Rebound effects following refurbishment – Scientific monitoring of the Karlsruhe-Rintheim neighbourhood concept
Comparison between the measured energy consumption and the calculated energy demand for the renovated wing R2, entrances E1 to E3. Only those flats of entrances E1 to E3 of wing R2 are compared that have the same tenancy terms. There is a large variance between the individual flats. The same renovation was carried out for each entrance but there are differences of up to 140%. This is not always a case of exceeding the requirement but also of falling short of the calculated value. All flats in wing R3 and the corresponding renovation variants displayed an average overshoot of 40% over the EnEV calculations.
© RWTH Aachen, E.ON ERC
|Location of local community||Stadt Karlsruhe, 76137 Karlsruhe, Baden-Württemberg|
|Settlement in figures||Rintheimer Feld neighbourhood: Area: 0.25 km²; approx. 2,500 residents|
|Developer, organizer||Volkswohnung GmbH|
|Settlement||Urban perimeter and block developments|
|Built-over area per block||534 m²|
|Heated building volume||7.212 m³|
|Apartment size||ca. 70 m²|
|State of construction and refurbishment||Poor– requires considerable refurbishment|
|Heating system||Individual stoves (energy sources) / individual central heating systems (natural gas)|
|Ownership structure||Renting through municipal building society|
The phenomenon of reduced saving potentials has been debated for some time under the term "rebound effect". This is usually caused by the increased use of energy services. The lack of data available has meant that a detailed, quantitative analysis of the rebound effect has not been taken into account in previous studies.
The research project intends to use the already existing, comprehensive measurement technology and building simulation models in order to close this gap. The measurement values will be supplemented with discussions with tenants and questionnaires. By examining the efficiency of the systems and the actions of tenants in detail it is intended to divide the rebound effect into economic, structural and technical effects, whereby structural effects can be caused by changes in the heating system or expanding the usable living space and technical effects result from inadequately adjusting the heating system to the changed building parameters. The systematic structure enables the rebound effect to be reduced from the beginning with future construction projects.
Initial situation, implementation
The three residential blocks being refurbished in Karlsruhe-Rintheim were completed in the 1950s. Each block consists of three building sections, each of which has ten accommodation units and separate entrances. The residential blocks are 51.63 metres long and 10.34 metres wide on the outside, which means that they have an A/V-ratio of 0.48. The first building block acts as a comparison building, since the standard refurbishment provided by Volkswohnung Karlsruhe was implemented here.
All three blocks use a different combination of structural parts (variation in the U-values of the heat-transferring enclosure surface) and system-related components (heating, ventilation and domestic water installations). A comprehensive measurement system was installed in all apartments to check the forecast energy savings.
- Further data collection and analysis to separate structural from technological effects
- Technical consultation and information sessions to share knowledge
- Analysis of structural and technological rebound effects
- User behaviour and usage profile comparison as per DIN
- Specific user profiles for dynamic models.
Results already achieved in the project "Integrated Neighbourhood Energy Concept Karlsruhe Rintheim – Scientific Monitoring" as a basis for the present project
Calculations for the heating load and EnEV (German Energy Saving Ordinance) were performed as part of the structural analyses. Heating load was determined on a room and building basis for the various cases. The results were collated with those of Ingenieurbüro KW2-Ingenieure and used for the tender and design of the technical systems. The increased indoor temperature desired by Volkswohnung leads to larger heating surfaces and a higher output of the heat generators.
All thermal bridges were analysed for the existing buildings. Using the FEM programme "Therm", the resulting length or point-based thermal bridge loss coefficients were calculated for existing thermal bridges and compiled in a thermal bridge catalogue. With the identified thermal bridges and an appropriate proof of equivalence, it is possible to use the determined values instead of resorting to the normative values for upcoming EnEV calculations for the renovation of buildings constructed in the 50s and 60s. The calculations and the comparison with the standard thermal bridge factors have shown that in this case and with a detailed validation, losses over thermal bridges were reduced throughout the stock of buildings. In case of the two renovation alternatives, a detailed calculation of the thermal bridge loss coefficients tended to have a negative impact on the final result.
EnEV 2007 permits two different validation methods: the simplified heating period balance method and the monthly balance method. The German Energy Saving Ordinance requires undercutting a maximum value of annual primary energy demand QP (primary requirement) and of transmission heat loss H / (secondary requirement). The validation method allows a variation in the calculation depth of individual parameters, such as the temperature correction factor and the system expenditure factor. All calculations of transmission heat loss and primary energy demand were carried out both with the normative values and the values as determined in detail. It was demonstrated that the requirements for transmission heat loss and primary energy demand were always met or significantly undercut for every renovation and calculation method. It must be noted that the detailed calculation methods take considerably longer.
The dynamic building and system simulation was performed using the freely available programming language Modelica. The object-oriented programming language allows nesting of individual models of an overall simulation to produce a structured hierarchy. An overall model of a building consists of a large number of sub-models differing in complexity. In the course of a project, the sub-areas of weather, building, heating (generation, storage, distribution, transfer and control strategy), and ventilation devices are considered in detail. In order to simulate the buildings of this research project, existing libraries were supplemented with different models. In the area of building services technology, models of a decentralised ventilation device with heat recovery, an exhaust air heat pump, a solar power system, a wall model with PCM, as well as various models of an underfloor heating were developed and validated where possible using the available measurement data. The effects that were determined by the measurement analysis were also demonstrated by the simulations, for instance, that an indoor temperature increase by 2 K in a flat would lead to an increase in energy demand by 200%.
An extensive set of measurement equipment was installed in the buildings to analyse both user behaviour and the efficiency of the system technology. In the front doors, the number of measurement points depends on the special nature of the energy-oriented refurbishment. There generally are basic measurements in flats that are always the same. In addition, there are measuring points for specific parameters such as control parameters, volume flow, etc. The measurement intervals vary between one and fifteen minutes depending on the measurement point. This results in about six million data points daily for evaluation. The gathered data are stored in an HDF5 database and are easily and quickly visualised with the HDF viewer.
The evaluation tool "HDF-Tables-EBC", based on Python-HDF-Tables, was further developed at the institute to analyse the data quickly and efficiently. In addition to the standard features of "HDF-Tables", the tool allows for the visualisation and evaluation of data gathered from field tests and simulations with mouse clicks. At this time, 15 different functions have been implemented in "HDF-Tables-EBC".
Innovative and new components for building services technology pursue the goal of energy optimisation. In order to be able to draw conclusions from actual consumption and the expected savings, some components (decentralised ventilation device and radiator variants) are additionally tested in a laboratory to provide a basis for assessment.
The decentralised ventilation device from Schüco is equipped with a multi-gas sensor which, in automatic mode, controls the external air rate in three stages. The function of the sensor and the impact of the switching thresholds were both experimentally analysed. The analysis of the VentoTherm ventilation devices led to the conclusion that the IAQ sensors (IAQ stands for Indoor Air Quality) show a good correlation with ambient CO2 concentrations as measured by an external device. In conclusion, IAQ sensors react to human emissions with a sufficient level of detail. The measurements in the bathroom laboratory setting demonstrated that there was no significant influence of humidity on the IAQ sensor signal. The control of the ventilation rate can thus be separated based on contamination and humidity levels.
In the product brochure of Kermi, the radiator ThermX2 is described as an "energy-saving radiator". In the course of the laboratory test, it was examined whether a serial, uniform flow was present and which type of air current was produced around the radiator and inside the room. During the heating up process, it can be clearly seen that the front radiator plate has an even horizontal flow. This flow behaviour does not change during the entire heating up process. No discrepancies were identified at the edges or around the valve connections. Only given a complete flow inside the frontal heating plate is the rear plate heated up. With a rapid increase in radiation temperature inside the room, lowering the ambient air temperature at a temperature perceived as constant can lead to savings.
Despite energy-oriented refurbishments, energy is required for heating and supply with domestic hot water. To secure demand, system temperatures merely around the ambient or indoor temperature are required , i.e. with low exergy. Natural heat sinks and sources are usually only usable in combination with thermal storage systems. Sensitive heat storage, especially in the form of water storage, is probably the most common thermal storage system. A possible optimisation approach for water storage systems is to produce an optimal temperature distribution inside the storage systems by means of suitable control concepts.
In order to reduce exergy losses, an alternative controller concept was proposed in which the temperature inside the storage system is adapted to the exterior temperature. In the representative simulation phase from early January to late March, this concept led to a primary energy saving of 3.4%. An exergetically optimised connection concept was investigated for the combination of the built-in waste heat and exhaust air heat pumps. On part of the air heat pump, the simulations showed an electrical energy demand saving of 26.3%. The exhaust air heat pump had a small additional demand of 2.1%.
An evaluation of the measurement data shows that the theoretical savings after renovation cannot be fully realised – there is a so-called rebound effect. This effect describes how every effect incurs after-effects that impair the initial positive effect, or even turn it into its opposite. The analysis of the available consumption data of the respective flats with their measured indoor temperatures partially explains the high consumption, as the indoor temperature in some rooms was up to 24 °C. This circumstance is, for example, one reason for the discrepancy between the calculated demand and actual energy consumption.
An increase of energy efficiency in the building sector does not necessarily lead to a reduction of total energy consumption and is therefore only of limited use as an instrument of energy and environmental policy. Additional measures are called for. A reduction of the theoretically expected energy savings due to the performance gap and the rebound effect has to be taken into account in the future when planning the structural and system technological components. The impact on energy policy goals must be questioned and implementation scenarios must be created.