Design of batch plants under demand uncertainty is studied through a mathematical programming approach. The problem is posed as a mixed integer linear program (MILP). Scheduling needs to be considered, since it forms an integral part of batch plant design. Instead of embedding the scheduling constraints in the design problem, leading to an extremely large problem, scheduling is considered as a subsequent stage solved after the Design SuperProblem. This decomposition has the advantage of including scheduling considerations of complex design problems without complicating the design problem itself. Infeasible scheduling problems are used to construct inequalities to eliminate the infeasibility-causing design solutions. Demand uncertainty is considered at the Design level as a set of scenarios, each of which is associated with an occurrence of uncertain parameters. Probabilities are assumed to be known a priori. The model is modified to address retrofit problems. Constraints are proposed for flexibility, and a regret analysis is investigated.