Question 2 of 7: Soil remediation and effluent-monitoring statistics

Question 2: Soil remediation and effluent-monitoring statistics (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.

(i) Remediating heavy-hydrocarbon soil contamination impacting a fishable surface water body

Technology 1 — excavation with off-site thermal desorption. Example: a former fuel-storage tank farm on a riverbank where weathered diesel/bunker-C has migrated into the bank soil and is visibly sheening into the river at low flow. Contaminated soil above the applicable CCME/provincial heavy-hydrocarbon (F3/F4 fraction) criterion is excavated under a sheet-pile or turbidity-curtain containment to prevent re-mobilization into the river during the dig, hauled to a licensed thermal desorption facility where the soil is heated (300–550 °C) in an oxygen-limited chamber to volatilize the hydrocarbons for capture/destruction in an afterburner, and the treated (clean) soil is returned or the excavation backfilled with clean fill. This is the fastest route to eliminating the ongoing source and is preferred where the shoreline geometry means continued leaching cannot be tolerated even short-term.

Technology 2 — in-situ bioremediation with a permeable reactive/nutrient barrier. Example: the same site but where excavation is impractical (contamination extends beneath an active structure or below the water table adjacent to the bank). Indigenous hydrocarbon-degrading bacteria are stimulated in place by injecting oxygen (air sparging or oxygen-release compound) and nutrients (N, P) through a line of wells installed parallel to the shoreline, forming a biologically active zone that intercepts and degrades the hydrocarbon plume before it reaches the river; groundwater monitoring wells confirm declining hydrocarbon concentrations and rule out daughter-product accumulation. This avoids disturbing the riverbank and the associated fish habitat/erosion risk of excavation, at the cost of a much longer remediation timeline (months to years, since biodegradation of heavier hydrocarbon fractions is slow).

Both approaches would typically be selected together in practice on a fish-bearing watercourse: thermal desorption of the worst, most-mobile source zone to stop the acute discharge immediately (satisfying a regulator's/DFO's immediate fish-habitat-protection concern), with in-situ bioremediation left running on the residual, less-mobile fringe of the plume as the lower-cost long-term polish.

(ii) Replication, method detection limit, and measurement frequency in compliance monitoring

Replication (R) is the number of independent samples analyzed for the same parameter at the same point in time and location. It quantifies (and controls) the analytical and sub-sampling variability inherent in any single measurement: a compliance decision based on one grab sample cannot distinguish a true exceedance from ordinary measurement noise, whereas $R$ replicates let the analyst compute a mean and a confidence interval and test whether that interval genuinely excludes the permit limit. Replication is especially important near a permit limit, where the cost of a false non-compliance call (unnecessary regulatory/enforcement action) or a false compliance call (an undetected real exceedance reaching the receiving water) is high; CCME and provincial guidance typically require duplicates or triplicates at a set frequency as part of a facility's quality-assurance program.

Method detection limit (MDL) is the lowest concentration of the analyte that a specific analytical method can reliably distinguish from zero with a stated statistical confidence (conventionally 99%, from repeated analyses of a low-level spike and Student's-t statistics). It matters for compliance because a permit limit set below the MDL of the method a lab is using cannot actually be verified — every result reports as “non-detect” regardless of the true concentration — so the MDL must be substantially lower than the limit it is being used to demonstrate compliance against (a common rule of thumb is MDL ≤ permit limit/3 to 1/10). Reporting values below the MDL as if they were exact numbers (rather than <MDL) overstates the precision of the measurement and can distort trend analysis or loading calculations.

Measurement frequency (ν) is how often the facility is required to sample and report (e.g. daily, weekly, monthly, or flow-proportional composite). It governs the statistical power to detect an exceedance and the representativeness of the reported average: a discharge with a highly variable load (e.g. a batch industrial process) needs a higher ν (or flow/time-proportional composite sampling) to avoid missing short-duration exceedances that a monthly grab sample would smooth over, while a stable, well-buffered municipal effluent can be adequately characterized at lower ν. Frequency is usually set in the permit as a function of discharge volume, receiving-water sensitivity, and historical compliance record (facilities with a good compliance history are often granted reduced ν, an incentive structure used by several Canadian provincial regulators).

Together, $R$, MDL and ν define the overall defensibility of a compliance monitoring program: $R$ controls confidence in any one reported value, the MDL sets the floor of what can be meaningfully reported, and ν controls how representative the sampled record is of the facility's true ongoing performance — a program strong in one but weak in the others (e.g. one very precise annual sample) still gives a poor picture of actual compliance.