Thanks for your reply, but I am afraid it doesn't solve my problem. I want to model a return loop, ie an output from a process that goes back as an input to the same process. In my case it concerns reuse, but it could also be a case of e.g. internal heat recovery in an industrial process. So my question is if it is possible to model such return loops directly or if I have to go via a dummy process. Hope my question is clear!
So I think, now I understand your problem. Thanks for the clarification.
I have tried to work with your problem and made two rough sketches (please see attached).
While Fig. 1 shows only the system without its sub-systems, Fig. 2 provides the details of sub-systems and recycling flow.
It is possible to model internal loops. But be aware that these flows can not be computed by a mass balance (see attached example) because the balance equation would be F1+F2+F5=F3+F4+F5, what is the same as F1+F2=F3+F4. I.e. the value of the internal loop has to be given in order to be displayed.
All the best,