A unique requirement of underwater vehicles' power/energy systems is that they remain neutrally buoyant over the course of a mission. Previous work published in the journal of Power Sources reported gross as opposed to neutrally-buoyant energy densities of an integrated solid oxide fuel cell/Rankinecycle based power system based on the exothermic reaction of aluminum with seawater. This paper corrects this shortcoming by presenting a model for estimating system mass and using it to update the key findings of the original paper in the context of the neutral buoyancy requirement. It also presents an expanded sensitivity analysis to illustrate the influence of various design and modeling assumptions. While energy density is very sensitive to turbine efficiency (sensitivity coefficient in excess of 0.60), it is relatively insensitive to all other major design parameters (sensitivity coefficients < 0.15) like compressor efficiency, inlet water temperature, scaling methodology, etc. The neutral buoyancy requirement introduces a significant ( similar to 15%) energy density penalty but overall the system still appears to offer factors of five to eight improvements in energy density (i.e., vehicle range/endurance) over present batterybased technologies. (c) 2013 Elsevier B.V. All rights reserved.