Energy Efficient Renovation of Belgian Houses - CyberLeninka

Available online at www.sciencedirect.com

ScienceDirect Energy Procedia 96 (2016) 158 – 169

SBE16 Tallinn and Helsinki Conference; Build Green and Renovate Deep, 5-7 October 2016, Tallinn and Helsinki

Energy efficient renovation of Belgian houses: sensitivity analysis for thermal bridges Stijn Van Craenendoncka*, Leen Lauriks a, Cedric Vuye a a

EMIB, University of Antwerp, Groenenborgerlaan 171, 2020 Antwerp, Belgium

Abstract In this paper, the influence of construction details on the overall thermal insulation and energy demand in Belgian houses is examined. The research was carried out for 12 archetypal Belgian single family houses. For each house, nine typical construction details were selected. The Flemish EPBD method was used for an energy performance simulation before and after refurbishment. It can be concluded from this research that window joints refurbishment has a bigger impact than any other joint refurbishment. Another observation is that that the reduction of energy consumption due to refurbishment of construction joints is not linked to joint length. The Authors. Published by Elsevier Ltd. © 2016 2016Published © by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license Peer-review under responsibility of the organizing committee of the SBE16 Tallinn and Helsinki Conference. (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of the SBE16 Tallinn and Helsinki Conference. Keywords: T hermal bridge; Renovation; Insulation, Energy demand

Nomenclature ai Ai AT C D DS EAP

weighting factor correcting for surface i adjacent to the ground or unheated spaces area of the surface i total area of heat loss surface compactness Detached Door Sill joint Energy Advice Procedure

* Corresponding author. Tel.: +32-3-265-19-77 E-mail address: [email protected]

1876-6102 © 2016 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of the SBE16 Tallinn and Helsinki Conference. doi:10.1016/j.egypro.2016.09.117

Stijn Van Craenendonck et al. / Energy Procedia 96 (2016) 158 – 169

EPC FC FF FG kS K Li NED N-RJ RD RE RJ SD StDv T Ui V WL WP WS Fi \i

Energy Performance Certificate Façade – Cellar joint Façade – storey Floor joint Façade – Ground floor joint mean heat transfer coefficient K-level (Global insulation measure) length of linear construction detail i Net Energy Demand for heating Refurbished situation with non-refurbished joints Rake Verge joint Roof Eaves joint Refurbished situation with refurbished joints Semi-Detached Standard Deviation Terraced heat transfer coefficient of the surface i according to NBN B62-301 protected volume Window Lintel joint Window Post joints Window Sill joint thermal transmittance coefficient for point construction detail i thermal transmittance coefficient for linear construction detail i

1. Introduction The influence of our buildings on climate change is considerable. Numerous studies showed that 40% of European CO2-emissions and energy consumption can be linked to our buildings . [1], [2] Reducing the environmental impact and energy consumption of buildings cannot be solely achieved by regulating new buildings. Interventions on existing buildings by means of renovation are also necessary. Since more than 80% of Flemish buildings was built before 1990, there is still a lot of potential for improvement. [3]. A method often used to assess refurbishment potential and strategy is the use of an envelope parts typology. Cuerda et al. [4] developed 8 façade typologies and 5 renovation typologies for Spanish buildings to assess potential energy savings through renovation. Dascalaki et al. [5] in turn used a typological approach to assess the energy performance of residential buildings in Greece. In [6], building typologies were used as a first step to develop standard prefabricated modules for retrofitting multi-family homes. In [7], the Institut Wohnen un Umwelt employed a typological approach to assess energy savings potential for German buildings. As outcome from the European TABULA-project [8], a standard framework for European building typology definition and energy assessment was developed. In all of these projects, only typologies for planar envelope sections were employed. However, typologies for construction details in existing buildings are still lacking. [7], [9], [10] Nevertheless, other research has shown that the influence of construction joints on the energy performance of buildings is considerable. Lahmidi and Leguillon [11] demonstrated in a French house that breaking thermal bridges can provides a primary energy saving of more than 7%, to up to more than 15% when combined with other measures. In Belgium, the SENVIVV-study [12] proved that thermal bridges can account for 10% in global thermal insulation in Belgian houses between 1990 and 1996. In [13], Erhorn-Kluttig and Erhorn provide an overview of several studies where typologies are used to calculate energy savings [13]. It is reasonable to assume that, due to stability and accessibility restrictions during refurbishment, it is not possible to achieve the energy performance of newly constructed details through renovation. How much the energy performan ce can be improved through refurbishment is still unknown. This refurbishment cannot only provide solutions for improving energy performance, but also for other local problems such as draft, moisture and mold. [14]. A correct assessment of the energy impact of thermal bridge

159

160


refurbishment can be used, in combination with existing building stock models , to project future energy consumption and assess the effectiveness of refurbishment measures. Similar perspectives are expressed in the Episcope project [15] and Build Up Thermal Bridges forum [16]. This paper identifies priorities for refurbishing construction details in Belgian houses. For this purpose, a construction detail typology was developed first. Th is typology was used to assess the energy savings potential and relative importance of retrofitting each construction detail for 12 archetypal Belgian houses. In section 2 of this paper, the research method is explained, as well as the energy simulation method and data sources. In section 3, the construction joint typology and corresponding refurbishment strategies are established. Section 4 presents the results of the energy simulation. The paper is finalized in section 5 with conclusion and further research perspectives. 2. Research Methodolog y 2.1. Research setup The archetypal Belgian houses used in this research originate from the Belgian housing typology developed for the TABULA-research project. [8] The houses in this project are split into several age groups. Only the four age groups pre 2005 were studied (pre 1946, 46-70, 71-90, 91-05). Later houses were not included because of their limit ed renovation frequency. Furthermore, the houses were split into three housing configurations (detached, semi-detached, terraced). The geometrical data of the houses in this TABULA housing typology was based on statistical processing of approximately 11000 Belgian houses in the Energy Advice Procedure (EAP) and Energy Performance Certificat e (EPC) databases. The roof, floor, façade, window and door surface areas, as well as protected volume and gross floor surface were averaged. A distinction was made between surfaces bordering outside, ground or unheated indoor spaces. Because of the large number of houses, these data can be assumed representative for the Belgian housing stock. An algorithm was developed to derive the dimensions of the archetypal houses based on the reported surface areas. From these dimensions, the lengths of all studied linear construction joints were calculated (more info in 2.3). The development of the construction joint typology was based on textbooks, from which nine typical construction joints for each age period were selected, see Fig. 1: x 4 window & door joints: window sill (WS), posts (WP) & lintel (WL), door sill (DS) x 3 façade joints: façade – cellar joint (FC), façade – ground floor joint (FG), façade – storey floor joint (FF) x 2 roof joints: roof eaves (RE), roof verge (RV)

Fig. 1: Location of studied construction joints

For each of the archetypal houses, a model was made according to the Flemish Energy Performance of Buildings (EPBD) energy performance evaluation method. The EPB 6.5.1. software [17] was used for building the models. Then, for each housing type, a refurbishment strategy was developed to comply with the Flemish EPB2016 demands. This allowed for two possible scenarios. In the firs t one, the planar envelope parts were refurbished without intervention on the construction joints. In the second one, the joints were refurbished as well. The effect of the construction joints was taken into account via their respective linear thermal tran smittance coefficient \. The \-values were calculated using the THERM 7.3 software [18] according to NBN EN ISO 10221:2007. [19]

161


2.2. EPBD energy performance evaluation method The Flemish EPBD evaluation method [20] was used to assess the influence of the construction joints on the energy performance of the entire building. This method uses a fixed inside temperature of 23°C for calculating cooling load and 18°C for other calculations, and typical monthly outside temperatures to represent the Belgian Cfd Köppen -Geiger climate [21]. Two of the resulting parameters from this method were used in this research to assess impact on energy performance: K-level and net energy demand. The K-level is an indication of the level of global thermal insulation of the building. It is determined from the mean heat transmittance coefficient (kS), calculated according to Eq. ( 1 ), and the compactness of the building, defined by Eq. ( 2 ). The linear construction joints examined in this paper can influence the calculated K-level through their length and their thermal transmittance coefficient. From ks and C parameters, the K-level is determined using Eq. ( 3 ): σሺ ܽ௜ ή ‫ܣ‬௜ ή ܷ௜ ሻ ൅ σሺ‫ܮ‬௜ ή ߰௜ ሻ ൅ σሺ ߯ ௜ ሻ ݇௦ ൌ (1) σ ‫ܣ‬௜ ‫ ܥ‬ൌ

ܸ

(2)

‫்ܣ‬

‫ ܭ‬ൌ ͳͲͲ ή ݇௦ ‫ ܥ‬൒ Ͷ ͵ͲͲ ή ݇௦ ‫ܭ‬ൌ ͳ ൏ ‫ ܥ‬൏ Ͷ ‫ܥ‬൅ʹ ‫ ܭ‬ൌ ͷͲ ή ݇௦ ‫ ܥ‬൑ ͳ

(3)

The maximu m allowed K-level for new buildings constructed in 2016 is K40, coming from K65 in 1992, when Klevel was introduced, and dropping by 5 points every few years . The fact that K-level depends on compactness of the building can be seen as a problem. The archetypal buildings don’t all have the exact same compactness, complicating comparing K-level between them. For this reason, another variable was studied: the net energy demand for heating (NED). NED is more directly linked to energy consumption, which also makes it a valuable variable to take into account. K-level cannot be pushed aside completely though, because it still provides the most direct link to thermal insulation, which can be impacted directly by construction joints. The NED is calculated by Eq. ( 4 ) and is influenced by construction details via both final energy consumption terms, which depend on the average heat transfer coefficient of the building skin, which is in term influenced by thermal bridging effects in construction joints. ‫ܧ‬௣ ǡ௛௘௔௧ǡ௠ ൌ ෍ ݂௣ ή ܳ௛௘௔௧ǡ௙௜௡௔௟ǡ௦௘௖௜ǡ௠ǡ௣௥௘௙ ൅ ݂௣ ή ܳ௛௘௔௧ǡ௙௜௡௔௟ǡ௦௘௖ ௜ǡ௠ǡ௡௣௥௘௙

(4)

௜

With:

Ep,heat,m fp Qheat,final,sec i,m,pref Qheat,final,sec i,m,npref

Monthly primary energy consumption for heating Conversion factor for primary energy per energy source Monthly final energy consumption by preferential heating device per energy sector i Monthly final energy consumption by non-preferential heating device per energy sector i

2.3. Building typology geometry As mentioned in the previous paragraph, construction details can influence the K-level by means of the total length of the construction joint. Therefore, it was important to make a correct estimation of the length of each typical construction joint in the archetypal Belgian buildings. From the reported surface areas in the TABULA-report, as well as the reported number of storeys from the TABULA Webtool [22], the dimensions of the building were calculated. To achieve a proper conversion from surface to lengths, some assumptions had to be made. Th ese assumptions were:

162


x x x x x x x x x

All houses have a rectangular ground plan; The number of storeys in a building is equal to the number of storeys reported, plus one for the attic floor; Every house has a cellar, accounting for the floor surface adjacent to an unheated indoor space; All buildings have a gable roof covering the entire building; All houses have an unheated area in the top part of the attic. The ceiling that connects to this roof is considered as roof surface bordering an unheated indoor space; The attic floor is included in the gross floor surface reported, the cellar floor is not ; There can be vides or stairwells in the house, as well as unheated spaces counted in the gross floor surface ; All buildings have a one-storey unheated indoor space adjacent to a free corner. Terraced houses don’t have a free corner and therefore have an unheated indoor sp ace adjacent to the rear façade; Door lintels and posts are have the same construction as window lintels and posts.

The lengths of all typical construction joints are represented in Table 1. Joint lengths remain more or less the same across time periods for each housing configuration. They also generally decrease from detached (D) to semi-detached (SD) and from semi-detached to terraced (T). It is noticable that the length of the door sill (DS) is the same for all time periods. This can be explained by the door area which has remained the same for all periods in the original data set, probably due to an assumption in the original TABULA-data. It should also be noted that the façade-floor joint length is lower for the detached buildings than the other housing configurations. This is due to the fact that the detached buildings reportedly have one reported storey, while the other configurations have two. T able 1: Construction joint lengths [m]

FG

D 7

pre 46 SD 7

T 8

D 5

46-70 SD 6

T 7

D 4

71-90 SD 6

T 8

D 4

91-05 SD 9

T 9

FC

55

39

27

55

36

28

58

39

24

55

35

23

WP

50

40

34

50

40

34

49

41

31

52

43

33

WS

28

21

17

28

21

16

28

22

15

30

23

15

WL

31

23

19

31

24

19

30

24

17

32

25

18

DS

2

2

2

2

2

2

2

2

2

2

2

2

RE

51

34

21

50

30

20

51

34

18

47

30

17

RV

9

10

12

8

10

12

8

8

11

10

13

15

FF

62

93

71

60

84

70

62

90

64

59

87

64

3. Construction detail selection

3.1. Definition of thermal bridge typologies The construction joint typology was developed using textbooks used in secondary and higher technical education, published in the period for which they were consulted (e.g. 46-70 typology was based on books published between 1946 and 1970). The textbook were on engineering and architecture, but also more on practical topics such as carpentry For the pre 46 period, 11 textbooks by 6 different authors were used [23]–[33]. For the second period (46-70) a total of 6 books by 4 different authors were consulted [34]–[39]. The typology for the 71-90 period was constructed by using a total of 6 books from 5 different authors [40]–[46], and one technical instruction for Stichting Bouwresearch (SBR) [47]. For the final period (91-05), the typology was entirely based on 9 technical instructions from the BBRI [48]–[56]. In total, 592 relevant reference pictures were collected. Some of these bo oks were written and published in the Netherlands, but the information was assumed to be relevant for the Belgian construction practice. For the definition of the typology, the construction joint typology was established by averaging the relevant collected pictures per time period. The final typology is represented in Fig. 2. Please note that the drawings only represent what was simulated, resulting in no glass are roo f finishing being drawn.


Fig. 2: Construction joint typology

163

164


3.2. Retrofitting solutions In this section, all refurbishment measures for planar envelope sections and construction joints of the archetypal buildings are discussed. The retrofitting strategy for each envelope section was the same for each time period, to make sure the results were comparable between all housing configurations and time periods. The strategies selected for the planar envelope sections are all common techniques. For the walls, the chosen strategy was external wall insulation. For original cavity walls, the outer cavity leaf was removed, as well as any thermal insulation present. The existing wall or inner cavity leaf was then covered with 14 cm rigid thermal insulation, to comply with the demands on energetic renovations in Flanders. The external finishing of the insulation is not taken into account, so the wall could be finished in all manner of ways. For the renovation of the roof, a wooden frame was installed between the underpurlins. Insulation was placed in the frame and between the rafters . This results in a total insulation thickness of 22 cm. The floor is insulated with 10 cm rigid insulation. This means that the floor level rises by 11 cm in the 91-05 period. In all other periods, the sabulous clay and soil under the original floor finishing is removed, a new concrete floor plate is cast into place, and insulation, screed and finishing are placed on top, in such a way that the original floor level is retained. Doors and windows are replaced in accordance with the new building regulations. Only one retrofitting strategy was developed and applied to each construction joint. The strategy was applied to the respective construction joints in each time period. The refurbished construction joint typologies of the 46-70 time period are shown in Fig. 3.

Fig. 3: Refurbished construction joints 46-70

In the construction details, the thermal bridges were broken by connecting insulation layers. The rafters needed to be elongated to make this connection in the RE-joint. Part of the gable wall was also removed to assure connection between insulation layers. In the FF-joint, a part of the floor beams was removed.

165


For the windows, the window frame was surrounded by a 2 cm thick layer of rigid insulation, with the frame fixed to the wall with steel brackets. The door sill was completely replaced in the refurbishments, connecting insulation layers with a thermal block. The new stone sill rests on a steel bracket on the concrete floor plate. In the façade-ground floor (FG) and façade–cellar (FC) details, connecting the insulation layers was not possible. The choice was made to elongate the insulation of the wall under the ground level, as to make the path of least resistance for the heat flow longer than 1 meter. This method is accepted within the Flemish EPBD-method. 3.3. Calculation of linear thermal transmittance The linear thermal transmittance (\) of all construction joints was calculated according to NBN EN ISO 10221:2007 [19] using the THERM 7.3 software [18]. The thermal conductivity coefficients of the materials used were mostly taken from the Flemish “Transmissiereferentiedocument” (Transmission reference document) [57]. The thermal conductivity coefficients are summarized in Table 2. Table 2: Thermal conductivity coefficients for interior (left column) and exterior (right column) of used materials OUi (W/mK)

OUe (W/mK)

OUi (W/mK)

OUe (W/mK)

Brick

0,46

0,91

Cinder block

N/A

0,45

Stucco

0,52

N/A

Drywall

0,24

Mortar

0,93

N/A

1,5

Light wood

0,13

0,18

Stone Concrete

2,91

3,5

Heavy wood

0,18

0,2

1,7

2,2

Underlayment

N/A

Screed

0,5

1,7

N/A

Rigid insulation*

0,06

N/A

Sabulous clay

1,7

N/A

Soft insulation *

0,06

N/A

Ceramic tiles

0,81

N/A

Soil

2

N/A

* non-refurbished

During renovation, all insulation materials were replaced and supplemented with PIR (polyisocyanurate)insulation. The conductivity coefficient for PIR in [57] (0,045 W/mK) is not representative for the products on the market, which have a reported value of 0,022-0,026 W/mK. Therefore, a value of OUi = 0,025 W/mK was chosen to represent the product. The walls of the pre 46 and 46-70 period, as well as the outer cavity leaf of the other periods are assumed to be constructed out of 18 x 5 cm bricks with a 1,5 cm mortar joint. The inner cavity leaves of the last two periods were constructed out of 29 x 14 cm cinder blocks with a 1,5 cm mortar joint. The wall cavity had a width of 5 cm in the 46-70 period, and 3 cm in the later periods. The cavities were assumed to be semi-ventilated with an opening of 700 mm²/m. This way, the total thermal transmittance (U) of both walls was in accordance with the values in the TABULA-report [8]. U-values for doors and windows before refurbishment were taken from [8] for each time period. After refurbishment, the U-values for doors and windows were assumed to be equal to the maximu m value according to Flemish regulations for newly constructed buildings in 2016 [58]. 4. Global energy performance

4.1. Thermal insulation level The K-level for the original, non-refurbished situation, was compared to the refurbished situation with nonrefurbished construction joints (N-RJ), and the refurbished situation with refurbished joints (RJ). The results are represented in Fig. 4. The black line represents the maximu m K-level according to Flemish regulations.

166


Fig. 4: K-level before and after refurbishment, with and without refurbished joints

The difference in K-level between the refurbished situations with and without refurbished joints is on average 4,0±1,0 (±StDv) points. Notice that K-level increases form the pre 46 period to the 46-70 period. This is due to the fact that the massive walls in the pre 46 period are thicker and have better insulating properties than the uninsulated cavity walls in the 46-70 period. A Wilcoxon signed rank test [55 p.186-189] performed per time period showed that the difference between both refurbishments is not significant. Therefore, from a K-level point of view, there is no evidence that it is reasonable to refurbish all nine of the in vestigated construction joints. To evaluate if there are joints whose refurbishment have a significant influence on K-level, the joints were divided into 4 subsets. The subsets were developed from refurbishments that are likely to be carried out at the same time. x x x x

Window joints: Window sill (WS), posts (WP) & lintel (WL), and door sill (DS); Wall base joints: Façade – ground floor (FG) and façade – cellar (FC) joints; Roof joints: Roof eaves (RE) and roof verge (RV) joints Intermediate floor: façade – storey floor (FF) joint

Multiple simulations were performed next, each with one subset of joints refurbished. The impact each refurbishment had on K-level was analyzed. For evaluation of the impact of joint length, all joint lengths were summated per subset. The graph in Fig. 5 presents the K-level improvement of each construction joint subset. For each subset, there were 3 data entries, one for each housing configuration (detached, semi-detached, terraced). Due to the fact that the original façade-intermediate floor (FF) joints are already quite well insulated, the improvement in thermal transmittance coefficient is small, resulting in no K-level improvement in all periods. For this reason, it isn’t included in Fig. 5. Not that the window joint refurbishment has a larger impact than other joint refurbishments. For the pre46 period, the impact is the largest with an average of 4,0±0,0 K-level points. Another large impact can be seen fro m refurbishment of the wall base in the 46-70 period, which is on average 2,7±0,6 K-level points. Roof and intermediat e floor refurbishments have only a small impact. Independent Samples Kru skal-Wallis test [55 p. 157-164] showed that there is indeed a significant difference in K-level improvement in the pre46, 46-70 and 91-05 periods. In the 71-90 period, no joint subset refurbishment was significantly different from the others. A linear regression analysis [55 p.190-208] was performed to evaluate the link between joint length and K-level improvement per time period. No such link was found, in dicating that all improvements of K-levels originate fro m improvements in linear thermal transmittance coefficients of the joints.


Fig. 5: K-level difference per construction joint subset refurbishment

4.2. Net energy demand An analysis of the net energy demand was also performed. The results of this analysis were very similar to the results of the K-level analysis. The NED in the RJ situation was 71.9±8.0 kWh/(m²·yr), coming form 210,0±63,5 kWh/(m²·yr) in the original situation. There was an average difference of 5,8±0,9 kWh/(m²·yr) between the NED reduction from the original to the refurbished with non-refurbished joints on the one side, and the NED reduction from the original to the refurbished with refurbished joints on the other side, which was proven not to be significant. The subset simulations explained in paragraph 4.1 were repeated and noted and compared the net energy demands resulting from each energy simulation. Again, a much larger impact of the window joint refurbishment compared to other joint refurbishments could be seen. The only exception was once again the wall base refurbishment in the 46-70 period. The differences were shown to be significant for all time periods. Again, no link between improvement in net energy demand and length of the joint was found. 5. Conclusion and research perspectives Although refurbishment of construction joints provided a drop in K-level of 4 points on average, and a drop in NED of almost 6 kWh/(m²·yr) on average, this difference was not statistically significant. Based on K-level and NED, refurbishment of window joints had a significantly larger influence than other joint refurbishments, especially in the pre46, 46-70 and 91-05 periods. Refurbishment of wall base joints proved to have a larger influence than other refurbishments in the 46-70 period. This last conclusion indicates that a reduction in K-level and NED stem from a reduction in linear transmission coefficient of the joints. This contradicts what seems to be logical, namely that refurbishing construction joints which have a greater length also has a greater impact on thermal insulation and building energy consumption. A remark that must be made is that it might be possible to draw more statistically significant conclusions when using the raw building data. The typological approach namely reduces the sample size from almost 11000 real buildings to 12 archetypal buildings. Unfortunately, the raw building data was not accessible for this research.

167

168


It should also be noted that the construction joint typology is based on literature review. Further investigation on whether the construction methods found in literature correspond to genuinely constructed joints is going on. The results in this paper concern the influence of construction joints on energy demand and t hermal insulation in buildings. Further research is needed on the influence these joints have on comfort and mold, as well as the cost of refurbishment of these joints. When these subjects are known, an effective cost -benefit analysis can be conducted.

References M. Economidou, “Europe’s buildings under the microscope: A country-by-country review of the energy performance of buildings,” 2011. R. Janssen, “ T owards Energy Efficient Buildings in Europe Final Report Towards Energy Efficient Buildings,” 2004. Algemene Directie Statistiek, “Census 2011 - Woningen.” 2014. E. Cuerda, M. Pérez, and J. Neila, “ Facade typologies as a tool for selecting refurbishment measures for the Spanish residential building stock,” Energy Build., vol. 76, pp. 119–129, Jun. 2014. [5] E. G. Dascalaki, K. G. Droutsa, C. a. Balaras, and S. Kontoyiannidis, “ Building typologies as a tool for assessing the energy performance of residential buildings - A case study for the Hellenic building stock,” Energy Build., vol. 43, no. 12, pp. 3400–3409, 2011. [6] P. Schwehr and R. Fischer, “ Building T ypology and Morphology of Swiss Multi-Family Homes 1919 - 1990,” Lucerne, 2010. [7] Institut Wohnen und Umwelt, “ Deutsche Gebäudetypologie - Systematik und Datensätze,” Darmstadt, Germany, 2005. [8] C. Cyx, N. Renders, M. Van Holm, and S. Verbeke, “ IEE TABULA - Typology Approach for Building Stock Energy Assessment,” Mol, Belgium, 2011. [9] BSRIA, “ iNSPiRe - Simulation Results of Reference Buildings,” 2014. [10] H. Kauko, M. J. Alonso, O. Stavset, and I. C. Claussen, “ Case Study on Residential Building Renovation and its Impact on the Energy Use and T hermal Comfort,” Energy Procedia, vol. 58, pp. 160–165, 2014. [11] H. Lahmidi and F. Leguillon, “ Study for the ASIEPI project: Thermal bridges influence on the primary energy consumption.” [12] WT CB, “ Isolatie, Ventilatie en Verwarming in nieuwbouwwoningen - Resultaten van een enquête,” 1999. [13] H. Erhorn-Kluttig and H. Erhorn, “Impact of thermal bridges on the energy performance of buildings,” P148, 2009. [14] L. Du, T . Prasauskas, V. Leivo, M. Turunen, M. Pekkonen, M. Kiviste, A. Aaltonen, D. Martuzevicius, and U. Haverinen -Shaughnessy, “ Assessment of indoor environmental quality in existing multi-family buildings in North–East Europe,” Environ. Int., vol. 79, pp. 74–84, 2015. [15] “ IEE Project Episcope.” [Online]. Available: http://episcope.eu/welcome/. [Accessed: 01-Aug-2016]. [16] “ Build Up T hermal Bridges forum.” [Online]. Available: http://www.buildup.eu/en/topics/thermal-bridges-forum. [Accessed: 01-Aug-2016]. [17] Vlaams Energieagentschap, “EPB Software 6.5.1.” 2015. [18] Lawrence Berkeley National Laboratory, “THERM 7.3.” 2015. [19] Bureau voor Normalisatie, “NBN EN ISO 10221: Thermal bridges in building construction — Heat flows and surface temperatures — Detailed calculations,” 2007. [20] Vlaamse regering, Bijlage 1 bij het besluit van de Vlaamse Regering houdende wijziging van het Energiebesluit van 19 november 2010 , wat betreft aanpassingen aan diverse bepalingen inzake de energieprestatieregelgeving (Dutch). 2016. [21] M. Kottek, J. Grieser, B. Beck, B. Rudolf, and F. Rubel, “ World map of the Köppen -Geiger climate classification updated,” Meteorologische Zeitschrift, 2006. [Online]. Available: http://koeppen-geiger.vu-wien.ac.at/present.htm. [Accessed: 01-Aug-2016]. [22] “ T ABULA Webtool,” 2015. [Online]. Available: http://webtool.building-typology.eu/#bm. [Accessed: 29-Mar-2016]. [23] J. G. Wattjes, Constructie van gebouwen: Deel 1, Second ed. Amsterdam: NV Uitgeversmaatschappij Kosmos, 1929. [24] J. G. Wattjes, Constructie van gebouwen: Deel 4, Second ed. Amsterdam, T he Netherlands: Uitgeverij Kosmos, 1932. [25] J. G. Wattjes, Constructie van gebouwen: Deel 9, Third ed. Amsterdam, The Netherlands: Uitgeverij Kosmos, 1930. [26] J. Oosterhoff, G. J. Arends, C. H. Van Eldik, and G. G. Nieuwmeijer, Bouwtechniek in Nederland: Constructies van ijzer en bet on, First edit. Delft: Delftse Universitaire Pers, 1988. [27] F. L. Berghuis and G. A. Scholten, Handboek voor bouwkundigen: Kappen, daken en verdere T immerwerken, Second edi. Leiden: A.W. Sijthoff, 1903. [28] F. L. Berghuis, Handboek voor Bouwkundigen: Kappen, Daken en verdere T immerwerken. Antwerpen, Belgium: De Nederlandsche Boekhandel. [29] J. Floribert, Cours de construction, First edit. Schaerbeek: St. Lucasschool, 1922. [30] G. A. Scholten, Handboek voor Timmerlieden, Opzichters en Aanstaande Architecten: Atlas van 46 platen, Second edi. ’s Gravenshage, The Netherlands: De Gebroeders Van Cleef, 1903. [31] L. Cloquet, T raité d’architecture: T ome premier, First ed. Liège, Belgium: Librairie polytechnique, Baudry et Cie, 1898. [32] L. Cloquet, T raité d’architecture: T ome Second, First ed. Liège, Belgium: Librairie polytechnique, Baudry et Cie, 1898. [33] L. Cloquet, T raité d’Architecture: Tome quatrieme, First ed. Liège, Belgium: Librairie Polytechnique Ch. Béranger, 1900. [34] C. Boer, Opbouw en afwerking van gebouwen, First ed. Amsterdam, The Netherlands: NV Uitgeverij Argus, 1958. [35] J. J. Vriend, Bouwen: Handboek voor de praktijk van het bouwen. Amsterdam, The Netherlands: NV Uitgeversmaatschappij Kosmos, 1955. [36] J. J. Vriend, Bouwen: Handboek voor de praktijk van het bouwen. Amsterdam, The Netherlands: NV Uitgeversmaatschappij Kosmos, 1953. [37] S. J. Van den Oever, Vorming en techniek: Constructieleer, Fifth ed. Haarlem, The Netherlands: NV Uitgeverij V/H A. Kemperman, 1965. [1] [2] [3] [4]

Stijn Van Craenendonck et al. / Energy Procedia 96 (2016) 158 – 169 [38] M. Mittag, Architekturdetails, Second edi. Detmold, Gemany: C. Bertlesmann Verlag, 1959. [39] M. Mittag, Baukonstruktionslehre, First edit. Güterloh, Germany: C. Bertlesmann Verlag, 1954. [40] M. Jacobs, A. Van T ol, H. Uffen, C. G. Engelsman, and C. Punselie, Algemene bouwkunde, Fifth edit. Delft: Uitgeverij Waltman, 1978. [41] H. Banz, Baukonstruktions - Details, First edit. Stuttgart, Germany: Heinrich Fink KG, 1973. [42] M. Dietens, Constructie van gebouwen: Deel 1, Third ed. Gent, Belgium: E. Story Scientia, 1978. [43] Jellema, Bouwkunde voor het hoger technisch onderwijs: Deel 2. Delft, The Netherlands: Uit geverij Waltman, 1971. [44] Jellema, Meischke, and Muller, Bouwkunde voor het hoger technisch onderwijs: Deel 9a, Third ed. Delft, The Netherlands: Uitge verij Waltman, 1977. [45] Jellema, Meischke, and Muller, Bouwkunde voor het hoger technisch onderwijs: Deel 9b, Fourth ed. Delft, The Netherlands: Uitgeverij Waltman, 1977. [46] N. Janssen and U. Janssen, Bauzeichnung und Architekturmodell, Fourth edi. Stuttgart, Germany: Karl Krämer Verlag, 1985. [47] Stichting Bouwresearch, Detaillering van gesloten dakbedekkingen, First ed. Deventer, The Netherlands: Kluwer T echnische boeken BV, 1981. [48] WT CB, “TVN 188 - Plaatsen van buitenschrijnwerk,” 1993. [49] WT CB, “TVN 189 - Dekvloeren,” 1993. [50] WT CB, “TVN 195 - Daken met natuurleien: Opbouw en uitvoering,” 1995. [51] WT CB, “TVN 196 - Balkons,” 1995. [52] WT CB, “TVN 202 - Daken met betonpannen,” 1996. [53] WT CB, “TVN 215 - Het platte dak: Opbouw, materialen, uitvoering, onderhoud,” 2000. [54] WT CB, “TVN 219: Dakbedekkingen met leien: Dakdetails, opbouw en uitvoering,” 2001. [55] WT CB, “TVN 175 - Daken met pannen in gebakken aarde: Opbouw - uitvoering,” 1989. [56] WT CB, “TVN 178 - Termische isolatie van gevels,” 1989. [57] “ T ransmissiereferentiedocument,” Belgisch Staatsblad, Brussels, Belgium, pp. 74848–74936, Dec-2010. [58] Vlaamse EnergieAgentschap, “ EPB-eisen vanaf 2016,” 2015. [59] J. H. McDonald, Handbook of Biological Statistics, 3rd ed. Baltimore, Maryland, 2015.

169