In e-commerce warehouses, replenishment operations involve the transportation of ordered items from reserve areas to forward picking areas. In this paper, we study where to store these items (called newly arrived items) in open storage locations in forward picking areas prior to picking to minimize the total travel distance for given picking orders. The locations occupied by items previously stored in the forward area are not available. The problem is formulated as an integer program and proved to be NP-hard. We propose a two-stage decomposition algorithm, showing that it provides high-quality solutions quickly. Compared with methods proposed in the literature and used in practice, the travel distance is substantially reduced. Also, our numerical study shows that considering the constraint that items have previously been stored, which has often been ignored in previous research, can improve the picking performance by more than $5\%$. We apply our proposed algorithm to real data from a Chinese 3PL retailer to show that it significantly outperforms the policy the company currently uses.