Progress 09/01/17 to 08/31/21
Outputs
Target Audience: CMS is continuing the Biogas R&D project, started under this SBIR award, into Commercial Product Development (PD). The continuing Biogas PD project is funded entirely by CMS and no USDA nor NIFA funds are being used. CMS has developed two pilot partners identified to test our membrane performance under field conditions: One will be an advancement of USDA funded research, upgrading raw, humidified biogas in a commercial sized rig. We are assessing our membranes viability in distributed and rural US porcine waste operations. We expect to provide capital and operational savings to customers compared to existing technology. The second pilot partner is with a growing biomethane operator in Europe: we aew planning to run a full scale cartridge in a first stage test rig with our membrane humidification system in place. Success in this pilot operation would give CMS an entry point to compete with major, international membrane providers. Concurrently, CMS is negotiating a manufacturing agreement to scale our OptipermTM Biogas line with a large industrial partner. A large scale coating, rolling, and packaging process would enable CMS to meet expanding marketing demand in 2023 with our pilot partners. We anticipate a FY2022 execution of this manufacturing agreement tied to our sales pipeline. We are making definite progress in commercializing the Biogas product development. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?Hiring and training of Chemical Engineering undergraduates from University of Delaware and other universities as interim or summer technicians. How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
Task 1. Optimize Biogas Membranes for Improved Flux and Selectivity We realized the first-generation biogas membrane (Polymer A and Polymer B) performance is heavily dependent on the humidity of the feed gas. The CMS biogas membrane needs water for the best carbon dioxide facilitated transport. The hydrophilic groups in the membrane can absorb water and let water go through the membrane quickly. Carbon dioxide is a water-soluble gas, its transport can be facilitated by the presence of water. By contrast, nitrogen or methane has extremely low solubilities in water and their permeance cannot be increased by the presence of water significantly. The humidity-permeance of test results confirmed our hypothesis: when the relative humidity of feed gas decreased from 84% to 44%, the permeance of CO2 dropped from 1546 GPU to 267 GPU and the CO2/N2 selectivity dropped from 29.6 to 12.0. Both permeance and selectivity drop when the humidity decreases. To improve the membrane module performance, we worked further on polymer synthesis to develop a second-generation polymer for CO2 separation. This polymer is an amine containing fluorinated polymer and we call it as Polymer C. Amine can act as a much stronger CO2 transport facilitation agent than water. Unlike water, which can only physically dissolve CO2, amine can chemically react with CO2 with presence of water. Compared with the first-generation polymer, the second-generation polymer has similar CO2 permeance but higher CO2/CH4 or CO2/N2 selectivity. Under optimized conditions, the Polymer C membrane can have a CO2 permeance of around 2000 GPU, CO2/N2 selectivity of 50-100 and CO2/CH4 selectivity of 35-70. One interesting feature of Polymer C is that at lower humidity, the CO2 permeance decreases but the selectivity of CO2/N2 or CO2/CH4 increases . So at lower humidity, the membrane performance loss is minimized (loss of permeance can be offset by the increase of selectivity). Also we found the Polymer C has a relatively low water permeance, so the humidity loss can also be less of a problem compared with the first generation polymer. Task 2: Long Term Test of Membranes The membrane needs to be exposed to biogas for years, the long term stability of the membrane is another deciding factor. In this task, we subjected the Polymer C membranes (the second-generation polymer) to multiple weeks of tests. We used humidified CO2/N2 (40/60) as the surrogate gas and the membranes were just exposed to air between gas tests. We can see, the permeance/selectivity combination of Polymer C remain stable during the aging test (relatively lower selectivity can be compensated with higher carbon dioxide permeance and vice versa). Compared with the first-generation polymers, it is obvious that the Polymer C has higher stability. Overall, the Polymer C membrane showed remarkable stability during the 4+ months aging test. It gave us much confidence on the success of real biogas membrane application. Task 3. Scale-up Composite Membranes The scale-up flat sheet coating method has been well developed in CMS and we coated biogas membranes from the Polymer C in similar ways to the Polymer B and Polymer A. Task 4. Develop, Fabricate and Optimize Membrane Modules: Multiple flat sheet spiral modules (1.5 square feet total area) have been made from the scale-up membranes. The module gas test showed stable performance and somehow lower performance than small sample test. The Polymer B (first generation) membrane module is also listed as a comparison. Consistent with the small sample results, we can see clearly that the Polymer C membrane has a significantly higher selectivity than Polymer B at the module level. At the module level, the stage cut is 15-25%, much higher than small sample tests (<3%) so the loss of water can affect the performance significantly and that is why we see significantly lower CO2 permeance compared with small membrane disk test. We still use N2 as a surrogate gas for CH4 to make the test easier. The permeance of N2 is close to that of CH4 based on our experience. The gas selectivity and permeance is calculated by the simple calculation method, taking average of the feed and retentate CO2 concentration as the feed side concentration and the permeate concentration as the permeate side concentration: We have recently fabricated much larger commercial length spiral module of 16 square feet area. This module is developed and tested in house in order to optimize for commercial installation. . Task 5. Build and Run the Pilot Test System The pilot system is designed to take biogas from the digester directly and compress it to 15- 30 psig. The compressed biogas is fed to the CMS biogas separation spiral module. We apply a vacuum (2-4 psia) to the permeate side of the membrane module to increase the driving force. The CO2 rich stream will come out from the permeate side of the module. It will go through the H2S scrubber (filled with iron sponge) and vent. The CH4 rich stream is from the retentate side and is collected in the leather bag. The pilot system was also equipped with a PLC to continuously monitor flow and composition to run without any operator input. We have developed two biogas rig systems. The first rig system was designed for the field test at the Dickinson farm site and the pilot test run was completed using the first rig (discussed in Task 6). Recently, another larger test rig was designed and developed that can accommodate larger commercial length modulesin order to further optimize the spiral membrane module design. The membrane module with larger area of 16 ft2 is studied in the second rig. The large modules can be tested up to 100 psig feed pressure and are designed to be humidified on either the feed or permeate side. Due to safety concern, no methane mixture is used for in house testing rather testing is carried out with CO2/N2 feed gas mixtures. Results of up to 6 hours run show that the performance of the membranes is stable and has potential for removal of CO2 from CO2/CH4 biogas mixtures at these commercial scales. Task 6. Run Long Term Field Test Previously we did the small samples of membrane and the biogas module exposure tests at the Wanner Farm of Lancaster, Pennsylvania. But later due to the difficulties of producing a stable biogas stream, the Wanner Farm had to terminate the collaboration with us. Fortunately, we found another farm, Dickinson College Farm at Harrisburg, Pennsylvania, is willing to collaborate with us to do the biogas upgrading test. The Dickinson college farm is using food waste from the cafeteria of the college as the feed to produce biogas. They can produce 3-5 cubic meters of biogas every day, which is enough for test runs of our rig. The produced biogas is stored in leather bags at atmospheric pressure and it will be the biogas feed source for our tests. The biogas test rig was installed in the Dickinson farm. The gas from the digester was fed to the pilot membrane system where it was scrubbed of H2S and compressed to be sent to the membrane unit. Testing was carried out for one month. Task 7. Engineering design and economic evaluation This field test work is synergistic with our parallel olefin-paraffin field tests at Delaware City Refinery. Economic analysis shows the payback time is 48 months for electricity generation vs. 33 months for biomethane production. The internal rates of return (IRR) are 25% and 37% for electricity generation and biomethane production, respectively. The advantage of the biomethane production plant is attributable to its lower capital cost and higher rate of sellable energy production. In the biogas to electricity plant only 28% of the energy in the biogas is converted to saleable electricity. The rest becomes unsaleable heat energy which must be used at the farm. This does not happen in the biogas to biomethane plant.
Publications
|
Progress 09/01/19 to 08/31/20
Outputs
Target Audience:
Nothing Reported
Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?
Nothing Reported
How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
Overall, the project has been going on well and we have been successful in in the membrane polymer development, scale up membrane coating, module fabrication, module performance optimization and rig building. Due to the COVD-19 outbreak, the rig pilot test and the farm field test were delayed. We plan to run these tests in the next fiscal year. Based on the small membrane disk sample and membrane module test data, we are confident that we can have a great demonstration of the biogas membrane upgrading system. Task 1. Optimize Biogas Membranes for Improved Flux and Selectivity We realized the first-generation biogas membrane (Polymer A and Polymer B) performance is heavily dependent on the humidity of the feed gas. The CMS biogas membrane needs water for the best carbon dioxide facilitated transport. The hydrophilic groups in the membrane can absorb water and let water go through the membrane quickly. Carbon dioxide is a water-soluble gas, its transport can be facilitated by the presence of water. By contrast, nitrogen or methane has extremely low solubilities in The humidity-permeance of test results confirmed our hypothesis: when the relative humidity of feed gas decreased from 84% to 44%, the permeance of CO2 dropped from 1546 GPU to 267 GPU and the CO2/N2 selectivity dropped from 29.6 to 12.0. Both permeance and selectivity drop when the humidity decreases. For the regular cross flow gas CMS module separations, as shown on Figure 1: we use the feed gas with 80-90% relatively humidity, so the CO2 gas flux close to the feed inlet can be high. But water is a small molecule which can permeate through the membrane rapidly. The gas on the feed side gradually loses water and becomes drier as the stage cut gets higher, thus the CO2 gas permeance becomes lower. The overall module gas separation performance will be much lower than the small membrane disk sample results if the humidity cannot be managed well. This is why we got underperformed data in the gas module test of the previous year. To improve the membrane module performance, we worked further on polymer synthesis to develop a second-generation polymer for CO2 separation. This polymer is an amine containing fluorinated polymer and we call it as Polymer C. Amine can act as a much stronger CO2 transport facilitation agent than water. Unlike water, which can only physically dissolve CO2, amine can chemically react with CO2 with presence of water. The CO2 transport mechanism is shown in Figure 3 as below: Compared with the first-generation polymer, the second-generation polymer has similar CO2 permeance but higher CO2/CH4 or CO2/N2 selectivity. Under optimized conditions, the Polymer C membrane can have a CO2 permeance of around 2000 GPU, CO2/N2 50-100 and CO2/CH4 35-70. One interesting feature of Polymer C is: at lower humidity, the CO2 permeance decreases but the CO2/N2 or CO2/CH4 increases.So at lower humidity, the membrane performance loss is minimized (loss of permeance can be offset by the increase of selectivity). Also we found the Polymer C has a relatively low water permeance, so the humidity loss can also be less of a problem compared with the first generation polymer. Task 2: Long Term Test of Membranes The biogas membrane need to be exposed to corrosive raw biogas for years, the long term stability of the membrane is another deciding factor. In this task, we subjected the Polymer C membranes (the second-generation polymer) to multiple weeks of tests. We used humidified CO2/N2 (40/60) as the surrogate gas and the membranes were just exposed to air between gas tests. The membrane aging performance are shown in Graph 3 and 4. The permeance/selectivity combination of Polymer C remain stable during the aging test (relatively lower selectivity can be compensated with higher carbon dioxide permeance and vice versa.). Compared with the first-generation polymers, it is obvious that the Polymer C polymer has higher stability. Overall, the Polymer C membrane showed remarkable stability during the 4+ months aging test. It gave us much confidence on the success of real biogas membrane application. Task 3. Scale-up Composite Membranes The scale-up flat sheet coating method has been well developed in CMS (Figure 4) and we coated biogas membranes from the Polymer C in similar ways to the Polymer B and polymer A. To test the coated sheets, we took 47mm diameters disk punches from the scale-up membrane and measure the gas permeance. The small samples of the scale-up membrane sheet showed good performance and homogeneity. We used PAN and PVDF as the porous support for the scale-up membranes. Both substrates had good performance of the membrane coating (Table 2). Also we noticed the big improvement of selectivity over the first generation polymer. We use PVDF for the biogas module fabrication due to its better chemical stability. Task 4. Develop, Fabricate and Optimize Membrane Modules: Multiple flat sheet spiral modules (1.5 square feet total area) have been made from the scale-up membranes. The module gas test showed stable performance and somehow lower performance than small sample test.Consistent with the small sample results, we can see clearly that the Polymer C membrane has a significantly higher selectivity than Polymer B at the module level. At the module level, the stage cut is 15-25%, much higher than small sample tests (<3%) so the loss of water can affect the performance significantly and that is why we see significantly lower CO2 permeance compared with small membrane disk test. We still use N2 as a surrogate gas for CH4 to make the test easier. The permeance of N2 is close to that of CH4 based on our experience. The gas selectivity and permeance is calculated by the simple calculation method, taking average of the feed and retentate CO2 concentration as the feed side concentration and the permeate concentration as the permeate side concentration: Then the CO2/CH4 or CO2/N2 selectivity can be calculated as: Task 5. Build and Run the Pilot Test System The pilot systemis designed to take biogas from the digester directly and compress it to 15- 30 psig. The compressed biogas is fed to the CMS CO2 module. We apply a vacuum (2-4 psia) to the permeate side of the membrane module to increase the driving force. The CO2 rich stream will come out from the permeate side of the module. It will go through the H2S scrubber (filled with iron sponge) and vent. The CH4 rich stream is from the retentate side and will be collected in the leather bag. The building of the rig systems has been completed (see Figure 6). Currently we are still working on the trouble shooting of the pilot test system. The pilot test run will start as soon as we complete the trouble shooting. We will first use a humidified CO2/N2 mixture as a biogas surrogate gas to test the CO2 removal. Task 6. Run Long Term Field Test Previously we did the small samples of membrane and the biogas module exposure tests at the Wanner Farm of Lancaster, Pennsylvania. But later due to the difficulties of producing a stable biogas stream, the Wanner Farm had to terminate the collaboration with us. Fortunately, we found another farm, Dickinson College Farm at Harrisburg, Pennsylvania, is willing to collaborate with us to do the biogas upgrading test. The Dickinson college farm is using food waste from the cafeteria of the college as the feed to produce biogas. They can produce 3-5 cubic meters of biogas every day, which is enough for test runs of our rig. The produced biogas is stored in leather bags at atmospheric pressure and it will be the biogas feed source for our tests (Figure 7). We will start the field test as soon as we finish the pilot test at CMS. ?
Publications
|
Progress 09/01/18 to 08/31/19
Outputs
Target Audience:
Nothing Reported
Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?
Nothing Reported
How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
Overall, the project is going well and we have made considerable progress in the scale up membrane coating, module fabrication and module performance optimization. We are expecting a successful demonstration of the biogas upgrading product and process feasibility at the end of this project. Working with our partner (Wanner Farm, 40 miles away from CMS) we plan to run a test for upgrading biogas (40% CO2 with 3000 ppm H2S) from a real digestion side stream in a field site (Wanner Farm). CMS will supply the pilot system for conducting the field test. CMS personnel will provide assistance with the installation and start-up of the system. The objective of this test is to demonstrate under field conditions that the membrane system consistently upgrading biogas with a separation factor of at least 35 and a gas permeance of at least 800 GPU at 60 psig feed pressure (CO2 20%). This field test work is synergistic with our parallel olefin-paraffin field tests at Delaware City Refinery only 10 miles from CMS. The proposed Phase II work will deal with membrane optimization, scale?up, pilot and field testing of a non?fouling membrane system that will upgrade biogas directly from the anaerobic digestion. The scale-up membrane chemical and fouling resistance will be confirmed under field test conditions. We will run a joint field test with the Wanner Farm. With our complete success in Phase I and in collaboration with our partners (Wanner Farm, Pall, and Air Product), we are well positioned for additional success in Phase II and subsequent commercialization. Task 1. Optimize Biogas Membranes for Improved Flux and Selectivity After testing various batches of CMS biogas polymers, we believe CMS50 and CMS72 have the best performance. We did further tests to optimize the coating and porous support conditions. First, we used 0.7% and 1.0% polymer solution to make the membranes. The PAN porous support and PVDF support were both tested. The 40% CO2 and 60% CH4 mixture was used as a biogas substitute. N2 was also used as a replacement for CH4 in some cases due to its very close permeability to CH4 in our biogas membrane separation. PVDF flat sheet was proven to be more chemically stable in the phase I research. The membrane coating on PVDF showed similar gas permeance as the ones on PAN but with somehow lower selectivity. We tried to optimize the thickness of the membrane by using lower concentration of the coating solutions. We found 0.25% polymer solution could still give similar gas selectivity (CO2/CH4 or CO2/N2) and much high gas permeance (Table 2). 30 psig feed pressure was used in the new membrane tests. Task 2: Long Term Test of Membranes Long term stability of the membrane is another important property for the biogas application since the membrane need to be exposed to corrosive raw biogas. In this task, we subjected the membranes showed good separation performance in task I to multiple days of tests. Initially we only used humidified CO2/N2 (40/60) as the surrogate gas and the membranes were just exposed to air between gas tests. The CO2 permeance of Membrane 1 declined to about half after 50 days of gas tests and the selectivity remained unchanged. It confirmed the instability of the PAN porous support. The membrane 2 (on PVDF) showed higher stability than membrane 1. The gas permeance declined initially and then stabilized after 40 days of tests. The gas permeance and selectivity remained almost unchanged between day 40 and day 80. The membrane 3 (CMS72-Li on PVDF) showed the best stability among all the membranes tested. The gas permeance only declined 16% after 130 days of aging. Task 3. Scale-up Composite Membranes The scale-up flat sheet coating method has been well developed in CMS and we coated biogas membranes from the CMS50 and CMS72 polymers. After the coating, first step to test the scale-up membrane sheet is to punch multiple small samples (47 mm diameter disk) and test the gas permeance. Most of the scale-up membranes showed good initial performance and homogeneity. We only used PVDF as the porous support for the scale-up membranes. Task 4. Develop, Fabricate and Optimize Membrane Modules: We successfully made multiple flat sheet spiral modules (1500 cm2) from the scale-up membranes. The module gas test showed stable performance and somehow lower performance than small sample tests. The stage cut is 10-20%, much higher than small sample tests (<3%). The gas selectivity and permeance is calculated by the simple calculation method, taking average of the feed and retentate CO2 concentration as the feed side concentration and the permeate concentration as the permeate side concentration: Cross Flow Model Calculation for Spiral Flat Sheet Modules: For high stage cut flat sheet module gas separations, cross flow model is needed to calculate the membrane permeance and selectivity. We used the method from the paper of Separation and Purification Technology 1998 (13) 209-225. When applying their method, we assume there is no significant pressure drop in our low pressure gas separation, then γ = . The membrane gas selectivity can be calculated. The biogas module 1 (1500 cm2 area) CO2/N2 (N2 as a surrogate gas for CH4) separation and calculation using the cross flow model discussed above:The biogas module 1 was exposed to raw biogas for 3 weeks. During this period, it was soaked with water from biogas stream. It not only did not cause any damage to the membrane module, but also increase the performance. Case 2 showed the vacuum permeate method had slight higher module performance than the atmospheric pressure permeate separation. Case 3 gave us an example that 38.5% CO2 in mixture can be reduced to less than 10% (8.6%) by the vacuum permeate method in one stage. Biogas Module 2, 0.3% CMS72-Li on PVDF, 1500 cm2 spiral module, 30-40% stage cut. Without sufficient hydration, the biogas membrane underperformed at the module level. Compared with the small sample and super low stage cut test results, the gas permeance of the non-humidified module is only 20% of the former (300 GPU vs. 1500 GPU). Long term water soaking can give better humidification and significant performance improvement to the biogas membrane module. Biogas Module 3 0.35% CMS50-NH3 on PVDF, 1500 cm2 spiral module, 30-40% stage cut. ?Having the same water management problem as the other biogas modules, the ammonium salt membrane seemed to have higher selectivity than Li membrane at the module level. Further optimization is needed to improve the membrane performance at the module level. Task 5. Build and Run the Pilot Test System We are still working on the membrane module performance improvement and the pilot test system will start as soon as we can better manage the module humidification problem and further optimize the module performance. Task 6. Run Long Term Field Test-Wanner Farm Biogas exposure test of the modules was done at the Wanner farm. Biogas modules were exposed to the raw biogas stream (containing 3000 ppm H2S and saturated with water) directly. As I mentioned in Task 4: after 3 weeks, the gas selectivity of the module remained same but the CO2 permeance almost doubled. We believe it was due to the water soaking, later we opened the module and found it contained significant amount of liquid water. The membrane was sufficiently hydrated and showed the maximum permeability. On the other hand, it showed the stability of the module to the raw biogas. Summary Overall, the project is going well and we have made big progress in the scale up membrane coating, module fabrication and module performance optimization. We are expecting a very successful demonstration of the biogas upgrading product and process feasibility at the end of this project.
Publications
|
Progress 09/01/17 to 08/31/18
Outputs
Target Audience:Animal manure can become a serious pollution problem if not managed properly. Every year over 1 billion tons of fresh animal manure is produced by livestock in US. Improperly stored or used, animal manure can pollute rivers and underground drinking water supplies. Inadequately sized and poorly-lined ponds or other storage structures allow manure to escape into the surrounding environment. Manure is washed into streams and waterways, helping to form the 230 oxygen-deprived "dead zones" along the US coast. An energy efficient and environmentally friendly solution to animal manure is converting it to biogas by an "anaerobic digestion" process. Anaerobic digestion is a series of fermentation processes in which microorganisms break down biodegradable material in the absence of oxygen to generate a gas mixture of methane, carbon dioxide and trace amounts of other contaminant gas like hydrogen sulfide. Although the total solid or liquid waste is not significantly reduced after the anaerobic digestion, the pollution potential of animal manure digestate is greatly reduced compared with untreated animal manure.3,4 Rich in nutrition but much less polluting, the digestate can be used as a good chemical fertilizer replacement. The generated manure gas (biogas), containing approximately 60% methane, 40% carbon dioxide, trace amounts of hydrogen sulfide (3,000 ppm) and saturated with moisture, is combustible and can be directly used as a fuel. In most of the US farms equipped with anaerobic digesters, non-purified biogas is only a low quality fuel for micro-turbine electricity generation which provides very limited economic returns. Changes/Problems:The major reasons foran extension request are that we need more time to prepare for the demonstration of the biogas upgrading at the long term field test at Wanner Farm. Without sufficient hydration, the biogas membrane underperformed at the module level. Compared with the small sample and super low stage cut test results, the gas permeance of the non-humidified module is less than the former. Long term water soaking can give better humidification and significant performance improvement to the biogas membrane module. We are still working on the membrane module performance improvement and the pilot test system has been delayed but will start as soon as we can better manage the module humidification problem and further optimize the module performance. What opportunities for training and professional development has the project provided?
Nothing Reported
How have the results been disseminated to communities of interest?Biogas exposure test of the modules was done at the Wanner farm. Biogas modules were exposed to the raw biogas stream (containing 3000 ppm H2S and saturated with water) directly. As I mentioned in Task 4: after 3 weeks, the gas selectivity of the module remained same but the CO2 permeance almost doubled. We believe it was due to the water soaking, later we opened the module and found it contained significant amount of liquid water. The membrane was sufficiently hydrated and showed the maximum permeability. On the other hand, it showed the stability of the module to the raw biogas. What do you plan to do during the next reporting period to accomplish the goals? Develop and fabricate membrane modules Build and run pilot test unit Conduct field test Engineering design and economic evaluation Accomplish: Develop and fabricate membrane modules Milestone 4: Fabricated at least two membrane modules for use in Tasks 5 and 6. Build and run pilot test unit Milestone 5: Design, build and debug pilot test system for use. Conduct field test Milestone 6: Field validation at Wanner Farm demonstrating the biogas upgrading performance of a CMS membrane module with projected flux loss of no more than 20% after 1 year. Engineering design and economic evaluation Milestone 7: Demonstrate a biogas upgrading cost savings of at least 30% compared to the next lowest cost technology.
Impacts
What was accomplished under these goals?
Overall, the project is going well and we have made considerable progress in the scale up membrane coating, module fabrication and module performance optimization. We are expecting a successful demonstration of the biogas upgrading product and process feasibility at the end of this project. Working with our partner (Wanner Farm, 40 miles away from CMS) we plan to run a test for upgrading biogas (40% CO2 with 3000 ppm H2S) from a real digestion side stream in a field site. CMS will supply the pilot system for conducting the field test. CMS personnel will provide assistance with the installation and start-up of the system. In addition, CMS will assist with data collection and data analysis. The objective of this test is to demonstrate under field conditions that the membrane system consistently upgrading biogas with a separation factor of at least 35 and a gas permeance of at least 800 GPU at 60 psig feed pressure (CO2 20%). This field test work is synergistic with our parallel olefin-paraffin field tests at Delaware City Refinery only 10 miles from CMS. Task 1. Optimize Biogas Membranes for Improved Flux and Selectivity After testing various batches of CMS biogas polymers, we believe CMS50 and CMS72 have the best performance. We did further tests to optimize the coating and porous support conditions. First, we used 0.7% and 1.0% polymer solution to make the membranes. The PAN porous support and PVDF support were both tested. The 40% CO2 and 60% CH4 mixture was used as a biogas substitute. N2 was also used as a replacement for CH4 in some cases due to its very close permeability to CH4 in our biogas membrane separation. Initially 60 psig was used as the feed pressure and permeate was tested at atmospheric pressure. The stage cut was managed at 3% or lower to ease the membrane performance calculations. Although PAN is a very good porous support material, PVDF flat sheet was proven to be more chemically stable in the phase I research. The membrane coating on PVDF showed similar gas permeance as the ones on PAN but with somehow lower selectivity. We tried to optimize the thickness of the membrane by using lower concentration of the coating solutions. We found 0.25% polymer solution could still give similar gas selectivity (CO2/CH4 or CO2/N2) and much high gas permeance (Table 2). 30 psig feed pressure was used in the new membrane tests. Task 2: Long Term Test of Membranes Long term stability of the membrane is another important property for the biogas application since the membrane need to be exposed to corrosive raw biogas. In this task, we subjected the membranes showed good separation performance in task I to multiple days of tests. Initially we only used humidified CO2/N2 (40/60) as the surrogate gas and the membranes were just exposed to air between gas tests. The CO2 permeance of Membrane 1 declined to about half after 50 days of gas tests and the selectivity remained unchanged. It confirmed the instability of the PAN porous support. The membrane 2 (on PVDF) showed higher stability than membrane 1. The gas permeance declined initially and then stabilized after 40 days of tests. The gas permeance and selectivity remained almost unchanged between day 40 and day 80. The membrane 3 (CMS72-Li on PVDF) showed the best stability among all the membranes tested. The gas permeance only declined 16% after 130 days of aging. Task 3. Scale-up Composite Membranes The scale-up flat sheet coating method has been well developed in CMS (Figure 4) and we coated biogas membranes from the CMS50 and CMS72 polymers. After the coating, first step to test the scale-up membrane sheet is to punch multiple small samples (47 mm diameter disk) and test the gas permeance. Most of the scale-up membranes showed good initial performance and homogeneity. We only used PVDF as the porous support for the scale-up membranes. Task 4. Develop, Fabricate and Optimize Membrane Modules: We successfully made multiple flat sheet spiral modules (1500 cm2) from the scale-up membranes. The module gas test showed stable performance and somehow lower performance than small sample tests. The stage cut is 10-20%, much higher than small sample tests (<3%). The gas selectivity and permeance is calculated by the simple calculation method, taking average of the feed and retentate CO2 concentration as the feed side concentration and the permeate concentration as the permeate side concentration. Cross Flow Model Calculation for Spiral Flat Sheet Modules: For high stage cut flat sheet module gas separations, cross flow model is needed to calculate the membrane permeance and selectivity. We used the method from the paper of Separation and Purification Technology 1998 (13) 209-225. A. The biogas module 1 (1500 cm2 area) CO2/N2 (N2 as a surrogate gas for CH4) separation and calculation using the cross flow model discussed above: Case 1 (moderate stage cut 25% and atmospheric pressure permeate) : Feed gas 38.5% CO2 in N2, feed pressure 30 psig, permeate atmospheric pressure. Retentate flow: 3.28 L/min 26.7% CO2 Permeated flow: 0.99 L/min, 74.3% CO2 Calculation: CO2 451 GPU and CO2/N2=20.5. Case 2 (low stage cut 13% and vacuum permeate) : Feed gas 38.5% CO2 in N2, feed pressure 0.6 psig, permeate 4.3 psia pressure Retentate flow: 2.75 L/min, 30.3% CO2 Permeate flow: 436 mL/min, 83.0% CO2 Calculation: CO2 476 GPU and CO2/N2= 28.0. Case 3 (high stage cut 50% and vacuum permeate) : Feed gas 38.5% CO2 in N2, feed pressure 0.1 psig, permeate 3.4 psia pressure Retentate flow: 225 mL/min, 8.6% CO2 Permeate flow: 225 mL/min, 69.0% CO2 Calculation: CO2 360 GPU and CO2/N2= 22.5 The biogas module 1 was exposed to raw biogas for 3 weeks. During this period, it was soaked with water from biogas stream. It not only did not cause any damage to the membrane module, but also increase the performance. Case 2 showed the vacuum permeate method had slight higher module performance than the atmospheric pressure permeate separation. Case 3 gave us an example that 38.5% CO2 in mixture can be reduced to less than 10% (8.6%) by the vacuum permeate method in one stage. B. Biogas Module 2, 0.3% CMS72-Li on PVDF, 1500 cm2 spiral module, 30-40% stage cut 30 psig feed pressure, 42.3% CO2 in N2 and atmospheric pressure permeate, No pre-humidification CO2 297 GPU and CO2/N2= 11.5 After humidified gas exposure for 2 hoursCO2 540 GPU and CO2/N2= 14.3 Water soaked for 2 weeks CO2 897 GPU and CO2/N2= 13.9 Without sufficient hydration, the biogas membrane underperformed at the module level. Compared with the small sample and super low stage cut test results, the gas permeance of the non-humidified module is only 20% of the former (300 GPU vs. 1500 GPU). Long term water soaking can give better humidification and significant performance improvement to the biogas membrane module. Task 5. Build and Run the Pilot Test System We are still working on the membrane module performance improvement and the pilot test system will start as soon as we can better manage the module humidification problem and further optimize the module performance. Task 6. Run Long Term Field Test-Wanner Farm Biogas exposure test of the modules was done at the Wanner farm. Biogas modules were exposed to the raw biogas stream (containing 3000 ppm H2S and saturated with water) directly. As I mentioned in Task 4: after 3 weeks, the gas selectivity of the module remained same but the CO2 permeance almost doubled. We believe it was due to the water soaking, later we opened the module and found it contained significant amount of liquid water. The membrane was sufficiently hydrated and showed the maximum permeability. On the other hand, it showed the stability of the module to the raw biogas.
Publications
|
|