Experimental Validation of a Multiphysics Model for a Cold Flow Unit

Abstract

Abstract Cold Flow (CF) technology is an alternative to conventional flow assurance measures (pipe insulation, heating, pigging and chemical injection) due to its lower cost, footprint and emissions. The CF concept enables the flow of oil and gas production fluids at thermodynamic equilibrium in uninsulated pipes by generating small, dry and inert hydrate and wax particles suspended in a cold liquid stream after the subsea wellhead. The interaction involving the cooling process of crude oil, seawater and dry particle formation is complicated to evaluate only through experimental assays. For this reason, two models were implemented to simulate the CF unit and predict the wax and hydrate deposits over time and pipe length. The first was a 1D model, where experimental temperatures were used to estimate the internal deposits for 0% and 2% WC conditions. For the 0% WC, a linear increase in the deposits over time was observed at the beginning of the cooler pipe. At the end of the cooler pipe, no significant increase in deposit thickness was observed due to thermodynamic equilibrium. Adding seawater in the slurry reduced the internal deposit thickness because water can interfere with the slurry heat capacity and mass transfer. For the second model, a transient 3D model was built based on the assembly of the medium-scale Cold Flow unit (2-inch and 300 m long cooler pipe). Simulations were implemented using the Finite Element Method in COMSOL Multiphysics 6.3. The turbulent flow and heat transfer in fluids physics were combined iteratively to predict the velocity profile, temperature and wax and hydrate deposition rate. The molecular diffusion, shear stripping and shear dispersion mechanisms and the Matzain model were considered in the deposition rate. For the model validation, the experimental temperatures were compared with the simulation results. At the beginning of the cooler process, the temperature predictions and deposit thickness matched the experimental results. After 3 hours of the deposition process, the Matzain model overestimated the deposition rate due to not considering the seeding (recirculation of the cooler pipe downstream) to accelerate the wax and hydrate formation reaction in CF technology. The transient 3D Multiphysics model could predict the CF slurry temperature with an absolute error lower than 3 °C. Thus, we expect the developed models will enhance the development of a subsea pilot-scale Cold Flow unit in the near future.