Available online at https://ijcpe.edu.iq and www.iasj.net Iraqi Journal of Chemical and Petroleum Engineering Vol.19 No.1 (March 2018) 61 – 66 ISSN: 1997-4884 Corresponding Authors: Zaid Adnan Abdel-Rahman, Email: zaid572010@yahoo.com Iraqi Journal of Chemical and Petroleum Engineering A Comparison between the Product-Refill and the Equalization Oxygen Pressure Swing Adsorption Processes Zaid Adnan Abdel-Rahman Chemical Engineering Department, College of Engineering, Tikrit University, Iraq Abstract This work presents a design for a pressure swing adsorption process (PSA) to separate oxygen from air with approximately 95% purity, suitable for different numbers of columns and arrangements. The product refill PSA process was found to perform 33% better (weight of zeolite required or productivity) than the pressure equalization process. The design is based on the adsorption equilibrium of a binary mixture of O2 and N2 for two of the most commonly used adsorbents, 5A & 13X, and extension from a single column approach. Zeolite 13X was found to perform 6% better than zeolite 5A. The most effective variables were determined to be the adsorption step time and the operational pressure. Increasing the adsorption step time from 1 to 5 minutes decreased the performance by approximately 5 times. Increasing the pressure from 2 to 6 bars increased the performance by approximately 3 times. Keywords: Oxygen production, PSA, process design, product refill, equalization 1- Introduction The pressure swing adsorption (PSA) process for oxygen separation from air is an important alternative unit operation in the chemical industry. It has advantages in energy consumption and cost over other methods, but is only used in approximately 20% of oxygen production especially on the small-to-moderate scale. However, the cryogenic process is currently used in approximately 80% of all oxygen production on the moderate-to-large scale. The analysis of previous works over the last three decades regarding the PSA process for oxygen separation from air using one-, two-, three-, and four-columns with different steps and arrangements allowed for the design of a process that produces high-purity oxygen with reasonable recovery and productivity. Two main arrangements of the PSA process were found: the product refill process and the pressure equalization process. The commercial zeolites 5A and 13X were the primary adsorbents used in these works. The end product is limited to 95% oxygen because of the presence of argon in air and because the zeolite adsorbents present have similar adsorption capacities for oxygen and argon [1-23]. The design of the PSA process was based on two principles. The first is the equilibrium between solutes in the fluid phase and the solute-enriched phase of the solid adsorbent. The understanding of the adsorbent material equilibrium properties, especially capacity as a function of pressure, is of primary importance. The second is that the fixed-bed process operates as a periodic or cyclic steady state, with several different steps constituting a cycle. Thus, knowledge of the transitional behavior of a single bed is necessary for the system design. Both time and space were used in the analysis, which is in contrast to many chemical engineering operations that can be analyzed at the steady state that only have a spatial dependence. For an optimal design, it is crucial to understand fixed- bed performance in relation to the adsorption equilibrium and rate behavior. An empirical approach is frequently used to design the PSA process, using the adsorption equilibrium capacity of the adsorbents, breakthrough time of a single column, and mass transfer zone (MTZ) approach [24-27]. The PSA process performance is represented by the oxygen purity of the product, recovery (oxygen in product to oxygen in feed) and productivity (continuous oxygen product flowrate to zeolite weight). The performance is affected by the following variables: pressure, cycle time or adsorption step time, purge flow ratio, feed flowrate, and product flowrate. No comparison study between the two PSA process arrangements, the product refill and the equalization process has been performed. The objectives of this study are to develop a simple empirical design model of the PSA process for the separation of oxygen from air that is suitable for different numbers of columns, to compare the two PSA process arrangements (product refill process and pressure equalization process), and to study the effect of system variables on the zeolite weight required and/or the process productivity for a specified function. Z. A. Abdel-Rahman / Iraqi Journal of Chemical and Petroleum Engineering91,9 (2018) 29-43 26 2- Process Design Approach In this work, the adsorption equilibrium, the single column breakthrough approach, stoichiometry and the experience of the different process steps times were used as a basis for the design of the PSA process for oxygen separation from air. 2.1. Equilibrium Isotherm The nitrogen adsorption equilibrium capacity for the adsorbents can be approximated by the following linear Henry like equation as follows: (1) The value of KZ equals 0.17 and 0.18 mole N2/kg zeolite.bar for 5A and 13X zeolites, respectively. The values were obtained by the analysis of the literature data [10-11, 23]. 2.2. Breakthrough of Single Column The single column breakthrough approach was used as a basis for the design using a factor, KB, representing the fraction of bed that was used, as shown in Fig. 1. Fig. 1. Single column breakthrough Breakthrough time was used as a guide for cycle time selection, which is approximately given by the following simple mass balance equations; ⁄ (2) ⁄ (3) 2.3. PSA System Fig. 2 shows the five PSA process cases used in present study. Fig. 2: The five PSA process cases used in present study Z. A. Abdel-Rahman / Iraqi Journal of Chemical and Petroleum Engineering91,9 (2018) 29-43 26 The five PSA process cases of the two main process arrangements were considered in this work as follows; a. Product refill or pressurizing arrangement PSA processes, 4-steps (AD-DP-PG-PP) system, were used for one-column with a product flow utilized time fraction KC=1/3, two-columns with KC=2/3 and three-columns with KC=1, as shown in Table 1. Table 1. Product refill process steps Columns Steps KC One-Columns One- Column AD AD DP PG PG PP 1/3 Two-Columns 1 st column AD AD DP PG PG PP 2/3 2 nd column DP PG PG PP AD AD Three-Columns 1 st column AD AD DP PG PG PP 1 2 nd column PG PP AD AD DP PG 3 rd column DP PG PG PP AD AD b. Equalization arrangement PSA processes, 6-steps (AD-ED-DP-PG-EP-FP) system, were used for two- columns with a product flow utilized time fraction KC=1/2 and four-columns with KC=1, as shown in Table 2. Table 2. Pressure equalization process steps Columns Steps KC Two-Columns 1 st column A D A D E D D P P G P G E P F P 1/2 2 nd column P G P G E P F P A D A D E D D P Four-Columns 1 st column A D A D E D D P P G P G E P F P 1 2 nd column E P F P A D A D E D D P P G P G 3 rd column P G P G E P F P A D A D E D D P 4 th column E D D P P G P G E P F P A D A D The total cycle time is the summation of the process steps times; ∑ (4) The 4-steps for the product refill PSA process cycle (τ=3τAD) is as follows: 1- The adsorption (AD) step was taken time τAD =1-5 minutes. 2- The depressurizing (DP) step was taken time τDP=0.5* τAD. 3- The purging (PG) step was taken time τPG= τAD. 4- The product pressurizing (PP) step was taken time τPP=0.5* τAD. The 6- steps PSA equalization process cycle (τ=4τAD) is as follows: 1- The adsorption (AD) step was taken time τAD =1-5 minutes. 2- The equalization depressurizing (ED) step was taken time τED=0.5* τAD. 3- The depressurizing (DP) step was taken time τDP=0.5* τAD. 4- The purging (PG) step was taken time τPG= τAD. 5- The equalization pressurizing (EP) step was taken time τEP=0.5* τAD. 6- The feed pressurizing (FP) step was taken time τFP=0.5* τAD. In the PSA process, the product flowrate is not actually continuous, but was assumed to be continuous to simplify the design. As a result a factor, KC, the utilized time fraction for the product flow, was introduced and calculated from the following equation; ⁄ (5) The PSA system approach introduced the product flow utilizes time fraction KC for different system arrangements, and the purge flow factor KP (fraction of product flow in a single column), as shown in Fig. 3. Fig. 3. PSA system For the PSA system, the product flowrate and productivity were calculated from the following equations; (6) (7) ⁄ (8) The recovery was difficult to calculate accurately because of intermediate steps and because the feed is not actually continuous. A case study of a PSA process of 1 kg/h production of approximately 95% oxygen purity was used in the present study. Four process variables were studied, the adsorption step time τAD (1-5 minutes), pressure P (2-6 bar), utilized bed factor KB (0.5-0.9), and purge factor KP (0.1-0.5). Z. A. Abdel-Rahman / Iraqi Journal of Chemical and Petroleum Engineering91,9 (2018) 29-43 26 3- Results And Discussion Figures (4) and (5) show the effect of the adsorption step time, zeolite type and PSA system arrangement on the process performance (weight of zeolite required or productivity), at constant values for P, KB & KP of 6 bar, 0.9, and 0.1, respectively. Zeolite 13X only had an approximately 6% better performance than 5A zeolite. This is in agreement with data obtained by Mofarahi & Shokroo [23]. The product refill 4-steps PSA process had an approximate 33% better performance than the pressure equalization 6-steps process. This is a novel point found by this work that has not been mentioned in the literature. Increasing the adsorption step time (τAD) from 1 to 5 minutes decreased the performance by an approximate factor of 5. The adsorption time was the most effective variable. The present work productivity range results (5-35 liter oxygen/kg zeolite h) are in agreement with most of the current literature. AD minute 0 1 2 3 4 5 6 m B k g 0 20 40 60 80 100 120 140 160 180 200 13X Product refill 13X Equalization 5A Product refill 5A Equalization Fig. 4. Effect of the adsorption step time, zeolite type and PSA system on zeolite weight at P=6 bar, KB=0.9 & KP =0.1. AD minute 0 1 2 3 4 5 6 P r o d u c ti v it y li te r O 2 /k g z e o li te h 0 5 10 15 20 25 30 35 13X Product refill 13X Equalization 5A Product refill 5A Equalization Fig. 5. Effect of the adsorption step time, zeolite type and PSA system on productivity at P=6 bar, KB=0.9 & KP =0.1 Figure (6) shows the effect of the operating pressure (P) on the process productivity, at constant values for τAD, KB and KP of 1 minute, 0.9, and 0.1, respectively. Increasing pressure from 2 to 6 bars increased the performance by an approximate factor of 3. The operating pressure was the most effective variable after the adsorption step time variable (τAD). This result is in agreement with several published works [19-22], whereas other works noted negative results for the effect of increasing pressure [1,9- 11,23]. This contradiction may be due to the variables being highly interacted. P bar 1 2 3 4 5 6 7 P r o d u c ti v it y li te r O 2 /k g z e o li te h 5 10 15 20 25 30 35 13X Product refill 13X Equalization Fig. 6. Effect of the operating pressure on productivity at τAD=1 minute, KB=0.9 & KP =0.1 Figure (7) shows the effect of the utilized bed factor (KB) on the process productivity at constant values for τAD, P and KP of 1 minute, 6 bars, and 0.1, respectively. Increasing the utilized bed factor from 0.5 to 0.9 approximately doubled the performance. KB 0.4 0.5 0.6 0.7 0.8 0.9 1.0 P r o d u c ti v it y l it e r O 2 /k g z e o li te .h 10 12 14 16 18 20 22 24 26 28 30 32 13X Product refill 13X Equalizaion Fig. 7. Effect of the bed volume utilized factor on productivity at τAD=1 minute, P=6 bar, & KP =0.1 Z. A. Abdel-Rahman / Iraqi Journal of Chemical and Petroleum Engineering91,9 (2018) 29-43 26 Figure (8) shows the effect of the purge factor (KP) on the process productivity at constant values for τAD, P and KB of 1 minute, 6 bars, and 0.9 respectively. Increasing the purge factor from 0.5 to 0.9 decreased the performance by approximately half. KP 0.0 0.1 0.2 0.3 0.4 0.5 0.6 P r o d u c ti v it y l it e r O 2 /k g z e o li te h 10 12 14 16 18 20 22 24 26 28 30 32 13X Product refill 13X Equalization Fig. 8: Effect of the purge factor on productivity at τAD=1 minute, P=6 bars, & KB =0.9 4- Conclusions 1- The product refill 4-steps PSA process had a 33% improved performance (weight of zeolite required or productivity) compared with the pressure equalization 6-steps process. This is a novel point found by the present study and has not reported in previous published works. 2- Zeolite 13X only had a 6% higher performance than 5A zeolite. 3- The most effective variables were the adsorption step time and pressure. Increasing the adsorption step time from 1 to 5 minutes decreased the performance by a factor of 5, and increasing the pressure from 2 to 6 bars increased the performance by a factor of 3. Nomenclatures Symbols KB Bed volume utilized fraction KC PSA system product flow utilized time fraction KP Purge flow fraction KZ Adsorption constant, mole N2/kg zeolite. Bar mB Total zeolite weight, kg mB1 Single column zeolite weight, kg n Number of steps N Number of columns P Pressure, bar q Equilibrium capacity, mole N2/kg zeolite QF Feed flowrate, m 3 /h QP Product flowrate, m 3 /h QP1 Single column product flowrate, m 3 /h VB Total zeolite bed volume, m 3 VB1 Single column zeolite bed volume, m 3 yF Feed oxygen fraction yP Product oxygen fraction Greek Symbols ρB Zeolite bed bulk density, kg/m 3 ρG Gas density, kg/m 3 τ Total cycle time, minute τAD Adsorption step time, minute τB Breakthrough time, minute τDP Depressurizing step time, minute τED Equalization depressurizing step time, minute τEP Equalization pressurizing step time, minute τFP Feed pressurizing step time, minute τi General step time, minute τPG Purging step time, minute τPP Product pressurizing step time, minute References [1] Farooq, S.; Ruthven, D. M.; & Boniface, H. A., Numerical simulation of pressure swing adsorption oxygen unit, chem. Eng. Sci., 44 (12), 2809-2816, 1989. [2] Ruthven, D. M.; and Farooq, S., Air Separation by Pressure Swing Adsorption, Gas Sep. Purif., 4, 141, 1990. [3] Liow J., Kenny C.N, The Backfill Cycle of the Pressure Swing Adsorption Process, AIChE J. 36, 53, 1990. [4] Zahur M., Air separation on carbon molecular sieves 4A, and 5A Zeolite by pressure swing adsorption, M.Sc. Thesis, King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia, 1991. [5] Chou, C., and Huang, W., Simulation of four bed pressure swing adsorption process for oxygen enrichment, Ind. Eng. Chem. Res., 33, 1250, 1994. [6] Rege, S.U., and Yang, R.T., Limits for Air Separation by Adsorption with LiX Zeolite, Ind. Eng. Chem. Res., 36, 5358-5365, 1997. [7] Teague K. T., and Edgar T. F., Predictive Dynamic Model of a Small Pressure Swing Adsorption Air Separation Unit, Ind. Eng. Chem. Res., 38, 3761- 3775, 1999 . [8] Shin, H.-S., Kim, D.-H., Koo, K.-K., Lee, T.-S., Performance of a Two-Bed Pressure Swing Adsorption Process with Incomplete Pressure Equalization, Adsorption 6, 233, 2000. Z. A. Abdel-Rahman / Iraqi Journal of Chemical and Petroleum Engineering91,9 (2018) 29-43 22 [9] Mendes, A. M. M.; Costa, C. A. V.; & Rodrigues, A. E., Analysis of Nonisobaric Steps in Nonlinear Bicomponent Pressure Swing Adsorption Systems, Application to Air Separation, Ind. Eng. Chem. Res., 39, 138, 2000. [10] Mendes A. M. M., Costa C. A.V., Rodrigues A. E., Oxygen Separation from Air by PSA: Modeling and Experimental Results , separation and purification technology, 24, 173, 2001. [11] Jee J. G., Lee J. S., and Lee C. H., Air Separation by a Small-Scale Two-Bed Medical O2 Pressure Swing Adsorption, Ind. Eng. Ch.. Res., 40, 3647-3658, 2001. [12] Jee, J. G.; Park, H. J. ; Haam, S. J. ; Lee, C. H. Effects of Nonisobaric and Isobaric Steps on O2 Pressure Swing Adsorption for an Aerator, Ind. Eng. Chem. Res., 41, 4383, 2002. [13] Jee J. G., Lee J. S., Moon, H. M., & Lee C. H., Adsorption dynamic of air on zeolite 13X and CMS beds for separation & purification, Adsorption, 11, 415, 2005. [14] Santos J. C., Portugal A. F., Magailaes F. D., and Mendes A., Simulation and Optimization of small Oxygen pressure swing adsorption unit, Ind. Eng. Chem. Res., 43, 8328-8338, 2004. [15] Yuwen, Z., Yuyuan, W., Jianying, G., Jilin, Z., The experimental study on the performance of a small- scale oxygen concentration by PSA, Separation and Purification Technology 42, 23–127, 2005. [16] Ahari, J. S., Pakseresht, S., Mahdyarfar, M., Shokri, S., Zamani, Y., pour, A. N. & Naderi, F., Predictive Dynamic Model of Air Separation by Pressure Swing Adsorption, chem.eng. tech., 29(1), 50-58, 2006 [17] Santos J. C., Portugal A. F., Magailaes F. D., and Mendes A, Optimization of Medical PSA Units for Oxygen Production. Ind., Eng., Chem., Res. 45, 1085, 2006. [18] Zahra, M.; Jafar, T., & Masoud, M., Study of a Four-bed Pressure Swing Adsorption for oxygen Separation from Air, Int. J. of chem. and Biomolecular Eng.1; 3, 2008. [19] Mofarahi M., Fathi L., and Towfighi J., oxygen separation from air by four bed pressure swing adsorption, Ind. Eng. Ch. Res, 48(11)5439-5444, 2009. [20] Beeyan , A. K.; Singh, K.; Vyas, R. K.; Kumar, S.; & Kumar, S., Parametric studies and simulation of PSA process for oxygen production from air, polish J. of chem. Tech., 12, 2, 18-28, 2010. [21] Abdel-Rahman, Z.A., Mhdi, A.H., and Ali, A.J., Two Spiral Tubes Pressure Swing Adsorption for Oxygen Separation from air, 1st National Conf. for Engineering Sciences, 7-8 November 2012. [22] Abdel-Rahman, Z.A., Mhdi, A.H., & Ali, A.J., Study the Behavior of Long Spiral Tube Adsorber for Oxygen Separation from Air, Eng. & Tech. J., 31 (17), 2471, 2013. [23] Mofarahi, M.; & Shokroo, E. J., Comparison of two adsorption processes for air separation using zeolite 5A and zeolite 13X, Petroleum & coal, 55(3), 216, 2013. [24] Smith, O. J., and Westerberg, A. W., the Optimal Design of Pressure Swing Adsorption Systems, Chem. Eng. Sci., 46, 12, 2967-2976, 1991. [25] Ruthven, D. M.; Farooq, S.; Knaebel, K. S., Pressure Swing Adsorption, VCH Publishers, New York, 1994. [26] Perry, R. H., Green, D. W., Maloney, J. O., Chemical Engineers handbook, McGraw-Hill, 1997. [27] Crittenden, B. and Thomas, W.J., Adsorption Technology & Design, Butterworth-Heinemann, Oxford, 1998.