A marine project consists of series of operations, with each operation subject to a predefined operational limit and duration. If actual weather conditions exceed the operational limit, the operation cannot be executed and hence downtime occurs. An accurate assessment of uncertainties and the expected downtime during a marine project is important in the tender and execution phase. This paper proposes a new probabilistic model for downtime estimation. It utilizes linked Markov chains that use actual metocean conditions to produce binary workability sequences for each operation. Synthetic time-series can be generated based on the statistics of the past observations and more project simulations are realizable, reducing the simulation uncertainty. The capabilities and limitations of the proposed approach are illustrated in a case study for a hypothetical project in the Tasman Sea.