【作者】 闵英杰；

【导师】 吴广宁；

【作者基本信息】 西南交通大学 ， 高电压与绝缘技术， 2011， 硕士

【摘要】 变压器是电力系统中最重要和最关键电气设备之一,其运行的安全可靠性直接关系到电力系统的安全与稳定,绝缘状态是变压器寿命决定性因素,而温度是影响绝缘状态的主要因素,尤其是变压器热点温度。因此研究变压器内部温升及其对绝缘老化的影响,对变压器的运行管理和寿命评估具有积极意义。论文研究了变压器内部温升计算方法,变压器实际使用寿命估算方法,以及热特征参数对温升和绝缘等值老化结果的影响。基于热电类比原理,将变压器内部热传递过程转换成易于求解的热路模型,建立了变压器顶层油温热路模型和热点温度热路模型,推导了顶层油温度和热点温度的微分方程,并基于IEC60076-7推荐的指数方程解法,得到了一种变压器内部温度计算方法,计算结果能与实测值很好的吻合,尤其是在负荷增加阶段,与实测值基本一致。在顶层油温热路模型和热点温度热路模型的基础上,建立了简化的热点温度热路模型,推导了热点温度与环境温度和负载系数的直接关系表达式,计算结果基本与实测值吻合。根据收集的变压器负荷和环境温度数据,形成典型负荷和典型环境温度。基于IEEE绝缘等值老化计算模型,建立了一种变压器实际使用寿命预测方法,该方法考虑了负荷和环境温度的变化,对变压器实际使用寿命进行50次计算,并利用威布尔分布对结果进行分析,得出最大概率的实际使用寿命值；考虑负荷逐年增长和气温升高的影响,预测了在这种条件下的变压器实际使用寿命。当平均负荷小于额定负荷,而且实时负荷超过额定负荷的次数很少时,实际使用寿命远大于设计使用寿命；但如果频繁超过额定负荷,虽然平均负荷小于额定负荷,负荷过冲会造成变压器温度急剧升高,加速变压器绝缘老化。研究了用于变压器内部温升和绝缘等值老化时间计算的热特征参数的物理意义,分析了热特征参数取值对计算结果的影响。以热特征参数在推荐值±10%的波动范围为例,利用蒙特卡洛法分别对常规负荷和高负荷的运行情况进行数据分析,研究热特征参数选取推荐值对结果精确度的影响。在常规负荷下,采用推荐值能基本满足估算精度要求；但是在高负荷下,推荐值与实际值的误差将给最终的计算结果带来更大的误差,如果需要得到较精确的结果,需通过温升试验进一步确定热特征参数实际取值。更多还原

【Abstract】 The transformer is one of the most important and the most key electrical equipments in power system, and the safe and reliable operation directly relates to the security and stability power system. The insulation condition is the decisive factor of transformer life, and the temperature is the key factor of insulation condition, especially the hot-spot temperature. Therefore, the research of transformer internal temperature and influence on the insulation aging is of positive significance for transformer operation management and life assessment. The calculation method of transformer internal temperature, the method of estimate the actual service life, and the influence of the thermal characteristic parameters to the results of the insulation equivalent aging are researched.Based on the thermoelectricity analogy method, the heat transfer process of internal transformer is transformed into the thermal circuit model, which is easy to be solved. The thermal circuit models for the top oil temperature and hot-spot temperature are established, and the top oil temperature and hot-spot temperature expressions are deduced. Based on the exponential equations solution recommended in IEC60076-7, a calculating method for transformer internal temperature is proposed. The calculation results are close to measured values, especially during the step of load increasing, the results are nearly the same to measured values. On the basis of the thermal circuit models for top oil temperature and the hot-spot temperature, the simple thermal circuit model for the hot-spot temperature is established. The direct expression of the hot-spot temperature to environmental temperature and load coefficient is proposed, and the results are similar to measured values.According to the collection of transformer load and environmental temperature data, typical load and typical environmental temperature are formed. Based on the IEEE insulation equivalent aging model, a transformer actual service life prediction method is proposed. The method considers the changes of load and environmental temperature, and 50 times calculations are preformed. Using Weibull distribution on the results, the actual service life maximum probability value is obtained. Considering the influence of load and temperatures expanding gradually, the actual use life is forecasted. When the average load is less than rated load and little overload, the actual service life is far longer than its design service life. If the overload is often happened, although the average load is less than rated load, load overshoot can cause transformer temperature sharp rise. The transformer insulation aging will accelerate, because of the transformer temperature rise overshoot due to overload.The thermal characteristic parameters of the calculations of transformer internal temperature and insulation equivalent aging are elaborated. The influence of each thermal characteristic parameter in calculation results is analyzed. By varying these variables under various load conditions and by Monte-Carlo simulation with uniform distribution and normal distribution of these variables, their influence in calculation accuracy is obtained. The influence is small under normal load, and the accuracy can meet the estimation needs by using recommended value. But it is great under heavy load, and the thermal characteristic parameters have to be obtained accurately to improve the calculation accuracy.更多还原