Energy performance gap in building retrofit : characterization and effect on the energy saving potential

(1)

Proceedings Chapter

Reference

Energy performance gap in building retrofit : characterization and effect on the energy saving potential

KHOURY, Jad, HOLLMULLER, Pierre, LACHAL, Bernard Marie

Abstract

The aim of this study is to characterize the energy performance gap in retrofit of multi-family residential buildings, and to assess its potential impact on the energy savings of the Geneva post-war building stock. In a first step, we analyse 10 recently retrofitted case studies, covering around 1'100 flats. For each retrofit, the theoretical and actual energy savings for space heating are calculated on the basis of: (i) measured final energy demand for heating (SH and DHW), before and after retrofit; (ii) design value for SH after retrofit, according to the SIA 380/1. As a major result of the study, a robust statistical correlation between theoretical and actual energy savings allows to characterize the energy performance gap. In a second step, this result is used to assess a realistic energy saving potential for Geneva's multifamily building stock. This assessment shows that, under current practice, only 42% of the theoretical energy saving potential of building retrofit could be achieved. Finally, the main reasons behind this gap are discussed, as well as its potential effect on the goals of the Energy Strategy 2050.

KHOURY, Jad, HOLLMULLER, Pierre, LACHAL, Bernard Marie. Energy performance gap in building retrofit : characterization and effect on the energy saving potential. In: 19.

Status-Seminar «Forschen für den Bau im Kontext von Energie und Umwelt». 2016.

Available at:

http://archive-ouverte.unige.ch/unige:86086

Disclaimer: layout of this document may differ from the published version.

1 / 1

(2)

19. Status-Seminar «Forschen für den Bau im Kontext von Energie und Umwelt»

Energy performance gap in building retrofit :

characterization and effect on the energy saving potential

Jad Khoury, Pierre Hollmuller, Bernard Lachal

University of Geneva, Energy group, Institute for Environmental Sciences & Institute Forel, Section of Earth and Environmental Sciences, www.unige.ch/energie

Contact : Jad Khoury, [email protected]

Zusammenfassung Résumé Abstract

Abstract

The aim of this study is to characterize the energy performance gap in retrofit of multi-family residential buildings, and to assess its potential impact on the energy savings of the Geneva post-war building stock. In a first step, we analyse 10 recently retrofitted case studies, covering around 1’100 flats. For each retrofit, the theoretical and actual energy savings for space heating are calculated on the basis of: (i) measured final energy demand for heating (SH and DHW), before and after retrofit; (ii) design value for SH after retrofit, according to the SIA 380/1. As a major result of the study, a robust statistical correlation between theoretical and actual energy savings allows to characterize the energy performance gap. In a second step, this result is used to assess a realistic energy saving potential for Geneva’s multifamily building stock. This assessment shows that, under current practice, only 42% of the theoretical energy saving potential of building retrofit could be achieved. Finally, the main reasons behind this gap are discussed, as well as its potential effect on the goals of the Energy Strategy 2050.

Résumé

Le but de cette étude est de caractériser l’écart de performance énergétique lié à la rénovation des bâtiments résidentiels collectifs, et d’évaluer son impact sur le potentiel d’économie d’énergie du parc de bâtiments d’après-guerre à Genève. En un premier temps, nous analysons 10 études de cas de rénovations récentes, totalisant environ 1'100 appartements. Pour chacune de ces rénovations, les économies d’énergie théorique et réelle ont été calculées sur la base de : (i) la demande d’énergie finale pour la production de chaleur (chauffage et ecs), mesurée avant et après rénovation ; (ii) la demande prévue de chauffage, calculée selon la norme SIA 380/1. Un résultat majeur de cette étude consiste en une caractérisation statistique de l’écart de performance, via une corrélation robuste entre économies théorique et réelle. En un deuxième temps, ce résultat est utilisé pour évaluer le potentiel réaliste d’économie d’énergie du parc résidentiel collectif du Canton de Genève. Il en ressort que, avec les pratiques actuelles, seuls 42% des économies théoriques liées à la rénovation des bâtiments pourraient être atteintes. Enfin, les principales raisons de cet écart sont discutées, ainsi que sa répercussion possible sur l’atteinte des objectifs de la Stratégie Energétique 2050.

(3)

1. Context and scope of the study

The Swiss Energy Strategy 2050 (ES2050) strongly relies on massive reduction of the heat demand for the building sector, with target energy savings of 45% by 2035, and 64% by 2050.

Given the slow transformation of the building stock, a strong focus should be set on building retrofit.

A major issue for succeeding with the ES2050 targets is to achieve a significant retrofit rate (quantity) and a high retrofit performance (quality). However, previous research conducted in Switzerland and elsewhere, mainly on new buildings, indicates a serious performance gap between design and actual energy demand in real-life condition of execution, operation and use.

Although the performance gap has recently gained interest in Switzerland, many questions still remain to be answered.

The main goals of this study are: (i) to characterize the energy performance gap in building retrofit, on the basis of a representative set of case studies; (ii) to assess its effect on the energy saving potential of the existing building stock. Focus is set on multifamily buildings in the canton of Geneva which are responsible for almost half of the thermal energy consumption of the canton as well as half of the CO2 emissions of the entire building stock.

This work is part of the SFOE Compare-Renove research project as well as of the SCCER Future Energy Efficient Buildings & Districts (FEEB&D).

2. Energy performance gap: definition and methodology

The notion of 'energy performance gap' is generally defined in as the difference between the measured and the expected energy performance of buildings (Struck et al., 2014; de Wilde, 2014;

de Wilde & Jones, 2014; Menezes et al., 2012; Zero Carbon Hub, 2013; van Dronkelaar et al., 2016; BFE 2016). Whereas most research has so far focused on new buildings, we will in this paper examine the performance gap linked to building retrofit, in particular of large multi-family buildings. We will for this sake define and examine two types of performance gap: the energy demand gap and the energy saving gap, latter specific to building retrofit (Figure 1).

Standard use conditions according to SIA 380/1

Limit value (Qh,li retrofit)

Different energy saving types

Qh norm,be

Qh norm,af

BEFORE retrofit

AFTER retrofit

Normative savings

Qh real,be

Theoretical savings

Actual savings Qh real,be

Qh norm,af

Qh real,af

Prebound effect

Rebound effect Real use conditions

Standard use conditions according to SIA 380/1

Real use conditions

Figure 1. Performance gap typology in building retrofit (see text for definition)

(4)

Definition of performance gap

The first type of performance gap, called energy demand gap, represents the gap between the space heating demand (Qh norm) calculated under standard conditions of use as defined in SIA 380/1 and the one derived from real consumption (Qh real) under real-life conditions of use and operation. These notions, presented in Figure 1, are defined as follows:

 Qh norm,be : Space heating demand before retrofit, calculated according to SIA380/1

 Qh norm,af : Space heating demand after retrofit, calculated according to SIA380/1

 Qh real,be : Space heating demand before retrofit, measured

 Qh real,af : Space heating demand after retrofit, measured

The second type of performance gap is the energy saving gap. This gap represents the difference between the expected and achieved savings that result from the energy retrofit of the existing building stock. In the present study, we distinguish between three types of energy savings, as presented below:

 Normative savings: ∆Qh norm = Qh norm,be - Qh norm,af

 Theoretical savings: ∆Qh theor = Qh real,be - Qh norm,af

 Actual savings: ∆Qh real = Qh real,be - Qh real,af

Normative savings are defined as the difference between the space heating demand before and after retrofit, calculated with the aid of a simulation software under standard conditions of use in accordance with SIA 380/1 (for example with a room temperature of 20°C).

Theoretical savings are considered as the difference between the real space heating demand before retrofit and the expected demand after retrofit, stated in the building permit and calculated under normal conditions of use according to SIA standards.

Actual savings are defined as the difference between the real space heating demand of a building before and after retrofit under real-life conditions of use and operation (for example with a room temperature of 22-23°C).

Figure 1 also shows two phenomena described by Sunikka-Blank and Galvin (2012). The first is the 'prebound' effect, when the calculated energy consumption before retrofit is higher than the real measured value, mainly due to inaccurate assumptions of the situation before retrofit (Majcen, 2016; Hoffmann and Ménard, 2015). The second is the well-known rebound effect, which can occur after a retrofit, when the real value is higher than the calculated value (Sorrell et al., 2009;

Haas and Biermayr, 2000; Greening et al., 2000; Berkhout et al., 2000). Reported reasons for the rebound effect include increased internal temperature or higher air exchange rate compared to the design values.

Methodology

In the forthcoming analysis, the expected space heating demand after retrofit (Qh norm,af) is given by the documents contained in the building permit requests of the analysed case studies. Note that the calculated demand before retrofit (Qh norm,be) is rarely available in these documents. As pointed out before, the latter is also considered too imprecise, and often overestimates the real demand before retrofit.

The real space heating demand before and after retrofit (Qh real,be ; Qh real,af) are derived from the final energy consumption of each building, during 3 consecutive years before and after retrofit.

The calculation of the effective space heating demand includes (Khoury, 2014):

 Taking into account of the efficiency of the heat production system, estimated as a function of its type (oil boiler, gas boiler, district heating, etc.) and its date of commissioning.

 Deduction of the DHW demand, estimated as a function of the number of occupants, by way of an average DHW demand of 1080 kWh/occupant/year (based on benchmarking of measured DHW demand of 65 residential buildings in Geneva).

 Application of a climatic correction (2659 degree-days, 18/12 basis).

(5)

In the absence of data on the calculated heat demand before retrofit (Qh norm,be) and given the uncertainty associated with the quality of such data, when it exists, we use the theoretical savings instead of normative savings to evaluate the expected savings resulting from retrofit case studies.

Finally, the gap between theoretical savings and actual savings is defined as the energy saving performance gap.

3. Retrofit case studies, main characteristics and heat demands

Criteria for selecting the case studies

The case studies were selected from the retrofit building permit requests which were submitted to the Energy Office of Canton Geneva, between March 2004 and May 2012. The different stages of the selection process are detailed by Khoury (2014) and are summarized in Table 1.

Selection criteria Description

Building category: Multi-family residential buildings

Type: 100% residential (or with commercial at ground floor level) Construction period: 1946-1980 (post-war period)

Retrofitted area: Energy reference area greater than 1500 m²

Year of retrofit: Between 2005 and 2009 (before the introduction of the new law on energy in Geneva in August 2010)

Nature of the retrofit: Energy-related retrofit of the thermal envelope, with or without retrofit of the heat production system, and without creation of additional dwellings / habitable areas.

Standard: Retrofit in accordance with SIA 380/1 (case of a justification by global performance), with or without Minergie labelling

Availability of information: Cases for which the energy related information of the building permit is available.

Table 1. Selection criteria of the case studies

Main characteristics of retrofit case studies

In total, 10 retrofit case studies were selected. These cases involve 50 residential multifamily buildings (1 building = 1 entrance), comprising about 1’100 dwellings and covering a total energy reference area of approximately 110’000 m². As will be seen, these buildings are fairly representative of those constructed during the post-war period (1946-1980) in the canton of Geneva and more generally in Switzerland, and also offer the most important energy saving potential.

Table 2 on the following page provides a summary of the general characteristics for these case studies. Photos of these buildings are presented in Figure 2.

(6)

Retrofit case studies B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 Year of construction 1961-70 1961-70 1961-70 1947-60 1961-70 1961-70 1961-70 1971-75 1971-75 1961-70

Year of last retrofit 2008 2008 2008 2007 2007 2008 2007 2007-09 2006 2009

Minergie label (<2009) - - -  -  - - - -

Location (municipality) Geneva Meyrin Meyrin Carouge Geneva Onex Chêne-Boug. Geneva Meyrin Geneva

Number of buildings (1) 1 11 4 2 4 2 14 5 4 3

Number of dwellings 28 153 96 60 160 63 222 208 96 45

Number of heated floors 7 4-5 7 7 9 9 6-7 7-11 12 8

Heated area ERA [m²] 1 640 14 210 7 560 3 846 14 016 5 357 27 709 19 971 9 708 7 518

Retrofit of the building envelope

nearly

complete complete nearly

complete complete facade complete nearly

complete partial partial partial

Roof -    -   - - -

Exterior wall         - -

Windows        -  

Ground   -  -  - - - -

Retrofit of the technical

equipment yes yes no yes no yes no no no no

Production   -  -  - - - -

Distribution   -  -  - - - -

Ventilation - - -  -  - - - -

Lighting (common)  - - - -  - - - -

Space heating demand (Qh)

Qh real,be [MJ/m²/yr] 708 662 561 455 448 435 422 405 402 332

Qh real,af [MJ/m²/yr] 341 327 387 283 301 286 355 349 341 291

(1) Number of buildings = number of entrances

Table 2. General characteristics of the retrofit case studies (10 case studies, 50 buildings)

B1. B2. B3. B4. B5.

B6. B7. B8. B9. B10.

Figure 2. Photos of the case studies

Space heating demand and comparison with the Geneva multi-family building stock

Figure 3 shows the distribution of space heating demand of the multi-family residential building stock in Geneva in 2010, grouped according to the three major construction periods, which results from a detailed survey on this matter (Khoury, 2014). The latter study mainly used the Geneva SITG database, which contains the measured final energy consumption for heating (SH and DHW) as well as the associated heated area of the multi-family residential buildings of the Canton.

The blue diamonds indicate the real space heating demand before retrofit (Qh real,be) of our 10 case studies. For 9 out of the 10 cases, the space demand before retrofit exceeds the median value of the multi-family residential buildings constructed between 1946 and 1980 (363 MJ/m²/yr).

Out of them, 7 cases exceed the 3^rd quartile (411 MJ/m²/yr), and 3 exceed the 9^th decile (468 MJ/m²/yr). The case studies are hence fairly representative of the buildings that need to be retrofitted in priority over the next few years.

(7)

100 200 300 400 500 600 700 800

0 2 4 6 8 10 12 14 16 18

Energy reference area [million m²]

Qh distribution (90th, 7...) Qh real before retrofit Qh real after retrofit Qh norm 380/1 after retrofit Qh limit retrofit SIA 380/1 Qh distribution (9th decile, 3rd quartile, median, 1st quartile and 1st decile)

B5

Construction periods

before 1946 1946 - 1980 1981 - 2010

344 363

281

Minergie-P:2009 rénovation 344 363

281

Qh,li: SIA retrofit (ed. 2009) Minergie-P retrofit (ed. 2009) Qh,li: SIA retrofit (ed. 2007)

B1 B2

B3

B4 B6B7 B8

B9

B10

Space heating demand [MJ/m²/yr]

0 100 200 300 400 500 600 700 800

0 2 4 6 8 10 12 14 16 18

Qh distribution (90th, 75th, 50th, 25th and 10th percentiles) Qh real before retrofit Qh real after retrofit Qh norm 380/1 after retrofit Qh limit retrofit SIA 380/1

B5

before 1946 1946 - 1980 1981 - 2010

344 363

281

Minergie-P:2009 rénovation 344 363

281

Qh,li: SIA retrofit (ed. 2009) Minergie-P retrofit (ed. 2009) Qh,li: SIA retrofit (ed. 2007)

B1 B2

B3

B4 B6B7 B8

B9

B10

Figure 3. Space heating demand of all multifamily residential buildings in the canton of Geneva in 2010, grouped according to the three major construction periods, as well as real and expected space heating

demand of the 10 case studies (values normalized at 2659 DJ/year). The limit value Qh,li for retrofit in accordance with standard SIA 380/1 (ed. 2007 and 2009) and Minergie-P (ed. 2009) is also indicated.

After retrofit, the expected space heating demand (Qh norm,af) calculated under normative conditions of use in accordance with standard SIA 380/1 (ed. 2007) is indicated by red triangles.

The yellow squares indicate the real demand of these buildings after retrofit (Qh real,af). Finally, the limit values Qh,li for retrofit defined by the SIA 380/1 (ed. 2007 and 2009) is also indicated, as well as the energy requirement for the building envelope according to the Minergie-P retrofit label (edition 2009). These limit values were calculated using an envelope factor of 0.7, as averaged over the 10 analysed case studies (50 buildings).

Note that the real space heating demand after retrofit is systematically higher than the one anticipated by calculation. This finding will be analysed and discussed in the next section.

4. Expected and actual energy performance of retrofit case studies

The analysis of performance gap will first focus on the energy demand performance gap, then on the energy saving performance gap.

Energy demand performance gap

Figure 4 presents the real space heating demand of the buildings before and after retrofit (Qh real,be & Qh real,af) as well as the expected demand, as calculated under normal conditions of use (Qh norm,af).

(8)

0 100 200 300 400 500 600 700 800

B1 B2 B3 B4 B5 B6 B7 B8 B9 B10

Qh real before retrofit Qh real after retrofit Qh norm 380/1 after retrofit

Actual and Theoretical energy savings (SH)

Figure 4. Real and expected space heating demand, before and after retrofit

We first note that the real space heating demand after retrofit is systematically higher than the expected, calculated value. Nonetheless, the real demand of cases B6 and B4, which are retrofitted to the Minergie label, are lower than the demand of the other buildings, which are retrofitted to the SIA 380/1 standard. This highlights the attraction effect of the Minergie label, a high-performance energy standard, which is also confirmed in the case of new buildings (Zgraggen, 2010).

After retrofit, the real space heating demand of the analysed building is between 283 and 387 MJ/m²/yr, with an average of 326 MJ/m²/yr. The relative difference between real values (Qh real,af) and expected values (Qh norm,af) varies between 43% and 142%, with the exception of case B4 which exceeds the expected objective by 310%. The average difference amounts to 116% and drops to 94% if case B4 is not included. These values are fairly close to the ones observed on new residential buildings (Zgraggen, 2010). In latter study, which concerns 10 buildings also situated in Geneva, the real demand (space heating and DHW) exceeds the expected value by an average of 70%, with variations between 30% and 120%.

These results are also in line with a recent study conducted on households in the Netherlands, using a database of around 100’000 buildings, which shows that buildings with a high energy efficiency rating tend to consume more than expected (Majcen, 2016), and with an ex-post evaluation of around 100 retrofitted single-family houses in France which reveals a discrepancy between the real consumptions and the ones simulated by an ex-ante model (Raynaud, 2014).

Energy saving performance gap

Figure 5 presents the actual and theoretical space heating savings of the 10 case studies (bars), as well as the ratio between them (line, secondary axis).

Since the calculated space heating demand after retrofit (Qh norm,af) is in most cases very close to the limit value Qh,li (see Figure 3), the theoretical savings depend mainly on the real heat demand before retrofit. For our case studies, the theoretical savings hence vary between 140 and 560 MJ/m² per year, according to the space heating demand before retrofit.

(9)

65% 64%

52%

45% 45% 47%

34% 35%

29% 29%

0%

10%

20%

30%

40%

50%

60%

70%

80%

0 100 200 300 400 500 600 700 800

B1 B2 B3 B4 B5 B6 B7 B8 B9 B10

Space heating savings [MJ/m²/yr]

Theoretical savings Actual savings

Ratio between actual and theoretical savings (%)

Figure 5. Theoretical and actual savings of space heating demand in multifamily building retrofits

With regard to the actual savings, the top ranking goes for cases 1 and 2 (366 and 336 MJ/m²/yr respectively), which were both conducted in accordance with the SIA standard. Cases 4 and 6, which aim to achieve the Minergie standard, arrive in 4^th and 5^th positions, with a reduction in space heating demand of 172 and 148 MJ/m²/yr respectively. It follows from this that the savings from a standard retrofit which complies with the SIA 380/1 standard may be higher than those achieved by a retrofit project based on the Minergie standard. This, of course, depends on the real value before retrofit. At the bottom of the rankings, cases 8, 9 and 10 concern partial energy retrofit (when only one element of the building envelope is retrofitted), with savings between 40 and 62 MJ/m²/yr.

In conclusion, the achieved fraction of the theoretical savings are in the range of 29% to 65%, with an average of 44% (Figure 5, secondary axis). This percentage reaches 65% and 64% for cases 1 and 2, which are implemented in accordance with the SIA standards, whereas the Minergie cases 4 and 6 come in later, with 45% and 47%.

Correlation between theoretical savings and actual savings

For the 10 analysed cases, the comparison between theoretical and actual savings of space heating (Figure 6) shows a fairly strong statistical relationship (R²=0.9897), which can be expressed as follows:

∆Qh real = 0.0009 * ∆Qh theor² + 0.17 * ∆Qh theor

Where : ∆Qh real (actual savings) = Qh real,be - Qh real,af, (MJ/m²/yr)

∆Qh theor (theoretical savings) = Qh real,be - Qh norm,af (MJ/m²/yr), between 100 and 600 MJ/m²/yr.

(10)

y = 0.0009x²+ 0.1701x R² = 0.9897

0 100 200 300 400 500 600

0 100 200 300 400 500 600 700

Actual energy savings [MJ/m²/yr]

Theoretical energy savings [MJ/m²/yr]

B1 B2

B4 B3

B5 B6

B9 B8 B7

B10

y= 0.0009x²+0.17x R² = 0.9897

Figure 6. Actual versus theoretical savings of space heating demand

This relation is established regardless of the nature of work carried out, and of the type of retrofit being considered (partial or complete retrofit, in accordance with SIA standards or aiming to achieve the Minergie label). It also shows that the higher the theoretical savings are, the higher the actual savings also are.

Hence, by aiming for theoretical savings in the order of 350 MJ/m²/yr, the actual savings will more probably be about 170 MJ/m²/yr, i.e. about half. However, by aiming for savings of 550 MJ/m²/yr, the actually achieved savings rise to about 365 MJ/m²/yr, in other words a realized fraction of almost 65%. On the other hand, this fraction falls to 35% when attempting to realize theoretical savings in the order of 200 MJ/m²/yr (observed for buildings partially retrofitted).

This relation takes into account the entire retrofit process, from the design stage (choice of solutions and use of simulation software) all the way through to the use of the buildings by the occupants and the operation by the energy managers. It can be viewed as a characterisation of the current retrofit and operation practices in the Canton, over a given period. Providing it is corroborated by additional case studies, this relation could be used as an average correction factor for the behaviour of users and entities involved in the retrofit, making it possible to estimate the actual savings based on theoretical savings.

5. Upscaling for Geneva multi-family building stock

Theoretical energy saving potential

For the existing multi-family residential building stock of the Canton Geneva, the theoretical saving potential of space heating demand is defined as the difference between the real space heating demand (Qh real,be) of each building that was derived from the Geneva SITG database, and the limit value to be achieved by retrofit (Qh,li).

The latter value depends on the intended objective (SIA 380/1 or Minergie-P, edition 2007 or 2009, see Figure 3) and on the building shape factor. Here, it is calculated with a shape factor of 0.7 (average value of the sample) and the requirements of SIA 380/1 for retrofit (edition 2009). Under these conditions, the expected space heating demand of retrofitted multi-family residential buildings should not exceed the limit of 104 MJ/m²/yr.

(11)

Therewith, the theoretical saving potential of the Geneva multi-family residential sector is estimated to 1’273 GWh/yr. When focusing on buildings from the post-war period (1946 -1980), which should undergo retrofit in a near future, this potential amounts to 677 GWh/yr, representing more than half (53%) of the total potential of this sector.

The results are shown in Figure 7 below.

0 100 200 300 400 500 600 700 800

0 2 4 6 8 10 12 14 16 18

before 1946 1946 - 1980 1981 - 2010

285 GWh/yr (realistic potential) 344 363

281

Qh,li: retrofit (SIA 380/1, 2009)

241 GWh/yr 19%

354 GWh/yr 28%

392 GWh/yr (untapped potential)

677 GWh/yr 53%

Qh real,be

Qh norm,af Qh real,af

Figure 7. Sorted space heating demand of the Geneva multifamily building stock, and realistic energy saving potential under current retrofit and operation practice

In order to take full advantage of the theoretical saving potential of post-war buildings (677 GWh/yr), the median space heating demand of these buildings (363 MJ/m²/yr) would need to be reduced by a factor of no less than 3.5. It is also evident from Figure 7 that the reduction objectives vary greatly from one building to another, depending on the situation before retrofit. For example, a building at the 9^th decile (space heating demand of 468 MJ/m²/yr) should have its demand reduced by a factor of 4.5, whereas this factor would drop to 2.7 for a building at the 1^st decile (276 MJ/m²/yr).

At this stage, it is worthwhile mentioning that unlocking retrofit potential of the entire stock presents difficulties which are not necessarily technical in nature, but rather linked to issues of costs and financing, organization and governance, as well as of conservation of heritage.

Realistic energy saving potential

If the performance of post-war buildings retrofitted during the period 2005-2009 are extrapolated to the entire stock, the realistic energy saving potential under current practice can be estimated by way of the correlation derived in section 4. This potential, which amounts to 285 GWh/yr, represents 42% of the theoretical potential calculated by way of the SIA 380/1 limit (cf. Figure 7, red hatched area).

This means that, even if the financial resources for retrofitting of these buildings are provided and all associated obstacles are removed, almost 60% of the theoretical potential will remain untapped unless current practices and use relating to building retrofit are evolving. Furthermore, since a low

(12)

theoretical saving potential results in an even smaller actual saving, the future retrofit of buildings with low space heating demand might become less significant.

In this regard, the primary issue not only concerns reducing of the space heating demand of buildings in order to satisfy ever more stringent energy requirements, but also to reduce the difference between theoretical and actual energy savings resulting from the retrofit.

6. Discussion: reasons behind the gap

There is a range of independent and interdependent factors which can explain why retrofitted buildings do not perform as well as predicted. These factors can occur at different stages of the building retrofit process. In this study, three parameter sets have been identified as the main causes for the observed discrepancy between theoretical and actual savings in building retrofit.

The underlying reasons are based upon experience and analysis of several real case studies such as the one conducted by Mermoud et al., 2012; Khoury et al., 2014 and Weber et al., 1991. These reasons are presented in Table 3 and are grouped as follows:

I. Inaccuracies due to the use of standard values according to SIA 380/1 II. Uncertainties in the input data used and model limitations

III. Other factors related to quality of execution, operation, measurements and user behaviour

Retrofit process

(selected phases) Main reasons behind the energy performance gap

Design phase I. Inaccuracies due to the use of standard values according to SIA 380/1 (input parameters related to the conditions of use)

- Occupancy (m²/pers)

- Electricity consumption (MJ/m²/yr) - Electricity factor reduction (-) - Utilization period (h/d) - Indoor temperature (°C) - Air flow rates (m³/m²/h)

II. Uncertainties in the input data used and model limitations - Design weather data (SIA 381/2, SIA 2028)

- Regulation type (per room, etc.) - Shading factor (Fs =Fs1+Fs2+Fs3)

- Calculation of surfaces (energy reference area, building envelope components, etc.)

- Thermal bridges, U values, b factors, etc.

Execution and operation phase

III. Other factors related to quality of execution, operation, measurements and user behaviour

- Poor quality of execution

- Design changes during the execution phase - Poor commissioning

- Malfunctioning / maladjustment of technical systems - Occupant behaviour

- Measurement limitations

- Uncertainties regarding the estimated DHW demand and the efficiency of heat production system, etc.

Table 3. Main reasons behind the energy performance gap during the design, execution and operation phase of the retrofit process

The first two sets of parameter have an impact on the energy model output (Qh norm, af), and thus on the calculated theoretical savings, during the design phase of the retrofit process; the third set has an impact on the real values during the execution and operation phase.

(13)

In fact, the use of SIA standard values for calculating the theoretical savings tend to overestimate the actual potential. This is particularly the case when we use the indoor air temperature (20°C) and ventilation rate values given by the SIA 380/1 (0.7 m³/m²/h for mechanical ventilation system without heat recovery) to calculate the space heating demand.

While the main goal of using the standard values according to SIA 380/1 is to compare on the same basis the calculated energy use of buildings with the limit and target values, the question remains whether it makes sense to use this standard method to estimate real energy savings at building level or to design thermal retrofit strategies and policies within the framework of the Swiss Energy Strategy 2050.

The reasons for the observed discrepancy may also be related to uncertainties in the input data used in the simulation and to model limitations (design weather data, energy reference area and surface of envelope components, shading factor, inaccurate U values or b factors, etc.), to factors related to the quality of execution, design changes during the execution phase, poor commissioning, technical problems (malfunctioning/maladjustment of heating system, etc.), as well as to user behaviour (both occupant and energy operator). In addition, other factors have to be taken into consideration during the operation phase, such as measurement limits/errors and uncertainties in the estimation of DHW demand and efficiency of the heat production system.

Regarding our case studies, the identification of the most determining factors behind the performance gap and the quantification of their relative importance via sensitivity analysis are still the subject of an ongoing study.

7. Conclusion

The aim of this study is to characterize the energy performance gap in building retrofit on the basis of a representative set of case studies and to assess its effect on the energy saving potential of the Geneva multi-family building stock. This sector is responsible for almost half of the thermal energy consumption of the canton as well as half of the CO2 emissions of the building stock.

In a first step, we analyse 10 recently retrofitted post-war multifamily building complexes, covering around 1’100 flats (with a total heated floor area of approximately 110’000 m²). For each retrofit, the theoretical and actual energy savings potential for space heating (SH) are calculated on the basis of: (i) real final energy demand for heating (SH and DHW), before and after retrofit; (ii) design values for SH after retrofit, according to the SIA 380/1. As a major result of the study, a robust statistical correlation between theoretical and actual energy savings allows to characterize the energy performance gap.

In a second step, this result is used to reassess a realistic energy saving potential for Geneva’s multifamily building stock. This assessment shows that, under current practices and even if all buildings would undergo low energy retrofit according to SIA standards, only 42% of the theoretical energy saving potential for space heating might be achieved in reality. Three parameter sets are identified in this paper as the main causes for the discrepancy observed: i) inaccuracies due to the use of SIA standard values in the method to calculate the theoretical savings, in particular the unrealistic indoor air temperature and ventilation rates; ii) uncertainties in the input data used for the simulation and model limitations (e.g. design weather data, regulation type, shading factor, calculated surfaces, etc.) and iii) other factors related to quality of execution, operation, measurements and user behaviour (both occupant and energy operator). The identification of the determining factors behind the performance gap and the quantification of their relative importance via sensitivity analysis are in progress and will be the subject of a separate paper.

In conclusion, while the main goal of using the standard method according to SIA 380/1 is to compare on the same basis the calculated energy use of buildings with the limit and target values, the question remains whether it makes sense to use this standard method to estimate real energy savings at building level, or to design thermal retrofit strategies and policies, for example within the framework of the Energy Strategy 2050 (ES2050).

(14)

In this regard, achieving the ambitious goals of the ES2050 set by the Federal government (particularly those that aim to reduce by half the space heating demand of the building sector by 2050) is hardly possible without reducing the observed gap between theoretical and actual energy savings. A better estimation of the actual savings is not only crucial for the authorities to design and implement effective energy efficiency strategies and policies for the building sector, but is also very important for tenants who are paying the expected savings on energy bills when buildings are retrofitted, for building owners who are undertaking deep energy retrofit and for other market players such as energy service companies (ESCOs) which are invited to have a key role to play in the coming years. Reaching these objectives also requires to improve the current practices of all the players involved in the building retrofit process and to avoid ruptures in the chain of responsibilities along the entire process, from the initial energy audit down to the use of buildings by the occupants and energy management people.

Finally note that this work would merit further development and consolidation with other retrofit cases conducted in Switzerland.

Acknowledgments

The authors would like to thank the authorities of Canton Geneva (OAC, OCEN, OCSTAT) for the provision of the data behind this study. The study was funded by SFOE and OCEN, in the frame of the Compare-Renove project, as well as by CTI, in the frame of the SCCER FEEB&D project (KTI.2014.0119).

References

[1] Struck, C., Benz, M., Dorer, V., Frei B., Hall, M., Menard, M., Moosberger, Sv., Orehounig, K., Sagerschnig C., Rochi, G., 2014. Performance Gap in der Schweiz – Brisanz, Ursachen und Einflüsse auf die Differenz von geplantem Energiebedarf und gemessenem Verbrauch in Gebäuden, 18. Status Seminar, ETH Zürich, September 2014.

[2] De Wilde, P., 2014. The gap between predicted and measured energy performance of buildings: A framework for investigation, Automation in Construction, Volume 41, Pages 40- 49.

[3] De Wilde, P., Jones, R. 2014. The building energy performance gap: Up close and personal.

Proceedings of the CIBSE ASHRAE Technical Symposium: Moving to a new world of building systems performance, Dublin, April 2014.

[4] Menezes, C., Cripps, A., Bouchlaghem, D., Buswell, R. 2012. Predicted vs. actual energy performance of non-domestic buildings: Using post-occupancy evaluation data to reduce the performance gap. Applied Energy, Volume 97, Pages 355-364.

[5] Zero Carbon Hub, 2013. Closing the Gap between design and as-built performance: new homes, Interim progress report, London, July 2013.

[6] Van Dronkelaar C., Dowson M., Spataru C., Mumovic D. 2016. A Review of the Regulatory Energy Performance Gap and Its Underlying Causes in Non-domestic Buildings, Frontiers in Mechanical Engineering, Volume 1, Article 17, January 2016.

[7] BFE 2016. Erfolgskontrolle Gebäudeenergiestandards 2014-2015, Bundesamt für Energie BFE, Sektion Gebäude, March 2016.

[8] Sunikka-Blank M., Galvin, R., 2012. Introducing the prebound effect: the gap between performance and actual energy consumption, Building Research & Information, Volume 40, Issue 3, Pages 260-273.

(15)

[9] Majcen, D., 2016. Predicting energy consumption and savings in the housing stock: A performance gap analysis in the Netherlands, PhD thesis, Delft University of Technology, OTB - Research for the Built Environment.

[10] Hoffmann C., Ménard M., 2015. Offene Fragen beim Einsatz der SIA 380/1 als

Prognoseinstru-ment bei Bestandsgebäuden, annual Report 2015 for the Swiss Federal Office of Energy (SFOE), November 2015.

[11] Sorrell S., Dimitropoulos J., Sommerville M., 2009. Empirical estimates of the direct rebound effect: A review, Energy Policy, Volume 37, Issue 4, Pages 1356-1371.

[12] Haas R., Biermayr, P., 2000. The rebound effect for space heating: Empirical evidence from Austria, Energy Policy, Volume 28, Issue 6, Pages 403-410.

[13] Greening L., Greene, D., Difiglio C., 2000. Energy efficiency and consumption - the rebound effect - a survey, Energy Policy, Volume 28, Issues 6-7, Pages 389-401.

[14] Berkhout P., Muskens J., Velthuijsen J., 2000. Defining the rebound effect, Energy Policy, Volume 28, Issues 6–7, Pages 425-432.

[15] Khoury, J. 2014. Rénovation énergétique des bâtiments résidentiels collectifs: état des lieux, retours d’expérience et potentiels du parc genevois, Thèse de doctorat, Université de

Genève. Institut des Sciences de l’Environnement et Institut Forel.

[16] Zgraggen, J.-M. 2010. Bâtiments résidentiels locatifs à haute performance énergétique:

objectifs et réalités, Thèse de doctorat, Université de Genève. Institut des Sciences de l’Environnement et Institut Forel.

[17] Raynaud, M. 2014. Evaluation ex-post de l’efficacité de solutions de rénovation énergétique en résidentiel, Thèse de doctorat, Ecole Nationale Supérieure des Mines de Paris, Centre Efficacité énergétique des Systèmes.

[18] Mermoud F., Khoury J., Lachal B., 2012. Suivi énergétique du bâtiment 40-42 de l'avenue du Gros-Chêne à Onex (GE), rénové selon le standard MINERGIE® Aspects techniques et économiques, Rapport pour le Service de l'énergie de l’Etat de Genève, Genève, Avril 2012.

[19] Khoury J., Mermoud F., Hollmuller P., Lachal B., 2014. Rénovation Minergie d’un bâtiment résidentiel de plus de 5’000 m² : bilan énergétique et financier. 18. Status Seminar, ETH Zürich, September 2014.

[20] Weber, W., Lachal, B., Gsponer, A., Guisan, O., 1991. Réhabilitation thermique d'un immeuble de 126 logements, construit en 1962 dans la banlieue de Genève. Conférence Internationale Energie solaire et bâtiment (CISBAT'91), Lausanne, Page 203.

Energy performance gap in building retrofit : characterization and effect on the energy saving potential

Proceedings Chapter

Reference

Energy performance gap in building retrofit : characterization and effect on the energy saving potential

Energy performance gap in building retrofit :

characterization and effect on the energy saving potential

1. Context and scope of the study

2. Energy performance gap: definition and methodology

3. Retrofit case studies, main characteristics and heat demands

4. Expected and actual energy performance of retrofit case studies

5. Upscaling for Geneva multi-family building stock

6. Discussion: reasons behind the gap

7. Conclusion

Acknowledgments

References

Other links