A very good read..
 

https://www.ipfw.edu/dotAsset/239446.pdf

Lots of info although it is r-134a and not 290 but it still has a lot of info

It looks like they used heat exchangers that were just barely as large as the simulation software said they should be. When WaterFurnace did an ASHRAE test on the rig, it performed very closely to the simulation projections. The unit pretty much did exactly what they built it to do. They just cut their specs too close to allow any losses in the system.

If the design team would have oversized one or both heat exchangers, it could have performed much better. They didn't insulate any refrigerant lines, either. In the article, it mentions that the design team had a miscommunication with WaterFurnace. WF changed both the flow rates and the evaporator water entering temperature on them during the project, to have the unit meet a higher ISO standard as well as an original ASHRAE standard. As expected, the prototype met the ASHRAE standard, but not the tougher ISO standard. It came really close, though.

The paper mentions that the team went on to help WF revise the prototype, replacing the brazed plate exchangers with coaxial. I wonder if a few seniors got jobs during the event...

I saw that too.

After all the very useful information on design and implementation, especially how straightforward it was to change part selection of BP HXs by adding or subtracting plates, the paper mentioned the abrupt and unexpected change in design criteria to coaxial HX, with no explanation as to why.

Very curious...

-AC