Factories for which steam is indispensable typically form an industrial park to share steam heating network (SHN) infrastructure. The energy systems in such parks are usually operated as microgrids. To improve operation efficiency, potential flexible and controllable resources should be utilized. This article proposes a novel method of utilizing the hydraulic inertia of SHNs to improve flexibility for these microgrids. A dynamic hydraulic model is established to characterize the steam storage property of SHNs. Based on the model, a joint dispatch problem formulated as a difference-of-convex problem is developed for the SHN and the electric power network (EPN). The convex-concave procedure method is employed to deal with the nonconvex constraints in the problem. Considering the EPN and the SHN are managed and operated independently by different entities, a modified generalized Benders decomposition (MGBD) algorithm is proposed to solve the problem in a decentralized manner for privacy preservation. Numerical results indicate that utilizing the hydraulic inertia of SHNs can reduce the total operating cost, the peak load of the tie transformer, and distributed renewable energy curtailment. The MGBD method outperforms the baseline methods in terms of convergence and efficiency.