Hello Parvathy,
I would probably model this as two reactions:
v1: A -> B
v2: A -> C
where you make the rates v1 and v2 dependent on some condition. I.e. this could just be a function f(p1, p2, p3, ..) which you add as a factor to the kinetic law or you could also use some event to change v1 and v2.
As an alternative: If you have a reaction A -> s1 * A + s2 * B, with s1 and s2 being the stoichiometric coefficients you could also make the coefficients non-constant, i.e. depening on a condition. This would shift your reaction either to A or B. As far as I know COPASI does not support variable stoichiometric coefficients (as do many other SBML tools). So I would probably go with the first suggestion here.
Hope this helps.
Best Matthias