Conference Publication Details
Mandatory Fields
D. Browne, K. Menzel
CIB W078 -W102 Joint Conference
Method for Validation of Building Simulation Results using Sensor Data
Optional Fields
Simulation, Data Modelling, Validation
Morland, Zarli
Sophia Antipolis, France
In general, current Building Energy Simulation Tools are used for pre-construction design and comparison of designs rather than a full exact varying representation of reality. To provide the best level of detail full CFD analysis for the entire building would be required. However this is currently by far outside the scope of current computing power for a building energy system. Because these simulation tools are designed for comparison of potential designs and because of the difficulty in predicting occupant behaviour, very often the predicted results do not correlate with the real actual performance when buildings are in operation.
From project experience encountered in the EU FP7 IntUBE project, a deficit has been encountered whereby the correlation between simulation results and real measured data is not entirely accurate. This paper discusses a method of validation, which will provide a means of comparing measured data (e.g. sensors and weather data), and simulated data (e.g. near future simulations).
This method for validation of building simulation results initially involves a comparison of data from building simulation and respective measured sensor readings. From this comparison, value is added from correction of simulation results, and/or input to simulation parameters. Further worth can also be provided by gaining knowledge for creation of simulation profiles which are difficult to predict before construction & operation. Additional value can also be derived from identifying conditions of poor results and relevant factors which can be corrected. Simulation data and actual data is available from a housing unit in Barcelona Spain and research building in Cork Ireland.
The expected result to be derived from this method is to give an indication of quality of simulated data results and provide feedback. If the difference between simulated and real data is too large, steps to improve results will be suggested. In future it is envisioned that automated adjustments may performed to simulation inputs to correct results. Aside from near future simulation validation, the tool may be able to provide long term commissioning feedback to detect and alert users to long term degradation of systems and possible maintenance or repair remedies.
Grant Details