Decompression schedule optimization with an isoprobabilistic risk of decompression sickness

Divers use decompression schedules to reduce the probability of occurrence of decompression sickness when returning to the surface at the end of a dive. The probability of decompression sickness resulting from these schedules varies across different dives and the models used to generate them. Usually the diver is unaware of this variance in risk. This paper describes an investigation into the feasibility of producing optimized iso-probabilistic decompression schedules that minimize the time it takes for a diver to reach the surface. The decompression schedules were optimized using the sequential quadratic programming method (SQP), which minimizes the ascent time for a given probability of decompression sickness. The U.S. linear-exponential multi-gas model was used to calculate an estimate of the probability of decompression sickness for a given dive. In particular 1.3-bar oxygen in helium rebreather bounce dives to between 18 m and 81 m were considered and compared against the UK Navy QinetiQ 90 tables for a similar estimate of probability of decompression sickness. The SQP method reliably produced schedules with fast and stable convergence to an optimized solution. Comparison of the optimized decompression schedules with the QinetiQ 90 schedules showed similar stop times for shallow dives to 18 m. For dives with a maximum depth of 39 m to 81 m, optimizing the decompression resulted in savings in decompression time of up to 30 min. This paper has shown that it is feasible to produce optimized iso-probabilistic decompression tables given a reliable risk model for decompression sickness and appropriate dive trials.