Unfortunately, when modeling things hardly ever go as planned at the first try and more than once you end up with results that you did not expect. At first glance everything in the symbolic model (i.e. the variables, constraints, and parameters in the model tree) might look OK, but still you are getting results that do not make sense or the solver concludes that your model is either infeasible or unbounded.
Luckily, for these cases, AIMMS provides a couple of tools that allow you investigate what is actually being sent to the solver (i.e. the complete generated problem). Investigating the information from these tools, allow you to figure out what part of your model is causing the unexpected results.