However the CentralHeatPump only seemed to go to mode 1,2,3... not mode 4 or 5.

I just ran CentralChillerHeaterSystem_Simultaneous_Cooling_Heating.idf annualy. OutputVariable Chiller Heater Operation Mode Unit <X> is 0 to 3 most of the time, but it's sometimes 4 and 5. You can find the idf file and the TimeStep result here.

Further, the inlet temperatures were quite high for the GHX

When I ran CentralChillerHeaterSystem_Simultaneous_Cooling_Heating.idf, Ground Heat Exchanger Inlet Temperature was below 20C in January. I used the weather file USA_IL_Chicago-OHare.725300_TMY2.epw. It might be different from the weather file you used. One thing to note is that you should check Ground Heat Exchanger Mass Flow Rate at the same time. Ground Heat Exchanger Inlet Temperature is meaningful only when there is water flow.

GHX energy transfer only negative; should be positive in heating mode, and negative in cooling mode.

You are right. I think the cause of the weird result (i.e., bugs in EnergyPlus) is more like GroundHeatExchanger:System rather than CentralHeatPumpSystem. The result of CentralChillerHeaterSystem_Simultaneous_Cooling_Heating.idf shows that

• When Chiller Heater Operation Mode Unit <X> = 2 (heating-only mode), Ground Heat Exchanger Heat Transfer Rate < 0.
• When CHILLED WATER LOOP:Plant Supply Side Cooling Demand Rate > 0 and Chiller Heater Operation Mode Unit <X> = 1 (cooling-only mode), Ground Heat Exchanger Heat Transfer Rate and Ground Heat Exchanger Mass Flow Rate are 0 even if COND CIRC PUMP:Pump Mass Flow Rate > 0.

I revised CentralChillerHeaterSystem_Simultaneous_Cooling_Heating.idf by replacing GroundHeatExchanger:System to GroundHeatExchanger:Surface which I copied from the other ExampleFile SurfaceGroundHeatExchanger.idf. Please find the revised idf file here. GHX energy transfer in the revised file is positive in heating mode and negative in cooling mode.