I'm using the dual duct terminals to model DOAS so that I can correctly do DCV (with other terminals, the OA rate will reduce but the total fan flow and power will not). It's mostly working except for some strange artifacts. On random dates, but always near hvac startup or shutdown, the OA rate spikes up to the maximum fan flow value. The return and relief air nodes, and the mixed air mass flow rate output variable, always show the correct flow (no spikes). The OA fraction reads larger than one (the ratio of the OA rate to the return air rate), and the supply outlet node flow has the spikes. Again, this happens RANDOMLY. There is no pattern that repeats weekly. Additionally, multiple air loops that share the same inputs show DIFFERENT behavior, reinforcing the indication that this behavior is random. I've tried all combinations of adding and removing the min OA schedules in the outdoor air controller and design specification outdoor air objects and nothing fixed the problem (in fact, adding the MinOA schedule to the design spec oa objects made it worse in some cases). Is this a bug or am I doing something wrong?