[en] We describe a stochastic dynamic programming model for maximising the revenue generated by regasification of LNG (liquefied natural gas) from storage tanks at importation terminals in relation to a natural gas spot market. We present three numerical resolution strategies: a posterior optimal strategy, a rolling intrinsic strategy and a full option strategy based on a least-squares Monte Carlo algorithm. We then compare model simulation results to the observed behaviour of three LNG importation terminals in the UK for the period April 2011 to April 2012, and find that there was low correlation between the observed regasification decisions of the operators and those suggested by the three simulated strategies. However, the actions suggested by the model simulations would have generated significantly higher revenues, suggesting that the facilities might have been operated sub-optimally. A further numerical experiment shows that increasing the storage and regasification capacities of a facility can significantly increase the achievable revenue, even without altering the amount of LNG received, by allowing operators more flexibility to defer regasification. - Highlights: • We present a revenue maximisation model for LNG (liquefied natural gas) storage tanks at import terminals. • Three resolution strategies: posterior optimal, rolling intrinsic and full option. • The full option strategy is based on a least-squares Monte Carlo algorithm. • Model simulations show potential for higher revenue in three UK LNG terminals. • Numerical experiments show how storage and regasification capacities affect revenue.