Predicting the energy demand of buildings during triad peaks in GB

•A stochastic model was developed to calculate the probability of having a “triad” peak on a daily and half-hourly basis.•Real weather data were analysed and included in the model to increase its forecasting accuracy.•An artificial neural network forecasting model was developed to predict the power demand of buildings during triad peaks.•The model was trained and tested on real “triad” peak data and building electricity demand data from Manchester.