甘油发酵动力学及工艺动态优化的研究
文献类型：学位论文
| 作者 | 谢东明 |
| 学位类别 | 博士 |
| 答辩日期 | 2000-07 |
| 授予单位 | 中国科学院研究生院 |
| 导师 | 刘德华 |
| 关键词 | 甘油发酵 动力学 气升式反应器 发酵工艺 动态优化 |
| 其他题名 | Kinetic models and dynamic process optimizations for glycerol production by fermentation |
| 学位专业 | 生物化学工程 |
| 中文摘要 | 提高甘油产率、降低残糖浓度、加快发酵速度，实现过程放大，是研究发酵法生产甘油面临的难题。为此，本文以“动态”的观点，从反应器操作以及温度、溶氧水平、营养物浓度等工艺参数的调控入手，通过实验并结合数值理论分析，对耐高渗酵母发酵生产甘油的反应器及工艺的优化进行了研究。首先，本文建立了甘油发酵过程的宏观动力学模型，为发酵工艺的优化及探求放大过程的规律提供了理论参考。其次，在前期好氧发酵中，对气升式反应器采用周期性振荡气升操作，并采用变温发酵、热冲击等动态操作，进一步提高了甘油产率；在发酵后期，采用CO_2气提厌氧发酵或真空微氧发酵，进一步降低残糖而不消耗甘油。随后，通过动力学模型的数值模拟，从理论上分析了温度及初始条件对发酵结果的影响及其随反应器的放大和通气量改变的变化规律。最后，根据甘油发酵是一状态时变过程的特点，运用最优控制理论，分析了几种反应器条件下温度、葡萄糖及玉米浆流加速率三参数的动态最优轨迹；在此基础上，进一步提出了在实际生产中易于实施的准最优控制策略。论文的内容包括如下几个方面：1.根据多组不同发酵条件下的平行实验数据，建立了耐高渗酵母进行甘油发酵的宏观动力学模型。并且假定模型的参数可分为只与微生物反应本身特性有关的固定参数以及还与反应器内溶氧、流动混合状况有关的可变参数两部分。仅仅对两个关键的可变参数重新进行估值，可将模型较好地拟合新反应器条件下的发酵过程，为发酵工艺的放大提供了新的理论参考。2.甘油发酵采用前期好氧和后期厌氧或微氧的二步发酵工艺，即通气量控制的溶氧水平整体上是“动态”的。在前期好氧发酵中，对气升式内环流反应器采用周期性振荡气升操作，与传统的稳态中心气升操作相比，振荡气升操作可以显著提高发酵甘油的产率。通过动力学模拟发现，振荡气升操作改善了反应器内的溶氧及流动混合状况。3.采用变温发酵方式和热冲击技术，在保持残糖消耗速率的前提下，进一步提高了甘油产率。变温发酵的技术方案是：在细胞对数生长期内采用较低的温度促进细胞生长；之后采用较高的温度促进甘油的产出；最后在35 ℃以下继续进行厌氧发酵促进残糖消耗。热冲击操作的技术方案是：在细胞指数生长期的中段对细胞进行45 ℃的热冲击，持续时间为30min。4.当好氧发酵后期浓度降到1-4%时，继续采用通气量为0.5vvm以上的CO_2气提厌氧发酵或真空度在0.02MPa以上的真空微氧发酵，可以较快地降低残糖，同时甘油浓度可不下降甚至继续增加。5.通过动力学模型的数值模拟，分析了温度及初始细胞、葡萄糖浓度和磷含量对发酵和种子培养结果的影响，并且发现由小型实验所得的参数优化条件需随反应器的放大作相应的调整，因而优化条件也是“动态”的。6.分析了温度、葡萄糖和玉米浆流加的动态最优或准最优控制策略。首先，运用Pontryagin的极大值原理，选择温度、玉米浆流加体积及葡萄糖流加体积与初始发酵体积之和作为控制变量，将奇异控制问题化为非奇异控制问题，采用交替优化的算法对常规共轭梯度法进行改进，求得了甘油发酵过程中温度、玉米浆及葡萄糖流加速率三个参数在理论中上的动态最优控制轨迹。在此基础上，将发酵过程划分为若干不同的稳态阶段，进一步分析了易于实施且逼近最优控制的准最优控制策略，其中准最优流加可采用多段恒速流加和多段脉冲流加两种方式。最后，温度单参数优化和葡萄糖、玉米浆脉冲流加量两参数优化的准最优控制策略得到了实验的初步验证。 |
| 英文摘要 | To improve the glycerol productivity, deplete the residual glucose concentration, shorten the fermentation time and find the basic rule of scale-up effect in glycerol production by fermentation, dynamic process optimizations were experimentally and numerically studied in this article by investigating the operation state of the reactor, the process parameters of temperature, dissolved-oxygen and concentration of nutrients. Firstly, macroscopic kinetic models of glycerol fermentation were established to provide theoretical guidance for process optimization and sale-up technology. Secondly, during the former aerobic fermentation, the Airlift Loop Reactor was performed in the periodic airlift operation and temperature-variation and heat-shock technology were applied in the process to further improve glycerol productivity; in the latter anaerobke fermentation, an anaerobic CO_2 stripping process or approximately-anaerobic vacuum fermentation was applied to deplete residual glucose but still improve or maintain glycerol concentration. Then by numerical simulation with the kinetic models, the effects of temperature and initial conditions on the fermentation results were analyzed and their changing trends with the scale up of the reactor or variation of the aeration rate were discussed. Finally, considering that glycerol fermentation is also a dynamic process, dynamic optimal profiles of temperature, feed rates of glucose and corn steep slurry were investigated based on optimal-control theory. To conveniently apply the dynamic profiles to practical process, a sub-optimal control strategy was proposed. 1. Kinetic models of glycerol production by fermentation with osmotolerant yeast Candida krusei were established, and the model parameters were estimated based on multiple batches of fermentation data under different culture conditions. It was assumed that the model parameters could be divided into two groups. One group is associated only with microbial reaction and cab be called as invariable parameters, while another group is associated further with both oxygen transfer and flowing and turbulent state in the reactor, thus is called as variable parameters. By re-estimating only two key variable parameters, the model simulation could agree well with the new experimental data when the reactor was scaled up or down or the aeration rate of the fermentation was greatly changed. 2. Glycerol production by fermentation should be performed firstly in a longer aerobic stage then in a shorter anaerobic one, It means the dissolved-oxygen programming in the whole fermentation process is dynamic, During aerobic fermentation, a periodic oscillating airlift operation was applied in the fermentation with an Airlift Reversible Loop Reactor (ARLR). It was found that the glycerol productivity in oscilating operation could be significantly improved compared with that in the traditional central airlift operation. Simulation results from the kinetic models suggested that the periodic oscillating airlift operation could intensity oxygen transfer and flowing and turbulent state. 3. Temperature-variation and heat shock treatments were investigated to further improve the glycerol yield and shorten the fermentation time. For temperature-variation process, the general temperature scheme should start at about 30 ℃ in a short former stage to improve cell growth rate, then be maintained at about 40 ℃ for a longer mid-stage to improve glycerol production, finally be lowered less than 35 ℃ in an further anaerobic process to accelerate consumption rate of residual glucose. For heat shock treatment, it should start at the mid term of the exponential growth stage and be controlled at 45 ℃ for 30min. 4. When the concentration of residual glucose was reduced to 1-4%, the aerobic glycerol fermentation was suggested by followed by a further anaerobic CO_2 stripping process with flow rate of 0.5vvm or by an approximately-anaerobic vacuum fermentation with vacuity of 0.2Mpa to deplete the residual glucose but still improve or maintain glycerol concentration. 5. By numerical simulation with the kinetic models, the effects of temperature and initial concentrations of cell, glucose and phosphorus on the final fermentation and seed culture results were analyzed and their changing trends with the scale up of the reactor or the variation of aeration rate were discussed. It was found that the optimized process conditions obtained in the little scale of reactors should be changed and be re-optimized as the reactor was scaled up, thus the optimized contidons is also dynamic during the scaling up of the fermentation. 6. Dynamic optimal and sub-optimal profiles of temperature, feed rates of glucose and corn steep slurry were investigated. Firstly, the optimal profiles were obtained by Pontryagin's maximum principle. By selecting temperature, the feed volume of corn steep slurry and the volume sum of both glucose feed and initial fermentation as the control variables, the possible singular problem was transformed to a nonsingular one, then it was solved by a novel alternative optimization technique based on conjugate gradient method. Secondly, to conveniently apply the control profiles to practical process, the sub-optimal control strategies were proposed in which the fermentation was divided into multiple different steady stages and the feed operation was performed in either multi-pulse or multi-constant-rate mode. Finally, The sub-optimal control strategies for only temperature optimization and for optimizations of multi-pulse feed amounts of both glucose and corn steep slurry were verified by the experiments. |
| 语种 | 中文 |
| 公开日期 | 2013-09-26 |
| 页码 | 207 |
| 源URL | [http://ir.ipe.ac.cn/handle/122111/1922]  |
| 专题 | 过程工程研究所_研究所（批量导入） |
| 推荐引用方式 GB/T 7714 | 谢东明. 甘油发酵动力学及工艺动态优化的研究[D]. 中国科学院研究生院. 2000. |
入库方式： OAI收割
来源：过程工程研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。