CO₂ Pipeline & Transport · Fundamentals

CO₂ Pipeline Materials & Fracture Arrest

Engineering reference for ASME B31.4 wall-thickness design and running-ductile-fracture arrest in dense-phase CO₂ pipelines. Covers API 5L grade selection, the Battelle/Maxey-Kiefner CVN method, the Mannucci grade correction, and the DNV-RP-J202 CO₂-specific multiplier that addresses the long decompression plateau unique to CO₂ service.

Design factor

F = 0.72

ASME B31.4 universal design factor for CO₂ in liquid/dense-phase service. No location-class derating, unlike B31.8 for natural gas.

CO₂ multiplier

×2.5 vs NG

DNV-RP-J202 / Cosham 2012 recommended factor on Battelle/Maxey-Kiefner CVN for dense-phase CO₂ — addresses the long Psat decompression plateau.

Mill tolerance

12.5% PSL2

API 5L PSL2 default. t_nominal = (t_design + CA) / (1 − 0.125) so worst-case mill thickness still meets design.

Run the calculation

Wall thickness + CVN

Compute design wall thickness, hoop stress, mass per metre, and required fracture-arrest CVN with full grade and CO₂ corrections.

A2: CO₂ Pipeline Wall Thickness →

1. Overview

CO₂ pipeline mechanical design has two coupled requirements:

  1. Static pressure containment: wall thickness sufficient to contain the design pressure with the appropriate code design factor and corrosion allowance — standard pressure-vessel logic.
  2. Running-ductile-fracture arrest: material toughness sufficient to stop a crack that initiates and starts to propagate axially along the pipe. Without arrest, a single defect can rupture kilometres of pipe in seconds.

The fracture-arrest requirement is what makes CO₂ pipeline materials engineering different from oil pipeline engineering. Dense-phase CO₂ has a thermodynamic peculiarity — the long pressure plateau during decompression — that demands much higher steel toughness than natural gas service. Standards governing this work:

StandardScope
ASME B31.4-2022Pipeline Transportation Systems for Liquids and Slurries — wall thickness formula and design factors for CO₂ service
API 5L PSL2 (46th Ed.)Line pipe specification — SMYS values, mill tolerances, supplementary CVN requirements
DNV-RP-J202 (2017)Design and Operation of CO₂ Pipelines — CO₂-specific CVN correction, materials envelope, qualification testing
BS 7910 / API 579Fitness-for-Service — fracture mechanics analysis for assumed flaws and inspection planning
NACE/AMPP MR0175Sour-service hardness limits (informational for pure CO₂; mandatory if H₂S present)
Industry experience: US CO₂ pipelines have operated for 50+ years with X65 line pipe in mostly trouble-free service (Cortez since 1984, Sheep Mountain since 1972). The fracture-arrest design margin in those vintage pipelines was confirmed by full-scale burst tests in the 1970s — the empirical foundation of the modern Battelle method.

2. ASME B31.4 Wall Thickness

The ASME B31.4-2019 §403.2.1 wall-thickness formula for liquid and dense-phase CO₂ service is:

t = P · D / (2 · S · E · F) t = required wall thickness (mm) P = internal design gauge pressure (MPa) D = pipe outside diameter (mm) S = SMYS (MPa) — from API 5L PSL2 specification E = longitudinal joint factor (dimensionless) F = design factor (dimensionless) — 0.72 for B31.4

Joint factor E

The longitudinal joint factor accounts for weld quality:

Pipe typeE
Seamless (API 5L)1.00
SAW (submerged-arc welded), API 5L PSL21.00
ERW (electric-resistance welded), API 5L PSL21.00
Furnace lap weld (legacy, pre-1970)0.60
Continuous weld (legacy)0.60

For modern dense-phase CO₂ pipelines, E = 1.00 is universal — both seamless and PSL2 SAW/ERW pipe meet this.

From tdesign to tnominal mill spec

The B31.4 formula gives the structurally required thickness. To obtain the orderable nominal mill spec, add corrosion allowance and divide by (1 − mill tolerance):

tnominal = (tdesign + CA) / (1 − tol) CA = corrosion allowance (mm) — typical 1.5 mm for dry pure CO₂ up to 6 mm for wet/sour service tol = mill tolerance (decimal) — 0.125 for API 5L PSL2 default 0.08 for tighter PSL2 supplementary

The selected pipe schedule must have wall thickness ≥ tnominal.

Wet CO₂ corrosion: Dry CO₂ (water content < 50 ppmv) is non-corrosive to carbon steel. Wet CO₂ produces carbonic acid and corrodes carbon steel at 1–10 mm/yr — completely unacceptable. CCUS pipelines must specify a maximum water content (typically 50 ppmv) to ensure dry service, or use 13Cr/duplex/CRA materials.

Hoop stress check

The Lamé thin-wall hoop stress at design pressure verifies the calculation:

σh = P · (D − t) / (2 · t) σh ≤ F · S · E · joint_factor (always satisfied by B31.4 formula)

For a properly designed pipe, σh ≈ F·S = 72% of SMYS. The actual % SMYS is a useful check on the calculation.

3. API 5L Grade Selection

API 5L PSL2 line pipe specifies SMYS (Specified Minimum Yield Strength) by grade designation:

GradeSMYS (MPa)UTS (MPa)SMYS (ksi)Typical CO₂ pipeline use
X4229041542.0Gathering, low-pressure trunks
X5235945552.0Distribution, medium-pressure trunks
X6041452060.0Trunk lines, moderate ID
X6544853565.0Most common — large trunks (Cortez, Sheep Mountain, Quest)
X7048357070.0Long high-pressure trunks where WT savings warrant high grade
X8055262580.0Premium high-pressure trunks; verify CVN feasibility
X90, X100625, 690695, 76090, 100Atypical for CO₂ — flag CVN and HE concerns

Pipe weight per metre

For a given outside diameter and wall thickness:

m = π · (D² − ID²) / 4 · ρsteel [kg/m] ID = D − 2·t (mm) ρsteel = 7,850 kg/m³ (carbon steel)

Pipe mass scales nearly linearly with WT for thin-wall pipe (D/t > 30) and quadratically for thick-wall — important for capital cost and transportation logistics.

D/t ratio

The diameter-to-thickness ratio is a useful classification:

D/t rangeClassificationTypical CO₂ application
≤ 50Thick-wallVery high-pressure trunks, high-grade pipe
50–100Standard line pipeMost CO₂ trunks operate here
100–140Thin-wallVerify buckling resistance; less common
> 140AtypicalRe-evaluate — may not be physically reasonable

4. Running-Ductile-Fracture Arrest

A running ductile fracture is an axial crack that propagates along a pressurized pipe at speeds of 100–300 m/s, releasing the stored elastic and pneumatic energy. Without arrest, a single defect can rupture kilometres of pipe in seconds — this is the worst-case CO₂ pipeline failure mode.

Arrest occurs when the crack speed drops below the decompression-wave speed at every pressure during the depressurization. The Battelle Two-Curve Method (BTCM, Maxey 1974) compares two curves on a velocity-pressure plane:

  • Crack velocity curve (steel side): increases with pressure; depends on material toughness, WT, and yield strength
  • Decompression curve (fluid side): outflow of pressurized fluid — pressure drops at the wave speed of the fluid

Arrest is achieved when the crack curve sits below the decompression curve everywhere — i.e., the crack cannot keep up with the decompression wave.

Battelle / Maxey-Kiefner short form for arrest CVN

The full BTCM requires explicit decompression-curve simulation. For screening, the Maxey 1974 short form gives the required CVN:

CVNarrest [J] = 1.26 × 10⁻⁴ · σh² · (R · t)1/3 σh = hoop stress at design pressure (MPa) R = pipe radius (mm) t = wall thickness (mm) Result is for natural-gas service only — see CO₂ multiplier in Section 5

This formula was empirically calibrated by Maxey 1974 from full-scale burst tests on natural gas pipelines in the 1970s. It captures the dominant scaling: CVN ∝ σh² (energy released grows as stress squared) and CVN ∝ (Rt)1/3 (geometry term).

Mannucci grade correction

Modern high-grade line pipe (X70+) has cleaner microstructure, finer grain size, and different fracture mechanics than the 1970s pipe used to calibrate Maxey's formula. Mannucci developed a multiplier:

Grade rangeMannucci correctionRationale
X42, X52, X60, X651.0Pipe similar to Maxey's calibration data
X70, X801.7Higher-strength pipe with cleaner inclusions; original BTCM under-predicts
X90, X100, X1202.5Very high-strength pipe with substantially different microstructure

The corrected CVN requirement is multiplied by Cgrade from this table — and for CO₂ service, by an additional CO₂ multiplier described in the next section.

5. The CO₂ Multiplier

For natural gas at typical pipeline conditions, decompression follows a smooth pressure decay set by the gas's adiabatic expansion. For dense-phase CO₂, decompression follows a very different path:

  1. Initial drop from operating pressure (~150 bara) until the saturation curve is reached at the local temperature.
  2. Plateau at the saturation pressure as the fluid undergoes phase change — pressure drops slowly while large amounts of CO₂ flash from liquid to vapor.
  3. Final decay after the fluid has fully vaporized.

The plateau is what makes CO₂ different. For 150 bara starting pressure and 35 °C, the plateau sits near 73.77 bar (the critical pressure) and persists over a long propagation distance. During this plateau, the crack-driving stress on the steel remains high while the steel's resistance is being tested.

The Cosham 2012 finding: Multiple full-scale CO₂ burst tests by Cosham (DNV) and others showed that BTCM-predicted CVN under-estimates the actual arrest CVN for dense-phase CO₂ by approximately 2–3×. The DNV-RP-J202 recommended multiplier is 2.5 — consistent with the Cosham field tests and conservative for design.

Final CVN requirement for dense-phase CO₂

CVNarrest, CO₂ [J] = CVNBTCM · Cgrade · Cfluid CVNBTCM = 1.26e−4 · σh² · (R·t)1/3 Cgrade = 1.0 (X42–X65) / 1.7 (X70–X80) / 2.5 (X90+) Cfluid = 2.5 for dense-phase CO₂ (DNV-RP-J202) Cfluid = 1.0 for natural gas (calibration condition)

When CVN exceeds mill capability

API 5L PSL2 supplementary CVN requirements typically allow up to ~250 J specification. For demanding CO₂ service, the calculated arrest CVN can exceed 400 J — beyond what's achievable in normal pipe heat treatment.

Three engineering responses:

  1. Thicker wall: Increasing t reduces σh (CVN ∝ σh²) and the required CVN drops faster than the WT increase. Diminishing returns above 1.5× design WT.
  2. Lower-grade pipe: X65 vs X80 reduces σh at design pressure, lowering the required CVN. Trade-off: thicker wall (more steel mass).
  3. Mechanical crack arrestors: Clamped welded sleeves around the pipe at intervals of 100–500 m physically block crack propagation. Standard practice for high-pressure CO₂ trunks; some designs include them every 250 m as best practice.

6. Worked Example

Problem: A 12" OD CO₂ trunk pipeline operates at 200 bara design pressure, 35 °C. Determine wall thickness, hoop stress, and required arrest CVN. Use API 5L X65 PSL2 pipe with E = 1.0, F = 0.72, CA = 1.5 mm, mill tolerance 12.5%.

Step 1: Convert design pressure to MPa.

P = (200 − 1.013) × 0.1 = 19.90 MPa (gauge)

Step 2: ASME B31.4 design wall thickness.

S = 448 MPa (X65 SMYS) tdesign = P·D / (2·S·E·F) = 19.90 × 323.85 / (2 × 448 × 1.0 × 0.72) = 6443.6 / 645.12 = 9.99 mm

Step 3: Add corrosion allowance + mill tolerance.

tnominal = (9.99 + 1.5) / (1 − 0.125) = 11.49 / 0.875 = 13.13 mm ID = 323.85 − 2 × 13.13 = 297.59 mm D/t = 323.85 / 13.13 = 24.7 ✓ (typical line pipe)

Step 4: Hoop stress at design pressure.

σh = P · (D − tdesign) / (2 · tdesign) = 19.90 × (323.85 − 9.99) / (2 × 9.99) = 6248 / 19.98 = 312.6 MPa % SMYS = 312.6 / 448 = 69.8% ✓ (≤ 72%)

Step 5: Pipe weight per metre.

Asteel = π × ((0.32385)² − (0.29759)²) / 4 = π × (0.10488 − 0.08856) / 4 = π × 0.01633 / 4 = 0.01282 m² m = 0.01282 × 7850 = 100.6 kg/m

Step 6: Battelle/Maxey-Kiefner CVN (NG baseline).

R = 323.85 / 2 = 161.93 mm CVNBTCM = 1.26e−4 · (312.6)² · (161.93 × 9.99)1/3 = 1.26e−4 · 97,719 · (1617.7)1/3 = 1.26e−4 · 97,719 · 11.74 = 144.5 J

Step 7: Apply Mannucci and CO₂ corrections.

Cgrade = 1.0 (X65 in Maxey calibration range) Cfluid = 2.5 (dense-phase CO₂, DNV-RP-J202) CVNarrest, CO₂ = 144.5 × 1.0 × 2.5 = 361 J (= 266 ft·lb)

Step 8: Sanity-check vs mill capability. 361 J exceeds typical PSL2 supplementary CVN of 250 J. Three engineering options: increase WT, reduce grade, or specify mechanical crack arrestors.

Recommended specification: API 5L X65 PSL2, OD 323.9 mm × WT 13.2 mm, full-body UT, supplementary CVN with mechanical crack arrestors at 250 m spacing in HCAs. Final spec must be qualified by full BTCM analysis with measured decompression curve.

7. Standards & References

  • ASME B31.4-2022, Pipeline Transportation Systems for Liquids and Slurries (CO₂ service)
  • API Specification 5L, 46th Edition (2018), Line Pipe, Product Specification Level PSL2
  • ISO 27913:2016, Carbon dioxide capture, transportation and geological storage — Pipeline transportation systems
  • DNV-RP-J202 (2017), Design and Operation of CO₂ Pipelines
  • Maxey, W.A. (1974). "Fracture Initiation, Propagation and Arrest," 5th Symposium on Line Pipe Research, AGA Cat. No. L30174.
  • Cosham, A., Eiber, R.J., Clouston, E.B. (2012). "The Decompression Behaviour of Carbon Dioxide in the Dense Phase," IPC2012-90461.
  • Mannucci, G., Demofonti, G. (2002). "Crack Arrestability of High-Strength Steel Line Pipe," 13th Biennial Joint Tech. Mtg.
  • BS 7910:2019, Guide to methods for assessing the acceptability of flaws in metallic structures
  • API 579-1/ASME FFS-1 (2021), Fitness-For-Service
  • NACE/AMPP MR0175/ISO 15156 (2020), Materials for use in H₂S-containing environments in oil and gas production
  • 49 CFR Part 195, Transportation of Hazardous Liquids by Pipeline (US PHMSA — applies to CO₂)

Frequently Asked Questions

Why does CO₂ pipeline fracture arrest require higher CVN than natural gas?

Dense-phase CO₂ has a long pressure plateau during decompression — as pressure drops it crosses the saturation line and stays nearly constant while the fluid changes phase. This plateau matches the steel's running-fracture driving force over a long propagation distance, requiring 2–3× the CVN of natural gas service. DNV-RP-J202 / Cosham 2012 recommend a 2.5× multiplier on the standard Battelle/Maxey-Kiefner result for dense-phase CO₂.

What is the ASME B31.4 design factor for CO₂ pipelines?

ASME B31.4 uses F = 0.72 universally for liquid (and dense-phase CO₂) service. Unlike B31.8 for natural gas, there is no location-class derating in B31.4 — the same factor applies in rural, suburban, and urban environments. This reflects the lower kinetic energy released in a CO₂ liquid pipeline failure compared to a high-pressure gas pipeline.

What pipe grade is recommended for CO₂ pipelines?

API 5L X65 PSL2 is the most common grade for dense-phase CO₂ trunk lines (Cortez, Sheep Mountain, Quest, Snøhvit). It balances yield strength (448 MPa) with weldability and toughness margin. X70 and X80 are used where wall-thickness reduction outweighs the higher toughness requirements; lower grades (X42–X60) are common for shorter lines and gathering networks. DNV-RP-J202 places X42-X80 in its preferred envelope.

How is mill tolerance applied to CO₂ pipeline wall thickness?

API 5L PSL2 allows up to 12.5% under-tolerance on wall thickness from manufacturing variation. The nominal pipe spec is computed as t_nominal = (t_design + corrosion_allowance) / (1 − 0.125) so that even a worst-case mill-tolerance reduction still meets the design thickness requirement. Stricter mill specs (e.g., 8% PSL2 alternative) are sometimes purchased for high-grade or thick-wall pipe.

When are CO₂ pipeline crack arrestors needed?

Mechanical crack arrestors (clamped sleeves welded around the pipe at intervals of 100–500 m) are needed when the calculated arrest CVN exceeds practical mill capability — typically > 250 J (185 ft·lb) for line pipe. For high-grade pipe at high operating pressure, the CVN required by the Battelle two-curve method × DNV CO₂ multiplier can exceed 400 J, which is unachievable in normal heat treatment. Arrestors solve this by mechanically blocking propagation between intact pipe segments.