WO2017180167A1 - Oil recovery using microporous hydrophobic membrane contactors - Google Patents

Oil recovery using microporous hydrophobic membrane contactors Download PDF

Info

Publication number
WO2017180167A1
WO2017180167A1 PCT/US2016/029301 US2016029301W WO2017180167A1 WO 2017180167 A1 WO2017180167 A1 WO 2017180167A1 US 2016029301 W US2016029301 W US 2016029301W WO 2017180167 A1 WO2017180167 A1 WO 2017180167A1
Authority
WO
WIPO (PCT)
Prior art keywords
oil
water
mixture
viscosity
membrane
Prior art date
Application number
PCT/US2016/029301
Other languages
French (fr)
Inventor
Frank Seibert
Aurore MERCELAT
Kerry Kinney
Lynn Katz
Original Assignee
Board Of Regents, The University Of Texas System
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Board Of Regents, The University Of Texas System filed Critical Board Of Regents, The University Of Texas System
Publication of WO2017180167A1 publication Critical patent/WO2017180167A1/en

Links

Classifications

    • BPERFORMING OPERATIONS; TRANSPORTING
    • B01PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
    • B01DSEPARATION
    • B01D17/00Separation of liquids, not provided for elsewhere, e.g. by thermal diffusion
    • B01D17/02Separation of non-miscible liquids
    • B01D17/04Breaking emulsions
    • B01D17/045Breaking emulsions with coalescers
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B01PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
    • B01DSEPARATION
    • B01D17/00Separation of liquids, not provided for elsewhere, e.g. by thermal diffusion
    • B01D17/08Thickening liquid suspensions by filtration
    • B01D17/10Thickening liquid suspensions by filtration with stationary filtering elements
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B01PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
    • B01DSEPARATION
    • B01D63/00Apparatus in general for separation processes using semi-permeable membranes
    • B01D63/02Hollow fibre modules
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B01PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
    • B01DSEPARATION
    • B01D2315/00Details relating to the membrane module operation
    • B01D2315/22Membrane contactor
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B01PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
    • B01DSEPARATION
    • B01D2325/00Details relating to properties of membranes
    • B01D2325/38Hydrophobic membranes

Definitions

  • This invention is in the field of oil and water separation techniques.
  • This invention relates generally to methods and systems for efficiently separating water and insoluble oil from mixtures using hydrophobic porous membrane contactors.
  • Oil/water separations are useful in processes associated with food, biochemical and chemical production as well as in the oil and gas industry.
  • produced water treatment is driven by regulations controlling well disposal and environmental releases as well as by the financial interest in recovering oil contained in the wastewater.
  • New technologies are needed to treat the oil and gas wastewater in an affordable and efficient way.
  • Membrane technologies show great promise for oil removal in produced water as they are affordable, flexible, have a small footprint and can recover small oil droplets not captured by conventional technologies.
  • Described herein are methods and systems for recovering oil from an oil and water mixture.
  • Some embodiments make use of hydrophobic hollow fiber membrane systems that allow permeation of oil from outside the hollow fibers to an interior of the hollow fiber. Due to their hydrophobic nature, the hollow fibers may exclude water from permeating into the interior of the fibers.
  • Various process conditions are disclosed, in some embodiments, as impacting a rate of oil flux into the interior of the hollow fibers and control over these process conditions allows for more efficient separation of the mixture.
  • methods are provided, such as methods for recovering oil from an oil and water mixture.
  • methods of this aspect may comprise determining a ratio of oil to water in a mixture comprising oil and water, such as a mixture comprising water and oil that are at least partly immiscible; identifying a transmembrane pressure, a process viscosity, or both for use in separating the oil and the water from the mixture; contacting a shell side of a hollow fiber with the mixture, such as a hollow fiber that comprises a porous and hydrophobic polymer material, and where at least a portion of the oil in the mixture wets the hollow fiber and is transported through the hollow fiber to a tube side of the hollow fiber, and where contacting includes applying the transmembrane pressure to the shell side of the hollow fiber, or changing a viscosity of the oil to the process viscosity, or both to enhance transport of oil through the hollow fiber; and collecting oil from the tube side of the hollow fiber.
  • the transmembrane pressure is identified as a first transmembrane pressure when the ratio of oil to water is less than a first threshold ratio.
  • the transmembrane pressure is identified as a second transmembrane pressure when the ratio of oil to water is greater than the first threshold ratio.
  • the first transmembrane pressure is less than the second transmembrane pressure.
  • the process viscosity is identified as a first viscosity when the ratio of oil to water is less than a second threshold ratio.
  • the process viscosity is identified as a second viscosity when the ratio of oil to water is greater than the second threshold ratio.
  • the first viscosity is greater than the second viscosity.
  • different operating regimes can be employed to more effectively separate the oil from the mixture. For example, in some embodiments, when a concentration of oil in the mixture is low, a higher process viscosity may more effectively result in removal of oil from the mixture. In some other embodiments, when a concentration of oil in the mixture is high, a lower process viscosity may more effectively result in removal of oil from the mixture.
  • identifying the transmembrane pressure, the process viscosity, or both includes determining an effective area ratio for the hollow fiber.
  • determining the effective area ratio includes computing where
  • a e /a t corresponds to the effective area ratio
  • a a constant
  • b a constant
  • C oil is the ratio of oil to water in the mixture.
  • a a' . TMP -1 6 . ⁇ 1 .0. Q 0.3
  • a' is a constant
  • TMP is the transmembrane pressure in Pa
  • is the process viscosity in Pa s
  • Q is a shell side flow rate of the mixture in m 3 /s
  • C oil is expressed as a volume fraction.
  • a' may be about 9.27 x 10 12 .
  • b b' . TMP -2.1. ⁇ 1 .1.
  • the transmembrane pressure is selected based on a concentration of oil in the mixture in order to allow for efficient separation of oil from the mixture.
  • the first transmembrane pressure is optionally selected from the range of 1 pounds per square inch to 40 pounds per square inch.
  • the second transmembrane pressure is selected from the range of 20 pounds per square inch to 100 pounds per square inch. In this way, different transmembrane pressures may be used depending on the concentration of oil in the mixture.
  • embodiments may employ a transmembrane pressure less than 40 pounds per square inch, such as a transmembrane pressure selected from the range of 1 psi to 40 psi, selected from the range of 5 psi to 40 psi, selected from the range of 10 psi to 40 psi, selected from the range of 1 psi to 30 psi, selected from the range of 5 psi to 30 psi, selected from the range of 10 psi to 30 psi, selected from the range of 1 psi to 35 psi, selected from the range of 1 psi to 25 psi, or selected from the range of 1 psi to 15 psi.
  • the listed pressure ranges are merely examples and that other transmembrane pressure selected from the range of 1 psi to 40 psi, selected from the range of 5 psi to 40 psi, selected from the range of 10 psi to 40 psi
  • embodiments may employ a transmembrane pressure greater than 20 pounds per square inch, such as a transmembrane pressure selected from the range of 20 psi to 100 psi, selected from the range of 20 psi to 80 psi, selected from the range of 20 psi to 60 psi, selected from the range of 30 psi to 100 psi, selected from the range of 30 psi to 80 psi, selected from the range of 30 psi to 60 psi, selected from the range of 40 psi to 100 psi, selected from the range of 40 psi to 80 psi, or selected from the range of 40 psi to 60 psi.
  • a transmembrane pressure greater than 20 pounds per square inch, such as a transmembrane pressure selected from the range of 20 psi to 100 psi, selected from the range of 20 psi to 80 psi, selected from the range of 20 p
  • the process viscosity is selected based on a concentration of oil in the mixture in order to allow for efficient separation of oil from the mixture.
  • the first process viscosity is optionally selected from the range of 0.2 cP to 100 cP.
  • the second viscosity is selected from the range of 0.2 cP to 100 cP. In this way, different process viscosities may be used depending on the concentration of oil in the mixture.
  • embodiments may adjust the oil viscosity to a value of between 0.2 cP to 100 cP, such as a viscosity selected from the range of 1 cP to 100 cP, selected from the range of 10 cP to 100 cP, selected from the range of 20 cP to 100 cP, selected from the range of 30 cP to 100 cP, selected from the range of 40 cP to 100 cP, selected from the range of 50 cP to 100 cP, selected from the range of 60 cP to 100 cP, selected from the range of 70 cP to 100 cP, or selected from the range of 80 cP to 100 cP.
  • the listed viscosity ranges are merely examples and that other viscosity values may be useful with oil concentrations less than 20%
  • embodiments may adjust the oil viscosity to a value of between 0.2 cP to 100 cP, such as a viscosity selected from the range of 0.2 psi to 90 cP, selected from the range of 0.2 psi to 80 cP, selected from the range of 0.2 cP to 70 cP, selected from the range of 0.2 cP to 60 cP, selected from the range of 0.2 cP to 50 cP, selected from the range of 0.2 cP to 40 psi, selected from the range of 0.2 cP to 35 cP, selected from the range of 0.2 cP to 30 cP, or selected from the range of 0.2 psi to 25 cP.
  • a viscosity selected from the range of 0.2 psi to 90 cP selected from the range of 0.2 psi to 80 cP, selected from the range of 0.2 cP to 70 cP, selected from the range of 0.2 cP to 60
  • the listed viscosity ranges are merely examples and that other viscosity values may be useful with oil concentrations greater than 20%.
  • selection of the process viscosity may vary as a function of length of the hollow fiber, as the oil concentration in the mixture present at the shell side of the hollow fiber may be reduced by passage of oil into the tube side of the hollow fiber.
  • the first viscosity is used on a first portion of the hollow fiber and a second viscosity is used on a second portion of the hollow fiber.
  • a single hollow fiber (or hollow fiber membrane comprising a plurality of hollow fibers) may have different process conditions allowing more efficient oil flux into the tube side of the hollow fiber than a single process condition.
  • a process viscosity close to an entrance port of a hollow fiber membrane contactor may be lower than the process viscosity close to an exit port of a hollow fiber membrane contactor.
  • temperature may be used to control the viscosity of oil in the mixture.
  • changing the viscosity of the oil may optionally include heating the mixture, heating the hollow fiber, or both. In some embodiments, heating the mixture, the hollow fiber, or both may result in a decrease of the oil viscosity.
  • changing the viscosity of the oil may optionally include cooling the mixture, cooling the hollow fiber, or both.
  • cooling the mixture, the hollow fiber, or both may result in an increase of the oil viscosity. It will be appreciated that a single hollow fiber that has different process conditions allowing more efficient oil flux into the tube side of the hollow fiber than a single process condition may utilize different temperature regimes for the different portions of the hollow fiber.
  • the oil and the water may be completely immiscible.
  • at least a portion of the oil may optionally be present in the mixture as microdroplets of oil suspended in water.
  • at least a portion of the microdroplets may have a cross-sectional dimension less than 1 ⁇ , less than 5 ⁇ , less than 10 ⁇ , less than 15 ⁇ , less than 20 ⁇ , or less than 50 ⁇ .
  • At least a portion of the microdroplets have a cross sectional dimension greater than 0.1 ⁇ , selected from the range of 0.1 ⁇ to 10 ⁇ , selected from the range of 0.1 ⁇ to 20 ⁇ , or selected from the range of 0.1 ⁇ to 30 ⁇ .
  • the ratio of oil to water in the mixture may correspond to a concentration of oil in water in the mixture.
  • the ratio of oil to water in the mixture may correspond to a concentration of water in oil in the mixture.
  • the ratio of oil to water in the mixture may correspond to a mass fraction of oil in the mixture.
  • the ratio of oil to water in the mixture may correspond to a mass fraction of water in the mixture.
  • the ratio of oil to water in the mixture may correspond to a mole fraction of oil in the mixture.
  • the ratio of oil to water in the mixture may correspond to a mole fraction of water in the mixture.
  • the ratio of oil to water in the mixture may correspond to a volume fraction of oil in the mixture.
  • the ratio of oil to water in the mixture may correspond to a volume fraction of water in the mixture.
  • determining the ratio of oil to water in the mixture includes determining a concentration of oil in water in the mixture, determining a concentration of water in oil in the mixture, determining a mass fraction of oil in the mixture, determining a mass fraction of water in the mixture, determining a mole fraction of oil in the mixture, determining a mole fraction of water in the mixture, determining a volume fraction of oil in the mixture, determining a volume fraction of water in the mixture, and any combination of these.
  • the first threshold ratio is selected from the range of about 1 part per million to about 20%.
  • useful first threshold ratios include, but are not limited to about 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 11%, 12%, 13%, 14%, 15%, 16%, 17%, 18%, 19% or 20%).
  • Other first threshold ratios are useful with some embodiments.
  • the second threshold ratio is selected from the range of about 1 part per million to about 2000 parts per million (ppm).
  • useful second threshold ratios include, but are not limited to about 1 ppm, 5 ppm, 10 ppm, 50 ppm, 100 ppm, 500 ppm, 1000 ppm, 1500 ppm, or 2000 ppm. Other second threshold ratios are useful with some embodiments.
  • the oil comprises one or more oils selected from the group consisting of biological oils, petroleum oils, food oils, paraffins, nonpolar hydrocarbons, saturated
  • the hollow fiber optionally comprises a porous material, such as a hydrophobic material or a material having a surface rendered hydrophobic.
  • a porous material such as a hydrophobic material or a material having a surface rendered hydrophobic.
  • oil may be preferentially attracted to the hydrophobic material of the hollow fiber while water may be preferentially repelled from the hydrophobic material of the hollow fiber.
  • the hollow fiber comprises a material selected from the group consisting of polypropylene, polyethylene, polytetrafluoroethylene, polyethylene terephthalate, polyvinylidene fluoride, polymethylpentene, and any combination of these.
  • the hollow fibers useful with the methods and systems described herein may be porous hollow fibers, such as mesoporous hollow fibers, microporous hollow fibers, or nanoporous hollow fibers.
  • Example hollow fibers include, but are not limited to, those including a plurality of pores, such as pores having cross-sectional dimensions selected from the range of 0.001 ⁇ to 1 ⁇ . It will be appreciated that pores in a single fiber may have dimensions distributed throughout this range or within subranges within this range.
  • the hollow fiber is present in a hollow fiber membrane.
  • contacting may include contacting a shell side of the hollow fiber membrane with the mixture.
  • collecting includes collecting oil from tube sides of the hollow fiber membrane.
  • the hollow fiber membrane is present in a membrane contactor.
  • the membrane contactor includes a shell side inlet, a shell side outlet, and a tube side outlet.
  • contacting the shell side of the hollow fiber includes flowing the mixture into the shell side inlet.
  • collecting includes collecting oil from the tube side outlet.
  • a system comprises: a pump for pumping a mixture comprising oil and water, such as where the oil and water are at least partially immiscible; a temperature control element for heating or cooling the mixture, such as a temperature control element that is positioned in thermal communication with the mixture; a hydrophobic porous membrane contactor, such as a hydrophobic membrane contactor that is positioned in fluid communication with the pump; an oil collection system positioned in fluid communication with the tube side outlet of the hydrophobic porous membrane contactor; and a controller that controls a flow rate of the mixture using the pump, and a temperature of the mixture using the temperature control element.
  • the hydrophobic membrane contactor comprises a hollow fiber membrane having a shell side in fluid communication with a shell side inlet of the hydrophobic porous membrane contactor and a shell side outlet of the hydrophobic porous membrane contactor.
  • the hollow fiber membrane has a tube side in fluid communication with a tube side outlet of the hydrophobic porous membrane contactor.
  • the hollow fiber membrane comprises a plurality of porous hydrophobic polymer hollow fibers.
  • the controller further controls a transmembrane pressure of the mixture applied to the shell side of the hollow fiber membrane, or a process viscosity of oil in the mixture, or both the transmembrane pressure and the process viscosity.
  • the controller may set the transmembrane pressure to a first transmembrane pressure when a ratio of oil to water in the mixture is less than a first threshold ratio.
  • the controller sets the transmembrane pressure to a second transmembrane pressure when the ratio of oil to water in the mixture is greater than the first threshold ratio.
  • the first transmembrane pressure is less than the second transmembrane pressure.
  • the controller sets the process viscosity to a first viscosity when a ratio of oil to water in the mixture is less than a first threshold ratio.
  • the controller sets the process viscosity to a second viscosity when the ratio of oil to water in the mixture is greater than the first threshold ratio.
  • the first viscosity is greater than the second viscosity.
  • physical parameter measurements and their control may also be useful with the methods and systems described herein. For example, it may be useful to monitor and/or control physical parameters such as flow rate, viscosity, pressure, temperature, concentration.
  • a system of this aspect may further comprise a pressure transmitter for measuring a pressure of the mixture, wherein the pressure transmitter is positioned in fluid communication with the mixture; and a feedback circuit between the pressure transmitter and the controller for providing the pressure of the mixture to the controller for use in controlling the transmembrane pressure.
  • multiple pressure transmitters may be used for measuring the pressure at various points within the system.
  • the transmembrane pressure may be estimated as an average pressure of a shell side inlet pressure and a shell side inlet pressure, such as when the tube side outlet pressure is at atmospheric or ambient pressure.
  • a system of this aspect may further comprise a temperature sensor for measuring a temperature of the mixture, such as a temperature sensor that is positioned in thermal communication with the mixture; and a feedback circuit between the temperature sensor and the controller for providing the temperature of the mixture to the controller for use in controlling the temperature of the mixture, or the process viscosity ,or both.
  • a temperature sensor for measuring a temperature of the mixture, such as a temperature sensor that is positioned in thermal communication with the mixture
  • a feedback circuit between the temperature sensor and the controller for providing the temperature of the mixture to the controller for use in controlling the temperature of the mixture, or the process viscosity ,or both.
  • a system of this aspect may further comprise a concentration sensor for determining a ratio of oil to water in the mixture, such as a concentration sensor that is positioned in communication with the mixture, e.g., fluid or optical communication; and a feedback circuit between the concentration sensor and the controller for providing the ratio of oil to water in the mixture to the controller for use in controlling the transmembrane pressure, or the process viscosity, or both.
  • systems of this aspect may comprise combinations concentration sensors, temperature sensors, and/or pressure transmitters and feedback circuits for providing measured values to a controller for use in controlling the separation process.
  • the hydrophobic porous membrane contactor is a first hydrophobic membrane contactor and the system further comprises a second hydrophobic porous membrane contactor in fluid communication with the shell side outlet of the first hydrophobic porous membrane contactor.
  • the transmembrane pressure is a first membrane contactor transmembrane pressure
  • the controller sets the first membrane contactor transmembrane pressure to the second transmembrane pressure
  • the controller sets a second membrane contactor
  • a dual stage separation system may be used to reduce the oil concentration in the mixture in two parts - a first reduction from a higher concentration by the first membrane contactor and a second reduction from a lower concentration by the second membrane contactor.
  • a first concentration of oil in the first hydrophobic porous membrane contactor is greater than a second concentration of oil in the second hydrophobic porous membrane contactor.
  • a dual stage separation system may allow for using different processing parameters at each membrane contactor.
  • the process viscosity is optionally a first membrane contactor process viscosity
  • the controller sets the first process viscosity to the second viscosity
  • the controller sets a second process viscosity of oil in the second
  • hydrophobic porous membrane contactor to the first viscosity
  • the temperature control element is in thermal communication with the hydrophobic porous membrane contactor.
  • the temperature control element is a heating element, or a cooling element, or both.
  • the temperature control element is arranged to control a first temperature of a first portion of the hydrophobic porous membrane contactor and to control a second temperature of a second portion of the hydrophobic porous membrane contactor. In this way, a single membrane contactor may be used with different process parameters across the membrane contactor to allow for efficient separation of oil at higher and lower concentrations, for example.
  • FIG. 1 A provides a schematic illustration of a hollow fiber.
  • FIG. IB provides a schematic illustration of a hollow fiber membrane.
  • FIG. 2 provides a schematic illustration of a membrane contactor, in accordance with some embodiments.
  • FIG. 3 provides a schematic cross-sectional illustration of a membrane contactor, in accordance with some embodiments.
  • FIG. 4A and FIG. 4B provide details of methods for recovering oil from an oil and water mixture, in accordance with some embodiments.
  • FIG. 5 provides data showing viscosity as a function of temperature for three different oils.
  • FIG. 6 provides a schematic illustration of a water/oil separation system, in accordance with some embodiments.
  • FIG. 7 provides data showing the effect of viscosity on oil permeation.
  • FIG. 8 provides data showing the effect of transmembrane pressure on oil flux corrected for viscosity.
  • FIG. 9 provides data showing the effect of influent flow rate on oil flux corrected for viscosity.
  • FIG. 10 provides data showing a comparison of module surface areas for oil flux as a function of transmembrane pressure.
  • FIG. 11 provides data showing a comparison of module surface areas for oil flux as a function of viscosity.
  • FIG. 12 provides data showing a comparison of module sizes for oil flux as a function of transmembrane pressure.
  • FIG. 13 provides data showing a comparison of fiber type for oil flux as a function of transmembrane pressure.
  • FIG. 14 provides data showing a parity plot comparing experimentally measured oil flux and predicted oil flux.
  • FIG. 15 provides a scanning electron micrograph image of a porous fiber and data showing a pore size distribution for a porous fiber.
  • FIG. 16 provides data showing droplet size distribution for droplets of oil in an oil/water mixture.
  • FIG. 17 provides a schematic illustration of a water/oil separation system, in accordance with some embodiments.
  • FIG. 18 provides a schematic illustration of a water/oil separation system, in accordance with some embodiments.
  • FIG. 19A and FIG. 19B provide data showing the effect of influent oil concentration on oil flux across a membrane.
  • FIG. 20A, FIG. 20B, and FIG. 20C provide data showing oil recovery as a function of time.
  • FIG. 21 A and FIG. 21B provide data showing the effect of transmembrane pressure on oil flux.
  • FIG. 22 provides data showing the effect of influent flow rate on oil recovery.
  • FIG. 23 A, FIG. 23B, and FIG. 23C provide data showing the effect of viscosity on oil flux.
  • FIG. 24 provides data showing the effect of viscosity on oil recovery as a function of time.
  • FIG. 25 A provides data showing the effect of surface area on oil flux.
  • FIG. 25B provides data showing the effect of surface area on oil recovery.
  • FIG. 26 provides schematic illustrations of oil droplet coalescence between fibers for two transmembrane pressures.
  • FIG. 27 provides a schematic representation of a hydrophilic water/oil separation system, in accordance with some embodiments.
  • FIG. 28 provides data showing an experimental flux ratio as a function of oil
  • FIG. 29 provides data showing the effect of influent flow rate on an oil flux ratio.
  • FIG. 30 provides data showing the effect of transmembrane pressure on an oil flux ratio for two oil concentrations.
  • FIG. 31 provides data showing the effect of viscosity on an oil flux ratio for different oil concentrations.
  • FIG. 32 provides data showing model residuals as a function of effective area ratio.
  • FIG. 33 provides data showing experimentally flux ratios as a function of estimated effective area ratios.
  • FIG. 34A, FIG. 34B, FIG. 34C, FIG. 34D, FIG. 34E, and FIG. 34F provide data showing examples of a model fitting experimental measurements.
  • FIG. 35 provides data comparing experimentally measured flux and predicted flux determined using a model.
  • FIG. 36A, FIG. 36B, FIG. 36C, FIG. 36D, FIG. 36E, FIG. 36F, FIG. 36G, FIG. 36H, FIG. 361, and FIG. 36 J provide data showing model predictions for oil permeation.
  • “Miscible” refers to a tendency of two or more liquids to mix into a single phase to form a homogeneous solution. “Immiscible” refers to a tendency of at least two liquids, when combined, to form at least two phases, such as first phase that includes a first of the at least two liquids and a second phase that includes a second of the at least two liquids. Oil and water are commonly identified as immiscible, such that mixing oil and water will result in two distinct phases when completely settled. It will be appreciated, however, that some oils are partly miscible in water and may dissolve, at least in part, in water. Similarly, water may be partly miscible in some oils and may dissolve, at least in part, in some oils. The “miscible” may be used
  • Viscosity refers to a property of a fluid indicative of the resistance of the fluid to flow or to deform by shear or tensile stress. In embodiments, a fluid with a higher viscosity may be observed to flow faster than a fluid with a lower viscosity.
  • a “process viscosity” may refers to a target viscosity used in a particular process, such as a separation process. In embodiments, a process viscosity corresponds to a set point for a system in which viscosity can be controlled, such as by controlling the temperature of a liquid.
  • FIG. 1 illustrates an example of a tube side and a shell side of a hollow fiber.
  • the shell side refers to the region outside of the hollow fibers included in the hollow fiber membrane
  • the tube side refers, collectively or individually, to the regions within the individual hollow fibers.
  • the tube sides of a plurality of hollow fibers in a hollow fiber membrane are arranged in fluid
  • a hollow fiber membrane may be arranged in a membrane contactor, which may include a shell side inlet, a shell side outlet, a tube side inlet and a tube side outlet.
  • Transmembrane pressure refers to a pressure differential between a tube side and a shell side of a hollow fiber membrane.
  • the pressure differential is expressed with the shell side of the hollow fiber membrane being a higher pressure than the tube side of the hollow fiber membrane.
  • Washing refers to the ability of a liquid to maintain contact with a solid surface.
  • a surface is wetted by a liquid when the liquid spreads across the surface and penetrates into at least a fraction of recessed regions, such as pores within the surface.
  • a hydrophilic porous surface may be wetted by water, for example.
  • a hydrophobic porous surface may be wetted by an oil, for example.
  • Polymer refers to a molecule composed of repeating structural units, referred to as monomers, connected by covalent chemical bonds or the polymerization product of one or more monomers. Polymers may be characterized by a high molecular weight, such as a molecular weight greater than 100 atomic mass units (amu), greater than 500 amu, greater than 1000 amu, greater than 10000 amu or greater than 100000 amu. In some embodiments, a polymer may be characterized by a molecular weight provided in g/mol or kg/mol, such as a molecular weight of about 200 kg/mol or about 80 kg/mol. The term polymer includes homopolymers, or polymers consisting essentially of a single repeating monomer subunit.
  • polymer also includes copolymers, which are formed when two or more different types of monomers are linked in the same polymer.
  • Copolymers may comprise two or more monomer subunits, and may include random, block, alternating, segmented, grafted, tapered and other copolymers.
  • Useful polymers include organic polymers that may be in amorphous, semi-amorphous, crystalline or partially crystalline states.
  • Example polymers include, but are not limited to, polypropylene, polyethylene, polytetrafluoroethylene, polyethylene terephthalate, polyvinylidene fluoride, polymethylpentene.
  • Porous refers to a characteristic of a material that includes recessed regions that may penetrate partly of fully through a body of the material. Pores in a porous material may be characterized by a cross-sectional dimension.
  • polymeric materials may be porous, and may include a plurality of pores extending through at least portions of the polymer structure.
  • Hydrophobic refers to a preference for a material to be attracted to a non-polar substance as compared to water and other polar substances.
  • This application generally relates to techniques and systems for separating oil and water in an oil and water mixture and for enhancing the separation process, such as by identifying process parameters which allow for more efficient separation of the oil and water than other process parameters.
  • the process parameters used for more efficient separation in low oil concentrations are different from those used for more efficient separation in high oil concentrations.
  • the present disclosure describes, in embodiments, operation of a hydrophobic porous membrane separator in two different operational regimes depending on the concentration of oil in the mixture being separated.
  • the operational parameters at high oil concentrations such as greater than 2%, greater than 10%, or greater than 20%, may be different from those used at low oil concentrations, such as less than 2%, less than 1%, or less than 0.5%.
  • the inventors have developed a model that allows, in various embodiments, determination of efficient operational parameters at any oil concentration.
  • FIG. 1 A provides a schematic illustration of a hollow fiber 100 illustrating various aspects of the fiber.
  • hollow fiber 100 includes a body 110 that separates an internal volume or tube side 120 of the fiber 100 from the external environment or shell side 130 of the fiber 100.
  • the body 110 may comprise a porous polymer material that is hydrophobic. Droplets of oil on the shell side 130, for example, that come into contact with the body 110 may wet the surface and fill into pores of the polymer material. As additional oil droplets contact the wetted and filled shell side 130 surface, the oil may be transported into the tube side of the hollow fiber allowing collection of the oil.
  • FIG. IB provides a schematic illustration of a hollow fiber membrane 150.
  • the hollow fiber membrane 150 comprises a plurality of hollow fibers 160 arranged as a web of interspersed fibers.
  • FIG. IB merely represents an illustration and that a higher or lower density of fibers will be present in a fiber membrane.
  • a majority of the fibers in a hollow fiber membrane have entrances to the tube side of the fibers that are proximal to one another as well as exits to the tube side of the fibers that are proximal to one another.
  • FIG. IB illustrates all fibers as extending from an upper side to a lower side of the membrane, other configurations are possible.
  • FIG. IB illustrates all fibers as extending from an upper side to a lower side of the membrane, other configurations are possible.
  • FIG. 2 provides a schematic illustration of a membrane contactor 200 useful with various embodiments described herein.
  • FIG. 2 illustrates a cut-away of regions of the membrane contactor 200 to illustrate internal components.
  • a collection tube 210 and distribution tube 220 are arranged in the center of membrane contactor 200.
  • a central baffle 225 may allow for varied flow within the shell side of membrane contactor 200.
  • a hollow fiber membrane may be wrapped around the distribution/collection tube. Individual hollow fibers 230 of the hollow fiber membrane are illustrated. Flow through the membrane contactor 200 may be achieved through four ports.
  • a shell side inlet 240 is illustrated at the bottom of FIG. 2, which represents an inlet for a mixture of oil and water into the membrane contactor 200 to allow flow on the shell side of the hollow fiber membrane.
  • a shell side outlet 250 is illustrated at the top of FIG. 2, which represents an outlet for a treated water that has had oil removed.
  • a tube side inlet 280 is illustrated near the bottom of FIG. 2 and is shown as plugged for use in various embodiments described herein. It will be appreciated that tube side inlet 280 may optionally be used in an unplugged configuration as well.
  • a tube side outlet 290 is illustrated near the top of FIG. 2 and represent an outlet for collection of oil that passes from the shell side of the hollow fiber membrane to the tube sides of the hollow fiber membrane.
  • FIG. 3 provides a schematic cross-sectional illustration of a membrane contactor 300 useful with various embodiments described herein. It will be appreciated that FIG. 3 is not drawn to scale but has portions of membrane contactor shown in an expanded or magnified and simplified view to better illustrate various aspects.
  • membrane contactor includes a shell side inlet port 310, which represents an inlet for a water and oil mixture, a shell side outlet port 320, which represents an outlet for treated water, a tube side inlet port 330 and a tube side outlet port 340. It will be appreciated that although ports 310, 320, 330 and 340 are identified as inlet and outlet ports, these notations may optionally be reversed.
  • the shell side 350 of membrane contactor is illustrated as including a varying concentration of oil and water (higher oil concentration near the bottom of FIG. 3 and lower oil concentration near the top of FIG. 3), which represents the removal of oil from the shell side 350 of membrane contactor 300 and passage to the tube side 360 of membrane contactor 300.
  • Porous hollow fibers 360 are illustrated as extending between the top and bottom of membrane contactor 300, with ends of porous hollow fibers 360 connected in a configuration that shows all tube side inlets connected in fluid communication with tube side inlet port 330 and all tube side outlets connected in fluid communication with tube side outlet port 340.
  • tube side inlet port 330 or tube side outlet port may be in a plugged configuration.
  • FIG. 4A provides an overview of an example method 400, in accordance with some embodiments, for recovering oil from an oil and water mixture.
  • a ratio of oil to water in the mixture is determined.
  • a transmembrane pressure and/or viscosity are determined for use in the separation process.
  • the mixture is contacted to the hollow fiber shell side, where oil from the mixture is allowed to pass to the tube side of the hollow fiber.
  • the oil is collected from the hollow fiber tube side.
  • FIG. 4B provides an overview 450 of details of identifying the transmembrane pressure or viscosity for use in the separation.
  • the ratio of oil to water in the mixture 455 is compared with a threshold ratio. If the ratio of oil to water in the mixture is greater than the threshold ratio, process flow continues to block 470. If the ratio of oil to water in the mixture is less than the threshold ratio, process flow continues to block 480.
  • a first threshold ratio If the ratio of oil to water in the mixture is greater than the threshold ratio, process flow continues to block 470. If the ratio of oil to water in the mixture is less than the threshold ratio, process flow continues to block 480.
  • a first threshold ratio If the ratio of oil to water in the mixture is greater than the threshold ratio, process flow continues to block 470. If the ratio of oil to water in the mixture is less than the threshold ratio, process flow continues to block 480.
  • a first threshold ratio If the ratio of oil to water in the mixture is greater than the threshold ratio, process flow continues to block 470. If the ratio of
  • transmembrane pressure or first viscosity is identified for use in the separation process.
  • a second transmembrane pressure or second viscosity is identified for use in the separation process.
  • the ratio of oil to water may be used with a model to directly identify process parameters, such as transmembrane pressure, viscosity, flow rate and/or temperature to use in enhancing the transport of oil through the hollow fiber.
  • EXAMPLE 1 PURE OIL CHARACTERIZATION OF A HYDROPHOBIC MICROPOROUS HOLLOW FIBER MEMBRANE CONTACTOR FOR A NEW APPLICATION IN OIL-WATER
  • Insoluble oil and water mixtures are generated in the oil and gas, biofuel and food industries as well as in natural environments contaminated by oil spills. Technologies are needed to cost effectively and reliably separate the oil from the water and obtain high-purity phases for downstream use or reuse. Separation of oil and water is particularly critical in many oil and gas industrial applications where wastewaters such as produced water are generated onshore and offshore and oil concentrations can range from 10 ppm to 200,000 ppm depending on the source.
  • the treatment of produced water allows the water to be reused in the oil-extraction system, revalorized for other uses such as irrigation or released into the environment. Oil recovery contributes to the economic feasibility of many treatment processes and in some instances, such as biodiesel production, oil purification is one objective of the process.
  • hydrophilic membranes While hydrophilic membranes have been used for produced water treatment, a major drawback of the technology is the buildup of a viscous fouling layer at the membrane surface, which progressively decreases the efficiency and increases the energy -input of the overall process. Hydrophilic membranes are typically used to permeate water while oil is retained. An increase in transmembrane pressure and influent flow rate increase water flux across hydrophilic membranes but also potentially lead to lower oil rejection. Increasing temperature may increase water flux by reducing the viscosity of the fouling layer on the membrane surface. [0096] Some materials may be pretreated to allow water permeation and oil retention.
  • Hydrophobic membranes may be used to break oil-water emulsions by coalescence. Some direct oil-water separations with hydrophobic membranes allow oil permeation and water rejection. Oil flux may increase with increasing transmembrane pressure for hydrophobic membranes. The increase of cross flow velocity may decrease oil flux across the membrane or not affect the process, depending on feed oil concentration. Other hydrophobic systems for oil and water separations may use membranes made from experimental materials, or of a small size, or where the oil is adsorbed and not recoverable. Additional research may allow understanding the fundamental mechanisms governing the separation of oil and water with larger scale hydrophobic membrane systems.
  • microporous membrane contactor performance improved over time to reach high oil recovery (above 95%) with injected oil over a two-week test run. While studies with actual and simulated feedstocks are useful, the fundamental mechanisms controlling the oil recovery are complex.
  • Equation 1 Hagen-Poiseuille law for streamline flow through channels:
  • the Kozeny-Carman equation is obtained for the case of a packed bed of spherical solids, which may be applied from membrane flux predictions.
  • the Blake-Kozeny equation may be applied to hollow fiber bundles to refine the correlation of A with porosity and linear correlations may exist between A and ⁇ . Such correlations allow refining of the application of the Blake-Kozeny equation to media with various porosities.
  • This Example characterizes the hydrophobic hollow fiber membrane contactor for oil and water separation and evaluates the potential for oil permeation without hydrocarbon fouling of the membrane surface. Characterization of the system under pure oil conditions may serve as a baseline for determining the maximum oil permeation attainable with the membrane contactor under various operating conditions. Development of a model that describes membrane behavior under various operating conditions and captures the impact of poorly quantified parameters such as porosity is also useful. Additional work may be useful to understand the fundamental mechanisms of oil/water separation with such systems. [0102] Materials and Methods. Membranes. The membranes used in this Example are manufactured by Liqui-Cel® (USA) and commercially available under the name Liqui-Cel® Extra Flow.
  • the modules are composed of hydrophobic microporous hollow-fibers with a central baffle to eliminate shell side bypassing and evenly distribute fluid perpendicularly to the fibers.
  • Modules with different intrinsic characteristics were investigated and are detailed in Table 1.
  • Membranes A and B have the same module sizes, but the hollow fibers are spaced more widely for Module B to obtain a surface area half of Membrane A.
  • Fiber types X50 and X40 differ in porosity and wall thickness.
  • Membrane D has a different module size and hence presents longer fibers with larger surface area. [0103] All of these membranes have similar designs that include a polypropylene hollow-fiber flat sheet rolled into a cylindrical casing allowing pseudo cross-flow filtration.
  • IsoparTM oils manufactured from ExxonMobil (USA) were used to simulate hydrocarbons of produced water.
  • the IsoparTM oils are controlled paraffinic mixtures and exist in different grades with varying viscosities as detailed in Table 2.
  • Table 2 Viscosities at 25 °C of IsoparTM grades used for the study.
  • FIG. 6 shows the schematic of the membrane system operated with pure oil. The feed was distributed through the system with a high shear pump (MTH pump, USA), and entered the membrane from the shell-side lower port. The retentate was sent back to the feed tank, while the permeate was collected in a graduated glass cylinder for quantification.
  • MTH pump high shear pump
  • the needle valve located downstream of the membrane module was used in coordination with the pump variable speed drive (Emerson, USA) to control transmembrane pressure and influent flow rate.
  • the transmembrane pressure was calculated from the reading of two pressure gauges (Solmony, USA). No pressure gauge was used on the membrane tube side since it was left open to the atmosphere and oil exited the system by gravity.
  • the influent flow rate was measured with a MicroMotion flow meter (Emerson, USA). Temperatures were obtained using a Rosemount RTD sensor (Emerson, USA).
  • FIG. 7 shows the effect of viscosity on oil permeation under different experimental conditions.
  • the data demonstrate that even within triplicates, where the same IsoparTM grade was used, viscosity varied because of ambient temperature variation.
  • the results suggest that oil flux is linearly related to the inverse of viscosity with coefficients of correlation higher than 0.99 for all three sets of conditions.
  • Membrane B utilizes a similar design (casing and fiber length) but contains half the surface area of Membrane A. In Membrane B, the fibers are twice as far apart from each other as in Membrane A. Experiments were conducted on both membranes to observe the effect of surface area on oil permeation. The results, presented in FIG. 10, show that the effects of transmembrane pressure and influent flow rate on Isopar M oil flux are consistent. Differences between the fluxes for the two membranes do not exceed 6.5% which falls within the standard deviations of the data sets.
  • the Isopar M pure oil flux may be proportional to the membrane surface area.
  • the relationship between oil flux and viscosity as a function of membrane surface area was more variable (FIG. 11).
  • the differences between the slopes of linear regression for surface areas of 0.7 m 2 and 1.4 m 2 were as high as 12.8%. Since results observed in FIG. 7 suggest a strong linear correlation between oil flux and viscosity at a given surface area, and data in FIG. 10 show linear dependence between oil flux and transmembrane pressure, such variability may be attributed to experimental error.
  • water at ppm levels was added to the pure oil feed after identification of static charging possibly increasing the variability of the results and decreasing oil flux as observed in FIG. 11.
  • Module size The membrane modules used in this study are commercially available for a range of module sizes. Pure oil experiments were conducted on both 2.5 inch-diameter x 8 inch- length and 4 inch-diameter x 28 inch-length modules. The effects of transmembrane pressure and influent flow rate on oil flux were studied for both module sizes and the results are shown in FIG. 12. The oil flux obtained for the larger module was consistently lower than those for the smaller module regardless of the experimental conditions.
  • the difference in slopes obtained from linear regressions of oil flux versus TMP for the two module sizes is 29.0%>. That variability is higher than any observed error in these experiments suggesting that experimental error is not responsible for the differences observed.
  • a possible explanation for such findings is module geometry.
  • the larger module not only has a higher surface area (20 m 2 ) but also contains longer fibers compared to the smaller module.
  • the module was placed vertically during experimentation and the permeated oil overflowed through the top port of the tube side to avoid short-circuiting. Therefore, the pressure on the tube side at the bottom of the larger module is expected to be higher than for shorter fibers due to static pressure.
  • Fiber Type The membranes used are manufactured with two types of fibers identified as X40 and X50.
  • the X50 fiber has a porosity of approximatively 40% and a wall thickness of 80 ⁇ , while the X40 has a lower porosity of approximatively 25% and a thicker wall of 100 ⁇ .
  • FIG. 13 shows the comparative results between X50 and X40 fibers for various transmembrane pressures and influent flows. The oil flux measured was lower for the X40 than the X50 membranes for the same set of operating conditions. The difference between the X40 and X50 fiber performance can be compared to predictive models accounting for differences in porosity and wall thickness.
  • Model The results presented in FIGs. 8-13 demonstrate the linear correlation between transmembrane pressure, inverse viscosity, and oil flux as represented in Equation 3. Influent flow rate was also confirmed to have no significant impact on the process for pure oil conditions.
  • Equation 3 Model describing oil permeation for X50 2.5 inch modules:
  • Table 4 Permeability constants for X50 and X40 fibers.
  • Equation 4 Equivalent mean particle diameter:
  • Table 5 presents the permeability constants for each of these models with adjusted A values for both X50 and X40 fibers.
  • the Hagen Poiseuille model largely overestimates the experimentally determined constants. The model is usually accurate for describing flux through membranes with parallel pores. However, membrane pores often present more complex pore geometry.
  • Table 5 Permeability constants for X50 and X40 fibers determined with the various models.
  • Equation 5 The most accurate model for the X50 fiber permeability was the Blake- Kozeny equation with the original Ergun constant of 150 as shown in Equation 5 below. Equation 5:
  • the X40 fiber permeability could not be properly predicted by the models since the viscosity may have been out of the usable range. A correction for the A constant to fit lower porosity membranes may be useful to determine.
  • the identification of the model best describing permeability for the X50 fiber may allow predictions of the maximum oil flux and may be helpful in evaluating oil/water separation performance of the hollow fiber membrane contactor.
  • the model may also be used as a performance guideline when utilizing the technology for oil purification purposes in the case of low water content in the feed.
  • FIG. 2 Membrane contactor in for oil/water separation configuration. Drawing adapted from Liqui-Cel®.
  • FIG. 6 Schematic of membrane system used for pure oil experiment.
  • FIG. 8 Effect of transmembrane pressure on oil flux corrected for viscosity.
  • symbols experimental data
  • dotted lines linear best fit.
  • Experimental conditions: IsoparTM M, Influent flow rate 3.8 liters/min; Membrane A. All experiments were conducted in triplicates; error bars are shown but are frequently smaller than the data symbols.
  • FIG. 9 Effect of influent flow rate on oil flux corrected for viscosity.
  • Experimental conditions IsoparTM M, Membrane A. All experiments were conducted in triplicates; error bars are shown but are frequently smaller than the data symbols.
  • FIG. 10 Comparison of module surface areas for oil flux vs transmembrane pressure.
  • symbols experimental data
  • dotted lines linear best fit.
  • Experimental conditions: IsoparTM M, Influent flow rate 3.8 liters/min unless stated otherwise, Membrane A and B. All experiments were conducted in triplicates; error bars are shown but are frequently smaller than the data symbols. Except for *, single replicate.
  • FIG. 11 Comparison of module surface areas for oil flux vs viscosity.
  • FIG. 13 Comparison of fiber type for oil flux vs transmembrane pressure.
  • Influent flow rate 3.8 L/min.
  • Influent flow rate 1.9, 3.8 and 5.7 L/min.
  • FIG. 14 Parity plot for the X50 fiber membrane contactor data set.
  • Efficient oil/water separation is important in many industries. Treatment of produced water, for instance, requires technologies capable of withstanding variable feeds and removing micron size oil droplets while maintaining long term performance.
  • Mechanism and key process parameters controlling the separation of oil and water in a hydrophobic, hollow- fiber membrane contactor are evaluated. Two ranges of oil feed concentration are identified. For high oil feed concentrations above 40% (v/v), increases in transmembrane pressure, detention time, and temperature are shown to increase oil flux across the fiber walls, which is consistent with the results from pure oil experiments (see above). However, for dilute mixtures (less than 2% (v/v)), experimental results indicate that an increase in transmembrane pressure and temperature lowers oil flux.
  • each membrane module may allow internal coalescence of oil droplets on the fiber surface followed by permeation of insoluble oil while maintaining water rejection.
  • the results may indicate that the stable formation of an oil film on the fiber walls is useful for the process.
  • several experiments carried out over two-week periods showed no signs of viscous fouling or flux reduction, suggesting that stable, long-term operation is possible.
  • Oil/water separations are useful in processes associated with food, biochemical and chemical production as well as in the oil and gas industry.
  • produced water treatment is driven by regulations controlling well disposal and environmental releases as well as by the financial interest in recovering oil contained in the wastewater.
  • New technologies are needed to treat the oil and gas wastewater in an affordable and efficient way.
  • Membrane technologies show great promise for oil removal in produced water as they are affordable, flexible, have a small footprint and can recover small oil droplets not captured by conventional
  • Membranes can be used for oil/water separation through direct separation or coalescence followed by gravity settling. Direct separation of oil and water is achieved in a membrane system through sieving and membrane material selectivity. For hydrophilic membrane surfaces, water is permeated through the pores while oil is rejected by the membrane material. For hydrophobic materials, the oil permeates and water is rejected.
  • process parameters may be useful to control oil and water separation in membranes including cross-flow velocity, transmembrane pressure, temperature, and oil concentration in the feed. For example, with increasing
  • Oil flux may increase through hydrophobic membranes with increasing TMP, as may be determined using the Hagen-Poiseuille model. Lower oil concentrations may decrease oil permeation because of a reduced probability for oil droplets to contact the membrane wall.
  • Increasing cross flow velocity may increase oil flux across the membrane.
  • a single phase of oil may be recovered along with high effluent water quality for a concentration of l%(v/v) oil.
  • TMP transmembrane pressure
  • Microchannel studies may be useful for simulating and observing coalescence mechanisms at pore scale and identification of important parameters.
  • a see-through microchannel may be used to observe coalescing and phase inversion mechanisms of oil droplets in water while permeating a hydrophobic PTFE pore. The formation of larger oil droplets may be observed, which may aid the subsequent oil/water gravity separation.
  • De-emulsification efficiency may improve with oil concentration and flow rate in the micro-channel, corresponding to operating a membrane with higher transmembrane pressure. Longer channels may also lead to longer contact times and higher coalescence efficiency.
  • operating with pore sizes small enough for the oil droplets to be able to contact the membrane surface but not so small as to re- disperse the emulsion may be useful.
  • Various membrane configurations may be used for the separation of oil and water including flat sheet, tubular, spiral and hollow fiber. These configurations may be used with hydrophilic material for water permeation and oil rejection. Permeation of oil through
  • hydrophobic hollow fiber membrane contactors may be more advantageous, however.
  • This Example evaluates a commercially available, hydrophobic, hollow-fiber membrane contactor for the separation of oil and water.
  • the membrane contactor extracted the majority of the available nonpolar oil for most of the feedstocks. Backed by this success, additional mixtures were tested for up to two weeks and demonstrated an increase of performance over time to reach high level recoveries above 95% without observation of the typical viscous fouling experienced by hydrophilic systems. While these results were promising, little was known on optimization of the process performance and separation mechanisms.
  • the hollow fiber geometry of the membrane module allows internal coalescence of oil droplets between fibers combined with selective permeation of oil through the membrane.
  • the present Example aims at identifying the effect of key process parameters and understanding the underlying mechanisms of oil/water separation in the hollow fiber membrane contactor. In particular, a broad range of oil concentrations, transmembrane pressures, influent flow rates and viscosities are investigated for controlled oil/water mixtures.
  • membrane contactor Commercially available in various sizes, manufactured by Liqui-Cel® and referred to as Extra Flow Contactors.
  • This contactor originally developed for liquid-liquid extraction, is commonly used for de-gassing liquid applications such as oxygen removal from water in the microelectronic industry.
  • the membranes have a hydrophobic microporous hollow- fiber design with a central baffle and shell-side distributor (FIG. 2).
  • the membrane contactors are available in various industrial module sizes including a 14-inch diameter version reported to process up to 550 gpm.
  • the module design consists of a hollow-fiber sheet/membrane rolled into a cylindrical casing allowing a pseudo cross-flow filtration. Four ports allow circulation of fluids on the shell-side (outside of fibers) and the tube side (inside of fibers).
  • the pore size distribution of the membranes was determined independently via scanning electron microscopy (Hitachi S5500 SEM/STEM, USA) with a gold and palladium coating of the membrane sample. Image analysis was conducted with the ImageJ image processing program developed by the National Institutes of Health (USA). As can be seen in FIG. 15, a range of pore sizes was observed with a mean equal to 0.047 ⁇ . Two fiber types are used in the manufacturing of these membrane contactors; the X50 fiber type was selected over the X40 type to maximize oil permeation across a more porous and thinner membrane wall (see above). [0152] Oil compounds and non-stabilized emulsion. IsoparTM, the synthetic oil used in this study, is a controlled mixture of paraffinic compounds manufactured by ExxonMobil. Various grades of IsoparTM oil have different viscosities as shown in Table 2.
  • FIG. 17 shows the "one pass through” configuration where oil-water emulsions were sent through the membrane system in a single pass. The oil permeated on the tube side was measured, while the retentate was discarded.
  • FIG. 18 shows the system operated in recycling mode. Two modules were used in series at all times to guarantee high oil removal on the shell side and allow recirculation of oil-free water back to the feed tank for long-term operation.
  • Influent flow rate and pressures were adjusted with a variable speed drive (Emerson, USA) controlling the pump and a needle valve located on the outlet line. Influent flow rates were measured with Micro Motion flow meters. Temperature was monitored with a Rosemount RTD sensor (Emerson, USA). Pressures on the shell side were collected with Rosemount pressure transmitters (Emerson, USA).
  • Oil-Water Experiments Tests were performed by feeding an oil-water emulsion to the shell side inlet of the module where oil permeated the membrane and was quantified. Oil concentration in the feed was controlled by the peristaltic pump and calculated on a volumetric basis. Transmembrane pressure was computed from the average of inlet and outlet pressures of the shell side, while the tube side was left open to atmospheric pressure. Oil flux across the membrane was the indicator used to characterize system performance. Oil recoveries were also calculated from the ratio of injected to permeated oil volumes. The experimental plan is presented in Table 7. Experiments where design to test the effect of one operating parameter while the other were maintain constant. The ranges of study of all parameters were chosen within the operation specifications of the membrane contactor.
  • the pure oil experiment defines the maximum flux across the membrane for a given set of operating conditions.
  • the flux measured across the membrane was consistently lower than the flux observed with pure oil feed, which suggests that the effective surface area was smaller than the total surface area.
  • the oil flux observed for experiments with lower oil content in the feed than the maximum pure oil flux was observed to decrease with decreasing oil
  • FIG. 19A also shows that an increase in transmembrane pressure leads to an increase in oil flux for oil concentrations above 5% (v/v).
  • the results show that at a transmembrane pressure of 20 psi, the curve reaches a plateau above 20% (v/v) oil concentration, while the 40 psi curve consistently increases.
  • Hydrophobic membranes may be used for coalescence of oil/water emulsions where oil droplets coalesce by contacting the hydrophobic pore material.
  • a gravity settler may be ultimately used to separate the two phases.
  • the particular geometry and material of the modules used in the present Example may allow coalescence to take place on the membrane fibers leading to the formation of an oil film coating the membrane wall.
  • the transmembrane pressure applied across the membrane wall subsequently drives the selective oil transfer through the porous wall.
  • the effective membrane surface area may be linked to the actual surface area of the oil film present on the fibers. Oil droplets may contact and coalesce with the oil layer rather than the non- wetted membrane suggesting that cohesion forces are greater than the adhesion forces.
  • the stability of the oil film on the fibers may be useful for oil permeation.
  • oil may be forced through the pores at a rate preventing the oil film from staying and growing on the fibers, permanently reducing the effective surface area.
  • operating with lower transmembrane pressure may lead to the formation and growth of a stable oil layer leading to consistently higher oil permeation through the pores as seen in FIG. 2 IB.
  • transmembrane pressure possibly reducing the effect of retention time for oil permeation.
  • Oil Viscosity Oil viscosity is expected to influence oil flux across membranes and typically the higher the oil viscosity, the lower the oil flux is expected to be. Tests with various IsoparTM grades were conducted with the membrane contactor to evaluate the effect of oil viscosity on the process. Under pure oil conditions, oil flux was shown to be inversely related to viscosity (see above); increased viscosity, led to reduced flux. At 80% (v/v) oil concentration, the flux seen in FIG. 23 A decreased with higher viscosity but not linearly as observed in the pure oil
  • transmembrane pressure and affected by viscosity.
  • oil content is low, there is less oil available to cover the membrane surface and rapid permeation of oil may hinder long-term flux by reducing the long-term effective area.
  • the stability of the film may provide greater effective surface area and higher oil flux even though oil is transported at a lower rate across the fiber.
  • changes in transmembrane pressure did not affect oil flux significantly. This observation suggests that the predominant mechanism contributing to oil film stability was the viscosity and not the transmembrane pressure under these conditions.
  • Coalescence may occur when oil droplets are small enough for the droplets to have a chance to contact the pore but not so small that the emulsion is re-dispersed.
  • the space between fibers of the studied module could be thought of as a long pore where droplets coalesce between the fibers.
  • the present results suggest that an increase of space between fibers from 40 to 80 ⁇ m is beneficial for the coalescence of the emulsion used in this work that had an average oil droplet size of 5.4 ⁇ m based on FIG. 16. This result does may contradict research conducted with coalescence in membrane pores. For example, if the pore is much larger than the oil droplet size, the drops may flow between the walls without coalescing.
  • the difference with conventional coalescence in membranes and the present coalescence mechanism may be the length of the coalescing channels.
  • Membrane pores may be only as long as the wall thickness, usually a few hundred micrometers.
  • the coalescing channels are the entire length of the module of approximatively 20 cm increasing the chance for the oil droplets to contact the fibers wall.
  • the increase of space between fibers may improve water flow distribution out of the coalescing area leading to better coalescence and stability of the film on the fiber.
  • the 0.7 m 2 module yields 73% higher oil flux than the 1.4 m 2 module for low transmembrane pressure (20 psi). That increase is 45% higher at higher transmembrane pressure (40 psi).
  • oil droplets 2610 are illustrated between porous hollow fibers 2620.
  • Partial surface layers 2630 comprising oil are illustrated as present on the surface of the hollow fibers. As the oil droplets contact the partial surface layers 2630, the oil may coalesce and pass through the pores of the fiber 2620.
  • Partial surface layers 2640 are illustrated where oil is not present and so pores in the hollow fibers may be blocked (e.g., by water) and transport of oil through or past the partial surface layers (black arrows in FIG. 26) may not be possible.
  • Comparison to an equivalent hydrophilic system Comparing the performance of hydrophilic and hydrophobic membrane systems can be challenging since the permeated phases across the membranes are different.
  • the equivalent water flux for a hydrophilic membrane would correspond to permeate the entirety of the water phase through 1.4 m 2 surface area, which equals a flux of 4.5x10 -5 m 3 /s m 2 .
  • the hydrophobic system studied here was shown to maintain such performance up to two weeks (FIG. 20C), which may not be the case of the equivalent hydrophilic system that may experience decrease of flux over-time due to viscous fouling.
  • Hollow-fiber membranes may be used for oil/water separation with water permeation across a modified PVDF membrane surface. Higher water flux may be expected than for typical PVDF membranes.
  • the highest flux attained under a transmembrane pressure of 50 psi and oil feed concentration of 500 ppm may be around 1.9 x 10 -5 m 3 /s m 2 prior to the observed decrease in performance due to fouling. Therefore, the results found here are comparable to the results for a hydrophilic system at start-up without the disadvantage of viscous fouling occurring over time, that would ultimately reduce the permeation flux. [0172] Therefore, the present hydrophobic membrane contactor represents a viable competitor to hydrophilic systems as comparable oil removal capability is obtained but also maintained over time.
  • Transmembrane pressure was shown to have different effects depending on the oil feed concentration range. For higher concentrations, increasing transmembrane pressure was confirmed to increase oil permeation. However, for oil concentrations below 2% (v/v), operating the system with lower transmembrane pressure was shown to be more beneficial by improving the oil film stability of the membrane wall.
  • transmembrane pressure which further supports the hypothesis that the development and stability of oil films on the membrane fibers are advantageous in optimizing the performance of the contactor.
  • the data presented in this Example may confirm the hypothesized mechanism that oil droplets coalescing on the fibers and progressively coating the membrane wall at low oil concentrations is advantageous for successful operation.
  • the stability of the oil film on the fibers is a useful mechanism at low oil concentration and leads to progressive improvement of performance over time.
  • the hydrophobic membrane contactor can withstand and perform efficiently for a wide range of oil concentrations for which fouling in hydrophilic systems would be prohibitive, which makes this technology a prime candidate for many applications ranging from oil removal in very dilute oil/water mixtures to oil purification processes.
  • FIG. 2 Extra flow Liqui-cel® membrane contactor in for oil/water separation configuration. Drawing adapted from Liqui-Cel®.
  • FIG. 15 Pore size analysis of membrane pores.
  • Left panel Scanning Electron
  • FIG. 16 Droplet size distribution of IsoparTM-water non-stabilized emulsion entering the system.
  • FIG. 17 “One pass through” system schematic.
  • FIG. 18 System schematic in recycling mode.
  • FIGs. 19A-19B Effect of influent oil concentration on oil flux across the membrane.
  • FIGs. 21A-21B Effect of transmembrane pressure on oil flux.
  • FIGs. 23A-23C Effect of viscosity on oil flux.
  • FIG. 24 Effect of viscosity on oil recovery.
  • Influent flow rate 3.78 L/min.
  • TMP 20 psi; Membrane A, Isopar V used initially, then switched to Isopar M, then back to Isopar V.
  • FIG. 25 A Effect of surface area on oil flux.
  • FIG. 25B Effect of surface area on recovery.
  • FIG. 26 Schematic of oil droplet coalescence between fibers for two transmembrane pressures.
  • FIG. 27 Schematic for equivalent hydrophilic system
  • EXAMPLE 3 MODELLING OF A MICROPOROUS HOLLOW FIBER MEMBRANE
  • a semi-empirical model is developed and relates the effective surface area to the operating parameters of the system such as transmembrane pressure, influent flow rate, oil viscosity, as well as to membrane characteristics.
  • the model predicts the required surface area for a chosen oil recovery under a given set of operating conditions. For instance, an oil/water influent containing 0.1% (v/v) of oil with a viscosity of 4 cP and entering the system at a rate of 240 1/hr would require 8 m 2 of membrane surface area to reach 95% oil removal if operated under a transmembrane pressure of 1.5 bar.
  • the model can be used to predict oil removal for a given surface area as well and, therefore, provides a valuable tool for design of a separation system with the membrane contactor.
  • a mechanistic model may be useful for describing the limiting permeation of water in a highly hydrophilic UF membrane system. Fouling behavior follows Hermia's models and water flux may become pressure independent over time. Viscosity, oil concentration and shear rate may be useful parameters affecting water flux across the hydrophilic membranes. Experimental data collected from hydrophilic systems may be compared to the transport models used for microfiltration such as Brownian diffusion, shear-induced diffusion or inertial lift. For example, a combination of Brownian and shear-induced diffusion in a model may obtain good agreement between predicted and measured steady state flux data in a hydrophilic microfiltration system treating industrial oily wastewater.
  • an analogy may be used between the solid particles in a liquid phase and the oil droplets in the water phase.
  • Oil droplets are analogous to solid particles that progressively block the pores or form a layer at the membrane surface leading to a reduced steady state liquid flux across the membrane.
  • the analogy does not hold since the "particles" (i.e., oil droplets) are actually permeated and not sieved as in conventional microfiltration. Permeation behavior in hydrophobic microfiltration of oil/water mixtures may differ from the typical trends observed in hydrophilic systems.
  • transmembrane pressure and viscosity may have opposite effects when the oil concentration is varied from high to low in the oil/water mixtures.
  • Use of hydrophobic systems for the separation of oil and water may be useful in identifying operating parameters involved in the filtration system.
  • This Example presents a semi-mechanistic, semi-empirical performance model for the design and operation of a microporous hydrophobic membrane contactor for the recovery of insoluble oil from oil/water mixtures.
  • the technology may be applicable for applications ranging from oil purification to the treatment of very dilute oil/water mixtures such as would be typical of the final oil removal step in produced water treatment.
  • the model is useful for determining membrane surface area and operating conditions needed to optimize process performance for a given oil/water stream.
  • Synthetic Oil Various IsoparTM grades were used as the synthetic oil for this study. IsoparTM oils are controlled mixtures of paraffinic compounds and manufactured by ExxonMobil (purchased from Nexeo Solutions, USA). Each IsoparTM grade has a different viscosity as shown in Table 2.
  • Oil-Water Experiments Experiments were conducted during which the separation of oil/water mixtures were monitored and assessed under various operating conditions presented in Table 10.
  • Table 10 Experimental conditions for oil/water experiments with the membrane contactor.
  • the product water was collected in a second tank. Oil permeation was measured on the tube side and the aqueous retentate was discarded.
  • a 2.5 inch diameter module was mounted in series with a 4 inch diameter "guard" module used to provide complete oil removal on the shell side and allow recirculation of oil-free water back to the feed tank (FIG. 18).
  • oil and water were emulsified and circulated in the system with the high shear regenerative turbine pump (MTH pump, USA).
  • a peristaltic pump (Thermo Fischer Scientific, USA), connected to the suction of the high shear pump, provided control of the oil injection rate.
  • the pump's variable speed drive and a needle valve located on the outlet line were used to control influent flow rates and pressures.
  • Influent flow rates were measured with mass flow meters (MicroMotion, USA). Temperatures were monitored with Rosemount RTD sensors (Emerson, USA). Pressures on the shell side were measured with Rosemount pressure transmitters (Emerson, USA). Mass quantities of oil injected to the system and permeating the membranes were monitored with three weigh scales (Arlyn, USA).
  • Tests were performed by circulating an oil-water mixture though the membrane contactor from the bottom to the top shell-side ports. Influent oil concentration was computed as a volume fraction. Permeated oil was collected and quantified with the scale from the top tube side port. Transmembrane pressure was calculated as the average pressure at the inlet and outlet ports, while the tube side was left open to atmospheric pressure. The performance of the system was evaluated based on oil flux and oil recovery.
  • Oil flux and surface area model development Effective surface area. In this study, a model describing oil flux across a hydrophobic membrane is described. The model compares to an approach used in packed column absorption.
  • the oil transferred across the membrane walls is the product of an intrinsic oil flux (Oil fluxpure) determined during the pure oil feed experiments and the effective surface area of the membrane.
  • the effective membrane area for coalescence may be significantly less than the actual membrane area, especially at high water concentrations.
  • the high concentrations of water may enter pores or cause the membrane surface to become hydrophilic which results in effectively blocking transport of the oil. Equation 6:
  • Equation 6 can be re-written to express oil flux across the membrane in terms of the intrinsic flux and the effective area ratio defined as the effective area over the total area of the membrane (a e /a p ) (Equation 7). Equation 7: where:
  • Oil FIUXOAV Oil flux for oil/water mixtures in m 3 /s m 2
  • Oil Fluxpure oii Oil flux for pure oil feed conditions in m 3 /s m 2
  • the membrane contactor performance depends on the intrinsic flux (Oil flux Pure ) and the effective area ratio, which depends on the entering oil rate.
  • the first step in developing the model was to characterize the intrinsic flux using membrane characteristics and operating parameters. Pure oil feed experiments allowed quantification of the intrinsic flux of the system where the total membrane surface is assumed to be available and maximized oil permeation is observed. Then, assuming the effective area ratio equals the ratio of measured flux for oil/water experiments to pure oil feed flux, a model was developed to predict the effective area ratio as a function of the operating parameters. Finally, a complete equation for oil flux prediction as a function of transmembrane pressure, viscosity, oil concentration, influent flow rates and membrane characteristics can be obtained and used to predict surface area requirements as a function of operating conditions. An example calculation can be found at the end of this Example.
  • Equation 8 The model describing the experimental data most accurately for pure oil feed filtration with the X50 membrane contactor is shown in Equation 8 (discussed previously above).
  • the model is the Ergun equation, which may be useful for describing flow through packed columns. Equation 8 :
  • Equation 9 Effect of oil concentration.
  • the experimental curve of effective area ratio plotted versus oil concentration seen in FIG. 28 displays saturation behavior seen in other applications such as the Michaelis-Menten kinetics equation or the Langmuir adsorption isotherm.
  • the membrane pores, which are directly related to oil permeation, can be considered as available sites, while the driving force for accumulation onto the membrane wall is the oil concentration.
  • the effective area ratio is linearly related. Oil concentration linearly increases the probability for the oil droplets to contact the membrane wall.
  • the curve reaches saturation where the effective surface area approaches the total surface area available for oil permeation. Therefore, a saturation model as presented in Equation 9 should provide a good fit to correlate effective area ratio to oil concentration (assuming that the ratio of flux to the pure oil feed flux is equal to the effective area ratio). Equation 9: with a and b constants.
  • TMP Transmembrane pressure in Pa
  • the residual plot seen in FIG. 32 shows the desired random scatter of data points with a slight unbalance to the left due to a higher number of experiments conducted for low oil concentrations.
  • the model was first fit with the lower oil concentration range leading to reduced residuals on the left hand side of the plot.
  • FIGs. 34A-34F shows a few examples of the model prediction compared to the experimentally measured data for various transmembrane pressures, viscosities, oil concentrations, and influent flow rates. The model fits are reasonable but in some cases the model underestimates the data. It should be noted the model is developed using chemical systems possessing a high interfacial tension (-50 dynes/cm) and should be used with some caution with systems with much lower interfacial tensions. [0220] The model can then be incorporated into Equation 7 and a model for oil flux prediction is obtained and shown in Equation 1 1 for modules with surface area less than or equal to 1.4 m 2 . Equation 1 1 :
  • Oil FIUX O/V Oil flux across the membrane in m 3 /s m 2 ,
  • FIGs. 36A-36J shows a few model simulations for every operating parameter affecting the process in the experimental ranges studied (Table 1 1). A number of conclusions can be drawn from FIGs. 36A-36J. A lower pressure may be more efficient for dilute mixtures ( ⁇ 2% (v/v) oil concentration), while higher transmembrane pressures will be preferable at high oil concentrations (40% (v/v) oil concentration) (FIG. 36A, FIG. 36B, and FIG. 36C).
  • the model can be used to compare performance as a function of influent flow rates. In this case, operating at a higher transmembrane pressure will lead to higher oil flux and recovery (FIG. 36G and FIG. 36H). Viscosity can also be used as an operating lever by changing temperature of a given influent stream for instance.
  • the model predicts the positive effect of increasing viscosity on oil permeation for dilute mixtures ( ⁇ 200 ppm oil concentration) and the negative effect of increasing viscosity for higher concentrations (FIG. 36D, FIG. 36E, FIG. 36F).
  • FIGs. 36I-36J show the decrease of effective surface area with increasing water content in the feed. The model predicts all the trends observed above accurately.
  • the effective surface area needed to reach a given oil recovery may then be determined with the selected operating conditions.
  • the advantage of using the 2.5 inch diameter module for the experimental data is that surface area is small and performance sensitivity may be accurately measured.
  • the small module can be thought as a slice of the larger modules and, therefore, the model regarded as a differential equation.
  • a possible limitation associated with the proposed scale-up approach is the difference of shell-side pressure drop occurring in modules of different sizes.
  • the model was developed with experimental data collected with the 2.5 x 8 inch module and flux predictions account for the pressure drop related to the small module size and working influent flow rates. If a larger module is selected for a higher surface area, the pressure drop profile may be different leading to less accuracy of the model prediction. Therefore, the model is useful in providing a first estimation of membrane performance.
  • Table 12 Surface area predictions for 95% oil recovery.
  • TMP Transmembrane pressure (Pa)
  • Oil FIUX O/W Oil flux across the membrane in m 3 /s m 2
  • Oil Fluxpure oii Oil flux across the membrane obtained with pure oil feed in m 3 /s m 2
  • the model can be used to predict membrane surface area needed to obtain desired insoluble oil removal from an oil/water feed mixture.
  • the effective area model developed in this work is based on studies using a high interfacial tension (-50 dynes/cm) and without solids so some caution should be applied. Additional work may address the comparative effects of lower interfacial tensions and solids.
  • FIG. 31 Effect of viscosity on flux ratio for various oil feed concentrations.
  • FIG. 32 Residual plot for the complete oil/water experimental data set presented in Table 10.
  • FIG. 33 Experimentally determined flux ratio vs a e /a t model estimated values. The model used is shown in Equation 8. The linear fit correlation factor of the solid line is 0.96.
  • Amelio et al Purification of biodiesel using a membrane contactor: Liquid-liquid extraction. Fuel Processing Technology 142, 352-360 (2016).

Abstract

Described herein are methods and systems for recovering oil from an oil and water mixture. Some embodiments make use of hydrophobic hollow fiber membrane systems that allow permeation of oil from outside the hollow fibers to an interior of the hollow fiber. Due to their hydrophobic nature, the hollow fibers may exclude water from permeating into the interior of the fibers.

Description

OIL RECOVERY USING MICROPOROUS HYDROPHOBIC
MEMBRANE CONTACTORS
CROSS REFERENCE TO RELATED APPLICATION
[0001] This application claims the benefit of and priority to U.S. Provisional Application 62/321,924, filed on April 13, 2016, which is hereby incorporated by reference in its entirety.
BACKGROUND
[0002] This invention is in the field of oil and water separation techniques. This invention relates generally to methods and systems for efficiently separating water and insoluble oil from mixtures using hydrophobic porous membrane contactors. [0003] Oil/water separations are useful in processes associated with food, biochemical and chemical production as well as in the oil and gas industry. In particular, produced water treatment is driven by regulations controlling well disposal and environmental releases as well as by the financial interest in recovering oil contained in the wastewater. New technologies are needed to treat the oil and gas wastewater in an affordable and efficient way. Membrane technologies show great promise for oil removal in produced water as they are affordable, flexible, have a small footprint and can recover small oil droplets not captured by conventional technologies.
SUMMARY
[0004] Described herein are methods and systems for recovering oil from an oil and water mixture. Some embodiments make use of hydrophobic hollow fiber membrane systems that allow permeation of oil from outside the hollow fibers to an interior of the hollow fiber. Due to their hydrophobic nature, the hollow fibers may exclude water from permeating into the interior of the fibers. Various process conditions are disclosed, in some embodiments, as impacting a rate of oil flux into the interior of the hollow fibers and control over these process conditions allows for more efficient separation of the mixture. [0005] In a first aspect, methods are provided, such as methods for recovering oil from an oil and water mixture. In some embodiments, methods of this aspect may comprise determining a ratio of oil to water in a mixture comprising oil and water, such as a mixture comprising water and oil that are at least partly immiscible; identifying a transmembrane pressure, a process viscosity, or both for use in separating the oil and the water from the mixture; contacting a shell side of a hollow fiber with the mixture, such as a hollow fiber that comprises a porous and hydrophobic polymer material, and where at least a portion of the oil in the mixture wets the hollow fiber and is transported through the hollow fiber to a tube side of the hollow fiber, and where contacting includes applying the transmembrane pressure to the shell side of the hollow fiber, or changing a viscosity of the oil to the process viscosity, or both to enhance transport of oil through the hollow fiber; and collecting oil from the tube side of the hollow fiber.
[0006] Optionally, the transmembrane pressure is identified as a first transmembrane pressure when the ratio of oil to water is less than a first threshold ratio. Optionally, the transmembrane pressure is identified as a second transmembrane pressure when the ratio of oil to water is greater than the first threshold ratio. Optionally, the first transmembrane pressure is less than the second transmembrane pressure. In this way, different operating regimes can be employed to more effectively separate the oil from the mixture. For example, in some embodiments, when a concentration of oil in the mixture is low, a lower transmembrane pressure may more effectively result in removal of oil from the mixture. In some other embodiments, when a concentration of oil in the mixture is high, a higher transmembrane pressure may more effectively result in removal of oil from the mixture.
[0007] Optionally, the process viscosity is identified as a first viscosity when the ratio of oil to water is less than a second threshold ratio. Optionally, the process viscosity is identified as a second viscosity when the ratio of oil to water is greater than the second threshold ratio.
Optionally, the first viscosity is greater than the second viscosity. In this way, different operating regimes can be employed to more effectively separate the oil from the mixture. For example, in some embodiments, when a concentration of oil in the mixture is low, a higher process viscosity may more effectively result in removal of oil from the mixture. In some other embodiments, when a concentration of oil in the mixture is high, a lower process viscosity may more effectively result in removal of oil from the mixture.
[0008] Optionally, identifying the transmembrane pressure, the process viscosity, or both includes determining an effective area ratio for the hollow fiber. For example, in some embodiments, determining the effective area ratio includes computing where
Figure imgf000004_0001
ae/at corresponds to the effective area ratio, a is a constant, b is a constant and Coil is the ratio of oil to water in the mixture. Optionally, a = a'. TMP-1 6. μ1 .0. Q0.3 ,where a' is a constant, TMP is the transmembrane pressure in Pa, μ is the process viscosity in Pa s, Q is a shell side flow rate of the mixture in m3/s, and Coil is expressed as a volume fraction. As one example, a' may be about 9.27 x 1012. Optionally, b = b'. TMP-2.1.μ1 .1.Q0.4, where b' is a constant, TMP is the transmembrane pressure in Pa, μ is the process viscosity in Pa- s, Q is a shell side flow rate of the mixture in m3/s, and Coil is expressed as a volume fraction. As one example, b' may be about 2.65 x 1016. [0009] In various embodiments, the transmembrane pressure is selected based on a concentration of oil in the mixture in order to allow for efficient separation of oil from the mixture. For example, the first transmembrane pressure is optionally selected from the range of 1 pounds per square inch to 40 pounds per square inch. Optionally, the second transmembrane pressure is selected from the range of 20 pounds per square inch to 100 pounds per square inch. In this way, different transmembrane pressures may be used depending on the concentration of oil in the mixture.
[0010] For example, when the concentration of oil in the mixture is in a low concentration regime, such as less than 20%, less than 10%, less than 5%, less than 2%, or less than 1%, embodiments may employ a transmembrane pressure less than 40 pounds per square inch, such as a transmembrane pressure selected from the range of 1 psi to 40 psi, selected from the range of 5 psi to 40 psi, selected from the range of 10 psi to 40 psi, selected from the range of 1 psi to 30 psi, selected from the range of 5 psi to 30 psi, selected from the range of 10 psi to 30 psi, selected from the range of 1 psi to 35 psi, selected from the range of 1 psi to 25 psi, or selected from the range of 1 psi to 15 psi. It will be appreciated that the listed pressure ranges are merely examples and that other transmembrane pressure values may be useful with oil concentrations less than 20%.
[0011] As another example, when the concentration of oil in the mixture is in a high
concentration regime, such as higher than 20% or higher than 40%, embodiments may employ a transmembrane pressure greater than 20 pounds per square inch, such as a transmembrane pressure selected from the range of 20 psi to 100 psi, selected from the range of 20 psi to 80 psi, selected from the range of 20 psi to 60 psi, selected from the range of 30 psi to 100 psi, selected from the range of 30 psi to 80 psi, selected from the range of 30 psi to 60 psi, selected from the range of 40 psi to 100 psi, selected from the range of 40 psi to 80 psi, or selected from the range of 40 psi to 60 psi. It will be appreciated that the listed pressure ranges are merely examples and that other transmembrane pressure values may be useful with oil concentrations greater than 20%. [0012] In various embodiments, the process viscosity is selected based on a concentration of oil in the mixture in order to allow for efficient separation of oil from the mixture. For example, the first process viscosity is optionally selected from the range of 0.2 cP to 100 cP. Optionally, the second viscosity is selected from the range of 0.2 cP to 100 cP. In this way, different process viscosities may be used depending on the concentration of oil in the mixture. [0013] For example, when the concentration of oil in the mixture is in a low concentration regime, such as less than 20%, less than 10%, less than 5%, less than 2%, or less than 1%, embodiments may adjust the oil viscosity to a value of between 0.2 cP to 100 cP, such as a viscosity selected from the range of 1 cP to 100 cP, selected from the range of 10 cP to 100 cP, selected from the range of 20 cP to 100 cP, selected from the range of 30 cP to 100 cP, selected from the range of 40 cP to 100 cP, selected from the range of 50 cP to 100 cP, selected from the range of 60 cP to 100 cP, selected from the range of 70 cP to 100 cP, or selected from the range of 80 cP to 100 cP. It will be appreciated that the listed viscosity ranges are merely examples and that other viscosity values may be useful with oil concentrations less than 20%.
[0014] As another example, when the concentration of oil in the mixture is in a high
concentration regime, such as higher than 20% or higher than 40%, embodiments may adjust the oil viscosity to a value of between 0.2 cP to 100 cP, such as a viscosity selected from the range of 0.2 psi to 90 cP, selected from the range of 0.2 psi to 80 cP, selected from the range of 0.2 cP to 70 cP, selected from the range of 0.2 cP to 60 cP, selected from the range of 0.2 cP to 50 cP, selected from the range of 0.2 cP to 40 psi, selected from the range of 0.2 cP to 35 cP, selected from the range of 0.2 cP to 30 cP, or selected from the range of 0.2 psi to 25 cP. It will be appreciated that the listed viscosity ranges are merely examples and that other viscosity values may be useful with oil concentrations greater than 20%. [0015] For example, in some embodiments, selection of the process viscosity may vary as a function of length of the hollow fiber, as the oil concentration in the mixture present at the shell side of the hollow fiber may be reduced by passage of oil into the tube side of the hollow fiber. Thus, in some embodiments, the first viscosity is used on a first portion of the hollow fiber and a second viscosity is used on a second portion of the hollow fiber. In this way, a single hollow fiber (or hollow fiber membrane comprising a plurality of hollow fibers) may have different process conditions allowing more efficient oil flux into the tube side of the hollow fiber than a single process condition. For example, a process viscosity close to an entrance port of a hollow fiber membrane contactor may be lower than the process viscosity close to an exit port of a hollow fiber membrane contactor. [0016] It will be appreciated that temperature may be used to control the viscosity of oil in the mixture. For example, changing the viscosity of the oil may optionally include heating the mixture, heating the hollow fiber, or both. In some embodiments, heating the mixture, the hollow fiber, or both may result in a decrease of the oil viscosity. As another example, changing the viscosity of the oil may optionally include cooling the mixture, cooling the hollow fiber, or both. In some embodiments, cooling the mixture, the hollow fiber, or both may result in an increase of the oil viscosity. It will be appreciated that a single hollow fiber that has different process conditions allowing more efficient oil flux into the tube side of the hollow fiber than a single process condition may utilize different temperature regimes for the different portions of the hollow fiber.
[0017] Various oil and water mixtures are useful with the methods and systems disclosed herein. For example, the oil and the water may be completely immiscible. Optionally, at least a portion of the oil may optionally be present in the mixture as microdroplets of oil suspended in water. For example, at least a portion of the microdroplets may have a cross-sectional dimension less than 1 μπι, less than 5 μπι, less than 10 μπι, less than 15 μπι, less than 20 μπι, or less than 50 μπι.
Optionally, at least a portion of the microdroplets have a cross sectional dimension greater than 0.1 μπι, selected from the range of 0.1 μπι to 10 μπι, selected from the range of 0.1 μπι to 20 μπι, or selected from the range of 0.1 μπι to 30 μπι.
[0018] Optionally, the ratio of oil to water in the mixture may correspond to a concentration of oil in water in the mixture. Optionally, the ratio of oil to water in the mixture may correspond to a concentration of water in oil in the mixture. Optionally, the ratio of oil to water in the mixture may correspond to a mass fraction of oil in the mixture. Optionally, the ratio of oil to water in the mixture may correspond to a mass fraction of water in the mixture. Optionally, the ratio of oil to water in the mixture may correspond to a mole fraction of oil in the mixture. Optionally, the ratio of oil to water in the mixture may correspond to a mole fraction of water in the mixture.
Optionally, the ratio of oil to water in the mixture may correspond to a volume fraction of oil in the mixture. Optionally, the ratio of oil to water in the mixture may correspond to a volume fraction of water in the mixture. Optionally, determining the ratio of oil to water in the mixture includes determining a concentration of oil in water in the mixture, determining a concentration of water in oil in the mixture, determining a mass fraction of oil in the mixture, determining a mass fraction of water in the mixture, determining a mole fraction of oil in the mixture, determining a mole fraction of water in the mixture, determining a volume fraction of oil in the mixture, determining a volume fraction of water in the mixture, and any combination of these.
[0019] Optionally, the first threshold ratio is selected from the range of about 1 part per million to about 20%. For example, useful first threshold ratios include, but are not limited to about 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 11%, 12%, 13%, 14%, 15%, 16%, 17%, 18%, 19% or 20%). Other first threshold ratios are useful with some embodiments. Optionally, the second threshold ratio is selected from the range of about 1 part per million to about 2000 parts per million (ppm). For example, useful second threshold ratios include, but are not limited to about 1 ppm, 5 ppm, 10 ppm, 50 ppm, 100 ppm, 500 ppm, 1000 ppm, 1500 ppm, or 2000 ppm. Other second threshold ratios are useful with some embodiments. [0020] Optionally, the oil comprises one or more oils selected from the group consisting of biological oils, petroleum oils, food oils, paraffins, nonpolar hydrocarbons, saturated
hydrocarbons, unsaturated hydrocarbons, aromatic hydrocarbons, water insoluble organic compounds, and any combination of these. [0021] In various embodiments, a variety of hollow fibers are useful with the methods and systems described herein. For example, the hollow fiber optionally comprises a porous material, such as a hydrophobic material or a material having a surface rendered hydrophobic. In this way, oil may be preferentially attracted to the hydrophobic material of the hollow fiber while water may be preferentially repelled from the hydrophobic material of the hollow fiber. Optionally, the hollow fiber comprises a material selected from the group consisting of polypropylene, polyethylene, polytetrafluoroethylene, polyethylene terephthalate, polyvinylidene fluoride, polymethylpentene, and any combination of these.
[0022] Advantageously, the hollow fibers useful with the methods and systems described herein may be porous hollow fibers, such as mesoporous hollow fibers, microporous hollow fibers, or nanoporous hollow fibers. Example hollow fibers include, but are not limited to, those including a plurality of pores, such as pores having cross-sectional dimensions selected from the range of 0.001 μπι to 1 μπι. It will be appreciated that pores in a single fiber may have dimensions distributed throughout this range or within subranges within this range.
[0023] Optionally, the hollow fiber is present in a hollow fiber membrane. For example, contacting may include contacting a shell side of the hollow fiber membrane with the mixture. Optionally, collecting includes collecting oil from tube sides of the hollow fiber membrane.
Optionally, the hollow fiber membrane is present in a membrane contactor. For example, in embodiments, the membrane contactor includes a shell side inlet, a shell side outlet, and a tube side outlet. Optionally, contacting the shell side of the hollow fiber includes flowing the mixture into the shell side inlet. Optionally, collecting includes collecting oil from the tube side outlet.
[0024] In another aspect, provided are systems for recovering oil from an oil and water mixture. In some embodiments, a system comprises: a pump for pumping a mixture comprising oil and water, such as where the oil and water are at least partially immiscible; a temperature control element for heating or cooling the mixture, such as a temperature control element that is positioned in thermal communication with the mixture; a hydrophobic porous membrane contactor, such as a hydrophobic membrane contactor that is positioned in fluid communication with the pump; an oil collection system positioned in fluid communication with the tube side outlet of the hydrophobic porous membrane contactor; and a controller that controls a flow rate of the mixture using the pump, and a temperature of the mixture using the temperature control element. Optionally, the hydrophobic membrane contactor comprises a hollow fiber membrane having a shell side in fluid communication with a shell side inlet of the hydrophobic porous membrane contactor and a shell side outlet of the hydrophobic porous membrane contactor. Optionally, the hollow fiber membrane has a tube side in fluid communication with a tube side outlet of the hydrophobic porous membrane contactor. Optionally, the hollow fiber membrane comprises a plurality of porous hydrophobic polymer hollow fibers.
[0025] In some embodiments, the controller further controls a transmembrane pressure of the mixture applied to the shell side of the hollow fiber membrane, or a process viscosity of oil in the mixture, or both the transmembrane pressure and the process viscosity. For example, the controller may set the transmembrane pressure to a first transmembrane pressure when a ratio of oil to water in the mixture is less than a first threshold ratio. Optionally, the controller sets the transmembrane pressure to a second transmembrane pressure when the ratio of oil to water in the mixture is greater than the first threshold ratio. Optionally, the first transmembrane pressure is less than the second transmembrane pressure.
[0026] As another example, the controller sets the process viscosity to a first viscosity when a ratio of oil to water in the mixture is less than a first threshold ratio. Optionally, the controller sets the process viscosity to a second viscosity when the ratio of oil to water in the mixture is greater than the first threshold ratio. Optionally, the first viscosity is greater than the second viscosity. [0027] Optionally, physical parameter measurements and their control may also be useful with the methods and systems described herein. For example, it may be useful to monitor and/or control physical parameters such as flow rate, viscosity, pressure, temperature, concentration.
[0028] In some embodiments, a system of this aspect may further comprise a pressure transmitter for measuring a pressure of the mixture, wherein the pressure transmitter is positioned in fluid communication with the mixture; and a feedback circuit between the pressure transmitter and the controller for providing the pressure of the mixture to the controller for use in controlling the transmembrane pressure. In some embodiments, multiple pressure transmitters may be used for measuring the pressure at various points within the system. For example, in some embodiments, the transmembrane pressure may be estimated as an average pressure of a shell side inlet pressure and a shell side inlet pressure, such as when the tube side outlet pressure is at atmospheric or ambient pressure.
[0029] In some embodiments, a system of this aspect may further comprise a temperature sensor for measuring a temperature of the mixture, such as a temperature sensor that is positioned in thermal communication with the mixture; and a feedback circuit between the temperature sensor and the controller for providing the temperature of the mixture to the controller for use in controlling the temperature of the mixture, or the process viscosity ,or both.
[0030] In some embodiments, a system of this aspect may further comprise a concentration sensor for determining a ratio of oil to water in the mixture, such as a concentration sensor that is positioned in communication with the mixture, e.g., fluid or optical communication; and a feedback circuit between the concentration sensor and the controller for providing the ratio of oil to water in the mixture to the controller for use in controlling the transmembrane pressure, or the process viscosity, or both. [0031] In some embodiments, systems of this aspect may comprise combinations concentration sensors, temperature sensors, and/or pressure transmitters and feedback circuits for providing measured values to a controller for use in controlling the separation process.
[0032] Optionally, the hydrophobic porous membrane contactor is a first hydrophobic membrane contactor and the system further comprises a second hydrophobic porous membrane contactor in fluid communication with the shell side outlet of the first hydrophobic porous membrane contactor. For example, in some embodiments, the transmembrane pressure is a first membrane contactor transmembrane pressure, the controller sets the first membrane contactor transmembrane pressure to the second transmembrane pressure, the controller sets a second membrane contactor
transmembrane pressure applied to the second hydrophobic porous membrane contactor to the first transmembrane pressure. In this way, a dual stage separation system may be used to reduce the oil concentration in the mixture in two parts - a first reduction from a higher concentration by the first membrane contactor and a second reduction from a lower concentration by the second membrane contactor. In a specific embodiment, a first concentration of oil in the first hydrophobic porous membrane contactor is greater than a second concentration of oil in the second hydrophobic porous membrane contactor.
[0033] Advantageously, a dual stage separation system may allow for using different processing parameters at each membrane contactor. For example, the process viscosity is optionally a first membrane contactor process viscosity, and the controller sets the first process viscosity to the second viscosity, and the controller sets a second process viscosity of oil in the second
hydrophobic porous membrane contactor to the first viscosity.
[0034] In embodiments, the temperature control element is in thermal communication with the hydrophobic porous membrane contactor. Optionally, the temperature control element is a heating element, or a cooling element, or both. Optionally, the temperature control element is arranged to control a first temperature of a first portion of the hydrophobic porous membrane contactor and to control a second temperature of a second portion of the hydrophobic porous membrane contactor. In this way, a single membrane contactor may be used with different process parameters across the membrane contactor to allow for efficient separation of oil at higher and lower concentrations, for example.
[0035] Without wishing to be bound by any particular theory, there can be discussion herein of beliefs or understandings of underlying principles relating to the invention. It is recognized that regardless of the ultimate correctness of any mechanistic explanation or hypothesis, an
embodiment of the invention can nonetheless be operative and useful. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] FIG. 1 A provides a schematic illustration of a hollow fiber. FIG. IB provides a schematic illustration of a hollow fiber membrane.
[0037] FIG. 2 provides a schematic illustration of a membrane contactor, in accordance with some embodiments. [0038] FIG. 3 provides a schematic cross-sectional illustration of a membrane contactor, in accordance with some embodiments.
[0039] FIG. 4A and FIG. 4B provide details of methods for recovering oil from an oil and water mixture, in accordance with some embodiments.
[0040] FIG. 5 provides data showing viscosity as a function of temperature for three different oils.
[0041] FIG. 6 provides a schematic illustration of a water/oil separation system, in accordance with some embodiments.
[0042] FIG. 7 provides data showing the effect of viscosity on oil permeation.
[0043] FIG. 8 provides data showing the effect of transmembrane pressure on oil flux corrected for viscosity.
[0044] FIG. 9 provides data showing the effect of influent flow rate on oil flux corrected for viscosity.
[0045] FIG. 10 provides data showing a comparison of module surface areas for oil flux as a function of transmembrane pressure. [0046] FIG. 11 provides data showing a comparison of module surface areas for oil flux as a function of viscosity.
[0047] FIG. 12 provides data showing a comparison of module sizes for oil flux as a function of transmembrane pressure.
[0048] FIG. 13 provides data showing a comparison of fiber type for oil flux as a function of transmembrane pressure.
[0049] FIG. 14 provides data showing a parity plot comparing experimentally measured oil flux and predicted oil flux.
[0050] FIG. 15 provides a scanning electron micrograph image of a porous fiber and data showing a pore size distribution for a porous fiber.
[0051] FIG. 16 provides data showing droplet size distribution for droplets of oil in an oil/water mixture.
[0052] FIG. 17 provides a schematic illustration of a water/oil separation system, in accordance with some embodiments.
[0053] FIG. 18 provides a schematic illustration of a water/oil separation system, in accordance with some embodiments.
[0054] FIG. 19A and FIG. 19B provide data showing the effect of influent oil concentration on oil flux across a membrane.
[0055] FIG. 20A, FIG. 20B, and FIG. 20C provide data showing oil recovery as a function of time.
[0056] FIG. 21 A and FIG. 21B provide data showing the effect of transmembrane pressure on oil flux.
[0057] FIG. 22 provides data showing the effect of influent flow rate on oil recovery.
[0058] FIG. 23 A, FIG. 23B, and FIG. 23C provide data showing the effect of viscosity on oil flux.
[0059] FIG. 24 provides data showing the effect of viscosity on oil recovery as a function of time.
[0060] FIG. 25 A provides data showing the effect of surface area on oil flux. FIG. 25B provides data showing the effect of surface area on oil recovery. [0061] FIG. 26 provides schematic illustrations of oil droplet coalescence between fibers for two transmembrane pressures.
[0062] FIG. 27 provides a schematic representation of a hydrophilic water/oil separation system, in accordance with some embodiments. [0063] FIG. 28 provides data showing an experimental flux ratio as a function of oil
concentration in the feed.
[0064] FIG. 29 provides data showing the effect of influent flow rate on an oil flux ratio.
[0065] FIG. 30 provides data showing the effect of transmembrane pressure on an oil flux ratio for two oil concentrations. [0066] FIG. 31 provides data showing the effect of viscosity on an oil flux ratio for different oil concentrations.
[0067] FIG. 32 provides data showing model residuals as a function of effective area ratio.
[0068] FIG. 33 provides data showing experimentally flux ratios as a function of estimated effective area ratios. [0069] FIG. 34A, FIG. 34B, FIG. 34C, FIG. 34D, FIG. 34E, and FIG. 34F provide data showing examples of a model fitting experimental measurements.
[0070] FIG. 35 provides data comparing experimentally measured flux and predicted flux determined using a model.
[0071] FIG. 36A, FIG. 36B, FIG. 36C, FIG. 36D, FIG. 36E, FIG. 36F, FIG. 36G, FIG. 36H, FIG. 361, and FIG. 36 J provide data showing model predictions for oil permeation.
DETAILED DESCRIPTION
[0072] In general the terms and phrases used herein have their art-recognized meaning, which can be found by reference to standard texts, journal references and contexts known to those skilled in the art. The following definitions are provided to clarify their specific use in the context of the invention.
[0073] "Miscible" refers to a tendency of two or more liquids to mix into a single phase to form a homogeneous solution. "Immiscible" refers to a tendency of at least two liquids, when combined, to form at least two phases, such as first phase that includes a first of the at least two liquids and a second phase that includes a second of the at least two liquids. Oil and water are commonly identified as immiscible, such that mixing oil and water will result in two distinct phases when completely settled. It will be appreciated, however, that some oils are partly miscible in water and may dissolve, at least in part, in water. Similarly, water may be partly miscible in some oils and may dissolve, at least in part, in some oils. The "miscible" may be used
interchangeably herein with the term "soluble" with respect to some embodiments. The term "immiscible" may be used interchangeably with the term "non-soluble" with respect to some embodiments.
[0074] "Viscosity" refers to a property of a fluid indicative of the resistance of the fluid to flow or to deform by shear or tensile stress. In embodiments, a fluid with a higher viscosity may be observed to flow faster than a fluid with a lower viscosity. A "process viscosity" may refers to a target viscosity used in a particular process, such as a separation process. In embodiments, a process viscosity corresponds to a set point for a system in which viscosity can be controlled, such as by controlling the temperature of a liquid.
[0075] "Tube side" refers to an enclosed region within a hollow fiber corresponding to an internal region within the fiber. "Shell side" refers to a region outside of a hollow fiber. FIG. 1 illustrates an example of a tube side and a shell side of a hollow fiber. When used, in
embodiments, with respect to a hollow fiber membrane, the shell side refers to the region outside of the hollow fibers included in the hollow fiber membrane, and the tube side refers, collectively or individually, to the regions within the individual hollow fibers. In some embodiments, the tube sides of a plurality of hollow fibers in a hollow fiber membrane are arranged in fluid
communication with one another and are not in fluid communication with the shell side of the hollow fiber membrane. A hollow fiber membrane may be arranged in a membrane contactor, which may include a shell side inlet, a shell side outlet, a tube side inlet and a tube side outlet.
[0076] "Transmembrane pressure" refers to a pressure differential between a tube side and a shell side of a hollow fiber membrane. In embodiments, the pressure differential is expressed with the shell side of the hollow fiber membrane being a higher pressure than the tube side of the hollow fiber membrane.
[0077] "Wetting" refers to the ability of a liquid to maintain contact with a solid surface. In embodiments, a surface is wetted by a liquid when the liquid spreads across the surface and penetrates into at least a fraction of recessed regions, such as pores within the surface. A hydrophilic porous surface may be wetted by water, for example. A hydrophobic porous surface may be wetted by an oil, for example.
[0078] "Polymer" refers to a molecule composed of repeating structural units, referred to as monomers, connected by covalent chemical bonds or the polymerization product of one or more monomers. Polymers may be characterized by a high molecular weight, such as a molecular weight greater than 100 atomic mass units (amu), greater than 500 amu, greater than 1000 amu, greater than 10000 amu or greater than 100000 amu. In some embodiments, a polymer may be characterized by a molecular weight provided in g/mol or kg/mol, such as a molecular weight of about 200 kg/mol or about 80 kg/mol. The term polymer includes homopolymers, or polymers consisting essentially of a single repeating monomer subunit. The term polymer also includes copolymers, which are formed when two or more different types of monomers are linked in the same polymer. Copolymers may comprise two or more monomer subunits, and may include random, block, alternating, segmented, grafted, tapered and other copolymers. Useful polymers include organic polymers that may be in amorphous, semi-amorphous, crystalline or partially crystalline states. Example polymers include, but are not limited to, polypropylene, polyethylene, polytetrafluoroethylene, polyethylene terephthalate, polyvinylidene fluoride, polymethylpentene.
[0079] "Porous" refers to a characteristic of a material that includes recessed regions that may penetrate partly of fully through a body of the material. Pores in a porous material may be characterized by a cross-sectional dimension. In embodiments, polymeric materials may be porous, and may include a plurality of pores extending through at least portions of the polymer structure.
[0080] "Hydrophobic" refers to a preference for a material to be attracted to a non-polar substance as compared to water and other polar substances. [0081] This application generally relates to techniques and systems for separating oil and water in an oil and water mixture and for enhancing the separation process, such as by identifying process parameters which allow for more efficient separation of the oil and water than other process parameters. In embodiments, the process parameters used for more efficient separation in low oil concentrations are different from those used for more efficient separation in high oil concentrations.
[0082] For example, at low oil concentrations it may be more efficient to use a lower transmembrane pressure for the separation process, while at high oil concentrations it may be more efficient to use a higher transmembrane pressure for the separation process. Use of a low transmembrane pressure for the separation process at low oil concentrations may be advantageous and provide an unexpected benefit because it generally is considered more efficient to use higher transmembrane pressures to enhance transfer of oil across a porous membrane. The inventors have discovered that, at low oil concentrations, such as less than 2% oil in water by volume, lower transmembrane pressures result in a higher oil flux across a hydrophobic porous membrane, such as a porous hydrophobic hollow fiber.
[0083] As another example, at low oil concentrations it may be more efficient to use a higher oil viscosity for the separation process, while at high oil concentrations it may be more efficient to use a lower oil viscosity for the separation process. Use of a high oil viscosity for the separation process at low oil concentrations may be advantageous and provide an unexpected benefit because it generally is considered more efficient to use lower viscosities to enhance transfer of oil across a porous membrane. The inventors have discovered that, at low oil concentrations, such as less than 2% oil in water by volume, higher oil viscosity results in a higher oil flux across a hydrophobic porous membrane, such as a porous hydrophobic hollow fiber.
[0084] Thus, the present disclosure describes, in embodiments, operation of a hydrophobic porous membrane separator in two different operational regimes depending on the concentration of oil in the mixture being separated. The operational parameters at high oil concentrations, such as greater than 2%, greater than 10%, or greater than 20%, may be different from those used at low oil concentrations, such as less than 2%, less than 1%, or less than 0.5%. Further, the inventors have developed a model that allows, in various embodiments, determination of efficient operational parameters at any oil concentration.
[0085] FIG. 1 A provides a schematic illustration of a hollow fiber 100 illustrating various aspects of the fiber. For example, hollow fiber 100 includes a body 110 that separates an internal volume or tube side 120 of the fiber 100 from the external environment or shell side 130 of the fiber 100. In embodiments, the body 110 may comprise a porous polymer material that is hydrophobic. Droplets of oil on the shell side 130, for example, that come into contact with the body 110 may wet the surface and fill into pores of the polymer material. As additional oil droplets contact the wetted and filled shell side 130 surface, the oil may be transported into the tube side of the hollow fiber allowing collection of the oil.
[0086] FIG. IB provides a schematic illustration of a hollow fiber membrane 150. Here, the hollow fiber membrane 150 comprises a plurality of hollow fibers 160 arranged as a web of interspersed fibers. It will be appreciated that FIG. IB merely represents an illustration and that a higher or lower density of fibers will be present in a fiber membrane. It example embodiments, a majority of the fibers in a hollow fiber membrane have entrances to the tube side of the fibers that are proximal to one another as well as exits to the tube side of the fibers that are proximal to one another. It will further be appreciated that while FIG. IB illustrates all fibers as extending from an upper side to a lower side of the membrane, other configurations are possible. [0087] FIG. 2 provides a schematic illustration of a membrane contactor 200 useful with various embodiments described herein. FIG. 2 illustrates a cut-away of regions of the membrane contactor 200 to illustrate internal components. A collection tube 210 and distribution tube 220 are arranged in the center of membrane contactor 200. A central baffle 225 may allow for varied flow within the shell side of membrane contactor 200. A hollow fiber membrane may be wrapped around the distribution/collection tube. Individual hollow fibers 230 of the hollow fiber membrane are illustrated. Flow through the membrane contactor 200 may be achieved through four ports. A shell side inlet 240 is illustrated at the bottom of FIG. 2, which represents an inlet for a mixture of oil and water into the membrane contactor 200 to allow flow on the shell side of the hollow fiber membrane. A shell side outlet 250 is illustrated at the top of FIG. 2, which represents an outlet for a treated water that has had oil removed. A tube side inlet 280 is illustrated near the bottom of FIG. 2 and is shown as plugged for use in various embodiments described herein. It will be appreciated that tube side inlet 280 may optionally be used in an unplugged configuration as well. A tube side outlet 290 is illustrated near the top of FIG. 2 and represent an outlet for collection of oil that passes from the shell side of the hollow fiber membrane to the tube sides of the hollow fiber membrane.
[0088] FIG. 3 provides a schematic cross-sectional illustration of a membrane contactor 300 useful with various embodiments described herein. It will be appreciated that FIG. 3 is not drawn to scale but has portions of membrane contactor shown in an expanded or magnified and simplified view to better illustrate various aspects. Using a similar flow configuration as shown in FIG. 2, membrane contactor includes a shell side inlet port 310, which represents an inlet for a water and oil mixture, a shell side outlet port 320, which represents an outlet for treated water, a tube side inlet port 330 and a tube side outlet port 340. It will be appreciated that although ports 310, 320, 330 and 340 are identified as inlet and outlet ports, these notations may optionally be reversed. The shell side 350 of membrane contactor is illustrated as including a varying concentration of oil and water (higher oil concentration near the bottom of FIG. 3 and lower oil concentration near the top of FIG. 3), which represents the removal of oil from the shell side 350 of membrane contactor 300 and passage to the tube side 360 of membrane contactor 300. Porous hollow fibers 360 are illustrated as extending between the top and bottom of membrane contactor 300, with ends of porous hollow fibers 360 connected in a configuration that shows all tube side inlets connected in fluid communication with tube side inlet port 330 and all tube side outlets connected in fluid communication with tube side outlet port 340. Although not shown, tube side inlet port 330 or tube side outlet port may be in a plugged configuration. [0089] FIG. 4A provides an overview of an example method 400, in accordance with some embodiments, for recovering oil from an oil and water mixture. At block 410, a ratio of oil to water in the mixture is determined. At block 420, a transmembrane pressure and/or viscosity are determined for use in the separation process. At block 430, the mixture is contacted to the hollow fiber shell side, where oil from the mixture is allowed to pass to the tube side of the hollow fiber. At block 440, the oil is collected from the hollow fiber tube side.
[0090] FIG. 4B provides an overview 450 of details of identifying the transmembrane pressure or viscosity for use in the separation. At block 460, the ratio of oil to water in the mixture 455 is compared with a threshold ratio. If the ratio of oil to water in the mixture is greater than the threshold ratio, process flow continues to block 470. If the ratio of oil to water in the mixture is less than the threshold ratio, process flow continues to block 480. At block 470, a first
transmembrane pressure or first viscosity is identified for use in the separation process. At block 480, a second transmembrane pressure or second viscosity is identified for use in the separation process.
[0091] It will be appreciated that, in some embodiments, no comparison of the ratio of oil to water with a threshold ratio is made and that the ratio of oil to water may be used with a model to directly identify process parameters, such as transmembrane pressure, viscosity, flow rate and/or temperature to use in enhancing the transport of oil through the hollow fiber.
[0092] The invention may be further understood by the following non-limiting examples.
EXAMPLE 1 : PURE OIL CHARACTERIZATION OF A HYDROPHOBIC MICROPOROUS HOLLOW FIBER MEMBRANE CONTACTOR FOR A NEW APPLICATION IN OIL-WATER
EMULSIONS SEPARATION
[0093] Membrane technologies developed for the separation of oil-water emulsions have focused on hydrophilic membranes that can become fouled by viscous oil layers that form on the membrane wall surface. On the contrary, hydrophobic membrane technologies are promising in allowing long-term stable performance. This Example describes the re-purposing of a
commercially available hydrophobic microporous hollow fiber membrane contactor for the recovery of insoluble oil from oil-water mixtures. This new technology overcomes the limitations associated with hydrophilic membrane systems, shows no sign of progressive performance decrease and is efficient at separating oil and water for a broad range of oil feed concentrations. This Example characterizes the effect of governing parameters for the base case of pure oil feed on oil permeation across the microporous membrane wall. The experimental results demonstrate that increased transmembrane pressure and available surface area provides a linear improvement in oil permeation across the membrane wall while the viscosity has the linear inverse effect. The oil flux is determined to be independent of the feed rate. Experimental data are compared to typical models of flow through porous media and the most accurate predictive model was identified for the present membrane contactor. The results described set a foundation for the study of oil-water separation in hydrophobic hollow fiber membrane contactors and provide an upper limit of oil flux for oil-water mixtures.
[0094] Introduction. Insoluble oil and water mixtures are generated in the oil and gas, biofuel and food industries as well as in natural environments contaminated by oil spills. Technologies are needed to cost effectively and reliably separate the oil from the water and obtain high-purity phases for downstream use or reuse. Separation of oil and water is particularly critical in many oil and gas industrial applications where wastewaters such as produced water are generated onshore and offshore and oil concentrations can range from 10 ppm to 200,000 ppm depending on the source. The treatment of produced water allows the water to be reused in the oil-extraction system, revalorized for other uses such as irrigation or released into the environment. Oil recovery contributes to the economic feasibility of many treatment processes and in some instances, such as biodiesel production, oil purification is one objective of the process. Conventional technologies, such as three phase separators, hydrocyclones, flotation systems or mixed media filters, can be used for oil water separation but have lower efficacy for fine oil droplet emulsions and may not provide the needed degree of purity. Membrane systems may provide an alternative to traditional separation methods and oil-water separation mechanisms are being investigated. Membranes have the advantages of allowing high purification levels, are a rather inexpensive and flexible technology and minimize the process footprint and weight for offshore applications.
[0095] While hydrophilic membranes have been used for produced water treatment, a major drawback of the technology is the buildup of a viscous fouling layer at the membrane surface, which progressively decreases the efficiency and increases the energy -input of the overall process. Hydrophilic membranes are typically used to permeate water while oil is retained. An increase in transmembrane pressure and influent flow rate increase water flux across hydrophilic membranes but also potentially lead to lower oil rejection. Increasing temperature may increase water flux by reducing the viscosity of the fouling layer on the membrane surface. [0096] Some materials may be pretreated to allow water permeation and oil retention.
Hydrophobic membranes may be used to break oil-water emulsions by coalescence. Some direct oil-water separations with hydrophobic membranes allow oil permeation and water rejection. Oil flux may increase with increasing transmembrane pressure for hydrophobic membranes. The increase of cross flow velocity may decrease oil flux across the membrane or not affect the process, depending on feed oil concentration. Other hydrophobic systems for oil and water separations may use membranes made from experimental materials, or of a small size, or where the oil is adsorbed and not recoverable. Additional research may allow understanding the fundamental mechanisms governing the separation of oil and water with larger scale hydrophobic membrane systems.
[0097] Studies were conducted with a commercially available hydrophobic membrane contactor. The research investigated the use of this membrane contactor for the recovery of submicron algae oil from algal slurries. Manufactured by Liqui-Cel®, the contactor was originally developed for liquid-liquid extraction and is now widely used for degasification in the microelectronics industry. Some oil/water studies operated by feeding the membrane system with a solution of concentrated non-flocculated lysed algae. The contactor extracted the majority of the available nonpolar oil for most of the feedstocks. U.S. patents 8,486,267, 8,491,792, 8,617,396, and 9, 149,772 provide additional information, and are hereby incorporated by reference in their entireties. Additional oil recovery performance data was obtained using the membrane contactor over multiple days. Tested with actual applications such as the recovery of nonpolar lipids from saltwater and freshwater micro-organisms, the technology showed promising results for the separation of oil and water. Contrary to hydrophilic membrane systems' typical fouling behavior, the hydrophobic
microporous membrane contactor performance improved over time to reach high oil recovery (above 95%) with injected oil over a two-week test run. While studies with actual and simulated feedstocks are useful, the fundamental mechanisms controlling the oil recovery are complex.
These data may not provide sufficient understanding to reliably design and engineer an optimized oil recovery system that may be applied to biological and petroleum based oils. As a result, fundamental and controlled studies may be useful for evaluating the effects of transmembrane pressure, flowrate (residence time), oil concentration and oil physical properties. One useful step in investigating the separation of oil and water with the hydrophobic membrane contactor is to understand the effect of key parameters on the system under the maximized condition of pure oil feed.
[0098] A typical model to describe permeation of a single phase through membranes is the Hagen-Poiseuille equation. For hydrophobic membranes and in the absence of typical viscous membrane fouling, oil permeation mechanisms may follow the Hagen-Poiseuille law (Equation 1). Equation 1 : Hagen-Poiseuille law for streamline flow through channels:
Figure imgf000020_0001
where:
J: flux,
dp: channel diameter (average pore size),
PT: applied transmembrane pressure,
μ: viscosity,
L: channel length (wall thickness),
ε: porosity,
Figure imgf000021_0003
[0099] Other models have also been used to describe the flow of liquids through porous media such as the Ergun equation. The Blake-Kozeny viscosity component of the Ergun equation may be applied in the case of hollow fibers to correlate fiber bundle permeability to pressure drop and viscosity. The equation may be used to model flow through spherical packing columns and is useful in other applications. The Blake-Kozeny equation applied to a hollow-fiber bundle results in Equation 2. Equation 2: Blake-Kozeny equation for flow through porous media:
Figure imgf000021_0001
with A=150, dp: Equivalent particle diameter and
Figure imgf000021_0002
[0100] A may be variable with change in particle shape, particle size, and porosity for flow through porous beds and a change to using A = 180 instead of 150 may be useful. With this correction, the Kozeny-Carman equation is obtained for the case of a packed bed of spherical solids, which may be applied from membrane flux predictions. The Blake-Kozeny equation may be applied to hollow fiber bundles to refine the correlation of A with porosity and linear correlations may exist between A and ε. Such correlations allow refining of the application of the Blake-Kozeny equation to media with various porosities.
[0101] This Example characterizes the hydrophobic hollow fiber membrane contactor for oil and water separation and evaluates the potential for oil permeation without hydrocarbon fouling of the membrane surface. Characterization of the system under pure oil conditions may serve as a baseline for determining the maximum oil permeation attainable with the membrane contactor under various operating conditions. Development of a model that describes membrane behavior under various operating conditions and captures the impact of poorly quantified parameters such as porosity is also useful. Additional work may be useful to understand the fundamental mechanisms of oil/water separation with such systems. [0102] Materials and Methods. Membranes. The membranes used in this Example are manufactured by Liqui-Cel® (USA) and commercially available under the name Liqui-Cel® Extra Flow. The modules are composed of hydrophobic microporous hollow-fibers with a central baffle to eliminate shell side bypassing and evenly distribute fluid perpendicularly to the fibers. Modules with different intrinsic characteristics were investigated and are detailed in Table 1. Membranes A and B have the same module sizes, but the hollow fibers are spaced more widely for Module B to obtain a surface area half of Membrane A. Fiber types X50 and X40 differ in porosity and wall thickness. Membrane D has a different module size and hence presents longer fibers with larger surface area. [0103] All of these membranes have similar designs that include a polypropylene hollow-fiber flat sheet rolled into a cylindrical casing allowing pseudo cross-flow filtration. While a distribution of pore sizes exists, the average measured pore size was 0.04 μπι and 0.03 μπι for the X50 and X40, respectively. The region outside of the fibers is referred to as the "shell side", while the volume inside is the "tube side" (FIG. 1 A). Four ports allow the distribution of the fluid in and out of the module; two ports for each side (FIG. 2).
[0104] Table 1 : Membrane Geometry.
Figure imgf000022_0001
[0105] Reagents. Three Isopar™ oils manufactured from ExxonMobil (USA) were used to simulate hydrocarbons of produced water. The Isopar™ oils are controlled paraffinic mixtures and exist in different grades with varying viscosities as detailed in Table 2.
[0106] Table 2: Viscosities at 25 °C of Isopar™ grades used for the study.
Figure imgf000022_0002
Figure imgf000023_0001
[0107] The effect of temperature on viscosity was measured for all three grades of Isopar™ with a modular compact rheometer (Paar Physica MCR300, USA). The results presented in FIG. 5 were used to estimate oil viscosity during experiments with varying temperature conditions. [0108] Experimental System. FIG. 6 shows the schematic of the membrane system operated with pure oil. The feed was distributed through the system with a high shear pump (MTH pump, USA), and entered the membrane from the shell-side lower port. The retentate was sent back to the feed tank, while the permeate was collected in a graduated glass cylinder for quantification. The needle valve located downstream of the membrane module was used in coordination with the pump variable speed drive (Emerson, USA) to control transmembrane pressure and influent flow rate. The transmembrane pressure was calculated from the reading of two pressure gauges (Solfrunt, USA). No pressure gauge was used on the membrane tube side since it was left open to the atmosphere and oil exited the system by gravity. The influent flow rate was measured with a MicroMotion flow meter (Emerson, USA). Temperatures were obtained using a Rosemount RTD sensor (Emerson, USA).
[0109] Experimental Procedure. Characterization of the 2.5 x 8 inch diameter modules involved recirculating pure Isopar™ through the system. The transmembrane pressure and influent flow rate were adjusted and maintained for the duration of the experiment and permeated oil was collected and quantified either with a graduated cylinder and stop watch or a weigh-scale and stop watch. At first, experiments conducted with the graduated cylinder were run until 12 L had permeated across the membrane with time measurements recorded every liter. Once repeatability over time was demonstrated, experiments were conducted for 8 L of oil permeated or 15 min. Mass was measured every 1.5 min using a scale. For the larger membrane contactor, 4 x 28 inch modules, the oil permeation rate was measured with a Micro Motion flow meter (Emerson, USA) for a duration of 10 min with flow measurements recorded every minute. In both cases, transmembrane pressures were calculated as the average of the inlet and outlet pressure of the shell side. All experiments were performed in triplicate. Table 3 summarizes the conditions of all experiments conducted.
[0110] Table 3 : Experimental conditions tested for membranes A, B, and C (2.5 inch modules)
Figure imgf000023_0002
Figure imgf000024_0001
[0111] Pure oil operation of the polypropylene membrane contactor was found to present some risk. A clicking noise was detected and identified as static discharge in the system. Friction along the fibers with elevated influent flow rate and the lack of a conductive material resulted in the build-up of an electrical charge in the module, which released periodically through a spark.
Adequate grounding of the system and addition of a conductive element to the fluid is necessary for safe operation under pure oil conditions. To prevent the reoccurrence of static charging, minimal amounts of water were added to the feed. The only experiments with added water presented here were conducted with Membrane B. [0112] Results and Discussion. The results may be characterized by three different studies. First, the effect of operating parameters such as temperature, viscosity, transmembrane pressure, and influent flow rate, were studied with the 2.5 inch diameter X50 small module Membrane A. Then, the performance observed between the various module contactors (Membrane A, B, C and D) were compared to investigate the effect of intrinsic membrane characteristics such as module size, surface area, and fiber type. Finally, the data were used to derive a model describing oil permeation in the membrane system.
[0113] Effect of operating parameters on performance of the 2.5 x 8 inch membrane module with X50 fibers (Membrane A). Temperature and viscosity. Oil viscosity change with temperature variations were measured and correlated for each Isopar™ grades as shown in FIG. 5. These viscosity variations affect pure oil experiments on two levels. Using the three different grades of Isopar™ results in intrinsic different viscosities between experiments. However, within one experiment where a single Isopar™ grade is used, the viscosity was also observed to fluctuate due to temperature variations. Daily temperature changes and heat buildup due to pipe friction and pump recirculation explain variations in viscosity during a single experiment with a fixed set of conditions (fixed Isopar™ grade, transmembrane, flow rate). FIG. 7 shows the effect of viscosity on oil permeation under different experimental conditions. The data demonstrate that even within triplicates, where the same Isopar™ grade was used, viscosity varied because of ambient temperature variation. The results suggest that oil flux is linearly related to the inverse of viscosity with coefficients of correlation higher than 0.99 for all three sets of conditions. Such a trend may be identified since a fluid with higher viscosity creates a higher resistance to permeation from friction when contacting the pore walls as described by the Hagen-Poiseuille equation. The plot additionally shows viscosity variations for a single Isopar™ grade due to temperature changes. To account for temperature changes between experiments, oil flux was subsequently corrected by multiplying viscosity and expressed in m3 cP/m2 s. FIG. 7 also shows that higher transmembrane pressures improved oil permeation.
[0114] Transmembrane pressure. To further explore the impact of transmembrane pressure, oil flux was measured for various transmembrane pressures applied across the fibers of membrane A, while influent flow rate and Isopar™ grade were held constant. Each experiment was run in triplicate and showed good repeatability with a maximum coefficient of variation of 2.9%. The results shown in FIG. 8 demonstrate a linear correlation between oil flux (corrected for viscosity) and transmembrane pressure with a correlation coefficient greater than 0.99. The permeation of oil may be directly related to the force applied on the liquid contacting the membrane wall and pores. A linear correlation between transmembrane pressure and oil flux across hydrophobic membranes for a pure oil system may be derived using the Hagen-Poiseuille equation. However, the relationship between transmembrane pressure and flux may change as the mixture is converted from a continuous oil phase to a continuous water phase in the case of oil/water mixtures. Thus, the relationship shown in FIG. 8 should only be applied to systems containing pure oil.
[0115] Influent flow rate. Experiments with three influent oil flow rates were conducted to determine the effect of detention time on oil permeation across the membrane. The results presented in FIG. 9 show that regardless of the flow along the membrane walls, the oil permeated across the fibers was constant. Variations between all three sets of experiments did not exceed 2.6%. When running under pure oil conditions, the membrane is only in contact with a single phase which allows complete wetting of the hollow-fibers. Theoretically, the total membrane surface area is effective for the pure oil tests. When working with oil-water mixtures, the influent flow rate is expected to impact the oil flux across the membrane as two phases will be in contact with the fiber surface. The effective surface area will be reduced and the chance of contacting oil drops to the membrane wall may strongly depend on the volume fraction of oil in the feed and residence time in the system. Thus, when water is added to the feed, the influent flow rate may affect oil permeation by changing the oil droplets retention time in the system. [0116] Effect of Membrane Characteristics on Performance. Surface area. Membrane B utilizes a similar design (casing and fiber length) but contains half the surface area of Membrane A. In Membrane B, the fibers are twice as far apart from each other as in Membrane A. Experiments were conducted on both membranes to observe the effect of surface area on oil permeation. The results, presented in FIG. 10, show that the effects of transmembrane pressure and influent flow rate on Isopar M oil flux are consistent. Differences between the fluxes for the two membranes do not exceed 6.5% which falls within the standard deviations of the data sets.
[0117] According to these observations, the Isopar M pure oil flux may be proportional to the membrane surface area. However, the relationship between oil flux and viscosity as a function of membrane surface area was more variable (FIG. 11). The differences between the slopes of linear regression for surface areas of 0.7 m2 and 1.4 m2 were as high as 12.8%. Since results observed in FIG. 7 suggest a strong linear correlation between oil flux and viscosity at a given surface area, and data in FIG. 10 show linear dependence between oil flux and transmembrane pressure, such variability may be attributed to experimental error. For the 0.7 m2 module experiments, water at ppm levels was added to the pure oil feed after identification of static charging possibly increasing the variability of the results and decreasing oil flux as observed in FIG. 11. Therefore, the results suggest that the effective surface area may equate to the available surface area in the case of pure oil feed for membranes of the same size. Such findings may allow prediction of oil flux values when operating conditions, membrane size, and surface area are known. [0118] Module size. The membrane modules used in this study are commercially available for a range of module sizes. Pure oil experiments were conducted on both 2.5 inch-diameter x 8 inch- length and 4 inch-diameter x 28 inch-length modules. The effects of transmembrane pressure and influent flow rate on oil flux were studied for both module sizes and the results are shown in FIG. 12. The oil flux obtained for the larger module was consistently lower than those for the smaller module regardless of the experimental conditions. The difference in slopes obtained from linear regressions of oil flux versus TMP for the two module sizes is 29.0%>. That variability is higher than any observed error in these experiments suggesting that experimental error is not responsible for the differences observed. Without wishing to be bound by any theory, a possible explanation for such findings is module geometry. The larger module, not only has a higher surface area (20 m2) but also contains longer fibers compared to the smaller module. The module was placed vertically during experimentation and the permeated oil overflowed through the top port of the tube side to avoid short-circuiting. Therefore, the pressure on the tube side at the bottom of the larger module is expected to be higher than for shorter fibers due to static pressure. Differences in internal pressure between the large and small module could not be measured since it occurs inside the module and cannot be detected at the tube side port. The approximate difference in fiber lengths between the small and the large module is 53 cm, contributing to an estimated maximum static pressure of 4091 Pa at the bottom of the module. Therefore, the error due to static pressure may only contribute up to 6% in the flux. Air bubbles strongly attached to the fiber walls, called dry points, may be reducing the effective surface area during pure oil permeation. In the case of the larger modules, the length of the fibers may also contribute to a more difficult release of the enclosed air bubbles, therefore, reducing the available surface area and flux across the membrane.
[0119] The results shown suggest that when increasing the available surface area, the oil flux increases proportionally for a given membrane size. In addition, the findings suggest that fiber and, therefore, module lengths play an important role in the relationship of transmembrane pressure and oil flux.
[0120] Fiber Type. The membranes used are manufactured with two types of fibers identified as X40 and X50. The X50 fiber has a porosity of approximatively 40% and a wall thickness of 80 μπι, while the X40 has a lower porosity of approximatively 25% and a thicker wall of 100 μπι. FIG. 13 shows the comparative results between X50 and X40 fibers for various transmembrane pressures and influent flows. The oil flux measured was lower for the X40 than the X50 membranes for the same set of operating conditions. The difference between the X40 and X50 fiber performance can be compared to predictive models accounting for differences in porosity and wall thickness. [0121] Model. The results presented in FIGs. 8-13 demonstrate the linear correlation between transmembrane pressure, inverse viscosity, and oil flux as represented in Equation 3. Influent flow rate was also confirmed to have no significant impact on the process for pure oil conditions.
[0122] Equation 3 : Model describing oil permeation for X50 2.5 inch modules:
Figure imgf000027_0001
[0123] Using the data obtained with Membranes A and C, the experimental permeability constants k can be computed and results are shown in Table 4.
[0124] Table 4: Permeability constants for X50 and X40 fibers.
Figure imgf000027_0002
Figure imgf000028_0003
[0125] The Hagen-Poiseuille and Kozeny-Carman permeability constants for each fiber type were computed. The equivalent particle diameter used for the Kozeny-Carman equation was determined with the assumption that the porous membrane surface is equivalent to packed spheres with interstices being the pores. Derived from the definition of porosity, the equivalent mean particle diameter is shown in Equation 4. Equation 4: Equivalent mean particle diameter:
Figure imgf000028_0001
[0126] Adjustments to the constant A with varying bundle porosity for a porosity range of 0.4 to 0.8 may be useful. For example, Equations include Αι=542*ε-128 and Α2=497*ε-103. Table 5 presents the permeability constants for each of these models with adjusted A values for both X50 and X40 fibers. The Hagen Poiseuille model largely overestimates the experimentally determined constants. The model is usually accurate for describing flux through membranes with parallel pores. However, membrane pores often present more complex pore geometry.
[0127] Table 5: Permeability constants for X50 and X40 fibers determined with the various models.
Figure imgf000028_0002
[0128] The use of models describing flow through porous media may lead to better membrane flux predictions. The Blake-Kozeny permeability constants are closer but none of the tested models predict the change of permeability from X50 to X40 fiber correctly. The Blake-Kozeny model with A=150 provides the best fit for the X50 fibers. The parity plot is presented in FIG. 14 and shows good agreement between experimental and predicted values from the Ergun model for the X50 fiber with a coefficient of correlation greater than 0.99. [0129] The X40 fiber experimental results observed may not be well described by any of the presented models. The porosity of the X40 fibers is 0.25, which may lies below the useful range. Adjusting the constant A with various particle shape, equivalent particle size, and porosity may be useful for improving model reliability. Therefore, additional experimental work may be useful for refining the correction of the constant A in the Blake-Kozeny equation to describe flux through membrane pores.
[0130] Use of hydrophobic hollow fiber membrane contactors for oil/water separation has been investigated by this example. The technology shows significant potential for separation of oil and water. Under the baseline conditions of a pure oil feed it was possible to identify the impact of key operating parameters on oil flux. These impacts include:
• Higher viscosity was shown to linearly affect the permeation of oil across the membrane contactor. Oil viscosity is related to temperature, which can, therefore, also change the system performance.
• An increase in transmembrane pressure was observed to linearly improve oil flux across the membrane.
• Influent feed flow rate was shown to have no effect on oil permeation. The membrane fibers were in contact with a single oil phase and, therefore, oil permeation was
independent of retention time.
• A decrease in half the available membrane surface area proved to yield proportionally
lower oil flux.
• Scale up between modules with different fiber length to surface area ratios was not solely based on scaling the surface areas. The smaller module performed more efficiently.
• X50 fibers were shown to allow higher oil flux over X40 fibers and are recommended for future use in oil/water separation.
• Various predictive permeation models were compared to the measured experimental
permeability. The most accurate model for the X50 fiber permeability was the Blake- Kozeny equation with the original Ergun constant of 150 as shown in Equation 5 below. Equation 5:
Figure imgf000029_0001
• The X40 fiber permeability could not be properly predicted by the models since the viscosity may have been out of the usable range. A correction for the A constant to fit lower porosity membranes may be useful to determine. The identification of the model best describing permeability for the X50 fiber may allow predictions of the maximum oil flux and may be helpful in evaluating oil/water separation performance of the hollow fiber membrane contactor. Furthermore, the model may also be used as a performance guideline when utilizing the technology for oil purification purposes in the case of low water content in the feed.
[0131] Example Calculation
Inputs:
Membrane material:
Figure imgf000030_0002
Process:
Figure imgf000030_0003
Oil flux calculation:
Figure imgf000030_0001
[0132] Figure Captions. FIG. 2: Membrane contactor in for oil/water separation configuration. Drawing adapted from Liqui-Cel®. [0133] FIG. 5: Viscosity vs temperature for three Isopar™ grades. Legend: symbols = experimental data, dotted lines = best fit: exponential correlation. All experiments were conducted in triplicates, error bars are shown but are frequently smaller than the data symbols.
[0134] FIG. 6: Schematic of membrane system used for pure oil experiment. [0135] FIG. 7: Effect of Viscosity on Oil Permeation. Legend: symbols = experimental data, dotted lines = linear best fit. Experimental Conditions: Isopar L, M, V; Influent flow rate= 3.8 liters/min; Membrane A.
[0136] FIG. 8: Effect of transmembrane pressure on oil flux corrected for viscosity. Legend: symbols = experimental data, dotted lines = linear best fit. Experimental conditions: Isopar™ M, Influent flow rate= 3.8 liters/min; Membrane A. All experiments were conducted in triplicates; error bars are shown but are frequently smaller than the data symbols.
[0137] FIG. 9: Effect of influent flow rate on oil flux corrected for viscosity. Experimental conditions: Isopar™ M, Membrane A. All experiments were conducted in triplicates; error bars are shown but are frequently smaller than the data symbols.
[0138] FIG. 10: Comparison of module surface areas for oil flux vs transmembrane pressure. Legend: symbols = experimental data, dotted lines = linear best fit. Experimental conditions: Isopar™ M, Influent flow rate = 3.8 liters/min unless stated otherwise, Membrane A and B. All experiments were conducted in triplicates; error bars are shown but are frequently smaller than the data symbols. Except for *, single replicate.
[0139] FIG. 11 : Comparison of module surface areas for oil flux vs viscosity. Legend:
symbols = experimental data, dotted lines = linear best fit. Experimental conditions: Isopar™ L, M and V, Influent flow rate = 3.8 liters/min, Membrane A and B
[0140] FIG. 12: Comparison of module sizes for oil flux vs transmembrane pressure. Legend: symbols = experimental data, dotted lines = linear best fit. Experimental conditions: Isopar™ M, Influent flow rate = 3.8 liters/min unless stated otherwise, Membrane A and D.
[0141] FIG. 13 : Comparison of fiber type for oil flux vs transmembrane pressure. Left Panel: Influent flow rate = 3.8 L/min. Right Panel: Influent flow rate= 1.9, 3.8 and 5.7 L/min.
Experimental conditions: Isopar™ M, Membrane A and C. [0142] FIG. 14: Parity plot for the X50 fiber membrane contactor data set.
EXAMPLE 2: NEW APPLICATION OF A MICROPOROUS HYDROPHOBIC HOLLOW
FIBER MEMBRANE CONTACTOR FOR OIL-WATER SEPARATION.
[0143] Efficient oil/water separation is important in many industries. Treatment of produced water, for instance, requires technologies capable of withstanding variable feeds and removing micron size oil droplets while maintaining long term performance. In this Example, mechanisms and key process parameters controlling the separation of oil and water in a hydrophobic, hollow- fiber membrane contactor are evaluated. Two ranges of oil feed concentration are identified. For high oil feed concentrations above 40% (v/v), increases in transmembrane pressure, detention time, and temperature are shown to increase oil flux across the fiber walls, which is consistent with the results from pure oil experiments (see above). However, for dilute mixtures (less than 2% (v/v)), experimental results indicate that an increase in transmembrane pressure and temperature lowers oil flux. The design of each membrane module may allow internal coalescence of oil droplets on the fiber surface followed by permeation of insoluble oil while maintaining water rejection. The results may indicate that the stable formation of an oil film on the fiber walls is useful for the process. In addition, several experiments carried out over two-week periods showed no signs of viscous fouling or flux reduction, suggesting that stable, long-term operation is possible. [0144] Introduction. Oil/water separations are useful in processes associated with food, biochemical and chemical production as well as in the oil and gas industry. In particular, produced water treatment is driven by regulations controlling well disposal and environmental releases as well as by the financial interest in recovering oil contained in the wastewater. New technologies are needed to treat the oil and gas wastewater in an affordable and efficient way. Membrane technologies show great promise for oil removal in produced water as they are affordable, flexible, have a small footprint and can recover small oil droplets not captured by conventional
technologies. Many types of membrane systems may be useful for oil/water separation including microfiltration, ultrafiltration, nanofiltration, as well as a broad range of hydrophilic and hydrophobic materials. [0145] Membranes can be used for oil/water separation through direct separation or coalescence followed by gravity settling. Direct separation of oil and water is achieved in a membrane system through sieving and membrane material selectivity. For hydrophilic membrane surfaces, water is permeated through the pores while oil is rejected by the membrane material. For hydrophobic materials, the oil permeates and water is rejected. Several process parameters may be useful to control oil and water separation in membranes including cross-flow velocity, transmembrane pressure, temperature, and oil concentration in the feed. For example, with increasing
transmembrane pressures, water flux through hydrophilic membranes may increase. However, higher oil concentrations in the emulsion may reduce water permeation across the membrane wall because of the progressive build-up of a viscous fouling layer at the hydrophilic membrane surface. This phenomenon may limit the long-term removal efficiency and increase the energy
requirements in hydrophilic systems. A higher cross-flow velocity may enhance water flux by disturbing the viscous layer at the membrane surface. Understanding the underlying mechanisms controlling viscous fouling and enhancing membrane surfaces may be useful for improving oil removal in hydrophilic systems. [0146] Oil flux may increase through hydrophobic membranes with increasing TMP, as may be determined using the Hagen-Poiseuille model. Lower oil concentrations may decrease oil permeation because of a reduced probability for oil droplets to contact the membrane wall.
Increasing cross flow velocity may increase oil flux across the membrane. A single phase of oil may be recovered along with high effluent water quality for a concentration of l%(v/v) oil.
[0147] For hydrophilic membrane systems, when transmembrane pressure reaches a critical breakthrough value, permeation of both phases may occur for cross-flow or dead end filtrations. As a result, applying a transmembrane pressure below the critical breakthrough value is useful for obtaining a pure permeate for both hydrophobic and hydrophilic materials used for conventional membrane filtration of oil and water. However, the effect of critical pressure breakthrough and dual phase permeation may also be useful for enhancing oil/water separation through membrane coalescence. The membrane system may be operated in dead-end or cross-flow filtration mode and the oil/water mixture may be forced through the pores where oil droplets coalesce by contacting the pore wall, leading to higher oil droplet sizes and enhanced subsequent gravity separation. An increase in droplet size after permeation of a polypropylene membrane may be observed. A higher transmembrane pressure (TMP) may increase permeation rates and improve coalescence and time of settling, while an increase in temperature may allow higher permeation and better separation efficiency. Microchannel studies may be useful for simulating and observing coalescence mechanisms at pore scale and identification of important parameters. A see-through microchannel may be used to observe coalescing and phase inversion mechanisms of oil droplets in water while permeating a hydrophobic PTFE pore. The formation of larger oil droplets may be observed, which may aid the subsequent oil/water gravity separation. De-emulsification efficiency may improve with oil concentration and flow rate in the micro-channel, corresponding to operating a membrane with higher transmembrane pressure. Longer channels may also lead to longer contact times and higher coalescence efficiency. Furthermore, operating with pore sizes small enough for the oil droplets to be able to contact the membrane surface but not so small as to re- disperse the emulsion may be useful.
[0148] Various membrane configurations may be used for the separation of oil and water including flat sheet, tubular, spiral and hollow fiber. These configurations may be used with hydrophilic material for water permeation and oil rejection. Permeation of oil through
hydrophobic hollow fiber membrane contactors may be more advantageous, however.
[0149] This Example evaluates a commercially available, hydrophobic, hollow-fiber membrane contactor for the separation of oil and water. Studies investigated the use of a microporous membrane contactor for the recovery of submicron lipids from saltwater and freshwater microbial mixtures, including concentrated lysed microalgae slurries. The membrane contactor extracted the majority of the available nonpolar oil for most of the feedstocks. Backed by this success, additional mixtures were tested for up to two weeks and demonstrated an increase of performance over time to reach high level recoveries above 95% without observation of the typical viscous fouling experienced by hydrophilic systems. While these results were promising, little was known on optimization of the process performance and separation mechanisms. Without wishing to be bound by any theory, it is believed that the hollow fiber geometry of the membrane module allows internal coalescence of oil droplets between fibers combined with selective permeation of oil through the membrane. The present Example aims at identifying the effect of key process parameters and understanding the underlying mechanisms of oil/water separation in the hollow fiber membrane contactor. In particular, a broad range of oil concentrations, transmembrane pressures, influent flow rates and viscosities are investigated for controlled oil/water mixtures.
[0150] Materials and Methods. Membranes. The technology used in this Example is a membrane contactor commercially available in various sizes, manufactured by Liqui-Cel® and referred to as Extra Flow Contactors. This contactor, originally developed for liquid-liquid extraction, is commonly used for de-gassing liquid applications such as oxygen removal from water in the microelectronic industry. The membranes have a hydrophobic microporous hollow- fiber design with a central baffle and shell-side distributor (FIG. 2). The membrane contactors are available in various industrial module sizes including a 14-inch diameter version reported to process up to 550 gpm. The module design consists of a hollow-fiber sheet/membrane rolled into a cylindrical casing allowing a pseudo cross-flow filtration. Four ports allow circulation of fluids on the shell-side (outside of fibers) and the tube side (inside of fibers).
[0151] The characteristics of the three types of membrane contactors available from Liqui-Cel® and used in this Example are summarized in Table 6. Using a smaller surface area membrane contactor such as membrane module A allows identification of the limitations of the membrane contactor and the relative effect of operating parameters (i.e. it may not be possible to compare performance over a range of parameter values if complete removal of the oil is observed over the entire range). Tests performed with membrane C allow for evaluation of the effect of surface area on system performance. Membrane B utilizes the same module size as Membrane A, but the hollow-fibers are spaced wider apart from each other to obtain half the surface area. The pore size distribution of the membranes was determined independently via scanning electron microscopy (Hitachi S5500 SEM/STEM, USA) with a gold and palladium coating of the membrane sample. Image analysis was conducted with the ImageJ image processing program developed by the National Institutes of Health (USA). As can be seen in FIG. 15, a range of pore sizes was observed with a mean equal to 0.047 μιη. Two fiber types are used in the manufacturing of these membrane contactors; the X50 fiber type was selected over the X40 type to maximize oil permeation across a more porous and thinner membrane wall (see above). [0152] Oil compounds and non-stabilized emulsion. Isopar™, the synthetic oil used in this study, is a controlled mixture of paraffinic compounds manufactured by ExxonMobil. Various grades of Isopar™ oil have different viscosities as shown in Table 2.
[0153] Table 6: Characteristics of membranes used in the Example.
Figure imgf000035_0001
[0154] To generate the oil emulsion feed for the membrane contactors, oil and water were mixed with an in-line high shear pump (MTH pump, USA) to obtain micron size oil droplets in a water feed. An inline particle analyzer (JM Canty, Inc., USA) was used to observe and measure the droplet size distribution of the emulsion entering the membrane system. The majority of the oil drops were smaller than 10 μπι with an average drop size of 5.4 μπι as shown in FIG. 16. A twenty ml cloudy sample obtained from the pump discharge which typically required 12-24 hours to settle in vial.
[0155] Membrane Filtration System. Two membrane configurations were used for this study. In the first configuration, the module was operated in a "one-pass through" mode while in the second configuration the system was operated in a closed loop mode to evaluate the performance over longer term operation. FIG. 17 shows the "one pass through" configuration where oil-water emulsions were sent through the membrane system in a single pass. The oil permeated on the tube side was measured, while the retentate was discarded. FIG. 18 shows the system operated in recycling mode. Two modules were used in series at all times to guarantee high oil removal on the shell side and allow recirculation of oil-free water back to the feed tank for long-term operation. The removal of oil in the first module was used to assess the performance of the system at each operating condition, while the second module in the series was primarily used to remove the remaining oil from the water when the system was operating in recirculation mode. For practical reasons, experiments with oil concentrations higher than 10% (v/v) were operated in the "one-pass through" mode while the lower concentration experiments were conducted in recycle mode. In both the single pass and closed loop configurations, oil was injected into the main water line using a peristaltic pump (Thermo Fischer Scientific, USA) to control the influent oil concentration. The high shear pump (MTH pump, USA) mixed and circulated the emulsion in the system. Influent flow rate and pressures were adjusted with a variable speed drive (Emerson, USA) controlling the pump and a needle valve located on the outlet line. Influent flow rates were measured with Micro Motion flow meters. Temperature was monitored with a Rosemount RTD sensor (Emerson, USA). Pressures on the shell side were collected with Rosemount pressure transmitters (Emerson, USA).
[0156] Three scales (Arlyn, USA) were used to measure oil mass injected in the system and recovered by each module in series. All instruments were connected to a DeltaV data acquisition system (Emerson, USA) allowing parameter control and real-time data collection for extended periods of time without interrupting operation.
[0157] Oil-Water Experiments. Tests were performed by feeding an oil-water emulsion to the shell side inlet of the module where oil permeated the membrane and was quantified. Oil concentration in the feed was controlled by the peristaltic pump and calculated on a volumetric basis. Transmembrane pressure was computed from the average of inlet and outlet pressures of the shell side, while the tube side was left open to atmospheric pressure. Oil flux across the membrane was the indicator used to characterize system performance. Oil recoveries were also calculated from the ratio of injected to permeated oil volumes. The experimental plan is presented in Table 7. Experiments where design to test the effect of one operating parameter while the other were maintain constant. The ranges of study of all parameters were chosen within the operation specifications of the membrane contactor.
[0158] Table 7: Experimental plan
Figure imgf000036_0001
Figure imgf000037_0001
[0159] Results and discussion. Influent Oil Concentration. Experiments with oil feed concentrations ranging from 90% (v/v) to 200 ppm were conducted to study the effect of increasing water content on oil permeation. Other parameters such as influent flow rate, transmembrane pressure and viscosity were maintained at constant values. As expected, the oil flux through the membrane decreased with decreasing influent oil concentration (FIG. 19A and FIG. 19B) as the probability of oil droplets contacting and permeating the membrane walls also decreased. The maximum oil flux attainable occurs when the influent is pure oil (oil concentration is 100% in FIG. 19A). Under such conditions, a single phase is in contact with the entire membrane wall surface and the effective coalescing surface area is believed to equate to the actual membrane surface area. When adding water to the feed, the oil flux was consistently lower than for the pure oil conditions.
[0160] The pure oil experiment defines the maximum flux across the membrane for a given set of operating conditions. When water is added to the feed, the oil content flowing into the membrane can be higher than that flux under pure oil conditions. That is the case for experiments with oil concentration superior to 20% (v/v) for the TMP = 20 psi curve presented in FIG. 19 A. However, in those occurrences, the flux measured across the membrane was consistently lower than the flux observed with pure oil feed, which suggests that the effective surface area was smaller than the total surface area. The oil flux observed for experiments with lower oil content in the feed than the maximum pure oil flux was observed to decrease with decreasing oil
concentration as well. These results suggest that the effective membrane surface area decreases with increasing water content. Therefore, for higher oil concentrations where the continuous phase is oil, the curves shown in FIG. 19A approach an asymptote to the pure oil flux value indicating that the effective surface area approaches the available surface area. At lower oil concentrations, below 2% (v/v), the oil flux is linearly related to oil concentration with a coefficient of
determination higher than 0.999 (see FIG. 19B). Under such conditions, oil-water emulsions are highly dispersed and the probability for oil droplets to contact the fibers dictates oil permeation.
[0161] FIG. 19A also shows that an increase in transmembrane pressure leads to an increase in oil flux for oil concentrations above 5% (v/v). The results show that at a transmembrane pressure of 20 psi, the curve reaches a plateau above 20% (v/v) oil concentration, while the 40 psi curve consistently increases. Without wishing to be bound by any theory, when increasing
transmembrane pressure is applied, water may be forced onto the pores and may block the permeation of oil. Therefore, for lower pressure applied, the effective surface area may approach the total surface area faster than with higher transmembrane pressure as the effect of pore blocking by water is reduced. [0162] Long Term Performance. Typical hydrophilic membranes may experience fouling within minutes of exposure to oil/water mixtures. For high oil concentrations such as 60% (v/v), the flux was constant for periods as long as 20 min (FIG. 20A) for the hydrophobic membrane contactors studied in this Example. Experiments at lower oil concentrations were also conducted for extended periods of time and the oil recovery was observed to actually improve over time for a period of up to two weeks (FIG. 20B and FIG. 20C).
[0163] Hydrophobic membranes may be used for coalescence of oil/water emulsions where oil droplets coalesce by contacting the hydrophobic pore material. In these cases, a gravity settler may be ultimately used to separate the two phases. The particular geometry and material of the modules used in the present Example may allow coalescence to take place on the membrane fibers leading to the formation of an oil film coating the membrane wall. The transmembrane pressure applied across the membrane wall subsequently drives the selective oil transfer through the porous wall. The effective membrane surface area may be linked to the actual surface area of the oil film present on the fibers. Oil droplets may contact and coalesce with the oil layer rather than the non- wetted membrane suggesting that cohesion forces are greater than the adhesion forces. For higher oil concentrations, the creation of the film may occur easily and rather instantaneously, but for more dilute mixtures, the growth and stability of the oil film on the fibers may be critical and dependent on oil concentration in the feed. This hypothesis is supported by the increase in oil recovery over time for dilute oil/water mixtures possibly due to the progressive oil film growth. Results shown in FIGs. 20B-20C also demonstrate that oil recoveries up to approximatively 100% are attainable if enough membrane surface area is provided and for extended periods of time.
[0164] Transmembrane pressure. After identifying two relevant concentration ranges that are governed by two distinct separation mechanisms, subsequent experiments were conducted at higher (Coil = 40% (v/v)) and lower (Coil = 2% (v/v)) concentrations to investigate the effect of transmembrane pressure on oil flux. At an oil concentration of 40% (v/v), the transmembrane pressure was found to be linearly related to oil flux (FIG 21 A). This result is consistent with previous experiments with pure oil feeds where increasing transmembrane pressure led to higher permeation rate. [0165] However at a lower oil concentration of 2% (v/v), the effect of transmembrane pressure on oil flux surprisingly reversed (FIG. 2 IB) and a lower transmembrane pressure yielded higher oil permeation. While this result may be counterintuitive, it is consistent with the oil permeation mechanisms proposed in this Example. Higher transmembrane pressure leads to higher compression force of the fluid onto the membrane wall. At higher transmembrane pressure, water is likely to be forced into the membrane pore entrance and prevent contact and permeation of oil through the fibers. As shown by FIG. 15, the pore size distribution of the fibers is wide and larger pores are expected to be more easily blocked by water with increasing pressures, thereby reducing the effective surface area of the membrane. This phenomenon can be explained by lower water breakthrough pressures for larger pore size calculated from the Young Laplace equation. In addition, the stability of the oil film on the fibers may be useful for oil permeation. At higher transmembrane pressure, oil may be forced through the pores at a rate preventing the oil film from staying and growing on the fibers, permanently reducing the effective surface area. In contrast, operating with lower transmembrane pressure may lead to the formation and growth of a stable oil layer leading to consistently higher oil permeation through the pores as seen in FIG. 2 IB. [0166] Influent flow rate. The effect of influent flow rate on oil permeation was studied and the results are shown in FIG. 22. For Coil = 40% (v/v), a 67% decrease in influent flow rate led to a 174%) increase in oil recovery. For Coil = 2% (v/v), a 67% decrease in influent flow rate led to a 37%) and 5% increase in oil recovery for transmembrane pressures of 40 psi and 20 psi respectively. Therefore, results suggest that influent flow rate is a useful control parameter for oil recovery at higher oil concentration and higher transmembrane pressures. For both oil
concentrations tested, lower influent flow rates improved oil recovery which may be explained by an increase in retention time for oil droplets in the membrane module. In the case of pure oil feed, the detention time did not impact the permeation of oil as the total surface area of the model was wetted with oil (see above). In the case of oil/water mixtures, longer retention times increases the chances for oil droplets to contact and coalesce on the membrane wall and be permeated across the fiber walls. At higher oil concentration, the fibers are assumed to be rapidly oil wet. Therefore, permeation is strongly dependent on retention time. At lower oil concentrations, oil deposition and formation of an oil film on the membrane wall seems to be the key component for oil permeation. At lower transmembrane pressure, the oil film is more stable leading to a higher effective membrane surface area. The affinity of oil to the membrane wall is stronger at lower
transmembrane pressure, possibly reducing the effect of retention time for oil permeation.
[0167] Oil Viscosity. Oil viscosity is expected to influence oil flux across membranes and typically the higher the oil viscosity, the lower the oil flux is expected to be. Tests with various Isopar™ grades were conducted with the membrane contactor to evaluate the effect of oil viscosity on the process. Under pure oil conditions, oil flux was shown to be inversely related to viscosity (see above); increased viscosity, led to reduced flux. At 80% (v/v) oil concentration, the flux seen in FIG. 23 A decreased with higher viscosity but not linearly as observed in the pure oil
experiments (see above ). At 2% (v/v) oil concentration, results presented in FIG. 23B, confirm that increasing viscosity leads to decreased oil flux. However, the magnitude of the viscosity effect decreased at higher transmembrane pressures. The effect of viscosity on flux inverts for more dilute oil/water mixtures (e.g. 200 ppm water). As seen in FIG. 23C and FIG. 24, increasing viscosity increases oil flux possibly enhancing the stability of the oil film on the fibers. Two mechanisms may be competing for the permeation of oil in the present system. When enough oil is supplied at the pore entrance (for high oil concentrations), permeation is improved by
transmembrane pressure and affected by viscosity. However, when the oil content is low, there is less oil available to cover the membrane surface and rapid permeation of oil may hinder long-term flux by reducing the long-term effective area. It may be possible to maintain a stable film at the surface of the fiber at higher viscosity and lower transmembrane pressure. The stability of the film may provide greater effective surface area and higher oil flux even though oil is transported at a lower rate across the fiber. Interestingly, at 2% (v/v) oil concentration and high viscosity, changes in transmembrane pressure did not affect oil flux significantly. This observation suggests that the predominant mechanism contributing to oil film stability was the viscosity and not the transmembrane pressure under these conditions.
[0168] Fiber spacing. The performance of two membrane contactors with different fiber spacing was compared and results are shown in FIG. 25 A and FIG. 25B. [0169] Not surprisingly, oil recoveries were found to be higher with higher available surface area. However, the differences in oil flux which is normalized by surface area in this FIGs. 25A- 25B, indicates that the smaller surface area membrane contactor was significantly more efficient in this case. The two membrane contactors differ with respect to the spacing between the fibers, which ranges from approximatively 40 μm for the 1.4 m2 module to approximatively 80 μm for the 0.7 m2 module. The results suggest that wider spacing leads to higher oil flux. Coalescence may occur when oil droplets are small enough for the droplets to have a chance to contact the pore but not so small that the emulsion is re-dispersed. The space between fibers of the studied module could be thought of as a long pore where droplets coalesce between the fibers. The present results suggest that an increase of space between fibers from 40 to 80 μm is beneficial for the coalescence of the emulsion used in this work that had an average oil droplet size of 5.4 μm based on FIG. 16. This result does may contradict research conducted with coalescence in membrane pores. For example, if the pore is much larger than the oil droplet size, the drops may flow between the walls without coalescing. However, the difference with conventional coalescence in membranes and the present coalescence mechanism may be the length of the coalescing channels. Membrane pores may be only as long as the wall thickness, usually a few hundred micrometers. In the present case, the coalescing channels are the entire length of the module of approximatively 20 cm increasing the chance for the oil droplets to contact the fibers wall. The increase of space between fibers may improve water flow distribution out of the coalescing area leading to better coalescence and stability of the film on the fiber. The 0.7 m2 module yields 73% higher oil flux than the 1.4 m2 module for low transmembrane pressure (20 psi). That increase is 45% higher at higher transmembrane pressure (40 psi). At higher pressure, the water may start blocking the larger pores, therefore, if water flows out more easily from the coalescing area, more efficient use of the effective surface area is made. At lower pressure, the water is even less compressed against the fiber wall leading to improved coalescence and permeation. A schematic detailing the proposed mechanism is shown in FIG. 26. Such results seem to confirm that oil film stability on the fibers is critical for the establishment of the effective surface area of the membrane contactor.
[0170] In. FIG. 26, oil droplets 2610 are illustrated between porous hollow fibers 2620. Partial surface layers 2630 comprising oil are illustrated as present on the surface of the hollow fibers. As the oil droplets contact the partial surface layers 2630, the oil may coalesce and pass through the pores of the fiber 2620. Partial surface layers 2640 are illustrated where oil is not present and so pores in the hollow fibers may be blocked (e.g., by water) and transport of oil through or past the partial surface layers (black arrows in FIG. 26) may not be possible. [0171] Comparison to an equivalent hydrophilic system. Comparing the performance of hydrophilic and hydrophobic membrane systems can be challenging since the permeated phases across the membranes are different. Experiments where oil recoveries approach 100% (such as results presented in FIG. 20C) with the membrane contactors can be compared to an equivalent hydrophilic system depicted in FIG. 27. If 100% oil recovery is assumed with the hydrophobic membrane, an equivalent system would consist of a hydrophilic membrane permeating pure water. Therefore, for a 200 ppm oil concentration experiment at transmembrane pressure of 20 psi and influent flow rate of 3.78 L/min with a membrane contactor of 1.4 m2, an oil flux of 8.7x10-9 m3/s m2 was measured. The equivalent water flux for a hydrophilic membrane would correspond to permeate the entirety of the water phase through 1.4 m2 surface area, which equals a flux of 4.5x10-5 m3/s m2. Moreover, the hydrophobic system studied here was shown to maintain such performance up to two weeks (FIG. 20C), which may not be the case of the equivalent hydrophilic system that may experience decrease of flux over-time due to viscous fouling. Hollow-fiber membranes may be used for oil/water separation with water permeation across a modified PVDF membrane surface. Higher water flux may be expected than for typical PVDF membranes. The highest flux attained under a transmembrane pressure of 50 psi and oil feed concentration of 500 ppm may be around 1.9 x 10-5 m3/s m2 prior to the observed decrease in performance due to fouling. Therefore, the results found here are comparable to the results for a hydrophilic system at start-up without the disadvantage of viscous fouling occurring over time, that would ultimately reduce the permeation flux. [0172] Therefore, the present hydrophobic membrane contactor represents a viable competitor to hydrophilic systems as comparable oil removal capability is obtained but also maintained over time.
[0173] Conclusions. The research presented in this Example examined the use of a hydrophobic hollow fiber membrane contactor for the removal of insoluble oil from controlled oil/water mixtures. Various operating parameters, including transmembrane pressure, influent flow rate, viscosity, and oil concentration, were investigated to reveal the underlying mechanisms presiding over the successful separation of oil and water. Significant findings from the study include: • Decreasing oil concentration was shown to decrease flux across membrane fibers, even in instances where the oil content was below the maximum amount of oil recovered under pure oil conditions. Experiments conducted over a range of conditions demonstrated that water addition decreased the effective surface area of the membrane contactor.
• Two concentration regions were identified that helped isolate the two fundamental
processes dictating the separation of oil and water. At low oil concentration (below 2% (v/v)), the stability of an oil film on the membrane contactor appeared as the limiting mechanism, while at higher oil concentrations, the system performance was in agreement with typical membrane separation behaviors.
• No viscous fouling or performance decrease was observed for up to two weeks of operation with the membrane contactors. High oil recoveries were attained when sufficient membrane area was used for a given oil feed concentration.
• Transmembrane pressure was shown to have different effects depending on the oil feed concentration range. For higher concentrations, increasing transmembrane pressure was confirmed to increase oil permeation. However, for oil concentrations below 2% (v/v), operating the system with lower transmembrane pressure was shown to be more beneficial by improving the oil film stability of the membrane wall.
• Increasing influent flow rate, which yields a lower detention time, was shown to decrease oil recoveries. This effect was less important for lower oil concentrations and
transmembrane pressure, which further supports the hypothesis that the development and stability of oil films on the membrane fibers are advantageous in optimizing the performance of the contactor.
• Analogous to the transmembrane pressure, viscosity was shown to impact the system
differently depending on the influent oil concentration range. The typical effect of decreasing permeation with increasing viscosity was observed for most of the oil concentration range. However, for very dilute mixtures (200 ppm), higher oil viscosity contributed to oil film stability and oil permeation.
• Wider-spaced fibers were shown to improve oil flow per unit surface area by possibly improving water flow distribution out of the coalescing areas of the hollow fibers.
• The data presented in this Example may confirm the hypothesized mechanism that oil droplets coalescing on the fibers and progressively coating the membrane wall at low oil concentrations is advantageous for successful operation. The stability of the oil film on the fibers is a useful mechanism at low oil concentration and leads to progressive improvement of performance over time.
• The hydrophobic membrane contactor can withstand and perform efficiently for a wide range of oil concentrations for which fouling in hydrophilic systems would be prohibitive, which makes this technology a prime candidate for many applications ranging from oil removal in very dilute oil/water mixtures to oil purification processes.
• These studies are based on systems possessing a high interfacial tension (-50 dynes/cm) and without solids so some caution should be applied. Additional work may address the comparative effects of lower interfacial tensions and solids. [0174] Figure Captions. FIG. 2: Extra flow Liqui-cel® membrane contactor in for oil/water separation configuration. Drawing adapted from Liqui-Cel®.
[0175] FIG. 15: Pore size analysis of membrane pores. Left panel: Scanning Electron
Micrograph image. Right panel: Pore size distribution obtained with ImageJ analysis.
[0176] FIG. 16: Droplet size distribution of Isopar™-water non-stabilized emulsion entering the system.
[0177] FIG. 17: "One pass through" system schematic. [0178] FIG. 18: System schematic in recycling mode.
[0179] FIGs. 19A-19B: Effect of influent oil concentration on oil flux across the membrane. FIG. 19 A: Oil flux as a function of oil concentration for transmembrane pressures = 20 psi and 40 psi. FIG. 19B: Results at lower oil concentrations, Transmembrane pressure = 20 psi.
Experimental conditions: Isopar M; Influent flow rate = 3.78 L/min; Membrane A.
[0180] FIGs. 20A-20C: Oil recovery vs time. Experimental conditions: Isopar M; Influent flow rate = 3.78 L/min; Transmembrane pressure = 20 psi. FIG. 20A: Coil = 60% (v/v), Membrane A, SA = 1.4 m2. FIG. 20B: Coil = 1000 ppm, Membrane C, SA= 20 m2. FIG. 20C: Coil = 200 ppm; Membrane A, SA= 1.4 m2.
[0181] FIGs. 21A-21B: Effect of transmembrane pressure on oil flux. FIG. 21 A: Coil = 40% (v/v). FIG. 21B: Coil = 2% (v/v). Experimental conditions: Isopar M; Influent flow rate = 3.78 L/min; Membrane A.
[0182] FIG. 22: Effect on influent flow rate on oil recovery. Experimental conditions: Isopar M; Coii = 2% and 40% (v/v); TMP = 20 psi and 40 psi; Membrane A. [0183] FIGs. 23A-23C: Effect of viscosity on oil flux. FIG. 23A: Coil = 80% (v/v), TMP = 20 psi, Membrane A. FIG. 23B: Coil = 2% (v/v), TMP = 20 psi and 40 psi, Membrane B. FIG. 23 C: Coii = 0.02% (v/v) = 200 ppm, TMP = 20 psi, Membrane A. Experimental conditions: Influent flow rate = 3.78 L/min. [0184] FIG. 24: Effect of viscosity on oil recovery. Experimental conditions: Influent flow rate = 3.78 L/min; TMP = 20 psi; Membrane A, Isopar V used initially, then switched to Isopar M, then back to Isopar V.
[0185] FIG. 25 A: Effect of surface area on oil flux. FIG. 25B: Effect of surface area on recovery. Experimental conditions: Isopar M; Coil = 2% (v/v); Influent flow rate = 3.78 L/min; Membrane A and B .
[0186] FIG. 26: Schematic of oil droplet coalescence between fibers for two transmembrane pressures.
[0187] FIG. 27: Schematic for equivalent hydrophilic system
EXAMPLE 3 : MODELLING OF A MICROPOROUS HOLLOW FIBER MEMBRANE
CONTACTOR FOR OIL/WATER SEPARATION
[0188] The separation of oil and water presents significant technical and economic challenges in many industries. In particular, technologies capable of separating insoluble oil and water are needed for oil and gas wastewater treatment, biofuel production, petrochemical processing, and in food manufacturing applications. Reliable and economical technologies with high oil removal capabilities over long term operation are required. Hydrophobic membrane systems are a promising technology for the recovery of pure oil as well as the production of high purity water streams from oil/water mixtures without the typical fouling problems observed with hydrophilic materials. In this Example, the successful use of a hydrophobic microporous hollow fiber membrane contactor for the separation of insoluble oil from water is detailed. Oil flux
performance is shown to be directly related to the effective surface area of the membrane contactor. A semi-empirical model is developed and relates the effective surface area to the operating parameters of the system such as transmembrane pressure, influent flow rate, oil viscosity, as well as to membrane characteristics. The model predicts the required surface area for a chosen oil recovery under a given set of operating conditions. For instance, an oil/water influent containing 0.1% (v/v) of oil with a viscosity of 4 cP and entering the system at a rate of 240 1/hr would require 8 m2 of membrane surface area to reach 95% oil removal if operated under a transmembrane pressure of 1.5 bar. The model can be used to predict oil removal for a given surface area as well and, therefore, provides a valuable tool for design of a separation system with the membrane contactor.
[0189] Introduction. Many technologies have been developed to separate oil from water in oily wastewaters. Conventional treatment techniques such as hydrocyclones, gravity separators, gas flotation and centrifuges are good solutions for many applications in the oil and gas, food or metallurgical industries. However, high capital and operating costs along with poor removal efficiencies for micron size oil drops have limited the application of these technologies. As a result, membrane systems have gained attention for their ability to efficiently capture smaller oil droplets (< 10 μm) and achieve the high oil removals required for subsequent disposal or reuse of water streams. Many types of membranes may be useful including microfiltration, ultrafiltration, nanofiltration, as well as different hydrophilic, and hydrophobic materials. Hydrophilic membranes have proven useful for producing high quality water permeate. However, a progressive decrease in performance due to viscous fouling of the membrane surface is generally observed. Hydrophobic systems that permeate oil rather than water present great promise for preventing the typical fouling of the membrane surface by a viscous oil layer observed with hydrophilic systems. Understanding the mechanisms at play and characterizing the performance in larger scale hydrophobic systems may be useful.
[0190] Studies investigated the use of hydrophobic microporous membrane contactors for the recovery of submicron algae oil from algal slurries and showed useful results in insoluble oil and water separation. The membrane contactor extracted the majority of the available nonpolar oil for most of the feedstocks. Originally developed for liquid-liquid extraction, the membrane contactor has significant commercial applications associated with de-gassing of liquids such as the oxygen removal from water in the microelectronic industry. Additional oil recovery performance data was obtained using the membrane contactor over multiple days. In contrast to the fouling behavior observed in hydrophilic membrane systems, the hydrophobic microporous membrane contactor performance improved over time to achieve high oil recovery (>95%) with injected oil over a two- week test run. While the studies with actual and simulated feedstocks are useful, the fundamental mechanisms controlling the oil recovery are complex. These data may not provide sufficient understanding to reliably design and engineer an optimized oil recovery system that may be applied to biological and petroleum based oils. As a result, fundamental and controlled studies may be useful for investigating a module of limited surface area to evaluate the effects of transmembrane pressure, influent flowrate (residence time), oil concentration and oil physical properties (see above). The findings allow the development of a semi-empirical model to predict oil flux and recovery across the membrane contactor for long term steady state operation. [0191] Modeling of membrane systems has focused on characterizing time dependent fouling behaviors in hydrophilic systems and defining breakthrough pressures. A mechanistic model may be useful for describing the limiting permeation of water in a highly hydrophilic UF membrane system. Fouling behavior follows Hermia's models and water flux may become pressure independent over time. Viscosity, oil concentration and shear rate may be useful parameters affecting water flux across the hydrophilic membranes. Experimental data collected from hydrophilic systems may be compared to the transport models used for microfiltration such as Brownian diffusion, shear-induced diffusion or inertial lift. For example, a combination of Brownian and shear-induced diffusion in a model may obtain good agreement between predicted and measured steady state flux data in a hydrophilic microfiltration system treating industrial oily wastewater. To develop models for hydrophilic microfiltration, an analogy may be used between the solid particles in a liquid phase and the oil droplets in the water phase. Oil droplets are analogous to solid particles that progressively block the pores or form a layer at the membrane surface leading to a reduced steady state liquid flux across the membrane. However, in the case of hydrophobic systems and for concentrations where the continuous phase is water, the analogy does not hold since the "particles" (i.e., oil droplets) are actually permeated and not sieved as in conventional microfiltration. Permeation behavior in hydrophobic microfiltration of oil/water mixtures may differ from the typical trends observed in hydrophilic systems. In particular, transmembrane pressure and viscosity may have opposite effects when the oil concentration is varied from high to low in the oil/water mixtures. Use of hydrophobic systems for the separation of oil and water may be useful in identifying operating parameters involved in the filtration system.
[0192] This Example presents a semi-mechanistic, semi-empirical performance model for the design and operation of a microporous hydrophobic membrane contactor for the recovery of insoluble oil from oil/water mixtures. The technology may be applicable for applications ranging from oil purification to the treatment of very dilute oil/water mixtures such as would be typical of the final oil removal step in produced water treatment. The model is useful for determining membrane surface area and operating conditions needed to optimize process performance for a given oil/water stream.
[0193] Materials and Methods. Experimental data. Membrane contactor. The membrane contactor is commercially available (Extra Flow Contactor) and manufactured by Liqui-Cel® (a 3M Product). Various module sizes exist. The largest modules are reported to process up to 550 gallons per minute of liquid in degassing applications. The membrane contactors comprise hydrophobic microporous hollow-fibers packed in a cylindrical module (FIG. 2). [0194] The geometric characteristics of the membrane contactors selected for this study are presented in Table 8. While modules containing two different available hollow-fiber
characteristics (X50 and X40) were initially evaluated, the contactors containing X50 fibers were selected because these modules were found to produce higher oil permeation rates while preventing breakthrough of water (see above). The 2.5 inch diameter module is the smallest commercially available module and is useful for observing changes in performance as a function of operating conditions because it is possible to achieve less than 100% removal for experimentally reasonable operating conditions. The module's sensitivity allowed for an improved understanding of the oil/water separation and identifying optimum operating conditions. [0195] Table 8: Membrane characteristics for the 2.5 inch diameter X50 Liqui-Cel® Extra Flow Contactor selected for the study.
Figure imgf000048_0001
[0196] Synthetic Oil. Various Isopar™ grades were used as the synthetic oil for this study. Isopar™ oils are controlled mixtures of paraffinic compounds and manufactured by ExxonMobil (purchased from Nexeo Solutions, USA). Each Isopar™ grade has a different viscosity as shown in Table 2.
[0197] Pure oil experiments. Experiments were conducted with pure oil feeds under various experimental conditions detailed in Table 9 to determine the membrane contactor's baseline performance (see above). The effect of transmembrane pressure, oil viscosity, and membrane characteristics on oil flux was observed and successfully compared to the Ergun equation.
[0198] Table 9: Experimental conditions for pure oil feed experiments with the membrane contactor.
Figure imgf000048_0002
[0199] Oil-Water Experiments. Experiments were conducted during which the separation of oil/water mixtures were monitored and assessed under various operating conditions presented in Table 10.
[0200] The oil and water emulsions were created with an in-line high shear regenerative turbine pump (MTH pump, USA) to obtain micron size oil droplets in the water feed. Chemical emulsifiers were not used in this study. To monitor the droplet size distribution of the oil/water mixtures, an inline particle analyzer (JM Canty, Inc., USA) was used downstream of the high shear pump. The average droplet size was 5.4 μιη and a distribution of pore sizes was observed where the vast majority of droplets were smaller than 10 μιη. A twenty ml cloudy sample obtained from the pump discharge which typically required 12-24 hours to settle in vial.
[0201] Table 10: Experimental conditions for oil/water experiments with the membrane contactor.
[0202] Membrane Filtration system. Two experimental flow path configurations were used in this study: a "one-pass through" and a "recycling mode". The "one pass through" configuration consisted of feeding the oil-water emulsion to the membrane system in a single pass (FIG. 17).
The product water was collected in a second tank. Oil permeation was measured on the tube side and the aqueous retentate was discarded. In the recycle mode configuration, a 2.5 inch diameter module was mounted in series with a 4 inch diameter "guard" module used to provide complete oil removal on the shell side and allow recirculation of oil-free water back to the feed tank (FIG. 18). For both configurations, oil and water were emulsified and circulated in the system with the high shear regenerative turbine pump (MTH pump, USA). A peristaltic pump (Thermo Fischer Scientific, USA), connected to the suction of the high shear pump, provided control of the oil injection rate. The pump's variable speed drive and a needle valve located on the outlet line were used to control influent flow rates and pressures. Influent flow rates were measured with mass flow meters (MicroMotion, USA). Temperatures were monitored with Rosemount RTD sensors (Emerson, USA). Pressures on the shell side were measured with Rosemount pressure transmitters (Emerson, USA). Mass quantities of oil injected to the system and permeating the membranes were monitored with three weigh scales (Arlyn, USA). A DeltaV data acquisition system
(Emerson, USA) was used to continuously acquire instrument measurements.
[0203] Tests were performed by circulating an oil-water mixture though the membrane contactor from the bottom to the top shell-side ports. Influent oil concentration was computed as a volume fraction. Permeated oil was collected and quantified with the scale from the top tube side port. Transmembrane pressure was calculated as the average pressure at the inlet and outlet ports, while the tube side was left open to atmospheric pressure. The performance of the system was evaluated based on oil flux and oil recovery. [0204] Oil flux and surface area model development. Effective surface area. In this study, a model describing oil flux across a hydrophobic membrane is described. The model compares to an approach used in packed column absorption. As shown in Equation 6, the oil transferred across the membrane walls is the product of an intrinsic oil flux (Oil fluxpure) determined during the pure oil feed experiments and the effective surface area of the membrane. The effective membrane area for coalescence may be significantly less than the actual membrane area, especially at high water concentrations. The high concentrations of water may enter pores or cause the membrane surface to become hydrophilic which results in effectively blocking transport of the oil. Equation 6:
Figure imgf000050_0002
[0205] Equation 6 can be re-written to express oil flux across the membrane in terms of the intrinsic flux and the effective area ratio defined as the effective area over the total area of the membrane (ae/ap) (Equation 7). Equation 7:
Figure imgf000050_0001
where:
Oil FIUXOAV: Oil flux for oil/water mixtures in m3/s m2
Oil Fluxpure oii: Oil flux for pure oil feed conditions in m3/s m2
ae: Effective surface area in m2
at: Total surface area in m2
ae/at: Effective area ratio.
[0206] The membrane contactor performance depends on the intrinsic flux (Oil fluxPure) and the effective area ratio, which depends on the entering oil rate. The first step in developing the model was to characterize the intrinsic flux using membrane characteristics and operating parameters. Pure oil feed experiments allowed quantification of the intrinsic flux of the system where the total membrane surface is assumed to be available and maximized oil permeation is observed. Then, assuming the effective area ratio equals the ratio of measured flux for oil/water experiments to pure oil feed flux, a model was developed to predict the effective area ratio as a function of the operating parameters. Finally, a complete equation for oil flux prediction as a function of transmembrane pressure, viscosity, oil concentration, influent flow rates and membrane characteristics can be obtained and used to predict surface area requirements as a function of operating conditions. An example calculation can be found at the end of this Example.
[0207] Pure oil conditions. The model describing the experimental data most accurately for pure oil feed filtration with the X50 membrane contactor is shown in Equation 8 (discussed previously above). The model is the Ergun equation, which may be useful for describing flow through packed columns. Equation 8 :
Figure imgf000051_0001
where:
oil flux for pure oil feed experiments in m3/s m2,
Figure imgf000051_0003
dp: Equivalent particle diameter (m):
Figure imgf000051_0002
TMP: Applied transmembrane pressure (Pa),
μ: Viscosity (Pa s),
L: Membrane wall thickness (m),
ε: porosity.
[0208] Oil/water mixtures. The effective surface area ratio was computed for the set of oil/water experiments described in Table 10 using Equation 7. The effects of transmembrane pressure, influent flow rate, viscosity, and oil concentration on effective area ratio were first analyzed. The trends observed for each parameter were then used to develop a model for the effective area ratio.
[0209] Effect of oil concentration. The experimental curve of effective area ratio plotted versus oil concentration seen in FIG. 28 displays saturation behavior seen in other applications such as the Michaelis-Menten kinetics equation or the Langmuir adsorption isotherm. The membrane pores, which are directly related to oil permeation, can be considered as available sites, while the driving force for accumulation onto the membrane wall is the oil concentration. At low oil concentrations, the effective area ratio is linearly related. Oil concentration linearly increases the probability for the oil droplets to contact the membrane wall. For higher oil concentrations, the curve reaches saturation where the effective surface area approaches the total surface area available for oil permeation. Therefore, a saturation model as presented in Equation 9 should provide a good fit to correlate effective area ratio to oil concentration (assuming that the ratio of flux to the pure oil feed flux is equal to the effective area ratio). Equation 9: with a and b constants.
Figure imgf000052_0001
[0210] Effect of other operating parameters. The effect of influent flow rate on oil recovery in the system for feed containing oil/water mixtures was attributed to changes in residence time of the oil droplets in the membrane module (see above). At higher feed oil content and higher influent flow rates, the effective surface area is shown to increase in FIG. 29. This may be comparable to the effective mass transfer area increasing with liquid rate in packed column absorption systems. The effect of influent flow is less dramatic at higher oil concentrations that approach the ideal case of pure oil in the feed. Indeed, under pure oil conditions, the influent flow rate was determined to have no impact on the oil flux (see above).
[0211] The effect of transmembrane pressure on oil permeation and recovery was investigated for oil concentrations ranging from 200 ppm to 90% (v/v) in the feed (see above). For high oil concentrations, an increase in transmembrane pressure was found to increase the oil flux across the fiber wall (see above). This phenomena is expected in membrane filtration systems (both for hydrophobic and hydrophilic membranes). For low oil concentrations (2% (v/v)), however, an increase in transmembrane pressure was surprisingly observed to decrease the oil flux across the fiber wall, which is likely a result of water entering and blocking the pore. Breakthrough of water will eventually occur if a higher transmembrane pressure is applied. This is consistent with the Young-Laplace relationship between pressure and the curvature of a water droplet at the entrance of a pore, which has a contact angle greater than 90 degrees. Larger pores may have smaller breakthrough pressures than smaller pores. While the critical pressure for the average pore size is well above relevant transmembrane pressure values for this system, there are pores within the distribution that are large enough to be at least partially filled (blocked) by water. Thus, the effective surface area could be affected by this phenomenon without water breakthrough. The effective membrane surface area calculated from Equation 6 above decreases with increasing transmembrane pressure with water in the feed (FIG. 30). The decrease is more drastic for dilute oil/water mixtures for which more pores are exposed to water and subject to blocking. Therefore, there is a trade-off between forcing more oil through the pores with higher transmembrane pressure and reducing the useful surface area of the membrane. It should be noted that the effect of interfacial tension, which may affect water breakthrough, was not studied.
[0212] Similarly, oil viscosity was shown to impact oil permeation in the membrane contactor depending on the feed oil concentration (see above). However, the effective surface area seems to improve with increasing viscosity. Again, two mechanisms are balancing each other and affect the system differently depending on the oil concentration in the feed. Increasing viscosity appears to help grow and stabilize the oil film on the membrane surface and thereby increase the effective membrane surface area. However, for high oil concentrations, transport mechanisms are strongly affected by viscosity as demonstrated by the Ergun equation. For very dilute mixtures, stability of the oil film seems to be the main contributor to higher oil flux.
[0213] Effective area ratio model and characterization. The data gathered and presented in above represented a total of 58 experiments obtained with the 2.5 inch diameter X50 membrane contactor. Table 1 1 summarizes the range of the parameters studied. MATLAB was used to fit the experimental data to the effective area ratio model presented in Equation 10. The nlinfit function was used to conduct a nonlinear regression for the over constrained system of 58 nonlinear equations. The results were obtained by iterative least squares estimation. [0214] Table 1 1 : Summary of the range of operating parameters used to develop the model
Figure imgf000053_0002
[0215] Equation 10:
Figure imgf000053_0001
where:
Coil: oil volume fraction,
TMP: Transmembrane pressure in Pa,
μ: Viscosity in Pa s,
Q: Influent flow rate in m3/s.
[0216] Transmembrane pressure, influent flow rate and viscosity were integrated into the model so that with changing oil concentration, the varying effect of each parameter was taken into consideration. The model was initially fit based on the lower oil feed concentration data (< 2%) to allow better accuracy in this concentration range. This approach resulted in a slightly
underestimated fit at higher concentrations. The attempt to fit all data at once led to overestimated values for low oil concentrations, which is undesirable for engineering design purposes. A conservative fit in this range provides for a margin of safety for design purposes. For this reason, the first approach was selected to obtain a better fit at lower concentrations.
[0217] The residual plot seen in FIG. 32 shows the desired random scatter of data points with a slight unbalance to the left due to a higher number of experiments conducted for low oil concentrations. In addition, the model was first fit with the lower oil concentration range leading to reduced residuals on the left hand side of the plot.
[0218] The plot of experimental effective area ratio versus estimated values shown in FIG. 33 shows a reliable linear correlation. Most of the predicted data points are within the 20% of the measured values. Therefore, the model may be used to obtain a good estimate of effective surface area across the membrane contactor for the set of experimental conditions examined in this study (Table 11).
[0219] Some of the uncertainty of the model arises from the higher variability of results obtained at lower oil concentrations. At low oil concentrations, the oil flux was observed to improve over time as steady state conditions were approached leading to possible higher error in selecting steady state oil flux (see above). FIGs. 34A-34F shows a few examples of the model prediction compared to the experimentally measured data for various transmembrane pressures, viscosities, oil concentrations, and influent flow rates. The model fits are reasonable but in some cases the model underestimates the data. It should be noted the model is developed using chemical systems possessing a high interfacial tension (-50 dynes/cm) and should be used with some caution with systems with much lower interfacial tensions. [0220] The model can then be incorporated into Equation 7 and a model for oil flux prediction is obtained and shown in Equation 1 1 for modules with surface area less than or equal to 1.4 m2. Equation 1 1 :
Figure imgf000055_0001
[0221] Rearranging and separating the terms which remain constant from those which were varied:
Figure imgf000055_0002
where:
Oil FIUXO/V: Oil flux across the membrane in m3/s m2,
dpore : Membrane pore size (m),
TMP: Applied transmembrane pressure (Pa),
μ: Viscosity (Pa s),
L: Membrane wall thickness (m) ,
ε: porosity
[0222] Comparison of experimental and predicted fluxes, shown in FIG. 35, indicate that the coefficient of correlation of the linear fit is 0.96 indicating the model predicts the oil flux reasonably well for the 2.5 inch diameter X50 membrane contactor. Some variability is observed for higher oil concentration predictions in particular for transmembrane pressures higher than 20 psi and oil concentrations above 10% (v/v). The model may be used as a guideline for other hollow fiber membrane contactors but would require some preliminary assessment and model validation. In particular, the pure oil feed flux component may differ from the Ergun equation when the porosity is below 40%. The permeability constant (here 150) has been shown to vary with porosity (see above).
[0223] Design example. The oil/water model shown in Equation 1 1 can be used for different purposes, such as choosing the most efficient combination of operating parameters or determining the surface area needed to achieve the desired oil removal with the system. FIGs. 36A-36J shows a few model simulations for every operating parameter affecting the process in the experimental ranges studied (Table 1 1). A number of conclusions can be drawn from FIGs. 36A-36J. A lower pressure may be more efficient for dilute mixtures (< 2% (v/v) oil concentration), while higher transmembrane pressures will be preferable at high oil concentrations (40% (v/v) oil concentration) (FIG. 36A, FIG. 36B, and FIG. 36C). Similarly, the model can be used to compare performance as a function of influent flow rates. In this case, operating at a higher transmembrane pressure will lead to higher oil flux and recovery (FIG. 36G and FIG. 36H). Viscosity can also be used as an operating lever by changing temperature of a given influent stream for instance. The model predicts the positive effect of increasing viscosity on oil permeation for dilute mixtures (< 200 ppm oil concentration) and the negative effect of increasing viscosity for higher concentrations (FIG. 36D, FIG. 36E, FIG. 36F). Finally, FIGs. 36I-36J show the decrease of effective surface area with increasing water content in the feed. The model predicts all the trends observed above accurately.
[0224] The effective surface area needed to reach a given oil recovery may then be determined with the selected operating conditions. The advantage of using the 2.5 inch diameter module for the experimental data is that surface area is small and performance sensitivity may be accurately measured. The small module can be thought as a slice of the larger modules and, therefore, the model regarded as a differential equation.
Figure imgf000056_0001
Figure imgf000057_0001
[0225] Determination of A:
Figure imgf000057_0003
[0226] Determination of Cf: (trial and error calculation)
Figure imgf000057_0002
[0227] Upon integration between the influent and target effluent concentrations, an expression is obtained for predicting the required membrane surface area for a given effluent concentration and another expression is obtained to determine oil concentration in the water effluent for a given surface area. Example calculations are presented below. Projected values of surface area needed to obtain a chosen oil recovery may then be calculated. A few examples of the model application is shown in Table 12.
[0228] A possible limitation associated with the proposed scale-up approach is the difference of shell-side pressure drop occurring in modules of different sizes. The model was developed with experimental data collected with the 2.5 x 8 inch module and flux predictions account for the pressure drop related to the small module size and working influent flow rates. If a larger module is selected for a higher surface area, the pressure drop profile may be different leading to less accuracy of the model prediction. Therefore, the model is useful in providing a first estimation of membrane performance.
[0229] Table 12: Surface area predictions for 95% oil recovery.
Figure imgf000057_0004
Figure imgf000058_0002
[0230] Nomenclature:
Coii: Oil concentration, v/v ratio is used in the model
TMP: Transmembrane pressure (Pa)
Q: Influent flow rate (m3/s)
μ: Viscosity (Pa s)
Flux: Oil flux (m3/s m2)
: Feed oil concentration (volumetric ratio)
Cf: Effluent oil concentration (volumetric ratio)
q0: Oil flow rate in feed (m3/s)
A: Design surface area (m2)
Oil FIUXO/W: Oil flux across the membrane in m3/s m2
Oil Fluxpure oii: Oil flux across the membrane obtained with pure oil feed in m3/s m2
: Equivalent particle diameter (m):
Figure imgf000058_0001
dpore: Membrane pore size (m)
L: Membrane wall thickness (m)
ε: porosity
[0231] Conclusions. Modeling of oil and water separation with membrane systems have mainly focused on hydrophilic materials. The present work allowed the development of the first model to date to predict oil permeation for the separation of oil/water mixtures with hydrophobic hollow fiber membrane contactors. The new modelling approach used the performance of the membrane contactor under pure oil feed conditions and the effective area ratio to predict oil flux with oil/water mixture operations. The model fits the experimental data well with a correlation coefficient of 0.95. The model reliably predicts oil permeation in the low oil concentration range and presents more uncertainty for high oil concentrations and high transmembrane pressures. Transmembrane pressure, viscosity, influent flow rate and oil concentrations are the input parameters that the operator can control to optimize the use of the membrane technology for oil/water separation. The model can be used to predict membrane surface area needed to obtain desired insoluble oil removal from an oil/water feed mixture. The effective area model developed in this work is based on studies using a high interfacial tension (-50 dynes/cm) and without solids so some caution should be applied. Additional work may address the comparative effects of lower interfacial tensions and solids.
[0232] Example calculations.
A. Determination of surface area required for 95% oil removal
Inputs:
Membrane material: Pore size
Porosity = 0.4
Figure imgf000059_0004
Wall thickness
Figure imgf000059_0005
Process for 95% removal:
Figure imgf000059_0001
[0233] Calculate Oil Flux^
Figure imgf000059_0002
[0234] Determination of a and β:
Figure imgf000059_0003
[0235] Determination of A:
Figure imgf000060_0001
[0236] Calculation of effluent concentration for a surface area of 16 m2 and TMP is reduced to 1 bar.
Inputs:
Membrane material: Pore size
Figure imgf000060_0004
Porosity = 0.4
Wall thickness
Figure imgf000060_0006
Membrane Area = 16 m2
[0237] Process for 95% removal:
Figure imgf000060_0005
[0238] Calculate Oil FluxPure
Figure imgf000060_0002
[0239] Determination of a and β
Figure imgf000060_0003
Figure imgf000061_0003
[0240] Determination of Cf by trial and error
Figure imgf000061_0001
[0241] Trial and error:
Figure imgf000061_0002
[0242] Table 13 : Trial and Error Calculation
Figure imgf000061_0007
Therefore,
Figure imgf000061_0005
Figure imgf000061_0004
[0243] Figure Captions. FIG. 28: Experimental flux ratio vs oil concentration in the feed. Experimental conditions: TMP = 20 psi, Influent flow rate: 3.8 L/min, Isopar M. Good fit for a saturation model:
Figure imgf000061_0006
[0244] FIG. 29: Effect of influent flow rate on flux ratio. Experimental conditions: Isopar M.
[0245] FIG. 30: Effect of transmembrane pressure on flux ratio for two oil feed concentrations. Experimental conditions: Influent flow rate = 3.8 L/min, Isopar M.
[0246] FIG. 31 : Effect of viscosity on flux ratio for various oil feed concentrations.
Experimental conditions: TMP = 20 psi, Influent flow rate = 3.8 L/min.
[0247] FIG. 32: Residual plot for the complete oil/water experimental data set presented in Table 10.
[0248] FIG. 33 : Experimentally determined flux ratio vs ae/at model estimated values. The model used is shown in Equation 8. The linear fit correlation factor of the solid line is 0.96. [0249] FIGs. 34A-34F: Examples of model fitting experimental data. Legend: symbols = experimental data, lines = model predicted values with equation 7. Experimental conditions: FIG. 34A and FIG. 34B: Isopar M, Influent flow rate = 3.8 L/min. FIG. 34C: Isopar M, Influent flow rate = 3.8 L/min. FIG. 34D: TMP = 20 psi, Influent flow rate = 3.8 L/min; FIGs. 34E-34F: Isopar M, FIG. 34E: Coil = 2%, TMP = 20 psi, FIG. 34F: Coil = 40%, TMP = 40 psi. [0250] FIG. 35: Experimental vs predicted flux for the X50 2.5 inch module contactor.
[0251] FIGs. 36A-36J: Model predictions for oil permeation. Experimental conditions: FIG. 36A, FIG. 36B, and FIG. 36C: Viscosity = 3 cP, Influent flow rate = 240 L/hr; FIG. 36D, FIG. 36E, and FIG. 36F: TMP = 1.5 bar, Influent flow rate = 240 L/hr; FIG. 36G and FIG. 36H: TMP = 1.5 bar, Viscosity = 3 cP; FIG. 361 and FIG. 36J: Viscosity = 3 cP, TMP = 1.5 bar, Influent flow rate = 240 L/hr.
REFERENCES
[0252] U.S. Patents 8,486,267, 8,491,792, 8,617,396, and 9, 149,772.
[0253] A. Amelio et al, Purification of biodiesel using a membrane contactor: Liquid-liquid extraction. Fuel Processing Technology 142, 352-360 (2016). [0254] A. G. Ezzati, Elham; Mohammadi, Toraj, Separation of water in oil emulsions using microfiltration. Desalination 185, 371-382 (2005).
[0255] B. S. Hu, K., Microfiltration of water in oil emulsions and evaluation of fouling mechanism. Chemical Engineering Journal 136, 210-220 (2008). [0256] M. K. Konishi, M.; Tamesui, N.; Omasa, T.; Shioya, S.; Ohtake, H., The separation of oil from an oil-water-bacteria mixture using a hydrophobic tubular membrane. Biochemical
Engineering Journal 24, 49-54 (2005).
[0257] K. Masoudnia, A. Raisi, A. Aroujalian, M. Fathizadeh, Treatment of Oily Wastewaters Using the Microfiltration Process: Effect of Operating Parameters and Membrane Fouling Study. Sep. Sci. Technol. 48, 1544-1555 (2013).
[0258] M. Padaki et al, Membrane technology enhancement in oil-water separation. A review. Desalination 357, 197-207 (2015).
[0259] N. P. R. Tirmizi, Bhavani; Wiencek, John, Demulsification of water/oil/solid emulsions by hollow-fiber membranes. Aiche J 42, 1263-1276 (1996).
[0260] E. N. Tummons, V. V. Tarabara, J. W. Chew, A. G. Fane, Behavior of oil droplets at the membrane surface during crossflow microfiltration of oil-water emulsions. Journal of Membrane Science 500, 211-224 (2016).
STATEMENTS REGARDING INCORPORATION BY REFERENCE AND VARIATIONS [0261] All references throughout this application, for example patent documents including issued or granted patents or equivalents, patent application publications, and non-patent literature documents or other source material, are hereby incorporated by reference herein in their entireties, as though individually incorporated by reference, to the extent each reference is at least partially not inconsistent with the disclosure in this application (for example, a reference that is partially inconsistent is incorporated by reference except for the partially inconsistent portion of the reference).
[0262] All patents and publications mentioned in the specification are indicative of the levels of skill of those skilled in the art to which the invention pertains. References cited herein are incorporated by reference herein in their entirety to indicate the state of the art, in some cases as of their filing date, and it is intended that this information can be employed herein, if needed, to exclude (for example, to disclaim) specific embodiments that are in the prior art.
[0263] When a group of substituents is disclosed herein, it is understood that all individual members of those groups and all subgroups and classes that can be formed using the substituents are disclosed separately. When a Markush group or other grouping is used herein, all individual members of the group and all combinations and subcombinations possible of the group are intended to be individually included in the disclosure. As used herein, "and/or" means that one, all, or any combination of items in a list separated by "and/or" are included in the list; for example "1, 2, and/or 3" is equivalent to "Ί ' or '2' or '3' or Ί and 2' or Ί and 3' or '2 and 3' or '1, 2, and 3'".
[0264] Every formulation or combination of components described or exemplified can be used to practice the invention, unless otherwise stated. Specific names of materials are intended to be exemplary, as it is known that one of ordinary skill in the art can name the same material differently. One of ordinary skill in the art will appreciate that methods, device elements, starting materials, and synthetic methods other than those specifically exemplified can be employed in the practice of the invention without resort to undue experimentation. All art-known functional equivalents, of any such methods, device elements, starting materials, and synthetic methods are intended to be included in this invention. Whenever a range is given in the specification, for example, a temperature range, a time range, or a composition range, all intermediate ranges and subranges, as well as all individual values included in the ranges given are intended to be included in the disclosure.
[0265] As used herein, "comprising" is synonymous with "including," "containing," or
"characterized by," and is inclusive or open-ended and does not exclude additional, unrecited elements or method steps. As used herein, "consisting of excludes any element, step, or ingredient not specified in the claim element. As used herein, "consisting essentially of does not exclude materials or steps that do not materially affect the basic and novel characteristics of the claim. Any recitation herein of the term "comprising", particularly in a description of components of a composition or in a description of elements of a device, is understood to encompass, in addition to the open-ended description, those compositions and methods consisting essentially of and consisting of the recited components or elements. The invention illustratively described herein suitably may be practiced in the absence of any element or elements, limitation, or limitations which is not specifically disclosed herein. [0266] The terms and expressions which have been employed are used as terms of description and not of limitation, and there is no intention in the use of such terms and expressions of excluding any equivalents of the features shown and described or portions thereof, but it is recognized that various modifications are possible within the scope of the invention claimed. Embodiments described herein can be combined in any combination, without limitation. Thus, it should be understood that although the present invention has been specifically disclosed by preferred embodiments and optional features, modification and variation of the concepts herein disclosed may be resorted to by those skilled in the art, and that such modifications and variations are considered to be within the scope of this invention as defined by the appended claims.

Claims

WHAT IS CLAIMED IS:
1. A method for recovering oil from an oil and water mixture, the method comprising:
determining a ratio of oil to water in a mixture comprising oil and water, wherein the water and the oil are at least partly immiscible;
identifying, using the ratio of oil to water, a transmembrane pressure, a process viscosity, or both for use in separating the oil and the water from the mixture,
wherein the transmembrane pressure is identified as a first transmembrane pressure when the ratio of oil to water is less than a first threshold ratio, wherein the transmembrane pressure is identified as a second transmembrane pressure when the ratio of oil to water is greater than the first threshold ratio, and wherein the first transmembrane pressure is less than the second transmembrane pressure,
wherein the process viscosity is identified as a first viscosity when the ratio of oil to water is less than a second threshold ratio, wherein the process viscosity is identified as a second viscosity when the ratio of oil to water is greater than the second threshold ratio, and wherein the first viscosity is greater than the second viscosity;
contacting a shell side of a hollow fiber with the mixture, wherein the hollow fiber comprises a porous and hydrophobic polymer material, wherein at least a portion of the oil in the mixture wets the hollow fiber and is transported through the hollow fiber to a tube side of the hollow fiber,
wherein contacting includes applying the transmembrane pressure to the shell side of the hollow fiber, or changing a viscosity of the oil to the process viscosity, or both to enhance transport of oil through the hollow fiber; and
collecting oil from the tube side of the hollow fiber.
2. The method of claim 1, wherein the first transmembrane pressure is selected from the range of 1 pounds per square inch to 40 pounds per square inch, and wherein the second transmembrane pressure is selected from the range of 20 pounds per square inch to 60 pounds per square inch.
3. The method of claim 1, wherein the first viscosity is selected from the range of 0.2 cP to 100 cP, and wherein the second viscosity is selected from the range of 0.2 cP to 30 cP.
4. The method of claim 1, wherein changing the viscosity of the oil includes heating the mixture, heating the hollow fiber, or both.
5. The method of claim 1, wherein changing the viscosity of the oil includes cooling the mixture, cooling the hollow fiber, or both.
6. The method of claim 1, wherein at least a portion of the oil is present in the mixture as microdroplets of oil suspended in water.
7. The method of claim 1, wherein at least a portion of the microdroplets have a cross-sectional dimension less than 1 μπι, less than 5 μπι, less than 10 μπι, less than 15 μπι, less than 20 μπι, or less than 50 μπι.
8. The method of claim 1, wherein determining the ratio of oil to water in the mixture includes determining a concentration of oil in water in the mixture, determining a concentration of water in oil in the mixture, determining a mass fraction of oil in the mixture, determining a mass fraction of water in the mixture, determining a mole fraction of oil in the mixture, determining a mole fraction of water in the mixture, determining a volume fraction of oil in the mixture, determining a volume fraction of water in the mixture, and any combination of these.
9. The method of claim 1, wherein the first threshold ratio is selected from the range of about 1 part per million to about 20%.
10. The method of claim 1, wherein the second threshold ratio is selected from the range of about 1 part per million to about 2000 parts per million.
11. The method of claim 1, wherein the oil comprises one or more oils selected from the group consisting of biological oils, petroleum oils, food oils, paraffins, nonpolar hydrocarbons, saturated hydrocarbons, unsaturated hydrocarbons, aromatic hydrocarbons, water insoluble organic compounds, and any combination of these.
12. The method of claim 1, wherein the oil and the water are completely immiscible.
13. The method of claim 1, wherein the hollow fiber comprises a porous polymer material selected from the group consisting of polypropylene, polyethylene,
polytetrafluoroethylene, polyethylene terephthalate, polyvinylidene fluoride, polymethylpentene, and any combination of these.
14. The method of claim 1, wherein the hollow fiber comprises includes a plurality of pores having cross-sectional dimensions selected from the range of 0.001 μπι to 1 μπι.
15. The method of claim 1, wherein the hollow fiber is present in a hollow fiber membrane, and wherein contacting includes contacting a shell side of the hollow fiber membrane with the mixture, and wherein collecting includes collecting oil from tube sides of the hollow fiber membrane.
16. The method of claim 15, wherein the hollow fiber membrane is present in a membrane contactor, wherein the membrane contactor includes a shell side inlet, a shell side outlet, and a tube side outlet, wherein contacting the shell side of the hollow fiber includes flowing the mixture into the shell side inlet, and wherein collecting includes collecting oil from the tube side outlet.
17. A system for recovering oil from an oil and water mixture, the system comprising:
a pump for pumping a mixture comprising oil and water, wherein the oil and water are at least partially immiscible;
a temperature control element for heating or cooling the mixture, wherein the temperature control element is positioned in thermal communication with the mixture;
a hydrophobic porous membrane contactor, wherein the hydrophobic membrane contactor is positioned in fluid communication with the pump, wherein the hydrophobic membrane contactor comprises a hollow fiber membrane, wherein the hollow fiber membrane has a shell side in fluid communication with a shell side inlet of the hydrophobic porous membrane contactor and a shell side outlet of the hydrophobic porous membrane contactor, wherein the hollow fiber membrane has a tube side in fluid communication with a tube side outlet of the hydrophobic porous membrane contactor, and wherein the hollow fiber membrane comprises a plurality of porous hydrophobic polymer hollow fibers;
an oil collection system positioned in fluid communication with the tube side outlet of the hydrophobic porous membrane contactor; and
a controller that controls a flow rate of the mixture using the pump, and a temperature of the mixture using the temperature control element,
wherein the controller further controls a transmembrane pressure of the mixture applied to the shell side of the hollow fiber membrane, or a process viscosity of oil in the mixture, or both the transmembrane pressure and the process viscosity; wherein the controller sets the transmembrane pressure to a first transmembrane pressure when a ratio of oil to water in the mixture is less than a first threshold ratio, wherein the controller sets the transmembrane pressure to a second transmembrane pressure when the ratio of oil to water in the mixture is greater than the first threshold ratio, and wherein the first
transmembrane pressure is less than the second transmembrane pressure,
wherein the controller sets the process viscosity to a first viscosity when a ratio of oil to water in the mixture is less than a first threshold ratio, wherein the controller sets the process viscosity to a second viscosity when the ratio of oil to water in the mixture is greater than the first threshold ratio, and wherein the first viscosity is greater than the second viscosity.
18. The system of claim 17, further comprising:
a pressure transducer for measuring a pressure of the mixture, wherein the pressure transducer is positioned in fluid communication with the mixture; and
a feedback circuit between the pressure transducer and the controller for providing the pressure of the mixture to the controller for use in controlling the transmembrane pressure.
19. The system of claim 17, further comprising:
a temperature sensor for measuring a temperature of the mixture, wherein the temperature sensor is positioned in thermal communication with the mixture; and
a feedback circuit between the temperature sensor and the controller for providing the temperature of the mixture to the controller for use in controlling the temperature of the mixture, or the process viscosity ,or both.
20. The system of claim 17, further comprising:
a concentration sensor for determining a ratio of oil to water in the mixture, wherein the concentration sensor is positioned in communication with the mixture; and
a feedback circuit between the concentration sensor and the controller for providing the ratio of oil to water in the mixture to the controller for use in controlling the transmembrane pressure, or the process viscosity, or both.
21. The system of claim 17, wherein the hydrophobic porous membrane contactor is a first hydrophobic membrane contactor and wherein the system further comprises a second hydrophobic porous membrane contactor in fluid communication with the shell side outlet of the first hydrophobic porous membrane contactor.
22. The system of claim 21, wherein the transmembrane pressure is a first membrane contactor transmembrane pressure, wherein the controller sets the first membrane contactor transmembrane pressure to the second transmembrane pressure, and wherein the controller sets a second membrane contactor transmembrane pressure applied to the second hydrophobic porous membrane contactor to the first transmembrane pressure.
23. The system of claim 21, wherein the process viscosity is a first membrane contactor process viscosity, wherein the controller sets the first process viscosity to the second viscosity, and wherein the controller sets a second process viscosity of oil in the second hydrophobic porous membrane contactor to the first viscosity.
24. The system of claim 21, wherein a first concentration of oil in the first hydrophobic porous membrane contactor is greater than a second concentration of oil in the second hydrophobic porous membrane contactor.
25. The system of claim 17, wherein the temperature control element is in thermal communication with the hydrophobic porous membrane contactor.
26. The system of claim 25, wherein the temperature control element is arranged to control a first temperature of a first portion of the hydrophobic porous membrane contactor and to control a second temperature of a second portion of the hydrophobic porous membrane contactor.
27. The system of claim 17, wherein the oil and water are completely immiscible.
PCT/US2016/029301 2016-04-13 2016-04-26 Oil recovery using microporous hydrophobic membrane contactors WO2017180167A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US201662321924P 2016-04-13 2016-04-13
US62/321,924 2016-04-13

Publications (1)

Publication Number Publication Date
WO2017180167A1 true WO2017180167A1 (en) 2017-10-19

Family

ID=60042089

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2016/029301 WO2017180167A1 (en) 2016-04-13 2016-04-26 Oil recovery using microporous hydrophobic membrane contactors

Country Status (1)

Country Link
WO (1) WO2017180167A1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170291121A1 (en) * 2016-04-07 2017-10-12 Nikolai Kocherginksy Membrane-based washing and deacidification of oils
CN114247396A (en) * 2020-09-23 2022-03-29 中国科学院大连化学物理研究所 Polytetrafluoroethylene hollow fiber membrane microchannel reactor and preparation and application thereof

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4229297A (en) * 1978-01-10 1980-10-21 Asahi Kasei Kogyo Kabushiki Kaisha Method of separating oil from oil-containing liquid
US6146535A (en) * 1994-06-24 2000-11-14 Sutherland; George Organic recovery method using hollow fiber microfilters
US20020179529A1 (en) * 2000-06-29 2002-12-05 Johnson Robert S. Water separation from solvent
US20050264302A1 (en) * 2004-05-04 2005-12-01 Kam Controls Incorporated Device for determining the composition of a fluid mixture
WO2014089443A1 (en) * 2012-12-07 2014-06-12 Advanced Water Recovery, Llc Dissolved air flotation, antisolvent crystallisation and membrane separation for separating buoyant materials and salts from water
US20150136697A1 (en) * 2013-11-18 2015-05-21 Toyo Engineering Corporation Oily water separator, oil-water separating method and filtration membrane unit

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4229297A (en) * 1978-01-10 1980-10-21 Asahi Kasei Kogyo Kabushiki Kaisha Method of separating oil from oil-containing liquid
US6146535A (en) * 1994-06-24 2000-11-14 Sutherland; George Organic recovery method using hollow fiber microfilters
US20020179529A1 (en) * 2000-06-29 2002-12-05 Johnson Robert S. Water separation from solvent
US20050264302A1 (en) * 2004-05-04 2005-12-01 Kam Controls Incorporated Device for determining the composition of a fluid mixture
WO2014089443A1 (en) * 2012-12-07 2014-06-12 Advanced Water Recovery, Llc Dissolved air flotation, antisolvent crystallisation and membrane separation for separating buoyant materials and salts from water
US20150136697A1 (en) * 2013-11-18 2015-05-21 Toyo Engineering Corporation Oily water separator, oil-water separating method and filtration membrane unit

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170291121A1 (en) * 2016-04-07 2017-10-12 Nikolai Kocherginksy Membrane-based washing and deacidification of oils
US10065132B2 (en) * 2016-04-07 2018-09-04 Nikolai Kocherginksy Membrane-based washing and deacidification of oils
CN114247396A (en) * 2020-09-23 2022-03-29 中国科学院大连化学物理研究所 Polytetrafluoroethylene hollow fiber membrane microchannel reactor and preparation and application thereof

Similar Documents

Publication Publication Date Title
Chakrabarty et al. Ultrafiltration of stable oil-in-water emulsion by polysulfone membrane
Tummons et al. Behavior of oil droplets at the membrane surface during crossflow microfiltration of oil–water emulsions
US9616361B2 (en) Systems for separation of particulates and associated methods and devices
Mohammadi et al. Modeling of membrane fouling and flux decline in reverse osmosis during separation of oil in water emulsions
Kong et al. Oil removal from oil-in-water emulsions using PVDF membranes
Tirmizi et al. Demulsification of water/oil/solid emulsions by hollow‐fiber membranes
Chen et al. Removal of oil, grease, and suspended solids from produced water with ceramic crossflow microfiltration
Hlavacek Break-up of oil-in-water emulsions induced by permeation through a microfiltration membrane
Ezzati et al. Separation of water in oil emulsions using microfiltration
Kajitvichyanukul et al. Membrane technologies for oil–water separation
He et al. The effects of salt concentration and foulant surface charge on hydrocarbon fouling of a poly (vinylidene fluoride) microfiltration membrane
Murić et al. Comparison of ceramic and polymeric ultrafiltration membranes for treating wastewater from metalworking industry
Fux et al. Microscale dynamics of oil droplets at a membrane surface: deformation, reversibility, and implications for fouling
US10124300B2 (en) Flux-enhanced hierarchical porous membrane for oil-water nanoemulsion separation
Echakouri et al. Experimental and computational fluid dynamics investigation of the deterioration of the rejection capacity of the membranes used in the filtration of oily water systems
Mercelat et al. Mechanisms for direct separation of oil from water with hydrophobic hollow fiber membrane contactors
Al-Alawy et al. Treatment of simulated oily wastewater by ultrafiltration and nanofiltration processes
WO2017180167A1 (en) Oil recovery using microporous hydrophobic membrane contactors
Hejase et al. Nanofiltration of saline oil-water emulsions: Combined and individual effects of salt concentration polarization and fouling by oil
Loh et al. Experimental study of the separation of oil in water emulsions by tangential flow microfiltration process. Part 1: analysis of oil rejection efficiency and flux decline
Kiss et al. Treatment of model oily waste water by microfiltration
Ghaffour et al. Use of ultrafiltration membranes in the treatment of refinery wastewaters
Mercelat Fundamental study of hydrophobic microporous membrane contactors for the recovery of insoluble oil from oil-water mixtures
CN106315765B (en) Method for separating and analyzing oil-containing polymer-containing component in oilfield sewage by gradient membrane
Ghandashtani et al. Simulation of Oil-Water Separation in Flat Sheet Membrane

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16898850

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 18/03/2019)

122 Ep: pct application non-entry in european phase

Ref document number: 16898850

Country of ref document: EP

Kind code of ref document: A1