Question 3 of 6: Field-to-Laboratory Sample Protocol

Question 3: Field-to-Laboratory Sample Protocol (20 marks)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

Sample collection methods. The collection technique must match the medium and the analyte. Water sampling ranges from a simple grab (bailer or dip sample from a well or stream) to depth-integrated or flow-proportioned automatic samplers producing a composite sample of a discharge. Soil sampling uses hand augers or split-spoon/core samplers to obtain an intact, representative profile without cross-contaminating one depth interval with material from another. Air sampling uses either active methods (a calibrated pump drawing a known volume through a sorbent tube or filter) or passive diffusion badges for a time-weighted-average exposure estimate. In every medium, the sampling design (SRS, stratified, systematic grid, judgmental – per Question 1) should match the suspected contamination pattern, with enough sampling points and replicates to characterize spatial or temporal variability rather than relying on a single, unsupported point value.

Sample preparation and preservation. Containers must be matched to the analyte – amber glass for organics sensitive to light or plastic leaching, acid-washed containers for trace metals – and pre-cleaned or certified contaminant-free. Preservation begins in the field: cooling to $4\,{}^{\circ}\text{C}$ on ice, chemical preservation (e.g., $HNO_3$ to pH below 2 for metals, $H_2SO_4$ for nutrients), and strict adherence to method-specified holding times before analysis, since many contaminants – volatile organics, dissolved oxygen, some metal species – change measurably within hours of collection if not preserved correctly. Volatile-organic samples are filled headspace-free to prevent loss to a vapour phase; samples requiring biological analysis are kept cold but never frozen.

Quality assurance and quality control. A written sampling and analysis plan defines the QA objectives – precision, accuracy, representativeness, completeness, and comparability, the classic "PARCC" criteria – before sampling begins. Field QC samples travel alongside the real samples: a field blank (clean water exposed to field conditions, checks for contamination during handling), a trip blank (checks for contamination during transport, especially for volatiles), field duplicates (two samples from the same location, quantify overall sampling-plus-analytical precision), and an equipment rinsate blank (checks decontamination of reusable sampling gear). At the laboratory, method blanks, matrix spikes and spike duplicates, laboratory control samples, and calibration against certified reference standards verify the analytical result independently of the field program.

Data management. A rigorous chain-of-custody record accompanies every sample from collection through disposal, documenting each transfer of possession so results remain legally and technically defensible. Data are entered into a database, ideally with automated range and unit checks, together with full metadata – location, date and time, sampler, method, holding time – retained alongside each result. Non-detect results are flagged and carried through statistical analysis using an accepted censored-data method, never silently substituted with zero or the detection limit without documentation, and a data-quality officer reviews results against the sampling plan's acceptance criteria before release.

Sources of error. Errors can enter at every stage: sampling error (a non-representative location, or too few samples relative to the population's true variability); collection error (cross-contamination between samples, disturbance of a soil profile, loss of volatiles during transfer); preservation/holding-time error (degradation before analysis); analytical error (instrument drift, matrix interference, calibration bias); and transcription/data-handling error (unit mistakes, mislabeled samples, database entry errors). A defensible protocol treats each of these as a distinct, separately-controlled risk rather than lumping them together as a generic "measurement uncertainty."