If more information on the diurnal patterns is desirable, strategies based on grab samples cannot be used. Instead, the 24-h, 48-h or 7-d measuring strategy has to be used. These strategies can also be used to estimate an EF if multiple 24 h, 48 h or 7 d periods are taken. To get an estimated EF with a relative error below 15%, 21 (dataset 1) or 27 (dataset 2) 24-h periods, 20 (dataset 1) or 29 (dataset 2) 48-h periods or 13 (dataset 1) or 15 (dataset 2) 7-d periods were necessary. It can be concluded that there is no large difference in the number of 24-h or 48-h periods that have to be sampled. The number of 7-d periods needed to get a good estimate of the EF is about 40% lower than the number of 24-h or 48-h periods.