Wax Deposition in North Sea Gas Condensate and Oil Systems: Comparison Between Operational Experience and Model Prediction

Abstract

Abstract Control of paraffin deposition is addressed in early phase of new field developments, as part of the fluid-related effects on the flow performance of the multiphase transportation system. Paraffins are precipitated by a decrease of the fluid temperature. During the design phase, a combination of experiments and modeling is used to predict potential wax related problems and determine possible solutions for wax control. High reliability is required from wax deposition models in order to avoid over protective operational guidelines, over design of the transportation system and thereby reduce cost in terms of insulation, pigging and chemical injection. The main objective of this study is to compare wax deposition predicted by a simulation model to operational data for two different North Sea fields. The first case contemplates a 10-inch, approximately 21 km long multiphase and subsea production pipeline from the High Pressure gas condensate field A to the central gas-processing complex. The production has now come to an end. The second case is a single phase, stabilized oil being transported through a 16 inch, approximately 43 km production pipeline from a processing platform at field B to field C for storage and offloading. OLGA 2000 with wax module has been used for the dynamic wax deposition simulations. The physical and thermodynamic fluid properties used in the simulations are also presented and discussed. Various model parameters and their influence on modeling results are evaluated. Field experience after many years of production is available from the gas condensate field A. Although wax deposition was predicted in the gas condensate line, no significant deposit has ever been observed. Wax control for the field B pipeline is currently based on regular pigging every few days combined with continuous injection of wax inhibitor. There has not been sufficient field data available for a very detailed comparison with modeling results. Anyway, operational experience from two very different fields gives some important insight into the strengths and weaknesses of current modeling tools and practice. A main conclusion of this paper is that wax deposition under field conditions seems to be less severe than predicted by the model for the multiphase gas condensate line, while it seems to be reasonable agreement between modeling and field experience for the single-phase oil case. Introduction Wax deposition occurs on the inside surface of a flowline when the pipe wall temperature falls below the Wax Appearance Temperature (WAT), or cloud point of a paraffinic hydrocarbon liquid flowing in the line, and is lower than the bulk fluid temperature. The current approach in prediction of wax control strategy for new fields developments, is based on two steps. First, necessary experiments and simulations are performed in order to establish the wax deposition potential for a given fluid composition and thermo-hydraulic conditions. The degree of deposition will depend on the amount of wax components dissolved in the oil available for deposition, the rate of heat loss to the surroundings (or the temperature gradient at the pipe wall), the shear rate at the wall, the viscosity, water cut etc., and can be estimated using commercial software packages. The prediction models are still not very accurate, but they are believed to provide reasonable estimates of the rate of wax deposition and thus provide a basis for assessing the needs for wax control. The wax control strategy is then defined by taking into account other flow performance constraints and wax specific requirements. There are several methods for wax control in pipelines, such as pigging, chemical injection, thermal insulation and active heating. Currently, insulation and pigging are the most widely used methods. In an early phase and design phase evaluation, a maximum wax layer thickness of 2–3 mm is often used as a criterion for when a pipeline should be pigged.