3.1 Corrosion Averaging & Minimum Thickness Evaluation

Corrosion Averaging & Minimum Thickness Evaluation

In pressure vessel in-service inspection, metal loss rarely occurs in a perfectly uniform pattern across an entire shell course or head. Instead, localized corrosion mechanisms—such as condensate grooving, acid under-deposit attack, impingement, and flow-accelerated degradation—create Locally Thinned Areas (LTAs). When an inspector encounters an area where the measured thickness falls below the minimum required thickness ($t_{\text{min}}$) established by the construction code (ASME Section VIII, Division 1), the vessel does not automatically fail inspection or require immediate retirement.

API 510 Section 7.4 establishes rigorous engineering evaluation procedures that allow inspectors to evaluate LTAs through corrosion averaging. By assessing the structural reinforcement provided by the surrounding thicker base metal and properly orienting the evaluation length along the principal stress plane, an inspector can determine whether the component remains safe for continued operation at its design Maximum Allowable Working Pressure (MAWP).

---

1. Mechanics of Stress and Orientation of Evaluation Length

To understand why API 510 imposes strict directional rules on thickness averaging, an inspector must understand the fundamental membrane stresses operating in thin-walled pressure vessels.

Principal Membrane Stresses in Cylindrical Shells

Under internal pressure ($P$), a cylindrical shell experiences two primary membrane stresses:

1. Circumferential (Hoop) Stress ($\sigma_h$): $\sigma_h = \frac{P \cdot R}{t}$ Acts tangentially around the circumference, tending to split the cylinder longitudinally along its length.

2. Longitudinal (Axial) Stress ($\sigma_L$): $\sigma_L = \frac{P \cdot R}{2t} = \frac{1}{2} \sigma_h$ Acts parallel to the vessel's longitudinal axis, tending to pull the cylinder apart end-to-end.

Because hoop stress is twice the magnitude of longitudinal stress, the longitudinal plane is the most highly stressed cross-section in a cylindrical vessel. Consequently, circumferential stress governs the required wall thickness for cylindrical shells under internal pressure.

+-----------------------------------------------------------------------------------------+
|                        PRINCIPAL STRESS VECTORS ON A CYLINDER                           |
|                                                                                         |
|               ^ Circumferential (Hoop) Stress (σ_h = P*R/t)   [MAX STRESS]              |
|               |                                                                         |
|         +-----+---------------------------------------------------+-----+               |
|         |     |                                                   |     |               |
|   <-----+=====[=========== CRITICAL LENGTH L AXIS ===============]=====+----->         |
|         |     |   (Evaluated longitudinally along the cylinder)   |     |  Longitudinal |
|         +-----+---------------------------------------------------+-----+  Stress (σ_L) |
|               |                                                                         |
|               v Circumferential (Hoop) Stress (σ_h)                                     |
|                                                                                         |
|   KEY CODE PRINCIPLE: The critical length L must be oriented LONGITUDINALLY along the    |
|   cylinder axis so that the hoop stress vector acts perpendicularly across L.           |
+-----------------------------------------------------------------------------------------+

Code Rules for Orienting the Critical Length $L$

According to API 510 Section 7.4.2.1:

• For Cylindrical Shells: The critical length $L$ must be oriented parallel to the longitudinal axis of the vessel. Averaging along the circumferential direction is prohibited for hoop stress qualification because doing so would average across the maximum tensile stress field rather than along the resisting cross-section.
• For Spherical Shells and Hemispherical Heads: Because membrane stress is equal in all tangential directions ($\sigma = \frac{P \cdot R}{2t}$), the critical length $L$ may be oriented along any great circle arc.
• For Torispherical and Ellipsoidal Heads: The critical length $L$ is evaluated along the arc of the head profile, with special attention to the high-stress knuckle or toroidal transition zone.

---

2. API 510 Critical Length ($L$) Formulas

The maximum length over which thickness measurements may be averaged is called the Critical Length ($L$). It depends strictly on the inside diameter ($D$) of the vessel section containing the thinned area.

Mathematical Definition of Critical Length $L$

Diameter Transition & Length Caps Table

To avoid calculation errors on the API 510 exam, memorize how the diameter cutoff triggers both the divisor change ($D/2$ vs $D/3$) and the upper ceiling cap ($20\text{ in.}$ vs $40\text{ in.}$):

[!IMPORTANT] Exam Watchpoint: Notice that for a $48\text{ in.}$ vessel, $D/2 = 24\text{ in.}$, but the maximum length is capped at $20\text{ in.}$. For a $72\text{ in.}$ vessel ($D > 60\text{ in.}$), the formula shifts immediately to $D/3 = 24\text{ in.}$, which is under the $40\text{ in.}$ cap. Confusing these two rules is one of the most common calculation traps on the API 510 exam!

---

3. Thickness Averaging Procedure and Grid Requirements

When performing thickness averaging across the critical length $L$, the inspector must follow a disciplined measurement protocol:

1. Locate the Minimum Point: Use ultrasonic thickness gauging (UT) or profile radiography to scan the locally thinned area and locate the absolute minimum thickness reading ($t_{\text{act, min}}$).
2. Lay Out Critical Length $L$: Center the line of length $L$ over the lowest thickness reading, oriented parallel to the longitudinal axis of the vessel.
3. Select Measurement Spacing: Take a series of equally spaced thickness readings along line $L$. API 510 recommends taking multiple readings (typically at least 5 to 10 points spaced no more than $1\text{ in.}$ to $2\text{ in.}$ apart) to accurately characterize the profile.
4. Calculate Average Thickness ($t_{\text{avg}}$): $t_{\text{avg}} = \frac{1}{N} \sum_{i=1}^{N} t_i = \frac{t_1 + t_2 + t_3 + \dots + t_N}{N}$
5. Evaluate Minimum Point Criteria: In addition to the average meeting $t_{\text{min}}$, the lowest single reading within the averaged length must not be less than the minimum thickness limits permitted by API 510 or Fitness-for-Service Level 1 criteria (typically not less than $0.5 \times t_{\text{required}}$ without an API 579 evaluation).

+-----------------------------------------------------------------------------------------+
|                         THICKNESS PROFILE ALONG CRITICAL LENGTH L                       |
|                                                                                         |
|   Wall Thickness (in.)                                                                  |
|   0.500 |--- Nominal Thickness (t_nom = 0.500") -----------------------------------    |
|         |                                                                               |
|   0.375 |--- Required Thickness (t_req = 0.375") ----------------------------------    |
|         |                                                                               |
|   0.350 |       t1=0.380"                                               t5=0.390"       |
|         |          \         t2=0.340"                 t4=0.350"       /               |
|   0.300 |           \           \       t3=0.290"       /             /                |
|         |            *-----------*-----(MIN POINT)-----*-------------*                  |
|         |                                  *                                            |
|         +----------------------------------|------------------------------------>       |
|         |<-               CRITICAL LENGTH L = 20.0"                         ->|       |
|                                                                                         |
|   AVERAGE CALCULATION:                                                                  |
|   t_avg = (0.380 + 0.340 + 0.290 + 0.350 + 0.390) / 5 = 1.750 / 5 = 0.350"             |
+-----------------------------------------------------------------------------------------+

---

4. Joint Efficiency ($E$) Rules for Local Thin Areas

One of the most powerful and tested concepts in API 510 is the application of weld joint efficiency ($E$) during the evaluation of local thin areas.

The Base Metal Exemption ($E = 1.0$)

In ASME Section VIII, Division 1, shell thickness calculations for welded vessels include a weld joint efficiency factor $E$ (e.g., $E = 0.85$ for spot radiography or $E = 0.70$ for no radiography) to account for potential weld defects.

However, API 510 Section 7.4.2.2 states:

When the locally thinned area is located entirely within the base metal and does not touch or cross a weld joint, the required thickness of the thinned area may be calculated using a joint efficiency of $E = 1.0$ (seamless vessel equivalent), provided the LTA is located away from the weld by a distance of at least the greater of:

1. $1.0\text{ in.}$ ($25\text{ mm}$), or
2. $2 \times t$ (where $t$ is the nominal wall thickness).

Rationale for $E = 1.0$

Weld joint efficiency is a penalty applied to the weld seam itself due to the possibility of volumetric or planar weld flaws. The base plate away from the weld is seamless wrought material and does not possess weld defects. Therefore, if corrosion occurs strictly in the base metal, penalizing the corroded plate with the weld joint efficiency of a remote seam is overly conservative.

Weld Inclusion Rule

If the locally thinned area lies on, across, or within the heat-affected zone (HAZ) of a weld (i.e., closer than $\max(1.0\text{ in.}, 2t)$ to the weld toe), the calculation of required thickness for that LTA must use the actual joint efficiency ($E$) of the affected weld seam.

---

5. Comprehensive Step-by-Step Calculation Example

Let us work through a complete, open-book style API 510 calculation problem demonstrating every aspect of corrosion averaging, critical length determination, joint efficiency selection, and remaining life assessment.

Problem Statement

A vertical process separator has the following design and inspection data:

• Design Pressure ($P$): $250\text{ psig}$
• Design Temperature: $400^\circ\text{F}$
• Inside Diameter ($D$): $48.0\text{ in.}$ (Inside Radius $R = 24.0\text{ in.}$)
• Allowable Stress ($S$): $20,000\text{ psi}$ (SA-516 Gr. 70)
• Original Longitudinal Weld Joint Efficiency ($E$): $0.85$ (Type 1 Butt Weld, Spot Radiographed per UW-52)
• Nominal Shell Thickness ($t_{\text{nom}}$): $0.500\text{ in.}$
• Corrosion Allowance ($CA$): $0.050\text{ in.}$ required for next 5-year run

During an internal turnaround inspection, an isolated local thin area is discovered on the shell. Ultrasonic readings along a longitudinal line centered on the lowest reading yield the following 5 measurements spaced 4 inches apart (total span $= 16\text{ in.}$): $\text{Readings: } [0.320\text{ in.},\; 0.290\text{ in.},\; 0.265\text{ in.},\; 0.295\text{ in.},\; 0.330\text{ in.}]$

Evaluate the acceptability of this LTA for two separate cases:

• Case A: The LTA is located in the middle of the base plate, $8\text{ in.}$ away from all welds.
• Case B: The LTA runs directly along the longitudinal weld seam.

---

Step 1: Determine the Critical Evaluation Length ($L$)

Check the vessel inside diameter: $D = 48.0\text{ in.} \le 60\text{ in.}$

Apply the critical length formula for $D \le 60\text{ in.}$: $L = \min\left(\frac{D}{2},\; 20\text{ in.}\right) = \min\left(\frac{48}{2},\; 20\right) = \min(24\text{ in.},\; 20\text{ in.}) = \mathbf{20.0\text{ in.}}$

The 16-inch span of readings fits completely within the allowable 20.0-inch critical length $L$.

---

Step 2: Calculate the Average Thickness ($t_{\text{avg}}$)

$t_{\text{avg}} = \frac{0.320 + 0.290 + 0.265 + 0.295 + 0.330}{5} = \frac{1.500}{5} = \mathbf{0.300\text{ in.}}$

---

Step 3: Evaluate Case A (LTA in Base Metal Away from Welds)

Check distance from weld: Distance $= 8.0\text{ in.}$ Threshold distance $= \max(1.0\text{ in.},; 2 \times t_{\text{nom}}) = \max(1.0,; 2 \times 0.500) = \max(1.0,; 1.0) = 1.0\text{ in.}$ Since $8.0\text{ in.} \ge 1.0\text{ in.}$, the base metal exemption applies, and we can use $E = 1.0$.

Calculate minimum required thickness ($t_{\text{min}}$) using the ASME Section VIII Div 1 circumferential stress formula (UG-27(c)(1)): $t_{\text{min}} = \frac{P \cdot R}{S \cdot E - 0.6 \cdot P}$ $t_{\text{min}} = \frac{250 \cdot 24.0}{(20,000 \cdot 1.0) - (0.6 \cdot 250)} = \frac{6,000}{20,000 - 150} = \frac{6,000}{19,850} = \mathbf{0.3023\text{ in.}}$

Total required thickness including future corrosion allowance ($CA = 0.050\text{ in.}$): $t_{\text{required}} = t_{\text{min}} + CA = 0.3023 + 0.050 = \mathbf{0.3523\text{ in.}}$

Comparison for Case A:

• Measured $t_{\text{avg}} = 0.300\text{ in.}$
• Minimum required base metal thickness $t_{\text{min}} = 0.3023\text{ in.}$
• Since $t_{\text{avg}} < t_{\text{min}}$, the vessel cannot continue operating at 250 psig without either derating the MAWP, performing an API 579 Level 2 FFS assessment, or executing a weld buildup repair.

Calculate the maximum allowable working pressure of the thinned base metal ($E = 1.0$): $MAWP_{\text{LTA}} = \frac{S \cdot E \cdot t_{\text{avg}}}{R + 0.6 \cdot t_{\text{avg}}} = \frac{20,000 \cdot 1.0 \cdot 0.300}{24.0 + (0.6 \cdot 0.300)} = \frac{6,000}{24.18} = \mathbf{248.14\text{ psig}}$

---

Step 4: Evaluate Case B (LTA in Longitudinal Weld Seam)

Because the LTA encompasses the weld seam, we must use the vessel's actual weld joint efficiency $E = 0.85$:

$t_{\text{min}} = \frac{P \cdot R}{S \cdot E - 0.6 \cdot P} = \frac{250 \cdot 24.0}{(20,000 \cdot 0.85) - (0.6 \cdot 250)} = \frac{6,000}{17,000 - 150} = \frac{6,000}{16,850} = \mathbf{0.3561\text{ in.}}$

Comparison for Case B:

• Measured $t_{\text{avg}} = 0.300\text{ in.}$
• Minimum required weld thickness $t_{\text{min}} = 0.3561\text{ in.}$
• The deficit is significantly larger ($0.300\text{ in.}$ vs $0.3561\text{ in.}$) because the joint efficiency penalty ($E = 0.85$) increases the required thickness by $17.8%$.

Calculate the MAWP for the thinned weld ($E = 0.85$): $MAWP_{\text{weld}} = \frac{S \cdot E \cdot t_{\text{avg}}}{R + 0.6 \cdot t_{\text{avg}}} = \frac{20,000 \cdot 0.85 \cdot 0.300}{24.0 + (0.6 \cdot 0.300)} = \frac{5,100}{24.18} = \mathbf{210.92\text{ psig}}$

This calculation clearly highlights why inspectors must verify the proximity of an LTA to any weld seam.

---

6. Common Exam Pitfalls & Calculation Traps