水泵水轮机典型工况切换过程内流瞬态特性与演变机理

The transient stability problem of pump turbine has become one of the hottest topics in the domestic and international energy engineering field. As the unsteady unit fluid under typical conditions is the root cause of the system instability accident, it is urgently necessary to investigate the transient inflow evolution mechanism and system synergy. This project will be conducted on the performance of water turbines which run in typical conditions, including the start and shutdown, the increase and decrease of load according to dynamic grid demand, and the full-load rejection. In combination with hydrodynamics, the three-dimensional transient flow field numerical simulation and flow-induced noise sound field calculation of the full-flow channel will be carried out. Furthermore, high precision acquisition of signals and capture of flow fields in test will be used to reveal the evolution mechanism of internal flow in full flow channel and its influence on external characteristics of system. A high-order dynamic sub-grid DES model is proposed and different grid wall sliding techniques are determined to approach the real flow field. The flow field and sound field instability characterization are separately decomposed, simplified and coupled in both time and frequency domains. The internal flow field atlas is combined to explore the mechanism of transient inflow evolution. The hybrid nonlinear probabilistic model will be built based on the aforementioned research and convolution neural network, which illustrates the relationship between internal fluid characteristics and the external operating index of the system. Considering different cases from other plants, this probabilistic model will be optimized and validated based on coupling model and multi-objective functions. This model realizes the prevention and prediction calculation of the unfavorable phenomena in pumped storage plant, such as the output and speed swing, the coupling of water, machine and electricity. The resultant conclusions and advanced experimental methods of this project will provide a theoretical reference and a practical scientific guidance for the safe and efficient operation of pumped storage plants.

水泵水轮机瞬态稳定性问题已成为国内外研究热点之一，机组过渡工况是引起系统不稳定性事故的根本原因，因而瞬态内流演变机理和系统协同性亟待深层研究。本项目以水泵水轮机典型过渡过程启停机、动态增减负荷以及甩负荷为研究对象，综合水力动力学理论、全流道三维瞬态流场模拟和流动诱导噪声计算、高精度信号采集与暂态流场捕捉试验，重点研究内流演变机理及其与系统的协同性。尝试提出高阶动态亚格子DES模型和确定各工况高次网格运动方程以逼近流场实况，通过对不稳定表征量在时频域内单独分解、简化和耦合，结合可视化内流图集，揭示暂态内流演变机理。基于卷积神经网络构造内流特征与系统外部运行指标之间混杂非线性概率模型，根据耦合模型与多目标函数优化、验证该模型。从而实现抽水蓄能电站出力和转速摆动，水、机和电耦合等不良现象的预防和预测计算。本项目的相关成果将为抽水蓄能电站安全高效运行提供理论参考与实际性科学指导。

抽水蓄能电站持续性大力建设为新能源的发展保驾护航，助力于国策“双碳目标”实现。水泵水轮机作为抽水蓄能电站的核心灵魂，频繁切换于发电与抽水工况时必经过渡过程，其运行稳定性成为了抽蓄技术发展的制约。本项目依托西北农林科技大学，采用理论分析、数值模拟与试验研究相结合方法，对水泵水轮机部分典型工况内流演变机理及其与系统的协同性展开研究。主要研究内容包含：1)基于RNG k-ε湍流模型，采用Ehrhard非线性湍流模型对雷诺应力进行非线性化处理，得到非线性涡粘模型的局部时均化PANS模型用于求解内流场，提升了高压力梯度区、大曲率回流以及复杂的分离流动等现象模拟能力。2)针对水泵水轮机过渡过程转速、压力等大波动问题，提出流量调节规律优化方法、阀门与导叶协同调节方式，降低了水泵水轮机参量变化范围，从而提高了其在甩负荷、增减负荷过渡工况的运行稳定性。3)基于计算流体动力学完成三维流场计算，分析湍动能、压力变化、涡旋强度等揭示水泵水轮机各工况下内流演变规律。将流场动态变化信息加载至结构场与声场，利用流固耦合方法计算模态特性，基于LMS得到甩负荷工况流动诱导噪声分布及辐射水平特性。4)针对关键部件转轮采用常见六种材料，基于阻尼效应、模态频率下降率、抗变形能力等对比不同工况下材料性能，初步揭示转轮内流变化与结构特征的关联机制。5)基于Suter变化法针对电站过流系统增减负荷、甩负荷等工况撰写计算程序求解转速、压力等变化，以数值仿真、试验研究结果优化程序，并应用于其他电站验证其正确性与普适性。本项目以1作/通讯发表SCI/EI 6篇，获得授权国家发明专利5项，中文核心3篇，培养青年教师1名，培养硕士研究生3名，本科生毕业设计/论文13人次，研究成果获得2021年陕西省电源行业成果转化奖，2021年度陕西省电源学会优秀学术论文一等奖。