Novel performance assessment method for superheated steam control of a coal-fired power plant under renewable energy accommodation condition

Abstract

For the large-scale accommodation of intermittent renewable energy, coal-fired power plants operate under rapid load change condition so that the superheated steam temperature exhibits highly non-linear characteristics. This study proposed novel solution procedures for non-stationary and operating data to quantitively assess the performance of secondary superheated steam control. De-mean and difference operations were employed for non-stationary data to stabilize the data sequences. Auto regressive models were then built to estimate the time delay such that the assessment index of the minimum variances can be calculated. The results show that the proposed auto regressive model has the highest accuracy and its order is relatively stable at three, when the sampling period is approximately 20–34 s. Taking into account the variations in secondary superheated steam temperature, which amounted to ±8 ℃ and ±2.5 ℃ during low-load and high-load conditions respectively, the assessment index of minimum variance under the low-load operation was 0.495, which was 41 % lower than that under the high-load operation. This indicates that the higher the load operates, the better the randomness performance of the superheated steam control is. These findings may help optimize the superheated steam control strategies in coal-fired power plants.