A finite element analysis methodology employing rp-adaptivity is developed. The p-refinement phase is facilitated through the addition and deletion of general serendipity element edge nodes. The r-refinement phase is based on the reduction of error in the solution using element area as the grid optimization design variable. A series of tiered linked list data structure representations are developed for storing and manipulating the model information. The Zienkiewicz-Zhu error estimator is utilized for determining localized error. A modified superconvergent patch recovery technique is implemented to recover highly accurate nodal gradients utilized in the error estimation phase. Numerical results are presented for an idealized fluid flow problem.