Ethylene Glycol and Its Mixtures with Water and Electrolytes: Thermodynamic and Transport Properties Peiming Wang,* Jerzy J
Total Page:16
File Type:pdf, Size:1020Kb
Article pubs.acs.org/IECR Ethylene Glycol and Its Mixtures with Water and Electrolytes: Thermodynamic and Transport Properties Peiming Wang,* Jerzy J. Kosinski, Andrzej Anderko, Ronald D. Springer, Malgorzata M. Lencka, and Jiangping Liu OLI Systems, Inc., 240 Cedar Knolls Road, Suite 301, Cedar Knolls, New Jersey 07927, United States *S Supporting Information ABSTRACT: A comprehensive thermodynamic model has been developed for calculating thermodynamic and transport properties of mixtures containing monoethylene glycol (MEG), water, and inorganic salts and gases. The model is based on the previously developed mixed-solvent electrolyte (MSE) framework, which has been designed for the simultaneous calculation of phase equilibria and speciation of electrolytes in aqueous, nonaqueous, and mixed solvents up to the saturation or pure solute limit. In the MSE framework, the standard-state properties of species are calculated from the Helgeson−Kirkham−Flowers equation of state, whereas the excess Gibbs energy includes a long-range electrostatic interaction term expressed by a Pitzer− Debye−Hückel equation, a virial coefficient-type term for interactions between ions and a short-range term for interactions involving neutral molecules. Model parameters have been established to reproduce the vapor pressures, solubilities of solids and gases, heat capacities, and densities for MEG + H2O + solute systems, where the solute is one or more of the following components: NaCl, KCl, CaCl2,Na2SO4,K2SO4, CaSO4, BaSO4,Na2CO3,K2CO3, NaHCO3, KHCO3, CaCO3, HCl, CO2,H2S, and O2. In particular, emphasis has been put on accurately representing the solubilities of mineral scales, which commonly appear in oil and gas environments. Additionally, the model predicts the pH of mixed-solvent solutions up to high MEG contents. On the basis of speciation obtained from the thermodynamic model, the electrical conductivity of the MEG + H2O + NaCl + NaHCO3 solutions is also calculated over wide ranges of solvent composition and salt concentration. Additionally, associated models have been established to compute the thermal conductivity, viscosity, and surface tension of aqueous MEG mixtures. 1. INTRODUCTION Because of the practical importance of MEG, a large number Ethylene glycol is an important industrial solvent and raw of research papers has been published in the literature about systems containing aqueous MEG, various salts, and dissolved material in a variety of processes. In the oil and gas industries, gases. The reported studies include data on phase equi- monoethylene glycol (MEG) is commonly used to reduce the − libria,4 50 speciation,26,51,52 caloric effects,53,54 den- risk of gas hydrate formation during the production and 11,12,53,55−57 33,58−62 sities, surface tension, and transport proper- transportation of hydrocarbons because gas hydrates pose ties such as viscosity and electrical and thermal conductiv- serious economic and safety problems by blocking pipelines or 31,53,54,63−70 1 ity. These data make it possible to develop plugging up wells and preventing gas production. At the same comprehensive thermodynamic and transport property models time, organic gas hydrate inhibitors such as MEG may cause for the simulation of chemical processes in which MEG plays a fl adverse scaling phenomena in drilling uids and produced role. The recent work of Fosbøl et al.71 on modeling the system water, which commonly contain high concentrations of MEG + water + CO2 +Na2CO3 + NaHCO3 represents an dissolved minerals. Therefore, the thermodynamic behavior of important advancement in the understanding and simulation of ff MEG-containing systems directly a ects gas hydrate and scale such systems but is limited to carbonates and bicarbonates as control, which are two of the key aspects of flow assurance in salt components. the petroleum industry. In addition, the presence of important Modeling phase equilibria in MEG-containing systems gas contaminants such as CO2,H2S, and O2 in crude oil and requires the use of a comprehensive thermodynamic model natural gas processing requires the knowledge of their for mixed-solvent electrolyte mixtures. At the same time, it is solubilities in fluids that may contain ethylene glycol. Moreover, highly desirable to have a computational framework that can be ethylene glycol is used as an antisolvent or an additive in used for predicting transport properties and surface tension as crystallization to obtain solid materials of desirable physical well as bulk-phase thermodynamic properties. Besides its quality and chemical purity.2,3 Therefore, a better under- obvious practical usefulness, such a framework would provide standing of salting-out effects and solubility behavior is additional insights into the properties of such systems. For necessary for the effective design and implementation of example, when analyzing transport phenomena (e.g., electrical optimum operating conditions and equipment for processes involving such complex systems. Accurate models are thus Received: June 19, 2013 clearly of vital importance to predict the phase and chemical Revised: October 10, 2013 behavior as well as other relevant thermophysical properties of Accepted: October 14, 2013 MEG-containing systems. Published: October 14, 2013 © 2013 American Chemical Society 15968 dx.doi.org/10.1021/ie4019353 | Ind. Eng. Chem. Res. 2013, 52, 15968−15987 Industrial & Engineering Chemistry Research Article and thermal conductivity, viscosity, and diffusivity), the dependent, symmetrical second virial coefficient-type expres- ionization behavior of electrolytes in MEG-water mixtures sion73 needs to be taken into account. Thus, a reasonable prediction G ex of speciation is important for the simultaneous representation II =−()∑∑∑nxxBI () RT i ijijx of phase equilibria and transport properties. i ij (2) In this study, we first apply a previously developed speciation-based thermodynamic model,72,73 referred to as the where Bij (Ix)=Bji (Ix), Bii = Bjj = 0 and the ionic strength MSE (mixed-solvent electrolyte) model, to selected MEG- dependence of Bij is given by containing systems with a particular focus on salts and gases BI()=+ b c exp( − I + a ) that commonly exist in oilfield waters. The MSE model was ij x ij ij x 1 (3) − previously shown to reproduce simultaneously vapor liquid, where bij and cij are adjustable parameters and a1 is set equal to − − solid liquid, and liquid liquid equilibria, speciation, caloric, 0.01. The parameters bij and cij are calculated as functions of and volumetric properties of electrolytes in water, organic, or temperature as 74,75 mixed solvents. In particular, the model is capable of 2 reproducing solubility variations with solvent and ionic bbij=+0, ij bTbTbTb 1, ij + 2, ij/ln + 3, ij +4, ij T (4) 76 composition in crystallization studies and accurately 2 represents phase equilibria in multicomponent inorganic ccij=+0, ij cTcTcTc 1, ij + 2, ij/ln + 3, ij +4, ij T (5) − systems containing multiple salts, acids, bases,74,77 84 and For most electrolyte systems, only the first three terms are ionic liquids.85 A combination of the MSE model and the necessary to represent the variations of thermodynamic extensive thermodynamic data that are available in the literature properties with temperature over a temperature range up to for MEG-containing systems provides an excellent opportunity 300 °C. Additional temperature-dependent parameters are for developing a comprehensive thermodynamic treatment. necessary only for a limited number of systems for which data After establishing the parameters of the thermodynamic model, analysis needs to be performed over an extended range of we develop associated models for electrical conductivity, temperatures.84 In cases where very high pressures are of thermal conductivity, viscosity, and surface tension. interest, a pressure dependence may also be introduced into the b and c parameters.84 2. THERMODYNAMIC FRAMEWORK ij ij The short-range interaction contribution is calculated from Details of the thermodynamic model have been described the UNIQUAC equation.87 When justified by experimental elsewhere.73,75 Here, we briefly introduce the fundamentals of data, the temperature dependence of the UNIQUAC energetic the model and specify the parameters that need to be parameters can be expressed using a quadratic function: determined on the basis of experimental data. The thermody- (0) (1) (2) 2 namic framework has been designed to provide a simultaneous aij=+aaTaT ij ij + ij (6) treatment of phase equilibria, ionic equilibria in the solution, fi and derivative thermodynamic properties such as enthalpy and In systems containing only strong electrolytes, only the speci c heat capacity. To achieve this objective, the framework consists ion-interaction parameters are needed to reproduce the of properties of the solutions. However, in nonelectrolyte systems such as the MEG + water binary, only the short-range (a) An excess Gibbs energy model that accounts for the parameters are needed. For electrolyte systems such as those nonideality of liquid systems containing ionic and neutral encountered in oilfield waters and brines where the ionic solute species in single or multicomponent solvents; strength and salt concentrations are significant, the specific ion- (b) A standard-state property model that determines the interaction contribution is the most important one to thermodynamic properties of individual species at infinite reproduce the properties of the solutions. When a chemical dilution and thus defines the reference state for process occurs