In the past, machine learning methods such as Convolutional Neural Networks (CNN) and Autoencoder (AE) have been employed for leakage detection in water distribution networks (WDNs). However, these methods primarily rely on local pressure information and often fail to capture long-term dependencies in pressure series. In this paper, we propose a transformer based model to detect leakages in WDNs. This model incorporates an attention mechanism to learn data distributions and account for correlations between historical and daily interval pressures, thereby emphasizing long-term dependencies in the pressure series. The performance of the proposed method is evaluated using detection accuracy and F1-score. Experimental studies conducted on simulated pressure datasets from two different WDNs demonstrate that the transformer based model significantly outperforms traditional CNN methods.