In the perspective of network operators, demand response (DR) is a form of shaping of the overall consumption in the system by peak clipping and shifting. Bearing the necessity and importance and the role of DR in the overall system in focus, the objective of the paper is to extend an analysis on the flexibility of the demand. The potential and the corresponding cost for demand flexibility of different levels in private households is determined from literature. Following that, consumption in a residential house in Nordic Europe climatic conditions was measured. In order to identify the patterns and categorize the consumption, a clustering algorithm is implemented on time-series load demand based on the trend using Wards agglomerative hierarchical clustering to formulate classifications. Using the classifications, a model was simulated in GAMS based Balmorel platform, which chooses optimal operation scheme for categorised loads. Two distinct models were created. One as in grid integrated, where market price acts as an incentive and second as islanded mode, where the resource availability is the motivation. The resulting consumption profiles and the decision making economic stand-point of flexibility offered by means of DR are discussed. The dendogram of the associated individual devices in the consumption pattern confirms that the weekend and weekday pattern of consumption are of different construct. However, same devices provide different flexibility whether operating on workdays or weekends. Modelling results indicate an overall increase in the efficiency of energy consumption with a demand-shifting possibility. In analysis, no attention was paid to type nor price of devices that allow to control individual appliances.