Brought to you by
Brought to you by
This chapter deals with the subject of how monitoring can be successfully carried out in the Bahnstadt district in Heidelberg, Germany, see Figure 9, which is currently the largest Passive House settlement within Europe. In this construction area, (due to lack of funding) it was not possible to carry out detailed measurements; however, monthly meter readings are available for the total heating consumption (heating, hot water and losses etc.) for entire development blocks and some larger buildings with over a hundred apartments each. For this reason, evaluation of the data will take place in the context of minimal monitoring where, with the aid of research findings from other projects, heating consumption will be calculated to a good approximation from these monthly values.
The Passive House Standard is mandatory in the whole district. Thus, a space heating demand of less than 15 kWh/(m²a) was one of the goals pursued already during the planning phase. For comparison, a recent study called „Energiekennwerte 2014“) by techem on the space heating consumption in the year 2013 shows: Buildings supplied by a district heating system had on average 112 kWh/(m²a) energy consumption for space heating.
The Passive House development area Bahnstadt in Heidelberg consists of several large development blocks, each supplied by a central district heat connection. This means that only one central district heat connection exists for billing purposes for up to five large apartment blocks. The supplier has no access to any other sub-meters that may exist in individual buildings. These main heat meters at the transfer stations were previously read by the public utilities company “Stadtwerke Heidelberg” during on-site visits roughly every six months. Readings of these electronic heat meters will subsequently take place regularly via a data network connection. Monthly meter readings took place and were provided to the Passive House Institute for an initial overview of the overall functioning of the settlement.
The consumption data compiled in Table 1 were available from the development blocks. Here, the treated floor area constitutes the useable areas defined according to the PHPP (in case of dwelling units: living areas); when classifying the results, account must be taken of the fact that for the characteristic values according to the German minimum energy standard EnEV, the areas AN for these buildings will be 28 % larger thus, the specific consumption values accordingly even lower:
|Type||Number||Treated floor area||Number of dwelling units|
|Residential development||5||61 981m²||698|
|Student hostel||2||15 457 m²||564|
|Kindergarten||1||1 027 m²||-|
|Office use||1||9 694 m²||-|
|Laboratory building||1||21 346 m²||-|
Table 1: Overview of the type of buildings in the studied construction sites
For evaluation to be meaningful, the buildings must have been fully occupied or utilised for at least one year, only then will it be possible to calculate a reliable annual consumption. So far, evaluable data of this scale is available from the development blocks with residential utilisation, the student hostels, and the kindergarten. These can be evaluated based on the existing data for the complete year 2014. However, this only applies to a limited extent for three development blocks, because these were only fully occupied one to three months later (January - Mach 2014). Exact details regarding the time of full occupancy are not available. It is not known whether the circumstances of full occupancy in these buildings led to over-consumption or under-consumption1.
1 If a dwelling unit in a building is not yet used, then the following two variants are possible:
(I) Increased consumption will result if heating already occurs at the housing level but internal heat sources (persons, electricity use) are not present. The activity relating to occupancy itself (such as open doors and windows) can also lead to considerable additional losses.
(II) Reduction of the consumption is also conceivable if no or reduced heating occurs before occupancy.
With the available monthly readings of the central heat meter, the consumptions for all heat uses are available as total values for each development block. These total values include the following consumption variables:
The individual consumption variables cannot be differentiated out of the total values of the monthly consumptions, therefore an empirical method must be used which will at least enable a good estimate of this breakdown. In doing so, it should be kept in mind that the examined buildings do not have any solar heating panels and hot water generation in the residential buildings takes place completely via district heat.
In the main summer months the energy expenditure for all those applications which are not related to heating of the building can be determined from the monthly consumption values. In the process, it is assumed that unintentional and undesirable heating in summer does not take place. In Passive House buildings there is no distinct summer heating demand – on account of their long time constants even during “cold snaps” lasting several weeks, such buildings will still exhibit comfortable indoor temperatures without any heating at all. The month with the lowest summer consumption must not be used because some apartments may not be used in the summer due to long holiday periods. Since these are large buildings with many apartments, a slight concurrence of the holiday periods can be assumed. The average consumption of the four summer months (June - September) is calculated and used as the consumption variable ‘Expenditures without heating’ for each month; in a Passive House building in this climate, the heating demand in these months is definitively zero. If this average consumption value for the summer is now extrapolated to the whole year, this will result in the ‘annual expenditure without heating’. Figure 3 shows this consumption for the evaluation year inside the box with the green dotted line. For the sake of abbreviation, this will be referred to as the ‘base consumption’. In this development block, the average summer consumption value for Summer 2014 was 3.72 kWh/(m² month). All consumption values of the other months which are now above the green box are assessed as the “heating consumption”.
In this simplest approach, the heat dissipated by the distribution pipes is assumed to be constant in the course of the year. The forward flow temperature of district heat is determined by the demand for year-round provision of hot water. Principally, the heat dissipated by the distribution pipes is influenced by the temperature difference between the surface of the pipework and the surrounding air (e.g. basement room, underground garage).
According to this method, the district heat consumption for the entire development block in the example shown in Figure 11 results as 23.3 kWh/(m²a) in a first approximation. The base consumption is 3.72 kWh/(m² month) x 12 months = 44.6 kWh/(m²a). This type of calculation basically allows the calculation of the heating consumption value from the little measured data that is available. However, for different reasons, this first approximation leads to overestimation of the heating consumption:
Altogether, due to the first approximation (“base consumption method”) the effects described here result in an overestimation of 1.4 to 2.5 kWh/(m²a) of the heating consumption plus the respective “unintentional” heating consumption for May. With the maximum value of 2.5 kWh/(m²a) and the project-specific heating consumption for May, the overestimation results as 2.9 to 3.7 kWh/(m²a). This consumption must be deducted in the second approximation that has now occurred in order to achieve a more realistic value for the district heat consumption for heating. A heating consumption (second approximation) of 23.3 kWh/(m²a) – 3.3 kWh/(m²a) = 20.0 kWh/(m²a) results for the development block (Figure 3). The major influences that lead to overestimation of the heating consumption in this method are thus taken into account. The values of the second approximation that are thus determined will subsequently be referred to as the “heating consumption” for the purpose of abbreviation again. The average measurement error here may be in the range ca. ±4 kWh/(m²a). The heating consumption in the useable area of over 80 000 m² that was measured here may be extremely small even with this (relatively large but in absolute terms extremely small) error margin. It is already apparent that the Passive House project with the Bahnstadt district in Heidelberg is extremely successful.
In accordance with the method described above, evaluation was first carried out for the available development blocks with residential utilisation. The average values of the same summer months (June - September) were always used for calculating the base consumption values (first approximation) – separately for each of the development blocks. For adjusting the district heat consumption for heating for the second approximation, the respective calculated values between 2.9 and 3.7 kWh/(m²a) were deducted and added to the “base consumption”. This shifts the allocation of the consumption values but not their total amount. In the study year 2014, the specific total district heat consumption values for the development blocks with residential use were between 46 and 68 kWh/(m²a). Allocation leads to base consumptions between 33.0 and 48.0 kWh/(m²a). The expenditure for heating is between 9.3 and 24.2 kWh/(m²a); the average value based on area density is 14.9 kWh/(m²a) for this. These consumption values are shown Figure 4 as a total value and as allocated values. For the evaluation of the consumption data carried out here, account must be taken of the fact that the achievable accuracy is limited. This method uses simplifications and assumptions, which means that in this analysis, accuracies less than ± 3 kWh/(m²a) [Feist 2004], which are achievable with detailed measurements, are unacceptable.
2The numbering of the development blocks are anonymized because of data protection
It is clear that there is relatively large scattering of the consumption values between the development blocks. In particular it must be considered that three of the seven development blocks (shown hatched) were not fully inhabited during the complete one-year period of 2014. Based on the available data, these three development blocks were only inhabited one to three months later. As mentioned in the explanation in chapter 1.2, it is understandable that this can lead to an increase or reduction in the consumption values. Only the next study year can clarify this.
The scale of the distribution across heating and the remaining consumptions for the “base heat” for hot water provision, distribution and storage in the Bahnstadt is within the typical range compared with Passive House projects studied previously. In order to illustrate this, the data from a building with 19 apartments that was centrally supplied with heat and investigated in detail are shown in Figure 5. The heating energy, hot water consumption, distribution heat and other detailed variables were analysed in more detail in the related study [Peper/Grove-Smith/Feist 2009]. In this project, it was exemplarily demonstrated that the heating energy had a 33 % share of the total supplied energy. The Bahnstadt objects presented here with shares between 20 and 36 % were thus within a realistic scale.
The potential for technical optimisation certainly also exists in this project, particularly with reference to storage and distribution losses. The two outliers with values above 60 kWh/(m²a) should be examined more closely in this respect and optimised if necessary.
For evaluation of the heating consumption values in particular, account must be taken of the fact that the consumption data depends significantly on the respective weather during the study period and the indoor temperature selected by the users. Thus it is unreasonable to expect that a building which has been balanced for 15.0 kWh/(m²a) during the planning should now give exactly this consumption value, for example. In addition, with the large number of apartments in a complex, it always comes down to the average consumption value as only this is conclusive. The actual weather conditions and actually set room temperatures should be taken into account for the actual consumption (see also [Peper 2012b]; these values cannot be known at the time of planning, therefore the planning team must use standardised design values). The measurement data are determined by these “boundary conditions” as well as by the characteristic values of the building.
Evaluation of the consumption data shows conclusively that extensive efforts by the City of Heidelberg to design a city district to a high standard of energy efficiency through provisions and quality assurance have proved successful. An extremely good result has been achieved here with heating consumption values of 14.9 kWh/(m²a) on average, for measurements mainly in the first year of operation and including hostels. The fact that this involved a very large number of apartments (over 1000) with a total studied living area of 75.000 m² is particularly impressive. With the great number of these buildings, it will be possible to show that wide scale implementation of highly energy-efficient buildings with many different stakeholders is quite possible and can be done successfully.
The following illustration, Figure 6, shows the heating consumptions of the residential buildings and the area-weighted average value separately. In the next study year, slightly different results should be expected on account of different weather conditions, the complete year-round utilisation data which will become available, and the absence of first-year effects. It must be observed whether both higher consumption values (20 and 24 kWh/(m²a) respectively) change noticeably after full occupancy, for example. However, it can be stated that these consumption values will also remain at a low level and should not be assessed as problematic at all.
The PHPP (Passive House Planning Package) was used for planning all buildings in the Bahnstadt district. This allows for energy-relevant optimisation of the building during the planning process. The PHPP was also used by the City of Heidelberg for quality assurance of the planning. Certification of the buildings by the Passive House Institute or an accredited certifier only took place in a few individual cases, and none of the buildings in the development blocks that were studied here were certified. Accurate and complete tracking of the changes made during the planning and particularly during the construction process are crucial for a realistic calculation. Experience has shown that if this is done with the necessary level of accuracy, the PHPP (among other things) delivers a realistic heating demand in accordance with the boundary conditions applied, such as climate data, occupancy density, internal heat gains, indoor temperature etc. In the case of projects that were studied in more detail, the comparison between consumption data and the PHPP demand calculations often shows quite good correlations, see Figure 7. As the most important parameters for the comparison with the measurement, the climate data and the indoor temperature – as explained above – must be determined for this purpose in accordance with the actually existing boundary conditions, and used in the PHPP.
As a next step, the consumption data of the studied residential buildings in the Bahnstadt should be compared with the planning data in the PHPP. In this way it will be possible to check the data for plausibility and to identify any outliers. For the study in the Bahnstadt, the actual weather data for Heidelberg in the observation period 2014 was required. It was necessary to have at least the monthly outdoor temperature and the monthly total global radiation (horizontal). Data from a measuring station in Heidelberg-Kirchheim was used for the outdoor temperature (http://heidelberg-kirchheim-wetter.de). This was at a distance of ca. 2.5 km. The comparison with measurements from Ludwigshafen and Speyer (both in Germany) only showed slight variances. A source could not be found for obtaining the global radiation data for Heidelberg, therefore the radiation data for the location Ludwigshafen-Mundenheim was obtained from ZIMEN, the measuring network of the German State of Rheinland-Pfalz (www.luft-rlp.de). The outdoor air temperatures for Heidelberg exhibited minimal deviations from the location in Ludwigshafen-Mundenheim compared with the other alternative (Speyer). The weather data set that was thus prepared will subsequently be referred to as “Wetter Heidelberg” and “Wetter LU/HD”.
The comparison of the weather data for Heidelberg that was available for 2014 and the standard climate data set “Mannheim” in PHPP used during the planning shows clearly that the period 2014 was extremely mild: the winter months were significantly warmer than in the climate data set “Mannheim”. Global radiation differed particularly during the summer months, which was not relevant here, see Figure 8.
In the present minimal monitoring, measured indoor temperatures were not available for the 1260 dwelling units. This essential parameter for adjusting the PHPP calculation for planning can thus only be assumed on the basis of other monitored projects. In other measurements in residential-use Passive Houses, indoor temperatures of about 21.5 °C on average were measured in winter [Peper 2012b]. For this reason this indoor temperature has also been used here and was applied as a boundary condition in the PHPP.
The residential buildings and student hostels studied here in minimal monitoring were balanced in a total of 30 PHPP calculations. Since only one heat meter exists for each development block, the demand values for heating from the individual PHPP calculations of the development block must be summarised into a comparative, area-weighted value. The data for each student hostel is available in a separate PHPP calculation. For the development blocks with residential buildings there are between three and five, and in one case, ten PHPP calculations. Each of these PHPP calculations took place using the weather data set for HD/LU as a boundary condition, and in a second step, the indoor temperature was increased from 20 to 21.5°C. The resulting balance value calculated thus for a development block can now be compared with the consumption value from the previous section. The PHPP calculations provided to the PHI could not be checked within the framework of this study. However, during the processing of the PHPP calculations, such as addition of the weather data set for 2014, a few points did come to light. Some of these have a noticeable effect on the heating demand and were therefore accommodated:
Further tests and changes to the PHPP calculations did not take place. The changes that were made were taken into account in the values presented below. Figure 17 shows the heating consumption values, see chapter 4.4, with the summarised PHPP calculations for each development block. The results with the 2014 weather data set Heidelberg are shown for PHPP demand values. The heating demand value is depicted for 20 °C as well as for 21.5 °C.
The consumption data can be most reasonably compared (orange and dark green bars) using the current boundary conditions described above (usual indoor temperature 21.5 °C and weather data for HD/LU) during the study period. For five of the seven development blocks, there are excellent correlations with variations between 0.3 and 3.9 kWh/(m²a). This is excellent for a comparison of consumption measurements, particularly as this involves minimal monitoring with expected measurement deviation on the same scale. It can therefore be assumed that these PHPP calculations are reliable. However, two of the studied projects (BS-07/08 and BS-13) show considerably larger differences between the measured consumption values and the PHPP calculations with the weather data set HD/LU at an indoor temperature of 21.5°C. These are also the development blocks with the highest measured consumption values. There are significant deviations here, for which different reasons are conceivable:
There may be a mixture of different reasons; it would be pure speculation to make a decision about this at this point. One of the objectives of the present study was to find such projects so that more exact examinations can follow. With the present results, this is possible.
Figure 18 illustrates the results of the PHPP calculations for space heating using various weather/ climate data sets. The calculation results of the Mannheim climate data set (PHPP standard) also clearly point out the mildness of the study year 2014. Accordingly, higher consumption values can be expected in cooler years. But even if the climate data set Mannheim (long-term average from previous years) is used, low consumption values of just 16 to 18.5 kWh/(m²a) result in comparison. Also, the influence of the respective higher indoor temperatures can be seen clearly for both climate or weather data sets: increases in the demand values for space heating of 15 % typically result for each Kelvin of indoor air temperature increase.
The following can be stated in conclusion: With a total useable area of more than 75 000 m², the buildings studied here consume only one third of the district heating of that of comparable existing buildings, with 55 kWh/(m²a) for heating, DHW distribution and storage losses together as an overall average. This is comparable with the results from a previous detailed study of two Passive House buildings with a district heating connection.
The district heating consumption for heating is around 15 kWh/(m²a) on average (±4). This is an excellent result for consumption in the first year of operation. According to various studies carried out in Switzerland, this first-year effect always leads to additional consumption which can be between 15 and 30 % of the regular consumption. Higher values were measured in the case of two development blocks that were first occupied during the measurement year; this was probably due to the influence of activities relating to moving in.
Comparison with the values in the PHPP planning (recalculated with the current weather data) gives an excellent correlation with measurement/calculation variances of less than ±4 kWh/(m²a), which remains within the limits of measurement accuracy; only the objects already mentioned above are beyond this mark.
In the first year of operation, the heating energy savings of about 87% envisaged for the present project through Passive House project planning was, in comparison to the measured mean value of 112 kWh/(m²a) by Techem company, already nearly achieved with 81% - including outliers.
[AKKP42] Passive House Institute: Protokollband Nr. 42 Ökonomische Bewertung von Energieeffizienzmaßnahmen, Darmstadt 2013 (German only).
[Bretzke_2009] Bretzke, A.: Benefits of the Passive House Standard in schools: cost Benefits of the Passive House Standard in schools: cost-effectiveness and user convenience, conference proceedings of the 13th International Passive House Conference, Frankfurt, Germany 2009
[Feist 2004] Feist, W.: Wärmeübergabeverluste im Licht der Baupraxis. In: Wärmeübergabe- und Verteilverluste im Passivhaus. Protokollband Nr. 28 des Arbeitsreises kostengünstige Passivhäuser Phase III, Passivaus Institut, Darmstadt 2004. Heat transmission losses in the light of construction practice, in: Heat transmission and distribution losses in Passive Houses. Protocol Volume No. 28 of the Research Group for Cost-effective Passive Houses Phase III, Passive House Institute, Darmstadt 2004 (German only)
[Peper 2008] Peper, S.: Passivhaus-Heizsysteme in der Praxis. Ergebnisse und Erfahrungen aus der Feldmessung. In: Protokollband 38 des Arbeitskreises kostengünstige Passivhäuser Phase IV; Passivhaus Institut; Darmstadt 2008. Passive House heating systems in practice: Results and experiences from field measurements. In: Protocol Volume No. 38 of the Research Group for Cost-effective Passive Houses Phase IV, Passive House Institute, Darmstadt 2008 (German only)
[Peper 2012a] Peper; S.: Messung zur Verbrauchskontrolle – „Minimalmonitoring“. In: Richtig messen in Energiesparhäusern; Protokollband 45 des Arbeitskreises kostengünstige Passivhäuser Phase V, Passivhaus Institut, 2012 Measurements for checking consumptions – Minimal Monitoring. In: Accurate measurements in energy-efficient buildings; Protocol Volume 45 of the Research Group for Cost-effective Passive Houses Phase V, Passive House Institute, Darmstadt 2012 (German only)
[Peper 2012b] Peper; S.: Messkonzepte, Störgrößen und adäquate Lösungen. In: Richtig messen in Energiesparhäusern; Protokollband des Arbeitskreises kostengünstige Passivhäuser Phase V, Passivhaus Institut, 2012 Monitoring concepts, disturbance variables and adequate solutions. In: Accurate measurements in energy-efficient buildings; Protocol Volume of the Research Group for Cost-effective Passive Houses Phase V, Passive House Institute, Darmstadt 2012 (German only)
[Peper/Grove-Smith/Feist 2009] Søren Peper; Jessica Grove-Smith; Prof. Dr. Wolfgang Feist: Sanierung mit Passivhauskomponenten, Messtechnische Untersuchung und Auswertung Tevesstraße Frankfurt a.M., Bericht im Auftrag des Hessischen Ministeriums für Wirtschaft, Verkehr und Landesentwicklung, Wiesbaden, Passivhaus Institut, Darmstadt, 2009. Download unter: www.passiv.de Scientific monitoring of the Tevesstrasse Passive House renovation in Frankfurt am Main. Report commissioned by the Ministry for Economy, Transport and Regional Development of the German State of Hesse, Wiesbaden, Passive House Institute, Darmstadt 2009, download from www.passiv.de
|The sole responsibility for the content of Passipedia lies with the authors.
While certain marked articles have been created with the support of the EU, they do not necessarily reflect the opinion of the European Union;
Neither the EACI nor the European Commission are responsible for any use that may be made of the information contained therein.