Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)Thermodynamic calculations are traditionally carried out under the assumption of specified input parameters. Errors associated to the results are not often estimated. Here, we propose a novel algorithm that propagates the uncertainty intervals on thermodynamic constants to the uncertainty in chemical equilibrium compositions. The computing uses a dataset of uncertainties on thermodynamic parameters for minerals, solution species and gases consistent with the SUPCRT92 database. Also the algorithm of nonlinear optimization is thoroughly described and realized on a base of the CRONO software. This code can be incorporated into reactive mass transport models as a core for calculating equilibrium compositions. The performance of the algorithm is tested in an experimental system involving Mont Tern's Opalinus Clay interacting with pore water. Its effectiveness is also evaluated against Monte Carlo simulations and Latin Hypercube sampling. (C) 2015 Elsevier Ltd. All rights reserved.79 118128Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)FAPESP [2011/12682-3]CNPq [164939/2014-8]