A novel methodology for optimizing thermal storage for biomass boiler heating systems is presented in this paper. Biomass boiler heating systems have gradually gained popularity in residential, commercial, and large-scale district heating plants due to the lower carbon footprint of biomass energy. One unanswered question in this field was the size of the thermal energy storage (TES) tank with respect to different boiler capacities and different types of heat demand profiles. Different standards and recommendations (based only on boiler capacities) have been proposed by various organizations around the world with large deviations on the sizing of the TES tank, but without a clear technical basis for these recommendations. These inconsistencies impede the effective design and deployment of biomass boiler heating systems. This study used a dynamic process simulation system based on field data from a 25-kW wood pellet boiler with radiant floor heating. By running the simulation for different boiler capacities (N) and different types of heat demand profiles (i.e. low, medium, high, intermittent, and variable), the optimum TES tank volume can be estimated by evaluating the TES tank discharge efficiency, boiler on time, boiler off-time, and maximum boiler output temperature. Linear correlations between optimum TES tank volume and boiler nominal capacity for different types of heat demand profiles were obtained. The results showed that the medium building heat demand (45% N) required the smallest TES tank volume while intermittent building demand requires the largest. High building demand (60% N) generates the highest TES tank discharge efficiency (up to 99%) but also requires large storage volume. From the economic point of view, it is recommended to size the thermal energy storage system such that the average building heat demand is around 45% of boiler nominal capacity. (C) 2018 Elsevier B.V. All rights reserved.