The Storage Location Assignment Problem (SLAP) is a key driver of warehouse efficiency, as it simultaneously impacts picking effort and replenishment frequency.
In previous work, we addressed SLAP as a max-flow min‐cost assignment between products and locations, solved via a Hungarian algorithm.
In many automated systems, however, adjacent storage positions can be merged to increase the volume allocated to a single product, offering fewer replenishments at the price of consuming more locations and sometimes degrading picking accessibility.

We introduce a merge-aware SLAP in which such mergeable neighbor groups are explicitly modeled.
Our solution framework has three stages:
(i) a baseline min-cost assignment providing a reference cost and flow;
(ii) a greedy activation of mergeable groups that preserves full feasibility and yields a structured MIP start; and
(iii) a mixed-integer linear “polishing'' model that maximizes effective storage usage under a tight cost constraint.

Computational results on a large industrial instance show that controlled merging can significantly increase the number of exploited locations while keeping the total cost close to the classical SLAP optimum, and reveal a clear trade-off frontier between picking and replenishment performance.