The electricity generation from renewable sources in the U.S. is projected to increase from 19% in 2019 to 38% in 2050. The major portion of this increase originates from highly variable energy generation sources, including solar and wind. This variability in electricity production, in combination with the variability of the load and system disturbances, demands higher flexibility in the grid, from the supply and/or demand side. Demand-side management (DSM) will be a more economical and environmentally-friendly option to provide flexibility services, compared to non-renewable energy dependent generation plants which are commonly used today during peak use periods. The characterization of different types of the load is essential to provide the quantification of potential flexibility of demand. This study uses a data-driven approach to model occupant-dependent residential appliance loads, based on several sources of high-resolution disaggregated energy data. The main components of this model for each end-use include, characterization of the load profile and duration of operation for a single load cycle, and the likelihood of use by the time of day. The uncertainties inherent in the appliance load curves are estimated from the different data sources, in an effort to make the model scalable to the regional level for grid-level modeling. End-use loads of interest include coincidental loads from occupant-dependent loads including clothes washers and dryers, dishwashers and stove/ovens. The resulting models will help to assess the capability of different types of loads to provide the flexibility services, such as capacity services, spinning reserves and/or voltage regulation, to the grid.