The large-scale deployment of smart meters means accumulating massive data on consumers' electricity consumption. The data collected from smart meters contains a lot of information that is useful to support the operation and management of low voltage distribution systems. However, data quality is key to the performance and accuracy of such support methods. For instance, missing data is unavoidable due to hardware or software failures in the meter, communication network, and end server. Incomplete data from smart meters ranging from a few samples to a few weeks block makes these data difficult to use. This paper deals with such missing data points in the smart meter data. Missing data can be solved by neglecting data with missing points, enhancing communication infrastructure, or using statistical imputation methods. Statistical imputation based methods in literature were evaluated in smart-meter data made public by the Belgian distribution system operator in this paper. Concretely, eight different imputation methods commonly used in the literature are implemented to estimate the missing points in the data set from actual smart-meter data. Linear imputation was suitable for shorter missing periods, while bilinear imputation performed better in larger blocks of missing data points.