1.2 Absolute vs. Relative Risk and Risk Acceptability Criteria

Absolute vs. Relative Risk Assessment Frameworks

API RP 580 Section 4.6 outlines two primary analytical methodologies for assessing risk across fixed equipment assets: Absolute Risk Assessment and Relative Risk Assessment. While both approaches share the ultimate goal of identifying high-risk assets and optimizing inspection strategies, they differ fundamentally in mathematical formulation, input data requirements, and application scope.

┌────────────────────────────────────────────────────────────────────────┐
│                     API RP 580 RISK METHODOLOGIES                      │
├───────────────────────────────────┬────────────────────────────────────┤
│       ABSOLUTE RISK METHOD        │        RELATIVE RISK METHOD        │
├───────────────────────────────────┼────────────────────────────────────┤
│ • Numeric units (ft²/yr, $/yr)    │ • Categorical ranking (High/Med/Low)│
│ • Rigorous probabilistic models    │ • Ordinal scoring (1 to 50 scale)  │
│ • Required for regulatory compliance│ • Ideal for turnaround scoping     │
│ • High data & resource intensity  │ • Efficient for large plant units  │
└───────────────────────────────────┴────────────────────────────────────┘

1. Absolute Risk Assessment

An Absolute Risk Assessment calculates equipment risk as a precise numerical quantity expressed in physical, safety, or economic units of measure:

• Safety & Health: Expected fatalities or serious injuries per year (e.g., $1.5 \times 10^{-5} \text{ fatalities/year}$).
• Flammable / Toxic Release Area: Annual expected impact area (e.g., $450 \text{ ft}^2/\text{year}$).
• Financial Loss: Total monetary risk (e.g., $$25,000/\text{year}$).

To achieve valid absolute risk numbers, the methodology must utilize rigorous, statistically validated failure rate databases, precise structural reliability models, and sophisticated fluid discharge and atmospheric dispersion algorithms (e.g., WHAZAN or PHAST software). Absolute risk assessment is necessary when evaluating facility safety against statutory regulatory criteria, societal risk curves (F-N curves), or strict corporate absolute financial risk caps.

2. Relative Risk Assessment

A Relative Risk Assessment evaluates and ranks equipment items in comparison to one another within a specific process unit, plant facility, or corporate enterprise. Rather than generating absolute unit values, relative risk methodologies output ordinal rankings (e.g., Equipment Item #1 to #250) or categorical risk levels (High, Medium-High, Medium, Low).

Relative risk frameworks are highly practical because they do not depend on knowing the exact absolute baseline failure frequency of a vessel. Instead, they evaluate relative factors—such as whether Piping Circuit A has twice the corrosion rate of Piping Circuit B or whether Vessel X contains 10 times the toxic fluid inventory of Vessel Y. This approach allows inspection engineers to rapidly filter plant assets, directing immediate inspection resources toward the highest relative risk drivers without spending months building rigorous dispersion models.

---

Risk Analysis Methodologies: Qualitative, Quantitative, and Semi-Quantitative

API RP 580 classifies risk assessment approaches into three broad categories along a continuum of analytical complexity and data intensity:

Low Data Requirements                                    High Data Requirements
Low Resource Intensity                                   High Resource Intensity
┌──────────────────────┬────────────────────────┬────────────────────────┐
│     QUALITATIVE      │    SEMI-QUANTITATIVE   │      QUANTITATIVE      │
├──────────────────────┼────────────────────────┼────────────────────────┤
│ • Engineering judgment│ • Structured scoring   │ • Detailed failure algorithms│
│ • 5x5 Categorical    │ • Bounded algorithms   │ • Dispersion physics   │
│   Risk Matrices      │ • Hybrid precision     │ • Numeric risk values  │
└──────────────────────┴────────────────────────┴────────────────────────┘

1. Qualitative Risk Analysis

Qualitative risk analysis relies on engineering judgment, expert consensus, and descriptive categorical scaling (e.g., Low, Medium, High). Input data is grouped into broad ranges, and Probability of Failure (POF) and Consequence of Failure (COF) are plotted on a discrete $5 \times 5$ Risk Matrix.

• Advantages: Fast implementation, low cost, requires minimal detailed historical process data.
• Limitations: Highly dependent on assessor experience, susceptible to subjective bias, lacks granularity to differentiate assets within the same matrix cell.

2. Quantitative Risk Analysis (QRA)

Quantitative risk analysis utilizes explicit mathematical and probabilistic algorithms to calculate exact numerical POF and COF values. Equipment thickness data, corrosion rates, NDE confidence levels, fluid thermodynamic properties, isolable inventory masses, and weather dispersion parameters are processed through structural reliability and physical discharge equations.

• Advantages: Highly objective, repeatable, generates absolute risk numbers suitable for cost-benefit economic optimization.
• Limitations: High implementation cost, intensive data collection requirements, requires specialized software and risk engineering expertise.

3. Semi-Quantitative Risk Analysis

Semi-Quantitative risk analysis combines the speed and structure of qualitative categorization with the mathematical consistency of quantitative algorithms. Input parameters are assigned numerical scores or bounded mathematical ranges (e.g., scoring thinning damage from 1 to 100 based on measured wall loss ratios), which are then mapped into discrete risk matrix categories.

• Advantages: Strikes an optimal balance between accuracy, repeatability, and computational effort; forms the core architecture of industry standards such as API RP 581.

---

Methodological Comparison Matrix

---

Risk Acceptability Criteria & The ALARP Principle

Defining Risk Acceptability Criteria is a fundamental requirement under API RP 580 Section 4. Risk criteria establish the boundaries separating acceptable operational risk from unacceptable risk requiring mandatory engineering mitigation.

The ALARP Framework

API RP 580 adopts the ALARP (As Low As Reasonably Practicable) framework, which divides risk into three distinct management regions:

▲ High Risk
│  ┌────────────────────────────────────────────────────────┐
│  │ UNACCEPTABLE REGION                                    │
│  │ Risk is intolerable. Mandatory immediate mitigation   │
│  │ regardless of cost.                                    │
│  └────────────────────────────────────────────────────────┘
│  ┌────────────────────────────────────────────────────────┐
│  │ ALARP (TOLERABLE) REGION                               │
│  │ Risk is tolerable ONLY IF further mitigation is       │
│  │ cost-disproportionate to the risk reduction gained.   │
│  └────────────────────────────────────────────────────────┘
│  ┌────────────────────────────────────────────────────────┐
│  │ BROADLY ACCEPTABLE REGION                              │
│  │ Risk is low. Routine inspection monitoring maintained. │
│  └────────────────────────────────────────────────────────┘
▼ Low Risk

1. Unacceptable Risk Region: Risk levels exceed corporate safety or financial thresholds. Operation without immediate risk mitigation (e.g., targeted NDE, process derating, inventory reduction, shutdown) is prohibited.
2. ALARP (Tolerable) Region: Risk is tolerated only if risk reduction measures are implemented unless an engineering economic evaluation demonstrates that the cost of further risk reduction is grossly disproportionate to the risk reduction benefit achieved.
3. Broadly Acceptable Region: Risk is sufficiently low that additional capital expenditure for risk reduction is unnecessary. Baseline regulatory and API code minimum inspections are maintained.

Mathematical Formulation of Iso-Risk Contours

On a log-log Risk Matrix where $\text{POF}$ is plotted on the vertical axis and $\text{COF}$ is plotted on the horizontal axis, a constant risk target threshold (Iso-Risk line) forms a straight diagonal boundary with a negative unit slope:

$\text{Risk}_{\text{target}} = \text{POF} \times \text{COF}$

Taking the natural logarithm of both sides:

$\ln(\text{POF}) = \ln(\text{Risk}_{\text{target}}) - \ln(\text{COF})$

Any equipment item lying above and to the right of the Iso-Risk line exceeds the acceptable risk criterion, whereas items below and to the left remain within acceptable boundaries.

---

Technical Worked Example: ALARP Evaluation & Mitigation Trade-Off

Scenario

A Hydrotreater Reactor Vessel operates with a calculated flammable release consequence $\text{COF} = 20,000 \text{ ft}^2$.

• Corporate Risk Acceptability Target: $\text{Risk}_{\text{target}} = 500 \text{ ft}^2/\text{year}$.
• Baseline Assessment: High-Temperature Sulfidation corrosion without recent internal thickness data yields a calculated $\text{POF} = 0.04 \text{ failures/year}$.

1. Initial Risk Calculation

$\text{Risk}_{\text{initial}} = 0.04 \times 20,000 \text{ ft}^2 = 800 \text{ ft}^2/\text{year}$ Result: Exceeds the $500 \text{ ft}^2/\text{yr}$ target threshold (Unacceptable Region).

2. Mitigation Evaluation Option A (Inspection Mitigation)

Execute targeted internal automated Ultrasonic Thickness (AUT) grid mapping during a minor outage at a cost of $$30,000$.

• AUT confirms uniform low corrosion rate, reducing inspection uncertainty and lowering the calculated Damage Factor.
• Updated $\text{POF}_{\text{AUT}} = 0.01 \text{ failures/year}$.
• Updated Risk: $\text{Risk}_{\text{AUT}} = 0.01 \times 20,000 = 200 \text{ ft}^2/\text{year}$ (Enters Acceptable Region).

3. Mitigation Evaluation Option B (Hardware Consequence Mitigation)

Install remote-actuated Emergency Isolation Valves (EIV) to isolate liquid inventory in the event of a leak at a cost of $$450,000$.

• EIV reduces isolable liquid inventory mass, lowering calculated release consequence to $\text{COF}_{\text{EIV}} = 6,000 \text{ ft}^2$.
• Updated Risk: $\text{Risk}_{\text{EIV}} = 0.04 \times 6,000 = 240 \text{ ft}^2/\text{year}$ (Enters Acceptable Region).

ALARP Decision Analysis

Both options successfully bring equipment risk below the $500 \text{ ft}^2/\text{yr}$ threshold. However, Option A achieves risk reduction at a cost of $$30,000$, whereas Option B costs $$450,000$ for a comparable risk reduction. Under the ALARP principle, Option A represents the most cost-effective risk reduction measure, demonstrating that Option B's excessive expenditure is grossly disproportionate to the incremental safety gain.