The IO Reference has recommended configurations for using thermal storage. In particular, I'm interested in the parallel configuration.

Parallel: In this configuration, the chiller is on a branch parallel to the storage unit branch when it is not charging. During charging mode, valves will be changed so that the chiller is in series upstream of the storage. To accomplish this in EnergyPlus, the chiller must be modeled using two different chiller objects to represent the same chiller. One chiller object on a parallel branch operates only when storage is not being charged. The other chiller object, in series upstream of the storage unit operates only during charging mode.

I was hoping that this could be accomplished with only one chiller (we are computing costs for the models we're generating and hoping to avoid the complication of two chillers representing one chiller in code) and a TemperingValve object. I think it will work if I use a setpoint manager with MaximumTemperature, or perhaps MinimumTemperature (not quite sure if I have it backwards or not) and correct schedules, I can get the tempering valve to charge the tank at night. I'll post whether or not it works in case it's of use to someone else.

But I'd really like to know before I try: Does anyone know if the code governing the TemperingValve object will be able to handle this since this isn't transferring hot water to a how water loop, which is what a TemperingValve is described to do in the IO Reference?