Energy and climate simulation in the Upper Lawn Pavilion, an experimental laboratory in the architecture of the Smithsons

Research Article Architecture and Human Behavior E-mail: [email protected] Energy and climate simulation in the Upper Lawn Pavilion, an experimental laboratory in the architecture of the Smithsons Rocío Escandón (), Juan José Sendra, Rafael Suárez Instituto Universitario de Arquitectura y Ciencias de la Construcción, Universidad de Sevilla, Av. Reina Mercedes 2, Seville 41012, Spain Abstract This article aims to provide a critical analysis of the work of the Smithsons, with a focus on its interpretation of architecture from the standpoint of environmental commitment, an aspect that is barely touched upon in the extensive literature analysing their work. The main aim is to reveal the strategies of environmental conditioning and the energy behaviour of their major work both from an architectural and environmental standpoint, as well as that of their own home and experimental laboratory: the Upper Lawn Pavilion. These features make it a building of great heritage value, classified as a Grade II listed building according to the Statutory List of Buildings of Special Architectural or Historic Interest. Energy models reproducing the case study (in its original construction, occupancy and thermal conditions) have been simulated using DesignBuilder software in order to carry out a quantitative assessment of the house’s environmental conditions, perceptively described by Alison Smithson in her diary. Indoor thermal conditions obtained from the energy simulations were studied, identifying the environmental benefits and deficiencies caused by the strategies applied by the Smithsons to their pavilion. It is concluded that its environmental behaviour is far from the current standards of adaptive comfort. Keywords energy simulation, Smithsons, solar pavilion, climate and architecture, building and environment Article History Received: 9 March 2014 Revised: 1 August 2014 Accepted: 5 September 2014 © Tsinghua University Press and Springer-Verlag Berlin Heidelberg 2014 1 Introduction Following the Second World War the purist propositions of the Modern Movement were transformed into a standard style, incapable of incorporating local tradition (Pallasmaa 1983). The architecture of the International Style, rationalist and functional, did not take climate or solar movement into account. It neglected local conditions and favoured habitability and environmental control over the implementation of resources which mechanised the environment. In the 1950s a group of young architects, grouped around Team X, proposed a review of the principles of the Modern Movement (Risselada et al. 2011) seeking an architecture which establishes a relationship with the environmental context. Alison and Peter Smithson were important figures within Team X and their work invites reflection on the connection of architecture with the environmental context: “Our intention was to shift architecture towards particularity; its forms to arise from attention to persons and place.” (Vidotto 1997) The Smithsons incorporated a new awareness of climate and energy resources into their architecture (Smithson and Smithson 2005) by integrating passive conditioning in the architectural project (Wong et al. 1994). This was reflected in most of their drawings, in their solar studies, as well as in their writings: “The stress is on the needs for immediacy of response and reaction to the changeable weather of England; the almost constant need for full or partial weather protection […] Northern Europe involves us inevitably in sun acceptance, amelioration of climate and, above all, of exclusion of rain.” (Smithson and Smithson 2001) To verify the correct operation of their innovative passive conditioning proposals they could only use their own experiments since there was no modern energy simulation software at that time. The history of the architecture from the Modern Movement usually found in bibliographies focuses on analysing the work from formal, functional or ideological perspectives, but few authors approach this analysis from BUILD SIMUL DOI 10.1007/s12273-014-0197-0 Escandón et al. / Building Simulation 2 List of symbols Tco indoor comfort operative temperature (℃) Text,ref outdoor reference temperature (outdoor temperature at the exact same point in time as the desired calculation of Tco) (℃) the perspective of the creation of habitable surroundings. From the 1960s onwards major authors, like Olgyay (1963), Banham (1969), and Givoni (1969), did recognise this. More recently other authors, such as Porteous (2002) and Hawkees (2008, 2012) have analysed important architectural work from an environmental standpoint. However, these works are not scientific studies but are limited to descriptive non-quantitative studies of the relationship between climate and architecture. There is little scientific work analysing important examples of the history of architecture from an environmental standpoint, most notably Requena-Ruiz (2012a, b), Taranto and Kacel (2013). Requena-Ruiz carried out a historical, but mostly technical and scientific analysis of Le Corbusier’s later work. The author used the potential of modern simulation techniques to carry out a quantitative analysis of the influence of Le Corbusier’s innovations, such as the “brise-soleil” and the “aerateur”, on environmental conditions (humidity, ventilation, solar exposure, radiation and luminance). The aim of this article is to complete the critical analysis already published on the Smithsons’ work (Hart 1994; Krucker 2002; Van Den Heuvel et al. 2004), and to do so from an environmental perspective which has barely been touched upon until now. This would reveal the energy behaviour of a specific case study through the simulation of energy models reproducing it and carrying out a quantitative assessment of their environmental conditions and the repercussions of the innovations suggested by the Smithsons for the passive conditioning of their dwellings. The Upper Lawn Pavilion in Wiltshire, Great Britain, built in 1962, was chosen as a case study given that it is one of their most representative works both in architectural and environmental terms given that it is “the only building designed, built, and inhabited by the Smithsons for their own use on weekends and holidays, and one of their few realised domestic projects” (English Heritage 2014). During the 20 years the Smithsons lived there it became a laboratory in which to test different passive conditioning strategies: “To find out what it is like to live in a house in England all the year round which presents glass walls to entire South, East and West, but a solid Wall to most of the North face (in conjunction with insulated roof), i.e. if it is a feasible proposition that solar heat can be obtained most of the year round and that its build up can offset heat loss.” (Smithson and Smithson 1986) The Upper Lawn Pavilion has been widely documented and exhaustively photographed by its authors, most notably in the diary kept by Alison Smithson during their stays at the pavilion (Smithson and Smithson 1986), which provides important information for the development of energy simulations on the conditions for operation and use of the pavilion. This case study is made all the more interesting by its emblematic status as one of the most representative pieces of neobrutalist architecture (Smithson et al. 2005), and by its cultural heritage legacy since it is a Grade II listed building. This means that it is “a historically or architecturally significant building that is of special interest, warranting every effort to preserve it”, according to the Statutory List of Buildings of Special Architectural or Historic Interest (English Heritage 2014). 2 Case study: The Upper Lawn Pavilion In 1959, Alison and Peter Smithson envisaged the project for the Upper Lawn Pavilion as a small holiday home where they could enjoy the passing of the seasons and be the subjects of passive conditioning strategies, different habitation models and new building materials and solutions. They designed a simple “climate home” powered by solar energy (Smithson and Smithson 1963, 1986). Its construction began in 1961, with the demolition of most of the existing eighteenth-century building, although the north-facing wall and part of the west wall were preserved and used to support the new structure (Krucker 2002). The Upper Lawn Pavilion is no longer conserved in its original condition since the Smithsons sold it in 1981 and it was neglected by the new owner. In 2002 it was restored to its original appearance by Sergison Bates Architects, but the material and environmental conditions were not restored, and so cannot be established through in situ measurements. 2.1 Location and description of climate The Upper Lawn Pavilion is found at Fonthill Abbey, in Tisbury in the southwest English county of Wiltshire. The exact geographical position of the pavilion is latitude 51°3ʹ36˝N, longitude 2°6ʹ36˝ W, and an altitude of 131 m. Escandón et al. / Building Simulation 3 The climate in Wiltshire is oceanic, with warm summers, cold winters and rain all year round. Average annual thermal oscillation is medium to low. The weather is highly variable and can change drastically from one day to the next. Temperatures are low in winter, with high humidity, while summers are warmer and drier, with occasional storms. Autumn and spring are usually mild, although with relatively high temperatures and occasional showers in spring and heavy rain in autumn (Table 1). 2.2 Envelope and construction systems The Upper Lawn Pavilion has a wood and concrete “balloon-frame” structure resting on the existing stone wall. Its envelope combines large surfaces of glazing (permeable to solar radiation and with low thermal inertia) (Table 2), with a pre-existing north-facing solid stone wall (opaque and with high thermal inertia), and lightweight enclosures and roofs with thermal insulation (Fig. 1), as is reflected in the plans and the building and thermal characterisation of the envelope (Fig. 2). 2.3 Passive conditioning strategies The Smithsons incorporated a series of passive conditioning strategies into the pavilion:  In opposition to traditional English cavity walls they incorporated low-cost lightweight coverings and enclosures, which were thermally insulated. Insulation materials had barely been incorporated into housing designs before the 1960s.  In order to capture solar energy they built the house with entirely glass facades which were oriented to the South, East and West, generating a “greenhouse effect”. It provided significant solar gains in winter, in combination with the pre-existing north-facing solid stone wall which accumulated heat thanks to its considerable thermal mass. For protection from excessive solar radiation in the summer, the Smithsons installed full height internal roller blinds on the south-facing glazing on the first floor.  The house was constantly connected with its surroundings as the ground floor could be opened up almost completely to the garden (using an enclosure composed of folding Table 1 Climate data for Wiltshire, period 1961–2010 (Met Office 2014) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year Average max temp. (℃) 6.8 7.2 9.8 12.7 16.2 19.3 21.6 21.2 18.4 14.4 10 7.5 13.8 Average min temp. (℃) 1.1 0.9 2.3 3.9 6.9 9.6 11.7 11.6 9.7 7.1 3.6 1.8 5.9 Average days of air frost 11.3 11.4 7.1 3.4 0.3 0 0 0 0 0.8 5.6 10.5 50.6 Average hours of sunlight 58 77 118 168 205 213 217 199 150 110 75 56 1646 Average rainfall (mm) 73.7 51.1 58.7 47.7 55.2 50.8 45 54.1 59.1 73.6 76.8 80.7 727 Average days of rainfall ≥ 1 mm 12.1 9.3 10.8 9.3 9.7 8.6 7.9 8.6 8.9 11.1 11.5 11.9 120 Table 2 Geometrical data of Upper Lawn Pavilion Orientation Height (m) Built surface (m2) Interior volume (m3) Facade surface area (m2) Glass surface area (m2) Percentage of glass surface (%) Percentage of glass surface — total enclosure (%) N 19.67 1.37 6.96 S 19.67 10.29 52.31 E 10.11 4.52 44.71 Ground floor W 2.15 35 75.25 10.11 4.52 44.71 34.75 N 24.41 6.44 26.38 S 24.41 14.12 57.85 E 12.02 6.87 57.15 Top floor W 2.30 35 80.50 12.02 7.21 59.98 47.54 N 44.08 7.81 17.72 S 44.08 24.41 55.38 E 22.13 11.39 51.47 Pavilion W 5.26 70 155.75 22.13 11.73 53.00 41.79 Escandón et al. / Building Simulation 4 glass panels that could be rotated 180°) when the outdoor temperature was pleasant and providing natural ventilation whenever necessary. The pavilion could therefore be adapted to occupancy patterns, which varied according to the time of year, opening up the ground floor to the garden or closing it off depending on the weather (Fig. 1). Fig. 1 Upper Lawn Pavilion 2.4 Thermal conditioning systems The Upper Lawn Pavilion had no active thermal conditioning system, except for a small coke-burning stove connected to the chimney of the original cottage and a portable electric convector heater, later to be supplemented by a second one (Hawkes 2012). In winter this electric convector heater complemented the heat gains from the large glass panes, compensating for the heat losses resulting from the low exterior temperatures. There was no active cooling system. As regards natural ventilation, although the carpentry on the ground floor allowed major air renovation, the top floor only had two small usable windows on the east- and west-facing facades. These barely represented 3.5% (1.16 m2) of the total openings of the top floor, which meant there was little chance of natural ventilation of this space, given the dominant southwest winds. Fig. 2 Upper Lawn Pavilion. Floorplans and table of envelope construction and thermal characteristics Escandón et al. / Building Simulation 5 3 Description of the methodology The methodology applied examined the interesting architectural work carried out by Alison and Peter Smithson, specifically their preoccupation for environmental conditioning, with a view to carrying out a detailed study of their passive conditioning strategies and the adaptation of their architecture to the climate in which it is found through the energy and climate analysis of their most experimental work: the Upper Lawn Pavilion. Several energy models were built reproducing the Upper Lawn Pavilion. Their simulations provided an energy and climate characterisation of the pavilion, quantifying its monthly, daily and hourly evolution. The study of solar exposure and the energy and climate simulation were carried out simultaneously. 3.1 Study of solar exposure The study of solar exposure in the Upper Lawn Pavilion was carried out with Autodesk Ecotect Analysis software [version 5.6] (2014) which provided an easy way to visualise solar incidence and radiation on the enclosures, as well as showing the solar position and the relative trajectory with respect to the model, at any given date, time and location. For this purpose, in addition to considering the building itself, the model took into account elements providing shade which the Smithsons had already taken into account in the design stage: the pre-existing wall and the large deciduous tree to the northwest, outside the plot (Fig. 3). The composition of the envelope and internal loads were not taken into account as they have no effect on the study of solar exposure and shade. The number of daily hours of sunlight inside the pavilion, solar exposure of major elements accumulating heat and the shade projected were all analysed. However, in order to Fig. 3 Upper Lawn Pavilion solar exposure model (Ecotect) calculate solar gains per glass surface this solar study needed to be combined with an energy and climate analysis. 3.2 Energy and climate simulation The energy and climate simulation was carried out using DesignBuilder software [version 2.3.6.005] (2014), with a 3D interface that makes use of thermal simulation engine EnergyPlus to obtain precise data on interior temperature and annual, monthly and hourly energy demands. EnergyPlus is the US DOE software for the modelling and dynamic simulation of phenomena relating to the environmental and energy behaviour of buildings, both mechanical (with HVAC systems) and passive (only harnessing natural resources such as the wind and solar radiation). For the construction of the energy models of the case study represented (Fig. 4), the following starting points have been proposed (O’Brien et al. 2011):  The pavilion has been considered as a single thermal zone with two adjacent and communicating areas, in order to carry out separate analyses of the behaviour of the ground and top floors.  The environmental conditions affecting the sun exposure of the pavilion have been assessed and therefore included in the models, the shade projected, both by the building and other surrounding elements such as vegetation and neighbouring walls.  The conditions for use and operation defined in Table 3 were applied homogeneously to both areas considered in the pavilion, given their spatial continuity. The climate data used for the energy simulation of the case study were taken from the Boscombe Down military airport weather station, 30 km from Tisbury. These data are included in the EnergyPlus database. As regards this case study, given that from 1981 onwards the pavilion was no longer preserved in the same way as it was built by the Smithsons it was not possible to carry out in situ measurements for the calibration of the simulation model. For this reason, in order to generate the protocols for occupancy, lighting, air tightness and natural ventilation (Table 3) associated with the lifestyle of the Smithsons, and Fig. 4 Upper Lawn Pavilion model (DesignBuilder) Escandón et al. / Building Simulation 6 in order to contrast the data obtained in the energy and climate simulation of the pavilion, this article is based mainly on the publication “Upper Lawn: Folly solar pavilion” (Smithson and Smithson 1986). As well as photographs of the pavilion and surroundings taken by the Smithsons themselves, this text includes the diary kept by Alison Smithson during their stays at Upper Lawn, from June 1962 to December 1981. In this diary Alison provides information that is essential to the development of this research on how they perceived the climate every day they spent there, the ways in which the pavilion was used depending on the season, the periods spent there, and the environmental problems and benefits of the layout of the house. It was generally occupied by the Smithsons and their three children at weekends and during the holidays, with a different use in summer and in winter. “This a story about a period of life […] it is a romantic vignette of a rural play-life of week-end hermits, in a hermitage that is an unassuming permanent-tent […] the property had been purchased for use the weekend after school […] Ground floor fully opened: August, 1963 […] Corner doors open: the Sunday breakfast table with long, high, children’s bench (later replaced by three more Van Gogh chairs): summer, 1971.” (Smithson and Smithson 1986) These data were used to build a model that aims to simulate and characterise the behaviour of the pavilion during the time it was inhabited by the Smithsons, in order to establish the thermal conditions occurring inside. The pavilion was originally built with no active HAVC system — it only had local heating provided by an electric convector heater — so that the interior thermal conditions were analysed in free evolution. An analysis was carried out on the hourly evolution of operative temperature in free evolution, during a typical day for each month. One Saturday of each month, with the environmental conditions that were most representative of that month, was established as the typical day as it was occupied for 24 hours. The outdoor environmental conditions were analysed for all the Saturdays of every month. Among these, the Saturday presenting the median outdoor temperature values from the total set of values for each month was considered to be the most representative. Through this methodology, more scattered and therefore less representative values were ruled out. For the summer period the operative temperature was previously analysed without introducing ventilation. This was to establish the timetable during which it is necessary to open the glass surface on the ground floor (Table 3), that is to say, when indoors the operative temperature exceeds 25℃. Based on this premise the energy model was simulated activating the maximum rate of natural ventilation permitted by the software (18 ACH) during the established timetable. The maximum ACH value permitted by the simulation software is established to try to emulate as faithfully as possible the behaviour of the ground floor of the pavilion when all the glass panels (15.9 m2) were open. The timetable bands inside/outside the operative temperature bands associated with adaptive comfort were also analysed. While taking into account that with the Table 3 Use and operational conditions. Activity protocols used to calculate demand (time ranges unspecified in the table correspond with 0% activity) Schedule Activity Value General Holiday (July) Occupancy 0.071 person/m2 126 W/person Friday: 18:00 to 24:00 Saturday: 00:00 to 24:00 Sunday: 00:00 to 18:00 100% 100% 100 % Monday – Sunday: 00:00 to 24:00 100% Lighting 5 W/m2 Friday: 18:00 to 22:00 Saturday: 07:00 to 22:00 Sunday: 07:00 to 18:00 100% 100% 100% Monday – Sunday: 07:00 to 22:00 100% Air tightness 1.5 ACH 00:00 to 24:00 100% 00:00 to 24:00 100% Natural ventilation 18 ACH Saturday (August): 11:00 to 18:00 Sunday (August): 11:00 to 18:00 100% 100% Monday – Sunday: 10:00 to 20:00 100% Escandón et al. / Building Simulation 7 construction of the Upper Lawn Pavilion the Smithsons partly renounced the standards of comfort of domestic architecture of their time (Smithson and Smithson 1994), this study aims to analyse the behaviour of the pavilion and carry out a comparison in relation to the present context of environmental culture. Establishing a comfort band was considered a simple approximation to the issue of thermal comfort as it is a somewhat subjective concept that is hard to assess. In order to establish this comfort band the ASHRAE adaptive index, one of the most widely used, was applied, following Eq. (1) (de Dear and Brager 1997; Ferrari and Zanotto 2012): Tco = 0.31·Text, ref + 17.8 (1) An acceptability range, corresponding to 80% of satisfied occupants and defined by a temperature interval of ± 3.5 ℃, was applied. When analysing the results obtained a distinction was made between the space on the ground floor and on the top floor since different types of behaviour were observed on these two levels. 4 Analysis and assessment of the results Alison and Peter Smithson considered the Upper Lawn a small “solar pavilion” and for this reason one of the first steps of this research was to carry out an analysis of solar exposure and radiation. Initially, solar trajectory was analysed to detect any possible shade from pre-existing elements throughout a summer and winter day (Fig. 5): as they are located to the north, the stone wall and the large beech tree barely project shade on the pavilion, except in the early morning in summer months (from April to August, between 4:00 and 6:00). A study was carried out of the incident solar radiation on the different facades throughout a typical day in the two most significant months of the year, January and July (Fig. 6). This revealed the solar gains on the glass surfaces of every orientation. Fig. 5 Shade projected every 60 minutes, during the time range of 4:00 to 20:00 h Fig. 6 Incident solar radiation (Wh) on facades The north-facing facade barely receives incident solar radiation, as it is shaded most of the time, both in winter and summer. However, the south-facing facade receives a great amount of incident solar radiation, particularly in summer, and thanks to its glass surface and the incidence of solar radiation it becomes the main surface for solar capture in the pavilion. During the winter the effects of the low angle insolation on the east and west facades in the morning and evening are also considerable in the solar pavilion (Fig. 7). The solar study was completed with an analysis of the number of daily hours of solar exposure inside the pavilion throughout the year (Fig. 8). At the latitude of Upper Lawn sunrise at the summer solstice is at approximately 03:55 GMT, with the corresponding sunset at 20:05. In April/August these times are 05:00 and 19:00 GMT respectively. The sunrise in the winter solstice is at approximately 08:05 GMT, and the corresponding sunset is at 15:55. A high number of hours of sun was observed and was slightly more noticeable on the top floor than on the ground floor due to its larger glass surface (Table 2). However, this study did not allow us to calculate the balance between solar gains during the day and thermal losses at night. Therefore, this analysis was complemented with the energy and climate simulations developed using DesignBuilder. The hourly evolution of the operative temperature in free evolution and on the solar gains through exterior glass surfaces on typical days in January and July was analysed. Fig. 7 Shade projected every 30 minutes in January Escandón et al. / Building Simulation 8 Fig. 8 Daily hours of solar exposure inside the pavilion Solar gains were quantified (kW), as was the increase in indoor temperature resulting from these gains. This increase in indoor temperature was quantified by comparing the results of the simulation of the original prototype model, with another model with light covers placed on all glazed surfaces. All this was to verify and quantify the energy behaviour of the pavilion and detect any possible interior thermal comfort issues. The aim of this study was to verify whether, as the Smithsons expected, solar gains could be obtained all year round to make up for thermal losses, even in winter. The analysis confirmed the existence of major solar gains in the pavilion, even in the colder periods, due to the high percentage of glass surfaces (Fig. 9). On the day analysed (January 12) solar gains on the top floor reach 6.5 kW, causing a temperature increase indoors of up to 13.5℃ in the middle hours of the day. However, from the analysis of the interior thermal behaviour of the pavilion it was deduced that the solar gains during the day did not make up for the thermal losses occurring at night (Fig. 10). In winter, during the hours in the middle of the day there is an important increase of operative temperature, with a thermal leap on the top floor between maximum exterior and interior temperature of +16℃ in January (Fig. 10), reaching operative temperatures around 14℃ on the ground floor and 23℃ on the top floor, as there are greater solar gains. This figure tells us that comfort temperatures were reached during several hours a day without any active heating system on the top floor. Despite this the temperature falls drastically as soon as solar gains start to disappear, with night-time minimum temperatures of 5℃ on the ground floor and 4℃ on the top floor, which shows that the losses through the glass surfaces and the carpentry, due to its lack of airtightness, are noticeably greater than the heat gains accumulated throughout the day. Fig. 9 Hourly evolution of solar gains through glazed surfaces, and the temperature increase they provoke inside the pavilion (January 12) Fig. 10 Hourly evolution of exterior and operative temperature (January 12) Comparing the results obtained for both floors of the pavilion, it can be observed that the top floor has little thermal inertia so the solar mass barely accumulated enough solar heat throughout the day to continue to provide it Escandón et al. / Building Simulation 9 during the night, when heat losses through the large glass surface and due to the lack of airtightness in the carpentry are very high. However, the existence of a large thermal mass on the ground floor, the pre-existing solid stone wall in the north-facing facade and the concrete floor, provides greater inertia and is able to return part of the heat accumulated throughout the day at night. Thus, interior temperatures reached throughout the day are not as high on the ground floor as on the top floor but the temperature is more constant over a full day. The solar exposure inside the pavilion throughout a typical winter day was observed in order to detect the direct radiation reaching the stone wall, both on the ground and top floors, and the concrete flooring on the ground floor (Fig. 11). In winter, when there is greater need for heat accumulation in the thermal mass of the envelope, the stone wall and concrete floor received direct solar radiation during most of the day. Specifically the lower part of the stone wall located to the north of the pavilion receives direct solar radiation up to a height of 1 m, while the incidence on the concrete floor is approximately 2.5 m deep. Although the accumulation surface on the horizontal plane is greater than that on the vertical one, the wall has more than double the capacity for accumulation than the flooring, mostly due to its greater thermal mass. The experience lived by the Smithsons in the pavilion and recorded in the diary kept by Alison already indicated that the pavilion was not functioning as they had expected: “1962. November 30. Cold; end of flies. […] 1974. May 3–5. Cold. […] 1976. December 10–12. Hoarfrost heavy Friday night: very cold. […] 1976. December 26–28. Flood lights on in evening: Jack Frost on windows inside (even behind downstairs shutters) as well as on outside. […] 1979. March 9–11. Upstairs Windows require half-hourly wipe due to condensation.” (Smithson and Smithson 1986) Alongside the problem of interior low temperatures during the night-time in winter the Smithsons recorded the poor operation of the top floor during the day in summer, autumn and spring: Fig. 11 Shade projected every 60 minutes inside the Upper Lawn Pavilion, during the time range of 6:00 to 20:00 h (January 12) “[…] it could be baking upstairs. […] 1967. October 12–25. Friday roasting hot inside Folly.” (Smithson and Smithson 1986) Unlike what happens in winter, in summer solar capture provokes a negative effect on the thermal behaviour of the pavilion, generating considerable overheating on the top floor (Fig. 12). The internal blinds used by the Smithsons on the top floor, did not provide enough solar protection to solve this problem. In July when the outdoor temperatures are higher, even when the blinds are used, operative temperatures of up to 35℃ are reached inside the top floor during the middle hours of the day. However, at night the heat accumulated throughout the day dissipates, and temperatures around 18℃ are obtained on both floors. This overheating problem would also have occurred on the ground floor in the summer but the Smithsons resolved this by ventilating it through the almost complete opening of glass surfaces (Fig. 1). It is observed that when introducing the maximum ventilation rate permitted by the program (18 ACH) to simulate the ground floor with all the glass panels open (15.9 m2), operative temperature falls, approaching the exterior temperature curve and remaining below 26℃ (Fig. 12). Following the analysis of the hourly evolution of operative temperature in all seasons an annual assessment was carried out on the results obtained (Fig. 13). When analysing the range of operative temperatures during a typical day for each month, the considerable oscillation between the top floor and ground floor temperatures is striking for the reasons mentioned earlier. Over a full year a problem of low temperature inside the house in winter, spring and autumn can be clearly observed. At such times the Smithsons very probably compensated for the low temperatures by using the electric Fig. 12 Hourly evolution of exterior and operative temperature (July 25)