8.1 Combining POF and COF: Risk Calculation and Risk Plotting

8.1 Combining POF and COF: Risk Calculation and Risk Plotting

Introduction to Combining POF and COF in API RP 580

API Recommended Practice 580 (4th Edition, August 2023 with Addenda 1, March 2025) defines Risk as the mathematical combination of the Probability of Failure (POF) and the Consequence of Failure (COF) resulting from an unplanned loss of containment (LOPC) event. In Section 12 ("Risk Determination, Assessment, and Management"), API RP 580 outlines how these two independent analytical dimensions are integrated to quantify equipment risk, plot risk profiles, and establish risk-based inspection priorities across refining, petrochemical, and chemical processing facilities.

Risk determination is the central analytical bridge between damage mechanism evaluation (API RP 571), inspection history review (API 510, API 570, API 653), and inspection planning (API RP 580 Section 13). By explicitly combining probability and consequence, RBI shifts plant maintenance from traditional calendar-based inspection schedules to a risk-engineered framework where resources are concentrated on assets presenting the highest potential business disruption, environmental damage, or safety impact.

---

Quantitative vs. Qualitative Risk Calculations (API RP 580 Section 12)

API RP 580 recognizes a continuum of risk assessment methodologies ranging from qualitative to fully quantitative. The method chosen for combining POF and COF depends on data availability, asset complexity, facility risk criteria, and regulatory requirements:

1. Quantitative Risk Calculation: In quantitative RBI models (such as API RP 581), risk is calculated as a continuous numerical value representing the expected loss per unit time: $\text{Risk}(t) = \text{POF}(t) \times \text{COF}$ Where $\text{POF}(t)$ is the annual failure frequency (\text{failures/year}) at operating time $t$, and $\text{COF}$ is the quantified outcome magnitude (\text{ft}^2, \text{m}^2, or $ per event).

2. Qualitative Risk Determination: In qualitative RBI models, POF and COF are evaluated using engineering judgment, historical operational experience, and categorized descriptor tables (e.g., High, Medium, Low). Combining POF and COF involves mapping discrete qualitative categories onto a two-dimensional grid to yield an ordinal risk ranking (e.g., Low, Medium, Medium-High, High).

3. Semi-Quantitative Risk Determination: Semi-quantitative approaches combine structured numerical scoring algorithms for POF (e.g., assigning points for damage rates, age, and maintenance practices) with broad consequence categorization brackets to plot risk on a matrix without requiring full rigorous thermodynamic release modeling.

---

Dimensional Units of Calculated Risk

A critical concept for the API 580 examination is understanding the dimensional units resulting from quantitative risk calculations. Depending on whether consequence is measured in terms of physical release area or financial monetary loss, risk is expressed in two primary unit systems:

1. Area-Based Risk (\text{ft}^2/\text{yr} or \text{m}^2/\text{yr})

When evaluating safety, health, flammable, or toxic consequences, $\text{COF}$ is quantified as a physical impact area (\text{ft}^2 or \text{m}^2) representing the envelope where personnel exposure or structural damage exceeds hazardous thresholds (e.g., thermal radiation of $5 \text{ kW/m}^2$, toxic IDLH concentration, or blast overpressure of $1 \text{ psi}$): $\text{Risk}_{\text{area}}(t) \left[\frac{\text{ft}^2}{\text{yr}}\right] = \text{POF}(t) \left[\frac{\text{failures}}{\text{yr}}\right] \times \text{COF}_{\text{area}} \left[\frac{\text{ft}^2}{\text{failure}}\right]$

2. Financial / Monetary Risk ($/\text{yr})

When evaluating economic impact—including business interruption, equipment replacement cost, environmental remediation, and regulatory fines—$\text{COF}$ is quantified in monetary terms ($): $\text{Risk}_{\text{financial}}(t) \left[\frac{\$}{\text{yr}}\right] = \text{POF}(t) \left[\frac{\text{failures}}{\text{yr}}\right] \times \text{COF}_{\text{financial}} \left[\frac{\$}{\text{failure}}\right]$

Understanding both unit systems enables facilities to set separate risk acceptance criteria for life safety vs business financial risk.

---

Iso-Risk Lines and Mathematical Trajectories

When plotting risk on a logarithmic scale with Probability of Failure on the vertical Y-axis and Consequence of Failure on the horizontal X-axis, lines of constant risk are referred to as Iso-Risk Lines or Iso-Risk Curves.

Mathematically, an iso-risk line represents the locus of points where the product of POF and COF equals a fixed constant $K$: $\text{POF} \times \text{COF} = K \implies \log(\text{POF}) + \log(\text{COF}) = \log(K)$ $\log(\text{POF}) = -\log(\text{COF}) + \log(K)$

On a log-log plot, this equation forms a straight diagonal line with a slope of $-1$. All points along a given iso-risk line represent identical numerical risk:

• An asset with a high POF ($1.0 \times 10^{-2}/\text{yr}$) and a low COF ($100 \text{ ft}^2$) has a risk of $1.0 \text{ ft}^2/\text{yr}$.
• An asset with a low POF ($1.0 \times 10^{-4}/\text{yr}$) and a high COF ($10,000 \text{ ft}^2$) has the exact same risk of $1.0 \text{ ft}^2/\text{yr}$.

Iso-risk lines allow risk analysts to immediately identify whether an asset's risk is driven primarily by probability (requiring inspection or degradation mitigation) or by consequence (requiring inventory reduction, isolation valves, or consequence mitigation).

---

Dynamic Risk Trajectories Over Operational Time

In API RP 580 risk models, risk is not static; it evolves dynamically over operational time $t$. The dynamic behavior of risk is governed by the time-dependent nature of Probability of Failure:

$\text{POF}(t) = \text{GFF} \times D_{\text{total}}(t) \times F_{\text{MS}}$

Where:

• $\text{GFF}$ is the baseline Generic Failure Frequency for the equipment type.
• $D_{\text{total}}(t)$ is the Total Damage Factor, which increases over time as active damage mechanisms (such as thinning, cracking, or creep) consume material design margins.
• $F_{\text{MS}}$ is the Management System Factor.

Because $D_{\text{total}}(t)$ increases monotonically with operating exposure time, $\text{POF}(t)$ climbs upward on the risk plot over time. Assuming process chemical inventory and operating conditions remain unchanged, $\text{COF}$ remains constant. Consequently, an equipment item moves vertically upward across iso-risk lines toward higher risk categories over time.

The Role of Inspection in Resetting Risk Trajectories

Performing non-destructive examination (NDE) does not physically repair thin metal wall or eliminate cracks. However, high-effectiveness inspections gather empirical data, which reduces uncertainty in calculated damage rates. In API quantitative models, updating inspection credit reduces the calculated Damage Factor $D_{\text{total}}(t)$, which drops the calculated $\text{POF}(t)$ downward back to a lower risk level, resetting the equipment's trajectory along the time axis.

---

Worked Engineering Example: Distillation Column Overhead Circuit Risk Plotting

Problem Statement

A crude unit Overhead Distillate Condenser Shell operating at $260^\circ\text{F}$ has the following parameter baselines:

• Equipment Type: Carbon Steel Heat Exchanger Shell ($\text{GFF} = 1.0 \times 10^{-4} \text{ failures/yr}$).
• Active Damage Mechanism: HCl Corrosion and Corrosion Under Insulation (CUI).
• Consequence of Failure: Flammable release area $\text{COF} = 25,000 \text{ ft}^2$; Financial consequence $\text{COF}_{\text{financial}} = $1,200,000$.
• Management System Factor: $F_{\text{MS}} = 1.0$.
• Initial Damage Factor ($t=0$ years after last inspection): $D_{\text{total}}(0) = 10.0$.
• Annual Damage Factor Increase Rate: $\Delta D_{\text{total}} = 15.0 \text{ per year}$.
• Facility Target Risk Acceptance Threshold: $\text{Risk}_{\text{target}} = 100 \text{ ft}^2/\text{yr}$.

Calculation Steps

1. Calculate Initial Risk at $t=0$: $\text{POF}(0) = (1.0 \times 10^{-4}) \times 10.0 \times 1.0 = 1.0 \times 10^{-3} \text{ failures/yr}$ $\text{Risk}_{\text{area}}(0) = (1.0 \times 10^{-3}) \times 25,000 \text{ ft}^2 = 25.0 \text{ ft}^2/\text{yr}$ Since $25.0 \text{ ft}^2/\text{yr} < 100 \text{ ft}^2/\text{yr}$, the asset is currently operating within acceptable risk limits.

2. Project Risk Trajectory at $t = 5$ Years: $D_{\text{total}}(5) = 10.0 + (15.0 \times 5) = 85.0$ $\text{POF}(5) = (1.0 \times 10^{-4}) \times 85.0 \times 1.0 = 8.5 \times 10^{-3} \text{ failures/yr}$ $\text{Risk}_{\text{area}}(5) = (8.5 \times 10^{-3}) \times 25,000 \text{ ft}^2 = 212.5 \text{ ft}^2/\text{yr}$ At 5 years, the calculated risk ($212.5 \text{ ft}^2/\text{yr}$) exceeds the target risk threshold ($100 \text{ ft}^2/\text{yr}$).

3. Determine Target Inspection Date ($t_{\text{target}}$): Set $\text{Risk}{\text{area}}(t{ ext{target}}) = 100 \text{ ft}^2/\text{yr}$: $\text{POF}(t_{\text{target}}) = \frac{100 \text{ ft}^2/\text{yr}}{25,000 \text{ ft}^2} = 4.0 \times 10^{-3} \text{ failures/yr}$ Solve for $D_{\text{total}}(t_{ ext{target}})$: $D_{\text{total}}(t_{\text{target}}) = \frac{4.0 \times 10^{-3}}{\text{GFF} \times F_{\text{MS}}} = \frac{4.0 \times 10^{-3}}{1.0 \times 10^{-4} \times 1.0} = 40.0$ Solve for $t_{\text{target}}$: $40.0 = 10.0 + 15.0 \times t_{\text{target}} \implies 15.0 \times t_{\text{target}} = 30.0 \implies t_{\text{target}} = 2.0 \text{ years}$

Conclusion: To prevent equipment risk from exceeding the facility target threshold of $100 \text{ ft}^2/\text{yr}$, a high-effectiveness inspection must be performed within 2.0 years.