Modeling Water-source Variable Refrigerant Volume Systems

Hey everyone, I have a question about the best software choice and modeling strategy for water-source VRV (or VRF) systems. Our proposed design will use a water-source variable refrigerant volume system with refrigerant heat recovery. The office building is a three story rectangle with long north- and south-facing exposures. Due to capacity limitations, each floor will be split onto two VRV condensing units, one serving the east half, and one serving the west half, so that each VRV system utilizes refrigerant heat recovery between the north and south faces. Along with the six VRV condensing units, dedicated outdoor air heat pump units will also be tied to the same water loop. The water loop will have variable primary pumping. The temperature of the water loop will be maintained by a gas-fired condensing boiler and a dry cooler on the roof. We have immediate access to Carrier HAP and eQUEST modeling software. HAP can model air-source VRV, but it cannot explicitly model water-source VRV, so I tried some alternate methods. First I approximated our design in HAP as a WSHP system, hoping that the heat recovery of our designed refrigerant systems would be captured on the modeled water loop, but the results did not reflect this. Also, the pumping energy of this modeled system appeared to be higher than it should have been, leading me to believe HAP was not accurately modeling the pumping turn-down. I then modeled in HAP an air-source VRV system with refrigerant heat recovery, and electric resistance auxiliary heating (the only option) in low ambient conditions. This reflected the refrigerant heat recovery performance but did not account for enough pumping energy, and the unloading performance of the modeled VRV condensing units responded directly to the ambient air temperature instead of the water loop. For eQUEST, I've seen some documents that show how it may be possible to insert equipment performance curves from manufacturer's data into the .inp file to model water-source VRV. However, due to time and budget concerns, we would like to maintain just one model throughout the project for both load calcs and energy modeling. While I'm not as familiar with eQUEST, my understanding is that its capabilities are not strong at producing zone-level load results used for equipment sizing, so if we need to maintain a HAP (or other) file for loads anyway, eQUEST might not be the best option for the energy model. While we don't yet own a Trane TRACE 700 seat at my current firm, we have also looked at purchasing this software as an option since I have some past experience with the program. CDS says TRACE cannot yet model water-source VRV explicitly, but they've seen users approximate the system by using an air-source VRV system and altering the condenser performance curves to reflect the water loop temperatures instead of the air dry bulb. My thought is we could then lump all the VRV units onto the same modeled system so that TRACE would account for refrigerant heat recovery of the entire system, approximating how our water loop would recover heat between the condensing units. TRACE would also allow us to obtain load calcs and system sizing from the same model. Does anyone have experience modeling this system? What is the best software to use? Have you submitted this system for credit EAp2? It seems like, due to the limited availability of software programs that can accurately model the system, LEED reviewers should have some precedent to fall back on as far as accepting approximate modeling methods. If you have submitted using a work-around, what did you run into? Thanks for the help.