Field assessment of thermal behaviour of historical dwellings in France

Building and Environment 45 (2010) 473–484

Contents lists available at ScienceDirect

Building and Environment journal homepage: www.elsevier.com/locate/buildenv

Field assessment of thermal behaviour of historical dwellings in France R. Cantin a, *, J. Burgholzer b, G. Guarracino a, B. Moujalled a, S. Tamelikecht c, B.G. Royet b a

ENTPE – LASH, DGCB CNRS URA 1652, Universite´ de Lyon, Rue Maurice Audin, 69518 Vaulx en Velin, France CETE de l’Est – Laboratoire Re´gional de Strasbourg, 11 rue Jean Mentelin BP9, 67035 Strasbourg, France c Maisons Paysannes de France, 8 passage des deux sœurs, 75009 Paris, France b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 23 March 2009 Received in revised form 1 July 2009 Accepted 6 July 2009

In many countries, there is a great number of old buildings with local architectural, patrimonial, aesthetic and historic interest. They are the products of the vernacular traditional architecture fully integrating the environmental, social and economic local constraints. Moreover, this built inheritance is more heterogeneous than the modern stock of existing buildings. The historical buildings were built with different architectural designs featuring local styles of construction, different techniques and historical expertise. By experience, the actors of the building sector know that the thermal behaviour of historical buildings are not those of modern buildings set up at the time of the industrial period. However, they do not have assessed these specific thermal characteristics of historical buildings. This paper describes the complexity of architectural designs of historical dwellings in France. A field investigation during one year highlights various thermal characteristics of 11 dwellings. It provides a new understanding of thermal behaviour of these historical dwellings. The results show the thermal characteristics of historical dwellings and their differences with modern architecture. Ó 2009 Elsevier Ltd. All rights reserved.

Keywords: Old building Energy consumption Thermal performance Complexity System

1. Introduction The present energy and environmental constraints require the improvement of the energy performance and environmental quality of buildings. The European Energy Performance Building Directive requires extending thermal regulations to existing buildings and local specificities. The energy optimisation of existing building stock is a complex issue in many countries [1,2]. The existing building stock increases slowly. For instance, in France, there are more than 30 million dwellings with an average annual growth of 1–2% approximately. To reduce the CO2 emission and energy consumption, the main effort must address this existing stock [1–4]. The energy performance of buildings is not the same for all the existing stock as shown in Table 1. Three main periods of architectural design characterise the existing building stock. In many countries, the middle of the 20th century is a transition period that marks real changes in the constructive modes of residential buildings (Fig. 1). In France, more than 10 million of dwellings, a third of the existing building stock, were constructed before 1948 [6].

* Corresponding author. Tel.: þ33 4 72 04 70 37; fax: þ33 4 72 04 70 41. E-mail address: [email protected] (R. Cantin). 0360-1323/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.buildenv.2009.07.010

The elements that make it possible to carry out this constructive distinction are:  new manufactured building materials that are easier to use (reinforced concrete floors, post-beams structures, concrete blocks) with hygrothermal properties that are different;  constraints of town planning due to the real estate market that made it difficult to build in relation to the environment (orientations according to the sun and winds);  economic and profitability constraints of the building sector related to a massive demand for dwellings after the World War II. From the thermal point of view, there is an important change from an architecture taking into account the climatic environment and using local resources and materials, to an industrialised architecture, using new building materials with different thermal properties. At the same time, designers with new techniques available tend to ignore the local conditions of each site. The 20th century sees the beginning of the production of modern buildings: they are mechanically ventilated, heated and artificially lit. The first international oil crisis and the first thermal building regulations also draw a line of essential fracture in the history of architecture. Since the 1970s, the requirements for the insulation of buildings are considerably reinforced and regularly updated.

474

R. Cantin et al. / Building and Environment 45 (2010) 473–484

Table 1 Energy consumptions ratios in kWh/m2 year of existing buildings in France [5]. Building sector

Use

Built before 1975

Built after 1975

Average

Residential

Heating Hot water Heating Hot water

328 36 209 19

80–100 40 155 40

210 37.5 196 29

Tertiary

In this paper, a building is qualified as ‘‘historical’’ if built before the emergence of the modern architectural movement (before 1948). Thus, historical existing buildings constitute the inheritance and the identity of historic cities. The history of architecture traces the changes in these buildings: a medieval period with a Romanesque and Gothic styles, a Renaissance architecture, a Baroque architecture, etc. For instance, during the Haussmann period in the 19th century, buildings complies with a set of rules regarding outside appearance. Neighbouring buildings have their floors at the same height and the use of quarry stone is mandatory along the avenues. The historical buildings are constructed with traditional techniques and the historical expertise acquired by specialised building craftsmen. Professional practices were non-industrialised and there was significant interaction between trades. The knowledge of designers benefited from multidisciplinary experiments. In this context, and from a heritage and architectural point of view, any energy retrofitting project of historical buildings is a risk in terms of sustainable development and a challenge in terms of the conservation of historical buildings [7–9].

Ancient architectural design

Transition period Industrialization

1930’s Thermal approach Local architectural designs

1950’s

Modern architectural design 1970’s Oil crisis

Thermal approach Global regulations

Fig. 1. Different thermal approaches and various architectural designs.

 built or vegetable masks: the planted drives (caducous foliage) of avenues limit the solar contribution in summer, particularly for the lower floors; they also reduce overheating of the roadway under trees.  joint ownership of façades in most cases, which makes it possible to reduce thermal surface losses of an apartment.  climatic variation between the street and the centre of the small urban block: these thermal differences (several degrees Celsius) are observed between the two façades of the building. In rural areas, the situation of old buildings generally tends to optimise solar savings and to reduce thermal losses, featuring a main façade oriented to the south, few or no openings towards the north, use of vegetation to create shading in summer, etc. The thermal behaviour of the building is generated by its situation and its architectural characteristics: orientation of rooms and spaces, distribution typology (crossing area or not), size of the rooms according to their use, etc. For the historical buildings, without existing efficient heating, ventilation and air conditioning systems (HVAC), the indoor environment generally tends to conserve indoor comfort. Some characteristics of the indoor environment which influence the thermal behaviour are identified:

2. Architectural characteristics of historical dwellings

 microclimate (wind, temperature, precipitation, exposure to sun, altitude, mountain or littoral area, urban or rural area, etc.)  close relief (a cliff may moderate the changes in air temperature; an embankment slope may bring thermal attenuations, etc.)  solar masks (other buildings, persistent vegetation, etc.)  built-up areas and joint ownership

 Rooms distribution: there is a separation between living rooms and service rooms. These rooms tend to be set according to the sun path: living rooms on the sunnier side, and service rooms on the colder side.  Thermal buffer spaces: back kitchens, cellars, storerooms and roofs are as many additional spaces, which constitute thermal attenuation zones limiting heat transfers with outside.  Historical dwellings are designed with a crossing ventilation. Without HVAC equipment, the air ventilation strategy is a natural ventilation. Thus, a crossing dwelling makes it possible to create an effective airflow to improve the indoor air quality.

In a historical urban environment, the predominant urban form is a small urban block, which developed in the French cities under the Haussmann period. This urban form shows several characteristics having consequences on the energy performance and environmental quality of the building:

Historical buildings have different bioclimatic and vernacular characteristics. An example of bioclimatism and vernacular architecture is described in different climatic zones in north-east India [13]. The houses, with solar passive features in building design, are constructed using locally available materials. Another example

The outdoor environment and the building site are influenced by the following constraints [10–12]:

Table 2 Characteristics of historical dwellings. N

Location

Latitude

Longitude

Altitude

Architectural features

Date of construction

B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 B11

Paris Strasbourg Toulouse Vouvray Paris Marne Nice Bretagne Corre`ze Normandie Alsace

48 870 N 48 330 N 43 360 N 48 170 N 48 870 N 49 050 N 43 950 N 47 900 N 45 240 N 49 420 N 48 820 N

2 330 E 7 380 E 1 250 E 1 380 E 2 330 E 3 950 E 6 900 E 3 270 W 1 510 E 0 450 E 7 730 E

34 m 153 m 136 m 142 m 34 m 74 m 450 m 96 m 580 m 20 m 150 m

Haussmann architecture, 5th floor, apartment Haussmann architecture, 1st floor, apartment Brick architecture, 3rd floor, apartment Stone, lime mortar, Ground floor, apartment Limestone, 1st floor, apartment Brick of clay, tile, House Stone, lime mortar, tile, House Granite, stubble, House Granite, slate, House Timber, cob, stubble, House Concrete, tile, insulation, House

1918 1898 18th century 1755 17th century 1870 18th century 17th century 15th century 1789 2004

R. Cantin et al. / Building and Environment 45 (2010) 473–484

475

Fig. 2. Location and architectural designs of case studies in France.

shows how, in Iceland, the vernacular architecture of farmhouses is characterised by turf walls and sod roofs, and thick turf walls influence thermal comfort in a positive way [14]. Another study presents the evolution of residential architecture in Old Havana and

Outdoor environment and site

Traditional techniques Envelope

Indoor environment

Windows

Equipment

Fig. 3. Levels of subsystems for investigation.

Occupants

its qualitative relation to climatic aspects, comfort and people’s way of life [15]. It shows how the compact urban morphology influence the natural ventilation and the thermal comfort in residential buildings in the Historical Centre of Old Havana in Cuba. Thus, different studies show the complexity interactions between the indoor environment of the historical building and its local outdoor environment. 3. Methodology 3.1. Selection of case studies In order to improve energy retrofitting strategies for historical buildings, it is necessary to know their thermal behaviour. It is in

476

R. Cantin et al. / Building and Environment 45 (2010) 473–484

Table 3 Organisation of collected data with subsystems. Subsystems

Data collected

Physical measures

Outdoor environment and site

Geographical parameters (latitude, longitude, altitude) Climate zone, average radiation Local climate, exterior temperature and relative humidity, site description (vegetation, urban area, solar masks), orientation Structural type, horizontal and vertical thermal bridge Physical characterisation of walls (conductivity, volumetric mass, thermal capacity) Window type, glass, size, surface, thickness, air infiltration HVAC characterisation, energy consumption, lighting fittings Plans, sections, façades, thermal zones (heating zones or not) Occupation and HVAC scenarios (day, week, year)

Field measurements Web data (Satel-Light) Recorded and monitoring data

Traditional techniques Envelope Windows Equipment Indoor environment Occupants

Table 4 List of monitoring sensors in each dwelling. N

Physical parameters

Monitoring sensors

A B C D E F G H

Outside air temperature Outside Relative humidity Inside air temperature (living room) Inside relative humidity (living room) Operative temperature (living room) Inside surface temperature of sunny wall Inside air temperature (sunny room) Inside relative humidity (sunny room)

Data logger HOBO H 08-032-02 Data logger HOBO H 08-007-02 Data logger HOBO H 08-007-02 HOBO TMC 6 HA

this way of view that their architecture and thermal behaviour is assessed in a field investigation [16–18]. For the field assessing the thermal behaviour of historical buildings, a sample of dwellings was selected over a wide geographical area in France (Table 2, Fig. 2). The assessment of each building was primarily based on architectural investigation. Each historical dwelling was selected according to the structural composition of its envelope, the construction techniques, the geometrical characteristics of its inhabited space, the location of the building on the plot of land, the glazed surfaces ratios, the orientation and the design of windows. Other criteria relate to the age and to the architectural design, the geographical and the urban location, and the number and the size of the floors. The sample is composed of ten historical dwellings and one modern residential building designed in 2004 according to the French thermal regulation of the year 2000.

Field measurements, thermographs Field measurements, bibliographical sources Field measurements, thermographs Field measurements questionnaires, energy invoices Field measurements Questionnaires

Each historical dwelling is a complex structure adapted to the way of life of occupants from the past, which combines historical construction techniques, various architectural designs and locally available materials.

3.2. Field investigation The assessment of the thermal behaviours of historical dwellings requires to take into consideration the heterogeneity and the complexity of their architectural designs. To understand the observed energy behaviours, the building must be considered as an open system including the multiple interactions between the building and its environment. The complexity of historical dwellings results from variable sustainable interactions with their environments. Thus, it is inadequate to assess the energy and thermal behaviour of complex buildings designed in different architectural periods with only the analytical approach. It is necessary to consider the dynamic, interdisciplinary and complex whole. In this context, a systemic approach, depending on cybernetics and system theory [19,20], can consider the historical dwelling in its complexity and its dynamics. It allows the interdisciplinary consideration of different levels of complexity with seven subsystems (Fig. 3):     

outdoor environment and site indoor environment traditional building techniques envelope windows

Fig. 4. Location of sensors (B9).

R. Cantin et al. / Building and Environment 45 (2010) 473–484

477

Fig. 5. Infra-red investigation of the heterogeneous envelope of a historical dwelling (B10).

Envelope Wall Width (10-2 m)

Conductivity walls (W/m. K )

90

8

80

7

70

6

Width

5

50 4

40

3

30 20

2

10

1

0

Conductivity

60

0 B1

B2

B3

B4

B5

B6

B7

B8

B9

B10

B11

Reference dwellings Fig. 6. Characteristics of walls.

Energy consumption (kWh/m².y)

U-value building (W/m².K)

5

200

4

150

3

100

2

50

1

0

U-value Building

Energy consumption

250

0 B1

B2

B3

B4

B5

B6

B7

B8

B9

B10

B11

Reference dwellings Fig. 7. Energy performance of dwellings.

 equipment  occupants A complex boundary can be modelled between the outdoor environment and the indoor environment [21]. This boundary is a subsystem composed of four subsystems: ‘‘envelope’’, ‘‘traditional techniques’’, ‘‘windows’’ and ‘‘equipment’’ (HVAC). The subsystem ‘‘occupants’’ is included in the subsystem ‘‘indoor environment’’. The systemic approach allows to emphasize the thermal characteristics of historical dwellings related to the subsystem ‘‘outdoor environment and site’’. This subsystem is a natural unit composed of

biotic factors in an area functioning together with all physical factors of the environment. For example, among the biotic factors, the trees modify the solar masks and create a microclimate with specific thermal and humidity conditions relative to a site. Among the physical factors, the geographical characteristics of a surrounding countryside influence the thermal exchanges, for example with the evaporation or convection phenomena (river, lake, local wind), or by the geological availability of local materials with specific hygrothermal properties [22–24]. The outside air temperature and outside relative humidity are measured with sensors (Table 4). They provide a thermal characterisation of the subsystem ‘‘outside environment and site’’.

478

R. Cantin et al. / Building and Environment 45 (2010) 473–484

IA* winter temperature (°C) 30

25

25

20

20

15

15

10

10

5

5

0

0 B1

-5

B2

B3

B4

B5

B6

B7

B8

B9

B10

B11

Indoor average temperature

Outdoor average temperature

OA* winter temperature (°C) 30

-5

Reference dwellings OA*: Outdoor Average – IA*: Indoor Average Fig. 8. Average winter temperature for dwellings.

Tout

Tin

26

Temperature (°C)

24 22 20 18 16 Fig. 10

14 12 16/2/05

11/3/05

4/3/05

27/4/05

20/5/05

12/6/05

6/7/05

29/7/05

21/8/05

Date Fig. 9. Example: indoor and outdoor temperatures (B10).

Tout

Tin

21

Temperature (°C)

20 19 18 17 16 15 14 11/4/05

15/4/05

19/4/05

23/4/05

27/4/05

1/5/05

6/5/05

Date Fig. 10. Indoor and outdoor temperatures between 10th April and 8th May 2005 (B10).

This systemic representation has the advantage to include in the investigation the totality of elements and their interdependence. It allows a global approach of the thermal behaviour, which is an emergence from a complex building in which neither the components nor the couplings are simple.

Each dwelling is investigated in situ, as an organised whole, with this decomposition in seven subsystems. The data groups are combined according to this systemic decomposition (Table 3). For each dwelling, physical thermal parameters were monitored continuously over one year (February 2005 – February 2006) to

R. Cantin et al. / Building and Environment 45 (2010) 473–484

OA* spring temperature (°C)

IA* spring temperature (°C)

30

25

25

20

20

15

15

10

10

5

5

0

Indoor average temperature

Outdoor average temperature

30

479

0 B1

B2

B3

B4

B5

B6

B7

B8

B9

B10

B11

Reference dwellings OA*: Outdoor Average – IA*: Indoor Average Fig. 11. Average spring temperature for dwellings.

IA* summer temperature (°C) 30

25

25

20

20

15

15

10

10

5

5

0

Indoor average temperature

Outdoor average temperature

OA* summer temperature (°C) 30

0 B1

B2

B3

B4

B5

B6

B7

B8

B9

B10

B11

Reference dwellings OA*: Outdoor Average – IA*: Indoor Average Fig. 12. Average summer temperature for dwellings.

Correlation coefficient ( )

100 90 80 70 60 50 40 30 20 10 0 B1

B2

B3

B4

B5

B6

B7

B8

B9

B10

B11

Dwellings Fig. 13. Correlation coefficients between indoor and outdoor temperature (without HVAC).

record their thermal dynamic behaviours. In addition, the location of sensors was selected according to the indoor environment (Fig. 4). This investigation was completed with an infra-red approach. Infra-red thermographs were mainly taken during the investigation of three subsystems: envelope, windows and traditional techniques. These infra-red investigations aim to characterise thermal heterogeneities of envelope that are otherwise undetectable. For instance, different construction techniques are hidden by the

coating on the walls of the building B10. It can be identify with the infra-red thermograph (Fig. 5). It allows to identify the type of air tightness of windows (differentiation according to façades), the modifications of thermal U-value of different surfaces of walls (old sealed openings, envelope) and the thermal bridges in some junctions of walls (traditional techniques). The infra-red thermograph was realised with a ThermaAM PM 595 of FLIR System with LCD screen and image processing software.

480

R. Cantin et al. / Building and Environment 45 (2010) 473–484

Correlation coefficient ( )

80 70 60 50 40 30 20 10 0 B1

B2

B3

B4

B5

B6

B7

B8

B9

B10

B11

Dwellings Fig. 14. Correlation coefficient between indoor and outdoor relative humidity (without HVAC).

HRout

HRin

2/5/05

31/5/05

Relative Humidity ( )

100

80

60

40

20

0 4/2/05

5/3/05

3/4/05

29/6/05

29/7/05

Date Fig. 15. Indoor and outdoor relative humidity in the modern building (B11).

Attic floor

0.26 m

0.39 m

0.39 m Floor 3

0.52 m Floor 2

0.65 m Floor 1

Basement

Fig. 16. Different thicknesses of walls in the subsystem ‘‘envelope’’ (B2).

0.73 m

R. Cantin et al. / Building and Environment 45 (2010) 473–484

481

Fig. 17. Specific thermal bridge with ventilated beam in brick façade (B2).

The selected monitoring configuration allowed to collect data for three different zones: outside local environment (climate and site), the living room and another room, the most lightened and exposed to the direct solar radiation. With this approach, the collected data can describe the thermal behaviours of each thermal zone during each season. Data were collected by sensors and stored on loggers, with a time step of 2 h, allowing measurements over the whole year (Table 4). The annual data were recorded with the monitoring sensors in each dwelling. They are used to characterise its thermal behaviour upon the different time periods. For each period, the eight sensors (Table 4) were placed in each dwelling according to the systemic organisation (Table 3). This year-long measurement in situ of 11 case studies provides a important database. Different analysis of collected data were made according to 4 time periods:

 one period of fifteen days in winter, in order to examine the compensation of the thermal losses by the heating system.  two periods of fifteen days in spring and autumn respectively, in order to examine the significant compensation of thermal losses by solar gains.  one period of fifteen days in hot season, in order to examine the building behaviour regarding to the summer comfort requirements.  different short time periods (48 h), located within these seasonal periods. The field investigation provide numerical characterisation of each historical dwelling: maximum and minimal values, standard deviations, average values, correlation between inside and outside air temperatures, inside and outside relative humidity. 4. Results and discussion 4.1. Assessment of thermal performance

35

Temperature (°C)

30 25 20 15 T (outdoor) 10 1h00

7h00

13h00

19h00

1h00

7h00

13h00

T (indoor) 19h00

Time Fig. 18. Outdoor and indoor temperatures the 28th and 29th June 2005 (B7).

The thermal behaviour is the result of interactions between the seven subsystems (Table 3). It is an emergent property analysed with quantitative indicators: energy consumption ratio, U-value, temperature, relative humidity (Figs. 6–10). The energy consumptions of these 10 historical dwellings, which are not insulated, approach the energy consumption of the modern dwelling B11, which is insulated (Figs. 6 and 7). The energy performance of historical dwellings in this sample is better than the average performance of the existing building stock built before 1975. Indeed, the average energy consumption of existing residential building is 364 kWh/m2 year (Table 1) and the energy consumption of these historical buildings in the sample is between 107 and 227 kwh/m2 year (Fig. 7). The energy consumption of the modern building B11 is 110 kWh/m2 year.

482

R. Cantin et al. / Building and Environment 45 (2010) 473–484

Table 5 Differences of architectural designs and thermal characteristics between historical dwellings and modern dwellings. Historical dwellings

Modern dwellings

Local materials with building design used over several centuries Location and orientation according to solar paths, winds and precipitations Indoor distribution with crossing rooms according uses, thermal buffer zones, specific optimised openings Materials which are very sensitive to water (variable U-Value) Significant uses of plaster and superficial coatings allowing the absorption of air humidity without structural damages (rooms without heating or cooling systems) Heavy structural masonry with strong thermal mass (façades and distribution walls). Different types of walls on the same floor according to uses (representative stones on the street side and wood site on the courtyard side) Different thickness of walls with their structural constraints according to the floors (reduced thickness for the higher floors) Wooden floors with good energy performance Materials filling partitions and floors with hygrothermal regulation properties Few thermal bridges in façades because of building design

Development of building systems according to new demographical, economical and industrial constraints No optimisation of climatic constraints: location, orientation, openings, etc. Choices influenced by urban regulations Indoor distribution according general models in order to create spaces independent of local environment. Industrial materials which are unaffected by water but which are sensitive to thermal dilatations. Few absorptive materials inside the building

Identical and prefabricated walls with limited thickness (concrete, inside insulation) Similar, industrialised, horizontal and vertical walls

Standardisation of building design without differences between structural walls and façades. Solid concrete floors Secondary materials with principally an aesthetical role Significant thermal bridges because the construction techniques use jointed prefabricated elements.

In winter, a satisfactory thermal behaviour of the subsystem ‘‘indoor environment’’ is obtained for most of the historical dwellings (Fig. 8). The indoor average temperatures of dwellings B3 and B4 are low (17.7  C and 14.2  C) because of a recurring inoccupation of housing. For the dwellings B6, B9 and B10, the indoor average temperatures are also low (16.3  C, 18.2  C, 14.4  C) because of an inadequate heating equipment control. In order to assess the thermal behaviour of each building without any heating or cooling system, the monitoring measures are selected in spring, between April and May 2005 (Figs. 9 and 10). In spring (Fig. 11), indoor thermal behaviour of historical dwellings is satisfactory without any subsystem ‘‘equipment’’(HVAC). The range of indoor average temperatures is between 17.6  C (B9) and 23  C (B8) with a standard deviation between 0.3 (B8) and 2.4 (B3). In summer (Fig. 12), the variations of average indoor air temperatures are less significant than the variations of average outdoor air temperatures. The range of indoor temperatures is between 23  C (B10) and 27  C (B1 & B5) with a standard deviation between 0.7 (B8) and 2.5 (B3). The range of outdoor temperatures is between 20.8  C (B8) and 25.7  C (B5) with a standard deviation between 2.7 (B10) and 5.3 (B11). These results show the capacity of these historical dwellings to attenuate the impact of the diurnal summer thermal variations. These thermal characteristics can be also explained by an optimised management of occupants, combining reduction of solar gains during daytime and cooling at night. This adaptive behaviour ensures a thermal regulation thanks to specific architectural designs of historical dwellings: openings windows ratios adapted to the orientation of the façades, crossing distribution of the indoor environment, high thermal mass of the built unit (wide walls), various devices allowing natural ventilation (openings windows permanent or not, thermal draught created by existing temperature gradients between the façade under the sun and the other façade in the shadow).

correlation coefficient. The correlation coefficient is determined between outdoor and indoor temperature without HVAC systems. The Fig. 13 shows that the thermal behaviour of historical dwellings (B1 to B10) is strongly dependent on the subsystem ‘‘environment and site’’ (near surrounding). For all these historical dwellings (B1 to B10), the correlation between the indoor and outdoor temperature is over 40%. For these historical dwellings, the average of correlation coefficients is 60%. Each historical dwelling works like a thermally open system in interaction with its environment and site. In most cases, the subsystems ‘‘envelope’’ and ‘‘windows’’ set up a permeable boundary, which receives and controls the thermal and microclimatic contributions. There are several air exchange and humidity flows by infiltration and natural ventilation across the boundary. Thus, the indoor environmental quality appears as the combination of complex interactions between the outdoor subsystem ‘‘environment and site’’ and the indoor subsystem ‘‘indoor distribution’’ and ‘‘occupants’’. The correlation between the indoor and outdoor temperature for the modern dwelling B11 is only about 10%. This modern dwelling is designed like a closed and insulated system where the exchanges of energy flows with the outdoor environment are mechanically controlled. The microclimate and the site factors have not been a priority for the designer. Indeed, this modern building B11 has the same standardised and industrialised architecture encountered in other modern buildings in France. Although it is in the coldest area in France, this building B11 has been directed East-West with only one small window to the south (Fig. 2). This modern building is tight and impermeable like a mechanically ventilated box. The monitoring data show a strong correlation between indoor and outdoor relative humidity for the historical dwellings (Fig. 14). In the contrary, the indoor relative humidity in the modern dwelling B11 is not directly dependant on the outdoor relative humidity (Figs. 14 and 15). The correlation coefficient of B11 is 3% while the correlation coefficients of historical dwellings exceed 34%.

4.2. Thermal correlation between outdoor and indoor environments

4.3. Other thermal considerations about historical dwellings

In these dwellings, the thermal interaction between outdoor and indoor environments can be identified with calculation of

The field investigation shows the complexity of architectural design of each historical design. For instance, the local materials

R. Cantin et al. / Building and Environment 45 (2010) 473–484

used in the envelope often have unknown characteristics. The theoretical or statistical values such as thermal conductivity of a wall must be considered with caution. These values given in the scientific literature are determined in laboratory where the boundary conditions of theoretical characterisation are known. In historical buildings, the characteristics of local materials are heterogeneous and their hygrothermal properties are variable and anisotropic. The infra-red thermograph shows this thermal heterogeneity. For example, it shows this heterogeneity in the envelope of the dwelling B10 made of different materials: stone (foundation and basement), timber and mud (vertical walls), and a thatched roof (Fig. 5). Envelopes of historical buildings may differ within the same building according to their geographical areas and their age, as well as in their vertical or horizontal constitution. Several materials may have been used in a same wall and sometimes there is different width according to floors. For instance, in the building B2, the width of wall is 0.65 m at the 1st floor and 0.39 m at the 4th floor (Fig. 16). This diversity generates difficulties for modelling the heterogeneous and anisotropic characteristics of the walls. Different types of materials in the walls were encountered: stone masonry (B4, B5, B7, B8, B9), brick masonry (B1, B2, B3), walls in timber (B5 and B10), and clay brick (B6). Stone masonry constituted the building type most common in France in the past. The old stone structure has a great material diversity. The stones were extracted locally because of the high cost of transport. Bricks are manufactured materials made of clay. Until 1930, the dimensions of solid bricks were not standardised and varied from one geographical area to another. The bricks are encountered in the geological area where clay deposits exist at low ground depth, and where stones are rare. The junctions between the masonry and timber or iron beams are generally carried out by arranging open spaces around these parts of structure. They avoid the rotting of wood or the oxidation of iron when in contact with masonry. This technique considerably limits heat exchange by conduction between the floor and the façade (Fig. 17). The construction techniques and indoor environment require the need for heavy components inside old buildings: partitions in brick or stone, heavy floors, etc. This inside mass supports the thermal inertia. The higher thermal mass of a wall provides the thermal capacity to store and release large energy flows in winter or in summer. In summer, the thermal inertia makes it possible to obtain a shift of the temperature, between indoor and outdoor, which can last several hours. For instance, in the dwelling B7 with stone walls and an average width of 0.50 m, there is a maximal daily variation in temperature which can reach 6 K (Fig. 18). In winter, a high thermal inertia allows to store solar energy during the day. Thus, the building components release stored energy with a delay of several hours. This delay makes it possible to reduce the heating period and to preserve a comfortable temperature during the night. With numerous sunny days in spring and in autumn, the thermal inertia make it also possible to reduce the heating period. The thermal investigation of historical dwellings has shown an important sensitivity to hygrothermal flow without any particular structural problem related to moisture. This sensitivity mainly results from used porous materials and specific building techniques in historical dwellings. In modern industrial buildings, the walls are generally tight. Their foundations and the materials are dry and protected against humidity. Physical monitoring does not show condensation problems on surface. The recorded temperature of surface walls curves is

483

systematically above the dew temperature with a margin ranging between þ6  C and þ15  C. The superficial coatings is often a porous material which resist structurally to the temperature and humidity variations without degradation. The thermal conductivity of coating is variable according to the humidity of materials, and the climatic variations of the environment. From the thermal point of view, the field investigation allows to identify several differences between historical and modern dwellings. The Table 5 shows these differences. 5. Conclusion This field investigation of 11 dwellings has underlined the specific thermal behaviour of historical buildings. A systemic approach has allowed to investigate the complexity of the architectural design of historical dwellings with seven subsystems: ‘‘outdoor environment and site’’, ‘‘traditional techniques’’, ‘‘envelope’’, ‘‘windows’’, ‘‘equipment’’ and ‘‘indoor environment’’ including ‘‘occupants’’. The emergent properties of historical dwellings have been identified with a year-long measurement of temperatures and relative humidity. This field investigation has shown that the energy consumption of these historical dwellings is lower than the average energy consumption of existing dwellings. This result is against as it would be expected. A strong thermal correlation has been underlined between outdoor and indoor environment, higher than in the modern dwelling. These results have shown that these historical dwellings are interactive systems, with bioclimatic properties more complex than the modern dwelling, with similar energy performance. Indeed, their design and architectural elements take into account climate, environmental and site conditions to maintain thermal comfort, without mechanical systems. Thus, it is necessary to investigate accurately each historical building before any retrofitting intervention. The lifespan of historical dwellings could be reduced because of inappropriate retrofitting project which can modify the thermal balance of historical buildings. References [1] European Commission. Implementation of the energy performance of building directive. Country reports 2008. Brussels: EPBD Buildings Platform, ; 2008. p. 220. [2] International Energy Agency. Energy conservation in buildings and community systems programme, Annex 46. Holistic assessment tool-kit on Energy Efficient Retrofit Measures for Government Buildings (EnERGo), ; 2005–2009. [3] Balaras CA, Droutsa K, Argiriou AA, Asimakopoulos DN. EPIQR surveys of apartment buildings in Europe. Energy and Buildings 2000;31(2):111–28. [4] Poel B, van Cruchten G, Balaras CA. Energy performance assessment of existing dwellings. Energy and Buildings 2007;39(4):393–403. [5] Observatoire de l’e´nergie. France: INSEE; 2003. [6] Age distribution of residential buildings. France: MELTLM, Account Housing; 2002. [7] Wood C, Tadj O, Bordass W, Cassar M, Palmer O, UK. Building regulations and historic buildings. English Heritage Publishing, ; 2004. [8] Naaranoja M, Uden L. Major problems in renovation projects in Finland. Building and Environment 2007;42(2):852–9. [9] Balaras CA, Gaglia AG, Georgopoulou E, Mirasgedis S, Sarafidis Y, Lalas DP. European residential building and empirical assessment of the Hellenic building stock, energy consumption, emissions and potential energy savings. Building and Environment 2007;42(3):1298–314. [10] Vissilia AM. Evaluation of a sustainable Greek vernacular settlement and its landscape: architectural typology and building physics. Building and Environment 2009;44(6):1095–106. [11] Ipekoglu B, Boke H, Cizer O. Assessment of material use in relation to climate in historical buildings. Building and Environment 2007;42(2):970–8. [12] Tassiopoulou T, Grindley PC, Probert SD. Thermal behaviour of an eighteencentury Athenian dwelling. Applied Energy 1996;53(4):383–98.

484

R. Cantin et al. / Building and Environment 45 (2010) 473–484

[13] Singh MK, Mahapatra S, Atreya SK. Bioclimatism and vernacular architecture of north-east India. Building and Environment 2009;44(5):878–88. [14] van Hoof J, van Dijken F. The historical turf farms of Iceland: architecture, building technology and the indoor environment. Building and Environment 2008;43(6):1023–30. [15] Tablada A, De Troyer F, Blocken B, Carmeliet J, Verschure H. On natural ventilation and thermal comfort in compact urban environments – the Old Havana case. Building and Environment 2009;44(9):1943–58. [16] Ipekoglu B. An architectural evaluation method for conservation of traditional dwellings. Building and Environment 2006;41(3):386–94. [17] Saı¨d MNA, Brown WC, Shirtliffe CJ, Maurenbrecher AHP. Monitoring of the building envelope of a heritage house: a case study. Energy and Buildings 1999;30(3):211–9. [18] Grinzato E, Bison PG, Marinetti S. Monitoring of historical buildings by the thermal method. Journal of Cultural Heritage 2002;3(1):21–9.

[19] von Bertalanffy L. The´orie ge´ne´rale des syste`mes. Paris: Bordas-Dunod; 1991. p. 298. [20] Le Moigne JL. La the´orie du Syste`me Ge´ne´ral. 4th ed. Paris: PUF; 1994. [21] Hendricks L, van der Linden K. Building envelopes are part of a whole: reconsidering traditional approaches. Building and Environment 2003;38(2):309–18. [22] Bidner T, Lanz B, Mitterer S, Recheis A, Hauser W, Franzen C, et al. Material properties, room-climate measurements and building history – an interdisciplinary project to set the conditions of restoration and re-use for the fortress of Kufstein, Tyrol, Austria. Building and Environment 2003;38(9–10):1133–41. [23] Beck K, Al-Mukthar M, Rozenbaum O, Rautureau M. Characterization, water transfer properties and deterioration in tuffeau: building material in the Loire Valley – France. Building and Environment 2003;38(9–10):1151–62. [24] Clarke JA, Yaneske PP. A rational approach to the harmonisation of the thermal properties of building materials. Building and Environment 2009;44(10): 2046–55.