How to interpret cycling equipment outputs?

edit

There are 2 control methodologies in EnergyPlus. A constant flow system where a load or set point is the target, or a cycling system where a load is the target.

A constant flow system "outputs" an average SAT, where the compressor cycles on and off to meet the load or temperature set point (the fan never cycles off). These systems are the CoilSystem or VAV types (and now AirloopHVAC:UnitarySystem). A water coil or DX coil can be used in these applications.

A cycling system "outputs" a compressor on SAT while averaging the flow rate (PLR*MaxFlow while RTF is used for power/energy calculations). These systems are the furnaces, heat pumps, and unitary systems. These systems typically use a DX coil and refrigerant compressor but can also be applied to water coils.

In the Q = mCpdT calculation, a constant flow system has a constant air flow (m) and the dT is the average (Tout(PLR) + Tin(1-PLR)). A DX system which uses a cycling fan has a constant dT while the air flow is averaged. Similar calcs for Q = mdH. When water coils are used, there is rarely any cycling involved and the water coil water-side flow rate is constant the entire time step (but will modulate to meet the load or temperature set point).

Although real systems would cycle (cycling compressor or water control valve opening and closing) over a long interval, on the order of 10's of minutes, EnergyPlus simulates this (DX) cycling over shorter time intervals using a part-load degradation curve to account for cycling losses. This methodology allows EP to hit a zone thermostat temperature set point (since no thermostat hysteresis exists in EP) or coil outlet air temperature set point.

The AirloopHVAC:UnitarySystem object just recently began to allow a cycling system to meet a set point temperature [which can be difficult with cycling systems since temperature is constant while flow is proportional to PLR...HINT: use constant fan with SP control for AirloopHVAC:UnitarySystem].

Interpretation of results will therefore depend on the system type, either constant fan operation or cycling fan operation.

I could be overruled here and I'm assuming we are talking about "Detailed" HVAC timestep level report variables in EnergyPlus. The temperature and flow that is reported at the node (presumably we are talking about nodes after DX coils, but it doesn't really matter) represent the actual temperature and flow that the simulation is feeding into the zone. There is no averaging of any kind unless we are talking about report variables at the zone timestep or longer frequency. Flow is determined by the requests from the demand side terminals. A target temperature is established by a setpoint manager (often SetpointMangerSingleZoneReheat for packaged units where RTF applies), and the DX coil attempts to hit that setpoint by figuring out what fraction of its capacity (RTF) would be needed to hit that temperature at the already established flow rate. Most DX equipment has limited ability to part load and meet a particular node (supply) temperature, but the simulation isn't modeling reality. It just degrades the coil's performance using an empirical correlation involving the RTF. SetpointMangerSingleZoneReheat in particular computes a variable supply temperature to hit a particular zone's load at the system's current air flow rate, even though a real system would simply be cycling on and off to meet the zone load.

It is possible to use EnergyPlus's EMS feature combined with very short time steps to model real cycling, but you won't get that out of the box with EnergyPlus, and I'd say it is still a research project. For one thing you don't capture the latent degradation when you do this.

Averaging is going to happen for temperature and flow if you view report variables above the "Detailed" reporting frequency. There is some ambiguity as to what the correct temperature is for time steps when there is no flow. I believe that EnergyPlus will just carry over the previous timestep value.