# How does ETB dis-aggregate 60-minute granularity data (either usage data or solar production data) down into 15-minute increments?

We disaggregate 60-min data down into 15-min intervals using a curve smoothing polynomial regression, which is a widely used statistical process in prediction and forecasting. Note: this process will never create a higher value than the highest value contained in the dataset. This is important when creating 15-minute intervals for kW demand because we don’t want to assume there was a higher kW interval than what was contained in the 60-minute data.