基于分解算法的混合气体多级膜分离系统全局优化

doi:10.11949/0438-1157.20250261

摘要/Abstract

摘要：

混合气体多级膜分离是一种高效的分离技术，通过多级膜系统的协同作用实现混合气体中各组分的高效分离。但对其进行优化设计时需要建立复杂的混合整数非线性规划（MINLP）模型，求解困难。本文提出了一种混合气体多级膜分离系统的全局优化方法，使用分解算法将复杂的MINLP问题分解为混合整数规划（MIP）问题和非线性规划（NLP）问题，通过MIP模型枚举分离序列，引入多线程并行计算方法，将各线程中的分离序列代入两个不同精度的NLP模型中序贯优化，综合所有优化结果后确定全局最优解。通过天然气脱硫案例得到的膜分离系统年总成本（tac）较文献最优结果降低7.35%，通过高炉煤气中捕集CO₂案例验证了该方法可拓展至可变压力的膜分离系统优化。

关键词: 分离, 膜, 优化设计, 超结构, 并行计算

Abstract:

Mixed gas multistage membrane separation is an efficient separation technology, which achieves efficient separation of various components in mixed gas through the synergistic effect of multistage membrane system. However, optimizing its design requires establishing complex Mixed-integer nonlinear programming (MINLP) models, which are difficult to solve. This paper presents a global optimization method for multi-stage membrane separation systems for gas mixture. Using a decomposition algorithm, the complex MINLP problem is decomposed into a mixed-integer programming (MIP) problem and two nonlinear programming (NLP) problems. The separation sequences are enumerated via the MIP model, followed by the implementation of a multithreaded parallel computing approach where each thread sequentially optimizes the assigned sequences through two NLP models of distinct accuracy levels. The global optimal solution is determined by integrating all optimization results. Through a case study of natural gas sweetening, the total annual cost (tac) of the membrane separation system obtained is 7.35% lower than the optimal result in the literature. In addition, a case of CO₂ capture from blast furnace gas verifies that the method can be extended to the optimization of membrane separation system with variable pressure.

Key words: separation, membrane, optimization design, superstructure, parallel computing

中图分类号:

TQ 028.1

王杰, 林渠成, 张先明. 基于分解算法的混合气体多级膜分离系统全局优化[J]. 化工学报, 2025, 76(9): 4670-4682.

Jie WANG, Qucheng LIN, Xianming ZHANG. Global optimization of mixed gas multistage membrane separation system based on decomposition algorithm[J]. CIESC Journal, 2025, 76(9): 4670-4682.

图/表 13

图1 多级膜分离系统的超结构

Fig.1 Superstructure of multistage membrane separation systems

图2 算法框图

Fig.2 Algorithm diagram

表1 案例1相关参数

Table 1 Related parameters of example 1

参数	数值
初始流量/(kmol/h)	36
进料压力/MPa	3.500
渗透压力/MPa	0.105
温度/K	313.15
CO₂纯度/%	2
CH₄回收率/%	90
进料组成/%（摩尔分数）	CO₂: 19, CH₄: 73,C₂₊: 7, H₂S: 1
渗透率/(kmol/(m²·MPa·h))	0.10656/0.005328/0.0021312/0.085248

表2 案例1目标函数相关参数

Table 2 Objective function-related parameters of Example 1

参数	描述	数值	参数	描述	数值
$C c p$	压缩机成本/(USD/kW)	1000	$K s g$	原始天然气成本/(USD/1000 m³)	35
$f a c ̂$	固定成本的年化因子/%	27	$K w k$	营运资本比率/%	10
$K h v$	天然气总热值/(MJ/m³)	43	$t m$	膜使用寿命/a	3
$K m$	膜成本/(USD/m²)	200	$t o p$	年工作时间/(d/a)	300
$K m r$	膜更换费用/(USD/m²)	90	$η'$	压缩机效率/%	70
$K m t$	维修率/(%/a)	5

表2 案例1目标函数相关参数

Table 2 Objective function-related parameters of Example 1

参数	描述	数值	参数	描述	数值
$C c p$	压缩机成本/(USD/kW)	1000	$K s g$	原始天然气成本/(USD/1000 m³)	35
$f a c ̂$	固定成本的年化因子/%	27	$K w k$	营运资本比率/%	10
$K h v$	天然气总热值/(MJ/m³)	43	$t m$	膜使用寿命/a	3
$K m$	膜成本/(USD/m²)	200	$t o p$	年工作时间/(d/a)	300
$K m r$	膜更换费用/(USD/m²)	90	$η'$	压缩机效率/%	70
$K m t$	维修率/(%/a)	5

图3 案例1最优膜分离系统[24]

Fig.3 Optimal membrane separation system of Example 1[24]

表3 案例1优化结果及文献对比

Table 3 Optimization results and literature comparison of Example 1

项目	Qi和Henson^[17]	Ramírez-Santos等^[22]	Taifan和Maravelias^[24]	本研究
第一级膜面积/m²	182.75	127.65	140.06	174.75
第二级膜面积/m²	197.92	181.08	151.40	111.58
第三级膜面积/m²	13.33	14.00	10.63	6.00
总膜面积/m²	394.00	322.73	302.09	292.33
压缩机功率/kW	12.61	12.87	10.66	7.31
CH₄纯度/%	88.71	89.09	88.84	88.75
年总成本/(USD/1000 m³)	10.971	9.095	8.501	7.876
求解时间/s	—	—	1185	698

表4 案例1优化结果与PRO Ⅱ模拟结果对比

Table 4 Comparison of optimization results of Example 1 with simulation results of PRO Ⅱ

项目	$F 1$	$F 1 R T O T - F 2$	$F 1 P T O T$	$F 2 R T O T$	$F 2 P T O T - F 3$	$F 3 R T O T$	$F 3 P T O T$
本研究	38.24	29.53	8.71	26.65	2.88	2.24	0.64
PRO Ⅱ	38.24	29.46	8.78	26.59	2.87	2.24	0.63
误差/%	0	0.24	-0.80	0.23	0.35	0	1.59

表4 案例1优化结果与PRO Ⅱ模拟结果对比

Table 4 Comparison of optimization results of Example 1 with simulation results of PRO Ⅱ

项目	$F 1$	$F 1 R T O T - F 2$	$F 1 P T O T$	$F 2 R T O T$	$F 2 P T O T - F 3$	$F 3 R T O T$	$F 3 P T O T$
本研究	38.24	29.53	8.71	26.65	2.88	2.24	0.64
PRO Ⅱ	38.24	29.46	8.78	26.59	2.87	2.24	0.63
误差/%	0	0.24	-0.80	0.23	0.35	0	1.59

表5 案例2相关参数

Table 5 Related parameters of Example 2

参数	数值
初始流量/(mol/s)	1240.0
初始进料流股压力/bar	1.00
温度/K	308.15
N₂纯度/%	1
CO₂回收率/%	90
进料组成/%（摩尔分数）	CO₂: 23.2, CO: 22.6,N₂: 50.3, H₂: 3.9
渗透率/(mol/(m²·bar·s))	1000/20/15/85

表6 案例2目标函数相关参数

Table 6 Objective function-related parameters of Example 2

参数	描述	数值	参数	描述	数值
$f a c ̂$	固定成本的年化因子/%	8.54	$M F c p$	压缩机材料因子	1.4
$C c p$	压缩机成本/EUR	10⁶	$I C F$	间接成本因素	1.8
$C v p$	真空泵成本因子/(EUR/kW)	1500	$t o p$	膜每年使用时间/(h/year)	8322
$C O E ̂$	电力成本/(EUR/(kW·h))	0.044	$U F 2000$	更新因子	1.42
$K m$	膜成本/(EUR/m²)	40	$γ$	绝热比	1.36
$K m f$	基础框架成本/EUR	286×10³	$η$	压缩机等熵效率/%	85
$K m r$	膜更换费用/(EUR/m²)	25	$λ$	真空泵等熵效率/%	85
$M C O 2$	CO₂摩尔质量/(g/mol)	44.01	$ν$	膜年更换率/%	20
$M D F c p$	压缩机模块因子	2.72	$φ$	机械效率/%	95

表6 案例2目标函数相关参数

Table 6 Objective function-related parameters of Example 2

参数	描述	数值	参数	描述	数值
$f a c ̂$	固定成本的年化因子/%	8.54	$M F c p$	压缩机材料因子	1.4
$C c p$	压缩机成本/EUR	10⁶	$I C F$	间接成本因素	1.8
$C v p$	真空泵成本因子/(EUR/kW)	1500	$t o p$	膜每年使用时间/(h/year)	8322
$C O E ̂$	电力成本/(EUR/(kW·h))	0.044	$U F 2000$	更新因子	1.42
$K m$	膜成本/(EUR/m²)	40	$γ$	绝热比	1.36
$K m f$	基础框架成本/EUR	286×10³	$η$	压缩机等熵效率/%	85
$K m r$	膜更换费用/(EUR/m²)	25	$λ$	真空泵等熵效率/%	85
$M C O 2$	CO₂摩尔质量/(g/mol)	44.01	$ν$	膜年更换率/%	20
$M D F c p$	压缩机模块因子	2.72	$φ$	机械效率/%	95

图4 案例2最优膜分离系统[22]

Fig.4 Optimal membrane separation system of Example 2[22]

表7 案例2优化结果及文献对比

Table 7 Otimization results and literature comparison of Example 2

项目	Ramírez-Santos等^[22]	恒压	变压
第一级膜面积/ m²	105192	92072	106672
第二级膜面积/ m²	3058	7333	7476
第三级膜面积/ m²	7229	3774	4490
总膜面积/ m²	115479	103179	118638
进料压力/bar	2.29	2.29	2.14
第一级渗透压力/bar	0.20	0.20	0.20
第二级渗透压力/bar	0.71	0.71	0.70
第三级渗透压力/bar	0.43	0.43	0.41
总功率/MW	—	7.67	7.53
年总成本/(EUR/t)	28.9	27.0	26.9
求解时间/s	—	16129	30172

表8 案例2在恒压下优化结果与PRO Ⅱ模拟结果对比

Table 8 Comparison of optimization results of Example 2 under constant pressure with simulation results of PRO Ⅱ

项目	$F 1$	$F 1 R T O T$	$F 1 P T O T - F 2$	$F 2 R T O T - F 3$	$F 2 P T O T$	$F 3 R T O T$	$F 3 P T O T$
本研究	1398	971	427	235	192	158	77.1
PRO Ⅱ	1404	975	429	240	189	164	75.6
误差/%	-0.4	-0.4	-0.5	-2.1	1.6	-3.7	2.0

表8 案例2在恒压下优化结果与PRO Ⅱ模拟结果对比

Table 8 Comparison of optimization results of Example 2 under constant pressure with simulation results of PRO Ⅱ

项目	$F 1$	$F 1 R T O T$	$F 1 P T O T - F 2$	$F 2 R T O T - F 3$	$F 2 P T O T$	$F 3 R T O T$	$F 3 P T O T$
本研究	1398	971	427	235	192	158	77.1
PRO Ⅱ	1404	975	429	240	189	164	75.6
误差/%	-0.4	-0.4	-0.5	-2.1	1.6	-3.7	2.0

表9 案例2在变压下优化结果与PRO Ⅱ模拟结果对比

Table 9 Comparison of optimization results of Example 2 under variable pressure with simulation results of PRO Ⅱ

项目	$F 1$	$F 1 R T O T$	$F 1 P T O T - F 2$	$F 2 R T O T - F 3$	$F 2 P T O T$	$F 3 R T O T$	$F 3 P T O T$
本研究	1417	971	446	269	177	177	92.1
PRO Ⅱ	1424	975	449	275	174	184	90.8
误差/%	-0.5	-0.4	-0.7	-2.2	1.7	-3.8	1.4

表9 案例2在变压下优化结果与PRO Ⅱ模拟结果对比

Table 9 Comparison of optimization results of Example 2 under variable pressure with simulation results of PRO Ⅱ

项目	$F 1$	$F 1 R T O T$	$F 1 P T O T - F 2$	$F 2 R T O T - F 3$	$F 2 P T O T$	$F 3 R T O T$	$F 3 P T O T$
本研究	1417	971	446	269	177	177	92.1
PRO Ⅱ	1424	975	449	275	174	184	90.8
误差/%	-0.5	-0.4	-0.7	-2.2	1.7	-3.8	1.4

参考文献 34

[1]	Godin J, Liu W Z, Ren S, et al. Advances in recovery and utilization of carbon dioxide: a brief review[J]. Journal of Environmental Chemical Engineering, 2021, 9(4): 105644.
[2]	Yu Y F, Yang X, Chen J, et al. Spontaneous and rapid self-healing ionogels membrane based on dual dynamic crosslinking networks strategy for high-efficiency CO₂ separation[J]. Separation and Purification Technology, 2025, 362: 131916.
[3]	NOAA Global Monitoring Laboratory. Trends in CO₂ - NOAA Global Monitoring Laboratory [EB/OL]. (2024-07-15) [2025-03-14]. .
[4]	IEA. CO₂ emissions in 2023[R/OL]. IEA: Paris, 2024. .
[5]	Li G Q, Kujawa J, Knozowska K, et al. The advancements in mixed matrix membranes containing functionalized MOFs and 2D materials for CO₂/N₂ separation and CO₂/CH₄ separation[J]. Carbon Capture Science & Technology, 2024, 13: 100267.
[6]	Faruque Hasan M M, Zantye M S, Kazi M K. Challenges and opportunities in carbon capture, utilization and storage: a process systems engineering perspective[J]. Computers & Chemical Engineering, 2022, 166: 107925.
[7]	Yousef A M, El-Maghlany W M, Eldrainy Y A, et al. New approach for biogas purification using cryogenic separation and distillation process for CO₂ capture[J]. Energy, 2018, 156: 328-351.
[8]	Ji L, Zhang L, Zheng X, et al. Simultaneous CO₂ absorption, mineralisation and carbonate crystallisation promoted by amines in a single process[J]. Journal of CO₂ Utilization, 2021, 51: 101653.
[9]	Abd A A, Othman M R, Naji S Z, et al. Methane enrichment in biogas mixture using pressure swing adsorption: process fundamental and design parameters[J]. Materials Today Sustainability, 2021, 11: 100063.
[10]	Qian Q H, Asinger P A, Lee M J, et al. MOF-based membranes for gas separations[J]. Chemical Reviews, 2020, 120(16): 8161-8266.
[11]	Li G Q, Kujawski W, Válek R, et al. A review - the development of hollow fibre membranes for gas separation processes[J]. International Journal of Greenhouse Gas Control, 2021, 104: 103195.
[12]	Hong T, Li Y, Wang S J, et al. Polyurethane-based gas separation membranes: a review and perspectives[J]. Separation and Purification Technology, 2022, 301: 122067.
[13]	Pasichnyk M, Stanovsky P, Polezhaev P, et al. Membrane technology for challenging separations: removal of CO₂, SO₂ and NO _x from flue and waste gases[J]. Separation and Purification Technology, 2023, 323: 124436.
[14]	Verma H, Jassby D, Maravelias C T. Superstructure-based optimization of membrane network systems for multicomponent liquid mixture separation[J]. Journal of Membrane Science, 2025, 717: 123574.
[15]	Samei M, Raisi A. Multi-stage gas separation process for separation of carbon dioxide from methane: modeling, simulation, and economic analysis[J]. Chemical Engineering and Processing - Process Intensification, 2022, 170: 108676.
[16]	Lindqvist K, Roussanaly S, Anantharaman R. Multi-stage membrane processes for CO₂ capture from cement industry[J]. Energy Procedia, 2014, 63: 6476-6483.
[17]	Qi R H, Henson M A. Membrane system design for multicomponent gas mixtures via mixed-integer nonlinear programming[J]. Computers & Chemical Engineering, 2000, 24(12): 2719-2737.
[18]	Arias A M, Mussati M C, Mores P L, et al. Optimization of multi-stage membrane systems for CO₂ capture from flue gas[J]. International Journal of Greenhouse Gas Control, 2016, 53: 371-390.
[19]	Chavez Velasco J A, Tumbalam Gooty R, Tawarmalani M, et al. Optimal design of membrane cascades for gaseous and liquid mixtures via MINLP[J]. Journal of Membrane Science, 2021, 636: 119514.
[20]	Ohs B, Lohaus J, Wessling M. Optimization of membrane based nitrogen removal from natural gas[J]. Journal of Membrane Science, 2016, 498: 291-301.
[21]	Gilassi S, Taghavi S M, Rodrigue D, et al. Optimizing membrane module for biogas separation[J]. International Journal of Greenhouse Gas Control, 2019, 83: 195-207.
[22]	Ramírez-Santos Á A, Bozorg M, Addis B, et al. Optimization of multistage membrane gas separation processes. Example of application to CO₂ capture from blast furnace gas[J]. Journal of Membrane Science, 2018, 566: 346-366.
[23]	Chiwaye N, Majozi T, Daramola M O. On optimisation of N₂ and CO₂-selective hybrid membrane process systems for post-combustion CO₂ capture from coal-fired power plants[J]. Journal of Membrane Science, 2021, 638: 119691.
[24]	Taifan G S P, Maravelias C T. Generalized optimization-based synthesis of membrane systems for multicomponent gas mixture separation[J]. Chemical Engineering Science, 2022, 252: 117482.
[25]	Taifan G S P, Maravelias C T. Integrated membrane material design and system synthesis[J]. Chemical Engineering Science, 2023, 269: 118406.
[26]	Chen Z W, Tawarmalani M, Agrawal R. Global minimization of power consumptions for multicomponent gas membrane cascades[J]. Computers & Chemical Engineering, 2024, 180: 108464.
[27]	Bocciardo D, Ferrari M C, Brandani S. Modelling and multi-stage design of membrane processes applied to carbon capture in coal-fired power plants[J]. Energy Procedia, 2013, 37: 932-940.
[28]	Uppaluri R V S, Smith R, Linke P, et al. On the simultaneous optimization of pressure and layout for gas permeation membrane systems[J]. Journal of Membrane Science, 2006, 280(1/2): 832-848.
[29]	Liao Z W, Hu Y X, Wang J D, et al. Systematic design and optimization of a membrane-cryogenic hybrid system for CO₂ capture[J]. ACS Sustainable Chemistry & Engineering, 2019, 7(20): 17186-17197.
[30]	Chang C L, Liao Z W, Costa A L H, et al. Globally optimal synthesis of heat exchanger networks. Part Ⅱ: non-minimal networks[J]. AIChE Journal, 2020, 66(7): e16264.
[31]	Merkel T C, Lin H Q, Wei X T, et al. Power plant post-combustion carbon dioxide capture: an opportunity for membranes[J]. Journal of Membrane Science, 2010, 359(1/2): 126-139.
[32]	Yuan M Y, Narakornpijit K, Haghpanah R, et al. Consideration of a nitrogen-selective membrane for postcombustion carbon capture through process modeling and optimization[J]. Journal of Membrane Science, 2014, 465: 177-184.
[33]	Zhai H B, Rubin E S. Techno-economic assessment of polymer membrane systems for postcombustion carbon capture at coal-fired power plants[J]. Environmental Science & Technology, 2013, 47(6): 3006-3014.
[34]	Zito P F, Brunetti A, Barbieri G. Multi-step membrane process for biogas upgrading[J]. Journal of Membrane Science, 2022, 652: 120454.