SINTEF Energy Research is running a project on this topicaimed at investigating the fundamental mechanisms in electrocoalescence in heavy crude oils. In an earlier project on model systems electro-hydrodynamic instability between drops was identified as the main mechanism behind water drop coalescence. Historicallyoil-water separation has been done in large sedimentation tanks. These tanks can be very large with diameters of some 5metres and tens of metres in length. Sometimes sedimentation is sought enhanced by introducing electric fields in water-in-oil emulsions to help small water drops to coalesce into bigger drops and sediment faster. Howeverthe basic mechanisms of coalescence in an electric field are not fully understoodand several empirical studies on separation efficiency have pointed in different directions: ac or dcac on dcac with specific frequencies and even pulsed dc. Latelysmaller more compact coalescers have been introduced to save space on offshore platforms and seabed installations. Water bridging electrode gaps resulting in short circuit is a problem both ac and dc. To prevent this short-circuiting solid insulation is used to cover the electrodes. The basic difference between a conventional sedimentation coalescer and a compact coalescer is that a conventional one is based on the emulsion being semi stagnantwhile shear flow is applied to give more frequent drop collisions in a compact coalescer. In order to achieve coalescence two things have to be achieved. Firstone has to make drops collide; here both liquids flow and electrostatic forces plays a role. Secondone has to make the drops coalesce; here mainly electrostatic forces play a role. This will be explained in more details in the following. If a liquid is charged like eg petrol in a tankthe charge will leak away with time. The speed is determined by the charge relaxation time constant of the liquid. This time constant can be found from the formulae: In a stagnant emulsion the drops will mainly be influenced by gravitation and electrostatic forces. The electrostatic forces are either electrophoretic or dielectrophoretic forces. Electrophoretic forces are caused by an electric field acting upon a net charge carried by the particle as shown in Fig.1a. For a charged drop between parallel plate electrodes these forces do not depend on position but will vary with the applied voltage. Between drops the force is inversely proportional to the square of the distance between the drops and either attractive or repulsivedepending on whether the drops have opposite or equal polarity charge. Dielectrophoretic forces are caused by the dipole moment of a charge neutral water drop induced from an inhomogeneous field. The force on a polarised charge neutral water drop will always try to attract the drop towards the high field region like shown in Fig.1b. Between parallel electrodes in a homogeneous electric fieldthe dielectrophoretic force is zero. Howeverbetween adjacent water drops in any field configuration the force from the induced dipole moment will always be attractive. The force is inversely proportional with the distance to the fourthand is only significant within a distance of the order of the diameter of the smaller drop. When the drops get very closethe attractive forces between the drops increases rapidly. When considering electrocoalescers with stagnant liquids and homogeneous electric fields the drops have to be charged for a long-range force to act upon them before they get within close range. Electrochemical charging is unlikelyso the drops must acquire charge from electrodeswhich possible only for uncovered electrodes. The acquired drop charge will be proportional to the electric field and the diameter of the drop. Drops will then move with the field towards the other electrode and will when they meet other drops either be attracted or repulsed. Attractive forces will not occur until they meet drops from the opposite electrode. In the case of an AC fieldthe force acting on the charged drops will be reversed and the drop will be retarded once the voltage changes polarity. A relevant question is how long the drop will stay chargedwhich is given by the time constant of the oilcompared to the time needed to travel from the electrode where the drops are charged to the middle of the gap where they can meet oppositely charged drops. As attractive forces between drops are short-range forces it can be advantageous to introduce shear flow or turbulence in the liquid to allow drops to get within the range of the attractive forces. It is well known that water drops become unstable in electric fields once the electrostatic forces exceed the capillary pressure as shown in Fig.2a. This occurs at a critical field that is reduced with increasing drop size. Once drops in an electric field get within close range electrostatic forces will attract them to each otherand they may eventually coalesce. Experiments have shown that electrocoalescence mainly is due to two different mechanisms: o o The smaller the drops arethe higher field is necessary. If one applies a dc field the electrostatic forces will be strong all the timewhile under ac the forces will vary with time. One can conclude that in order to have a high probability of coalescence a voltage with a high rms value will be most efficient. From this point of view there is no optimum frequencya high dc field is the best provided it does not exceed the stability limit for the large drops. Coalescence is a bulk effect in an emulsion with a high electric field that is activated once drops get close. Potentially some drop resonance could be imagined but that would only apply to drops within a narrow size distribution. For low water cuts when drops are far apart in stagnant oil electrocoalescence will occur rarely as they do not get within close range of each other. Again shear flow may improve coalescence efficiency. Having concluded that dc fields are most advantageousit must immediately be commented that this is only the case for uncovered electrodes. Electrode covering introduces effects that totally change the phenomena. Again the time constant of the oil becomes important. As shown in Fig.3 the emulsion with a low conductivity is in series with the solid electrode barrier insulation with a high resistivity. Under dc the voltage distribution is governed by the resistivities. Because the resistivity of oil is much lower than of the insulator the oil will effectively short circuit the gap with the emulsion. All voltage will end up over the solid insulation and there will be no electric field in the emulsion to drive the coalescence. Now ac may become advantageous. If an ac voltage with a half period time that is short compared to the time constant of the emulsion or the oil is appliedthe voltage distribution is governed by the permittivity of the solid insulation and the oil emulsionwhich are of the same magnitude. This will give a more linear voltage distribution across the gapand a high field in the emulsion. Based on this analysis again there is no optimum frequencyone only has to apply an ac voltage above a certain frequency. When using a sine ac voltage some efficiency will be lost because of the time variation. In principle this can be avoided by using a square wave ac fieldhaving a high rms value with a frequency high enough to get a capacitive voltage distribution. We assume that the effect of eg pulsing dc is mainly the effect of a redistribution of the electric field after the relaxation occurring under dcand can see no advantage of this compared to the ac square wavewhich in principle also is a pulsed dc. In order for the pulsed dc to be as efficient as the square wave it has to be pulsed continuouslywhich basically results in an ac with or without an dc offset. Even if some understanding has been obtained there is much work to do to grasp the complexity of electrocoalescence in a real heavy crude oil. Work will continue for three years in a research carried out at SINTEF Energy Research with the Ugelstad lab at NTNU and CNRS (G2E Lab) in Grenoble. The project is funded through a contract with The Research Council of Norway. The partners are AibelAker-Kværner Process SystemsStatoilHydroBPShellPetrobras and Saudi Aramco.
At low electric fields electrostatic attractive forces may help to ‘glue’ particles to each otherand coalescence will occur when the interfacial film eventually breaks.
At higher electric fields an instability will occur at the larger of the two drops‘shooting’ the water of this drop into the smaller oneas shown in Fig.2b.
Lars LundgaardGunnar Berg and Petter E Røkke are with SINTEF Energy ResearchTrondheimNorway. www.sintef.no
"
