Shape sensing with optical fiber sensors is getting attention for its potential in many different applications. However, the failure of a sensing core during operational conditions is a possibility that must be considered since it may jeopardize the shape reconstruction. In this study, we perform Monte Carlo simulation of the curve reconstruction process for a shape sensing scheme to study how the sensing performance changes when one of the sensing core fails in presence of several noise levels using a simple compensation methodology. Results show that the accuracy degrades, but measurement is still possible. Moreover, the results seem to be dependent on the curve evolution in the three-dimensional space, suggesting that further research is still needed in this field.