1. Overview & Design Criteria
Pipeline line sizing balances competing objectives: minimize capital cost (smaller pipe is cheaper) against minimize operating cost (larger pipe reduces pressure drop and compression/pumping energy). The optimal diameter depends on velocity constraints, pressure drop limits, and economic analysis.
Velocity constraints
Maximum & Minimum
Max: erosional velocity, noise. Min: solids settling, liquid holdup, gas sweep.
Pressure drop
Allowable ΔP
Available pressure budget; affects compression power and booster station spacing.
Economic diameter
NPV optimization
Minimize total cost (CAPEX + OPEX) over project life; energy cost vs pipe cost.
Code compliance
ASME B31.8, B31.4
Design pressure, wall thickness, material selection, location class factors.
Design Criteria Summary
| Criterion | Concern | Typical Limit |
|---|---|---|
| Maximum velocity | Erosion, noise, vibration, pressure drop | Gas: 60–100 ft/s (plant); 20–50 ft/s (pipeline) Liquid: 6–10 ft/s (discharge); 3–5 ft/s (suction) |
| Minimum velocity | Liquid holdup, solids deposition, gas pocketing | Gas: 10 ft/s (to sweep liquids) Liquid: 3–5 ft/s (to suspend solids) |
| Erosional velocity | Pipe wall erosion (especially with sand/solids) | Ve = C/√ρ (API RP 14E) C = 100–150 depending on service |
| Pressure drop | Available pressure, compression/pump power | Gathering: 5–20 psi/mi Transmission: 1–5 psi/mi Plant piping: 0.5–2 psi/100 ft |
| Reynolds number | Flow regime (laminar vs turbulent) | Re > 4000 for fully turbulent (typical pipeline) |
| Noise | Personnel safety, equipment damage | < 85 dBA at 1 meter distance |
| Mach number (gas) | Compressibility effects, choking | < 0.3 for incompressible flow assumption < 0.5 for flare headers (avoid noise/backpressure) |
Basic Velocity Equations
Gas Velocity (Volumetric Flow):
V = 0.002122 × (Qacfh) / D²
V = 0.0509 × (Qmmscfd × Z × T) / (P × D²)
Where:
V = Gas velocity (ft/s)
Qacfh = Actual volumetric flow rate (acfh)
Qmmscfd = Standard volumetric flow rate (MMscfd, 14.7 psia, 60°F)
P = Pressure (psia)
T = Temperature (°R)
Z = Compressibility factor
D = Pipe inside diameter (inches)
Liquid Velocity:
V = 0.4085 × Q / D²
Where:
V = Liquid velocity (ft/s)
Q = Flow rate (GPM)
D = Pipe inside diameter (inches)
Mass Velocity (Gas or Liquid):
V = (ṁ / ρ) / A = ṁ / (ρ × π × D²/4)
Where:
ṁ = Mass flow rate (lb/s)
ρ = Density (lb/ft³)
A = Cross-sectional area (ft²)
D = Inside diameter (ft)
Design philosophy: Line sizing is iterative. Start with velocity criteria to get initial diameter, check pressure drop, evaluate economics, verify wall thickness meets code requirements, then iterate if necessary. Long pipelines are typically sized by allowable pressure drop; short plant piping by velocity limits.
Recommended Gas Velocities
| Service | Velocity (ft/s) | Basis / Notes |
|---|---|---|
| Compressor suction | 30–60 | Low ΔP critical to minimize compression power |
| Compressor discharge | 50–80 | Higher density allows higher velocity |
| Process plant headers | 60–100 | Short runs, moderate ΔP acceptable |
| Transmission pipelines | 20–50 | Long distance, optimize ΔP vs diameter |
| Gathering lines | 10–30 | Low pressure, must sweep liquids |
| Flare headers | Ma < 0.5 | Mach number limit; noise and backpressure concerns |
| Relief valve inlet | ΔP < 3% Pset | Per API 520; pressure drop limit, not velocity |
| Custody transfer | Optimize for meter | Ultrasonic: 10–40 ft/s; Orifice: Re-dependent |
Recommended Liquid Velocities
| Service | Velocity (ft/s) | Basis / Notes |
|---|---|---|
| Pump suction | 3–5 | NPSH critical; minimize friction losses |
| Pump discharge | 6–10 | Higher velocity acceptable with available head |
| Gravity flow | 1–3 | Limited by available static head |
| Process lines (general) | 5–8 | Balance cost vs ΔP; typical plant piping |
| Crude oil pipelines | 3–7 | Economic optimization; pump station spacing |
| Refined products | 4–8 | Lower viscosity than crude |
| Water service (cooling, utility) | 5–10 | Economic velocity; erosion not typically limiting |
| Slurries, viscous fluids | > 5 | Keep solids suspended; prevent settling |
| Boiler feedwater | 8–15 | High velocity to minimize pipe size, erosion acceptable |
Erosional Velocity (API RP 14E)
API RP 14E provides an empirical equation for maximum gas velocity to prevent erosion of pipe walls, especially critical when entrained liquids or solids are present:
API RP 14E Erosional Velocity:
Ve = C / √ρ
Where:
Ve = Erosional velocity (ft/s)
C = Empirical constant (dimensionless)
ρ = Gas density at flowing conditions (lb/ft³)
C-factor guidelines:
C = 100: Continuous service, clean dry gas (API RP 14E baseline)
C = 125: Intermittent service, clean gas with some liquids
C = 75: Corrosive service (H₂S, CO₂) - industry practice*
C = 60: Service with solids/sand production - industry practice*
*API RP 14E recommends reducing C below 100 for corrosive or erosive
conditions but does not prescribe specific values. Values shown are
commonly used in industry. Some operators use C = 100 for all cases.
Example:
Natural gas at 800 psia, 80°F → ρ ≈ 3.5 lb/ft³ (with Z ≈ 0.88)
Ve = 125 / √3.5 = 125 / 1.87 = 66.8 ft/s
Two-phase flow (gas-liquid mixture):
Use mixture density: ρm = (WL + WG) / (WL/ρL + WG/ρG)
⚠ Erosional velocity is not a hard limit: API RP 14E states that exceeding erosional velocity does not automatically cause erosion—it's a guideline for preliminary design. Actual erosion depends on fluid composition, metallurgy, operating history, and inspection results. Use C = 100 for conservative design with sand production or corrosive service.
2. Flow Equations
Flow equations relate flow rate, pressure drop, pipe diameter, and fluid properties. For gas pipelines, several empirical equations have been developed based on field data. For liquid lines, Darcy-Weisbach or Hazen-Williams equations are used.
General Flow Equation (Gas Pipelines)
General Gas Flow Equation (AGA):
Q = C × √[(P₁² - P₂²) / (G × Tavg × Le × Zavg)] × Dn
Where:
Q = Flow rate (various units, equation-dependent)
P₁, P₂ = Upstream, downstream pressure (psia)
D = Pipe inside diameter (inches)
G = Gas specific gravity (air = 1.0)
Tavg = Average temperature (°R)
Le = Equivalent length including fittings (miles or feet)
Zavg = Average compressibility factor
C, n = Constants depending on equation (Weymouth, Panhandle A/B, etc.)
Exponent n varies:
Weymouth: n = 2.667
Panhandle A: n = 2.6182
Panhandle B: n = 2.530
Colebrook-White: n = 2.5 (fully turbulent)
Weymouth Equation
Developed for large-diameter, high-pressure pipelines. Based on smooth pipe flow data. Suitable for fully turbulent flow (Re > 10⁷).
Weymouth Equation (US units):
Q = 433.5 × E × (Tb/Pb) × [(P₁² - P₂²) × D16/3 / (G × Tavg × Le × Zavg)]1/2
Where:
Q = Flow rate (scfd at base conditions)
E = Pipeline efficiency factor (0.92–0.95 typical)
Tb = Base temperature (°R, typically 520°R = 60°F)
Pb = Base pressure (psia, typically 14.73 or 14.7)
P₁, P₂ = Inlet, outlet pressure (psia)
D = Inside diameter (inches)
G = Gas specific gravity
Tavg = Average flowing temperature (°R)
Le = Equivalent length (miles)
Zavg = Average compressibility factor
Applications:
- Transmission pipelines, D > 12 inches
- High Reynolds number (fully turbulent)
- Smooth pipe conditions
- Conservative (predicts lower flow than Panhandle for same ΔP)
Limitations:
- Less accurate for smaller diameter pipe (< 6 inches)
- Does not explicitly account for friction factor variation
Panhandle A Equation
Developed from data on natural gas pipelines 6–24 inches in diameter. Accounts for partial turbulence. More optimistic than Weymouth.
Panhandle A Equation (US units):
Q = 435.87 × E × (Tb/Pb) × (P₁² - P₂²)0.5394 × D2.6182 / (G0.4604 × Tavg0.5394 × Le0.5394 × Zavg0.5394 × μ0.0788)
Where:
μ = Gas viscosity (centipoise)
Other variables same as Weymouth
Applications:
- Medium to large diameter pipe (6–24 inches)
- Partially turbulent flow (Re = 5×10⁶ to 10⁷)
- Gathering and transmission systems
Typical results:
Panhandle A predicts 5–10% higher flow than Weymouth for same conditions
Panhandle B Equation
Modified Panhandle equation for fully turbulent flow in newer, smoother pipelines. Higher efficiency than Panhandle A.
Panhandle B Equation (US units):
Q = 737 × E × (Tb/Pb) × (P₁² - P₂²)0.51 × D2.53 / (G0.49 × Tavg0.51 × Le0.51 × Zavg0.51)
Applications:
- High-pressure transmission lines
- Smooth, well-maintained pipe
- Fully turbulent flow (Re > 10⁷)
Efficiency factor E:
E = 1.0 for ideal, new pipe
E = 0.95 typical for well-maintained pipeline
E = 0.92 for average condition
E = 0.85–0.90 for older pipe with deposits
Colebrook-White / AGA Equation
Theoretically rigorous equation based on Darcy-Weisbach with Colebrook-White friction factor. Iterative solution required.
AGA / Colebrook-White Equation:
ΔP = f × (L / D) × (ρ × V² / 2) (Darcy-Weisbach)
Friction factor f from Colebrook-White:
1/√f = -2 × log₁₀[(ε/(3.7×D)) + (2.51/(Re×√f))]
Where:
f = Darcy friction factor
ε = Absolute pipe roughness (feet)
D = Inside diameter (feet)
Re = Reynolds number = ρ V D / μ
Iterative procedure:
1. Assume f (start with f = 0.02)
2. Calculate Re = ρ V D / μ
3. Calculate f from Colebrook-White
4. Repeat steps 2–3 until f converges (typically 3–5 iterations)
Advantages:
- Theoretically sound
- Valid for all flow regimes (laminar, transition, turbulent)
- Accounts for pipe roughness explicitly
Pipe roughness (ε) typical values:
New steel pipe: 0.0018 in (0.00015 ft)
Commercial steel: 0.002 in (0.000167 ft)
Slightly corroded: 0.006 in (0.0005 ft)
Heavily corroded: 0.04 in (0.0033 ft)
Comparison of Flow Equations
| Equation | Best Application | Re Range | Relative Flow Prediction |
|---|---|---|---|
| Weymouth | Large transmission (D > 12") | Re > 10⁷ | Most conservative (lowest Q) |
| Panhandle A | Medium pipe (6–24"), gathering | 5×10⁶ to 10⁷ | Moderate (5–10% > Weymouth) |
| Panhandle B | High-pressure, smooth pipe | Re > 10⁷ | Optimistic (10–15% > Weymouth) |
| Colebrook-White (AGA) | General purpose, any diameter | All regimes | Depends on roughness (varies ±10%) |
Liquid Flow (Darcy-Weisbach)
Darcy-Weisbach Equation (Liquids):
ΔP = f × (L / D) × (ρ × V² / 2)
Or in head loss form:
hL = f × (L / D) × (V² / 2g)
Where:
ΔP = Pressure drop (lbf/ft²; divide by 144 for psi)
hL = Head loss (ft)
f = Darcy friction factor (from Moody diagram or Colebrook-White)
L = Pipe length (ft)
D = Inside diameter (ft)
ρ = Liquid density (lb/ft³)
V = Velocity (ft/s)
g = Gravitational constant (32.2 ft/s²)
Hazen-Williams Equation (Alternative for Water):
Q = 0.442 × CHW × D2.63 × S0.54
Where:
Q = Flow rate (GPM)
CHW = Hazen-Williams coefficient (100–150; 140 for PVC, 120 for steel)
D = Inside diameter (inches)
S = Hydraulic slope (ft/ft) = hL / L
Note: Hazen-Williams is empirical and only valid for water. Use Darcy-Weisbach for hydrocarbons and other liquids.
Equation selection: For gas pipelines, Weymouth is conservative and widely used for transmission lines; Panhandle B is common for high-pressure, well-maintained systems. AGA/Colebrook-White is rigorous but requires iterative solution. For liquids, always use Darcy-Weisbach with appropriate friction factor correlation.
3. Pressure Drop Calculations
Pressure drop in pipelines consists of friction losses (dominant in long horizontal lines), elevation changes (hydrostatic head), and acceleration losses (negligible except in high-velocity gas lines or two-phase flow).
Total Pressure Drop Components
Total Pressure Drop:
ΔPtotal = ΔPfriction + ΔPelevation + ΔPacceleration
Friction losses (gas):
From flow equations (Weymouth, Panhandle, Darcy-Weisbach)
Elevation change (gas):
ΔPelev = ρavg × g × Δh / 144
ΔPelev ≈ 0.0375 × (Pavg / Tavg) × G × Δh / Zavg
Where:
Δh = Elevation change (ft, positive = uphill)
Pavg = Average pressure (psia)
Tavg = Average temperature (°R)
Elevation change (liquid):
ΔPelev = ρ × g × Δh / 144 = 0.433 × SG × Δh (psi)
Where:
SG = Specific gravity (water = 1.0)
Δh = Elevation change (ft)
Acceleration (typically negligible):
Important only when velocity changes significantly (high ΔP/P or two-phase)
Friction Factor Determination
| Flow Regime | Re Range | Friction Factor Correlation |
|---|---|---|
| Laminar | Re < 2100 | f = 64 / Re |
| Transition | 2100 < Re < 4000 | Unstable; avoid if possible; use Colebrook-White |
| Turbulent (smooth) | Re > 4000, ε/D → 0 | 1/√f = 2.0 × log₁₀(Re×√f) - 0.8 (Prandtl) |
| Turbulent (rough) | Re > 4000 | 1/√f = -2 × log₁₀[(ε/(3.7D)) + (2.51/(Re√f))] (Colebrook-White) |
| Fully turbulent (rough) | Re > 10⁶, rough pipe | 1/√f = -2 × log₁₀[ε/(3.7D)] (independent of Re) |
Reynolds Number
Reynolds Number:
Re = ρ V D / μ
Gas (using mass velocity):
Re = 48 × ṁ / (D × μ)
Re = 0.0004778 × (Qscfd × G) / (D × μ)
Where:
ṁ = Mass flow rate (lb/s)
D = Inside diameter (inches)
μ = Viscosity (centipoise, cP)
Qscfd = Standard flow rate (scfd)
G = Specific gravity
Liquid:
Re = 3160 × Q × SG / (D × μ)
Where:
Q = Flow rate (GPM)
D = Inside diameter (inches)
SG = Specific gravity
μ = Viscosity (cP)
Typical Re values:
Gathering lines: 10⁵ to 10⁶
Transmission pipelines: 10⁶ to 10⁸
Plant piping: 10⁴ to 10⁶
Temperature Effects on Gas Flow
Gas temperature affects density, viscosity, and compressibility. In long pipelines, gas temperature varies due to Joule-Thomson cooling and heat transfer with surroundings.
Joule-Thomson Cooling:
ΔTJT = μJT × ΔP
Where:
μJT = Joule-Thomson coefficient (°F/psi)
ΔP = Pressure drop (psi)
For natural gas:
μJT ≈ 7 to 10°F per 100 psi (at typical pipeline conditions)
Example:
Pressure drop of 500 psi → Temperature drop of 35–50°F
Heat transfer with surroundings:
Buried pipeline: T approaches ground temperature over distance
Thermal diffusion length: Lth ≈ 10–50 miles
Above ground: T approaches ambient air temperature
Faster equilibration (~5–20 miles)
Design approach:
- Short lines (< 5 miles): assume isothermal
- Long lines: use average temperature or segment pipeline for detailed analysis
- Include Joule-Thomson effect in high-ΔP systems
Equivalent Length for Fittings
Fittings, valves, and bends cause additional pressure drop. Express as equivalent length of straight pipe.
| Component | Equivalent Length (Le/D) | K Factor (Alternative) |
|---|---|---|
| 90° standard elbow | 30 | K = 0.9 |
| 45° elbow | 16 | K = 0.4 |
| Long-radius 90° elbow | 20 | K = 0.6 |
| Gate valve (fully open) | 8 | K = 0.2 |
| Ball valve (fully open) | 3 | K = 0.05 |
| Globe valve (fully open) | 340 | K = 10 |
| Swing check valve | 100 | K = 2.5 |
| Tee (flow through run) | 20 | K = 0.6 |
| Tee (flow through branch) | 60 | K = 1.8 |
| Pipe entrance (sharp) | — | K = 0.5 |
| Pipe entrance (rounded) | — | K = 0.05 |
| Pipe exit | — | K = 1.0 |
Pressure drop from K factors:
ΔPfitting = K × (ρ × V² / 2) / 144
Where:
K = Loss coefficient (from table)
ρ = Density (lb/ft³)
V = Velocity (ft/s)
Result in psi
Equivalent length approach:
Le,total = Lpipe + Σ(Le/D)i × D
Where:
Lpipe = Actual pipe length (ft)
(Le/D)i = Equivalent length ratio for fitting i
D = Pipe inside diameter (ft)
Then use Le,total in flow equation instead of Lpipe
Fittings significance: In long pipelines (> 10 miles), fittings contribute < 1% to total pressure drop—can be neglected. In short plant piping (< 100 ft with many fittings), fittings can contribute 20–50% of total ΔP—must be included.
4. Economic Optimization
The optimal pipe diameter minimizes total cost over the project lifetime. Total cost = capital cost (CAPEX) + present value of operating cost (OPEX). Larger pipe costs more upfront but saves energy; smaller pipe is cheaper initially but requires more compression/pumping power.
Cost Components
| Component | Effect of Larger Diameter | Magnitude |
|---|---|---|
| Pipe material cost | Increases (more steel, higher weight) | Major (40–60% of CAPEX) |
| Installation cost (trenching, welding, coating) | Increases (wider trench, more weld passes) | Major (30–50% of CAPEX) |
| Valves, fittings, flanges | Increases (exponentially with size) | Moderate (5–15% of CAPEX) |
| Right-of-way (ROW) cost | Minimal increase (wider trench) | Minor (< 5% of CAPEX) |
| Compressor/pump capital cost | Decreases (lower ΔP → smaller equipment) | Major (20–40% of total CAPEX for long lines) |
| Compressor/pump operating cost (energy) | Decreases (lower ΔP → less power) | Major (dominant OPEX for long pipelines) |
| Maintenance cost | Slightly decreases (lower velocity, less wear) | Minor |
Economic Diameter Equation
Total Cost (NPV basis):
TC = CAPEXpipe + CAPEXcompression + PV(OPEXenergy)
Pipe CAPEX (approximate):
CAPEXpipe = Cpipe × L × D1.5 to 2.0
Where:
Cpipe = Unit cost ($/ft per inch-diameter, varies with location, terrain, market)
L = Length (ft or miles)
D = Diameter (inches)
Exponent typically 1.5–2.0 (larger diameter has lower $/ton but higher weight)
Compression CAPEX (approximate):
CAPEXcomp = Ccomp × HP
Where:
HP = Compression horsepower (decreases with larger D)
Ccomp = $/HP installed (~$1500–$3000/HP for gas turbine compressor)
Energy OPEX (present value):
PV(OPEXenergy) = (Annual energy cost) × [(1 - (1+i)-n) / i]
Annual energy cost = HP × 8760 hrs/yr × Utilization × (Fuel cost / BHP-hr)
Where:
i = Discount rate (e.g., 0.08 = 8%)
n = Project life (years, typically 20–30)
Utilization = Operating hours / 8760 (e.g., 0.95 for high utilization)
Optimal diameter:
d(TC)/d(D) = 0 (minimize total cost)
Typically solved iteratively or via optimization algorithm
Economic Velocity Rules of Thumb
Economic velocity is the velocity that minimizes total cost. Correlations have been developed based on typical costs:
Peters & Timmerhaus Economic Velocity:
Vopt = K × ρ-0.37 × D0.13
Where:
K = Constant depending on fluid and service
ρ = Density (lb/ft³)
D = Diameter (ft)
Typical results:
Gas (natural gas at pipeline conditions): Vopt ≈ 60–100 ft/s
Liquid (hydrocarbons): Vopt ≈ 5–8 ft/s
Water: Vopt ≈ 6–10 ft/s
Note: Economic velocity often coincides with erosional velocity or typical design velocity limits, which is why rule-of-thumb velocities work well for preliminary sizing.
Economic Analysis Example
Scenario: 100-mile natural gas pipeline, 500 MMscfd, P1 = 1200 psia, P2 = 800 psia, gas SG = 0.60, compare 20-inch vs 24-inch pipe.
| Item | 20-inch Pipeline | 24-inch Pipeline |
|---|---|---|
| Pipe cost (material + install) | $50M | $70M |
| Pressure drop (psi) | 600 | 350 |
| Compressor HP required | 18,000 HP | 10,500 HP |
| Compressor CAPEX (@$2000/HP) | $36M | $21M |
| Annual fuel cost (@$4/MMBTU, ~10,000 BTU/HP-hr) | $6.3M/yr | $3.7M/yr |
| PV of fuel (20 yrs, 8% discount) | $62M | $36M |
| Total NPV cost | $148M | $127M |
Result: 24-inch pipe is more economical despite $20M higher pipe cost, because it saves $26M in total compression cost (CAPEX + OPEX).
Factors Affecting Economic Diameter
- Energy cost: Higher fuel/electricity cost → larger optimal diameter
- Utilization factor: High utilization (near-continuous operation) → larger diameter justified
- Project lifetime: Longer project life → energy savings dominate → larger diameter
- Discount rate: High discount rate → less weight to future energy savings → smaller diameter
- Distance: Long pipelines → energy cost dominates → detailed economic analysis critical
- Short lines: Short plant piping → energy cost minor → size on velocity, upsize for future expansion
Rule of thumb: For short lines (< 1 mile), size on velocity and pressure drop; energy cost is negligible. For long pipelines (> 10 miles with high throughput), always perform economic optimization—energy cost dominates and optimal diameter can differ significantly from velocity-based sizing.
5. Practical Considerations
Standard Pipe Sizes
Pipe is manufactured in standard nominal sizes. Always round up to the next standard size after calculating required diameter.
| Nominal Size (inches) | Sch 40 ID (inches) | Sch 80 ID (inches) | STD ID (pipeline) |
|---|---|---|---|
| 2 | 2.067 | 1.939 | — |
| 3 | 3.068 | 2.900 | — |
| 4 | 4.026 | 3.826 | — |
| 6 | 6.065 | 5.761 | 6.065 |
| 8 | 7.981 | 7.625 | 7.981 |
| 10 | 10.020 | 9.562 | 10.136 (0.250" wall) |
| 12 | 11.938 | 11.374 | 11.750 (0.250" wall) |
| 16 | 15.000 | — | 15.250 (0.375" wall) |
| 20 | 18.812 | — | 19.250 (0.375" wall) |
| 24 | 22.624 | — | 23.250 (0.375" wall) |
| 30 | — | — | 29.250 (0.375" wall) |
| 36 | — | — | 35.250 (0.375" wall) |
Note: Process plant piping uses ASME B36.10 schedule sizes (Sch 40, 80, etc.). Transmission pipelines use ASME B36.10 for D ≤ 12 inches, then API 5L line pipe with wall thickness calculated per ASME B31.8 design formula.
Wall Thickness Selection (ASME B31.8)
ASME B31.8 Design Formula:
t = (P × D) / (2 × S × E × F × T)
Where:
t = Wall thickness (inches)
P = Design pressure (psig)
D = Outside diameter (inches)
S = SMYS (specified minimum yield strength, psi; e.g., 35,000 psi for Grade B; 52,000–70,000 for X52–X70)
E = Longitudinal joint factor (1.0 for seamless or ERW)
F = Design factor (location class factor, ASME B31.8 Table 841.1.6-1):
F = 0.72 for Class 1 (rural, ≤10 buildings per mile within 220 yd each side)
F = 0.60 for Class 2 (suburban, 11-46 buildings per mile)
F = 0.50 for Class 3 (urban, >46 buildings or 4+ story buildings)
F = 0.40 for Class 4 (high-consequence areas, multi-story buildings)
Note: F = 0.80 allowed for Class 1 Division 1 under limited conditions per B31.8 §841.1.6
T = Temperature derating factor (1.0 for T ≤ 250°F)
Example:
24-inch pipeline, MAOP = 1440 psig, Grade X70 (SMYS = 70,000 psi), Class 1 location
t = (1440 × 24) / (2 × 70,000 × 1.0 × 0.72 × 1.0)
t = 34,560 / 100,800 = 0.343 inches
Add corrosion allowance (0.0625" typical) and round up to available wall thickness:
trequired = 0.343 + 0.0625 = 0.406 inches → use 0.500" wall
Percent SMYS:
Hoop stress % SMYS = (P × D) / (2 × t × S) × 100%
Class 1 location: max 72% SMYS (Design Factor F = 0.72)
Process Piping Wall Thickness (ASME B31.3)
For process plant piping (not transmission pipelines), ASME B31.3 uses a different formula that accounts for pipe curvature effects through the Y coefficient:
ASME B31.3 Pressure Design Formula:
t = (P × D) / (2 × (S × E + P × Y))
Or rearranged for MAOP:
P = (2 × S × E × t) / (D - 2 × Y × t)
Where:
t = Minimum wall thickness (inches)
P = Design pressure (psig)
D = Outside diameter (inches)
S = Allowable stress (psi, from B31.3 Table A-1)
E = Quality factor for longitudinal weld joints (1.0 for seamless)
Y = Coefficient from B31.3 Table 304.1.1:
Y = 0.4 for ferritic steels at T < 900°F (typical process conditions)
Y = 0.5 for austenitic stainless steels
Y = 0.7 for ferritic steels at T ≥ 900°F
Manufacturing tolerance:
For most pipe: add 12.5% to calculated thickness (mill tolerance)
tnominal = tcalculated / 0.875
Example:
6-inch Sch 40 pipe (OD = 6.625", t = 0.280"), A106 Grade B (S = 20,000 psi), E = 1.0, Y = 0.4
P = (2 × 20,000 × 1.0 × 0.280) / (6.625 - 2 × 0.4 × 0.280)
P = 11,200 / 6.401 = 1,750 psig
Key difference from B31.8:
- B31.8: Uses design factors (F) based on location class, higher base stresses (SMYS)
- B31.3: Uses allowable stresses (lower than SMYS), Y coefficient for thick-wall correction
- B31.8 is for gas transmission; B31.3 is for process plant piping
Material Selection
| Material / Grade | SMYS (psi) | Application |
|---|---|---|
| API 5L Grade B | 35,000 | Low-pressure gathering, older pipelines |
| API 5L X42 | 42,000 | Gathering lines, moderate pressure |
| API 5L X52 | 52,000 | Transmission lines, common for 800–1200 psig |
| API 5L X60 | 60,000 | High-pressure transmission |
| API 5L X65 | 65,000 | High-pressure transmission, modern standard |
| API 5L X70 | 70,000 | Very high-pressure transmission (>1400 psig) |
| API 5L X80 | 80,000 | Ultra-high-pressure, specialty applications |
| ASTM A106 Grade B | 35,000 | Process plant piping, seamless |
| ASTM A53 Grade B | 35,000 | Process plant piping, ERW or seamless |
Future Capacity Planning
Pipelines are expensive to replace. Consider future flow growth when sizing:
- 20-year forecast: Review production forecasts, market growth, expansion plans
- Upsize strategy: If unsure, upsize by one standard diameter (e.g., 12" → 16" instead of 12" → 14")
- Looping option: For very long pipelines, consider initial smaller diameter with future looping (parallel pipe)
- Compression addition: Transmission lines can add booster compression stations to increase capacity 20–50%
- Cost tradeoff: Larger initial pipe costs ~20–40% more, but retrofit/replacement costs 3–5× more later
Two-Phase Flow Considerations
Gas-liquid two-phase flow is complex and requires specialized methods (Beggs-Brill, OLGA, etc.).
API RP 14E for Two-Phase Lines:
Vm = Vsg + Vsl
Where:
Vm = Mixture velocity (ft/s)
Vsg = Superficial gas velocity = Qg / A (ft/s)
Vsl = Superficial liquid velocity = Ql / A (ft/s)
Qg, Ql = Gas, liquid volumetric flow rates (ft³/s at P, T)
A = Pipe cross-sectional area (ft²)
Erosional velocity (mixture):
ρm = (WL + WG) / (WL/ρL + WG/ρG)
Ve = C / √ρm
Slug flow concerns:
- Terrain slugging in hilly pipelines (slugs accumulate at low points)
- Riser slugging at pipe outlets
- Requires slug catcher or inlet separator sized for liquid surge
Design approach:
1. Use two-phase flow correlation (Beggs-Brill, etc.) to calculate pressure drop
2. Check erosional velocity with mixture density
3. Verify flow regime (stratified, slug, annular) and design accordingly
4. Size downstream equipment (separators, slug catchers) for worst-case liquid holdup
⚠ Two-phase sizing complexity: Single-phase equations (Weymouth, Panhandle) do NOT apply to two-phase flow. Use multiphase correlations (Beggs-Brill, Duns-Ros, Hagedorn-Brown) or commercial software (PIPESIM, OLGA, HYSYS). Flow regime maps and holdup calculations are essential. Slug catchers can be very large (1000+ bbl) for long gathering lines.
References & Standards
- ASME B31.8 – Gas Transmission and Distribution Piping Systems
- ASME B31.4 – Pipeline Transportation Systems for Liquids and Slurries
- ASME B31.3 – Process Piping
- API RP 14E – Recommended Practice for Design and Installation of Offshore Production Platform Piping Systems
- API 5L – Specification for Line Pipe
- GPSA – Section 17 (Fluid Flow), Section 18 (Pipeline Design)
- AGA Reports – AGA Report No. 3 (Orifice Metering), AGA Report No. 8 (Compressibility)
- Crane TP-410 – Flow of Fluids Through Valves, Fittings, and Pipe
- Ludwig's Applied Process Design – Volume 1, Chapter 6 (Fluid Flow)
- Pipeline Design & Construction – McAllister (textbook)
Ready to use the calculator?
→ Launch Calculator