It is an operational practice that refineries (here producers) work in an optimised way and produce diesel and gasoline with a mass fraction of sulfur close to the specification. However, when this quality parameter is analyzed by refineries and, later, after custody transfer operations, it is re-analyzed by storage terminals (here called consumers), these measurement results do not always meet the specification. This study proposes a new methodology based on data reconciliation connected to the guard bands concept to establish upper acceptance limits for producers offering a comfortable giveaway to consumers. From the minimized reconciled expanded uncertainties, process model constraints, and the upper specification limits, the upper acceptance limits for the producers were optimised to ensure that the reconciled values meet the specification. Consumer and producer risks and histograms were calculated. Validation data sets proved that the proposed methodology could be applied because all the reconciled values comply with the specifications.