A fundamental error of sampling occurs when a sample is extracted from a larger mass of ore or lot. The act of sampling requires that all particles belonging to the lot have an equal chance of being extracted. This type of sampling is called unbiased sampling. The sampling equipment (herein referred to as the sampler) should be constructed with some additional mechanical characteristics to guarantee that the samples will be unbiased. Samplers, extracting increments from a continuous flow, such as a conveyor belt, collect aliquots at regular time intervals. These time intervals relate to the number of aliquots extracted from a given mass of lot, whilst passing through the continuous flow. The frequency of extractions will influence the precision or the error of the sample (integration error). Further, the ore variability influences the amount collected from the continuous flow. Normally, the sampling protocol utilised is designed for the extractions obtained from the conveyor belt, omitting any error committed during this operation, ie the error of extracting from the continuous flow. This study aims to analyse the error of sampling a continuous flow by using a variogram with added on auxiliary functions. Each different piece of material passing through the flow will exhibit a distinct temporal variogram, which should influence the frequency of the samples being extracted for each chosen sampling error. The methodology used is presented in the form of a case study from a coal processing plant.