Outdoor testbeds for studies on performance of building envelopes: review and recommendations

Abstract

High-performance building envelope materials, components and systems have been developed to achieve reduction in energy demand and meanwhile maintaining a satisfactory indoor environmental quality. The performance of building envelopes is generally assessed through numerical and experimental methods. Experimental work conducted with outdoor testbeds provides live data under real weather conditions, and therefore become an effective approach to evaluating the performance façade technologies. This paper reviewed 82 outdoor testbeds employed in studies on performance of building envelopes. They are categorized according to scale and form. Representatives in each category are discussed in details, and the main features are compared and summarized systematically. The functions of testbeds are analyzed according to the dedicated experimental purposes. Different construction materials and structures are also compared. On this basis, recommendations are made for future design, construction and application of outdoor testbeds.

1. Introduction

Building sector accounts for more than one third of the global energy consumption and carbon emissions. According to the International Energy Outlook Report, the global energy consumption will grow by 56% from 2010 to 2040. As the rapid growth of global floor area especially in developing countries, more energy has been spent on maintaining thermal, visual, and acoustic comfort for occupants. In order to put the building sector on track with the Net Zero Emissions by 2050 Scenario, it is essential to reduce the overall energy consumption of buildings under the premise of ensuring high indoor environmental quality (Agency 2013; Walsh, Urban, and Herkel 2009).

Since energy loss through building envelopes occupies 20–60% of the total energy use in buildings (Balali and Valipour 2020), building envelopes play a crucial role in reducing building energy consumption. Compared to traditional ones, advanced façade systems have shown remarkable potential in reducing energy demand and meanwhile providing insulation, moderating ventilation, selectively admitting desirable solar heat gain and natural daylight, etc. (Akram et al. 2023; Heidari Matin and Eydgahi 2022; Liang et al. 2019). However, most of the advanced technologies are still in concept or prototype stage. Alkhatib et al. (Alkhatib et al. 2021) listed 24 groups of advanced façade technologies, out of which only 1/5 are available in industry. One of the main reasons is that there is still a big challenge in characterizing and understanding the real performance of these technologies.

Evaluating the effectiveness of an innovative facade component/system is a complex task. It is therefore necessary to develop appropriate methods that correctly estimate the actual behavior of building components in an economical way. Research on characterizing and assessing building envelopes components/systems can be broadly categorized into numerical simulation and experimental measurements. Experimental measurements may be further categorized according to the facilities adopted, which are indoor laboratory facilities, outdoor test facilities, or real-scale buildings (Cattarin et al. 2016). Indoor Laboratory facilities are suitable for measuring particular performance indicators under precisely controlled boundary conditions, such as the “hot-box” method for measuring U-values (Misiopecki, Grynning, and Gustavsen 2023) and calorimetric measurement for determining g-values (ISO 19467 2017). Outdoor test facilities are test chambers under boundary conditions determined by real weather and sometimes occupant behavior (Cattarin et al. 2016; Judkoff and Neymark 2006; Loutzenhiser et al. 2008; Roels et al. 2015; Strachan and Vandaele 2008). Real-scale buildings are often used as representative of a certain architectural type to assess the applicability of a particular façade technology, with occupants living or working in sometimes (Korsnes, Berker, and Woods 2018). Indoor laboratory facilities can produce and replicate identical boundary conditions. In comparison, real-scale buildings may provide the most realistic and complex boundary conditions, i.e., both real weather for exterior walls and dynamic effects between different internal thermal zones for interior walls), which are essential for understanding the performance of a building envelope component in the real world. The outdoor test facilities make the perfect intermediate complement, because the dynamic outdoor weather condition is taken into account and meanwhile the simplicity of the test chamber reduces the level of complexity in building service system and heat transfer through boundary surfaces. The data obtained can be used to evaluate the indoor laboratory results, to validate building energy consumption simulation tools, etc.

There have been some scholars who studied outdoor testbeds for evaluating building envelope technologies. Janssens (Janssens 2016) presented an inventory of 25 full-scale test facilities for evaluation of building energy performances. Details were provided for each facility on main functionalities, objectives, lay-out, typical equipment and operation, examples of measuring campaigns and corresponding analysis methods, although there was a lack of systematic classification and comparison. Cattarin et al. (Cattarin et al. 2016) proposed a classification method focused on the test cell construction concept and test principle adopted, which helped understand the possibilities and limitations of the outdoor test cells. Emphasis was mainly put on test cells, so not many examples of real-scale test buildings were mentioned. Nevertheless, it is the most comprehensive review on outdoor test cells to the authors’ knowledge. La Ferla (La Ferla 2020a) conducted a comprehensive review of different areas of research on adaptive façade systems and provides both general and specific knowledge about numerical and experimental research methods in this field, although roof systems were not discussed. Jin and Meng (Jin and Meng 2020) pointed out the future direction of the development of outdoor test cells after reviewing thirty-three outdoor test cells in and out of China, but a classification solely based on the geographical location is limited and real-scale test buildings were excluded.

As intelligent façade technologies develop swiftly over the past decade, more outdoor testbeds were built and upgraded to perform complex performance assessment. Unlike experiments performed with indoor laboratory facilities, which usually follow well established standards and procedures, studies carried out in outdoor test facilities and real-scale buildings are more diverse and scattered, and most of them are specially customized, making it difficult for cross-comparison. It is therefore worthwhile to systematically review these testbeds, in order to characterize their construction features and functions, understand their historical revolution, and identify their limitations. This will eventually contribute to establishing the future development direction for the design and construction of outdoor testbeds for more complex and advanced building envelope performance assessment.

This paper reviewed 82 outdoor testbeds published between 1928 and 2023 for assessing the performance of building envelope material, components, and systems. Section 2 provides a classification of the outdoor testbeds according to scale and form, and features of each category are discussed and compared. Section 3 investigates the experimental purpose of studies based on these testbeds. Section 4 made comparison in terms of materials and construction. Based on the previous review, Section 5 provides recommendations for the design, construction and use of future outdoor testbeds.

2. Outdoor experimental facilities classified by typology

2.1. Classification

Outdoor experimental facilities are used for on-site measurements with boundary conditions determined by the weather and sometimes affected by the behavior of the occupants. In this study, we mainly focus on outdoor experimental facilities with a dimension larger than 2 × 2 × 2 m, as these facilities tend to be more complex and provide more profound implications for the assessments.

As shown in Figure 1, the outdoor experimental testbeds may be divided into three general categories by scale and form: stand-alone test cells, embedded test rooms, and real-scale test houses/buildings. Stand-alone test cells are chambers built specially for experimental purposes. They are usually small in size (average area of the test room around 3.0 × 3.0 m), and may contain more than one test rooms and some supplementary space. Depending on the number of test rooms in one cell and the total number of test cells, they are further divided into single-room test cells, multi-room test cells and multiple test cells (a group of individual test cells). Stand-alone test cells may be customized according to various research objectives and require relatively short construction time. Most or all the surfaces of the test cells are exposed to the outdoor environment, which is not a real representation of boundary condition for an office or residential space. In addition, such test cells usually employ simplified service system. Occupants’ effect is excluded in most cases. Embedded test rooms are incorporated in an existing building. Apart from the test room for experiments, the rest space of the building is in practical use such as offices. They usually have one or two surfaces exposed to the outdoor environment, which are used as test surfaces. The rest surfaces subjected to thermal effects from other internal thermal zones, which can be considered as real-world boundary conditions. Real-scale test houses/buildings are houses or buildings that are either initially dedicated to experimental purpose, or originally for practical use such as residence and converted for experiments later. The former requires longer construction time and more funding support compared to the latter. They usually have real-scale HVAC system as well that may participate in the research for more complex tasks. During the test, the house or building may be occupied or not, according to the specific research objective.

Figure 1. Classification of outdoor testbeds for building envelope performance assessment.

2.2. Stand-alone test cells

Stand-alone test cells are individual test chambers. They are usually small in size (average area of test room is around 3 × 3 m), and may include more than one equipment rooms or other supporting areas. According to the number of test rooms and the way they are conjugated, they can be classified into: single-room test cells, multi-room test cells, and multiple test cells.

2.2.1. Single-room test cells

A single-room test cell contains only one room that simulates a working or residential space. Façade material or component is usually installed on one surface and tested. Table 1 summarizes detailed information of 24 single-room test cells. Some of the single-room test cells also include a supporting space such as buffer zone (approximately 1 m wide), or space for equipment and storage.

Table 1. Examples of single-room test cells.

Picture	Name of test cell	Construction time	Country	Facility Owner	Size/Dimensions (W × L × H)/Layout	Features	Tested systems & aspects	HVAC system
	PASSYS (Aleo et al. 2001; Assimakopoulos et al. 2007; Hahne and Pfluger 1996; Van Dijk and Van der Linden 1993)	1985	Ten European Union countries	Commission of the European Communities	2.75 × 5.0 × 2.75m (test room interior dimension)	35 Test cells located in different European countries, each has a test room and service room	Steady-state evaluation of thermal performance of passive solar building components. Validation of building design simulation tools.	A heating/cooling system with an electric heater, and air/water heat exchanger, and a fan.
	PASLINK (Baker and Van Dijk 2008; Dimoudi, Androutsopoulos, and Lykoudis 2004; García-Gáfaro et al. 2020; Goethals et al. 2012; Kotsiris et al. 2012; Leal and Maldonado 2008)	1989	Ten European Union countries	Commission of the European Communities	13.8 m²*2.75m	Test cells located in different European countries, each has a test room and service room	Characterization of dynamic behavior of building components. Validation of building design simulation tools.	A heating/cooling system with an electric heater, and air/water heat exchanger, and a fan.
	Calorimetric Façade and Rooftop Test Facility (Kersken 2021a; Strachan et al. 2015)	2001	Germany	IBP	3.8 × 3.8 × 3.5m 3.2 × 2.3 m (max specimen size)	Test chamber can be turned by 360⁰ and tilted from a vertical to a horizontal specimen.	Determination of the energetic and visual performance of transparent large-scale façade and roof elements	Air conditioner to maintain air temperature to avoid activation of thermal mass
	Prefabricated outdoor room facility (Dehra 2019)	2004	Canada	Concordia University	N/A	Prefabricated structure with two double façade sections	Acoustic and thermo-fluid insulation performance of an airflow window with a photovoltaic solar wall	The transportation of heat was achieved through manually operated intake and set of exhaust dampers. The exhaust damper allowed the pre-conditioned outdoor air directly into the room.
	The Cube (Kalyanova et al. 2009; Kalyanova and Heiselberg 2008; Marszal et al. 2009)	2006	Denmark	University of Aalborg	6.0 × 6.0 × 6.0m	1 Test room; 1 instrument room; 1 engine room. Flexible façade configurations, Manually controlled blinds.	Detailed thermal performance assessment of double skin façade and different types of shading devices.	A ventilation system with the heating and cooling unit is installed in the experiment room. A cooling unit is installed in the experiment room activated.
	PASLINK test cell at CIEMAT (Jiménez, Porcar, and Heras 2009; Jiménez, Porcar, and Heras 2008; Jiménez and Madsen 2008)	2008	Spain	PSA-CIEMAT	6.13 m²	1 test room; rotatable base	Thermal properties (U-value, g-value) of building envelope components;	The same as other PASLINK test cells
	Bamboo structure wall test cell (Wang 2011)	2011	China	Hunan University	2.0 × 2.0 × 2.0m	1 test room	Coupled heat and moisture transfer characteristics of bamboo structure wall	None.
	LGI test cell (Bigot 2011; Guichard et al. 2014)	2011	France	Reunion University	Volume of 30 m³	Single glazed window, inclined roof	Thermal performance of PCM integrated roof	Not specified.
	Btga-box (Schweiker et al. 2012)	2012	Germany	Wuppertal University	2.82 × 7.06 m	A prefab concrete building cell, south surface Exchangeable. electrically driven external shading device	Thermal and visual performance of windows and vacuum insulation panels.	Mechanically (fan) or natural ventilated.
	Test cell at University of Florence (Alcamo and De Lucia 2014; Luible et al. 2018; Romano, Gallo, and Donato 2021)	2014	Italy	University of Florence	2.8 × 2.8m (Test room)	Insulated wooden structure with rotatable base, an upgrade of PASSYS test cell	Thermal performance test of façade components	The same as PASSYS test cell
	Chang’an University test cell (Qian, Guan, et al. 2016)	2016	China	Chang’an University	3.0 × 3.0 × 3.0m	1 test room, concrete and brick wall with insulation	hygrothermal behavior of facades of new building	Two electric radiators for heating.
	Delft University of Technology test cell (Stec and Van Paassen, 2005; Stec and Van Paassen 2002)	2018	Netherlands	TU Delft	3.1 × 1.9 × 2.68m	thermally light construction with a double skin façade. 1 test room with a monitoring room	Thermal performance and control strategy of double skin facade	An electric radiator (1750 W) supplies heating to the test-cell. A Daikin Split-Sky indoor air conditioner supplies air-cooling.
	Kovilpatti experimental room (Karthick et al. 2020)	2019	India	National Engineering College	3.0 × 3.0 × 6.0m	Steel and wooden structure, with sloped roof. 1 test room	Thermal performance and energy conversion efficiency of semi-transparent photovoltaic (STPV) module as rooftop window,	None.
	Anseong testbed (Ko, Oh, et al. 2020)	2020	Korea	KyungHee University	14.57m^2×2.9m	Well insulated testbed, no ventilation, double glazed windows	Energy consumption of suspended particle device (SPD) windows	Heat pump cooling capacity of 2500 W, heating capacity of 2700 W.
	MATELab (Luna-Navarro and Overend 2021)	2021	UK	University of Cambridge	4.5 × 5 × 2.5m	1 test room or can be partitioned into two test rooms. South, east and west facades have reconfigurable wooden frames.	Occupant-facade interaction, influence of façade on indoor environmental quality (IEQ)	Heat-pump with heat-recovery unit made of two identical and separate systems for heating, cooling and ventilation. Under Floor Air Displacement (UFAD) system for ventilation
	Experimental platform in 2iE (Neya et al. 2021)	2021	Burkina Faso	International Institute of Water and Environmental Engineering (2iE)	3.06 × 4.06 × 4.33m	0.14m thick Compressed Stabilized Earth Brick (CEB) walls with single glazing on north and south.	influence of insulation and thermal mass on building thermal performance	None.

In 1986, the European Commission launched the Passive Solar Energy Components and Systems Testing (PASSYS) project (Aleo et al. 2001; Assimakopoulos et al. 2007; Hahne and Pfluger 1996; Van Dijk and Van der Linden 1993) (Figure 2), which was the world’s largest research project on methods for assessing the thermal performance of building components at the time. A total of 35 outdoor test cells were built during the project period, 29 of which were designed according to a unified standard and each contained a test room (with an area of 13.8 m² and a clear height of 2.75 m) and an adjoining equipment room.

Figure 2. PASSYS (Baker and Van Dijk 2008).

Using the PASSYS test cells as a prototype, more upgraded test cells have been established. For example, the LABIMED test cell (Alcamo and De Lucia 2014; Luible et al. 2018; Romano, Gallo, and Donato 2021) and CIEMAT (Jiménez, Porcar, and Heras 2009; Jiménez, Porcar, and Heras 2008; Jiménez and Madsen 2008) maintained the PASSYS floor plan with an upgrade at the bottom of the structure. Figure 3 shows the layout of LABIMED test cell that included one test room and an auxiliary space/equipment room. It could rotate on a circular platform to allow for targeted testing (Figure 4). IBP's calorimetric test facility (Kersken 2021; Kersken et al. 2021) (Figure 5) could rotate 360°and tilt up to 90°, allowing a façade component being tested at any slope and orientation.

Figure 3. Layout of LABIMED test cell (Alcamo and De Lucia 2014).

Figure 4. LABIMED test cell (Alcamo and De Lucia 2014).

Figure 5. Calorimetric test facility (Kersken 2021).

A few test cells were designed to simulate an office environment and occupants were involved during the test, such as the CEC (Ko, Schiavon, et al. 2020) and daylight laboratory (Bakker et al. 2014). During the daylighting test (Bakker et al. 2014), participants in the experiments were required to sit at a distance of 1.5 m from the window for several hours. Different window configurations were tested and the resulting the indoor environmental conditions and energy use were assessed.

2.2.2. Multi-room test cells

A multi-room test cell contains two or more rooms that simulate identical working or living spaces. Different building envelope components were usually installed and tested in those rooms respectively. Auxiliary spaces such as plant rooms may be contained in the test cell. Table 2 summarizes detailed information of 23 multi-room test cells.

Table 2. Examples of multi-room test cells.

Picture	Name of test cell	Construction time	Country	Facility owner	Size /Dimensions (W × L × H)/Layout	Features	Tested systems & aspects/parameters	HVAC systems
	MoWiTT (Klems and Keller 1986; Klems 1992, 1988)	1979	USA	LBNL	2.44 × 3.05 × 2.34m (Interior dimensions for each test room)	2 Test rooms with removable exterior wall and roof panels. 1 Service zone.	Thermal and energy performance of windows (including skylights)	Forced-flow, temperature-controlled air supply system with a fan coil unit and a chiller for heating and cooling.
	The VLIET -test building (Abuku, Blocken, and Roels 2009; Desta, Langmans, and Roels 2011; Janssens and Hens 2007; Langmans and Roels 2015; Roels and Deurinck 2011; Saelens, Roels, and Hens 2004)	1996	Belgium	KU Leuven	25.2 × 7.2 × 2.7m	2 Test rooms: the south east side (8 test walls and 4 flat roofs) and the north west side (12 walls and 6 duo pitched roofs)	Hygrothermal behavior of facades and roofs; thermal mass; ventilation	Not specified.
	EMPA facility (Manz 2004; Simmler, Binder, and Vonbank 2000; Simmler and Binder 2008)	1999	Switzerland	EMPA	2.9 × 4.6 × 2.4m	Wooden structure, 2 test rooms	Determination of the total solar energy transmittance of glazing with venetian blind shading/double facades; thermal performance of double facades	Mechanical ventilation
	ERS test facility (Kunwar et al. 2019; Loutzenhiser, Maxwell, and Manz 2007; Price and Smith 2000)	2000	USA	University of Iowa	8 Test rooms (23.63 m²)	3 Pairs of test rooms are located at the perimeter of the building (east, south, and west) and the other 2 test rooms are inside the facility. Different types of windows installed for each pair of test rooms, interior shading installed	Empirical validation of daylighting algorithms (illuminance, light power reduction, associated thermal load interaction) for various windows and shading devices	The primary system for cooling was air-cooled liquid chillers. Air handling units (AHUs) were utilized for cooling. Rooms had variable air volume (VAV) terminal unit to control the amount of supply air required to maintain the setpoint temperature in each room.
	LBNL test cell (Lee et al. 2015, Lee, DiBartolomeo, and Selkowitz 2006, 2005)	2003	USA	LBNL	88.40m²	3 Test rooms	Control strategies; visual comfort;	A packaged air-conditioning (heat pump) unit provides space conditioning to the thermal guard space. Dedicated fan coil units provide space conditioning to each individual test room.
	VERU (Schweiker et al. 2014)	2004	Germany	Fraunhofer Institute for Building Physics IBP	150m² 6 square test rooms	Reinforced concrete 3-storey test building, a pair of test room on each floor. Each test room can change internal depth and they can be combined. Flexible façade configurations	Integral investigation of façades, spaces, and building services to give practical information on energy consumption, visual and thermal comfort.	Fully conditioned with centralized system, thermally-activated slabs
	ISE test facility (Wienold and Christoffersen 2006, 2005)	2005	Germany	ISE & SBi	7.3 × 4.6 × 3 m	2 Identical experimental rooms, the distance from floor to the suspended ceiling can be changed	Glare effect from different window sizes and glazing typologies	Not specified.
	Test Cells (Silva, Almeida, and Mendonça 2006a, Silva et al. 2006b)	2006	Portugal	University of Minho	N/A	3 Different test cells(a passive solar optimized construction, a Portuguese conventional construction, and a super-insulated lightweight construction)	Thermal performance of integrating PCM in floors of a lightweight structure. Energetic and environmental performance of sustainable building technologies (high thermal mass and appropriate shading, etc.)	None.
	BSRTU test facility (Buxbaum et al. 2008; Janssens 2016)	2008	Austria	BSRTU	10.3 × 18.2 × 2.7m	Wooden frame structure, separation of roof areas allows for test of 14 different roof assemblies; variable spacing of partition walls	Hygrothermal behavior of facades; interactions of building component with indoor comfort conditions	Double air conditioning system.
	BESTLab (Favoino et al. 2018)	2010	France	EDF research center of Les Renardières	3.3 × 3.6 × 2.9m or 3.9 × 3.4 × 4.6 m	12 Test rooms (6 for vertical facades, 6 for sloped roofs), over insulated walls.	Thermal behavior and energy characterization of technologies associated or connected to building envelope components	HVAC split unit system with a 2650 W cooling and a 3050 W heating capacity, 460 m^3/h flow rate.
	Lwf-FAÇADE-facility (Laboratory for Intelligent Building Technology 2011)	2011	Germany	Rosenheim University of Applied Sciences	13 × 5.5 × 6m	3 Test rooms surrounded by an air-conditioned service room; can rotate and tilt; facade openings of 3 × 4 m individually adaptable	Façade assessment: physical behavior, thermal comfort, ventilation, control strategies, visual comfort	Heating and cooling supplied from suspended ceiling of each test room.
	LOBSTER (Safizadeh, Schweiker, and Wagner 2018; Schweiker et al. 2014)	2014	Germany	Karlsruhe Institute of Technology	4 × 6 × 3m	Rotatable structure, 2 test rooms, room surfaces are cooled or warmed by hydronic capillary tubes, operable window	Thermal comfort and occupant behavior	Two separated heat pumps for radiative heating and cooling, Ceiling fans or mechanically ventilated
	ZEB test cell (Cattarin et al. 2018; Goia, Schlemminger, and Gustavsen 2017; Gullbrekken, Kvande, and Time 2017)	2015	Norway	NTNU & SINTEF	12.6 × 10.8 × 4.7m	2 South exposed test rooms contained in a guarded cell	Energy performance and indoor environmental quality of building envelope components and building equipment components	Air conditioning of the cells bases on a balanced mechanical ventilation layout. The supply section consists of an adjustable fan, a water-based cooling coil, a PID controlled electric coil, followed by a silencer. Air system of the guard volume consists of a silence, an adjustable circulation fan, a water-based cooling coil, a PID controlled electric coil, and a second silencer.
	MATRIX (Ascione, De Masi, et al. 2016)	2015	Italy	University of Sannio	6.0 × 6.0 × 5.5m	2 Test rooms; can rotate of 360°; steel frame structure with wooden wall and roof; removable and configurable walls and windows; various mechanical systems	Hygrothermal characterization of facades/materials; energy performance of different envelope/HVAC system; evaluations of solutions for indoor comfort	The test cell is equipped with two different systems for heating and cooling, based on hydronic technologies (fan coil units, radiant systems).
	AEL test offices (Xiong and Tzempelikos 2016)	2016	USA	Purdue University	5.0 × 5.2 × 3.4m	2 Identical, side-by-side test rooms equipped with reconfigurable facade, shading and lighting systems	The impact of facade design and control options on indoor environmental conditions and energy use	N/A
	FACT (Bianco et al. 2017)	2017	France	CEA–INES	10 Different test cells (9 m^2 each)	Two-storey building, metallic structure, 5 test cells on ground floor and 5 on first floor. 8 Test cells each presents two façades of test with an area of 7.70 m² (2.3 × 3.3 m), 2 test cells each present only one façade of test south exposed.	Perform energy and IEQ test as well as to evaluate technically innovative building envelope solutions (aerogel based envelope component, cladding unit for deep renovation, etc.)	An HVAC system with a reversible heat pump.
	BCA SkyLab test cell (Yang et al. 2018)	2018	Singapore	Nanyang Technological University	43.5m^2×3.5m (each test room)	360-Degree rotatable structure 1 test room, 1 reference room, and 1 control room	Energy efficiency of façade technologies, thermal comfort, control strategies of cooling systems	A fan coil unit (FCU) ACMV system was installed. The FCU operates at constant fan speed, and is controlled by a thermostat that is linked to the actuator that regulates the chilled water flow rate through the cooling coil in the FCU. The thermostat controlled the chilled water flow rate according to room temperature set point that is determined by a proportional–integral–derivative (PID) control. The conditioned air is then supplied to the cell through four 4-way spread type ceiling air diffusers and subsequently returned through the six slot type diffusers on the ceiling.
	Test cell in Sophia Antipolis (Souayfane, Biwole, and Fardoun 2018)	2018	France	Université Cote d’Azur	9.3m² Floor area each (test/reference room)	Well insulated wall; 1 test room, 1 reference room, 1 data acquisition room	Thermal behavior of a TIM–PCM wall	The test cell is left in free floating mode; the reference room is equipped with a cooling/heating system maintaining the room air temperature at 23 °C in summer and winter.
	SCUT test cell (Lu et al. 2019)	2019	China	South China University of Technology	6.1 × 3.1 × 3.2m	2 Side-by-side test rooms, air conditioner installed	Effect of the inner wall surface radiation rate on the indoor thermal comfort	Air conditioners for cooling
	CELLS (EPFL 2020)	2020	Switzerland	EPFL	Not specified.	Two identical office-like test rooms, side by side, are divided by a buffer zone. Each room has two windows on opposite facades.	Thermal properties and performance of building envelope, control strategies. Visual and thermal comfort. Performance of building-integrated PV panels.	heat pump and ceiling radiant panels. Mechanical ventilation.
	Smart TinyLab (Mohammadi et al. 2022)	2021	The Netherlands	Saxion University of Applied Sciences	3.6 × 4.8 × 2.9 m (each test room)	Two test rooms plus an installation room, rotatable.	Assessment of energy efficiency and indoor climate conditions resulted from building products, HVAC components, smart technologies, user-related aspects	Heating/cooling is provided by the air and the floor heating, generated by a single heat pump with a buffer. Variable volume ventilation.
	BeTOP (Berardi and Soudian 2023; Soudian and Berardi 2022)	2022	Canada	Toronto Metropolitan University	45m²	2 Test rooms and a vestibule; allows for simultaneous testing of eight façade modules, and two large ceiling modules; easy removal and integration of different façade systems.	Thermal and comfort performance of envelopes (e.g., ventilated façade, radiant ceiling) and the dynamic interactions between building enclosure components, indoor environments, and mechanical systems	Primary heating and cooling system include two air source heat pumps (one for each room). The second heating system is a hydronic radiant floor system that can be activated upon demand.
	Vertical greening measurement platform (Zhang, Zhang, and Meng 2022)	2022	China	South China University of Technology	3.6 × 6.5 × 3.1m overall 2.0 × 1.9 × 2.0m (Each test room)	2 Test rooms contained in a butter room	Cooling benefits (in radiation reduction and thermal comfort) of vertical green facades	Not specified.

Most early multi-room test cells contained 2 test rooms, and some of them were rotatable. For example, LBNL constructed the MoWiTT facility (Klems and Keller 1986; Klems 1992, 1988) in 1979 (Figure 6), which was rotatable and contained 2 test rooms and 1 service room. MATRIX (Ascione, De Masi, et al. 2016) was mounted on a rotatable steel structure that was installed on a reinforced platform. Lwf façade facility (Feldmeier 2006; Schaefle, Feldmeier, and Lux 2012) (Figure 7) was also a rotatable facility, which sat 2 m above the ground level. The external frame enabled the test cell to rotated toward different directions and at any inclination angles.

Figure 6. MoWiTT (a) Field configuration; (b) Plan (Klems 1988).

Figure 7. Lwf façade facility (a) Exterior view; (b) Plan (Janssens 2016).

The BSRTU test cell (Buxbaum et al. 2008) (Figure 8) was a flexible rectangular box with structural columns placed 1 m back from the wall assembly. Wall or roof sections can be installed between this frame structure according to experimental requirements, allowing for flexible adaptation of the layout of the room.

Figure 8. Layout of the BSRTU test facility (Janssens 2016) (Grey blocks indicate the load-bearing structures, red blocks indicate the exterior walls where the facade components will be tested; and the blue lines indicate partition walls).

More complex test cells featured with more than two test rooms were built to accelerate testing speed for comparable studies. The facility at a community college campus in Ankeny (Loutzenhiser, Maxwell, and Manz 2007; Price and Smith 2000) (Figure 9) contained 8 test rooms, a computer room, an office, 2 classrooms, a mechanical room and a service room. The LBNL test cell (Lee et al. 2015, Lee, DiBartolomeo, and Selkowitz 2006, 2005) (Figure 10) consisted of three identical side-by-side test cells built with almost identical construction materials to mimic a commercial office environment. The VLIET-test building (Abuku, Blocken, and Roels 2009; Desta, Langmans, and Roels 2011; Janssens and Hens 2007; Langmans and Roels 2015; Roels and Deurinck 2011; Saelens, Roels, and Hens 2004) had multiple flexible walls and test chambers simultaneously. The EMPA facility (Manz 2004; Simmler, Binder, and Vonbank 2000; Simmler and Binder 2008), ZEB test cell (Cattarin et al. 2018; Goia, Schlemminger, and Gustavsen 2017; Gullbrekken, Kvande, and Time 2017), SCUT test cell (Qian, Guan, et al. 2016), LOBSTER (Schweiker et al. 2014), etc., also consist of 2 or 3 test cells side by side.

Figure 9. Floor plan of the facility at community college campus in Ankeny (Price and Smith 2000).

Figure 10. (a) Floor plan and (b) Cross-section of LBNL test cell (Lee et al. 2005).

Apart from testing single façade components, it was also possible to evaluate entire façade system. For example, KUBIK (Figure 11) (Chica et al. 2011; Elguezabal et al. 2020; Garay et al. 2015) was designed to test a completely demountable envelope. It allowed reconfiguration and assessment of the façade at the building level. Moreover, INVISO Project (Industrialized Sustainable Housing)(Arranz et al. 2020) (Figure 12) aimed to develop a new industrial, lightweight, sustainable and energy-efficient building system, which was capable of harnessing, storing and managing solar passive energy, reducing heating energy demand, and avoiding overheating problems.

Figure 11. Replacing the façade for KUBIK (Janssens 2016).

Figure 12. INVISO industrialized sustainable construction system project (Arranz et al. 2020).

2.2.3. Multiple test cells

Multiple test cells are a group of identical test cells designed and constructed according to a common standard. They are located in the same site for comparative studies so that the performance indicators are comparable regardless of the climate. Table 3 summarizes detailed information of 13 multiple test cells.

Table 3. Examples of multiple test cells.

Picture	Name of test cell	Construction time	Country	Facility Owner	Size /Dimensions (W × L × H)/Layout	Features	Tested systems & aspects/parameters	HVAC system
	TWINS (Serra, Zanghirella, and Perino 2010)	2004	Italy	Politecnico di Torino	1.6 × 3.6 × 2.5m	2 Test cells	Evaluating active transparent/opaque facade thermal performances and testing responsive building envelope elements integrated with HVAC systems	Air conditioning system, independent ventilation system (external centrifugal fans driven by inverters and connected to the cell by PVC ducts).
	Solar air collector testbed (Yao 2010)	2010	China	TU Dalian	3.0 × 3.0 × 3.0m (each cell)	1 Reference cell, 2 test cells, masonry structure,	Thermal performance of solar air heating building	Solar air heating system
	PASSYS test cells at UIBK (Ochs et al. 2015; Siegele, Ochs, and Feist et al. n.d.)	2010	Austria	University of Innsbruck	7.56 m² each	2 PASSYS test cells	Thermal performance of active façade components	A façade integrated micro-heat pump (µ-HP) in combination with mechanical ventilation with heat recovery (MVHR)
	EGUZKI and ILARGI test cells (García-Gáfaro et al. 2022, 2020, 2012)	2011	Spain	LCCE	13.8 m²×2.75m	2 Test cells; PASLINK test cells	ventilation; thermal mass; PV;	The same as PASLINK test cell.
	Living wall test cells (Chen, Liu, et al. 2015)	2014	China	Huazhong University of Science and Technology	2.5 × 2.2 × 3.3m	2 Test cells; the distance between the living wall and the building wall is adjustable,	Thermal behavior of a living wall system	Air conditioner installed but not used.
	Solar air heating system test rooms (Chen, Liu, et al. 2015)	2015	China	TU Dalian	3.9 × 3.9 × 3.0m	1 Test cell; 1 reference cell	Thermal performance of façade with closed loop solar air collector system coupled with temperature control material	Closed-loop solar air heating system on test cell, no HVAC for reference cell.
	Geslab Facility (Alonso et al. 2016; La Ferla, Román, and Calzada 2020b)	2015	Spain	Polytechnic University of Madrid	3.3 × 3.3 × 3 m, with 2.1 × 2.1 × 2.1 m interior dimension	11 Test cells with well insulated opaque envelope, and different glazing/facade types;	Thermal performance of refurbishment façade solutions (Alonso et al. 2016); the thermal and comfort performance (thermodynamic performance, radiant and thermal asymmetry, local indoor thermal comfort levels) of radiant glass façade technology (La Ferla 2020a)	A single zone system constant flow (INVERTER heat pump) was used to heat and cool the modules with pre-conditioned air (Alonso et al. 2016); HVAC split unit system with a 2650 W cooling and a 3050 W heating capacity, 460 m3/h flow rate, for heating in reference cell; no HVAC system for test cell. (La Ferla 2020a)
	Test cells in Torreón (Rojas et al. 2016)	2016	Mexico	Universidad Nacional Autonoma de Mexico	2.8 × 2.8 × 2.7m	Two non-air-conditioned test cells with different envelopes of the same thickness for monolithic concrete buildings, one is mono-layered and other is two-layered.	Thermal performance of two envelope constructive systems for walls and roofs of monolithic concrete buildings	A fan was operating for night ventilation.
	University of Seville test cells (Calama-González, León-Rodríguez, and Suárez 2022; Domínguez-Torres et al. 2022)	2016	Spain	University of Seville	2.4 × 3.2 × 2.8m (Each test room)	Steel structure, 2 test cells, each contain 2 test rooms. Represents a social housing stock	Thermal comfort; ventilation	Each cell has an interior thermal control system, consisting of a reversible heat pump (air–water) and three fancoils. Fan for ventilation/ help simulate natural ventilation
	COOLSKIN test facility (Bröthaler et al. 2021)	2018	Austria	TU Graz	13.49 m^2	Concrete structure, 1 reference-room, 1 climatized test-room, and one supply container	PV output and cooling demand of an PV integrated façade module	Test-room was cooled by a plate heat exchanger on the ceiling, heated by a floor integrated pipework, and directly conditioned with a fan coil unit. The reference room was only supplied with warm water from the supply container. A heat load of 300 W was activated to simulate the internal thermal loads.
	Test cells (Ruiz-Valero et al. 2021)	2021	Philippines	Santo Domingo Campus	3.0 × 3.0 × 3.0m	4 guarded test cells, steel structure, well insulated wooden wall, rotatable foundation	Thermal and energy performance of facades of different thermal mass and glazed areas.	a split air conditioning system for heating and cooling
	GEMINI (Ciampi et al. 2021)	2021	Italy	University of Campania Luigi Vanvitelli	2.2 × 2.8 × 2.4m (each test cell)	Steel frame structure, 2 test cells; simulate the basic indoor environment module	Thermal characterization of double-skin façades and smart window with different geometries, layout, materials and technologies.	An air conditioning system for heating and cooling
	Comparative experimental platform in cold zone (Lin and Song 2021)	2021	China	Tsinghua University	3.0 × 3.0 × 3.2m (each test cell)	2 Test cells, steel frame structure, light wood envelopes.	Thermal and energy performance assessment of double skin façade design parameters and ventilation mode	An air conditioning system for heating and cooling

Santo Domingo Campus test cells (Ruiz-Valero et al. 2021) (Figure 13) consisted of four experimental cells with an external dimension of 3.00*3.00*3.00m. These test cells had wheels that allow rotation to evaluate the facades in different orientations. HUST test cells (Chen, Liu, et al. 2015) (Figure 14) were designed for testing a living wall, and the distance between the living wall and the exterior wall was adjustable between 30 and 600 mm. Slovak University test cells (Curpek, Hraska, and Cekon 2018; Čurpek and Čekon 2020), University of Seville test cells (Calama-González, León-Rodríguez, and Suárez 2022; Domínguez-Torres et al. 2022), EGUZKI and ILARGI (García-Gáfaro et al. 2022, 2020, 2012), and TWINS (Serra, Zanghirella, and Perino 2010) are all multiple test cells.

Figure 13. Building process of Santo Domingo Campus test cells (Ruiz-Valero et al. 2021).

Figure 14. Exterior and wall bulk view of HUST test cells (Chen, Liu, et al. 2015).

In summary, stand-alone test cells exhibit large variety in the scale of tested components, ranging from a single small piece of glazing to an entire façade system prototype. For example, with KUBIK (Chica et al. 2011; Elguezabal et al. 2020; Garay et al. 2015) the whole facade structure can be tested. Tests of this scale are more comprehensive and realistic. Early stand-alone test cells have no dedicated functions. Later on, some test rooms have been further defined to represent a typical working space (such as MATELab (Luna-Navarro and Overend 2021)) or residential space (such as the University of Seville test cells (Calama-González, León-Rodríguez, and Suárez 2022; Domínguez-Torres et al. 2022). This helps clarify the test objectives and corresponding needs for mechanical systems and other experimental instrument.

Due to the simple construction and flexibility of stand-alone test cells, the function of test rooms may also be altered according to experimental requirements. Most stand-alone test cells have test rooms directly exposed to the exterior environment. This could lead to over-heating problems that may not be present in a real-world situation, thereby resulting in some bias especially in measuring the performance indicators. Multi-story test cells such as KUBIK (Chica et al. 2011; Elguezabal et al. 2020; Garay et al. 2015) and VERU (Hauser, Sinnesbichler, and Eberl 2010; Heusler 2011; Sinnesbichler 2007; Sinnesbichler and Eberl 2009), or those with a buffering zone may alleviate this problem.

2.3. Embedded test rooms

Embedded test rooms are located in the perimeter space of a building. Therefore, at least one surface has access to daylight and view, or to the outdoor environment. Apart from the room itself that are used for experimental purposes, the rest part of the building operates for normal use (such as offices) in most cases. Table 4 lists 12 embedded test rooms with detailed information.

Table 4. Examples of embedded test rooms.

Picture	Name of test cell	Construction time	Country	Facility owner	Size/Dimensions (W × L × H)/Layout	Features	Tested systems & aspects	HVAC system
	Duplex flat (Bozonnet, Doya, and Allard 2011)	1995	France	Private	100m²	Duplex apartment under a sloped roof in a 4-storey building of 87 collective dwelling, sloped roof. Steel cladding with insulation.	Thermal performance of roof with cool paint in summer	No air conditioning (as common in most dwellings in France)
	Building No. 5 (Daemei et al. 2021)	1998	Iran	Private	3.2 × 3.5 × 2.9m each	Two rooms in a 2-storey residential building with concrete block walls, single glazed window.	Thermal performance of green wall	All mechanical devices were shut down and there were no active internal heat gain sources during test.
	FH/VC/HB buildings (Mahdavi et al. 2008)	N/A	Austria	A university, an international organization and a governmental organization	Single/double/triple-occupancy offices	48 Offices selected from three buildings, they are on different floors and facing different orientations, manually controlled external/internal shading.	Occupants’ operation of lighting and shading systems in response to irradiance in office buildings	Fan coil/ radiator units under the window for fine adjustment of temperature.
	Field Exposure of Wall Facility (FEWF) (Maref et al. 2007)	2006	Canada	National Research Council of Canadian Institute for Research in Construction	Maximum façade component size for testing: 4.8 × 2.1 m	Two storey house facility consisting of a test opening measuring 4.8 m wide by 2.1 m high on the ground floor of the west façade for testing one large sample or several small samples. Concrete wall with insulation.	Hygrothermal behavior of exterior wall assemblies	Hybrid heating (zone-controlled hydronic radiant floor heating and zone-controlled forced-air heating) Hybrid ventilation (mechanical air exchange and passive ventilation openings in the basement walls and flues)
	Controlled environmental chamber (Ko et, Schiavon, al. 2020)	2008	USA	University of California	5.5 × 5.5 × 2.5m	An office-like test room divided into two spaces by a curtain, one with windows and the other without	The influence of having a window with a view on thermal and emotional responses as well as on cognitive performance.	Air handling system.
	IEQ Lab (De Dear et al. 2013)	2013	Australia	University of Sydney	85m² (60 m^2+25m^2)	2 Test rooms sharing 1 outdoor simulation corridor; operable windows, different facade ventilation modes	Indoor environmental quality (thermal, visual, acoustic, air quality)	Fully-conditioned with variable air volume (VAV), under floor air distribution (UFAD) and chilled beams (active and passive) Natural or hybrid ventilation.
	Daylight laboratory (Bakker et al. 2014)	2014	Netherlands	TU Eindhoven	5.4 × 5.4 × 2.7m	West wall is all glazed, on the interior side equipped with a prototype automated dynamic facade in the form of nontransparent movable roller shades	Control strategies and occupant influence on visual performance of dynamic facade	Air-based heating system and electric Radiators, VAV for cooling.
	Office rooms (Bian and Luo 2017)	2017 (Lab)	China	South China University of Technology	Not specified	Two offices in perimeter zones on the 3rd floor of a six-story open-plan concrete building. One has east-facing side-lit window with manual controlled louver blinds; one has south facing with encorbelment overhung.	Visual comfort	Not specified.
	LESO Building Physics Lab (Benedetti et al. 2019)	2018 (Lab)	Switzerland	EPFL	Not specified	Two identical office rooms in a university building, south oriented and single-occupied.	Visual performance of automated shading and lighting control systems	Not specified.
	Electricity substation (Lee and Jim 2019)	2019	Hong Kong	Electricity substation	N/A	Reinforced concrete building; Climber green wall installed on the northeast oriented façade	Radiation properties and shading-induced cooling energy savings	Not specified.

In NRC-IRC team’s Field Exposure of Walls Facility (FEWF) (Figure 15) (Maref et al. 2012, Maref, Armstrong, and Rousseau 2010), a test bay facing west measuring 7.5 m wide and 3.2 m high is contained in the house. It was possible to test one large façade sample or two small samples. The test bay may be separated thermally with an insulated cavity. Each experiment required full removal of the test bay and complete reconstruction, and re-instrumentation.

Figure 15. (a) Schematic of FEWF (Janssens 2016); (b) Insulating concrete form (ICF) wall specimens before installation of siding (Maref et al. 2012).

The ARFRISOL project (Bosqued et al. 2006; Heras et al. 2006; Palero et al. 2006; Soutullo et al. 2010) (Figure 16) included 5 office buildings of more than 1,000 m². CDI CIESOL UAL (Bosqued et al. 2006; Castilla et al. 2011, 2013; Heras et al. 2006; Palero et al. 2006; Soutullo et al. 2010), CDI Fd. Barredo Asturias (Janssens 2016), and CDI PSA (Bosqued et al. 2006; Enríquez et al. 2012; Heras et al. 2006; Palero et al. 2006; Soutullo et al. 2010) were newly-built. CDI-70-CIEMAT (Bosqued et al. 2006; Heras et al. 2006; Palero et al. 2006a; Soutullo et al. 2010) and CDI-CEDER Cubo Solana (Bosqued et al. 2006; Heras et al. 2006; Palero et al. 2006; Soutullo et al. 2010) were renovations of old buildings. They were located in five different representative areas of the Spanish climate. For each building, a few representative rooms were monitored during the experiment. Most of the remaining rooms are used as offices, and a few other rooms were used to demonstrate the role of bioclimatic buildings, solar thermal and photovoltaics in saving energy in future buildings. ESP-r (Beausoleil-Morrison et al. 2008; Beyer and Kelly 2008; Clarke et al. 1999; Kelly, Tuohy, and Hawkes 2014) is similar to the ARFRISOL project. Mahdavi et al. (Mahdavi et al. 2008) converted office rooms in three different buildings into test rooms (Figure 17). FH included a variety of test room sizes, i.e., ten single-occupancy offices, two double-occupancy and one triple-occupancy offices.

Figure 16. Location of the buildings in the ARFRISOL project (Janssens 2016).

Figure 17. (a) FH, (b) VC and (c) HB (Mahdavi et al. 2008).

One on hand, embedded test rooms are particularly suitable for post-occupancy assessment (POA) or applicability tests of a façade technology for a certain architectural type; on the other hand, it put several constraints in terms of location and orientation of the room, function of the building, service system, etc. Construction time is saved, although installation of testing components and corresponding measurement equipment is still required. It is essential to understand the construction details and thermal behavior of the hosting building and calibrate the boundary condition of the test room, before new experiments are carried out. Some embedded test rooms such as in Duplex flat (Bozonnet, Doya, and Allard 2011) were on the top floor of the building to test roof components. Test rooms on the top floor have a higher solar radiation exposure and hence it is necessary to be clear about the purpose of the experiment before determining its location in a hosting building.

2.4. Real-scale test houses/buildings

Real-scale test houses/buildings are those entirely dedicated to experimental purposes. Table 5 lists 16 real-scale test houses/buildings for detailed information. They can be broadly divided into two groups according to the primary design purpose: experimental living lab and privately owned properties under monitoring. The former are real-scale laboratory houses/buildings specially designed and built for empirical studies. Examples include INCAS platform (Gouy-Pailler et al. 2011; Ibrahim et al. 2015; Spitz et al. 2012), EnergyFlexHouse (Jensen et al. 2016; Tahersima 2012), TWIN Houses (Mantesi et al. 2019; Strachan et al. 2015), Botticelli project (Causone et al. 2019; Erba et al. 2019), FLEXLAB (Jia, Pang, and Haves 2018; Lee and Hong 2019; Pang et al. 2018), etc. They could be either used to test a particular technology, or in most cases, to understand the overall performance of the house at a system level. The latter, such as Westfield house (Kelly and Cockroft 2011), the Camden Passive House (Ridley et al. 2013), Passive House Duplex (Sage-Lauck and Sailor 2014), and PH houses (Colclough et al. 2018), are existing or newly-built housing with advanced monitoring systems. Their performances are monitored while occupants actually live in. In many cases the houses were designed to satisfy a particular standard such as the nearly Zero Emission Buildings (BNZEB (Ascione, Bianco, et al. 2016; Ascione et al. 2019, 2022)), or to represent a typical housing in its located area (the affordable house (Fensterseifer et al. 2022), and they were used to characterize the performance of a state-of-the-art architectural typology or implementation of an advanced system.

Table 5. Examples of real-scale test houses/buildings.

Picture	Name of test cell	Construction time	Country	Property Owner	Size/Layout	Features	Tested systems & aspects/parameters	VAC system
	Westfield house (Kelly and Cockroft 2011)	1938-1969	UK	Private	8 Houses of 1-3 bedrooms	Double glazed windows; insulated brick walls; natural slate/concrete tile roof; constructed between 1938 and 1969	Efficiency of air source heat pumps	A retrofitted air source heat pump (ASHP) heating system comprised the ASHP unit feeding hot water radiators directly via insulated pipes running under the flooring.
	CBCC Office Building (Mejri, Del Barrio, and Ghrab-Morcos 2011)	1999	France	Confort Bois Contruction Company	110m² In total, including 4 offices, 1 meeting room, 1 kitchen, 1 bathroom ;1 rest room	Double glazed windows; massive wooden walls; 20% window-to-floor-area ratio	Building energy performance (indoor air dry temperature, total heating power, electricity power for lighting and equipment)	Heating is achieved by electrical convectors with manual regulation.
	Limbuš passive house (Mlakar and Štrancar 2011)	2006	Slovenia	Jožef Stefan Institute	113 m²	A single-family passive house; triple glazed windows; well insulated; wooden frame structure.	Overheating characterization (Indoor air temperature, passive solar gain, internal heat gain from heating systems)	A floor heating system with maximal temperature of 29◦C and maximal power of 1800 W. A ventilation system includes two infrared-based preheaters of 100 and 200 W and involves a water-to-air heat exchanger.
	INCAS platform (Favre and Peuportier 2014; Gouy-Pailler et al. 2011; Ibrahim et al. 2015; Spitz et al. 2012)	2008	France	French National Solar Energy Institute (INES)	4 Houses, 180 m² Each	Typical of French residential houses in similar shapes but different in exterior wall construction techniques (cavity wall/ brick wall…);double glazed windows;	Energy efficiency of aerogel-based insulation rendering (for brick-monomur walls); Influential parameters (sensitivity and uncertainty) on indoor air temperature (for cavity walls)	A controlled mechanical ventilation system with a heat exchanger and by-pass.
	EnergyFlexHouse (Iqbal et al. 2015; Jensen et al. 2016; Tahersima 2012)	2009	Denmark	Danish Technological Institute	216 m²	Two-storey, single-family house, nZEB, PV integrated roof,	Performance of intelligent components and systems for the future flexible energy system; advanced control of HVAC system; interaction/ influence of occupants	Hydronic central heating system, an electrically driven heat pump as the only heat source which is connected to a network of sub-floor heating pipes and radiators.
	Napevomo (Ango 2011; Rouault et al. 2014)	2010	Spain	Arts et Métiers ParisTech	N/A	Initially designed as a single-family house for “Solar Decathlon Europe” competition	Performance validation of a PCM energy exchange and storage system for air cooling	A latent heat thermal energy storage device, a fan coupled with a variable electric supply and an upstream air tank.
	TWIN Houses (Mantesi et al. 2019; Strachan et al. 2015)	2010	Germany	Fraunhofer Institute IBP	each ground floor :82 m²	Residential layout, detached, identical, located side-by-side; Double-glazed windows with electric roller blind, well insulated external brick wall.	Comparison of different heating systems; Empirical validation of building performance simulation models	The two houses are equipped with different heating systems: electric radiators vs underfloor heating.
	Welsh Passive Houses (Ridley et al. 2014)	2010	UK	RMIT University	87 m² (3 Bedrooms)+67 m²(2 bedrooms)	Certified to Passivhaus Standards Two houses built side by side differ in terms of glazed area (55% vs 22% glazed on south wall), the use of thermal stores, area of installed PV and occupant behavior (opening window)	Overall energy and carbon performance, indoor thermal comfort and air quality. Effects of occupants.	Dwelling 1: Heater battery in Mechanical Ventilation and Heat Recovery (MVHR), towel rails in bathroom. Heating straight from boiler. Dwelling 2: Heater battery in MVHR. Heating from thermal store.
	KUBIK (Chica et al. 2011; Elguezabal et al. 2020; Garay et al. 2015)	2010	Spain	Tecnalia	500m^2, 10 *10 *10m	Steel structure, totally demountable and allows reconfiguration of the scenarios at construction level, by exchanging the components of the envelope, the roof, the floors and the partitions.	Evaluation of energy performance, acoustic performance and air tightness evaluation, taking into account the holistic interaction of the envelope, the intelligent management of the climatisation and lighting systems and the supply from nonrenewable and renewable energy sources.	A hydronic system and a Variable Air Volume (VAV) system.
	NBERT retrofit test-bed (Murphy et al. 2021; NBERT 2013; O’Donovan, O’Sullivan, and Murphy 2017)	Built in 1974, retrofitted in 2011	Ireland	Munster Technological University	223 m²	Highly insulated concrete office building, contains 1 large open plan office area, 2 small cellular offices, 1 conference room;1 seminar room	Thermal comfort, ventilative and passive cooling, energy systems, and micro-grid applications	Heating is supplied by an air source heat pump while a zonal temperature controller and a thermal buffer tank are used to optimize both the energy efficiency and thermal comfort conditions of the building.
	The Camden Passive House (Ridley et al. 2013)	2013	UK	Private	101 m² 2 bedrooms	Timber structure with well insulated wall, inward-tilting triple glazed windows on south and west facades, external blind with automatic solar control	Heating season performance in terms of internal temperature and energy consumption	Ventilated by a Paul Thermos 200 MVHR unit. Space heating is provided by a 1 kW heater battery in the supply air duct of the ventilation system, complemented by heated towel rails on demand in the bathrooms.
	Botticelli project (Causone et al. 2019; Erba et al. 2019)	2013	Italy	Politecnico di Milano	144 m²	Typical single family residential house, satisfies nZEB standards high-performance brick envelope, PV integrated roof	Overall building performance in terms of thermal comfort and energy; individual technology tests (heat exchangers, energy storage systems, control strategies)	HVAC system with earth to air heat exchanger.
	Passive House Duplex (Sage-Lauck and Sailor 2014)	2014	USA	Private	145 m²+145 m²	Two mirror-image apartments, each contains 3 bedrooms;2 bathrooms; a common living room; a kitchen. Wooden structure, well insulated wall. Meet the Passive House standard	Performance of phase-change material integrated wall in terms of thermal comfort and energy	Heating and cooling for common areas are provided by Mitsubishimini-split heat pumps. Additional heating in bathrooms is provided via fan-forced, electric wall heaters. Mechanical ventilation is achieved via Heat Recovery Ventilators.
	FLEXLAB (Jia, Pang, and Haves 2018; Lee and Hong 2019; Pang et al. 2018)	2014	USA	Laurence Berkeley National Lab	370 m²	17 Interior cubicles; 10 perimeter offices	Thermal comfort; ventilation; thermal mass; control strategies	Radiant ceiling panel (RCP) system and the radiant slab (RS) system.
	Trondheim Living Lab (Goia, Finocchiaro, and Gustavsen 2015; Korsnes, Berker, and Woods 2018)	2015	Norway	Norwegian University of Science and Technology	100 m^2	Single family house, wooden-frame structure, highly insulated envelope, automated double skin (ventilated window) on south façade; PCM-based boards on roof, PV roof, solar thermal panels on south façade, automated window opening and shading control	Overall performance in terms of indoor environmental quality, energy use; effects of users’ patterns and habits, qualification of solar energy exploitation (by PV roofs and façade integrated solar thermal panels), efficiency in energy conversion and storage.	Floor heating or one high-temperature (55 °C) radiator is used for heating. Fresh air supply is managed by a compact air handling unit that integrates a heat recovery system with rotatory wheel.
	BNZEB (Ascione et al. 2022, 2019; Ascione, Bianco, et al. 2016)	2015	Italy	University of Naples	70m² (2 Bed rooms; 1 kitchen; 1 living room; 1 bathroom;1 technical hub)	Single-storey dwelling designed to nZEB; Wooden structure, double glazed window; PV roof, solar collector	Overall performance in terms of thermal comfort, energy performance, occupants’ behavior	An aerothermal heat pump and a backup DX multi-split are installed for heating and cooling. Air-to-air heat pump provides mechanical ventilation.
	PH houses (Colclough et al. 2018)	2016-2017	Ireland	Private	Each 102 m²	Three two-storey timber framed semi-detached dwellings, Built to satisfy Passive House standards, well insulated façade, triple glazed window, 5 m^2 PV on roof	Overall performance in terms of thermal comfort, indoor air quality, energy consumption, operational costs.	Heating: MVHR HP & 2 electric radiators (550 W). Ventilation system: MVHR.
	An affordable house (Fensterseifer et al. 2022)	2022	Brazil	Private	55 m²	Affordable house built with soil-cement bricks with poor thermal insulating properties, a deciduous vine used in the green façade installed on a west wall	Dynamic thermal characterization of a green façade to improve thermal performance of an affordable housing	Not specified.

Apart from CBCC Office Building (Mejri, Del Barrio, and Ghrab-Morcos 2011) and FLEXLAB (Jia, Pang, and Haves 2018; Lee and Hong 2019; Pang et al. 2018), all others simulated residential environment. This is probably because an office environment is relatively easier to characterize (with more regular occupancy, less freedom of control in building service system) and hence an embedded test-room in an office building would be sufficient. In comparison, residential housing tends to be affected by more variable occupant behaviors.

Most facilities are unique for absolute measurements, but some contain two identical houses, either situated a few meters apart like the Energy Flex Lab (Iqbal et al. 2015; Jensen et al. 2016; Tahersima 2012) and Passive House Testing-building (Mlakar and Štrancar 2011), or joint by a common wall like the 2 mirror-image apartment (Sage-Lauck and Sailor 2014), which is essentially a two-story building divided into two mirror-image apartments that share a wall on the north–south axis. These buildings may be used for comparative studies. However, when occupants are introduced, different behaviors patterns would make it rather difficult for comparative studies.

Real-scale test buildings undertake the most comprehensive tests, and often take occupancy interactions into consideration when monitoring energy consumption and indoor environmental comforts. Calibration work is needed to understand the thermal behavior of a real-scale test building, before tests of building envelope technologies can be performed. Since the building is relatively complex than test cells or embedded test rooms, a large number of sensors need to be installed for calibration and measurements. Special attention should be paid to minimize the impacts of the measurement equipment itself on the hygrothermal behavior of the test building, such as additional building thermal bridges, local thermal loads, etc. Real-scale test buildings require a dedicated site and ample research funding for design, construction, operation, and maintenance. Additionally, it also needs to involve municipal supports if residential housing were to be employed for experimental studies in many cases. These are probably why the number of real-scale test houses/buildings is significantly lower compared to the number of stand-alone test cells and embedded test rooms.

2.5. Summary

This section reviewed 82 outdoor testbeds for assessing the performance of building envelopes between 1938 and 2023. The testbeds are categorized into three major categories according to scale and form, i.e., stand-alone test cells, embedded test rooms, and real-scale test buildings. Representatives in each category are discussed in details. According to the above analysis, the advantages and disadvantages of testbeds in each category are summarized in Table 6.

Table 6. Summary of advantages and disadvantages of testbeds for assessing building envelopes.

Testbed type	Advantages	Disadvantages
Stand-alone test cells	Simple and quick construction Large flexibility in material selection, test variables (e.g., orientation), etc. Relatively simple calibration work Low cost	Boundary condition may not be a good approximation of the real-world conditions. Size may be limited and not suitable for tests with occupants
Embedded test rooms	Particularly suitable for POA or applicability test of a façade for a specified architectural type Both experiments with/without occupants may be carried out Low construction time	Low flexibility in function of building, service system, orientation, etc. Boundary conditions strongly depend on location of room in the building, and hence have large impacts on test results. Complex calibration work, particularly on existing building service system.
Real-scale test buildings	Realistic representation of indoor space and boundary conditions Both experiments with/without occupants may be carried out	Dedicated site is required High budget and longer construction period are required Complex calibration work Monitoring work may be affected by occupants, and comparative studies with occupants may be difficult

Due to the evolution of façade technologies and increasing number of performance parameters to be considered for experiments, the outdoor test facilities have gradually become more complex and diverse. PASSYS and PASLINK projects were undoubtedly exceptional pioneers for test cells, based on which more functions such as rotatable base and tiltable roof were further included later on. Test buildings are constructed from a more practical perspective. Although the outdoor testbeds share the same basic unit – a cube test room, there is a lack of link between different categories, even though test cells could potentially be accommodated by test buildings or test buildings may be decomposed into individual test cells.

3. Experimental objectives of outdoor testbeds

The outdoor testbeds have been employed for investigating the façade design parameters and building performance indicators, as shown in Figure 18. Some test facilities were constructed to test specific technologies, and others provide opportunities for testing of different technologies and systems, as shown in Table 5. Generally, the experimental work that have been done aimed to understand the underlying interactive mechanism between the parameters and indicators.

Figure 18. Classification of experimental purposes.

3.1. Design parameters

Façade geometric parameters are important design factors that may significantly affect building performance. Stand-alone test cells are most frequently used for exploring these parameters. For example, Jaber et al. (Jaber and Ajib 2011) achieved a 27.6% annual energy saving by optimizing the façade orientation, window size, insulation thicknesses, etc. MATRIX (Ascione, De Masi, et al. 2016) was employed to investigate facades components of different sizes, and the thicknesses of exterior wall. Hans et al. (Simmler and Binder 2008) showed the influence of various blind configurations, including slat tilt angle and solar reflectance of the slat surface, on the total solar energy transmittance. The calorimetric test facility (Kersken 2021) allowed components to be tested in relation to different angles of solar incidence.

As science and technology continue to develop, more and more new materials and technologies appear. The types of facades have evolved from static to dynamic, from passive to active. Outdoor test facilities have been widely used for understanding the complex effects of the physical properties of facades, including static properties such as U-value (Jiménez, Porcar, and Heras 2008; Kotsiris et al. 2012; Mandilaras et al. 2014), g-value (SHGC) (Clarke et al. 2002; Jiménez, Porcar, and Heras 2008; Ko, Oh, et al. 2020), and other dynamic properties. For instance, Baker (Baker and Van Dijk 2008) used dynamic methods to obtain the thermal and solar characteristics of building components. Dimoudi et al. (Dimoudi, Androutsopoulos, and Lykoudis 2004) tested a prototype ventilated wall coupled with dynamic insulation panels and found that the permeability of the intermediate materials must be considered when using a multi-layer component.

The interaction between occupants, façade, and building service system is another important factor that highly affect the indoor environmental quality and energy demand of buildings. Occupant interaction with building envelopes varies from as simple as opening/closing windows based on their thermal sensation, to adoption of advanced control strategies for dynamic façade technologies for improving thermal comfort, visual comfort or indoor air quality. In the study by Bakker et al. (Bakker et al. 2014), several experimenters were involved and completed the questionnaire to help understand the logic and patterns of the control-oriented user behavior. Alessandra et al. (Luna-Navarro and Overend 2021) captured the effects of facade systems on occupants environmental perception and interaction in MATELab. Ko et al. (Ko, Schiavon et al. 2020) investigated the thermal perception and emotional and cognitive impacts of having a view to the outdoors via a window in a working environment. This will not only help to better predict building performance, but also support the effective operation of building envelopes.

3.2. Performance indicators

3.2.1. Indoor environmental quality (IEQ) performance indicators

IEQ performance, which includes thermal performance, visual performances and air quality, directly affects occupants’ comfort. The indicators for the above aspects may all be assessed with outdoor testbeds.

Thermal performance can be evaluated by: (1) directly measurement of thermal/solar related properties, such as determination of U-values, g-values with the PASSYS and PASLINK test cells; (2) recording the dynamic variations of key factors such as indoor air temperature, relative humidity, air flow rate, wall surface temperature, based on which metrics such as PMV or PPD may be calculated. La Ferla et al. (La Ferla, Román, and Calzada 2020b) demonstrated that radiant glass façade technology as a heating device can improve the level of thermal comfort by avoiding radiant asymmetry through a uniform distribution of radiant temperature. In this study, envelope surface temperature, operative, air, radiant temperatures, and radiant asymmetry temperature are measured, and PMV was calculated accordingly. Nicolini et al. (Nicolini et al. 2023) analyzed the thermal performance of a façade enclosure integrated by vegetation under windy and rainy climatic conditions, by measuring the wall surface temperatures. Yang et al. (Yang et al. 2018) proposed a state-space thermal model incorporating thermal comfort and humidity for model predictive control in buildings, and its performance is evaluated through a case study on the BCA SkyLab test bed facility in Singapore.

Visual performance can be determined by: (1) measuring luminance/illuminance at different points; (2) taking circular fisheye photos, for calculating daylight glare index (DGI) and daylight glare probability (DGP). Bian and Luo (Bian and Luo 2017) recorded luminance and illuminance with meters at different points of table level around a participant, and captured circular fisheye images that equal the field of view as seen by the participant. HDR luminance image were produced, from which DGI/DGP were calculated. Benedetti et al. (Benedetti et al. 2019) developed HDR vision sensors to capture luminance maps continuously, from which vertical illuminance and DGP may be extracted.

Assessment of air quality usually includes measurements of ventilation rate, relative humidity, vapor pressure excess, and CO₂ concentrations. Examples can be found in Ridley et al. (Ridley et al. 2014) and Colclough et al. (Colclough et al. 2018). Epidemiological evidence suggests that the presence of mold growth in buildings can have a detrimental effect on the well-being of occupants. A study by Clarke et al. (Clarke et al. 1999) looked into infectious aerosols and developed a predictive tool to prevent the likelihood and extent of mold infestation, and its predictive capability was tested against monitored data and mycological samples taken from a mold-infested house.

In addition to the above-mentioned objective measurements, subjective questionnaires may be collected from participants/occupants for understanding their thermal comfort (Schweiker et al. 2012), visual comfort (Benedetti et al. 2019; Bian and Luo 2017), and satisfaction with air quality (Colclough et al. 2018).

3.2.2. Energetic performance indicators

With the aim of assessing the energy efficiency of advanced building envelopes at system/building level under real weather conditions, they are often tested using outdoor testbeds installed with HVAC systems. According to Tables 1–5, 90% of the reviewed testbeds provided specific descriptions on their HVAC systems, and the corresponding energy consumption for heating, cooling, and ventilation can be measured. The rest either intentionally excluded the effects of HVAC systems to simulate a common local situation like Bozonnet, Doya, and Allard 2011) and (Fensterseifer et al. 2022), or had research focuses excluding energetic performances like Wienold and Christoffersen (Wienold and Christoffersen 2006).

The HVAC systems are generally simpler and more flexible for single-room, multi-room, and multiple test cells, as shown in Tables 1–3. A ventilation system with heating and cooling unit, electric radiators, air conditioners, and fans were mostly adopted. Floor heating (Mohammadi et al. 2022; Soudian and Berardi 2022) and ceiling heating/cooling (Lwf-FAÇADE-facility (Laboratory for Intelligent Building Technology 2011) were also possible. More complex HVAC systems can be found in embedded test rooms and real-scale test houses/buildings. For instance, EnergyFlexHouse (Iqbal et al. 2015; Jensen et al. 2016; Tahersima 2012) adopts hydronic central heating system with an electrically driven heat pump as the heat source, which is connected to a network of sub-floor heating pipes and radiators. KUBIK had a hydronic system and a Variable Air Volume (VAV) system. CIESOL building (Bosqued et al. 2006; Castilla et al. 2013; Heras et al. 2006.; Palero et al. 2006; Soutullo et al. 2010) is installed with HVAC systems based on solar cooling using a solar collector field, a hot water storage system, a boiler, and absorption machine with its refrigeration tower. Radiant ceiling panel (RCP) system and the radiant slab (RS) system are equipped in Flexlab (Jia, Pang, and Haves 2018; Lee and Hong 2019; Pang et al. 2018). Apart from those, simpler heating devices such as electric radiators could also be used as additional heating source (Colclough et al. 2018). Therefore, with such embedded test rooms and real-scale test houses/buildings, a more realistic indoor environment can be achieved, and detailed measurements and analysis can be carried out for a specific and innovative type of HVAC system. The only concern is that if the test house/building was not initially designed solely for experimental purpose, the addition of monitoring system may encounter some constraints in installation and introduce uncertainties to experimental data. For example, in the Camden Passive House (Ridley et al. 2013), the monitoring system measures the domestic hot water consumption and the heat input to the cylinder from the solar thermal system, but the heat input to the cylinder from the boiler and the cylinder temperature are not directly measured. Similarly, it would have been ideal to directly measure the total heat output of the boiler, in order to correctly calculate the boiler efficiency. However, the installation of such a monitoring system to the Viessmann system post installation would be difficult and compromise the warranty.

HVAC systems may play different roles in helping assess the performance of building envelopes. Firstly, they can be used to simulate internal heat gain when occupants are absent. For instance, in COOLSKIN test facility (Bröthaler et al. 2021) a heat load of 300 W was activated to simulate the internal thermal loads. Secondly, they keep room air temperature constant to deactivate thermal mass, like in PASSYS project. Thirdly, they are operated according to rule-based control manners and used to measure energy consumption. Dimoudi et al. (Dimoudi, Androutsopoulos, and Lykoudis 2004) and Shahrzad et al. (Soudian and Berardi 2022) evaluated the cooling performance of cool selective coating and ventilated facades, respectively. Lee et al. (Lee, DiBartolomeo, and Selkowitz 2006) demonstrated the first generation of prototype electrochromic window and daylighting control system, and monitored the cooling load and visual comfort of this technology. Fouthly, they are used to study the combinatorial energetic and comfort performance of advanced building envelopes with occupant interaction and advanced HVAC and lighting control strategies. (Yang et al. 2018) and (Castilla et al. 2013) both investigated thermal effects of model predictive control strategies in building automation and control. Mahdavi et al. (Mahdavi et al. 2008) observed user control actions pertaining to lighting, and assessed lighting energy saving potential due to consideration of occupancy and behavioral patterns in office buildings.

From the energy storage point of view, the study of material properties is also very important. Ibrahim et al. (Neya et al. 2021) tested the relationship between the thickness of the brick materials and thermal mass. Umberto et al. (Berardi and Soudian 2023) studied a radiant ceiling panel system with thermal energy storage in BeTOP. (Rouault et al. 2014) conducted a study on a low temperature phase-change materials (PCMs) thermal energy exchange and storage system. Sage-Lauck et al. (Sage-Lauck and Sailor 2014) found that installation of the PCMs had a positive effect on thermal comfort. Marine et al. (Auzeby et al. 2016) demonstrated the potential contribution of PCMs in reducing overheating risks.

With the help of outdoor testbeds, solar thermal collector and PV for energy harvesting were tested to understand the total energy flow through and captured by building envelopes. Heras et al. (Heras et al. 2001) developed optimal solar collectors for building integration in ARCHINT project. Experiments have been carried out With KUBIK (Garay et al. 2015) on solar thermal collector on facades and obtained a patent for a Passive Solar Collector Module for building envelopes (09812437.3, EP 2520 870 A1). García-Gáfaro et al. (García-Gáfaro et al. 2022) demonstrated the potential of one particular photovoltaic forced ventilated façade (PV-FVF) in latent thermal energy storage (TES). Jakub Čurpek and Miroslav Čekon (Čurpek and Čekon 2020) integrated a ventilated BiPV façade system with PCM, and investigated its influence on the thermal and energy performance. Other tests of photovoltaics on facades (Bröthaler et al. 2021; Dehra 2019) were also beneficial for understanding the integrative mechanism of PV and the entire façade system, and increasing the energy-efficiency of buildings.

4. Materials and construction

Research on facades carried out with stand-alone test cells, embedded test rooms and real-scale test buildings have shown different emphasis in terms of experimental scope. According to Tables 1–5, stand-alone test cells are mainly oriented toward component and/or room level, while embedded test rooms and real-scale test buildings are more oriented toward room and/or building level, and provides more opportunity to take into account effects of occupants’ behaviors. This difference in a way affects the choice of materials and construction techniques.

Stand-alone test cells were usually designed as a simple chamber or their aggregation. The size of a single test room is of the smallest suitable dimension for the activities of experimenters. More than 90% of the 52 reviewed stand-alone test cells are constructed of light steel, timber or light steel-wood construction. For example, SCUT test cells and GESLAB test cells (Alonso et al. 2016; La Ferla, Román, and Calzada 2020) are light steel structures. LABIMED test cells (Romano, Gallo, and Donato 2021; Luible et al. 2018; Alcamo and De Lucia 2014) are made of steel and wood. Such structures are economical, light in weight, require short construction period on site. Most test cells are placed on platforms with no surrounding shade, such as in an open place like the Smart TinyLab (Mohammadi et al. 2022), or on the rooftop of an existing building like the SCUT test cell (Lu et al. 2019). For the latter case, a light weight structure is necessary, because usually the exiting building roof was not initially designed to take additional large concentrated load. Due to their site and experimental purpose, the stand-alone test cells are usually exposed to the exterior environment entirely, which brings a challenge for maintenance especially when the major materials are steel or timber. No protection/maintenance routines were mentioned in the literature so far, which is probably the reason why around 25% of the stand-alone test cells were only reported once in a study.

Embedded test rooms typically maintain the structure and material of the original hosting building. According to Table 4, at least half of the reviewed embedded test rooms are hosted by concrete buildings such as the FEWF (Maref et al. 2007), FH/VC/HB buildings (Mahdavi et al. 2008), and Building No. 5 (Daemei et al. 2021). Reinforced concrete structures have stronger weather resistance and require less maintenance work. These test rooms also offered superior opportunities to evaluate the façade performance in buildings with high thermal mass. Since host buildings usually have already been in use for some time, careful inspection should be done on its current condition, including insulation, infiltration, possible damaged connections and so on, before new experiments are performed. Embedded test rooms show less flexibility compared to test cells, mainly because of the constraints brought by the host building that the structure/wall of concrete cannot be easily altered. Out of the 12 test rooms listed in Table 4, only 4 was reported to have changes on their envelopes, i.e., for the Duplex flat (Bozonnet, Doya, and Allard 2011) a cool selective coating was applied on the roof, and tests were carried out in order to quantify the improvement in the indoor thermal environment; FEWF (Maref et al. 2007) hosted a test room with one façade interchangeable; Building No. 5 (Daemei et al. 2021) and Electricity substation (Lee and Jim 2019) both were installed with green walls. All other test rooms have operable or switchable facades/shadings that were measured as is.

According to Table 5, around 2/3 of the real-scale test buildings mainly adopt timber/steel structure, and the rest are masonry/concrete structures. For test houses that are initially designed for experimental purposes, steel or timber structures requires a shorter construction cycle and offers higher reconfigurability. For example, KUBIK (Chica et al. 2011; Elguezabal et al. 2020; Garay et al. 2015) adopted steel structure that was totally demountable and allowed reconfiguration of the scenarios at construction level. The components of the envelope, the roof, the floors and the partitions can all be exchangeable. For test houses/buildings that offer both residential and experimental use, the choice of construction materials is mainly dependent on the architectural typology and its construction time. For instance, the Westfield house (Kelly and Cockroft 2011) constructed between 1938 and 1969 were masonry structures, while the Camden Passive House (Ridley et al. 2013) adopted timber structure. These test houses/buildings either have refurbished or new high-performance envelopes that were measured as is without changing.

5. Design, construction, and use of future outdoor test facilities

5.1. Standardization of design and use

Through previous study on existing outdoor testbeds, it is not difficult to find that the test facilities can be used either as a “customized” experimental instrument for a particular test, or as a “universal” instrument to provide experimental testing of the performance of various facade materials and components under different climatic conditions. Both from an economic point of view and for the purpose of continuation and expansion of research activities, the future outdoor testbeds should aim toward the latter. The PASSYS (Aleo et al. 2001; Assimakopoulos et al. 2007; Hahne and Pfluger 1996; Van Dijk and Van der Linden 1993) and PASLINK (Baker and Van Dijk 2008; Dimoudi, Androutsopoulos, and Lykoudis 2004; García-Gáfaro et al. 2020; Goethals et al. 2012; Kotsiris et al. 2012; Leal and Maldonado 2008) projects in Europe are good examples. They provide standards not only in facility design but also in testing procedures. However, after the two projects, few projects continue with this methodology. More customized test cells were constructed at a particular location by a specific institution. This makes the measurement result valid only for one climate type, and less possible for cross-comparison between different climates.

In terms of testbed types, stand-alone test cells and real-scale test houses/buildings are more suitable for standardization design than embedded test rooms, because important factors such as boundary conditions and orientations of embedded test rooms are largely dependent on the hosting building that is difficult for adjustment. Stand-alone test cells require a relatively lower budget than real-scale test houses/buildings. Single and multiple test cells show the highest flexibility, since additional ones may be quickly constructed as further study requires. Multi-room test cells with an auxiliary zone for the service system are more comprehensive, because they satisfy the need of both absolute and comparative experiments. Standardized real-scale test buildings, such as Flexlab (Jia, Pang, and Haves 2018; Lee and Hong 2019; Pang et al. 2018), are ideal for obtaining results that best reflect the real performance, because more advanced HVAC systems may be installed and measurement can be carried out not only at the component level, but also at the system/building level. No matter which type of testbeds is chosen, functions should be well defined and a consistent design and construction process should be developed accordingly. Geometric parameters, layout of the test rooms and auxiliary space, materials, connections, and service systems should be identical to make sure the experimental results are comparable among testbeds located in different places. Corresponding testing devices and measuring procedures should also follow the same guideline.

At the same time, scientific alliances can be formed between research institutions in different countries to help the promotion of standardized experimental facilities. Experimental results can be collected through a common database for cross-comparison and complementarity on the basis of data sharing and even deeper mining of technical data.

5.2. Flexibility enhancement based on prefabricated modular design strategies

The research for the construction of the outdoor laboratory takes place in the era of rapid development of prefabricated assembly technology, when there is a continuous and dynamic interaction between diverse requirements and standardized technical routes. As a mockup facility, the role of a future outdoor testbeds is to undertake the most diverse research tasks, including assessing complex and advanced building envelope systems in response to various modes of energy use. Most existing facilities only have one elevation or part of it (e.g., window) that is reconfigurable, while all other parts are fixed. Particularly, few has possibility of various testing roof components. Adaptability of test room dimension is also limited. In addition, the test facilities are all designed as is, without possibility of extension. The best design strategy is to develop modular components that is efficient for installation, disassembly, replacement, and reconfiguration It may also facilitate obtaining primary products.

This may be achieved by two levels of modules, i.e., component modules and space modules, which have already been successfully applied in small scale prefabricated construction such as Kiosk K67 (Figure 19) (Stierli, Kulić, and Klarin 2018). Component modules provide the flexibility of changing tested modules, connections, wall construction, and the shape and dimension of a single test room. Space modules provide flexibility in the layout and relationship between test rooms. In doing so, components are easily replaceable and spaces are easily combinable, which is a viable way to obtain superior adaptability of the laboratory space and its envelope systems. Single-room test cells may be conjugated to form multi-room test cells, which may then be upgraded to outdoor test buildings to satisfy experimental objectives of different levels of complexity. It should be noted that apart from composition of test room modules, compatibility of auxiliary components such as staircases for multi-story test buildings and HVAC systems should also be considered.

Figure 19. Decomposition and composition of Kiosk K67 (Stierli, Kulić, and Klarin 2018).

5.3. Establishment of long-term measurement mechanism

Most studies with outdoor testbeds were conducted in days, weeks or months, while only a few real-scale test houses ran for more than a year. Welsh Passive Houses (Ridley et al. 2014) and Camden Passive House (Ridley et al. 2013) were monitored for 2 years for understanding the performance of newly built houses in Welsh and London according to Passive House Standards. An IEQ repository was developed that involved over 20 dwellings to provide valuable insights into the long-term performance of the dwellings in both the Republic of Ireland and Northern Ireland (UK) (Colclough et al. 2018). In fact, short-term monitoring, or measuring of the first 1–2 years of newly built houses are not always sufficiently representative of the long-term building performance. The quality of maintenance, degradation of construction materials and building service systems, replacement of certain components, change of occupancy or occupants’ behavior all possibly affect the performance of a building in its service life. In addition to short-term characterization, long-term measurement mechanism is required to be established for fully understanding the individual performance of building envelopes and its combinatorial effects on the entire house/building at the system level.

6. Conclusions

This paper analyzed 82 outdoor testbeds for assessing the performance of building envelopes between 1938 and 2023. It aims to provide a comprehensive review and summary of existing outdoor test facilities, and to propose useful suggestions for future design and use of them.

The testbeds are categorized into three major categories according to scale and form, i.e., stand-alone test cells, embedded test rooms, and real-scale test buildings. Representatives in each category are discussed in details, and the main features are compared and summarized systematically. The functions of testbeds are analyzed according to the dedicated experimental purposes. The historical evolvement of the testbeds was discussed to identify the main trends in the development of outdoor testbeds. Different construction materials and structures were also compared to identify suitable choices for testbeds of different scales and research objectives.

On the basis of previous review and analysis, recommendations are proposed for future design, construction and application of outdoor test facilities for assessing performance of building envelopes. Both from an economic point of view and for the purpose of continuation and expansion of research activities, future outdoor testbeds should develop toward a more universal route with diverse testing capabilities. Standardized design, construction, maintenance, and testing procedures should be established to ensure the long-term validity and cross-comparability of studies based on testbeds located in different places. Prefabricated modular design strategies can be employed to enhance the flexibility in the scale and functions of testbeds. A long-term testing mechanism is essential for better understanding of the performance of building envelopes at system level.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study was funded by and the National Natural Science Foundation of China [Grant No. 52278039] and the Research Fund of Tongji Architectural Design (Group) Co., Ltd. [2023J-JB04]

Notes on contributors

Qian Jin

Qian Jin, PhD, is an Associate Professor. Qiuting Sun, BA, is a Master Student. Gang Meng, PhD is an Associate Professor.