1. Introduction
The price of housing has been steadily increasing in New Zealand, a trajectory sometimes punctuated by steep increases. This has translated to higher mortgage repayments and higher rental costs for tenants. Across the last 15 years, the average weekly rental has risen by 84.3%, whilst average household disposable income only increased by 67.6% [
1]. Low income households frequently must respond to rising housing costs by increasing the occupancy rate of the house: taking in more paying tenants, moving to a smaller house with fewer bedrooms, or moving in with family to share the housing cost. This leads to crowding where the number of occupants of a house exceeds the recommended number according to occupancy standards.
In New Zealand, the indigenous people (Māori) and those with Pacific Island heritage (Pacifica), are among those most severely impacted by crowding [
2]. In the Auckland region, the Manukau area is one of the higher deprivation areas, as well as having a higher percentage of Māori and Pacific peoples: 17.1% and 25.3%, respectively [
3].
Crowding enables airborne respiratory diseases to spread more efficiently due to the greater number and closer proximity of occupants. This effect intensifies when families must sleep in one room to save on heating costs. The effect is most pronounced for infants and young children, leading to a greater incidence of respiratory infections and hospital admissions for this group [
4,
5].
In addition, New Zealand is a country with high relative humidity (dampness), and a relatively old housing stock with a high proportion constructed of wood [
6]. Centralised HVAC systems are rare, with after-market solutions quite common, such as the installation of single-room split unit heating and cooling systems (heat pumps) and positive air pressure ventilation systems. These devices and systems are costly to install; however, heat pumps are increasingly being installed by landlords, in order to comply with the New Zealand Healthy Homes 2019 Regulations [
7]. Other forms of space heating involve less expensive upfront costs, but higher running costs, meaning that such devices remain switched off in order to save power. Therefore, many homes exhibit poor indoor environmental quality (IEQ): low thermal comfort (cold), high relative humidity (damp), airborne pathogens, and other indoor air quality (IAQ) problems, including harmful particulate matters caused by mould spores. These homes increase the risk to occupants of respiratory conditions and preventable illnesses [
7].
Poor thermal comfort is a particular problem in New Zealand houses. In their 2021 study of 2000 children in New Zealand, Morton et al. (2021) found that 50% of children routinely slept in bedrooms that were too cold [
8]. Approximately 1000 of the children in their surveys experienced ambient temperatures that fell below 19
overnight, with some children’s bedrooms registering temperatures below 4
overnight [
8].
Engineering studies [
9,
10,
11,
12,
13,
14,
15] enable numerical and computational analysis of thermal comfort to propose potential solutions to address the issues described above. In particular the computational fluid dynamics (CFD) method can inform of thermal comfort issues, and can be deployed in a number of software packages. Pérez et al. [
9] used CFD in Star (Computational Continuum Mechanics (CCM+)) to investigate the effect of fresh inlet air on IEQ (including thermal comfort) in a multipurpose room. In Wahhad et al. [
11], CFD analysis was conducted in ANSYS-Fluent to examine the disparity between outdoor and indoor temperatures, as well as the distribution of temperatures within the indoor space. The results of their CFD analysis align with the experimental data, having an 8.23% maximum accuracy error, using k-ε and Reynolds stress models. Raczkowski et al. [
12] used the standard k–ε model in Autodesk Inventor to study the temperature profiles when mixing indoor with outdoor air during heating, in terms of the Predicted Mean Vote (PMV) and Predicted Percentage Dissatisfied (PPD) thermal comfort parameters. They found that the PMV and PPD confirmed their measured data, reflecting that CFD can be used to accurately predict performance of thermal parameters. Al-Rawi et al. [
13] conducted CFD analysis in ANSYS Fluent to investigate the behavior of air velocity contours in a residential home office when one occupant coughs. They also examined thermal comfort parameters including temperature, PMV and PPD at different times, and validated the CFD results with data taken using the Testo 400 IAQ and comfort kit (Testo SE & Co. KGaA, Lenzkirch, Germany). These data closely agreed with the results of the CFD model, further illustrating the usefulness of CFD for predicting these thermal comfort parameters.
In this paper we present a survey of the case study house, including house dimensions, occupancy, readings of air temperature, air velocity, relative humidity, moisture content and thermal imaging for the living areas of the house. Then, a model of the house was designed and analysed in terms of air temperature and air velocity. We then outline a potential Heating, Ventilation and Air- Conditioning (HVAC) system, a ducted heat pump system, on a simulated model of the house, assessing the system’s potential impact on this case-study house, in terms of temperature, air velocities, relative humidity and thermal comfort parameters, using ANSYS -CFX 2021 R1 (Ansys, Inc., Canonsburg, PA, USA). This study reflects analysis of IEQ using CFD methods and a proposed solution for a house in realistic sub-optimal living conditions in New Zealand where houses are frequently too cold, have high relative humidity and may have high occupancy in some or all rooms. To the best of our knowledge, this is the first such study of a home with poor IEQ and high occupancy for which this analysis has been performed.
2. Materials and Methods
This study selects a single house from a collection of 24 home surveys that were conducted for the winter months (June to August in New Zealand), in the Manukau area of Auckland, New Zealand. The case study was selected because the house exhibited poor IEQ and, due to the distribution of occupants in each room, one room in the house would be described as crowded, according to the Canadian National Occupancy Standards [
16]. Ethical approval (E15/FET/14) was obtained for this survey, and collection of data for the house was done in accordance with ethical protocols, with data stored in accordance with the New Zealand Privacy Act.
This house contained four adults and one infant aged nine months old. Due to high electricity bills, the occupants revealed they were not running the heaters in the house.
Table 1 below presents the IEQ measures taken for the case study house, along with the specifications of the instrument used to collect each measure.
Table 2 presents the IEQ measures for the case study house and
Figure 1 presents the thermal images for areas the house. The air velocity reading was zero for all rooms, therefore this is not included in the table. As shown in
Table 2, the ambient temperature in all rooms was slightly cold, and below the minimum of 19.4
recommended by the American Society of Heating, Refrigeration and Airconditioning Engineers (ASHRAE) [
17].
The thermal images, which were obtained as part of the home survey, provide richer information on temperature, as they show the high and low points of temperature in a given area of the room (e.g., at the corner). They help to explain why some rooms can feel colder than the measured temperature, where streams of cooler air enter the room through uninsulated portions of the wall or ceiling. The temperature measured at each area of the room shown in
Figure 1 was, on average, lower than the ambient temperature in the room, and the scale (at the bottom of each panel) shows the coolest part of the area in bedrooms 1 and 3 reached as low as 8
.
The house did exhibit dampness, with the indoor relative humidity measure exceeding the relative humidity outside. As
Table 2 illustrates, all measures for relative humidity exceeded 70%, with the average relative humidity for all rooms being 73.68%; this is higher than the 30–50% recommended level for relative humidity [
18]. Therefore, the measurement obtained from this house shows the potential for increased mould and fungi growth on the building materials [
19].
The thermal images were taken using the FLIR thermal imaging camera to determine the temperatures on those surfaces.
Figure 1 shows the top corner surface images for the walls of each room in the house. As described above, the surface wall temperature in bedrooms 1 and 3 (
Figure 1a,c) were colder compared to bedroom 2. As shown in
Table 2, bedrooms 1 and 3 each had one occupant whilst bedroom 2 had two adults and one infant (see
Figure 1b). In bedroom 1 the average surface temperatures for each wall ranged from 12.3
to 14.8
, with the coldest surface area reading 8
and the warmest 19
, while bedroom 3 had an average temperature range between 12.3
and 16.3
, with the coldest surface area reading 8
and the warmest 21
. Bedroom 2′s average temperature for each wall measured between 15.1
and 16.6
with a smaller range: the coldest surface area reading 11
and the warmest was 20
. This greater warmth could be due to the number of occupants in the room.
Figure 1d,e shows the surface temperature for the living room and kitchen. Respectively, these ranged from 13.2
to 15.7
and 13.9
to 15.2
.
Non-invasive (pin-less) moisture meter readings were also taken for all four walls using the PROTIMETER MMS2, General Electric Company, Boston, MA, USA.
Table 2 displays these readings. The Wood Moisture Equivalent (WME) value can be seen in
Table 2. For this house, the WME values range from 110 to 131, which are in the dry range [
20]. This indicated that the house did not have any water ingress issues or structural defects that affected weather-tightness of the building envelope.
4. Results and Discussion
The results, shown in
Figure 5, indicate that the velocity ranges from 0 m/s to 0.76 m/s for the fluid body as shown at three different planes: the finished floor level; 1 m above the finished floor level; and 1.5 m above the finished floor level. The velocity contours show a good distribution of air flow going to the bedrooms, with increased velocity at the centre of the house, venting outside (exhausting) via the bathroom. The outcome of the velocity contours and vectors encourage us to investigate the temperature distribution at the same planes.
Figure 6 shows the temperature distribution with the duct temperature set to 25 °C (298 K). The results show the temperature ranged between 21 °C and 25 °C (285 K and 298 K). The lower temperature represents the wall temperature, which is to be expected; this was also observed in the thermal images, but the variation was to a much greater extent, as described earlier. The flow rate was designed to improve each room’s thermal comfort, with particular attention given to improving bedroom 2, containing the infant and two adults. The ambient temperatures for the three bedrooms were: 20.82
in bedroom 1; 24.64
in bedroom 2; and 24.21
in bedroom 3. The living room’s average temperature was 20.87
and the kitchen was 22.03
. Therefore, our proposed system increased the ambient temperature in all rooms, and set them within the range recommended by ASHRAE standard 55-2017 [
17], with an average temperature for the house of 21
, as illustrated in
Figure 6.
If, alternatively, the inlet temperature was set to 20
, consistent with the New Zealand Healthy Homes Standards 2019, then the system performs worse in terms of overall temperature. However, the Healthy Homes Standards 2019 [
7] require that the main living area be heated to, or above, 18
[
23]. As shown in
Figure 7, the main living room would meet the standard, but the other bedrooms would be lower in temperature.
Figure 8 shows the relative humidity profile in the house, at the heights: 0.1 m, 1 m and 1.5 m from the finished floor level. The results show a huge improvement to this IEQ parameter for each room in the house. The average relative humidity for all rooms in the house fell below 50% (whereas, in the house with no intervention, these readings were above 70%). The bathroom had the highest relative humidity (around 48%) reflecting that it is the ventilation room, and assuming that the window is kept open.
The results for the proposed system can be further assessed using the PMV (predicted mean vote) and PPD (predicted percentage dissatisfied) thermal comfort parameters [
24,
25]. The PMV is an index used to predict how occupants of an indoor space will vote in terms of their thermal sensation. The PMV scale ranges from
(cold) to
(hot) and can be embedded into ANSYS-CFX [
10,
13] using the user CEL (in the C programming language) function with the PMV calculation described in Equation (6) below [
13].
where
M is the metabolic value, equal to 60 W/m
2, or 1.0 met, reflective of non-exertion in activities. Aspects of the body are incorporated, such as
for weight of the person,
to reflect the body’s surface area ratio,
to reflect the clothing’s surface temperature,
is the thermal resistance of the clothing (m
2k/W), and
reflects the thermal insulation afforded by the clothing (set to 0.1555
). The coefficient of convective heat transfer is
, air temperature is
(°C) and the mean radiant temperature is
(°C). Relative air velocity is given by
(m/s), and
is the water vapour’s partial pressure (as per the saturation curve).
Figure 9 presents our results for the PMV in the house, where the scale on the side shows that all values for PMV were between
and
, indicating very good performance in relation to thermal comfort. The highest PMV score was
(in the kitchen) and the lowest was
, occurring at the windows and bathroom (venting area) for the house based on the input of 25
and relative humidity of 45%. Importantly, the bedrooms were in the slightly warm range, which is preferable, particularly for young children.
Figure 10 shows the PMV when heating the inlet air to 20
and relative humidity 45%. The living room performs acceptably (shown by the lighter blue areas), but the bedrooms are colder, indicating more occupants feeling the temperature was colder than they might prefer.
The PPD indicates what percentage of the occupants would not be satisfied with the thermal comfort in the given space. It never reaches zero, as differences in preferred levels of heat and metabolic rate mean the same temperature in the room will not perfectly satisfy all occupants. However, a higher PPD means the room is deemed too hot or too cold for a larger proportion of its occupants. ASHRAE Standard 55-2017 considers a PPD of under 10% reflects an acceptable level of thermal comfort.
Figure 11 shows the PPD contours for the house assessed based on both standards EN ISO 7730 and ASHRAE 55-2017 using Equation (7).
The CFD results show that when the inlet temperature is heated to 25, the PPD for the house of between 5% and 9%, which is within the acceptable range, given the current temperature and relative humidity. Therefore, the proposed system shows significant improvement of the indoor thermal conditions for the occupants.
When using the lower inlet temperature of 20
, consistent with the Healthy Homes Standards 2019, the PPD values are much higher, as evidenced by
Figure 12, where the scale shows much higher values of PPD, with an average PPD of 25%. This means a significant proportion of occupants may find the temperature too cold in the house. Consequently, although the living room can reach the minimum value of 18
, as set out by the Healthy Homes Standards 2019, the overall temperatures in the house’s living spaces may result in less satisfaction with thermal comfort for the household’s occupants.
Figure 13 shows the comparison between the thermal comfort performance of the two inlet temperature sets (25
, and 20
). The optimal location is the area outlined in green in
Figure 12, with a PMV of
, and a PPD of 10% or less. Running the system with the inlet heat of 25
and relative humidity 45% locates comfortably within this green region of thermal comfort, whereas using a lower inlet temperature of 20
leaves the house outside of the ideal zone of thermal comfort. This indicates that targeting a temperature of 18
for the main living area, in a ducted heat pump system, may leave the rest of the household uncomfortable (slightly cold).
5. Conclusions
Poor quality housing represents a form of deprivation that disproportionately impacts those from low income and marginalised communities. Poor quality housing generates indoor spaces which have poor indoor environmental quality (IEQ) measures, such as high relative humidity, low temperature, and are associated with higher incidence of respiratory illness, particularly among infants and young children. In this paper we presented a sample house from a collection of houses surveyed in the Manukau, Auckland region of New Zealand, which is an area with lower average income. This case-study house performed poorly in terms of most IEQ measures: all rooms were colder than recommended for indoor living spaces; the relative humidity was higher than recommended, and the thermal image results showed significant cold-spots and variations in temperature in living spaces; however, the moisture content of the walls was in the acceptable range. We then simulated the effect of a ducted heat pump if it were applied to this house. The simulation for this proposed HVAC system, conducted in ANSYS CFX 2021 R1 showed that when the system used an inlet temperature of and relative humidity 45%, it was capable of lifting the indoor temperature to above , including at the walls of the room, thereby eliminating cold spots and improving thermal comfort, particularly in the living areas (bedrooms and living room). The proposed system also reduced relative humidity to below 50%, which would conform to the acceptable range defined by ASHRAE Standard 55-2017, and improved the performance of the house in terms of the PMV and PPD measures of thermal comfort. When using a lower inlet temperature of to lift the living room’s temperature to the target of , the rest of the house performed worse in terms of thermal comfort, and it was likely the occupants would not be satisfied with the thermal comfort.