This report present a study by FOI to review the current FOI model in Reglemente Am- och minröj - Skyddsåtgärder (Regulations for Ammunition and Mine Clearance - Protective Measures) and to further develop it. In the regulation, the model is presented without traceability, which makes it necessary to validate and review the fundamental assumptions of the FOI model. The review also identifies areas for further development. The work on the development is then presented as suggestions for updating theexisting models in order to improve the risk calculation method for fragment ejection. The review and further development of the method are presented in detail in order to create increased traceability and reliability in its use. The FOI model is a method for calculating risk distances in case of fragment ejection, and it is used for explosive charges where no established risk distances exist, as well as for explosive charges where, for various reasons, a better understanding of the potential risks is desired. The method consists of simple computational components (sub-models) that are suitable for practical use, and the same requirements for simplicity and usability also apply to the further development. A literature review is presented, and the underlying material for the sub-models are found. However, the references for the sub-models do not themselves indicate what the sub-models are based on. A further review shows that the current model for fragment deceleration underestimates the fragments' velocity along the trajectory when compared to numerical simulations. For the models concerning the fragments' maximum range and penetration capability, a need is identified to extend the validity range to lower ejection velocities respectively impact velocities, as well as to include more mass classes to facilitate better usage. In the conducted further development, it is proposed, among other things, that:  the ejection velocity be calculated using the Gurney  the fragments' maximum range be calculated numerically, with the optimal launch angle used for each combination of mass and ejection velocity  the fragments' velocity as a function of distance be calculated by dividing the deceleration into two different velocity intervals where a constant drag coefficient is assumed for each interval  the validity range for the model for penetration be adjusted to include lower impact velocities. Furthermore, the sub-models for maximum range, deceleration, and penetration for different fragment materials and shapes are provided, which are largely missing in the current version. This creates a coherent method in which the output from one sub-model can be used as input for the next sub-model. For natural fragmentation and the formation of fragments, no basis for the fragments' size has been identified during the literature review. It is therefore currently not possible to say how well the current sub-model for fragment size corresponds to reality. For the same reason, no specific change to this sub-model is proposed.