![Aerial view of the schools where data collection takes place. Photos adapted from [6 , 7] .](https://www.researchgate.net/publication/377475255/figure/fig2/AS:11431281222472795@1707235699494/Aerial-view-of-the-schools-where-data-collection-takes-place-Photos-adapted-from-6-7_Q320.jpg)
Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Data in Brief 53 (2024) 110 045
Contents lists available at ScienceDirect
Data in Brief
journal homepage: www.elsevier.com/locate/dib
Data Article
Dataset of indoor air parameter measurements
relating to indoor air quality and thermal
comfort in South African primary school
classrooms of various building infrastructure
types
Rita Elise van der Walt
a
,Sara Susanna Grobbelaar
b
,
Marthinus Johannes Booysen
b , ∗
a
Department of Electrical and Electronic Engineering, Stellenbosch University, South Africa
b
Department of Industrial Engineering, Stellenbosch University, South Africa
a r t i c l e i n f o
Article history:
Received 14 November 2023
Revised 16 December 2023
Accepted 4 January 2024
Available online 17 January 2024
Dataset link: Indoor air quality
measurements in South African primary
school classrooms of various building
infrastructure types (Original data)
Keywo rds:
Indoor air quality
Carbon dioxide
Particulate matter
Thermal environment
Portable classrooms
Brick classrooms
Learning environment
Developing country
a b s t r a c t
In the resource-constrained South African education sector,
infrastructure considered temporary or a backup in other
countries is used as permanent classrooms, primarily but
not exclusively in lower-income areas. Children’s cognitive
performance and comfort are directly impacted by indoor
air quality. Temperature, relative humidity, particulate mat-
ter and CO2 levels, substantial determinants of air quality
and thermal comfort, have not been investigated across dif-
ferent classroom building and infrastructure types. We mea-
sure these parameters with 11- mi n intervals in 24 class-
rooms at schools in Stellenbosch, South Africa. These class-
rooms consist of a range of different infrastructure types.
Container classrooms with and without insulation, mobile
(prefabricated) classrooms, and brick classrooms of different
configurations are included. Measurements are concurrently
sampled over ten months (249 days, still ongoing) across
multiple seasons with relevant metadata, including ambient
weather conditions, school days and times, and electricity
availability in the (South) African context, which impacts air
∗Corresponding author.
E-mail address: mjbooysen@sun.ac.za (M.J. Booysen).
https://doi.org/10.1016/j.dib.2024.110045
2352-3409/© 2024 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND
license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ )
2 R.E. van der Wal t, S.S. Grobbelaar and M.J. Booysen / Data in Brief 53 (2024) 110 0 4 5
conditioning usage. This dataset provides valuable insights
into true learning conditions in South African classrooms.
©2024 The Author(s). Published by Elsevier Inc.
This is an open access article under the CC BY-NC-ND
license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ )
Specifications Table
Subject Electrical and Electronic Engineering
Specific subject area A data-in-brief article presenting concurrent indoor air environment
measurements in South African classrooms of different building infrastructure
types and categories.
Data format Raw .xlsx file including sheets for codebook, metadata, other information, and
data in csv format.
Type of data Tabl e including measurement data (with metadata and flags)
Data collection Measurements in 24 classrooms in two South African primary schools using
custom sensing devices measuring CO2
(COZIR-LP-50 0 0), T, RH (DHT22), PM1.0
,
PM2.5
, PM10
(PMS5003). Power meter data is used to see whether power was
on at sample time. Weather data from the NEMS dataset (High resolution, for
machine learning) for Stellenbosch Central was used for weather conditions.
School days and hours are recorded. LoRaWAN-based communication is
utilised for live and continuous data capture.
Data source location All schools where data collection took place are in Stellenbosch, We stern Cape,
South Africa (33.9372 °S, 18.8596 °E).
Data accessibility Mendeley Repository
Repository name: Indoor air quality measurements in South African primary
school classrooms of various building infrastructure types. [1]
Data identification number: 10.17632/tys2gscdv7.6
Direct URL to data: https://data.mendeley.com/datasets/tys2gscdv7/6
At the time of publication, data collection was ongoing. The Mendeley
repository may be updated accordingly, and readers are encouraged to
download the latest version of the dataset. The information in this paper
relates to Version 6 of the repository.
1. Value of the Data
• Indoor air quality and thermal comfort parameters including CO2
and temperature sig-
nificantly impact concentration, comfort, performance, and motivation. Other parameters
including particulate matter have potential undesirable health impacts. Children are even
more susceptible to poor indoor air or thermal conditions than adults. Given the substan-
tial amount of time spent in classrooms while learning, and the cognitive nature of typical
school activities, obtaining information on these conditions and investigating the learning
environment becomes critical.
• South African schools utilise classrooms of various infrastructure types. Classroom infras-
tructure, typically considered as merely for temporary use or ‘backup’ classrooms else-
where, is used permanently in both high-income and low-income schools. The learn-
ing environments resulting from the different building infrastructures must be investi-
gated and quantified. This dataset allows for the assessment of the potential impact on
schoolchildren’s learning abilities, and how classroom infrastructure potentially disadvan-
tages schoolchildren who receive schooling in these classrooms.
• For the first time, this dataset concurrently measures indoor air environment parameters
in classrooms of different types, across various seasons. Weather parameters, electrical
availability, school days, lecture times and break times, and other metadata and relevant
information, including room dimensions and materials, are also captured. The vast num-
ber of concurrent measurements, given that classrooms are in similar orientations and
are exposed to the same ambient conditions, makes this data invaluable for direct com-
R.E. van der Wal t, S.S. Grobbelaar and M.J. Booysen / Data in Brief 53 (2024) 110 0 4 5 3
parisons and analyses of indoor air quality and thermal conditions in different classroom
types.
• The captured weather information, room dimensions and materials, in combination with
indoor measurements, provide the opportunity to inspect the influence of ambient con-
ditions on resulting indoor environmental conditions. The insulative effect of the wood
panels on the container classrooms, and how it affects the impact of different ambient
conditions on the indoor environment, can be assessed and modelled. This holds poten-
tial for the development of a thermal model or other thermal analyses, relating different
infrastructure parameters to the measurements, occupancy data and weather parameters.
• The flagging of data according to school days, school hours, and break times allows for
distinguishing and comparisons between school days and non-school days. The impact of
occupancy, human activity and emitted body heat on indoor conditions can be investi-
gated. Additionally, free-running conditions, when occupants do not influence or alter the
indoor air environment, can be investigated for direct infrastructure comparisons and the
resulting indoor conditions. This data will be invaluable when utilised in the development
of thermal models.
• Capturing electrical availability and load shedding allows for assessing the impact of load
shedding and the lack of air conditioning on the resulting indoor air environment for
different classroom types. This data can also be utilised to model the potential impact of
load shedding on pupils and their academic performance. Displaying the significant effect
that power outages may have on immediate indoor conditions may highlight the severity
of indoor conditions in schools where air conditioning systems are unavailable.
2. Background
Mobile/prefabricated and container classrooms are used extensively as permanent classrooms
in South African schools. No existing work quantifies the potentially unfavourable indoor air con-
ditions, despite the undesirable impact of poor environmental conditions being greater on chil-
dren than their adult counterparts [2–4] . It becomes critical to evaluate the indoor air environ-
ment in classrooms and its potential effects. Despite it being generally acknowledged that prefab
classrooms should serve as a short-term solution, schools are still equipped with them. Addi-
tionally, some schools also use container classrooms (converted shipping containers) as perma-
nent classrooms due to their cost-effectiveness. These mobile/prefab and container classrooms
are not exclusively located in low-income schools but are prevalent in nearly all South African
public schools, including high-income schools. The high prevalence of temporary classrooms in
South African schools makes it critical to understand the indoor air learning environment in
these classrooms.
3. Data Description
The dataset consists of a .xlsx file, “dataset and codebook.xlsx”, which consists of four sheets.
The sheet named “codebook” contains the index, value, and description of each field in the
dataset, and the sheet named “dataset” contains the current 372 084 data points and all mea-
surements and flags associated with it. Captured parameters and flags include the measure-
ment vector, information on ambient conditions, infrastructure type and category, room iden-
tifiers, hardware identifiers, school identifiers, pupil ages, electrical availability at the sample
time, and occupancy at the time of sample using school hours and break times. Table 1 provides
an overview of the field names, values, and descriptions of the data fields. This information is
also available in the “codebooks” sheet in the .xlsx file.
The sheet named “metadata” contains relevant metadata including classroom dimensions and
building materials for each room in the dataset. This includes an illustration to identify which
4 R.E. van der Wal t, S.S. Grobbelaar and M.J. Booysen / Data in Brief 53 (2024) 110 0 4 5
Tabl e 1
Adaption of codebook and description of parameters in the dataset, as of 7 November 2023, Version 6.
Parameter Description
Date Date [yyyy-mm-dd]
Time Time (GMT + 2) [HH:MM:SS]
Classroom Type Container, No Insulation
Container, With Insulation
Mobile/Prefab
Brick, First Floor
Brick, Second Floor
Brick, Single Story
Classroom Category Temporary (Container and prefab classrooms)
Permanent (Brick building classrooms)
Room Number Classroom number of that type
Device Code Sensing device code (hardware identification number)
School No 1 – School 1
2 – School 2
…- …
N – School N
Grade Grade of learners in classroom (0 to 7).
0 refers to pre-primary
Measured T Measured indoor temperature [ °C]
Measured RH Measured indoor relative humidity [%]
Measured CO2 Measured indoor carbon dioxide [ppm]
Measured PM1.0 Measured indoor PM1.0
[μg/m3
]
Measured PM2.5 Measured indoor PM2.5
[μg/m3
]
Measured PM10 Measured indoor PM10
[μg/m3
]
School Day Y – School day
N – Weekend, public holiday, or school holiday (no school)
School Hours Y –During school hours on a school day
N –Not during school hours
Break Time Y –During break time on a school day
N – Lesson time/after hours on a school day, or non-school
day
Power On On –Power on
Off – Power off/Load shedding
Outdoor Temperature Outdoor/ambient temperature [ °C]
Fig. 1. Different classroom configurations and definitions of dimensions in relation to the door face.
parameters constitute the width, height, and length, relative to the wall face with the door.
Table 2 and the corresponding Fig. 1 provides an adapted extract from this sheet.
The sheet named “other information” contains additional relevant information including lec-
ture times, school hours and break times for each school.
R.E. van der Wal t, S.S. Grobbelaar and M.J. Booysen / Data in Brief 53 (2024) 110 0 4 5 5
Tabl e 2
Adaption of metadata, classroom dimensions and materials. Corresponds to dataset Version 6.
Classroom Type School Room Numbers Classroom Configuration w [m] l [m] h [m] Ceiling Wall Floor
Container, No Insulation 1 1, 2 Fig. 1 a 7.12 5 11.825 2.685 Sheets Sheets Non-slip epoxy
Container, With Insulation 1 1 Fig. 1 a 7.125 11.825 2.685 Sheets Sheets Non-slip epoxy
Mobile/Prefab 2 1 Fig. 1 a 8.435 7. 24 2 2.650 Sheets Sheets Laminated
wood
2 Fig. 1 b 7.165 7.6 65 h1: 2.480
h2: 3.135
Sheets Sheets Vinyl
Brick, First Floor 1 1, 6 Fig. 1 a 6.840 6.930 3.060 Concrete,
painted
Brick Linoleum
2 Fig. 1 a 6.865 9.110 3.065 Concrete,
painted
Brick Linoleum
2 3, 4, 5 Fig. 1 a 6.828 6.882 2.941 Concrete,
painted
Brick Wood
Brick, Second Floor 1 1, 2, 3, 4, 8 Fig. 1 a 6.850 6.935 3.160 Drywall,
painted
Brick Vinyl
2 5, 6, 7 Fig. 1 a 6.930 7.0 55 3.040 Suspended
ceiling
Brick Wood
Brick, Single Story 1 1, 2, 3, 4, 5 Fig. 1 a 6.865 8.163 3.075 Drywall,
painted
Brick Linoleum
6 R.E. van der Wal t, S.S. Grobbelaar and M.J. Booysen / Data in Brief 53 (2024) 110 0 4 5
Fig. 2. Aerial view of the schools where data collection takes place. Photos adapted from [6 , 7] .
4. Experimental Design, Materials and Methods
Data collection commences at two quintile-5 (high affluence) primary schools in Stellenbosch,
Western Cape, South Africa (33.9372 °S, 18.8596 °E). All classrooms have a similar and compa-
rable number of pupils (27 to 32). The greatest line-of-sight distance between any two class-
rooms is 2.74 km, and ambient conditions in all classrooms can therefore be assumed to be
the same, using the data obtained from the NEMS dataset (High-resolution, for machine learn-
ing) for Stellenbosch Central [5] . All classroom dimensions are captured using a laser distance
measurer, and building materials are recorded. The schools are in the same load-shedding zone
(electricity is switched off periodically, typically for 2 to 5 h at a time, to reduce strain on the
South African electrical grid). In the context of this research, load shedding affects the use of
air conditioning and other mechanical HVAC systems. Therefore, electricity (air conditioning) is
off simultaneously for all classrooms in the dataset. All classrooms have access to working air
conditioning or HVAC systems/mechanisms. Fig. 2 depicts the satellite view of the two schools
and the geographic location of the various classroom types, showing that all classrooms have
similar orientations.
The classroom types and their infrastructure description and sample sizes are shown in
Table 3 and Fig. 3 shows images of the different classroom types.
Tabl e 3
Classroom type description (Number of rooms as on 7 November 2023, dataset Version 6).
Name and Description Number of
Rooms
Category Image
Container, No Insulation. Shipping containers converted to
be used as school classrooms, without insulation
2 Temporary Fig 3 a
Container, With Insulation. Shipping containers which were
converted to be used as school classrooms, with decorative
wood panels on the front and back. Wood panels were
installed for decorative purposes but leave a resulting
insulating effect on indoor conditions.
1 Temporary Fig 3 b
Mobile/Prefab. Prefabricated classrooms which were
designed and manufactured by dedicated companies,
intended as classrooms. Insulated with polyurethane foam
and coated with Aluzinc painted steel.
2 Temporary Fig 3 e
Brick, First Floor. First-floor classroom in a two-story brick
building.
6 Permanent Fig 3 c
Brick, Second Floor. Second-floor classroom in a two-story
brick building.
8 Permanent Fig 3 c
Brick, Single Story. Classroom in a single-story brick
building.
5 Permanent Fig 3 d
R.E. van der Wal t, S.S. Grobbelaar and M.J. Booysen / Data in Brief 53 (2024) 110 0 4 5 7
Fig. 3. Images of classroom types. Image credit: R.E. van der Wal t, 2023.
Tabl e 4
Sensing module specifications.
Parameter Sensor Module Resolution Accuracy Unit
T DHT22 0.1 °C ±0.5 °C °C
RH DHT22 0.1 % ±1 % %
CO2 COZIR-LP-50 0 0 1 ppm ±30 ppm ( ±3 %) ppm
PM1.0
, PM2.5
, PM10 PMS5003 1 μg/m3 98 % μg/m3
A sensing device to measure T, RH, CO2
, PM1.0
, PM2.5
and PM10
was developed, manufactured,
and installed. Table 4 summarises the integrated sensing modules used in the sensing device and
their most relevant specifications.
The sensing modules in Table 4 were used to create an integrated sensing device, depicted
in Fig. 4 a. Sensors were programmed with their classroom number and type to transmit 12-
byte LoRaWAN packets at 11-minute intervals, per the LoRa Alliance Fair Use Policy. This policy
dictates that any end device (sensing device) may not exceed 30 s of airtime in 24 h [8] . By
transmitting 12-byte packets every 11 min, with the configured radio parameters, the LoRaWAN
network usage is well within compliance with these limitations.
8 R.E. van der Wal t, S.S. Grobbelaar and M.J. Booysen / Data in Brief 53 (2024) 110 0 4 5
Fig. 4. Sensing device and typical classroom setup.
Tabl e 5
Number of sensing devices for each classroom type, as of 7 November 2023, dataset Version 6.
Classroom Type Number of Sensing Devices
Container, No Insulation 8
Container, With Insulation 4
Mobile/Prefab 8
Brick, First Floor 7
Brick, Second Floor 8
Brick, Single Story 8
This measurement resolution is comparable to that of related studies of indoor air in class-
rooms by Jovanovic et al., 2014 (10 min) [9] and Lovec et al., 2020 (15 min) [10] . The 11-minute
sampling rate is also of higher resolution than other related publications, including Bunyasi et al.,
2022, (40 min) [11] , Essah et al., 2016 and Gibberd et al., 2013 (60 min) [12 , 13] . Several class-
rooms of interest have multiple sensors placed in each room, increasing the effective sampling
resolution of classroom measurements. This is primarily done in classroom types for which there
are not many different rooms available, predominantly the container and mobile/prefab class-
rooms, which have four sensing devices per room. Given the vast number of classrooms for
the classroom types, each logging interval consists of multiple captured measurements for each
classroom type. Table 5 shows the number of logging devices for each classroom type.
The LoRaWAN gateway allowed for continuous and remote data collection without requiring
site visits. This minimised disturbances to school activity.
The sensing module was placed in an enclosure with large openings, allowing for sufficient
airflow and ventilation and ensuring accurate measurements. All sensor modules were installed
facing the outside of the device and were located next to an opening in the enclosure.
To ensure uninterrupted measurements during load shedding, sensors were powered with a
power bank as battery backup. Sensing devices were placed above ground level and away from
walls. Fig. 4 b depicts the typical classroom setup.
4.1. Weather data
Weather and climatic data with hourly resolution from the NEMS-dataset [5] were obtained
from a Stellenbosch weather station. Captured parameters include temperature (2 m elevation
corrected), relative humidity (2 m), precipitation amount, sunshine duration (minutes), solar ra-
diation, pressure (mean sea level), wind speed (10 m) and wind gusts. This data is not included
in the published dataset but the rights to it can be purchased from MeteoBlu [5] .
4.2. School days and school hours
Data collection commenced in two South African public schools which both follow the West-
ern Cape school calendar [14] . School hours and break times were obtained directly from the
R.E. van der Wal t, S.S. Grobbelaar and M.J. Booysen / Data in Brief 53 (2024) 110 0 4 5 9
Tabl e 6
Lesson and break times for each school. Lessons last 30 min and break times are 20 min.
Description School 1 School 2
School day commences 07:45 07:45
Lesson times 07:50–09:50 07:50–10:20
Break time 09:50–10:10 10:20–10:40
Lesson times 10:10–12:10 10:40 –12:40
Break time 12:10–12:30 12:40–13:00
Lesson times 12:3 0–14: 00 13 :0 0–14:0 0
School day concludes 14 :0 0 14: 00
schools. Measurements are flagged and sorted according to these dates and times. All lecture
periods for both schools last 30 min, with the first lesson commencing at 07:50 on school days
and the final lesson concluding at 14:00. Pupils do not move among classrooms between lessons
and remain in the classroom for all lessons during the day. Classrooms are therefore occupied by
the same pupils for the full school day duration, apart from break times. The flag ‘Break Time’
in the dataset indicates measurements taken during break times. Table 6 shows the lesson and
break times for each school.
4.3. Load shedding and power outages
Considering the ongoing South African energy crisis and load shedding, capturing data on
electrical availability becomes relevant because it affects the use of air conditioning or mechani-
cal HVAC systems. We, therefore, flag measurements taken while the power is off (and therefore
air conditioning measures are not usable), and the potential resulting impact on the indoor air
environment. Power meter data captures electrical availability with a 30-minute resolution to
obtain this information. All schools participating in this study are in the same electrical ‘zone’
and experience load shedding at the same time.
4.4. Dataset composition
Data collection commenced in February 2023, near the end of the South African summer sea-
son. Therefore, at the time of this publication, most measurements and data points are taken in
the winter season. Fig. 5 shows the number of measurements taken per month, and the compo-
sition of measurements for each month concerning classroom type and category.
Fig. 5. Dataset monthly composition and growth as of 7 November 2023, dataset Version 6.
10 R.E. van der Walt, S.S. Grobbelaar and M.J. Booysen / Data in Brief 53 (2024) 110 0 4 5
At first glance at the monthly dataset growth from Fig. 5 , the increase in monthly sample
volume may make it appear that the earlier months of data collection did not fully capture the
indoor conditions. However, this increase in volume is merely because of the larger number of
sensing device in use and near-full-resolution data capture (measurements per hour) has been
captured in the earlier months as well.
Table 7 shows the resulting dataset and its composition by filter and classification.
Tabl e 7
Dataset composition by different classifications and flags as on 7 November 2023, dataset Version 6.
Filter Classification Data days Data points Weight [%]
Class Category Temporary Classrooms 249 173 182 46.5
Permanent Classrooms 239 198 902 53.5
Class Type Container, No Insulation 240 60 873 16.4
Container, With Insulation 211 43 179 11.6
Mobile/Prefab 232 69 13 0 18. 6
Brick, First Floor 230 79 646 21.4
Brick, Second Floor 231 54 234 14 .6
Brick, Single Story 221 65 022 17.5
School Day School Days 141 230 17 7 61.9
Non-School Days 108 141 907 38.1
School Hours School Hours 14 0 55 351 14. 9
Not School Hours 249 316 733 85.1
Electrical Availability Power On 248 313 337 84.6
Power Off (Load Shedding) 215 57 243 15.4
Total: 372 084
Table 8 shows the number of samples containing measurement data for each measured pa-
rameter.
Tabl e 8
Dataset composition by measurement parameter as on 7 November 2023, dataset Version 6.
Parameter Data points Data days
CO2 338 195 240
T 372 079 249
RH 372 082 249
PM1.0 354 202 249
PM2.5 354 202 249
PM10 354 202 249
Limitations
Occasionally, devices or the gateway were plugged out by students or school staff, leading
to some gaps in measurement data from that specific device, classroom, or school. In these
situations, effort was made to remediate it as soon as possible. However, as a result, not all
classrooms or devices have continuous data. The vast number of devices and classrooms does,
however, make up for this.
Not all devices are equipped with all sensing modules. Although most sensing devices are
equipped with CO2
, PM and T/RH sensor modules, some data points do not have measurement
data for all parameters. Refer to Table 6 for the number of data points for each filter, flag, or pa-
rameter. Initial CO2
measurements were captured with a different CO2
sensor, which was found
to be inaccurate, and these measurements have been removed from the dataset. Earlier mea-
surements, therefore, have fewer CO2
measurements.
R.E. van der Wal t, S.S. Grobbelaar and M.J. Booysen / Data in Brief 53 (2024) 110 0 4 5 11
Ethics Statement
The authors confirm that the ethical requirements for publishing data in Data in Brief have
been read and understood and further confirm that the data collected does not involve human
subjects, animal experiments, nor data collected from social media platforms.
Data Availability
Indoor air quality measurements in South African primary school classrooms of various
building infrastructure types (Original data) (Mendeley Data)
CRediT Author Statement
Rita Elise van der Walt: Conceptualization, Data curation, Formal analysis, Investigation,
Methodology, Software, Visualization, Writing – original draft; Sara Susanna Grobbelaar: Con-
ceptualization, Formal analysis, Investigation, Methodology, Writing –review & editing, Supervi-
sion; Marthinus Johannes Booysen: Conceptualization, Formal analysis, Investigation, Method-
ology, Writing –review & editing, Supervision, Funding acquisition, Project administration.
Acknowledgements
This research did not receive any specific grant from funding agencies in the public, com-
mercial, or not-for-profit sectors. We would like to thank the schools that participated in this
study.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal rela-
tionships that could have appeared to influence the work reported in this paper.
References
[1] Rita van der Walt, Susara Grobbelaar, M.J. Booysen, Indoor Air Quality Measurements in South African Primary
School Classrooms of Various Building Infrastructure Types, 2023 Mendeley Data, V6, doi: 10.17632/tys2gscdv7.6 .
[2] S.S. Korsavi , A. Montazami , D. Mumovic , Indoor air quality (IAQ) in naturally ventilated primary schools in the UK:
occupant-related factors, Build. Environ. 180 (2020) 106992 .
[3] S. Sadrizadeh , R. Yao , F. Yuan , H. Awbi , W. Bahnfleth , Y. Bi , G. Cao , C. Croitoru , R. de Dear , F. Haghighat , P. Kumar ,
M. Malayeri , F. Nasiri
, M. Ruud , P. Sadeghian , P. Wargocki , J. Xiong , W. Yu , B. Li , Indoor air quality and health in
schools: a critical review for developing the roadmap for the future school environment, J. Build. Eng. 57 (2022)
104 908 .
[4] G. Torriani , G. Lamberti , G. Salvadori , F. Fantozzi , F. Babich , Thermal comfort and adaptive capacities: differences
among students at various school stages, Build. Environ. 237 (2023) 110 34 0 .
[5] NOAA Environment Monitoring SystemNemsaf01 Dataset, 2023 URL www.meteoblue.com .
[6] Google. Krigeville, Stellenbosch, 2023. URL https://www.google.com/maps/@-33.9465254, 18.8556574, 281m/data=
!3m1!1e3?entry=ttu .
[7] Google. Mostertsdrift, Stellenbosch, 2023. URL https://www.google.com/maps/@-33.9334234, 18.8804177,162m/data=
!3m1!1e3?entry=ttu .
[8] The Things Network. Duty cycle, 2016. URL https://www.thethingsnetwork.org/docs/lorawan/duty-cycle/ .
[9] M. Jovanovi´
c , B. Vu ˇ
ci´
cevi´
c , V. Turanjanin , M. Živkovi´
c , V. Spasojevi ´
c , Investigation of indoor and outdoor air quality
of the classrooms at a school in serbia, Energy 77 (2014) 42–48 .
[10] V. Lovec , M. Premrov , V. Žegarac Leskovar , Thermal comfort and indoor air quality after a partially energy-efficient
renovation of a prefabricated concrete kindergarten constructed in 1980’s in Slovenia, Prostor 28 (2020) 346–359 .
12 R.E. van der Walt, S.S. Grobbelaar and M.J. Booysen / Data in Brief 53 (2024) 110 0 4 5
[11] E.W. Bunyasi , K. Middelkoop , A. Koch , Z. Hoosen , H. Mulenga , A.K.K. Luabeya , J. Shenje , S.C. Mendelsohn , M. Tame ris ,
T.J. Scriba , D.F. Warner , R. Wood , J.R. Andrews , M. Hatherill , Molecular detection of airborne mycobacterium tuber-
culosis in south African high schools, Am. J. Respir. Crit. Care Med. 205 (2022) 350–356 .
[12] E. Essah , W. Kwawu , L. Motsatsi , Indoor Microclimate in a South Africa School: Impact of Indoor Environmental
Factors, 2016 .
[13] J. Gibberd , L. Motsatsi , Are Environmental Conditions in South African
Classrooms Conducive for Learning?, 2013 .
[14] Western Cape Education DepartmentSchool Calendar and Public Holidays, 2023 URL https://wcedonline.
westerncape.gov.za/school- calendar- and- public- holidays .
