How can I determine why there is a difference in results between models/versions?
When you see a difference in results between models, or between results from different versions, there's often a quick way to diagnose the cause of the difference.
In Vista, select one room and one day and view the results side by side. First look at the default graph of air temperature and space conditioning sensible.
1. If (over part of the day) the temperature is controlled and the load is different: Look at the sensible heat balance and identify the variables that have changed. This often leads directly to the cause of the difference.
2. If the temperature is free floating (and the load is therefore zero): Look at the sensible heat balance and identify the variables that have changed. In this case the cause can be less obvious because some of the variables will have changed as a knock-on effect of the change in temperature.
The task is now to distinguish cause from effect. Let's suppose it's a cold day and the temperature has fallen (adjust the recipe for other combinations). If the change is due to an increase in the fabric heat loss properties (say an increased U-value) the external conduction loss will have increased (despite the fall in temperature). On the other hand, if it's due to a reduction in gain (solar gain, internal gain, internal conduction, etc) the external conduction loss will have decreased (as a result of the fall in temperature). The same argument can be applied to ventilation heat loss. Once you've established the immediate cause of the difference you can investigate the reasons and get to the root of the problem.
