We present robust optimization techniques for dynamic systems which are affected by time-varying uncertainties. After reviewing existing techniques from the field of reachability analysis and ellipsoidal calculus, we discuss how to over-estimate the influence of uncertainty in nonlinear dynamic systems. The corresponding strategies lead to a framework which can be used to solve min-max optimal control problems in a conservative approximation. The technique is illustrated by applying it to a robust optimal control problem for a nonlinear jacketed tubular reactor. Inside this reactor a highly nonlinear and exothermic chemical reaction takes place which is uncertain due to fouling at the reactor wall. We regard safety constraints on the temperature which must be satisfied for all possible scenarios. (C) 2012 Elsevier Ltd. All rights reserved.