Water splitting through the Hybrid Sulfur (HyS) process powered by solar energy is a promising pathway to the production of green hydrogen. The main challenges to the development of this process are related to the intrinsic variability of the solar resource, which, besides requiring the deployment of innovative process solutions, introduces significant elements of uncertainty in the plant design. In this paper, the Polynomial Chaos Expansion (PCE) method is applied for the uncertainty quantification (UQ) in this kind of systems. In particular, a forward analysis dealing with the evaluation of the output probability distributions is performed. This is carried out in terms of the input probability distributions, and the analysis is focused on how uncertainty is propagated from the input to the output. Moreover, a comparison between the PCE method and the standard Monte Carlo analysis (using the Latin Hypercube Sampling method) is performed. The obtained results show the advantage of the PCE approach in terms of convergence rate and the number of function evaluations. Finally, a sensitivity analysis through Sobol’ indices has been carried out, which highlighted the influence of each variation in the input on the output.
Uncertainty quantification in a hydrogen production system based on the solar hybrid sulfur process
Turchetti L.;Liberatore R.
2020-01-01
Abstract
Water splitting through the Hybrid Sulfur (HyS) process powered by solar energy is a promising pathway to the production of green hydrogen. The main challenges to the development of this process are related to the intrinsic variability of the solar resource, which, besides requiring the deployment of innovative process solutions, introduces significant elements of uncertainty in the plant design. In this paper, the Polynomial Chaos Expansion (PCE) method is applied for the uncertainty quantification (UQ) in this kind of systems. In particular, a forward analysis dealing with the evaluation of the output probability distributions is performed. This is carried out in terms of the input probability distributions, and the analysis is focused on how uncertainty is propagated from the input to the output. Moreover, a comparison between the PCE method and the standard Monte Carlo analysis (using the Latin Hypercube Sampling method) is performed. The obtained results show the advantage of the PCE approach in terms of convergence rate and the number of function evaluations. Finally, a sensitivity analysis through Sobol’ indices has been carried out, which highlighted the influence of each variation in the input on the output.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.