(480a) Prediction of the Fugacity of Hydrogen Gas Dissolved in Organic Solvents with Emphasis on the High-Temperature High-Pressure Subcooled Liquid State

The so-called Nelson Chart is used to determine susceptibility of steel piping to High Temperature Hydrogen Attack (HTHA) based on the type of steel, temperature and the amount of hydrogen in a process stream. Although the chart is presented in terms of the partial pressure of hydrogen in the stream, this variable is unsuitable for evaluating streams where you have subcooled liquid and no vapor phase is present. In this situation the desired thermodynamic property reflecting the concentration of hydrogen in the stream is the fugacity of dissolved hydrogen. This study investigated the accuracy of rigorous methods of calculating the fugacity of hydrogen dissolved in hydrocarbon streams typical of the petroleum refining industry. Several equations of state were used to calculate the fugacity of hydrogen dissolved in saturated and subcooled liquid hydrocarbon phases. The predictions of two correlative equations of state (EOS), Peng-Robinson and PC-SAFT, were compared to experimental data for the fugacity of dissolved hydrogen. In addition, values calculated using two other ‘predictive’ methods, PSRK and VTPR, were examined.

The dissolved hydrogen fugacity data were obtained by integrating binary experimental data for the partial molar volume (PMV) of hydrogen dissolved in organic non-polar solvents. A literature survey identified hydrogen high-temperature high-pressure PMV data for six solvents: n-pentane, 2,3-dimethylbutane, benzene, cyclohexane, n-octane, 1,4-diethylbenzene, m-xylene and n-decane. As a preliminary step, the Peng-Robinson and PC-SAFT EOS were fit to binary vapor-liquid equilibrium (VLE) data for hydrogen with the respective solvents. The resulting fits of these two equations, as well as the predictions of the three other models were compared to the VLE data. The results of the comparison indicate that the correlative models adequately represent the VLE data, albeit with large values of the binary interaction coefficients. The 2 fitted models and predictive models were then used to calculate values of the partial molar volume and fugacity of hydrogen in the binary systems over a wide range of temperature, pressure and hydrogen concentration.

Using the EOS fit to VLE data and the two predictive models, additional calculations were performed to predict values of hydrogen fugacity in multicomponent liquid phase streams typical of the refining industry with products such as naphtha, kerosene and diesel. The results of this study are useful in validating short-cut methods for predicting hydrogen fugacity in subcooled streams to be used in HTHA susceptibility.