Definition
Risk-Weighted Inspection Coverage Index (RWICI) measures whether inspection effort is proportional to consequence exposure (by risk tier) across the asset population.
The Measurement Gap
A petrochemical complex ran autonomous drone inspections across 2,400 piping circuits, 180 pressure vessels, and 340 structural steel elements. The inspection program manager was proud of the numbers: 94% of scheduled inspections completed on time, coverage across every process unit, and a 40% cost reduction compared to the previous rope-access and scaffold-based program. The KPIs were green. The board presentation looked excellent.
Then a 16-inch high-pressure steam header failed at a pipe support where corrosion under insulation had progressed to wall loss below minimum thickness. The failure caused an emergency shutdown, a steam release that triggered shelter-in-place for the adjacent unit, and a regulatory investigation. Total cost including lost production, emergency repair, and regulatory compliance: $3.2 million.
The investigation found that the failed pipe support was in Risk Tier 1: the highest consequence category in the complex's risk matrix. It was located in a congested area between two reactor vessels where drone access was physically constrained. The drone inspection program had classified this area as "limited access" and scheduled it for inspection once per 24 months instead of the 12-month cycle applied to open-access areas. The same limitation applied to 47 other Risk Tier 1 piping circuits in congested areas across the complex.
Meanwhile, Risk Tier 3 piping circuits in open pipe racks (low-pressure utility services with minimal consequence of failure) were being inspected every 6 months because the drones could easily access them. The inspection program was inspecting low-risk assets twice as frequently as high-risk assets. Not because someone made that decision consciously. Because the drone program optimized for access efficiency, and access is inversely correlated with risk in a petrochemical complex. The hardest things to reach are often the most critical.
Here is what the existing metrics showed: Inspection Completion Rate was 94%. Green. It counted every completed inspection equally, regardless of the risk tier of the asset inspected.
PM Compliance
tracked whether inspections were completed on schedule, and they were, because the schedule itself was the problem. A schedule that inspects low-risk assets frequently and high-risk assets infrequently is still 100% compliant if every scheduled inspection gets done.
The measurement gap: there is no metric that tells you whether inspection coverage is proportional to the risk profile of the asset population, which means high-consequence assets may be underinspected relative to their risk while low-consequence assets absorb disproportionate inspection effort.
The petrochemical complex was not running a bad inspection program. It was running a program that optimized for the wrong variable: coverage volume rather than risk-weighted coverage. Every metric they tracked rewarded completing inspections. No metric asked whether those inspections were allocated to where the risk actually lived.
Formula
Risk-Weighted Inspection Coverage Index (RWICI) measures whether inspection effort is proportional to consequence exposure (by risk tier) across the asset population. It answers the question IDAR cannot: not "are we finding what's there?" but "are we looking where it matters?"
In plain language: Divide your assets into risk tiers using whatever criticality framework you already have. For each tier, compute what percentage of total inspection effort goes to that tier. Compare it to what percentage of total risk that tier represents. A ratio of 1.0 means the tier gets exactly the inspection effort its risk warrants. Below 1.0 means underinspection relative to risk. Above 1.0 means overinspection.
The formula:
RWICI_tier = (IE_tier / IE_total) / (R_tier / R_total)
Where:
- IE_tier = inspection effort allocated to the tier (inspection hours, frequency, or cost: use whichever your program tracks consistently; hours are preferred because they normalize across methods)
- IE_total = total inspection effort across all tiers
- R_tier = quantified risk exposure for the tier (sum of consequence × probability across all assets in the tier)
- R_total = total risk exposure across all tiers
Worked example:
The petrochemical complex classifies its 2,400 piping circuits into four risk tiers:
Tier 1 (Critical): 180 circuits, 52% of total risk, 1,240 annual inspection hours (18% of effort). RWICI = 0.35.
Tier 2 (High): 460 circuits, 28% of total risk, 2,100 hours (31% of effort). RWICI = 1.11.
Tier 3 (Medium): 820 circuits, 14% of total risk, 2,350 hours (34% of effort). RWICI = 2.43.
Tier 4 (Low): 940 circuits, 6% of total risk, 1,150 hours (17% of effort). RWICI = 2.83.
Read the RWICI column. Tier 1 (the highest-risk circuits, representing 52% of total risk) receives 18% of inspection effort. RWICI = 0.35. These assets are getting roughly one-third of the inspection coverage their risk warrants. Tier 4 (the lowest-risk circuits, representing 6% of total risk) receives 17% of inspection effort. RWICI = 2.83. These assets are getting nearly three times the coverage their risk warrants.
The access-bias pattern is visible in the numbers: Tier 1 circuits are in congested areas with limited drone access. Tier 4 circuits are in open pipe racks. The inspection program has been optimizing for throughput, and throughput correlates with access, and access inversely correlates with risk.
The gap between RWICI of 0.35 on Tier 1 and RWICI of 2.83 on Tier 4 is the quantified access bias. It tells you exactly how much inspection effort needs to be reallocated (and from where to where) to bring coverage in line with risk.
Edge cases:
Assets with no assigned risk tier: exclude from RWICI and flag as "unclassified." The existence of unclassified assets in a risk-managed inspection program is itself a governance finding.
Inspection types with different detection capabilities: weight inspection effort by detection effectiveness if IDAR data is available. An hour of drone inspection with IDAR of 60% provides less effective coverage than an hour of manual inspection with IDAR of 90%. The quality-adjusted effort would be: Effective_IE = IE_hours × IDAR_composite.
Risk quantification methodology varies: RWICI works with any risk quantification approach as long as it produces a per-asset or per-tier score. API 580/581, FMEA-based risk ranking, semi-quantitative risk matrices: all produce scores that can be summed within tiers. The important thing is consistency: use the same methodology across all tiers.
Proposed formula, pending validation pilots.