石化行业固定储罐VOCs排放估算方法研究

发布时间：2019-06-09 12:38

【摘要】：石油及其产品在中国的能源结构中占有相当的比重,中国已经发展成为生产和消费的石油大国。石油及其产品是多种碳氢化合物的混合物,其中的轻组分具有较强的挥发性,给企业和社会在安全、环境、能源等方面产生了严重的影响。挥发性有机物是对大气对流层非常重要的痕量组分,在大气化学过程中扮演者极其重要的角色,对二次有机污染物形成、大气氧化能力、人体健康等方面都有着重要的影响。石化固定储罐VOC排放量的估算方法各国有所不同。本文主要采用BP神经网络估算方法,研究了存储真实蒸气压、储罐直径、周转量,摩尔分子量、储罐的颜色等参数对储罐排放量的影响,运用BP神经网络可以快速简单、有效的估算出储罐的VOCs的排放量,具有重要的指导价值。(1)通过石化储罐四种VOCs排放量估算方法的分析,确定影响储罐排放量的主要影响因素是存储的日平均温差、存储物质的真实蒸气压、存储物质的周转量、存储物资的摩尔分子量等,次要的影响因素有储罐的容积、存储物料的高度、储罐的颜色等。(2)以上海市某石化企业的汽油储罐和甲醇罐为案例对象,计算储罐的总损失与财政部结果进行比较,其中EPA推荐的TANKS模型估算结果最为接近,汽油罐的静置损失误差为39.16%,工作损失误差为2.14%,甲醇罐的静置损失误差为33.24%,工作损失误差为50.95%;日本的《减少污染物排放和转移登记物质排放的成功事例》(以下简称事例)的计算方法与财政部结果的差距次之,汽油罐静置损失误差为83.14%,工作损失误差为106.23%,甲醇罐的静置损失误差为54.04%,工作损失误差为142.10%;《石油库节能导则》推荐公式计算结果误差最大。(3)运用BP神经网络模拟石化企业固定储罐的VOCs排放量,并验证BP模拟结果的准确性,其中静置损失的平均误差为5.57%,工作损失的平均误差3.46%。
[Abstract]:Petroleum and its products account for a considerable proportion of China's energy structure, and China has developed into a major oil producer and consumer. Petroleum and its products are mixtures of many hydrocarbons, in which the light components have strong volatility, which has a serious impact on enterprises and society in terms of safety, environment, energy and so on. Volatile organic compounds (VOCs) are very important trace components in the tropospheric layer. Volatile organic compounds (VOCs) play an extremely important role in the process of atmospheric chemistry and have important effects on the formation of secondary organic pollutants, atmospheric oxidation ability, human health and so on. The estimation methods of VOC emission from petrochemical fixed storage tanks are different from country to country. In this paper, the effects of storage real vapor pressure, tank diameter, turnover, molecular weight and color of storage tank on tank emission are studied by using BP neural network estimation method. BP neural network can be used to be fast and simple. It is of great guiding value to estimate the VOCs emission of storage tank effectively. (1) through the analysis of four VOCs emission estimation methods of petrochemical storage tank, it is determined that the main influencing factor of tank emission is the daily average temperature difference of storage. The real vapor pressure of the storage material, the turnover of the storage material, the molar molecular weight of the storage material, etc., the secondary influencing factors are the volume of the storage tank, the height of the storage material, (2) taking the gasoline tank and methanol tank of a petrochemical enterprise in Shanghai as an example, the total loss of the tank is compared with that of the Ministry of Finance, among which the TANKS model recommended by EPA is the closest. The static loss error of gasoline tank is 39.16%, the working loss error is 2.14%, the static loss error of methanol tank is 33.24%, and the working loss error is 50.95%. The difference between the calculation method of Japan's "success stories of reducing Pollutant emissions and transfer of registered substances" (hereinafter referred to as the example) and the results of the Ministry of Finance is the second, with the static loss error of gasoline tanks being 83.14%. The working loss error is 106.23%, the static loss error of methanol tank is 54.04%, and the working loss error is 142.10%. The recommended formula of Oil Depot Energy Saving Guide has the largest error. (3) BP neural network is used to simulate the VOCs emission of fixed storage tanks in petrochemical enterprises, and the accuracy of BP simulation results is verified. The average error of static loss is 5.57% and the average error of work loss is 3.46%.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：X74

【参考文献】