The Federal Office for Building and Regional Planning

Publication Quality test of software for issuing energy performance certificates

BBSR-Online-Publikation 21/09, Eds.: BMVBS/BBSR, September 2009

Series: BBSR-Online-Publikation Published: 2009

Editing:
Institut für angewandte Informatik im Bauwesen (IAIB) (contractor)
Prof. Dr. Dr.-Ing. Klaus Fehlauer (project leader), Heiko Winkler

Federal Institute for Research on Building, Urban Affairs and Spatial Development, Bonn
Horst-P. Schettler-Köhler (project leader) horst.schettler@bbr.bund.de

Executive Summary

Shortly after the introduction of the Energy Saving Directive of 2007 (EnEV 2007) the Institute for Applied Informatics in Civil Engineering (IAIB e.V.) was issued the task to survey the existing software implementing this directive and to conduct quality test by the Federal Office for Building and Regional Planning (BBR).

The goal of this project was to investigate whether the software was correctly calculating and issuing the energy performance certificates (EPC). Also, methods for measuring and ensuring the quality of the software needed to be developed in order to ensure some transparency concerning the products that exist. One of the results was to be a testing module that made it possible to test the products for faults on a sufficiently high level to ensure their quality. In this summary primary focus will be on the results for software for non-residential buildings.

Analysed Software

Overall, there were 15 software products for calculating the energy demand for non-residential building and 13 Products for issuing energy performance certificates based on energy consumption data were tested.

For issuing EPCs for residential buildings (according to DIN V 4108-6 and DIN V 4701-10) 13 Products were tested for issuing EPC abased on energy demand calculation and energy consumption.

Quality features and test method

  • Conformity to EnEV and norm (degree of implementation) for residential and non-residential buildings

In order to get an overview on the degree of implementation a list of requirements was compiled, using the EnEV and norms. From this a questionnaire divided into 3 categories was derived.

When the list of test criteria was generated it was soon evident that only for a part of the proposed questions needed to be answered positively in order to ensure the ability of the software to issue an EPC based on the calculations for the energy demand. This is why the there are two question types.

  • Key questions: they ask for requirements. Answering them positively ensures the availability of a key feature for issuing the EPC or the ability to calculate the energy demand. Some of the key questions ask for the presence of necessary provision devices going as far as the settings for specific parameters and operation modes (like low consumption mode at night).
  • Comfort questions: they mostly ask for features that enable the user to conduct his calculation and the issuing of EPCs with greater comfort (like the single zone model).

All questions of the catalogues are proposed in a positivistic fashion. This means that the answer "yes" means, the requirement is fully met.

Possible answers beyond "yes" are "partially", "no" and "n/a".

After carefully answering all questions of the questionnaire a score is calculated from the provided answers.

  • Comparative calculation for residential buildings (RB)

As a reference building for the calculation a brick-built apartment building from the 1930s was selected. Programming errors were detected by stepwise checking of all input and output including all intermediate and final results. Deviations were investigated.

  • Comparative calculation for non-residential buildings (NRB)

The analysis several research and statistical data revealed that a mid-terrace house with store on the ground floor and offices on the upper floors represents the most common type of a NRB in Germany. Therefore, a NRB of that type was selected as test building for all testing in our research.

Since for NRB the focus of the norm DIN V 18599 is mainly on building equipment and appliances (HVACR) an emphasis was placed on a variation of the HVACR while leaving the parameters describing the building envelope and the lighting as they were. The first four HVACR variants were selected in such a way that they could be evaluated as such. Four more variants are simply combinations of several supply systems.

The main difference between the norms for residential buildings and non-residential buildings is that the energy consumption according to DIN V 4108-6 and DIN V 4701-10 can, for the most part, be calculated by hand. The demand calculation for non-residential buildings is far more complex. This makes calculation by hand and checking it very difficult and time consuming if not even impossible.

In order to have reference values for evaluating the calculation accuracy an expectation value was derived using statistical means.

Since nine of the programs tested, used the kernel provided by the Fraunhofer Institute for building physics, a study needed to be conducted whether or not the results programs provided statistically independent results. The variance of the results from the kernel based programs and the non-kernel programs was similar, which is why a statistical independence of the results from the kernel programs could be assumed.

Rating of the results for the programs for residential buildings

Some of the requirements not met as well as most of all the errors detected according to the EnEV directive were due to the fact that most of the new requirements were not yet implemented. In many cases these were merely changes of values for parameters that had not yet been changed from the last EnEV directive in 2004.

Examples for this are: changed calculation f the usable area of older buildings with story heights of less than 2.5 m and greater than 3 m and irritating notes regarding the ratio of opaque and non-opaque outer surface area of a build when calculating the summer heat protection certificate.

When answering the questions from question catalogue B (degree of implementation of the EnEV directive for residential buildings) a wide range of results was identified. Only one program was able to sufficiently fulfill all requirements. All-in-all, it can be said that the EnEV 2007 requirements were only insufficiently incorporated into the software.

Regarding the usability of the programs the results range between 50 and 70 percent of the designated criteria. While most programs feature online updates and a support hotline only few offer important control features for inputs like input checklists, input help i.e. in dialogue windows like diagrams or last but not least input and revision support like input status, tutorials and case studies.

By comparing the results of individual programs for the demand calculation, a number of identical results were identified. Beyond that, a distinct statistical spread was identified for heat loss due to ventilation. The reason for this was a differently calculated heated air volume in the respective programs. Furthermore, there were deviations for the permissible annual primary energy demand QP`` as well as the heat transmission loss HT'. The reasons for this were not investigated more closely. All-in-all must be said that there was a wide statistical spread for the annual primary energy demand QP``.

For the calculations for the energy performance certificates there were no serious shortcomings (see Annex O). Only because some products still used older climate profiles that could not be updated it was not possible to perform weather corrections. Thus it was not possible to calculate averages for the Energy usage and to calculate the characteristic value for the energy consumption.

Assessment of the results for the programs for non-residential buildings

For the degree of implementation of the EnEV and the DIN V 18599 most programs reach about 75%. If only the key questions are considered only one program reaches about 95%, five more programs reach a score of above 80%. This means that the probability that with these programs the issuing of an energy performance certificate based on demand calculation for non-residential builds is most likely possible for all cases without having to fear that specific HVACR combination cannot be modeled.

For the assessment of the accuracy of the calculation for a number of specific values (intermediate or final results) the statistical expectation value was derived using all results from all programs and their deviation from the expectation value was calculated.

After the evaluation and determining the position of the results for the specific values an overview was compiled for each HVACR combination. From the evaluations of all variants can be derived that some products have their strengths in certain areas, while others tend to be less precise.

This impression is confirmed when counting the hits inside the 5 % and 10 % corridors for all HVACR variants. At the same time the transparency value that was calculated for each program informs about the availability of the specific values required for the evaluation of the products.

For a general assessment of the serviceability of the tested programs for non-residential buildings the results for the degree of implementation of the specifics of energy savings directive (EnEV) and the norm (DIN V 18599) on one hand and the accuracy of the calculation on the other are put in direct relation to each other.

The resulting diagram shows that there are currently 6 programs for non-residential buildings that offer a high serviceability regarding accuracy and conformity. Four more of the tested programs only partly were able to cope with the given scenario and therefore appear to be only partly serviceable However there is a good chance that these programs may improve rapidly due to the observed development dynamics.

At the same programs could be identified that are not sufficiently developed due to the lack of quality and transparency.


The abstract is part of the German publication "Qualitätsprüfung für Energieausweis-Software", BBSR-Online-Publikation 21/09, Hrsg.: BMVBS/BBSR, September 2009, Bonn
ISSN 1868-0097, urn:nbn:de:0093-ON2109R226
Download (PDF, 14MB, File meets accessibility standards)

This Page

Also interesting

Pagination