NOVEL COMPOSITE MEMBRANES AND PROCESS FOR NATURAL GAS UPGRADING

Bikson, Ben ; Giglia, Sal ; Hao, Jibin

Innovative Membrane Systems, Inc. (United States)

关键词: Natural Gas;    Water Vapor;    Flow Rate;    Fibers;    Economics;   

DOI  :  10.2172/823967
RP-ID  :  NONE
RP-ID  :  FC26-99FT40497
RP-ID  :  823967

美国|英语

来源: UNT Digital Library

【 摘 要 】

In the second phase of this project, the newly developed membrane module for natural gas dehydration was tested and evaluated in a pilot plant located at a commercial natural gas treatment site. This phase was undertaken jointly with UOP LLC, our commercialization partner. The field test demonstrated that a commercial-size membrane module for natural gas dehydration was successfully manufactured. The membrane module operated reliably over 1000 psi differential pressure across the membrane in the field test. The effects of feed gas pressure, permeate gas pressure, feed flow rate, purge ratio (flow rate ratio of permeate outlet to feed), and feed gas dew point on the membrane module performance were determined and found to meet the design expectations. Although water vapor permeance was lower than expected, substantial natural gas dehydration was demonstrated with low purge ratio. For example, dew point was suppressed by as much as 30 F with only about 2 {approx} 3% purge ratio. However the bore side pressure drops were significantly higher than the projected value from the fluid dynamic calculation. It is likely that not all the fibers were open in either the sweep or the permeate tube sheet end. This could help to explain the relatively low water vapor permeances that were measured in the field. An economic evaluation of the membrane process and the traditional Triethylene Glycol (TEG) process to dehydrate natural gas was performed and the economics of the two processes were compared. Two sets of membrane module performance properties were used in the economic analysis of the membrane process. One was from the results of this field test and the other from the results of the previous small-scale test with a medium pressure membrane variant conducted at 750 psig. The membrane process was competitive with the TEG process for the natural gas feed flow rate below 10 MMSCFD for the membrane with previously measured water vapor permeance. The membrane process was competitive for the feed flow rate below 1 MMSCFD even for the membrane with the water vapor permeance of this field test.