Reliable Wax Predictions for Flow Assurance

Abstract

Abstract A number of wax models are currently in use by the oil industry which are based on parameters that were empirically determined to match available data for black oils. These data are often not very precise. The recently developed model of Coutinho is, however, based on high accuracy thermodynamic data. The paper describes how the Coutinho model can be used in conjunction with conventional equations of state to perform wax equilibrium calculations for black oils. Examples are given showing how well the model can predict both wax appearance temperature and the amount of wax precipitated at varying temperatures with or without n-paraffin analyses. The examples include the effect of pressure on live oils. Improved thermodynamic modelling of wax formation in turn allows better prediction of wax deposition rates for flow assurance. Introduction With the on-going trend to deep water developments, flow assurance has become a major technical and economic issue. The avoidance or remediation of wax deposition is one key aspect of flow assurance. The ability to predict wax deposition rates depends on a number of factors one of which is examined in this paper: the thermodynamic equilibrium between oil and wax. Wax is a solid phase formed from the components of the oil that have the highest melting points. For temperatures of operational interest, i.e. above ~0°C, wax consists predominantly of the C20+ n-paraffins. A number of engineering models have been proposed for calculating oil-wax equilibria, for example the work of Won [1], Hansen et al. [2], Erickson et al. [3], Pedersen [4], Rønningsen et al.[5], Lira-Galeana et al. [6] and Pan et al. [7]. The authors of all these models propose a number of correlations to predict the key thermodynamic parameters, but there is no direct experimental evidence to show that the assumptions made are correct. Instead the authors rely on experimental data for wax formation from oils to validate their models, predominantly measurements of wax appearance temperature (WAT). However, in a recent survey for Deepstar, Monger-McClure et al. [8] suggested that uncertainties in WAT for good modern measurements may be ±5°F. For older measurements the uncertainties can be considerably higher. Thus using data of this kind can only provide an approximate method to evaluate proposed models; it is not possible on this basis to discriminate in any detail between models. In order to put wax calculations on a firmer footing, Coutinho and co-workers have developed a wax model that is directly based on high-quality laboratory data for the properties of liquid and solid hydrocarbons and their mixtures [9,10]. The model is summarised in the Appendix. Coutinho went on to show that the model accurately predicts the waxing behaviour of diesel fuels, jet fuels [11] and crude oils [12]. The Coutinho model exists in two variants, the Wilson and Uniquac wax models. The Wilson model is simpler to apply as it treats the wax phase as a single solid solution of n-paraffins. The Uniquac model is more realistic in that it predicts that the wax phase splits into a number of coexisting solid solution phases; experimental evidence confirms this to be the case [13]. Both variants require the n-paraffin distribution of the oil to be specified; however, in cases where this is not available, it can be estimated from the total wax content using a method devised by Coutinho and Daridon [12]. The method can therefore make optimum use of whatever data are available for a particular oil.