I have a three composite system of the form $H_{\text{tot}}=H_{ab}\otimes H_c$ where the system $C$ is behaving as the dissipator or the environment (I can model it as a thermal bath). And it is coupled only to system $B$ but not $A$. While $A$ is coupled with $B$ and entangles with it under time evolution. At $t=0$ is can take a composite state completely separable $H_{\text{tot}}(0)=H_a\otimes H_b \otimes H_c$. My objective is to solve the Master equation (more precisely the Lindbladian form) for $\rho_{ab}.$
But when I do that (with the partition as $H_{ab}|H_c$), the sub-system $A$ trivially disappears from the equations of motion. Because it does not couple with $C$ directly but only acts via $B$ indirectly. What is the right way to model this kind of interaction?