Grounding and stranding of tankers are some of the main causes of pollution to the sea environment, due to the usually large amount of oil which is spilled if rupture of the tanks should occur. Contrary to other types of accidents, the resulting damage for such situations is not directly visible. This is due to a difficult, if not impossible, access and investigation of the bottom area until the ship is actually removed from the stranding location. Tools for the assessment of the structural damage in these situations, with particular focus on penetration of the indenter into the ship bottom and reduction of the flexural capacity of the hull girder, are therefore of primary importance. In this respect, the present work adopts a procedure for identification of the type of indenter along with the assessment of its penetration into the bottom. This procedure is based on measurements of the draughts of the ship at the stranding site during ebb tide. The assessment of the damage must be carried out in a variety of potential stranding situations, i.e. different types of indenters and different locations along the bottom of the ship. This is required in order to establish a framework for near-real-time decision support and intervention which is to be applied in real situations. In addition to this, uncertainties affect the measurements. These considerations call for the adoption of a probabilistic framework. In the present study Bayesian Networks provide this framework. On the basis of the history of the draught measurements during tide, the posterior probability density functions (pdfs) of the indenter and of its penetration into the bottom are computed and updated as more measurements become available. The modeling of the BN as well as results obtained by application of this procedure are illustrated.