuel poor if it spends 10% or more of its annual income on heating fuel [8]. Appendix A 

Table A4. Fuel-poverty levels according to income bands and percentage of household income spent on heating with the base-case fabric.

Band	Annual Household Income	% in Fuel Poverty	Lower Limit of Household Income	% of Income Spent on Heating with Electricity with the Base-Case Fabric	Fuel Poverty Status of Household with Base-Case Fabric
1	<GBP 10,399	55%	GBP 10,399	31%	Fuel poor
2	GBP 10,400–GBP 15,599	33%	GBP 10,400	31%
3	GBP 15,600–GBP 20,799	23%	GBP 15,600	21%
4	GBP 20,800–GBP 31,199	7%	GBP 20,800	16%
5	GBP 31,200–GBP 46,799	1%	GBP 31,200	10%
6	>GBP 46,800	<1%	GBP 46,800	7%	Not fuel poor

Source of fuel poverty income bands: [8]

shows the categories of annual household income with their corresponding levels of fuel poverty. For this study, these categories have been labelled band 1–6, with bands 1 and 6 indicating the highest (55%) and lowest (less than 1%) percentage of houses in fuel poverty, respectively. Households with incomes in bands 1–5 would experience fuel poverty if they lived in an F-rated house with uninsulated fabric and single-glazed windows (see Appendix A Table A4). Households in bands 1–2 would spend 31% of their income on heating fuel if they lived in a house with the base-case fabric. Households in bands 3, 4 and 5 would spend 21%, 16% and 10% of their income. If households in bands 1 and 2 occupied the house with the optimised fabric, they would spend 17% of their income on heating. Households in band 3 would spend 11% of their income on heating. These households would still experience fuel poverty (see Appendix A 

Table A5. Fuel poverty levels according to income bands and percentage of household income spent on heating with the optimised and Passivhaus fabrics.

Band	Annual Household Income	Lower Limit of Household Income	% of Income Spent on Heating with Electricity with the Optimised Fabric	Fuel Poverty Status of Household with Optimised Fabric	% of Income Spent on Heating with Electricity with the Passivhaus Fabric	Fuel Poverty Status of Household with Passivhaus Fabric
1	<GBP 10,399	GBP 10,399	17%	Fuel poor	16%	Fuel poor
2	GBP 10,400–GBP 15,599	GBP 10,400	17%		16%
3	GBP 15,600–GBP 20,799	GBP 15,600	11%		10%
4	GBP 20,800–GBP 31,199	GBP 20,800	8%	Not fuel poor	8%	Not fuel poor
5	GBP 31,200–GBP 46,799	GBP 31,200	6%		5%
6	>GBP 46,800	GBP 46,800	4%		3%

). Households in bands 4 and 5 would spend 8% and 6% on heating, respectively. These households would no longer be fuel poor due to fabric upgrades, in line with building regulations. Households in bands 1–3 would still be in fuel poverty when living in the house with the Passivhaus fabric (see Appendix A Table A5). Households in bands 1 and 2 would spend 16% of their income on heating, while households in band 3 would spend 10% of their income on heating. Households in bands 4 and 5 would not experience fuel poverty, as they would spend 8% and 6% of their income on heating, respectively. Only households with an annual income of at least GBP 46,800 can live in the house with all fabrics and remain outside of fuel poverty. Additionally, only households with annual incomes of minimum GBP 20,800 can live in the house with the improved fabrics and remain outside of fuel poverty (see Appendix A Table A5).

4.3. Cost-Effectiveness of Retrofit Measures

Table 10. Lifetime of recommended measures.

Component	Lifetime (Years)
Solid wall insulation	36
Loft insulation	42
Glazing	20
Solid floor insulation	42

Source: [76].

shows the lifetime for each recommended measure. Figure 14 shows a comparison between installation cost and NPVs per measure, according to both retrofit standards.

4.4. External Solid Wall Insulation

ESWI was found to be cost-effective when installed in an uninsulated, hard-to-treat fabric to building regulation standards. With an NPV of GBP 5768, this measure was determined cost-effective regardless of the insulation type. The NPV of the ESWI installed to Passivhaus standards, regardless of insulation type, was negative (-GBP 11,331). The savings due to Passivhaus ESWI were only 3% more than the savings due to building regulations for ESWI. The cost of installing ESWI to building regulation standards was 73% less than installing ESWI to Passivhaus standards. ESWI installed to building regulation standards was the most cost-effective wall retrofit solution.

4.5. Loft Insulation

Loft insulation installed to building regulation standards was found to be cost-effective. The use of different material and installation methods determined the actual values of the NPVs. The NPVs of DIY loft insulation were higher than those of professionally installed loft insulation. With professional installation, the NPV of loft insulation was GBP 5697, GBP 6313 and GBP 6369 with PIR/PUR, mineral wool/glass wool and cellulose/EPS insulation, respectively. This indicates that insulating the loft with cellulose/EPS insulation is the most cost-effective solution when loft insulation is performed professionally. With DIY installation, the NPVs of loft insulation were GBP 6193, GBP 6587 and GBP 6623 for PUR/PIR, mineral wool/glass wool insulation and cellulose/EPS insulation. When the loft is DIY insulated, cellulose/EPS insulation is the most cost-effective material. Overall, DIY loft insulation with cellulose/EPS insulation is the most cost-effective solution when insulating the loft to building regulation standards. Similarly, insulating the loft to Passivhaus standards was generally estimated as cost-effective, with DIY installation being more cost-effective than professional installations. Variations in NPVs were due to the mode of installation and material choices. With professional installation, the NPVs of loft insulation were GBP 6067, GBP 6683 and GBP 6739 for PIR/PUR, mineral wool/glass wool and cellulose/EPS insulation, respectively. The NPVs of DIY installation were GBP 6692, GBP 7086 and GBP 7122 for PIR/PUR, mineral wool/glass wool and cellulose/EPS insulation respectively. The DIY installation of cellulose/EPS insulation to Passivhaus standards was the most cost-effective solution for loft insulation.

4.6. Glazing

NPV estimations showed that double-glazing retrofit was cost-effective, with only slight differences due to the type of frame. The NPVs of double-glazing options were GBP 2287, GBP 2295, GBP 2300, GBP 2311 with an aluminium-cladded timber frame, a uPVC frame, an aluminium frame, and a wooden frame. The NPV of installing double glazing without a frame was GBP 2457. Installing post-2006 double glazing with a timber frame is the most cost-effective option. Triple glazing was not cost-effective. The NPVs of triple glazing retrofit options were - GBP 3106, - GBP 3114, - GBP 3117, and - GBP 3122 for a timber frame, an aluminium frame, a uPVC frame and an aluminium-cladded timber frame, respectively. Installing triple glazing without the frame was not cost-effective with an NPV of - GBP 3004. Although triple glazing retrofit was generally determined to be not cost-effective, installing triple glazing without a frame was the least cost ineffective option.

4.7. Solid Floor Insulation

SFI was determined to be cost-effective with DIY installation. The DIY installation of SFI to building regulation standards was cost-effective, with an NPV of GBP 3240. The DIY installation of SFI to Passivhaus standards was cost-effective, with an NPV of GBP 2453. Similarly, the professional installation of SFI to building regulation standards was estimated to be cost effective, with an NPV of GBP 1764. However, the professional installation of SFI to Passivhaus standards was not cost-effective, with an NPV of -GBP 4404. The demand savings of SFI installed to Passivhaus standards were only 11% more than those of the SFI installed according to building regulations. The cost of installing SFI to Passivhaus standards was 75% higher than the cost of installing SFI according to building regulations.

4.8. Cost-Effectiveness of Installing All Retrofit Measures

The lifetime of glazing is 20 years [76]. Consequently, glazing units are recommended for replacement before 2050. This consideration increased the total retrofit cost from GBP 12,496 to GBP 15,040. The installation of wall, loft and solid floor insulation, as well as double glazing upgrades according to building standards, was determined to be cost-effective, with an NPV of GBP 12,101. Installing all measures to Passivhaus standards was not found to be cost-effective, with an NPV of - GBP 21,780. The cost of replacing triple glazed windows after 20 years was considered, and this increased the retrofit cost from GBP 43,461 to GBP 51,152. The demand savings from all measures installed according to Passivhaus standards was 8% more than demand savings from measures installed according to building regulations. Still, the total cost of retrofit to Passivhaus standards was 71% higher than the total cost of retrofit according to building regulations.

5. Limitations of Study

This study relies on thermal modelling methods, which do not accurately simulate real-life situations [77]. This limitation was reduced by validating the model through a comparison of its outputs with real-life data recorded via long-term monitoring [78]. Nevertheless, this limitation indicates the need for real-life pilot projects, which investigate the impacts of fabric retrofit on hard-to-treat residences, in preparation for heat pump use.

6. Conclusions

6.1. Economic Recommendations

It is cost-effective to improve wall, loft and solid floor insulation, as well as glazing, in line with building regulations. Dowson et al’s 2012 study suggests that improving solid walls and glazing in old, hard-to-treat dwellings is not cost-effective [3]. However, this study determined that installing individual measures of wall insulation, loft insulation, solid floor insulation and post-2006 double glazing in a pre-1919 house is cost-effective. Solid wall insulation is typically regarded as expensive [79]. This study demonstrates that energy savings due to external solid wall insulation make it a cost-effective measure. In addition, solid wall insulation and loft insulation were the most cost-effective measures. House owners would benefit from improving insulation and glazing in their homes. These measures would be functional for 20–42 years (as the improved glazing would need replacement after 20 years). Retrofit measures in pre-1919 properties can cost between GBP 30,000 and GBP 131,000 [11]. Homeowners find these costs too high, especially because they may not live in such homes long enough to experience a return on their investment. However, this study has shown that homeowners can improve the SAP rating of a house from band F to band D with insulation and glazing upgrades, with an approximate maximum cost of GBP 15,000. If such a retrofit program is be organised centrally, it might be possible to implement it under the medium range cost estimations of GBP 15,000, achieving significant cost savings. Homeowners would get a 46% reduction in annual heating bills. Fuel-poor households, with an annual income of GBP 20,800 or more, would no longer be fuel poor after implementing insulation and glazing improvements on an uninsulated fabric, in line with building regulations.

Improving insulation and glazing to Passivhaus standards is not cost-effective. The installation costs of all measures were too high in relation to the financial savings they would generate over 30 years. However, insulating the loft to Passivhaus standards was the most cost-effective loft insulation solution. Homeowners may consider insulating the walls and floors of their homes according to building regulations and insulate the lofts to Passivhaus standards. This would cost approximately GBP 13,600, which is just around GBP 1000 more than the median cost (circa GBP 12,500) of adding insulation and glazing upgrades, in line with building regulations. Passivhaus retrofit can be cost-effective with the installation of fabric retrofit measures, a mechanical ventilation with heat recovery (MVHR) system, solar thermal panels and a gas boiler for the production of hot water [72]. If homeowners intend to improve glazing and insulation, it will be cost-effective for them to do so according to building regulations.

Insulation and glazing improvements to Passivhaus standards would generate a 50% reduction in annual heating bills. This is only 4% less than the reduction in bills that can be achieved by upgrading insulation and glazing improvements according to building regulations. It seems logical to upgrade insulation and glazing in line with building regulations, as this is cost effective and almost halves annual heating bills. As with the upgrades in line with building regulations, only fuel-poor households with an income of GBP 20,800 or more would be able to escape fuel poverty with fabric improvements to Passivhaus standards.

6.2. Other Recommendations

6.2.1. Social, Policy and Planning Barriers

Pre-1919 houses have been termed ‘hard-to-treat’ because people are generally not open to installing the required retrofit measures, as they find them costly and/or disruptive [3,5]. A standard feature of hard-to-treat homes is the external solid wall construction, and it can be laborious to implement solid wall insulation [12]. External solid wall insulation is regarded as one of the more challenging measures to implement, primarily because it is requires capital investment [12]. This study shows that ESWI is a cost-effective measure, yielding the highest reductions in annual demand when installed in an uninsulated pre-1919 terrace house. The cost barrier can be mitigated if the government provides grants and subsidies that can encourage people to invest in adding this measure to their homes if required [11]. ESWI is a retrofit measure that can improve thermal comfort, as well as exterior façade appearance, and reduce heating bills [47]. Policy reforms which promote education on the economic and environmental impacts of this measure should be undertaken [20,80]. Installers can provide information on potential demand and energy bill reductions due to ESWI, which homeowners can access and consider [80].

A second barrier to the uptake of ESWI, is obtaining planning permission from the local authorities for certain houses in conservation areas [53,81]. Homeowners should factor in the time for planning applications to be approved. Prior research into the heritage of a property should be carried out and considered to determine if applications for Listed Building Consent are required [12]. Owners of properties in conservations areas must obtain planning permission before carrying out ESWI [53]. The government should revisit regulations and create avenues for building specialists to recommend retrofit measures, which can improve energy-efficiency while conserving the architectural character [11]. Additionally, ESWI can enhance the appearance of certain heritage properties [81].

Another barrier is the possibility of poor installation, which can lead to health concerns, such as moisture buildup and overheating [11]. This barrier is compounded if installers’ skills are not current [80]. Policy is essential in promoting education among installers and should provide support for skills certification and product standards [18]. More well-trained installers would perform more efficient installations, thereby encouraging more consumers to embrace domestic retrofit solutions. However, companies find that employees go through training and use their skills elsewhere [80]. This barrier can be overcome if the government encourages more opportunities to demonstrate experience, acquire accreditation and achieve higher levels of formal skill distribution [80]. A lack of awareness of the order of installation of measures can limit the benefits of fabric retrofit measures. If glazing is installed with ESWI, the windows should be retrofitted first to reduce the risk of damaging newly installed insulation [81]. If insulation and a new heating system are to be fitted, insulation should be installed first. This would allow for adequate heating system sizing after the assessment of the new space heating requirements due to the insulation. Sufficient consumer and installer education can provide the necessary information on the potency and installation order of measures.

6.2.2. Flammability and Durability of Insulation Materials

Retrofit measures are installed to improve the indoor comfort and, therefore, the health and wellbeing of occupants [82]. The fire load of modern materials should be considered [83]. The common insulation materials are ranked in order of increasing flammability as follows: glass wool (least flammable), stone/mineral wool, phenolic foam, polyisocyanurate (PIR) and polyurethane (PUR) foam and polystyrene (most flammable) [83].

Table 11. Flammability of insulation materials.

Insulation	Reaction to Fire by Euroclass Range
Glass wool	A1–A2
Stone wool/Mineral wool	A1–A2
Phenolic	B–C
Polyisocyanurate	C–D
Polyurethane	D–E
Expanded polystyrene	E–F
Extruded polystyrene	E–F

Source: [83].

shows the classification of the flammability of insulation types according to the Euroclass range, with A1 and F being the least and most flammable classes, respectively. This study shows that 100 mm of PUR insulation improved the thermal performance of the exterior walls more than the other insulation types. The 270 mm PIR insulation also improved the thermal performance of the roof more than other insulation materials. However, these are the most flammable insulation materials. The results showed that 100 mm of glass wool and 100 mm of EPS generated only slightly higher losses through the external walls. The results showed that the 270 mm mineral wool quilt (stone wool) insulation generated marginally higher heat losses through the roof than PUR insulation. This indicates that glass wool and mineral wool can be used as substitutes for wall and loft insulation. However, glass wool is the least flammable and safest insulation material.

The durability of insulation materials is affected by their resistance to mould growth [84]. Some insulation types can also aggravate the corrosion of wires, pipes and fasteners [84]. Glass wool does not foster corrosion; however, cellulose is treated with chemicals, which can catalyse corrosion in certain conditions. Mineral wool, glass wool and foam (phenolic, EPS, PIR, PUR) insulation materials are inorganic and, therefore, do not encourage mould growth. Cellulose promotes mould growth, as it is made of organic material. The recommended substitutes present a win–win solution, as they have low to medium flammability and are also resistant to mould growth. None of the recommended insulation materials would promote corrosion of other building fittings.

6.2.3. Improved Indoor Thermal Comfort

In December 2021, the Northern Ireland Statistics and Research Agency (NISRA) reported that the number of deaths in the winter period December–21 March 2020 had increased by 1120 from the previous winter period [85]. NEA Northern Ireland states that over 500 of such deaths can be directly linked to living in a cold, damp home. Warm and comfortable indoor conditions can reduce the occurrence of respiratory diseases, cancer and death, especially among fuel-poor people [86]. Well-insulated interiors can reduce moisture buildup and damp [11]. This study shows that insulation and glazing upgrades reduce heat demand, which indicates the capacity of the fabric to store heat and promote thermal comfort.

6.2.4. Fuel-Poverty Alleviation

Fabric retrofit can reduce heat demand and, consequently, heating bills [79,87]. Lower heating bills can help fuel-poor households spend less than 10% of their annual income on heating, thereby relieving them of fuel poverty [88]. Solid wall insulation is likely to alleviate fuel poverty and related health impacts [89]. This suggests the important role of fabric improvements in mitigating fuel poverty. This study shows that households who earn less than GBP 46,800 and live in an uninsulated, pre-1919 terrace house will experience fuel poverty. However, with insulation and glazing upgrades in line with building regulations, households that earn GBP 20,800 or more can come escape fuel poverty. Although this indicates the capacity of these measures to lessen the distribution of fuel poverty, houses that earn less than GBP 20,800 will still experience fuel poverty.

This study used a Northern Irish pre-1919 uninsulated terraced house with a SAP rating of F to assess the potential of fabric-first retrofit to prepare a hard-to-treat, older house for heat pump use. The above recommendations can be considered in similar contexts.