Table of Contents
Heating load in Passive Houses
Due to their extremely high level of energy efficiency, the heating demand in Passive Houses (typically no higher than 15 kWh/(m²a)) is negligible – only about 10% of the energy used in conventional buildings. This has been verified time and again in field measurements: by building to the Passive House Standard, it is actually possible to achieve energy savings as high as 90% compared to conventional buildings [AkkP-28]. Yet a particular building’s exact heating demand, guaranteed to be low in any case, is really secondary to ensuring that a Passive House building functions properly.
In this sense, the heating load is the decisive factor: a certain amount of heat can be distributed with very little effort via the supply air coming from a Passive House building’s ventilation system. The ventilation system thus serves a dual function (fresh air and heating) and reduces the investment required for heat distribution to a minimum. These cost savings can then be reinvested in energy efficiency measures, including the ventilation system itself. However, the amount of heat that can be distributed via the fresh air system without any additional costs is limited. For a house with an average occupancy of 1 person per 30 m² of living area, the capacity available is around 10 W/m² - irrespective of the climate (see box).
The specific values for heating loads are not identical with the ones for energy (measured in kilowatt hours (kWh)), the numbers for which are often easier to come by.
The Passive House energy threshold of 15 kWh/(m²yr) for heating typically relates to a heating load of 10W/m² in typical Central European climates, however, it is only supposed to serve as a rough benchmark which may vary with different climatic conditions: in Stockholm a house with a heating load of 10W/m² may use more like 20kWh/(m²yr); in Rome it might be as low as 10kWh/(m²yr).
The Passive House criteria allow buildings to go by either criterion - the 15 kWh/(m²yr) heat demand OR the 10W/m² heating load.
Climate-independent Passive House requirement
The minimum fresh air flow rate for one person is 30 m3/h (according to the DIN 1946 – health criterion). At 21°C and standard pressure, air has a heat capacity of 0.33 Wh/(m3K). Fresh air can only be heated by a maximum of 30 K (to 51°C) in order to avoid dust carbonisation or the burning of small dust particles in the air.
This results in the following capacity needed per person:
Ppers = 30 m3/h/pers * 0.33 Wh/(m3K) * 30 K = 300 W/pers
This shows that the heating of the supply air can provide 300 Watt per person. Assuming 30 m² of living area per person, this would result in 10 W per m² of living area, regardless of the climate. This is an output unit, i.e. the values are based on the day with the highest heat output required (heating load). In order to meet this criterion, a Passive House will require different levels of insulation depending on the climate zone: more in Stockholm, less in Rome.
A building whose maximum heating load does not exceed the available capacity of the supply air distribution system will no longer require other heat distribution pipes and radiation systems in addition to the ventilation system (with the exception, perhaps, of a small, emergency heating unit in the bathroom). This, in turn, results in simplified building services systems so that the total investment costs for building services in a Passive House do not necessarily have to be higher than the costs in a conventional building. Nevertheless, they typically are somewhat higher as more efficient heat generators and ventilation systems with heat recovery are generally more expensive than conventional building services systems.
Heating load calculation
Whether the described functional simplification can actually be realised in a particular project depends considerably on the maximum heat outputs required (i.e. the heating load) in that particular case. Therefore, it is essential to reliably ascertain the heating loads in advance. Today’s Passive House owners find themselves in a similar situation as home owners just before the oil crisis: the heating system has to be accurately dimensioned even though the resulting annual demand results in such low costs (either due to the low price of oil before the crisis or due to the low energy use in Passive Houses). That is why it is essential to have a reliable method for determining the heating load.
The following requirements apply for this method:
- The heating loads must be calculated conservatively, i.e. they must ensure a comfortable level of heating for the buildings planned.
- The calculated heating loads should not, however, contain excessive safety margins as this would make the structural and technical requirements unreasonable to reach and the specific advantages of buildings with very small heating loads would no longer be apparent.
- If possible, the method should be simple to apply
- The boundary conditions required for a particular design should be easily available. It would therefore seem appropriate to apply the existing standards for determining the space heating load [EN 12831]. However, it soon became evident that this standardised method led to extreme over-dimensioning in the case of highly efficient buildings such as Passive Houses. The reasons (besides the easily modifiable “special features” relating to oddly selected additional conditions but not to the method itself, such as U-values that always have to be set at a minimum of 0.3 W/(m²K)) are as follows:
- Internal heat sources and solar gains that are especially significant during very low outdoor temperatures are not taken into account adequately in EN 12831. However, even in the design case, these free heat gains play a very significant role, particularly in buildings with a very small heating load. “No internal loads” only applies if there are no occupants present, which in turn leads to very minimal requirements. If occupants requiring a high standard of comfort are present, internal heat gains exist on a regular basis; possibly in small amounts but not zero. This makes a crucial difference, especially in the case of well-insulated buildings.
- Buildings with very small heating loads often have very high time constants (anywhere from 5 to 30 days or more). Due to this, brief periods of extreme weather conditions are irrelevant for the Passive House (these are virtually disregarded by the building) and the design parameters refer primarily to longer periods of time. This fact was also known to the originators of the older standards (such as DIN 4701), but it was not extended to buildings with very long time constants and, in the end, it was even ignored in the newer standards.
- Room-by-room determination of the heating load is associated with a high degree of uncertainty even in conventional buildings, which results from the fact that the internal heat flows, caused by relatively small temperature differences between rooms, can be more significant than the heat losses towards the outside. This effect is even more significant in the Passive House. For this reason, room-by-room determination of the heating load usually doesn't make sense in Passive Houses; individual calculations for each apartment or building are more reliable and are usually sufficient. Details can be found in [AkkP 25].
- Heating load calculations are usually based on floor areas calculated according to interior dimensions, thus disregarding thermal bridge effects and going against the rest of the entire planning procedure, for which the use of exterior dimensions has become established.
Problems calculating the heating load in very well-insulated buildings
It was shown that the heat output in very well-insulated buildings, actually measured through scientifically monitored projects, had a much lower upper limit for output than the design output in accordance with the [DIN 4701] standard, even at extremely low outdoor temperatures. This was first published in [Feist/Werner 1993] based on the measured daily heat consumption values in the Passive House in Darmstadt-Kranichstein. According to these results, there is a kink in the horizontal curve for the measured daily mean heating loads below about 0 °C. This correlation was correctly explained in [Feist/Werner 1993] by solar energy gains during the colder weather periods and discussed in more detail in [Feist 2005]. Accurate heating load results simply cannot be obtained without taking into account solar gains, particularly in very well-insulated buildings.
It was thus necessary to develop a better algorithm for the heating load calculation as compared with the established method. A transient simulation of the thermal characteristics of the building is always methodically sound. Not only the heating loads but also the annual demands can be reliably calculated in this way. The validity of such transient calculation models is checked in comparison with measured values in [Feist/Loga 1997] and [Kaufmann/Feist 2001], based on the time courses of indoor, surface, and building component temperatures. These investigations showed that there was a very good correlation between the measured results and simulation results. It is therefore permissible to carry out basic examinations by means of thermal simulations of buildings. Of course, the results must also be checked experimentally, especially if they differ from the conventional view. The main disadvantage of using transient thermal simulations of buildings is the high complexity of the models on which they are based. This also results in a comparatively high error rate, apart from the great amount of work required for the preparation and evaluation of such simulation calculations. The number of parameters used in the simulation model is very large and thus the potential sources of error also increase during the generation of the models. It was therefore desirable to make available another simple method for ascertaining the heating load, similar to the existing monthly method for determining the annual heating and cooling demands according to ISO 13790.
Development of the evaluation model
This theme was treated by the Research Group for Cost-Effective Passive Houses and solved in cooperation with the University of Stuttgart (Institute of Thermodynamics and Thermal Engineering) and ebök engineering consultants. The key approaches were examined using the transient model DYNBIL in a dissertation by Carsten Bisanz [Bisanz 1999]. The model developed during this cooperation is principally based on energy balances in accordance with DIN EN 832, but with boundary conditions that take into account the key climatic periods of the region under consideration, instead of the annual or monthly data. It has proven essential that dimensioning be based on at least two different climate data sets, namely a “cold and sunny design period” and a “moderately cold and cloudy period”. Especially in the case of buildings with very small heating demands, it is highly uncertain whether the maximum heating load will actually occur during extremely cold periods or during a very cloudy but only moderately cold period. The relevant design boundary conditions must be determined functionally for each climate with the aid of dynamic building simulations using test data sets. This was done in [Bisanz 1999] for the German test reference years, and an initial theoretical validation of the method was carried out with the aid of simulations.
The procedure developed in [Bisanz 1999] fulfils the requirements for simplicity and easy availability of the respective boundary conditions; therefore, in 1999, this procedure and the selected boundary conditions were adopted experimentally in the second version of the Passive House Planning Package [PHPP 1999]. At the time, this procedure depended on successful implementation, but since been applied in thousands of successful Passive House projects.
Verification of the heating load method: Model/Practical experience
Within the framework of the IEA SHC TASK 28 / ECBCS ANNEX 38 Research Project, detailed time-resolved data was obtained from scientifically substantiated metrological studies of the temperature characteristics and heating consumption of building projects with a total of over 200 Passive House dwelling units [Feist 2005]. This data has been analysed in a variety of studies with respect to various aspects, including heating consumption, other energy consumption figures, thermal comfort parameters as well as average and maximum heating loads.
The field results lead to a uniform evaluation:
- Not only do quality assured Passive Houses built actually meet the extremely small annual heating demand previously calculated by means of simulations and/or balancing procedures, the extremely low heating load resulting from the functional dimensioning is sufficient for the heating of these buildings.
- On the basis of the measured values, it is possible to confirm the fact that internal heat sources and passive solar gains must also be taken into account in the heating load calculation, especially for well-insulated buildings.
- Performance during special situations can be validated metrologically using measured temperatures and heating loads in very specific individual cases (e.g. unoccupied apartments unheated throughout the winter). These measurements also confirm the simulation.
- Very stable temperatures in buildings with excellent insulation, especially in Passive Houses, are confirmed by the case studies as well as the simulations. This increases error tolerance for extreme weather events (“worst winter in a hundred years”) as well as errors in dimensioning. Of course, this knowledge must be handled very carefully: if the cumulated errors exceed a certain level, even the Passive House Standard’s tolerance level will be exceeded and the errors will have an even greater effect. It is therefore advisable to take the planning task and quality assurance seriously and to use the tolerance level resulting from a correctly planned and built Passive House to ensure and account for variances in occupant behaviour.
- As regards the central question of this study, the calculation approaches based on the method published in [Bisanz 1999] have proved very successful for all buildings under consideration. This method has thus undergone a special performance test, as Passive House buildings with their extremely small heating loads are particularly susceptible to influencing factors such as solar radiation. Only buildings with such low heating loads allow the testing of such a method, as such small influences are usually masked by other effects in buildings with high heating loads.
The present study is an example of how thorough field measurements in combination with scientifically substantiated evaluations can provide helpful results. Such results are statistically proven and go beyond common estimates based on “gut instinct”. These results can be transferred to easy-to-use methods for specialists, thus facilitating their work.
[Feist 2005] Feist, W.: Heizlast in Passivhäusern – Validierung durch Messungen. Endbericht. IEA SHC TASK 28 / ECBCS ANNEX 38. Passivhaus Institut, Darmstadt 2005
[Feist 2005] Feist, W.: Heating load in Passive Houses – Validation through measurement. Final Report. IEA SHC TASK 28 / ECBCS ANNEX 38. Passive House Institute, Darmstadt 2005
[AkkP 28] Wärmeübergabe- und Verteilverluste im Passivhaus; Protokollband Nr. 28 des Arbeitskreises kostengünstige Passivhäuser Phase III; Passivhaus Institut; Darmstadt 2004
[AkkP 28] Transmission and distribution heat losses in the Passive House; Research Group for Cost-effective Passive Houses Phase III, Protocol Volume No. 28, Passive House Institute, Darmstadt 2004
[DIN EN 12831] DIN EN 12831: Heizungssysteme in Gebäuden – Verfahren zur Berechnung der Norm-Heizlast Deutsche Fassung EN 12831; Beuth Verlag; Berlin
[DIN EN 12831] Heating systems in buildings - method for calculation of the design heating load EN 12831; Beuth Verlag; Berlin
[AkkP-25] Temperaturdifferenzierung in der Wohnung; Protokollband Nr. 25 des Arbeitskreises kostengünstige Passivhäuser Phase III; Passivhaus Institut; Darmstadt 2004
[AkkP 25] Temperature differentiation in apartments, Research Group for Cost-effective Passive Houses Phase III, Protocol Volume No. 25, Passive House Institute, Darmstadt 2004
[DIN 4701] Deutsches Institut für Normung: DIN 4701: Regeln für die Berechnung des Wärmebedarfs von Gebäuden; Beuth Verlag; Berlin 1995
[DIN 4701] German Institute for Standardisation: DIN 4701: Regulations for calculating the heating demand of buildings; Beuth Verlag; Berlin 1995
[Feist/Werner 1993] Feist, W. und Werner, J.: Erste Messergebnisse aus dem Passivhaus Darmstadt Kranichstein; gi 114 (1993) Heft 5 Seite 240 ff
[Feist/Werner 1993] Feist, W. and Werner, J: Initial measurement results from the Passive House in Darmstadt Kranichstein; gi 114 (1993), Issue 5, page 240 ff
[Feist/Loga 1997] Feist, W. und Loga, T.: Vergleich von Messung und Simulation. In: Arbeitskreis kostengünstige Passivhäuser, Protokollband Nr. 5, Passivhaus Institut, Darmstadt 1997
[Feist/Loga 1997] Feist, W. and Loga, T.: Comparison of measurements and simulation. Research Group for Cost-effective Passive Houses, Protocol Volume No.5, Passive House Institute, Darmstadt 1997
[Kaufmann/Feist 2001] Kaufmann, B. und Feist, W.: Vergleich von Messung und Simulation am Beispiel eines Passivhauses in Hannover- Kronsberg. CEPHEUS-Projektinformation Nr. 21, Passivhaus Institut, enercity, Hannover 2001
[Kaufmann/Feist 2001] Kaufmann, B. und Feist, W.: Comparison of measurements and simulation using the example of a Passive House in Hannover-Kronsberg. CEPHEUS-Project Information No. 21, Passive House Institute, enercity, Hannover 2001
[Bisanz 1999] Bisanz, C.: Heizlastauslegung im Niedrigenergie- und Passivhaus, 1. Auflage, Darmstadt, Januar 1999
[Bisanz 1999] Bisanz, C.: Dimensioning the heating load in low-energy and Passive Houses, 1st edition, Darmstadt, January 1999
[PHPP 1999] Feist, W.; Baffia, E. und Schnieders, J.: Passivhaus Projektierungspaket 1999; Passivhaus Institut, Darmstadt, Januar 1999
[PHPP 1999] Feist, W.; Baffia, E. and Schnieders, J.: Passive House Planning Package 1999; Passive House Institute, Darmstadt, January 1999