IRI-Corrology®: Sour Water User Manual

Practical guide for running IRI-Corrology®: Sour Water, interpreting NH₄HS susceptibility scores, and using remediation pathways.

IRI-Corrology®: Sour Water User Manual

A framework for screening failure likelihood, interpreting findings, and determining remediation paths for assets exposed to alkaline sour water corrosion.

Manual Focus	Description
Primary objective	Support integrity prioritization and inspection planning
Decision style	Comparative susceptibility signal with engineering judgment
Recommended workflow	Baseline -> driver diagnosis -> pathway evaluation -> documented action

Use this page as the standard runbook during live sour-water analysis.

Designed for fast, reliable execution in NH₄HS workflows. Keep one baseline pinned, test one lever at a time, and document every decision with residual uncertainty.

Manual Map

Purpose

This manual explains how to use IRI-Corrology®: Sour Water for practical integrity screening and how to interpret outputs with confidence.

It is written for day-to-day engineering use: run setup, result interpretation, remediation-path usage, and reporting.

At a Glance

Primary purpose: Prioritize NH₄HS corrosion susceptibility and support inspection planning decisions.
Output: A comparative susceptibility score that supports engineering judgment. It is not a standalone fitness-for-service (FFS) assessment or complete API RBI risk value.
Recommended workflow: Establish a baseline case, then evaluate controlled what-if scenarios to understand how potential changes influence the projected score.

Quick Start

Run Flow

Open IRI-Corrology®: Sour Water.
Choose Predictive or Inspection mode.
Enter required inputs.
Click Analyze.
Review susceptibility score, corrosion-rate context, and susceptibility drivers.
Pin the baseline before running multiple what-if cases.

Recommended Operating Practice

Define a single objective for each analysis set.
Change only one variable or variable group in each what-if scenario.
Record one clear engineering conclusion for each analysis set.

Critical Mode Warning

Predictive and Inspection modes are not directly comparable without context translation.

When switching modes, treat the next run as a new analytical context and re-establish baseline before drawing conclusions.

Risk calculation basis

Both modes project from the last measured thickness to the target year. Inspection recency continues to inform Bayesian confidence, but it does not apply an additional reduction to the severe-state contribution. This avoids counting the same inspection evidence twice.

Before You Run

Complete the following checks to ensure reliable results and meaningful comparisons.

✔ Confirm that the intended analysis mode (Predictive or Inspection) is selected.
✔ Confirm that inspection-date context is representative for the assessed period.
✔ Verify whether the corrosion rate is measured or model-derived, and interpret the results accordingly.
✔ Verify that wet H₂S cracking and morphology settings are consistent with available inspection evidence.
✔ Ensure that all units and operating ranges are consistent with site records.
✔ Confirm that the assessment remains within carbon-steel NH₄HS sour-water scope and does not represent a multi-alloy case.

Wet H₂S Cracking Inspection Inputs

For the HIC/SOHIC wet H₂S cracking assessment, keep the age anchor separate from the compact inspection-history inputs:

Last qualifying Level A, B, or C wet H₂S cracking inspection year sets cracking age. Use the most recent qualifying inspection that found no cracking or confirmed repair of observed cracking.
Cracking inspection class and effective cracking inspection count provide the historical inspection effectiveness and count used for the base damage-factor lookup. Classes D and E remain valid history categories, but do not reset cracking age.

Sensitivity and remediation projections retain the qualifying inspection year. An inspection-upgrade recommendation can change the projected inspection class and monitoring credit, but does not reset cracking age until new qualifying inspection evidence is recorded.

Understanding the Model Outputs

Susceptibility Score: A summarized integrity indicator used for screening and prioritization.
Corrosion-Rate Context: The measured or modeled corrosion-rate basis used for the assessment.
Susceptibility Drivers: Relative contribution from Severity, Uncertainty, and Detection factors.
Baseline Comparison: Directional change relative to a prior case or pinned baseline case.

Interpretation Guidance

Treat the Susceptibility Score as a decision-support indicator rather than a standalone acceptance criterion.

Use the Susceptibility Drivers to determine whether improvement is best achieved through operating-envelope changes, improved inspection confidence, inspection quality, or a combination of these elements.

Status Pill Colour Mapping

The user interface uses status pills to provide a quick visual indication of the score band:

● Very Low
● Low
● Moderate
● High

These status pills are displayed in pathway cards and summary panels to indicate the mapped score band for the calculated logPf value. They provide a visual summary only and should always be interpreted together with the numerical results and the identified susceptibility drivers.

Reading the Results

Review the overall trend before focusing on the absolute score.
Identify the dominant driver family.
Determine whether the dominant contribution is operational (Severity) or related to data quality (Uncertainty or Detection).
Select the most appropriate response:

operational changes,
inspection improvements, or
a combination of both.

Decision Ladder

Is the score increasing compared with the baseline?
Which driver family contributes most to the result?
Is the available evidence sufficient to support a decision?
Select the appropriate response:

operational adjustment,
improved inspection or data quality,
or a combined mitigation strategy.

Decision Flowchart

Decision intent: Move from diagnostic output to one explicit action path, then verify its impact against the baseline.

Short Interpretation Examples

Example A: Severity Dominant

Pattern: High severity contribution, moderate uncertainty.

Interpretation: NH₄HS chemistry or velocity is likely the primary susceptibility driver.

Action focus: Adjust H₂S / NH₃ / pH₂S / velocity window where operationally feasible, then re-run analysis and evaluate score shift.

Example B: Uncertainty Dominant

Pattern: Uncertainty and detection limitations exceed severity.

Interpretation: Confidence is limited; the score may be overestimated or underestimated.

Action focus: Improve inspection quality, coverage, and data recency before changing operating conditions.

Decision Guidance by Driver Dominance

Driver-to-Action Matrix

Dominant Driver Pattern	Typical Meaning	First Recommended Move	Expected Direction
Severity > Uncertainty/Detection	Actual operating conditions likely drive the likelihood	Stabilize sour-water chemistry or velocity window and reassess	Score reduction
Uncertainty high	Data quality or age limits confidence	Improve inspection evidence and data quality	Confidence increase
Detection high	Detectability limitation	Adjust inspection method, interval, or coverage	Detectability improvement
Mixed profile	Multiple coupled mechanisms	Stage actions and validate after each step	Controlled convergence

Remediation Paths (Advanced Users)

What Remediation Paths Do

Remediation Paths mode converts susceptibility-driver outputs into actionable pathways and projects scenario impact.

Typical pathway families:

Defend Asset Life: operational change pathways (for example H₂S / NH₃ / pH₂S / velocity-window changes).
Inspection Evidence Refresh: inspection-evidence pathways that reset uncertainty recency and update Bayesian prior confidence; the selected wet-H₂S evidence-quality class remains fixed and this pathway adds no monitoring or inspection credit.
Monitoring and Inspection Upgrade: controls and monitoring pathways for trend protection, detection, and conservative tracking.

Defend Asset Life sensitivity is a decision-support preview. Full Optimization is the only sensitivity mode eligible to produce operating targets; Bounded and Diagnostic results are directional only, and Not Actionable means that no sensitivity matrix was run. Even in Full Optimization, targets are committed only when the backend permits recommendations and the best tested scenario produces a strict score reduction from the baseline. Otherwise, review the matrix without changing operating conditions.

Remediation Quick Reference (Example)

Pathway Families

Pathway Family	Typical Trigger	Typical Action	Expected Direction
Defend Asset Life	Severity-dominant susceptibility, aggressive sour-water chemistry or elevated velocity conditions	Adjust operating envelope	Score reduction
Eliminate Uncertainty	Uncertainty or detection drivers dominate	Upgrade inspection campaign and evidence quality	Confidence increase
Strengthen Monitoring	Trend instability or hidden acceleration concern	Enable monitoring and apply conservative control posture	Variability reduction

How to Use Remediation Paths

Run Analyze first and review the baseline score.
Ensure exactly one material is selected. Multi-material comparison remains available for normal analysis, but remediation must be evaluated and applied one material at a time because process-condition targets are shared across the active case.
Click Activate Remediation Paths.
Wait until the status changes to Remediation Paths Mode Active.
Review the Problem detected statement and the grouped pathway cards.
Select one or more pathway checkboxes.
Click Apply Selected Actions to commit eligible selected pathways. Defend Asset Life operating targets are applied only when the sensitivity result is recommendation-eligible and strictly lowers the baseline score.
Compare the recalculated scenario status and score to baseline.
Keep only practical pathways. Applied pathways remain marked Applied, accumulate in the report basket automatically, and cannot be selected twice; add engineering rationale separately when needed. The report retains the cumulative path list, while only the Prior/Baseline and latest Current cases are retained.

Changing inputs does not automatically recalculate the remediation projection. After editing inputs, click Analyze to exit and reset Remediation Paths mode, then review the new susceptibility drivers before activating remediation again. Applying selected actions is the explicit exception: the tool reruns Analyze automatically, validates the committed controls, adds the pathways to the report basket, and keeps Remediation Paths mode active.

What to Read in Each Pathway Card

Action: what operational or inspection step is proposed.
Result: expected outcome and projected score direction.
Engineering recommendation: API-aligned inspection/coverage guidance when relevant.
Top matrix scenarios: best candidate chemistry/velocity combinations for Defend Asset Life pathways.

Report Integration

After evaluating pathways, click Append Engineering Justification to Report to include selected pathway rationale in generated reporting output.

Practical Selection Guidance

Pick pathways that are implementable within your outage, access, and budget constraints.
Prefer pathways that reduce score and improve evidence quality together.
Do not select conflicting pathways only because each looks good in isolation.
Re-run Analyze after implementing real field changes to verify actual effect.

What Good Looks Like

Selected actions are technically feasible and operationally acceptable.
The projected reduction is meaningful and not achieved through unrealistic assumptions.
Data confidence improves, not only the score.
Follow-up inspection and monitoring steps are explicitly scheduled.

Important Limitations

Pathway projection is a decision-support estimate, not a guaranteed field outcome.
If no prior analysis exists, remediation evaluation cannot run.
Pathway logic is based on configured rule sets and model assumptions; engineering review is still mandatory.

Best practice: apply one practical pathway set at a time, then re-run and validate the real effect against baseline.

Common Mistakes to Avoid

Switching between Predictive and Inspection modes and interpreting values as directly equivalent.
Changing many inputs at once, then over-interpreting ranking order.
Treating one score as pass/fail without context.
Ignoring data quality warnings when uncertainty is dominant.

Troubleshooting

Open Troubleshooting Checklist

No result: verify required fields and date/year validity.
Unexpected ranking: verify mode, baseline, and toggles.
Counter-intuitive trend: verify corrosion-rate source, inspection context, chemistry inputs, and wet H₂S cracking flags.

Reporting Template (Short)

When documenting a run, include:

Run mode and context (asset, date, operating window).
Baseline score and dominant drivers.
Selected remediation pathway(s) and why.
Expected outcome and residual uncertainty.
Planned validation step (next inspection/monitoring check).

Report Quality Gate

What changed is explicitly identified.
Why it changed is technically justified.
Residual uncertainty is stated.
Verification step and timing are scheduled.

Status rule: mark Ready for Approval only when all items are checked.

Key Terms

Open Key Terms

Susceptibility Score

The reported susceptibility score is a failure-likelihood/integrity assessment indicator and should not be interpreted as a complete API RBI risk value derived from a likelihood × consequence framework. Consequence of failure is not included within the scope of this score.

Severity

Extent of expected damage from the identified degradation mechanism.

Uncertainty

Degree to which confidence is limited by data quality, completeness, or time gap.

Detection Limitation

Limitations in identifying damage due to inspection method or coverage.

Low Confidence in Historical Data

Data reliability issue due to incompleteness or inconsistency.

Outdated Inspection Data

Data age issue (staleness over time).

Localized Damages Present

Presence of localized morphology. The screening calculation applies a morphology factor of 10 to the Bayesian thinning damage factor, using an API 581-aligned screening surrogate for localized or pitting-type damage. In ranked drivers, this amplification is attributed to Severity because it acts on effective ART and the Bayesian thinning damage factor; it is not represented only as an inspection-detection limitation. This reflects approaches used in some API 581 applications where a pitting corrosion-rate contribution is amplified for screening. It is not a universal API 581 requirement, a direct localized-pitting corrosion-rate calculation, or a substitute for localized minimum-wall measurement or FFS assessment. The susceptibility score can therefore increase materially even when the corrosion rate is unchanged. The displayed remaining life remains a uniform-thinning estimate and should not be interpreted as localized-pitting FFS life.

Wet H₂S Cracking Present

Indicates potential wet H₂S cracking susceptibility. In NH₄HS systems, this condition does not automatically trigger an FFS requirement hard stop. A hard stop applies only when cracks are confirmed and remain unresolved. In such cases, wet H₂S cracking DF and cracking score are not calculated; the displayed susceptibility score is thinning-only until FFS disposition is completed.

Need Deeper Technical Detail?

For more details on IRI functionality, please contact Corrology Engineering Team.