Modifications to the subroutine OPALQP for dealing with large problems OPALQP is a Fortran subroutine implementing a sequential quadratic programming method for nonlinearly constrained optimization. It was originally designed for small to medium-sized problems on the assumption that the central calculation on each iteration -- the solution of a quadratic program -- would involve dense, rather than sparse, matrices. We describe a number of practical steps that have been considered for extending the applicability of OPALQP to larger problems. Among the modifications discussed are: the introduction of sparse data structures for the Jacobian matrix of constraint normals, the use of limited memory updates for approximating the Hessian of the Lagrangian, and the use of exact, rather than approximate, second derivatives. While none of these ideas are new in themselves, it is helpful to review the extent to which, individually and collectively, they can enhance the performance of an existing code. Such modifications are important because there is still comparatively little readily available software which has been designed from the outset for large-scale nonlinear optimization.