Paraffinic crude oils are often transported at conditions where heavy hydrocarbons contained in the oil may precipitate as a solid waxy material. This precipitation may plug the processing equipment. In the design of oil production facilities, it is therefore essential to be able to predict the conditions where wax crystallization may take place and the amounts of wax which may form. Several models have been proposed in the literature for the prediction of wax precipitation from natural crude oils. But they generally overestimate considerably the amount of wax precipitation below the wax appearance temperature (WAT).
The objective of this study has been to develop a reliable compositional thermodynamic model to compute wax appearance temperature as well as the amount and the composition of waxy deposits as a function of temperature and to verify the proposed model with experimental results. This paper will describe the thermodynamic model and the complementary laboratory test methods. The agreement between the measured and predicted values for two crudes of different production areas is good.
The model described hereafter calculates the thermodynamic equilibrium between gas, liquid and solid phases. It represents the non-ideality of the gas and of the liquid phases with an equation of state linked with the group contribution model proposed by Abdoul and Peneloux. It has a modular design leading to flexibility; i.e., for the representation of the solid phase, it describes the behavior of this phase either by assuming it to be an ideal mixture or a non-ideal system which can be described by a Margules equation containing a single parameter characteristic of binary interactions.
This parameter can be fitted on available experimental data. The fitted parameter can be further used for predictive calculations of thermodynamic equilibrium for the crude when its composition is varied (i.e., when adding a solvent). It is thus necessary to experimentally determine the curve giving the amount of waxy deposit versus temperature at atmospheric pressure for a given crude. This has been performed both by Nuclear Magnetic Resonance (NMR) and Differential Scanning Calorimetry (DSC).
Besides the thermodynamic equations, such a model requires an analytical representation of the crude oil and the evaluation of the thermodynamic properties of the different pseudo-components. The predictive computation (i.e., assuming an ideal solid solution) of the onset crystallization temperature has lead to values very close to the measured ones. Through the use of the Margules equation, it has been checked that the model is able to predict the decrease of both the onset crystallization temperature and the deposit amount when light fractions (i.e., a condensate) are added to the studied crudes.
DSC and NMR are used to determine wax precipitation from crude oils by measuring the amount of wax at different temperatures below the WAT, NMR and DSC data will be then compared.
Proton pulsed measurements have been already used by Pedersen and al. The measurements have been made with a Bruker Minispec PC20 pulsed-NMR instrument. The experiments have been carried out at constant temperature. The NMR signals are proportional to the mobility of protons in the solution. We can so distinguish between protons present in the liquid state and protons present in the solid state. P. 283^