Simulation software is widely used in building research and a large variety of software exists. The major advantage that simulation software provides for the researcher is the ability to quickly evaluate a huge number of buildings, or variations of a building. The basis of using computer simulation in this research was that it is a very good way to estimate how a building may react under different stimulus, and essentially this is what the decay method analyses: the response of the building overnight under varying initial temperature conditions. An example from the literature is that of Morrissey, Moore & Horne (2011) – a study in which the authors simulated how changes in orientation effected building heating and cooling loads. Much time and effort in conducting field experiments was saved using a computer simulation program to assist in the initial determination of the potential steps in the decay method set up and analysis. This allowed for increased quality and direction of the field experiments, and as time was saved, increased the number of experiments that could be run.
Furthermore, once the field experiments were completed, the data could be fed back into the computer simulation to assist in answering any additional questions about any anomalous readings in the field. The computer simulation allows for full control over various physical elements, such as removing the influence of the sun or wind, or disabling heat flow through certain areas of the building.
73 Choice of Simulation Software
The choice of simulation software was made on the basis of suitability for the study, access to knowledge and training, and cost.
The critical elements required were:
• A robust, proven, thermal simulation engine, having passed the Building Energy Simulation Test (BESTEST). Software that has passed this test is recognised for its quality and accuracy by the International Energy Agency.
• To be fully capable of simulating the testing conditions of at least the co-heating test.
This required the ability to fully customise thermostat settings, and create and modify weather inputs. If the co-heating test could be simulated, it was expected that the software would handle any other combinations of thermostat settings and weather conditions.
• The ability to output data at regular intervals, in a spreadsheet format or similar, regarding internal temperature, power input, and external conditions.
• Fast access to troubleshooting and training.
• Low cost.
Additional desirable elements included:
• Ability to bulk process multiple simulations.
• Ability to automate complex analysis of results (linked to the format of data output).
• Flexibility in disabling and enabling certain elements of building physics, such as removing the effects of sun or wind.
• Ability to create an Energy Rating Certificate to compare test results with NatHERS star rating standards.
Three Australian-based, NatHERS approved software, AccuRate, BERS Pro and FirstRate 5, and EnergyPlus, operated by the United States Department of Energy, were identified as possible software. Table 3.1 and Table 3.2 show a comparison of the software against the criteria laid out above.
74
Table 3.1 Critical requirements of simulation software
Simulation
*Note: Costs to researcher specific to this PhD study.
75
Table 3.2 Desired additional features of simulation software
Simulation
FirstRate 5, BERSPro and AccuRate all use the same Chenath engine to drive the underlying calculations, and each has its own front-end graphical user interface. Due to the studentship arrangement with CSIRO, both FirstRate 5 and BERSPro were discounted. Only the use of AccuRate, and the version of the Chenath engine provided with the AccuRate software, was explored.
The scratch file generated by AccuRate containing the direct inputs required by Chenath is open to be edited directly by users outside of the AccuRate program. To run this file, the engine must be called directly from the command line or run through the CSIRO’s batch simulation software, AccuBatch. Technically, this does allow for full customisation within the limitations of what Chenath accepts as inputs. The main limitation is that thermostat settings in the scratch file are based on a single 24 hour profile, repeated for each day of the year.
EnergyPlus is itself a backend program, and all inputs and outputs are based on .csv files, making it exceptionally customisable. All the equations that inform the EnergyPlus thermal
76 engine are available in the Engineering Reference (NREL 2015). EnergyPlus is available as a free download, as are all updates. Senior supervisor of this PhD study, Dr Ian Ridley, is a long time user of EnergyPlus; thus there was great access to training and help with troubleshooting the models. Through implementing a program developed by Phillip Biddulph at University College London, bulk processing of EnergyPlus simulations was possible. The .csv file output also enabled automated analysis using macros written in Excel.
EnergyPlus comes with a weather file builder in the box, and this is flexible enough to build weather files with one minute readings if this is available without gaps in the data source.
The thermostat settings are also customisable for each minute of each day, for every day of the year. This allows for on site temperature data to be input directly for specific time periods and mimicked exactly by the building model. The only shortcoming was that EnergyPlus does not automatically generate Energy Rating Certificates. It is technically possible to manually generate them if the base assumptions made in the Chenath engine were matched. This was not seen as critical to the research and not pursued.
Both AccuRate and EnergyPlus are highly respected software packages with varying strengths. The main advantages of customising weather and temperature inputs for EnergyPlus up to a 60-second frequency and defining unique profiles for each day, made it the best choice for this research project. Based on this assessment, EnergyPlus was selected as the building simulation software for this research.