Exploration and Production Technologies
Treating Coalbed Natural Gas Produced Water for Beneficial Use by MFI Zeolite Membranes


Project Goal
The objectives of the project are an improved understanding of the mechanisms of reverse osmosis (RO) processes on zeolite membranes and the factors determining the performance of the membranes. Improvement of the membranes and establishment of optimal operating conditions to enhance water flux and ion rejection will be addressed. The final goal is the evaluation of the technical and economic feasibilities of long-term RO operation.

Petroleum Recovery Research Center (PRRC), New Mexico Institute of Mining and Technology, Socorro, NM

Research conducted at PRRC has resulted in development of new zeolite membranes to treat coalbed natural gas (CBNG) produced water for beneficial use. The zeolite membrane has excellent performance factors for water separation, but the low volume of water flow and high cost of the membrane make it uneconomical at present. Improving methods to synthesize the zeolite membranes and to optimize operating conditions for water and ion exchange are expected to provide a long-term, cost-effective membrane. Results of the PRRC research include:

  • Ion separation by zeolite membranes was demonstrated by a cross-flow experiments. Ion separation by zeolite membranes was attributed to size exclusion of large hydrated ions and competitive diffusion of ion and water molecules in zeolite microchannel.
  • A phenomelogical model was established to describe the ion/water transport in zeolite membranes.
  • Synthesis of high aluminum ZSM-5 membranes with enhanced water flux (~2.0 kg/m2.h) and ion rejection (>90 percent) for concentrated solutions (>4x104 mg/L NaCl);
  • Zeolite membranes show high mechanical strength at high operating pressures of up to 10 MPa;
  • Mechanism of ion separation by zeolite membranes is attributed to size exclusion by intracrystalline zeolitic pores and electrostatic repulsion by intercrystalline pores.
  • Synthesis of 80 tubular zeolite membranes (L=25 cm) for development of the RO demonstration;
  • Development of RO demonstration with membrane surface area of 0.3 m2;
  • Long-term RO permeation with synthetic and real produced water that was carried out for over 90 days;
  • A slow decline in water flux was observed. The membrane fouling was attributed to scale deposition and organic molecule adsorption on zeolite surface.
  • Development of an in-line oxidative regeneration process with ~10 percent H2O2 solution for organic fouling removal (researchers found that 90 percent of water flux can be recovered).

Researchers focused on the long-term desalination performance of zeolite membranes on actual produced water: investigation of the fouling mechanism was the focus. The foulant species in produced water were investigated by analyzing the chemical composition in produced water, such as organic concentration, total dissolved solid (TDS), and solid suspensions. Long-term desalination was carried out for about 98 days through a cross-flow operation process. The water flux increased initially in the first two weeks, then declined with elapsed operating time. The decline in ion rejection was observed throughout the desalination process. The foulants formed on the membrane surface were studied by scanning electron microscopy (SEM) with an energy-dispersive spectrometer (EDS), in which the chemical composition of inorganic scale was revealed to be CaCO3, MgCO3, CaSO4, and MgSO4. The solute rejection behavior is markedly influenced by the scale formation, particularly the enhanced concentration-polarization, whereby the scale layer (~15 micron) hinders back diffusion of solutes from the membrane surface to the bulk solution. It is recommended that pretreatment to remove divalent ions is necessary to minimize the scale formation and membrane fouling.

To unravel the mechanisms of organic fouling and chemical regeneration, it is critical to understand the chemical composition of organic foulants in produced water and interactions of the foulants and the zeolite membrane surface. Fouling experiments were performed with organics that simulate organic species in produced water, such as fatty acid, phenol, and toluene solutions. The cleaning experiments for membrane regeneration after organic fouling were performed with diluted H2O2 solution under specified experimental conditions.

The synthetic zeolite membranes are easy to clean, have a longer life, and are mechanically and chemically more stable than other in-situ synthetic membranes. The membrane can resist acid corrosion and operate under high-pressure conditions up to 10 MPa. Performance of the modified zeolite membrane is better than traditional polymer membranes for separation of produced water. The goal of the project is to provide an economical membrane system that can reduce oilfield produced water salinity to less than 10 percent of normal levels and make the water available for beneficial use, particularly in agriculture. In-situ monitoring of produced water using zeolite-impregnated optical fibers could result in significant cost savings in CBNG produced-water monitoring.

Salinities for CBNG produced waters from New Mexico are high at about 170,000 ppm total dissolved solids (TDS). Most of this produced water is disposed of in deep injection wells, which is a costly process and inherently wasteful in an arid state where water supply is crucial.

The project’s objectives include improving operating conditions to optimize water flow, evaluating the technical and economic feasibility of scaling zeolite membranes from laboratory to field use, and achieving a better understanding of the RO process and membrane performance.

This project is a continuation of previous work performed under DE-FC26-00BC15326. Previously, DOE-funded projects at PRRC have shown the technical superiority of zeolite membranes and RO purification techniques for processing CBNG produced water. Laboratory-tested produced water samples from CBNG and oilfield development can be cleaned sufficiently to provide irrigation water. However, the low volume water flux and high membrane cost make the zeolite membranes uneconomical to produce at this point. Improvements in membrane synthesis and post-synthesis modification for enhanced water flux and ion rejection have been the main tasks of this period of work.

To clean produced water cost-effectively for surface disposal or beneficial uses, molecular sieve zeolite membranes were synthesized and tested for produced water desalination through a RO process. The experimental data are being used to establish a mass-transfer model and to evaluate technical and economic feasibilities of the new technology for CBNG produced-water treatment.

The fundamental studies in Phase 1 include mass-transport behaviors of water, ions, and dissolved organics passed through the zeolite membranes of varied Si/Al ratios, and effects of operation temperature, pressure, feed velocity, feed chemical composition, and pH on the RO performance. In Phase 2, research focuses on optimizing the RO operating conditions and developing a cost-effective method to repair the undesirable nanoscale intercrystal pores. In Phase 3, a RO unit with a 0.3-m² tubular membrane module was established and long-term operations testing for produced water purification were tested. Technical and economic evaluations were performed technical and economic evaluations and system design.

Phase 1—Completed September 30, 2005—milestones include the following:

  • Defect-free MFI-type zeolite membranes and FAU-type zeolite membranes were synthesized on porous a-alumina discs and commercial tubular substrates.
  • The RO experiments were performed for a variety of single-salt solutions and CBNG produced water.
  • Mechanistic studies showed that desalination by RO through the MFI membranes is controlled mainly by the size exclusion of large hydrated ions at the pore entrance and affected by water/ion competitive diffusion in the zeolite channels.
  • A phenomenological model was established to describe the water and ion transport behaviors during the RO process. Increasing the operating temperature and pressure significantly enhanced RO efficiency.

A comparison of properties exhibited by traditional polymer vs. zeolite membranes.

A comparison of properties exhibited by traditional polymer vs. zeolite membranes:

  • The dissolved organics could be effectively removed during the RO process, but the organic sorption in zeolite decreased ion rejection and water flux. However, membranes fouled by organics can be totally regenerated by heating at 250 °C.
  • A new zeolite-fiber microsensor has been developed for in-situ water monitoring in produced-water management.

Phase 2 (Year 2, ending September 30, 2006) accomplishments to date include the following:

  • In the fourth quarter of calendar year 2005, MFI membranes with a Si/Al ratio of 50 were synthesized by seeding and secondary growth technique. Meanwhile, efforts were made to reduce the membrane thickness for enhanced water flux. With the membrane thickness reduced to ~1.5 µm, the water flux was increased to ~1kg/m2/hour, while the ion rejection remained at >80 percent.
  • Tests showed that high Al/Si ratio MFI membranes exhibited better RO performance. Data summarizing the separation of 0.10 M NaCl solution on pure silicalite and MFI membranes with Si/Al = 50 (ZSM-5) showed that both membranes have high ion rejection rates—more than 90 percent. However, the permeability of the ZSM-5 membrane is nearly twice that of the pure silicalite membrane because the ZSM-5 zeolite had higher hydrophilicity.
  • Membrane modification using Al3+ oligomers was conducted by counter-diffusion treatment. Membrane modification using Zr4+ oligomers was carried out by online deposition. A map of chemical and operational conditions was established for effective membrane modification. Both methods achieved significant improvement in ion rejection on the modified membranes but caused decline in water flux.
  • Test results showed that dissolved organics and dissolved salts can be removed simultaneously during the RO process, but organic sorption decreased the ion rejection and water flux for silicalite membranes.
  • Zeolite membranes modified by metal oligimer deposition have shown long-term stability for over 500 hours at RO operating condition (high pressure and high ion concentration).
  • Improvements in membrane synthesis have achieved considerable progress in past months. Synthesis of high aluminum content ZSM-5 membrane (Si/Al=~40) resulted in improvement in membrane performance with water flux of ~3.2 kg/m2.h and ion rejection of ~94.0 percent for a 0.10 M NaCl solution.
  • RO permeation of high concentration solutions on the newly developed ZSM-5 membrane indicated that the hydrophilic zeolite membrane exhibited superior desalination capability for concentrated solutions at ultra-high pressure (~10 MPa). Over 90 percent ion rejection was achieved on a 2.9x104 mg/L NaCl solution. Unlike polymeric membranes, the water flux increases linearly with operation pressure in the tested pressure range (2.0 MPa~10.0 MPa), suggesting high water recovery could be achieved by RO processes on zeolite membranes.

Phase 3 (Year 3, ending September 30, 2007) accomplishments to date include:

  • About 80 tubular zeolite membranes (L=25 cm) have been synthesized and RO performance of each individual membrane has been tested. Zeolite membranes with ion rejection of >95% were selected for construction of the demonstration. Upscaling the membrane fabrication will be critical to technology transfer. Several factors including reproducibility and substrate quality for manufacturing zeolite membranes have been investigated
  • A zeolite RO demonstration with membrane surface area of 0.3 m2 was established. The RO permeation gave an overall 98.5 percent ion rejection and 1.5 kg/m2.h water flux at a transmembrane pressure of 3.4MPa for 0.1M NaCl solution.
  • A long-term permeation test with actual produced water was carried out for over three months. Water flux declined slowly during the permeation period, indicating the potential for fouling by dissolved components in produced water. A membrane fouling mechanism study revealed that scale deposition and organic sorption are the main reasons for flux decline.
  • With elapsed RO permeation time, multivalent ions in produced water (i.e., Ca2+, Mg2+, SO42-, and CO32-) will concentrate on the membrane surface and deposit salt crystals that cover the whole surface of the zeolite membrane and increase the mass transport resistance. An average scale deposition rate of 0.15 um/day was observed on zeolite membranes when testing with CBM produced water in the San Juan Basin of NM. Dissolved organics with small molecules also cause considerable decline in water flux. The flux decline caused by organic sorption cannot be regenerated by back-washing or mechanical cleaning. Periodic chemical cleaning is necessary to recover the water throughput.
  • An efficient in-line oxidative regeneration process for zeolite membranes was developed by using H2O2solution at room temperature. Over 90 percent of water flux was recovered after a 10 min. treatment of 10 percent H2O2 solution at room temperature. Pretreatment for divalent ion removal and periodic regeneration by strong oxidant treatment are necessary for long-term operation of produced water desalination.

Current Status (December 2008) 
This project is completed and the final report is listed below under "Additional Information".

The objectives of the proposed research include: (1) to conduct extensive fundamental investigations in order to understand the mechanism of the RO process on real zeolite membranes and factors determining the membrane performance (Phase I); (2) based on fundamental understandings obtained in Phase 1, to improve the membranes and optimize operating conditions to enhance water flux and ion rejection (Phase II); and (3) to perform long-term RO operation on tubular membranes to study membrane stability and collect experimental data necessary for reliable evaluations of technical and economic feasibilities (Phase III). Currently, all tasks in Phase I and Phase II are completed. Most of the proposed tasks in Phase III, except the economic evaluation, are finished. Researchers are currently working on the upscale of membrane synthesis and economic evaluation. The detailed description of the research at each phase is summarized as:

(1) Ion removal by MFI-type pure silicalite membranes was demonstrated in our laboratory. The fundamental study indicated that ion separation by MFI-type zeolite membranes is attributed to size exclusion of large hydrated ions and competitive diffusion of ion/water inside the microchannel. A phenomenological model is derived for describing the molecular transport of ions and water through zeolite membranes.

(2) Research in Phase II has resulted in the development of a new synthesis method for hydrophilic zeolite membranes and a post-synthesis modification technique for enhanced RO separation performance. Considerable enhancement in water flux (~2.0 kg/m2.h) and ion rejection (>95%) was achieved by optimizing the membrane synthesis process and post-synthesis modification. The performance of zeolite membranes was evaluated by averaging the ion rejection and water flux of all the zeolite membranes synthesized by conventional in-situ crystallization and two-step hydrothermal treatment. The enhancement in desalination performance and reproducibility suggest the potential for practical application of this technology.

(3) Research in Phase III was aimed at upscaling the synthesis of zeolite membranes to provide fundamental data for evaluating the economic efficiency of zeolite membrane technology. About 80 tubular zeolite membranes have been synthesized, with 73 having good ion separation efficiency (>95%). Each membrane was tested for RO separation on 0.1 M NaCl solution for membrane quality evaluation. A RO demonstration unit with membrane surface area of 0.3m2 has been developed. The RO demonstration includes three main parts: feed solution reservoir, high pressure pump, and separation unit. Long-term permeation of actual produced water was tested for over 3 months. Variations in water flux and ion rejection were studied and the fouling mechanism was explained by scale deposition and organic sorption.

(4) Eight papers have been submitted to or published in peer-reviewed journals. The achievement of membrane synthesis by two-step hydrothermal crystallization has resulted in a pending patent. The research approach sponsored by this program was also presented at several conferences, including the annual meeting of the Society of Petroleum Engineers (SPE2004, 2007), annual meeting of the American Institute of Chemical Engineering (AIChE2004), the Annual Forum of the Ground Water Protection Council (2005, Portland Oregon), and the Carbon Sequestration Southwest Partnership Workshop (August 20, 2006, Farmington).

This project was selected under DOE’s Federal Lands Access and Produced Water Management solicitation, DE-PS26-04NT15460, January 2004.

Project Start: October 1, 2004 
Project End: March 30, 2008

Anticipated DOE Contribution: $855,204 
Performer Contribution: $285,069 (33 percent of total)

Contact Information
NETL - Jesse Garcia (jesse.garcia@netl.doe.gov or 918-699-2036)
PRRC - Robert Lee (lee@prrc.nmt.edu or 505-835-5408)

Additional Information

Final Project Report [PDF]

LX. Li, JH. Dong, and T.M. Nenoff, “Transport of Water and Alkali Metal Ions through MFI Zeolite Membranes during Reverse Osmosis,” Separation and Purification Technology, 53 (2007), pp. 42-48.

LX. Li, N. Liu, B.J. McPherson, and R. Lee, “Influence of Counter Ions on the Reverse Osmosis through MFI Zeolite Membranes: Implications for Produced Water Desalination,” Desalination, 2007, accepted.

LX. Li, N. Liu, B.J. McPherson, and R. Lee, “Enhanced Water Permeation of Reverse Osmosis through MFI-type Zeolite Membranes with high Aluminum Contents," Industrial & Engineering Chemistry Research, 2007, in press.

The RO demo unit with membrane surface area of 0.3 m2 and produced water samples before and after purification. The produced water samples were provided by Basin Disposal Inc. and the Bureau of Land Management in Farmington, NM.

Close-up microphotograph of the zeolite membrane and diagrams of the crystal growth system that allows water to pass through but prevents salts from passing through the membrane.