Iron Count Trend After a Monitoring Program Change
A field's internal corrosion monitoring program has used iron counts from produced water as the primary indicator for five years. Baseline iron is 2-4 ppm. Six months ago the operator changed sampling from a separator outlet to a treater outlet and changed the analytical lab. Iron counts now read 9-12 ppm. Coupon rates are unchanged at 0.3 mpy. No pigging or chemical changes occurred. The operator wants to declare a corrosion upset and launch an investigation. A new operations supervisor argues the data are not comparable and the trend is an artifact.
1. How should the team determine whether the iron increase is real or an artifact of the program change?
Read the worked answer
The team should treat the change as a data comparability problem first, because two variables changed at once (sampling point and lab).
Steps:
- Re-establish a baseline: collect paired samples from both the old separator outlet and the new treater outlet for a defined period (e.g., 4-8 weeks) and analyze them at both the old and new labs. This separates the effect of location from the effect of analytical method.
- Check lab method: confirm detection limits, digestion procedure, and calibration. A lab change can shift iron results by several ppm without any process change.
- Check sampling point: the treater outlet may carry more water, different solids, or different partitioning of iron, so the same corrosion rate can produce a different iron concentration.
- Cross-check with independent indicators: coupon rates (0.3 mpy, unchanged), ER probes, and UT surveys. If all independent indicators are flat, the iron rise is more likely an artifact.
The judgment is that a single indicator changing at the same time as the measurement method is not evidence of a corrosion upset until comparability is restored.
2. What should the team do with the historical trend data and the monitoring program going forward?
Read the worked answer
- Do not splice the two datasets into a single trend line without a documented correction. Mark the program change on all trend plots and keep the old and new series separate until a bridge period establishes the offset.
- Document the change in the monitoring program records, including the reason, the bridge data, and the comparability conclusion, so future reviewers understand the discontinuity.
- Standardize the sampling point and lab going forward, or run both in parallel for a defined period if a change is needed again.
- Retain the old data as valid for its period; do not discard it, but do not use it to set alarm limits for the new program.
- Re-set alarm thresholds based on the bridged baseline once comparability is established.
The key point is that data management discipline - documenting method changes and bridging baselines - is what makes a trend usable for decisions.
3. Draft the conclusion the team should report, including whether an investigation is warranted.
Read the worked answer
Conclusion: the iron increase is most likely an artifact of the monitoring program change, not a corrosion upset, because the only independent corrosion-rate indicator (coupons at 0.3 mpy) is unchanged and two measurement variables changed simultaneously.
Report should state:
- The observed iron rise and its timing relative to the program change.
- The bridge-test results comparing old/new sampling points and labs.
- The coupon and UT data showing no rate increase.
- A recommendation: no full corrosion investigation at this time, but continue the bridge monitoring for one more quarter and re-evaluate.
- A corrective action: document the change and standardize the program.
A full investigation would be warranted only if the bridge test shows the iron rise persists at the old sampling point and lab, or if coupons/UT begin to trend upward. This keeps the response proportionate to the evidence.