Twenty-Eight Calls, Twenty-Eight Right
234 chloride samples screened across an assessment and a remediation, working to five different depth-based limits. On every sample clear enough to grade it, the screen and the laboratory landed on the same side of the line.
EXECUTIVE SUMMARY
This site applied five chloride limits, rising from 300 mg/kg near surface to 3,400 mg/kg below four metres, so the same reading meant different things at different depths. 234 samples were screened and 38 sent to an accredited laboratory. 10 of those sat too close to their limit on one side or the other to settle anything. On the 28 that could grade the screen, it got every one right — 8 exceedances flagged, 20 samples cleared, no misses and no false alarms. Eight measures score 16 of 16, rating Effective. Of the flags raised, 8 of the 23 flags raised across the site were ever submitted.
AT A GLANCE
234 screened • 38 laboratory-checked • 28 carrying a scoreable call
1.THE CHALLENGE
A conductivity meter measures how well soil conducts electricity, not how much salt is in it, and on this site that ambiguity would have been compounded five times over. The limit changed with depth — 300 mg/kg near surface, 3,400 mg/kg below four metres — so a single reading of, say, 600 mg/kg was a clear exceedance in one sample and comfortably clean in the next. Nothing in a conductivity number tells a crew which of those it is looking at.
2.THE APPROACH
Sal measures the chloride itself and reports it in milligrams per kilogram, the same units as the limit, so the App compares each result against the criterion that applies at that sample's depth as it is recorded. A result falling inside 20% of the criterion is not called either way, and a sample is only used to grade the screen where the laboratory result is outside that band too — a single grab that close to the line settles nothing on either side.
3. THE DATA
Of the 38 checked samples, 10 sat inside the band and 28 could grade the screen. The laboratory found 8 of those over their limits and the screen flagged all 8 — nothing clean flagged, nothing impacted passed.
WHERE THIS RESULT APPLIES
- Salt delineation across five depth-based criteria on one site
- Excavation floors and walls confirmed before the crew leaves
- Programmes where laboratory turnaround sets the schedule
- Deciding which samples are worth an accredited analysis
4.THE VERIFICATION
With the answer in the units of the limit, and the App holding the right limit for each depth, the crew sorted soil at the face rather than waiting on the laboratory. No false alarms is the half of that which shows up on an invoice.
5.THE DECISION
With the answer in the units of the limit, and the App holding the right limit for each depth, the crew sorted soil at the face rather than waiting on the laboratory. No false alarms is the half of that which shows up on an invoice.
6.THE VALUE
The screen sorted soil at the face against five different limits and was right every time the laboratory could settle the question. The band did its job too: samples reading close to the line were resolved on site rather than becoming disagreements in a report months later.
BASIS & LIMITS — WHAT THIS CASE DOES AND DOES NOT ESTABLISH
- 6 of the checked samples read within 20% of their limit and were re-addressed on site — re-tested, duplicated or stepped out — so the call was
- resolved before the crew left. A further 4 sit outside the scorecard because the laboratory result landed in that band, where a second grab could fall
- either side of the line. The remaining 28 carried a call to grade.
- 8 of the 23 flags raised across the site went to the laboratory; the rest were addressed by excavation. Catch and Flag Trust of 1.00 each rest on 8
- confirmed samples, with a 95% interval running down to 0.68.
- Closeness needs field duplicates and this programme ran none, so it is reported rather than scored. TRIUM was on site for initial training only, and
- sample locations and laboratory
.svg)
