A framework for demand point and solution space aggregation analysis for location models

Many location problems can be formulated as minimizing some location objective function subject to upper bounds on other location constraint functions. When such functions are subadditive and nondecreasing in the distances (a common occurrence), worst-case demand point aggregation error bounds are known. We show how to solve a relaxation and a restriction of the aggregated problem in such a way as to obtain lower and upper bounds on the optimal value of the original problem. We consider some applications to covering and related problems.