节点文献

煤化工行业水系统集成优化及循环水软化技术研究

The Integrated Optimization of the Water Systems in Coal Chemical Industries and the Softening Technology for Circulating Water

【作者】 徐宏勇

【导师】 刘勇弟;

【作者基本信息】 华东理工大学 , 环境工程, 2014, 博士

【摘要】 本文基于系统工程理论,开展煤化工行业水系统集成优化理论与技术研究。首先着眼于全厂水系统的宏观层面,开发了适用于水网络优化的“遗传算法”,并对全厂水网络进行了优化改造;其次针对循环水系统,开展了循环水用水网络优化及软化基础理论与技术研究。针对水网络优化过程的数学问题实质,开发了“小生境退火遗传算法”。首先针对水网络优化求解空间庞杂的问题,开展了求解空间的简化工作,利用浓度约束、流量约束、质量约束等条件,以及夹点技术,对用水网络超结构模型及循环水网络超结构模型进行了精简工作,从而提高了算法的收敛性、准确性、稳定性。其次针对水网络优化约束条件复杂的问题,构建了全新的、具有自适应性的适应度函数,该适应度函数可以很好地平衡约束函数和目标函数,确保算法在求解前后过程中均做出准确的“遗传行为”。通过实例验证,表明该算法适用于单杂质、多杂质、管路约束、循环水网络优化等多种情况。以自主算法为核心,开发了一套用水网络优化软件。该软件具备独立的视窗界面,降低了用户优化操作的难度;设置了约束条件自动生成模块,解决了用户无法独立生成约束条件的困难,提升了算法的易用性;构建了流程图输出模块,便于用户对优化结果的理解。将软件应用于某煤化工企业,优化后一次水节约率为21%;除盐水节约率为29%;废水减排率为66%。针对软化工艺出水浊度高、沉淀产物细小的问题,开展了成核机理、软化速率、粒径分布的研究。在碳酸钙“聚集成核”模型的基础上,假设破碎过程为“二元随机”过程,建立了“聚集-破碎成核”模型,在上述模型的基础上,采用多重Monte carlo算法考察了不同过饱和度的情况下,体系的成核诱导时间及颗粒粒径分布情况,结果表明“聚集-破碎成核”模型的模拟误差(9.8%)小于“聚集成核”模型的模拟误差(24.3%),更接近于实际成核过程,模拟结果显示过饱和度对于细小颗粒的产生具有重要的影响。通过改造万通907型电位滴定仪,建立了定组成实验装置,考察了不同条件(温度、过饱和度、离子强度、pH)下软化速率及产物粒径分布的变化情况,结果表明,随着上述参数的升高,软化速率均有不同程度的提升,但软化产物中细小颗粒所占的比例也逐步增大,因此在软化过程中,当温度等参数升高时,应添加晶种等措施,以抑制成核现象的发生。针对循环水外排水、地下水软化效率不高的问题,开展了水质分析、混凝剂筛选、添加晶种的研究。水质分析结果表明,由于水样中存在着天然有机物或阻垢剂,导致水样中方解石、白云石、文石等均存在着不同程度的过饱和现象,难以结晶析出。研究表明通过添加混凝剂,可以部分去除水体中有机物(40%)或者阻垢剂(>90%),通过添加混凝剂和晶种可以提升软化程度20%至30%。通过软化处理,外排水Ca2+浓度可以降到50mg/L至20mg/L之间;地下水Ca2+浓度可以降到30mg/L左右,单从硬度指标评价,均可以满足循环冷却水的要求。基于上述研究,提出了“强化软化”策略,即在传统的软化过程中添加混凝药剂和晶种,在加药点附近控制过饱和度,从而可以有效提升软化效果,改善产物沉淀效果,该策略不但可以应用于软化过程,也可以应用于其他难溶盐的沉淀处理。将上述策略应用于该企业循环水补水软化方案设计中,预计改造后每日可减少采水900吨,减少排水900吨,约20个月可以收回项目投资。

【Abstract】 In this dissertation, the integrated optimization of the water system in coal chemical industries were performed based on the theory of system engineering. For the whole plant water network, the "genetic algorithm" available for the water network optimization was developed, while for the circulating water system, the optimization of the water network and the basic theory and technology on softening were studied in detail.According to the essence of the math problem during the water network optimization process, a "niche annealing genetic algorithm" was developed. The simplification of the solution space was firstly carried out in the optimization of the water network. The super-structure models of the water-consuming network and the circulating cooling water network were simplified using the constraints of concentration, flow and quality, as well as pinch technology, thus improving the convergence, accuracy and stability of the algorithm. Then a new self-adaptive fitness function was constructed for the complex constraints during the optimization of the water network. This fitness function could well balance the constraint functions and the objective function, thus ensured that the solving process of the algorithm kept accurate "genetic behavior". It has been validated by examples, that this algorithm was applicable in many situations, such as single impurity, multiple impurities, pipeline constraints, recycled water network optimization and so on.Based on the independent algorithm, a new software for water network optimization was developed. This software had a separate window interface, which reduced the difficulty of optimization operation for users. It had also been set up automatic constraint generation modules, thus solved the problems that the users could not generate constraints independently and improved the usability of the algorithm. Furthermore, a flow chart output module was constructed in this software for users to understand the optimization results easily. The application of the software in some coal chemical enterprise showed that the water was saved about21%, the demineralized water was saved about29%and the waste water was reduced66%.For the high turbidity and fines in softened water, nucleation mechanism, softening rate and the particle size distribution were studied in this dissertation. According to the model of calcium carbonate "aggregation model", and assuming that the breakup process was "binary random" process, the model of "aggregation-breakup model" was established. Based on the above model, the induction time for nucleation and the distribution of particle sizes under different saturations were studied using multiple Monte Carlo algorithm. The results showed that the simulation error of the "aggregation-breakup model"(9.8%) was much lower than that of the "aggregation model"(24.3%). Thus the simulated results might be closer to the actual nucleation process. The simulation results showed that the degree of supersaturation had significant effect on the generation of fine particles.The softening rates and the size distributions of product particles under different conditions (temperature, supersaturation, ionic strength, and pH) were investigated using the fixed component experimental apparatus, which was established by transforming the titrator Metrohm907. The results showed that, with the increase of these operation parameters, the softening rates increased at some degree, while the proportion of fine particles in the softening products gradually increased as well. Because the high proportion of fine particles would have a negative effect on the subsequent precipitation and separation, some measures such as adding seeds to suppress the occurrence of nucleation should be taken, when the parameters such as temperature in the softening process increased.Considering the poor softening efficiencies of discharge from circulating water system and the groundwater, studies were carried out on water quality analysis, coagulant screening and the addition of seeds. The existence of natural organics or scale inhibitors in the water samples, which were proved from the results of water quality analysis, resulted in different supersaturation degrees of calcite, dolomite, aragonite, etc. Thus they were difficult to be crystallized. The experimental results showed that organics of40%or scale inhibitors of more than90%could be removed by adding a coagulant, and the extent of softening could be enhanced by20%to30%by adding coagulants and seeds.A strategy of "enhanced softening" was proposed through the combination of the above research results, that is, adding coagulation agents and seeds in the traditional processes, and controlling the supersaturation at near the dosing points, which could effectively enhance the softening efficiency and improve the sedimentation. When the above strategy was applied to the design of the softening plan for make-up water of the circulating water in this enterprise, it is estimated that the water supply and the water drainage would be expected to be reduced for900and900tons, respectively, thus the investment would be recovered after20months.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络