Question-and-Answer Resource for the Building Energy Modeling Community
 | 1 | initial version |
OK, I believe I tracked this down to the chilled water loop.
The following is the CHW loop as it appears in OpenStudio 1.14.0. Notice in particular that the demand side only has the two chillers but no bypass. This is what I want since I am trying to simulate a variable primary only loop.
The following is the CHW loop as it appears in OpenStudio 2.2.0 after opening the 1.14.0 .OSM file from within it. Notice in particular that the demand side still has the two chillers but now it also has a bypass. This no longer simulates what I intend to simulate. The bypass, which appears out of nowhere during the 1.14.0 to 2.2.0 conversion, makes the loop behave totally different.
When I delete this bypass in OpenStudio 2.2.0, then the results almost fall back in line exactly. Still not perfectly, but nowhere near as drastically different as with the bypass in place. For example the unmet heating hours are spot on identical, the unmet cooling hours are 1.67 hours higher for 2.2.0 than for 1.14.0. I can live with that a whole lot better than if they were seven times higher. Still, a small nagging doubt remains - why is 8.7.0 even slightly different than 8.6.0 - shouldn't all bugs that affect something so fundamentally standard as a chilled water loop have been ironed out by this high a software revision?
