Energy and Buildings 42 (2010) 13141319

Contents lists available at ScienceDirect

Energy and Buildings

journal homepage: www.elsevier.com/locate/enbuild

A performance comparison of passive and low-energy buildings

Ardeshir Mahdavi , Eva-Maria Doppelbauer
Department of Building Physics and Building EcologyVienna University of Technology, Austria

a r t i c l e

i n f o

a b s t r a c t
This paper compares apartments in two residential blocks in Vienna; one passive and the other one lowenergy. These blocks were constructed simultaneously in the same location and with comparable building construction features and oor plans. The main difference between the two blocks (other than the higher thermal insulation level in the passive building) lies in the ventilation system: passive buildings deploy controlled ventilation, whereas the low-energy buildings rely mostly on user-operated natural (window) ventilation. We measured indoor environmental conditions (indoor air temperature, relative humidity, and CO2 concentration) in two units of each block over a period of ve months. Additionally, the buildings were compared in view of operation and embodied energy use, CO2 emissions, and construction costs. 2010 Elsevier B.V. All rights reserved.

Article history: Received 24 February 2010 Accepted 24 February 2010 Keywords: Passive house Low-energy house Indoor climate Energy performance

1. Introduction While there is a general consensus that buildings energy use and environmental impact must be reduced, there have been many discussions in the recent years as to the proper means and ways of achieving this. Specically, the relative advantages and disadvantages of low-energy, passive, and energy-plus building approaches have been at the center of much debate. There are varying denitions for the terms passive house and lowenergy house in different countries. In Austria, a building may be declared as a low-energy house if its annual heating demand (HWB in kWh m2 a1 ) is equal or less than 17(1 + 2.5/lc ), wherein lc denotes the characteristic length (heated volume divided by the surface area of the building) [1]. Passive houses have not yet been conclusively dened, but certain benchmarks have been suggested: a passive house is considered to be a very-low-energy house with a HWB of less than 10 (1 + 2.5/lc ). Moreover, a passive house aims at doing without a dedicated heating system, which would require HWB values under 10 kWh m2 a1 and a highly tight envelope (n50 < 0.6 h1 ). Another denition [2] postulates that a passive houses annual heating demand must be equal or less than 15 kWh m2 a1 , the heating load must be equal or less than 10 W m2 , the primary energy use must be equal or less than 120 kWh m2 a1 and the tightness level (n50) must be equal or less than 0.60 h1 . To compare passive and low-energy buildings objectively, reliable information regarding their actual performance (energy, indoor climate, environmental impact, cost) would be more helpful

than ideological pronouncements. Thus, the present papers contribution to this discussion circles around a comparison of low-energy and passive buildings based on empirical data. Toward this end, we considered two recently constructed energy-efcient apartment house blocks in Vienna, Austria. One is a passive with controlled ventilation; the second is a low-energy, which is ventilated mainly naturally. Both were constructed simultaneously by the same company in the same location and with comparable general building construction features and similar oor plans. The main difference between the two building types (other than the higher degree of thermal insulation in the passive building) lies in the ventilation system: passive buildings deploy controlled ventilation, whereas the low-energy buildings rely mostly on user-operated natural (window) ventilation. Thus, these buildings provide a suitable case in point for the comparison of indoor environment, energy performance, and construction costs of low-energy and passive buildings. We measured indoor environmental conditions (indoor air temperature, relative humidity, and CO2 concentration) in two units of each block over a period of ve months. Subsequently, selected inhabitants of the observed buildings were interviewed to assess their satisfaction levels with their apartments air quality and environmental systems.

2. Methods 2.1. Selection of buildings The selected buildings are located in Vienna, Austria. The construction was completed September 2007, with inhabitants moving in shortly thereafter. The site contains ve buildings, one of which is a passive house with 27 apartments; the remaining four buildings are low-energy houses containing a total of 111 apartments.

Corresponding author. E-mail address: amahdavi@tuwien.ac.at (A. Mahdavi). 0378-7788/$ see front matter 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.enbuild.2010.02.025

A. Mahdavi, E.-M. Doppelbauer / Energy and Buildings 42 (2010) 13141319 Table 1 Overview of the selected apartments. Block Passive Low-energy Apartment PH 1 PH 2 LH 1 LH 2 Area [m2 ] 59.4 89.5 51.6 84.5 No. of residents 1 3 1 5 Main orientation West West, North, East East West U-value external walls [W m2 K1 ] 0.13 0.13 0.40 0.40

1315

U-value windows [W m2 K1 ] 0.8 0.8 1.34 1.34

The apartment sizes range from 48 to 120 m2 living space. Each apartment has one to four rooms. Building oor plans differ only marginally. We conducted our studies in the passive house and in one the low-energy house blocks. In each block, measurements were conducted in one small and one large apartment. Both selected apartments in the passive house are situated on the second oor of the building. In case of the low-energy house, the smaller apartment was located on the second oor and the larger apartment on ground oor (Table 1). 2.2. Construction and building systems The exterior walls of the passive house buildings consist of 18 cm of ferrocement and 30 cm of EPS F-15 insulation. The windows have high-quality three-panel-glazing with well-insulated frames. The house is equipped with a mixed central and semicentral ventilation system. A central ventilation device is situated in a mechanical equipment room in the basement. The central device has an outdoor-air lter, ventilators for delivery and return air and efcient waste heat recovery. Inhabitants can control the air change rate as well as the temperature. The additional heating of the supply air can be regulated via a thermostat. The required rest heating energy and the energy for warm water supply is provided by a district heating network. An energy-efcient pump is used. The air change rate can be regulated from off (0.1 h1 ) to eco (0.3 h1 ), normal (0.45 h1 ), and party (0.6 h1 ). In case of the low-energy house, the exterior walls consist of 18 cm of ferrocement and 11 cm insulation. The apartments are equipped in this case merely with a basic ventilation system, which provides a minimal air change rate. Thus, the apartments must be ventilated mainly via manual operation of the windows. The basic ventilation system works as follows: a supply-air device is installed above each window. Working as a ventilation slot, this device automatically opens and closes depending on the measured humidity levels of the indoor air. In the open position, a ventilator in the bathroom draws the air from the interior of the apartment and directs it to the outside. Gaps below door panels allow for the air to ow from one room to another. Outdoor air entering into the rooms is warmed as it is mixed with indoor air at the ceiling level. The buildings heating energy is supplied by a district heating network. 2.3. Monitored indoor conditions Indoor air temperature, relative humidity, and CO2 concentration levels were measured continuously (every 515 min) via sensors in each apartment situated in the living rooms (LR) and the sleeping rooms (SR). The measurements extended over a period of ve months (from early February to the end of June 2009). Thus, data for both colder and warmer outside conditions could be collected. 2.4. Additional data Additional data was collected pertaining to metered energy use, construction costs, embodied energy assumptions, and CO2 emission assumptions pertaining to the energy mix deployed. Upon conclusion of the measurements, the inhabitants were interviewed

to assess their satisfaction with the apartments energy systems, air quality, and the ventilation systems. 3. Results and discussion 3.1. Histograms Fig. 1 shows the frequency distribution of measured CO2 concentrations (in all observed apartments over the entire monitoring period). It reveals that CO2 concentrations lie within reasonable ranges in all apartments as only a small fraction of the recorded data lies above the Pettenkofer threshold of 1000 ppm. Nonetheless, PH 1 has overall slightly lower CO2 concentrations than LH 1 and instances of high concentrations in PH 2 are much less frequent than in LH 2. The differences are more discernable if the data are visualized in a cumulative manner. As Fig. 2 (cumulative depiction of CO2 concentrations for all apartments in April) shows, CO2 concentrations in the low-occupancy apartments (PH 1 and LH 1) are generally lower than those in the other apartments. Nonetheless, passive apartment PH 2 can be shown to perform slightly better than LH 2, particularly in the colder periods of the year (see Fig. 3) and most signicantly in the sleeping rooms during the night.

Fig. 1. CO2 concentration in all apartments.

Fig. 2. CO2 concentration (cumulative) in sleeping room of all apartments in April.

1316

A. Mahdavi, E.-M. Doppelbauer / Energy and Buildings 42 (2010) 13141319 Table 2 Fraction of time (in %) within comfort zones. Month Room Apartment PH 1 February March April May June Entire period April, May, June SR LR SR LR SR LR SR LR SR LR SR, LR SR, LR 3.9 0.2 62.2 91.0 100.0 100.0 60.9 81.7 60.5 90.7 65.1 82.3 LH 1 94.3 45.9 64.8 40.3 62.9 35.7 57.3 PH 2 1.0 39.7 21.6 73.2 84.0 100.0 41.6 77.6 70.0 97.4 60.6 78.4 LH 2 5.0 1.5 67.4 22.5 99.3 87.8 61.7 58.2 86.4 79.4 56.9 78.8

Fig. 3. CO2 concentration (cumulative), sleeping room of 3 apartments (February).

Fig. 4 shows the measured indoor air relative humidity values (RH). These values are within a reasonable range most of the time. The RH values in LH 1 vary more widely that those measured in the other apartments. Moreover, the multiple-occupancy apartments PH 2 and LH 2 show higher RH values as compared to the smaller apartments. Fig. 5 shows the measured indoor air temperature values. It suggests a similar distribution pattern in all apartments, with the exception of LH 1, which displays a shift toward higher temperatures. This specic circumstance may be a consequence of user behavior, i.e. absence of night-time window ventilation and day-time window shading [3].

3.2. Psychrometric charts To assess the thermal comfort conditions inside the apartments, we calculated the fraction of time (in %), when measured indoor air temperature and relative humidity values were inside the comfort zone. The extent of the comfort zone was obtained as follows. For each month the mean outdoor temperature was calculated. This value was used to compute the neutrality temperature and the extent of the comfort zones on psychrometric charts according to Ref. [4]. Table 2 summarizes the results (percentage of time within comfort zone) for individual months as well as the total monitoring period. Note that for these results different times of the day were considered depending on the room usage. The relevant times of the day for the living rooms was considered to be from 8:00 to 22:00, whereas the relevant times for the sleeping room was from 22:00 to 8:00. Since we could not obtain results for LH 1 in the months of February and March, two different percentages were calculated. The rst one covers the whole monitoring period and the second one covers the period from April to June. PH 1, PH 2, and LH 2 display comparable results that are signicantly better than those from LH 1, which shows very high temperatures (in May and June temperatures even rose above 30 C). This specic situation may be explained by considering the orientation, window size, and the shading circumstances. The window size turns out to be a major factor in particular when we compare temperatures in the sleeping and living rooms of LH 1 (Figs. 6 and 7), which have the same orientation and shading conditions. However, the living room window size (in relation to oor area) is signicantly larger than that of the sleeping room. As Fig. 7 shows, the temperature in the sleeping room of LH 1 does not rise above 27 C in June. (Note that this difference prevails during both day and night). Table 2 also suggests a very low fraction of time in comfort zone for all apartments in February. The adaptive comfort theory suggests in this case comfort temperatures signicantly lower than those maintained by the occupants. 3.3. PMV and PPD Calculating the PMV (predicted mean vote) and PPD (predicted percentage of dissatised) is an alternative approach in assessing thermal comfort [5]. Our underlying assumptions for the calculation of PMV and PPD are summarized in Table 3. For the purpose of the calculations, mean radiant temperature values were assumed to be equal to the (measured) room air temperatures. Table 4 summarizes PPD values for each month and each room as well as for the

Fig. 4. Relative humidity values in all apartments.

Fig. 5. Indoor air temperatures in all apartments.

A. Mahdavi, E.-M. Doppelbauer / Energy and Buildings 42 (2010) 13141319

1317

Fig. 8. PMV values, PH 1, living room.

Fig. 6. Indoor conditions, living room, LH 1, June.

Fig. 9. PMV values, PH 1, sleeping room.

Fig. 7. Indoor conditions, sleeping room, LH 1, sleeping room, June.

Table 3 Assumptions for the calculation of PMV and PPD. Room Hours Winter (Feb, Mar, Apr) Clo SR LR 22:008:00 8:0022:00 2.5 1.0 Met 0.8 1.3 Summer (May, Jun) Clo 1.5 0.5 Met 0.8 1.3

v [m s ]
0 0.15

v [m s1 ]
0 0.15

Table 4 Predicted percentage of dissatised [%]. Month Room Apartment PH 1 February March April May June Entire period April, May, June SR LR SR LR SR LR SR LR SR LR SR, LR SR, LR 16.2 5.6 17.5 5.3 28.7 7.9 17.4 6.4 22.5 5.8 13.3 14.8 LH 1 27.2 21.5 11.8 17.4 13.2 17.9 18.2 PH 2 16.1 5.4 22.0 5.5 37.1 9.4 20.8 6.5 19.4 5.8 14.8 16.5 LH 2 20.0 12.4 16.8 8.2 12.9 13.3 11.0 9.3 13.6 7.1 12.5 11.2

whole period. The table implies that the results for the living rooms are considerably better than the results for the sleeping rooms except for LH 1. Generally, all apartments perform rather well. Figs. 8 and 9 show PMV values for PH 1 during the entire measurement period for the living and sleeping rooms. Nearly all living room data lies within a range of 0.5 to +0.5. The corresponding values for the sleeping room suggest warmer conditions. The other three apartments display similar conditions except LH 1, which shows higher ranges in both rooms. The reason for this is again the window size, orientation and shading situation. In summary, PMV values in living rooms are fairly reasonable in all apartments, whereas sleeping rooms display higher PMV values. It is possible that the actual clo-values maintained by inhabitants were lower than our assumptions (see Table 3). 3.4. Evaluation by inhabitants Given the small number of interviewees (ve inhabitants of the passive house and six inhabitants of the low-energy house), the results are of limited representative value. Thus, they are not treated here in detail. Nonetheless, they do imply that the inhabitants in both building types are generally satised with indoor conditions. Moreover, the level of acceptance regarding the ventilation system in the passive house is quite high. High indoor air temperatures in the summer period represent a problem for some residents, whereby the inhabitants of south-facing apartments are particularly affected. PH inhabitants had some issues with too low humidity values in winter, whereas LH inhabitants complained to the contrary. None of the interviewees showed any signs of the sick building syndrome.

1318

A. Mahdavi, E.-M. Doppelbauer / Energy and Buildings 42 (2010) 13141319

Table 5 Energy use data for 2008 (totals for the apartments and per net oor area). Type of energy Unit Apartment PH 1 Heating Electrical (apartment component) Electrical (general component, entire block) Electrical (total) [kWh] [kWh m2 ] [kWh] [kWh m2 ] [kWh] [kWh m2 ] [kWh m2 ] 20.76 498.9 8.4 669.4 11.28 22786.3 9.48 32.78 30.12 PH 2 1358.7 15.2 2085 23.30 LH 1 1153.8 22.4 1388.8 26.94 7441.4 3.18 53.85 LH 2 3922.2 46.4 4281.3 50.67

Table 6 Total CO2 emissions in kg (calculated for 2008). Type of energy Apartment PH 1 District heating Electrical power Total 112 728 840 PH 2 306 1731 2037 LH 1 260 916 1176 LH 2 882 2685 3567

energy mix in this specic case (heating energy is provided through a district heating network). 3.7. Embodied energy Another aspect of energy consumption is the embodied energy of a building. Embodied energy is the energy that was used during the production process of a product including raw material extraction, transport, manufacture and deconstruction. We estimated the additional embodied energy and CO2 emissions due to additional insulation, better windows and additional ventilation system of the passive house (Table 7) based on data obtained from [6] and [7]. Subsequently we estimated the amortization period due to passive buildings lower energy demand and CO2 emissions during operation time. Table 8 shows that the amortization time for the additional (embodied) energy investment in passive apartments is around 13 years with respect to embodied energy and 25 years regarding CO2 emissions. 3.8. Costs The initial costs for the construction of the passive house block was in the present case approximately 5% higher than the comparable low-energy house block. The major contributing factors to the passive buildings higher construction costs are the ventilation system, the higher performing windows, and the additional thermal insulation (particularly exterior walls). On the other hand, the passive buildings energy use during the operation time is lower. A simple payback analysis (using energy price information given in Refs. [8] and [9]) resulted in an amortization period of roughly 818 years for the additional construction investment in the pasTable 9 Amortization time estimation for additional initial investments for passive buildings. Unit PH 1 8.40 20.76 0.78 3.60 4.39 52.08 2.93 17.8 PH 2 15.20 32.78 1.42 5.69 7.11 52.08 6.56 7.9 LH 1 22.40 30.12 2.09 5.23 7.32 LH 2 46.40 53.85 4.32 9.35 13.67

Table 7 Additional embodied energy and CO2 emissions associated with the construction of passive buildings. Embodied energy [kWh m2 ] Insulation Windows Ventilation system Total 47.7 6.1 15.0 68.8 CO2 emission [kg m2 ] 25.7 1.6 3.2 30.5

3.5. Energy use Energy use data (heating and electrical) for the four observed apartments (2008) is given in Table 5. The heating energy use of passive houses is signicantly lower, due to better insulation and lower ventilation heat recovery. As to the electrical energy use, a distinction can be made between general areas of each house (including the houses ventilation system) and the apartments. The general electrical energy use is higher in the passive house due to the energy requirements of the ventilation system. However, in the present case, passive house apartments require less electrical energy. The reason for this is not obvious. Contributing factors could be user behavior, more effective daylight usage, and higher efciency devices and appliances. Considering both components together, the electrical energy use of the low-energy apartments was, in the present case, somewhat higher than the passive house apartments. 3.6. CO2 emissions Table 6 includes computed CO2 emissions associated with the apartments energy use (heating and electrical). It can be asserted that the passive house apartments produce signicantly less CO2 emissions than the low-energy apartments. Note that all values of Table 6 may be said to be rather low, given the nature of the

District heating Electrical energy District heating Electrical energy Total energy costs Additional initial costs Energy cost savings Amortization time

kWh m2 a1 kWh m2 a1 D m2 a1 D m2 a1 D m2 a1 D m2 D m2 a1 years

Table 8 Amortization periods pertaining to passive buildings additional embodied energy and CO2 emissions. Apartment Additional investment Annual reduction Amortization [years] Energy PH 1 68.8 kWh m2 23.4 kWh m2 2.9 30.5 kg m2 5.7 kg m2 5.4 68.8 kWh m2 52.3 kWh m2 1.3 CO2 Energy PH 2 30.5 kg m2 17.1 kg m2 1.8 CO2

A. Mahdavi, E.-M. Doppelbauer / Energy and Buildings 42 (2010) 13141319

1319

sive buildings (see Table 9). Note that this calculation does not address rising energy prices or other energy sources, which could have resulted in shorter amortization times. 4. Conclusion A comparison of passive house apartments with low-energy apartments in Vienna was conducted based on monitored indoor environmental conditions, user evaluation, metered energy use, calculated embodied energy and CO2 emissions, as well as construction costs data. The results suggest that both passive and low-energy apartments performed well in view of thermal conditions and indoor air quality, even though the performance of passive apartments was slightly better. Specically, it can be asserted that the use of a ventilation system clearly contributed to lower levels of carbon dioxide concentrations in particular during cold periods and especially in multiple-occupancy apartments. The inhabitants of both buildings were generally satised with indoor conditions and building systems. Passive house apartments were shown to consume approximately 65% less heating energy and 35% less electrical energy as compared to low-energy apartments. Moreover, passive apartments CO2 emissions (estimated based on metered operation energy usage) were approximately 2540% less than low-energy houses. Considering the construction implications of passive house apartments in view of higher embodied energy and CO2 emissions, relatively short amortization times were calculated (approximately one to ve years). The initial costs penalty associated with the construction of passive apartments (as compared to low-energy apartments) was 5%, resulting in an estimated payback period of 818 years. The latter estimation must be seen in the context of a

twofold uncertainty: given a different source of energy than the one relevant in the present case (district heating) or in case of drastic future energy price increase scenarios, the payback period could be signicantly shorter. On the other hand, ifin contrast to the present case studythe electrical energy use of passive buildings would have be higher than the low-energy buildings, the payback time would have been longer. In summary, our study suggests that passive buildingsas compared to low-energy buildingsuse signicantly less heating energy and offer slightly better indoor conditions. Thereby, the required additional expenditure of resources (as represented by embodied energy) and environmental impact (as represented by CO2 emissions) are offset in a rather short period. Moreover, the required addition construction cost does not appear to be either excessive or prohibitive. References
[1] OENORM B 8110-6, Wrmeschutz im Hochbau - Teil 6: Grundlagen und Nachweisverfahren - Heizwrmebedarf und Khlbedarf, (2007) pp. 13,14. [2] Passivhaus Institut Darmstadt, URL: www.passiv.de (visited in August 2009). [3] Wagner, W., Prein A., Felberbauer K.-P., Energietechnische und baubiologische Begleituntersuchung der Bauprojekte: Berichtsteil Passivauswohnanlage Dreherstrasse, Zwischenbericht, AEEInstitut fr nachhaltige Technologien, Gleisdorf, Austria (2008). [4] S. Szokolay, Introduction to Architectural Science: The Basis of Sustainable Design, Architectural Press, Oxford, Great Britain, 2004. [5] P.O. Fanger, Thermal Comfort, McGraw-Hill Inc., US, 1970. [6] M. Schuss, Life-Cycle-Analyse von Passivhusern, Technische Universitaet Wien, Austria, 2004. [7] T. Waltjen, kologischer Bauteilkatalog. Bewertete gngige Konstruktionen, Springer Verlag Wien, Austria, 1999. [8] IWO-Austria, http://www.iwo-austria.at/index.php?id=126 (visited in January 2010). [9] Wien Energie, Personal communication (January 2010).