Hello,
The empirical trough model issue that we fixed in SAM 2013.1.15 is a little complicated to explain:
The empirical trough model uses an iterative algorithm to calculate the startup energy required at the start of each day. The error involved the value used in the iteration for the system's average temperature in the time step before start-up, which should have been fixed at the system's average temperature in that time step, but instead was allowed to fluctuate and tended to follow the iterated system temperature value in the current time step. This error caused the startup energy value to be significantly lower than it should have been.
Best regards,
Paul.