
换热网络是能量回收利用中的一个重要系统。在石油化工生产过程中,常常会遇到某些物流需要加热,而某些物流需要冷却,如果用热流来加热冷流,这样就可以回收能量。此外,为了保证过程物流达到指定的温度要求,往往还需要设置一些辅助的加热设备和冷却设备,换热流程中的换热器、加热器、冷却器、混合器和分流器的组合便构成了换热网络。对换热网络结构已定或已在运行的换热网络,如何优化改进使其达到最优,或在换热网络中的某些过程物流的流量、进出口温度发生变化时,要求换热网络的换热仍能满足工艺上的换热要求,并能在最优或接近最优状态下运行,要解决这些问题,就提出了换热网络优化节能的问题。
现有石油化工生产中的一些老装置,由于初建时比较落后,对节能要求不高,过程物流的换热不合理,有许多需要优化改进的,有些生产工艺由于生产能力或条件的变化,也对已有换热网络产生影响,也需对已有换热网络进行调整改进。所以换热网络优化改进的应用面是非常广泛的,也是非常重要的。无论设计一个新的换热网络还是对已有换热网络进行改进,都要用到换热网络的优化合成技术、换热网络分析技术、换热器优化设计技术、换热网络流程模拟技术、换热网络灵敏度分析、换热网络弹性分析等技术。
第1章,绪论,阐述了节能对我国国民经济建设的重要性,对国内外在换热网络的节能领域的研究状况进行了介绍。
第2章,智能法合成最优换热网络,通过对换热网络的热力学分析,给出能量体用量与换热网络有效能损失的关系,由此得到合成最大热回收量网络的物流匹配换热规则1;通过数学推导,给出换热网络中各换热设备的传热温差分布与网络总换热面积的关系,由此得到在热回收量一定条件下,合成最小换热面积网络的物流匹配换热规则2。此规则与规则1相同。这两个规则是从合成最优换热网络的两个不同寻优目标得到的,而且相同,得到合成最优换热网络的理论基础,
并得出这样的结论:热回收量最大和在热回收量最大条件下的最小换热面积网络可以同时合成。利用夹点概念,给出解析法求换热网络可能的最大热回收量 \(Q_{\mathrm{max}}\) 和量小能量体用量的数学模型,给出换热网络中实际回收量 \(Q_{r}\) 和能量体用量( \(Q_{\mathrm{cu}}\) 或 \(Q_{\mathrm{th}}\) )的极限。为衡量合成的换热网络的优劣提出了一个标准。依据规则,考虑网络结构的合理性,由实际合成经验,开发出一套合成最优换热网络的人工智能图表,并给出利用此图表合成最优换热网络的步骤,解决了解析法合成最优换热网络的问题。对本章方法编制了FORTRAN- 77实用计算机程序,计算了许多实例,与文献的计算结果作了比较,令人满意。使用本章方法可以合成物流有相变的换热网络;在合成过程中,根据需要,可以合成允许物流分流或不允许物流分流的换热网络;根据工程要求,也可以限制过程中的某些冷、热物流之间相互匹配换热。本章对使用多种目标函数进行了研究,发现以最大热回收量和在最大热回量条件下同时使传热设备数最少为目标函数(obj2)与以年费用最小为目标函数所得的最优换热网络是相同的。所以在合成最优换热网络时,当年费用(obj3)不能计算时,以obj2作目标函数同样可以得到最优网络。本章对最小允许接近温度 \(t_{\mathrm{min}}\) 的取值也进行了研究,发现 \(t_{\mathrm{min}}\) 在 \(5\sim 20^{\circ}\mathrm{F}\) 之间取值均能得到好的网络。本章方法与前人工作进行了比较,本章开发的方法吸取了前人方法的优点,克服了前人方法的缺点,使最优换热网络的合成问题大大简化。
第3章,换热网络优化分析,建立了换热网络优化分析的非线性规划(NLP)的数学模型,可对结构已定的换热网络中各换热单元的进出口温度进行优化,如果物流有分流,同时可优化各分流流量。利用复合形Box方法和线性规划方法的组合求解,成功地解决了换热网络的NLP问题。
第4章,换热器优化设计,建立了标准系列换热器优化设计的数学模型和优化设计方法,采用“智能”方法调整换热器的结构参数,设计出在满足换热要求条件下的具有最小换热面积的换热器。针对常减压油品换热,建立了物性数据库,同时建立了标准系列换热器结构参数数据库,使整个优化设计过程在输入尽可能少的数据下自动完成。另外在软件的编制中增加了换热器核算和进出口温度求解的功能。
第5章,换热网络流程优化模拟,建立了换热网络流程优化模拟的数学模型,即在换热流程模拟过程中,如果物流有分流,对各股分流的流量进行最优化,使换热网络回收最大的热量。换热网络流程优化模拟也是一个非线性规划(NLP)问题,对各路分流的流量采用复合形Box方法进行优化,经过优化模拟后获得的操作参数可使换热网络获得更大的热回收量。在换热网络流程模拟过程中,采用序贯模块法进行求解,首先采用系统分解方法把换热网络分解成可以独立求解的子系统,并排出计算顺序,然后按顺序一一求解。
第6章,换热网络的灵敏度分析,提出了一个通用的计算任意结构换热网络灵敏度系数及灵敏度的数学模型,并给出了求解数学模型的方法。编制了计算机程序并用实例进行了考核计算。换热网络的灵敏度及灵敏度系数对分析干扰变量对控制变量和目标函数的影响是非常重要的。根据换热网络的灵敏度系数,可对换热网络的操作性能进行分析,对换热网络提出最佳改进方案。
第7章,换热网络的弹性分析,提出了一个换热网络弹性分析的有效方法,该方法以换热网络过程模拟为基础并把灵敏度系数用于过程模拟的调节和加速收敛的过程中。换热网络的模拟过程模仿了一个过程的调节控制过程,在此过程中,具有干扰变量、控制变量和调节变量,干扰变量可取不同的值,控制变量也可以重新设置。整个弹性分析的过程(包括弹性指数的计算和固定状态的可行性分析)完全象是一个实际过程的操作,回避了以往复杂的数学规划求解问题。本法求解过程具有真实的物理意义,并编制了通用计算机程序并用实例进行了检验。
第8章,换热网络的优化改进和控制,通过对换热网络的定性分析,提出了换热网络优化改进的四条规则,即,(1)控制变量应对调节变量反应灵敏;(2)选择的调节变量数应小于或等于控制变量数;(3)一组控制变量应有一组调节变量相匹配;(4)如果在标准操作条件下现有换热网络是最佳的,而且改进过程中对它作最小的改变就使换热网络适合当前的或多周期操作条件,那么该改进费用一定是最小的,因此这种改进就是最佳的。根据上述规则,结合换热网络的灵敏度分析,建立了换热网络优化改进及控制的数学模型和求解方法,给出了换热网络优化改进的步骤,通过实例计算及应用,证明本章所提方法正确、可靠、有效,可在解决实际工程问题中使用。
第9章,多变量控制回路的识别,基于灵敏度系数定义了一个关系矩阵,根据关系矩阵提出了识别控制回路的方法,介绍了控制回路表和控制回路级。基于控制回路级分析了控制回路的相互影响,并举例说明如何使用本方法。
第10章,换热网络优化节能技术的应用,应用本书提出的换热网络优化节能技术,对中石油和中石化的60多套常减压装置的换热网络进行了优化节能研究,提出了改造方案。本书以某厂的常减压换热网络为例,阐述本书某些技术的应用过程。对某厂的常减压换热网络通过换热网络的合成、分析、换热器优化设计及流程模拟,找出了原有换热网络存在的问题,提出了对原有换热网络优化改进的方案,并进行了优化设计。改进后换热网络的初底油换后温度由原来的 \(272^{\circ}C\) (标定数据)提高到 \(308^{\circ}C\) ,每年直接节省燃料油7868吨,年节能效益为767万元,设备投资200万元,改造投资回收期为3.1个月,节能效益十分可观,为某厂常减压换热网络的节能技术改造提供了可靠的技术依据,1997年在装置上实施,本书技术也分别在吉化第二常减压换热网络的优化节能改造和齐鲁石化胜利炼油厂第三常减压换热网络的优化扩产节能技术改造中得到应用,每年分别获的上千万元的节能效益。
以上模型及方法均用Fortran77语言编程,数据及信息输入窗口用Visual \(\mathrm{C + + }\) 在Windows下编制,人机界面友好,使用方便,结果表格输出,各软件均用实例进行认真考核,结果准确可靠,并在工程技术研究和改造过程中得到应用,取得了很大的经济效益。
总之,本书建立的数学模型、求解方法和计算机软件可在工程研究及设计中推广应用。
关键词:换热网络 换热器 传热 节能 过程模拟 优化设计 控制 灵敏度分析 优化
ABSTRACT
Heat Exchanger Network (HEN) is an important subsystem of energy recovery process. In chemical production, some streams need to be heated, while others need to be cooled. If the hot streams are used to heat the cold streams, the energy can be recovered. In order to ensure the streams reach the designated temperatures, some auxiliary heating devices and cooling devices are needed. So the combination of heat exchangers, heaters, coolers, mixers and splitters in heat exchange process constitute the heat exchanger network. The retrofit of heat exchanger network is how to improve a heat exchanger network with a certain structure or operated and make it to be optimal while the flows and the inlet and outlet temperatures of the streams are changed and the heat exchanger network can also meet the technological demand and operation under the optimum condition or close to optimum.
For some old devices in chemical production, the heat exchanger network need to be improved because of inefficiency of heat exchange for its backward technological conditions or influences of change of productivity and technological conditions. So the application of the HEN retrofit is wide and important. Whatever designing a new HEN or improving an old one, the optimal synthesis technology, analysis technology, optimal design technology of heat exchangers, process simulation technology of HEN, and so on, are needed.
In Chapter 2, the relationship of utility usage with the available energy lost of HEN by thermodynamic analysis is set up. The rule 1 of matches between cold and hot streams is obtained from the relationship for synthesizing HEN with maximum heat recovery. By the mathematics derivation, the relationship that the distribution of heat transfer temperature differences of heat exchange units in HEN affect on total heat transfer areas of the networks is given. From this rule 2 of matches between cold and hot streams is obtained for synthesizing minimum heat transfer areas with the heat recovery being constant. The contents of the rule 2 are as the same as the rule 1. Because they are obtained from two different optimizing objectives and are the same, they are theoretical basis of synthesis of optimum HEN. Hence, we come into the conclusion that the HEN with maximum heat recovery and HEN with minimum heat transfer areas under maximum heat recovery can be synthesis at the same time. On the basis of the rule 1 and the rule 2 and for the sake of the reasonableness of HEN structure, a set of artificial intelligent graphs is developed to synthesizing optimal HEN by the practical experience. The procedures of synthesizing optimal HEN are given with the graphs. Solving the optimal HEN can be synthesized analytically. The computer programs completed for the method. Many examples are tested, and the results are compared with previous authors' and satisfactory. Networks with streams can be synthesized with the method. According to process requirement, some hot streams are also restricted to be matched with some cold streams. Three objective functions used in synthesis are studied. We discover that optimal HEN obtained using an objective function (obj2) making heat recovery maximum and the number of heat transfer units minimum under the condition of maximum heat recovery is the same as using another objective function (obj3) making annual cost minimum. When optimal HEN is synthesized and obj3 can not be calculated, it can be synthesized using obj2 as an objective function. Studying minimum allowable approach temperature (tmin), we discover that the best HEN can be synthesized when gets a value between \(5\sim 20^{\circ}\mathrm{F}\).
In Chapter 3, the mathematical model of non- linear planning for optimal analysis of HEN is set up. The model can optimize the inlet and outlet temperatures of each heat exchange units in HEN with a certain structure, and if the streams are split, it can optimize the flow of every branch at the same time. By using the combination of complex BOX method and linear programming method, the NLP problem of the HEN analysis can be solved successfully.
In Chapter 4, the mathematical model and optimal design method are established for the standard series of the heat exchangers. The artificial intelligent method is used to regulate the structure parameters of the heat exchanger and design the heat exchangers with minimum heat exchange area under the designated conditions. The database of oil product properties and structure parameters of the standard series heat exchanger are set up in order to make the whole optimal design finish automatically. In the computer software, the functions of checking heat exchangers and solution of the inlet and outlet temperatures of the streams passing the heat exchangers are added.
In Chapter 5, the mathematical model of optimal simulation of the HEN is established. In the simulation of heat exchange process, if the streams have branches, the flow of every branch is optimized in order to recover the maximum energy. The optimal simulation of the HEN is a NLP problem. The flow of every branch is optimized with the complex BOX method, and the operation parameters after simulation can make the HEN recover more energy. In the simulation of the HEN, a sequential model method is used. First, the HEN is decomposed into subsystems which can be solved independently with the systematic decomposition method, and arranged their computing order, then solved in sequence.
In Chapter 6, a common mathematical model of calculation of sensitivity coefficients and sensitivities of heat exchanger networks with various configurations is proposed, and the method to solve the model is given. The computer program has been made and tested by the example. The mathematical formulations used in the calculation can be formed automatically by the program based on the basic input information about the structure of HEN. Necessary data used in theoretical analysis of operability of HEN can be provided.
In Chapter 7, a simple and effective approach to calculation of flexibility index (CFI) of HEN is proposed based on process simulation of HEN with help of sensitivity coefficients. The computer program is made. If there are manipulated variables in the CFI, when disturbance variables (DVs) change from their normal values and induce the change in the controlled variables (CVs) from their setpoints, the manipulated variables (MVs) will be adjusted in order to make CVs return to their setpoints. The curves of derivations of CVs from their setpoints vs. the number of adjustment of MVs can be drawn on the screen. Flexibility test at a fixed state can be done.
In Chapter 8, a model of selection of optimal manipulated variables and retrofitted parameters of heat exchanger networks (HEN) is proposed based on sensitivity analysis. The procedure of the retrofit of HEN is given. Two examples are done in order to illustrate the use of the approach.
In Chapter 9, a relation matrix is defined based on sensitivity coefficients. An approach to identification of control loops is proposed according to the relation matrix. Control loop table and control loop orders are introduced. Effect of control loops on each other is analyzed based on the control loop order. Two examples are used to illustrate the usage of the approach.
In Chapter 10, the software package for HEN are applied to make an optimal energy saving research for HEN in the plant of the refinery of JCIC and find out some problems in the existing HEN. After making synthesis of HEN, analysis of HEN, the methods are proposed for optimal design and process simulation of Heat exchangers in optimally retrofitting existing HEN. The leaving HEN temperature of the oil from the bottom of the initial column is risen from \(272^{\circ}C\) (determined data) to \(308^{\circ}C\) . The fuel oil saved directly is 7878 ton per year, benefit of energy saving is 7,670,000 yuan per year, equipment investment is 2,000,000 yuan, the return time of the investment is 4 months, the benefit of energy saving technique HEN in Normal- lower pressure plant of the refinery of JCIC.
The model and method are all programmed with Fortran 77 language. The data and information input windows are programmed with Visual \(\mathrm{C + + }\) under Windows 3.x and Windows 95. The software is tested with practical examples, and the results are accurate and reliable. It is applied in the engineering research and innovation, obtaining great economic benefit.
In a word, the mathematical models, solving methods and computer software set up in this paper can be applied in the engineering research and design widely.
KEY WORDS: Heat exchanger networks. Heat exchangers. Heat transfer. Energy saving. Process simulation. Optimal design. Sensitivity analysis
摘要 ABSTRACT. V
第1章 绪论. 1
1.1 前言. 1
1.2 文献综述. 2
1.2.1 换热网络的优化合成. 2
1.2.2 换热网络的分析. 9
1.2.3 换热网络的流程模拟. 9
1.2.4 换热网络的优化改进、弹性分析及最优控制. 10
1.2.5 非线性最优化方法. 13
参考文献. 14
第2章 智能法合成最优换热网络. 17
§2.1 前言. 18
§2.2 换热网络的能量体用量与有效能损失的关系. 19
2.2.1 工艺流的子处理. 19
2.2.2 热负荷及有效能的计算. 20
2.2.3 能量体用量与系统有效能损失的关系. 21
§2.3 网络总换热面积与传热温差分布的关系. 24
§2.4 最大热回收量及最小能量体用量的计算. 26
2.4.1 最大热回收量 Qmax 的求取. 26
2.4.2 冷、热能量体最小用量 Qcmin、Qthmin 的计算. 32
§2.5 最优换热网络的合成. 32
2.5.1 合成最优换热网络的智能图表. 33
2.5.2 合成最优换热网络的步骤. 41
§2.6 换热网络合成实例. 43
§2.7 讨论. 48
2.7.1 最小允许接近温度 tmin 对合成换热网络的影响. 48
2.7.2 是否允许物流分流对合成换热网络的影响. 49
2.7.3 与前人工作的比较. 50
§2.8 结论. 52
符号说明. 52
参考文献. 54
第3章 换热网络的分析. 56
§3.1 换热网络分析要解决的问题. 56
§3.2 换热网络分析的数学模型. 56
§3.3 线性规划的数学模型. 59
3.3.1 基础模型. 59
3.3.2 求解方法[2]. 59
§3.4 复合形BOX方法的数学模型和求解方法[2]. 63
3.4.1 复合形BOX法的基本思想. 63
3.4.2 复合形法的计算步骤. 63
3.4.3 使用复合形BOX法时的技巧和注意事项. 65
§3.5 换热网络分析求解过程. 66
3.5.1 换热网络分析基础数据信息. 66
3.5.2 换热网络分析数学模型的转化. 69
§3.6 换热网络分析实例. 71
§3.7 结论. 79
参考文献. 79
第4章 换热器的优化设计. 80
§4.1 问题的提出. 80
§4.2 数学模型. 80
4.2.1 基础物性的计算[1]. 80
4.2.2 传热膜系数、总传热系数和压降的计算[2]. 84
§4.3 换热器设计、核算的基本过程. 91
4.3.1 换热器标准系列数据库的建立. 91
4.3.2 换热器最优化设计过程. 92
4.3.3 换热器核算过程. 110
§4.4 结论. 111
主要符号. 111
参考文献. 114
第5章 换热网络的流程优化模拟. 115
§5.1 问题的提出. 115
§5.2 数学模型及求解过程. 115
5.2.1 换热温度求解模型和过程. 115
5.2.2 分流器求解模型及求解过程. 119
5.2.3 混合器的求解过程. 119
5.2.4 优化求解方法. 120
5.2.5 换热网络的分解过程. 121
5.2.6 模拟计算的收敛判断过程. 121
5.2.7 优化模拟计算步骤. 122
§5.3 换热网络流程优化模拟实例. 122
§5.4 结果分析与讨论. 130
§5.5 结论. 131
参考文献. 131
第6章 具有任意结构换热网络的灵敏度分析. 132
§6.1 前言. 132
§6.2 数学模型. 132
6.2.1 过程方程. 132
6.2.2 决策变量、状态变量和参数的确定. 134
6.2.3 灵敏度系数和灵敏度的计算. 135
§6.3 求解步骤. 142
§6.4 实例计算. 143
§6.5 结果分析. 143
§6.6 结论. 145
参考文献. 145
第7章 换热网络的弹性分析. 147
§7.1 数学模型. 147
§7.2 FI的计算步骤. 148
§7.3 换热网络的过程模拟. 150
7.3.1 换热网络单元模型. 150
7.3.2 换热网络过程模拟步骤. 152
§7.4 实例. 152
7.4.1 例1. 152
7.4.2 例2(非凸性问题). 153
§7.5 结论. 155
参考文献. 155
第8章 换热网络的优化改进和控制. 156
§8.1 最佳调节变量选择的数学模型. 156
§8.2 更改参数的选择. 158
§8.3 改进后的年费用. 159
§8.4 换热网络的改进步骤. 159
§8.5 实例. 159
§8.6 结论. 164
参考文献. 166
第9章 多变量控制回路的识别. 167
§9.1 前言. 167
§9.2 关系矩阵的定义. 167
§9.3 关系矩阵的性质. 168
§9.4 控制回路的识别. 169
§9.5 控制回路的相互作用及控制时序. 170
§9.6 实例. 170
§9.7 结论. 173
参考文献. 173
第10章 换热网络优化节能技术的应用. 174
常减压换热网络的优化节能研究. 174
10.1 原有换热网络的标定与分析. 174
10.1.1 原有换热网络的流程叙述. 174
10.1.2 原有换热网络的标定数据. 176
10.1.3 原有换热网络换热器一览表. 176
10.1.4 物流基础物性. 177
10.1.5 物流基础数据库. 178
10.1.6 热回收量计算. 179
10.1.7 物流换热要求表. 179
10.2 原有换热网络的优化改进. 180
10.2.1 最大热回收量的计算. 180
10.2.2 改进后的换热网络. 180
10.2.3 改进后换热网络中换热器的核算及设计. 182
10.2.4 新增换热设备及投资一览表. 182
10.2.5 改进后换热网络的流程模拟. 183
10.2.6 常减压塔取热比计算. 186
10.3 改进后换热网络的操作弹性分析. 186
10.4 节能经济效益分析. 189