Energy use – measurement results

The heat requirement of a building is mathematically determined – the actual consumption depends on many other influences, e.g. user behaviour and the weather. Years of experience have provided statistically reliable measurements of actual consumption values. The reliability of the Passive House concept can be judged from these results.

Even for identically constructed houses, there are significant differences in consumption due to varying user behaviour. Therefore the consumption values for a sufficiently large number of identical homes must be measured so that the usage-dependent influences can be averaged out, thereby enabling a comparison of the building quality. Fig. 1 shows an overview of measurement results from 41 low-energy houses and a total of 106 Passive Houses. A number of findings can be drawn from these results.

Fig. 1. Overview of consumption measurements: This diagram shows measure- ments for the heating energy consumption of four developments: One low- energy development (left) and three passive house developments to the right.

For comparison we will take a closer look at the low-energy housing development of 41 terraced houses in Niedernhausen. The individual values of the heat meter readings for the year 1994 are shown in Fig. 2 (measurements by: [Loga 1997] ). The average value of all homes measured is 65.56 kWh/(m²a) (Notes about the unit applied).

Fig. 2: Consumption statistics for a low-energy housing development: the housing estate with 41 low-energy houses in Niedernhausen was first occupied in 1992. The average consumption of 65 kWh/(m²a) matches the calculated demand of 68 kWh/(m²a) [PHPP] within the achievable accuracy. The inserted curve is the respective Gaussian distribution - Note for mathematics experts. The consumption measurements were carried out by T. Loga and M. Großklos [Loga 1997] .

This average value is considerably lower than the average heat consumption in existing housing stock in Germany, which amounts to about 160 kWh/(m²a). The consumption in the low-energy housing development is around 60% below the average consumption . Besides, the construction standard of this development, although built in 1991, is still better than the energy requirements of the current German energy code (EnEV).).

Fig. 2 also shows that the individual values are spread out around the average value. The influence of the user on the actual consumption is undeniable – it is even quite high. This is true not only for low-energy buildings, but also for poorly insulated houses. The standard deviation (a measure for the average deviation of the individual values from the average value) amounts to 13.6 kWh/(m²a) or 21% of the average consumption value for this development.

The deviations due to user behaviour average out to a great extent if an average value is used, even more so for a large number of identically constructed units. The average consumption value for this housing development has an uncertainty of only about ±2 kWh/(m²a). It is therefore statistically secured that the low-energy standard leads to significant energy savings of 60% ± 1% compared to average previously constructed houses.

This project was the first Passive House housing development in Germany (built in 1997, by Rasch & Partner) and consists of 22 homes. Fig. 3 documents the heat meter readings of the 1998/99 winter season. The average value was determined as 13.4 kWh/(m²a). This means that the average consumption of the Passive House estate is 80% lower than that of the low-energy development in Niedernhausen.

Fig. 3: Consumption statistics for the Passive House estate: the housing estate with 22 Passive Houses was built in 1997. The average consumption of 13.4 kWh/(m²a) correlates very well with the calculated demand of 13 kWh/(m²a) [PHPP]. Note for mathematics experts. (Measurements Wiesbaden-Dotzheim: [Ebel 2003] ; [Feist/Loga/Großklos 2000] ).

The standard deviation of the individual values is ±5.3 kWh/(m²a) and is much lower than that of the low-energy housing development. However, relative to the much smaller average consumption, the user behaviour is more noticeable. The average value is statistically determined by an error of only ±1.1 kWh/(m²a). The energy savings due to the Passive House standard are statistically secured as

(80% ±5%) in savings compared to the low energy standard,
(92% ±1%) in savings compared to average heating consumption in Germany.

The Passive House housing development in Hannover Kronsberg consists of 32 essentially identical terraced houses according to the Passive House standard in mixed construction. It was built in 1998/99; all units were designed individually. These were part of the CEPHEUS. Fig. 4 documents the measurements of the heat meters in the heating season of 2001/2002. The average value amounts to 12.8 kWh/(m²a). Therefore the consumption in this Passive House development is about 81% less than the low-energy development Niedernhausen ([Peper/Feist 2002] ).

Fig. 4: Consumption statistics for the Hannover Kronsberg Passive House settlement: The development of 32 passive houses was occupied in 1999. Average consumption during the third year of operation (2001/2002) was 12.8 kWh/(m²a). The heat requirement according to the [PHPP] calculations amounts to 13.5 kWh/(m²a). Note for mathematics experts.

The average consumption values of all occupied Passive Houses in the housing development as measured by means of heat meters were as follows:

1. Heating period 1999/2000: 14.9 kWh/(m²a)
2. Heating period 2000/2001: 13.3 kWh/(m²a)
3. Heating period 2001/2002: 12.8 kWh/(m²a)

The extremely low heat consumption values for the Passive House housing development in Hannover Kronsberg is therefore also statistically secured - the standard deviation of the individual values is 6.6 kWh/(m²a), the average value is determined with an error of only ±1.2 kWh/(m²a).

The Passive House housing development in Stuttgart Feuerbach (“Schelmenäcker Weg”) with a total of 52 terraced and detached houses was finished in the year 2000 by the Rudolf architectural practice. Fig. 5 documents the consumption values of the 2001/2002 heating season. The average consumption value amounts to 12.8 kWh/(m²a) [Reiß/Erhorn 2003] . For this housing development, there are a few outliers that are clearly identifiable as such.

Fig. 5: Consumption statistics for the Stuttgart/Feuerbach Passive House settlement: The development of 52 passive houses was finished in the year 2000 (Rudolf Architects). Average consumption was at 12.8 kWh/(m²a). The heat requirement calculations according to the [PHPP] amount to 13.5 kWh/(m²a). Note for mathematics experts

The extremely low heat consumption values for the Passive House housing development in Stuttgart Feuerbach is also statistically secured – the standard deviation of the individual values is 5.5 kWh/(m²a), the average value is determined with an error of only ±0.8 kWh/(m²a).

Further empirical studies have independently confirmed these conclusions in other areas of the world: * A publication by Treberspurg et al., found in the conference book of the 14th Passive House Conference, analyzes the measured energy consumption of six Passive House apartment complexes in Vienna; the average space heating consumption was shown to be under 10 kWh/(m²a). * A preliminary result from the measurement of the world's largest Passive House complex with 354 living units (finished in 2009) located in Innsbruck shows that even there, all expectations in terms of energy usage were met. (Lodenareal Passive House complex)

The measurements in the existing Passive House housing developments show that:

• the individual measures taken, i.e. thermal insulation, triple low-e glazing and heat recovery are effective. Deviations of more than about 1 kWh/(m²a) would have been identifiable in the average values, but do not occur.
• the calculation method and the boundary conditions used function satisfactorily in practice. The deviations between the calculated values and the measured values are very small.
• Additional heat losses, like losses due to the control system or high window ventilation losses cannot have a significant influence according to the consumption statistics monitored; they must lie within the limits determined with ±1 kWh/(m²a) and are therefore insignificant.
  

Different users, even if they live in identically constructed houses, frequently have very different consumptions: deviations of ±50% from the average value are not exceptional. This is not only the case in houses constructed according to the Passive House principles. The main reason is the different thermostat settings during the heating season. The maximum usage-dependent consumption values in Passive Houses are still significantly less than the lowest consumption in average new buildings.

The Passive House concept demonstrably and repeatedly leads to very high energy savings which amount to more than 90% compared to existing buildings and 80% on average compared to the requirements for new constructions. These results have been confirmed in later monitoring projects by other authors for other passive house developments as well [Treberspurg 2010]

1 kWh (kilowatt hour) is a unit of energy. One litre of fuel oil or 1 m³ natural gas purchased have a heating value of almost 10 kWh. Electricity consumptions are measured worldwide in kWh - so this should be a unit somewhat familiar to almost everybody ( 1 kWh = 3,6 MJ and 1 kWh = 3412 BtU).

The specific heating value q_H has been used all throughout for the comparison: q_H = Q_H / A_EBF.

Q_H is the measured (useable) heating consumption. The heating distribution at the heat transmission point was measured for all measurement projects (generally by means of heat meters; more details about that can be found in the scientific reports referred to in the text).

Example: heat meters with an m-bus output were used in all 32 terraced houses in the Passive House development in Hannover-Kronsberg. The entire measuring technique is described in [Peper/Feist 2001] . In addition, there is an overall heat meter in the central supply for each row of houses for checking the results.

This measurement includes distribution and heat transmission losses. However, it does not include the heat generator losses.

A_EBF is the treated floor area. For all results specified here, this is the heated living space; thus the area reference was selected upon which heating bills and all published statistics are based. It is important to note that the area A_N based on the calculations prescribed by the EnEV, is approximately 20 to 30% larger than the actual living space. Energy parameters based on A_N appear to be much lower than they actually are. For this reason, only the actual net floor area reference is consistently used for our calculations.

We have inserted the appropriate Gaussian or normal distribution (Wikipedia Website) into the diagrams:

1/(σ√2√π)∫ e ^{- ½ (x-μ)²/σ²},

to be precise, their inverse, i.e. the “Norminv” function which is available in spreadsheet programs. For this,

• μ describes the average value
• and σ the standard deviation.
  

Additionally, the number of the respective buildings in each sample group has been given as n. A look at the graphs shows that the heating consumption values of identically constructed objects are very well approximated by a normal distribution. More details can be found in the CEPHEUS Final Report [Schnieders/Feist 2001] (Chi-Quadrat-Test; Kolmogoroff-Smirnov-Test). In this paper it is also discussed in more detail, what effect it has if the distribution is cut off at zero. Obviously, in the case of Passive Houses, this does not make much a difference. But, for a zero-Energy-House only the positive deviations will remain; adding non-zero to zero just by statistical effects.

[Ebel 2003] Ebel, W.; Großklos, M.; Knissel, J.; Loga, T. und Müller, K.: Wohnen in Passiv- und Niedrigenergiehäusern – Eine vergleichende Analyse der Nutzungsfaktoren am Beispiel der „Gartenhofsiedlung Lummerlund“ in Wiesbaden-Dotzheim, Endbericht / Energie; Institut Wohnen und Umwelt; Darmstadt 2003.

[Feist/Loga/Großklos 2000] Feist, W.; Loga, T. und Großklos, M.: Durch Messungen bestätigt – Jahresheizenergieverbrauch bei 22 Passivhäusern in Wiesbaden unter 15 kWh/m² Wohnfläche, in BundesBauBlatt, 3/2000, S. 23-27.

[Loga 1997] Loga, Tobias; Müller, Kornelia; Menje, Horst: Die Niedrigenergiesiedlung Distelweg in Niedernhausen, Ergebnisse des Messprogramms, 1. Auflage, Institut Wohnen und Umwelt, 1997.

[Peper/Feist 2001] Peper, Sören; Feist, Wolfgang: Messtechnische Untersuchung und Auswertung - Klimaneutrale Passivhaussiedlung Hannover-Kronsberg; 1. Auflage, Proklima, Hannover 2001; You may downlod an English version on the Passive House Institute website.

[Peper/Feist 2002] Peper, Sören; Feist, Wolfgang: Klimaneutrale Passivhaussiedlung Hannover-Kronsberg Analyse im dritten Betriebsjahr; 1. Auflage, Proklima, Hannover 2002; You may downlod an English version on the Passive House Institute website.

[Reiß/Erhorn 2003] Reiß, Johann und Erhorn, Hans: Messtechnische Validierung des Energiekonzeptes einer großtechnisch umgesetzten Passivhausentwicklung in Stuttgart-Feuerbach, IBP-Bericht WB 117/2003, Fraunhofer-Institut für Bauphysik, Stuttgart 2003.

[Schnieders/Feist 2001] Schnieders, Jürgen; Feist, Wolfgang; Pfluger, Rainer; Kah, Oliver: CEPHEUS - wissenschaftliche Begeleitung und Auswertung, Endbericht, Projektinformation Nr. 22, 1. Auflage, Passivhaus Institut, 2001

[Treberspurg 2010] Treberspurg, Martin; Smutny, Roman; Grünner, Roman: Energy monitoring in existing Passive House housing estates in Austria, proceedings of the 14th Passive House Conference, pp. 35-42, 1. edition, passive house institute, Darmstadt 2010