API RP 14E Erosional Velocity Calculator (Line Sizing)

Returns the erosional velocity limit V_e = C/√ρ_m from the API RP 14E ¶2.5 form, the actual line velocity V = 0.4085·Q/d² from your flow rate and bore, and the ratio between them with a PASS/FAIL verdict. The empirical constant C — which RP 14E sets by service, continuous or intermittent and solids-free or not — stays a user input.

Method last updated 2026-07-30 (calculation changelog) · fixture-verified on every build — most recently 2026-07-31.

Line sizing in oil and gas facilities has been argued about for decades, and one equation has done most of the arguing: V_e = C/√ρ. It appears in API RP 14E, it is two symbols long, it fits on a napkin, and it has sized a very large fraction of the world's production piping. It is also, by the admission of the recommended practice itself and of essentially everyone who has studied it since, an empirical rule of thumb rather than an erosion model.

What it actually does is set a velocity ceiling that falls as the fluid gets denser. The reasoning is that erosion damage relates to the kinetic energy the fluid carries into the wall at a change of direction, and kinetic energy goes as ρ·V²; holding ρ·V² constant gives V ∝ 1/√ρ, which is the equation. That is a defensible dimensional argument and a poor mechanistic one, because it contains nothing about the actual erosion drivers: no particle size or concentration, no material hardness, no bend radius, no impingement angle, no flow regime. All of that is compressed into the single constant C.

The consequence is worth stating plainly, because it cuts both ways. On clean, solids-free service the equation is widely regarded as over-conservative — it has driven a great deal of oversized, more expensive piping, and it has also caused real problems of its own, because a line sized too large runs too slow and starts accumulating liquids, slugging, and corroding under the deposits. On sand-bearing service it is not conservative at all: the presence of solids changes the erosion mechanism entirely, and no value of C makes a single-parameter equation predict sand erosion. The calculator emits that second warning on every run.

Which is why C is your input here. RP 14E gives values by service — different for continuous and intermittent operation, and explicitly caveated for solids-bearing flow — and picking the right one for your fluid, your materials and your project's practice is the engineering decision. This tool does the arithmetic exactly, reports both velocities and the margin, and stays out of the choice that matters.

Method

Two velocities are computed and compared. The erosional velocity limit, API RP 14E ¶2.5:

V_e = C / √ρ_m

and the actual line velocity from volumetric flow and inside diameter, in the standard gpm/inch form:

V = 0.4085 · Q / d²

ratio = V / V_e → PASS when ratio ≤ 1

Units are fixed: ρ_m is the flowing mixture density in lb/ft³, Q the volumetric flow at flowing conditions in gpm, d the inside diameter in inches, and both velocities come out in ft/s. The 0.4085 constant is a pure unit identity converting gpm and inches of bore to ft/s — it is not a code value and carries no empirical content, unlike C, which carries all of it.

The word mixture in the density term is where most misuse of this equation begins. For a two-phase line, ρ_m is the density of the flowing gas–liquid mixture at the pressure and temperature at the point being checked, not the liquid density and not the gas density — and it changes along the line as pressure drops. A line that passes at the separator outlet can fail at the low-pressure end, because the mixture density there is lower, the limit is therefore higher, but the actual velocity has risen faster still as the gas expands. Check the demanding end, which for gas-bearing service is normally the outlet.

Entering Q = 0 puts the card in limit-only mode: it returns V_e and suppresses the operating check and the verdict, which is what you want when you are establishing a velocity ceiling before a line size exists. With a flow entered, a second warning fires when the actual velocity falls below roughly 3 ft/s — the low-velocity failure mode of liquid holdup and slugging in two-phase lines, which oversizing produces and which the erosional limit alone will never catch.

Limitations — what this calculator is not

• It is an empirical rule of thumb, not an erosion model. There is no particle size, no solids concentration, no material hardness, no bend radius, no impingement angle and no flow regime in the equation — all of it is compressed into C. Treat the result as a sizing screen with a long history, not as a prediction of metal loss rate.
• Solids-bearing service is outside it. When sand is present, erosion is dominated by particle impact and no single-parameter velocity limit describes it. The engine warns on every run; a real sand-erosion assessment (DNV-RP-O501 or equivalent) models particle flux, impact angle and target material, and can give a limit far below or, in some clean-geometry cases, above what RP 14E returns.
• Widely regarded as over-conservative for clean, solids-free service, and that conservatism has costs of its own: oversized lines run slow, and slow two-phase lines accumulate liquids, slug, and corrode under deposits. The low-velocity warning at roughly 3 ft/s is there because the failure mode at the other end of the range is real.
• ρ_m must be the flowing mixture density at the location being checked, and it changes along the line. A single check at one point does not qualify a line whose pressure — and therefore whose gas density and volumetric rate — varies significantly end to end.
• Straight-pipe velocity only. The limit is nominally about protecting fittings, bends, tees and control valve outlets where the flow changes direction, but the calculation is done on the straight-line bulk velocity. Local velocities at a restriction, a partially closed valve, or the inside of a short-radius elbow are higher than what is entered here, and that is where erosion actually occurs.
• No material, corrosion or CO₂ effects. Erosion-corrosion — where the flow strips a protective film and lets corrosion proceed at the bare surface — is a different and often more aggressive mechanism, and a velocity that is safe for pure erosion may not be safe for it. Corrosion-resistant alloys, coatings and inhibitor films all change the answer, and none of them enter here.
• Not a code requirement. RP 14E is a recommended practice for offshore production platform piping systems, and the erosional velocity equation is guidance within it. Many operators use it well outside that scope by convention; whether it applies to your line, and with what C, is a project decision.

Quick reference — how mixture density moves the limit

Because the limit falls with √ρ, the same constant produces very different ceilings across the density range a production facility actually sees. Values below are the equation evaluated at an illustrative C = 100 — they are arithmetic on the stated form, not RP 14E table data, and the constant for your service is yours to select:

Two readings follow. First, on dense liquid service the erosional limit is usually not the binding constraint at all — hydraulics gets there first, and sizing to V_e alone can produce a line that meets the screen and fails on pressure drop. Second, on light gas the limit is permissive enough that hitting it usually means something else is wrong: check noise, vibration and the local velocity at fittings before concluding the line is acceptable because the ratio came out under 1.

Worked example — fixture-verified

A clean, solids-free water line: NPS 6 schedule 40 (6.065 in bore) carrying 800 gpm at 62.4 lb/ft³. The project applies an erosional constant of C = 100 for this continuous, solids-free service.

Step by step

1. Density term: √ρ_m = √62.4 = 7.89937.
2. Erosional limit: V_e = C / √ρ_m = 100 / 7.89937 = 12.659 ft/s.
3. Bore squared: d² = 6.065² = 36.784225 in².
4. Actual velocity: V = 0.4085 · Q / d² = 0.4085 · 800 / 36.784225 = 326.8 / 36.784225 = 8.884 ft/s.
5. Ratio: 8.884 / 12.659 = 0.7018 — the line runs at about 70% of the erosional ceiling.
6. Verdict: PASS, with 3.775 ft/s of velocity headroom before the limit is reached.

Comfortably inside the screen, and comfortably clear of the low end too — 8.9 ft/s is well above the roughly 3 ft/s where holdup and slugging start to be a concern, so this line is not oversized either. That is the position to aim for on a two-phase or intermittently-flowing service: inside the erosional ceiling with margin, and above the deposition and holdup floor with margin. Note also how much room the ratio leaves: this line could carry about 1,140 gpm before reaching Vₑ, which is worth knowing before somebody proposes a debottleneck.

Worked example 2 — limit-only mode, before the line size exists

The same fluid and the same constant, but no flow rate entered. This is how the card gets used at the start of a sizing exercise: establish the velocity ceiling the service imposes, then choose a bore that keeps the design flow underneath it.

Step by step

1. The limit depends only on the constant and the mixture density — the bore does not enter it: V_e = 100 / √62.4 = 12.659 ft/s.
2. With no flow entered, the actual velocity and the ratio are suppressed and the card returns no verdict — status is COMPUTED rather than PASS or FAIL.
3. To use the limit for sizing, invert the velocity relation: the minimum bore for a given flow is d = √(0.4085·Q / V_e). At 800 gpm that is √(326.8/12.659) = √25.816 = 5.08 in.
4. Any bore above about 5.1 in therefore satisfies the screen at 800 gpm — which is why the 6.065 in schedule 40 line in the main example passes with room to spare.

Limit-only mode is the honest way to use this equation, because it makes the two decisions visible and separate: what ceiling the service imposes, and what bore you chose against it. The minimum bore the screen permits is almost never the bore you should install — it leaves nothing for flow growth, nothing for the pressure-drop check, and nothing for the local velocity at the fittings the limit is nominally protecting. Treat the number as a ceiling to stay under, not a target to size to.

Fixture case “limit-only mode (Q=0)” (tolerance 0.001) — locked in the same release gate as the example above.

Sources & citations

• API RP 14E, Recommended Practice for Design and Installation of Offshore Production Platform Piping Systems — ¶2.5, erosional velocity Ve = C/√ρm, including the C values by service; those values are user-supplied and not reproduced here.
• DNV-RP-O501, Managing Sand Production and Erosion — particle-based erosion modelling for solids-bearing service, where the RP 14E screen does not apply.
• ASME B31.3, Process Piping — ¶300.2 and ¶302, where erosion allowances enter as a mechanical allowance c in the wall thickness calculation rather than as a velocity limit.
• Constants: 0.4085 converts gpm and inches of bore to ft/s and is a unit identity, not a code value.

Per the source & citation policy, allowable-stress and factor table values are user-supplied. Where a page does reproduce specific ASME data (the B16.5 ratings, the quick-reference tables), it does so under ASME authorization and states the source table and conditions inline.

FAQ

What value of C should I use?

The one RP 14E specifies for your service, which is why it is an input and not a default. The recommended practice distinguishes continuous from intermittent service and treats solids-bearing flow separately, and many operators carry their own project values that differ from the published ones based on their operating experience and material selection. Two cautions worth repeating: a C intended for solids-free service applied to a sand-bearing line is not conservative, and a C chosen for carbon steel is not automatically right for a corrosion-resistant alloy, where a higher limit is often justified. This tool never selects it.

Why is this equation so often criticised?

Because it predicts erosion using a single empirical constant and a density, and erosion does not work that way. The mechanism depends on particle size, concentration and hardness, on impact angle and velocity at the target, on the target material's own hardness and its ability to hold a protective film, and on the geometry that turns the flow. RP 14E's form captures none of that. The criticism is not that it is useless — it is a fast screen with decades of field correlation behind it for clean production service — but that it is routinely used far outside the conditions it was correlated for, and quoted with a confidence the equation cannot support. Use it as a first screen; escalate to erosion modelling where solids or high-value assets are involved.

Should I use the liquid density or the gas density on a two-phase line?

Neither on its own — the flowing mixture density at the pressure and temperature of the point being checked, which is the mass flow divided by the volumetric flow at those conditions. It is a moving target: as pressure falls along the line the gas expands, the mixture density drops, the erosional limit rises, and the actual velocity rises faster. The result is that the demanding location on a gas-bearing line is normally the low-pressure end, not the inlet. If you check only one point, check that one — and if the density varies by more than a modest amount along the line, check several.

The line passes but the velocity is very low. Is that fine?

Often it is not, and it is the failure mode oversizing produces. Below roughly 3 ft/s in a two-phase line, liquids stop being swept along and start accumulating in low points and along the invert; that gives you slugging, unstable operation, and water sitting against the steel where it drives under-deposit corrosion far more effectively than the flow was ever going to erode it. The calculator raises a warning in that region for exactly this reason. The design target is a band, not a ceiling: fast enough to keep the line swept, slow enough to stay under the erosional limit and the pressure-drop budget.

Does passing this screen mean the line will not erode?

No. It means the bulk velocity in the straight run is under a rule-of-thumb ceiling. Erosion happens where the flow cannot follow the pipe — the outer radius of short-radius elbows, tees, reducers, downstream of control valves and orifices, and anywhere a partially closed valve is throttling. Local velocities in those places are considerably higher than the line velocity computed here, which is why erosion allowances, sweep radii, and material selection at fittings are separate design decisions. If sand is present, none of this is a screening problem at all: model it.

How does the erosional limit relate to slurry deposition velocity?

They are the two ends of the same design window on a solids-bearing line, and they push in opposite directions. The deposition velocity is a floor — go below it and solids settle out, build a bed and eventually plug the line. The erosional velocity is a ceiling — go above it and the solids you are keeping suspended start removing metal at every change of direction. A slurry line has to live between them, and on a demanding service the window can be narrow enough that the design is a genuine compromise rather than a choice. Size the floor with the Durand deposition velocity calculator and treat both results as constraints on the same bore.

Related calculators

• Slurry Critical Deposition Velocity Calculator (Durand–Condolios) — The floor at the other end of the same velocity window
• ASME B31.4 / B31.8 / B31.11 Pipeline Wall Thickness Calculator — Erosion allowance enters the wall as a mechanical allowance
• Pipe Support Spacing Calculator — Maximum Deflection-Limited Span — Line size chosen on velocity still has to span between supports