Dataset on the solution to the guaranteed-service time inventory model across 38 real-world supply chains with normally distributed demand

The placement of safety stock to guarantee a service time has received considerable attention due to its impact on reducing inventory costs across multi-echelon supply chains. This work compiles and analyses the structure of 38 real-world supply chains reported in the literature by characterising their depth, degree, and stage types (distribution, manufacturing, part, retail, and transport). We provide the detailed solutions of the guaranteed-service-time inventory model under normally distributed demand, obtained using branch-and-cut algorithms implemented in commercial optimisers. The resulting dataset includes the optimal safety stock levels, the corresponding CPU time, and full variable values for every instance, with the largest model containing 4,050 variables and 18,890 constraints. All data are provided in CSV format to enable flexible reuse, and the source code for reproducing the plots is openly available. This dataset offers a rigorous benchmark for researchers developing new methods for solving the guaranteed-service-time inventory model, particularly for the four instances where a 25% optimality gap remains.