Sampling location — whether the point is truly representative of the stream (e.g. downstream of adequate mixing, not immediately after a chemical addition point).
Sampling frequency and duration — too short a survey period misses diurnal, batch or day-of-week production variability.
Sample type — grab vs. time-proportional vs. flow-proportional composite; the wrong choice for the parameter/flow pattern biases the result.
Sample preservation and holding time — degradation, volatilization or biological activity between collection and analysis can change concentrations before the lab ever sees the sample.
Analytical method and QA/QC — method detection limits, interferences and laboratory error all affect result reliability.
Production schedule during the survey — sampling during an atypical (low, high or shut-down) production period does not represent normal operation.
Weather/infiltration — stormwater inflow/infiltration can dilute concentrations and distort flow measurements if the survey spans a wet-weather event.
Equipment calibration and chain of custody — uncalibrated flow meters or a broken chain of custody undermine the defensibility of the data.
