Pile Driving Formula

The problem the formula solves

You have a pile driver on a construction site. The hammer drops onto the pile head, the pile sinks a few millimetres, the hammer lifts and falls again. After thousands of blows the pile is in the ground and the foreman wants to know one thing: has it reached enough soil resistance to carry the design load? The static load test that would answer definitively takes a day to set up, a day to load, and tens of thousands of dollars. So instead, the foreman watches the rate at which the pile is going in.

If the pile sinks 50 mm per blow, it is moving through soft ground and has not found resistance. If it sinks 1 mm per blow, it is being held back by enormous shaft friction and tip pressure — the soil is pushing back hard, and the pile has found its bearing layer. Somewhere between these extremes is the set that corresponds to the design capacity. A pile-driving formula is the mathematical statement of that intuition: bearing capacity is some function of penetration per blow, given the hammer's kinetic energy.

The earliest such rule is the Engineering News formula, published by A.M. Wellington in 1888 in the New York-based Engineering News magazine. It has been around for nearly 140 years, and despite being known to be inaccurate it is still the single most widely used field-control formula in the world.

The Engineering News Record formula

           W · H
Q_a  =  ─────────       ← allowable bearing capacity
          s + c

where
   W  =  ram weight (e.g. lb or kN)
   H  =  drop height (e.g. ft or m)
   s  =  average set per blow at end of driving (same length units as H, c)
   c  =  empirical loss constant
            c = 0.1 inch (2.5 mm) for steam, air, or diesel hammers
            c = 1.0 inch (25 mm) for drop hammers

The formula already includes a factor of safety FS = 6 against
ultimate capacity Q_ult, i.e.   Q_ult ≈ 6 · Q_a.

Numerator W·H is the hammer's potential energy at the start of the drop — the kinetic energy at impact, assuming a frictionless fall. Denominator s + c is the equivalent plunge distance after accounting for elastic rebound c. The ratio is force per blow, which Wellington's calibration against late-19th-century static load tests showed to give a serviceable estimate of allowable capacity if you divide by a (built-in) factor of six.

Worked example: a 12-m steel H-pile

A driven HP 14×73 steel H-pile, 12 m long, is being installed for a bridge pier with a single-acting steam hammer:

Hammer:          Vulcan No. 1 — ram weight W = 5,000 lb (22.2 kN)
                 Stroke H = 3 ft (0.91 m), max rated energy 15,000 ft·lb (20 kJ)
Pile:            HP 14×73 steel, 12 m long, ~109 cm² cross section
End of driving:  Inspector marks pile, counts blows over the last 1 ft (305 mm).
                 Result: 60 blows per ft.
                 → set s = 12 inch / 60 = 0.20 inch (5.1 mm) per blow

Apply ENR (English units, c = 0.1 in for steam hammer):
   Q_a = (W·H) / (s + c)
       = (5,000 lb · 36 in) / (0.20 in + 0.10 in)
       = 180,000 in·lb / 0.30 in
       = 600,000 lb
       = 600 kips
       = 2,670 kN allowable

Ultimate (FS = 6 baked in):
   Q_ult ≈ 6 · 2,670 = 16,000 kN

Sanity check: static load test on a representative pile gave
   Q_ult,test = 12,500 kN  →  ENR over-predicted by ≈ 25%.
The contract spec says "drive each pile to set ≤ 0.20 in/blow under
the specified hammer or to refusal, whichever comes first."

This 25% over-prediction is typical. Field crews use the ENR-derived set as a stopping criterion; static load tests or PDA on selected piles verify that the assumed capacity is actually present. If a load test reveals significantly less capacity than ENR predicts, the contract's target set is tightened (e.g. from 0.20 in/blow to 0.12 in/blow) and remaining piles are driven harder.

Other historical pile-driving formulas

Formula	Year	Form	Notes
Sanders	1851	Qa = W·H / (8·s)	Earliest U.S. rule; assumes very rigid pile, no rebound term. FS ≈ 8.
Engineering News (ENR)	1888	Qa = W·H / (s + c)	Wellington's formula; FS = 6; c distinguishes hammer type.
Hiley	1925	Qu = (η·W·H) / (s + ½(c₁+c₂+c₃))	Adds hammer efficiency η and separate pile/cap/soil compression terms. More rational but harder to calibrate; widely used in UK and Australia.
Janbu	1953	Qu = η·W·H / (ku·s)	Norwegian; ku a function of pile/hammer weight ratio. Common in Scandinavia.
Danish (Sorensen-Hansen)	1957	Qu = η·W·H / (s + ½·se)	se = elastic compression of pile under blow. Adopted in Danish code.
Gates	1957	Qu = 7·√(η·W·H) · log(10/s)	Best-fit empirical regression against 100+ load tests. Lower scatter than ENR; recommended by FHWA pre-1990.
Modified ENR	1965	Qu = (1.25·W·H / (s+c)) · (W + n²Wp) / (W + Wp)	Accounts for pile weight Wp and coefficient of restitution n. FHWA's preferred formula 1965–1985.

Despite the procession of refinements, none of these formulas reduces predictive scatter below about 30–40% coefficient of variation. The fundamental problem is that pile driving is a wave-propagation phenomenon, and a single number (set per blow) cannot capture the rich information in the actual force-time and velocity-time histories at the pile head. Closing that gap required moving from algebra to integrated stress-wave measurement.

The wave equation method

In 1958, E.A.L. Smith of Raymond Concrete Pile Company published a one-dimensional wave-equation model of pile driving. The pile is modelled as a series of lumped masses (a few hundred along its length) connected by elastic springs; the hammer is modelled as a falling mass that strikes the pile head through a cushion (also a spring); the soil resistance is modelled as a static elasto-plastic spring at each pile element plus a velocity-dependent damping force. The differential equation is integrated numerically in time, blow by blow.

Outputs include:

• Bearing graph: a curve of ultimate capacity Q_ult versus blow count per unit penetration, for the specific hammer-pile-soil combination. Field crews compare the measured blow count against the graph to read off capacity.
• Peak stresses: tension at the pile tip during rebound, compression at the head during impact. If predicted peak compression exceeds the pile's strength, the hammer is too heavy or the cushion too stiff.
• Hammer efficiency: the fraction of rated energy delivered to the pile (typically 30–80% for diesel hammers, 60–90% for steam hammers).

The commercial implementation is GRLWEAP (Goble Rausche Likins Wave Equation Analysis of Piles), now in version 2010+. Every major driven-pile project starts with a GRLWEAP run that produces a target blow count and stress envelope.

High-strain dynamic pile testing (PDA)

The Pile Driving Analyser, developed at Case Western Reserve by Goble and colleagues in the 1970s, takes the wave equation idea full circle: instead of predicting the wave behaviour, measure it directly.

Strain gauges (typically two diametrically opposite, to capture bending) and accelerometers (two more) are bolted to the pile shaft about two pile diameters below the head. As each hammer blow lands, the gauges record force F(t) and the accelerometers double-integrate to give velocity v(t), at 10–50 microsecond resolution over the ~40 ms blow event. The two signals reveal:

• Wave-up and wave-down: F(t) and Z·v(t), where Z = E·A/c is the pile's characteristic impedance (E = Young's modulus, A = cross-section, c = wave speed ≈ 5,000 m/s in steel). Separating these tells the engineer how much wave is travelling down (input from hammer) versus up (reflected from soil resistance).
• Static bearing capacity by the Case method: R_SP = ½[F(t₁) + F(t₁ + 2L/c) + Z·(v(t₁) − v(t₁ + 2L/c))] − J_c·(F(t₁) − Z·v(t₁)), where t₁ is the wave's first peak, L is pile length, and J_c is the Case damping coefficient (≈ 0.4–0.7 depending on soil type).
• Refined capacity by signal matching (CAPWAP): a wave-equation model is run repeatedly with the measured input force as boundary condition, adjusting soil resistance distributions along the shaft and at the tip until the predicted velocity matches the measured velocity. The result is a high-quality estimate of shaft friction (versus depth) and tip resistance separately.

PDA + CAPWAP is the gold standard for capacity verification on driven piles. Static load tests remain the absolute reference (Q_ult by direct measurement), but they are expensive and slow; one static test per 100 PDA tests is common practice on a large project.

Hammer types in modern practice

Hammer	Mechanism	Energy (typ.)	Efficiency	Use
Drop hammer	Free-fall ram on cable	5–50 kJ	50–70%	Light timber piles, small jobs, marine
Single-acting steam/air	Steam/air lifts ram, gravity drops	10–200 kJ	60–80%	Mid-size concrete and steel piles
Double-acting steam/air	Steam/air both lifts and drives	20–250 kJ	50–75%	Higher cycle rate, sheet pile driving
Diesel hammer (single)	Internal-combustion ram-stroke	20–300 kJ	30–60%	Mainstream onshore driven piles
Diesel hammer (double-acting)	Compression assists return	30–400 kJ	30–60%	Heavy concrete piles, port works
Hydraulic hammer	Hydraulic ram, computer-controlled drop	50–1,000 kJ	85–95%	Offshore monopiles, mega-piles (wind turbines)
Vibratory hammer	Rotating eccentrics, harmonic shaking	—	—	Sheet piles, granular soils — not for bearing piles

Hammer choice affects both achievable capacity and energy transferred per blow. Hydraulic hammers, with their computer-controlled stroke and very high efficiency, are now standard on offshore wind monopile installations where ram weights exceed 100 tonnes and ratings exceed 4,000 kJ per blow.

How set is measured in the field

PILE SHAFT WITH SET PAPER (side elevation)

     ┌─┐         hammer
     │ │ ────────────
     │ │
     ╧═╧═══════════ pile cap
     │ │
     │█│ ◄── pile inspector marks the pile shaft
     │█│        in 1-inch increments
     │█│
     │█│
     │█│
     │█│       N_blows counted while hammer drops
     │█│       from 1 mark to next
     │ │
     │█│       set s = (1 inch) / N_blows
     │ │       or
     │█│       set s = (300 mm) / N_blows over 12 inch
     │ │
   ──┴─┴── ground surface
     ░░░░
     ░░░░░░░ soil resistance

In practice the inspector marks the pile shaft at 6 inch (150 mm) or 12 inch (305 mm) intervals with chalk or paint. As the hammer drives the pile down, the inspector counts the number of blows it takes for the topmost mark to descend to the cap. From that count, set per blow is computed. Common driving criteria:

• Practical refusal: ≤ 2 mm per blow (≥ 150 blows per 300 mm).
• Hard driving: 2–5 mm per blow.
• Moderate driving: 5–12 mm per blow.
• Easy driving: ≥ 12 mm per blow — keep going.

Note that the ENR-based target set depends on the contract-specified hammer; switching to a different hammer requires a new target. This is why hammer changes are typically prohibited mid-project without an engineer's review.

Why ENR is wrong by 50% on a bad day

The ENR formula bakes in five strong simplifying assumptions, each of which can be wrong by a factor of two:

1. Hammer efficiency is constant. ENR uses W·H as if 100% of the rated energy reaches the pile. In reality, diesel hammer efficiency varies from 30% to 60% depending on cushion condition, fuel injection, pile head alignment, and even ambient temperature. A pile inspector who measures 0.20 in/blow has no idea whether the actual delivered energy was 100% or 40% of rated.
2. Pile is rigid. The formula treats the pile as a rigid block driven by a single impact. In reality the pile is an elastic rod and the hammer blow generates a stress wave that takes 5–10 ms to traverse a long pile. The dynamics of that wave control how much soil resistance is mobilised on a given blow.
3. Soil resistance is the same under dynamic and static loading. Many soils show velocity-dependent damping — they resist faster than they would resist a slow static load. The dynamic resistance during driving can be 20–80% higher than the static capacity, especially in saturated clays where pore-pressure effects are large. ENR (and other formulas) do not separate these.
4. Loss constant c is fixed. The elastic rebound of pile, cap, and soil depends on pile length, pile stiffness, and cushion thickness — none of which are accounted for in a fixed c = 0.1 inch.
5. Soil setup is ignored. Many fine-grained soils gain capacity over hours to days after driving (the disturbed clay around the pile re-consolidates and gains shear strength). ENR captures only end-of-driving capacity; long-term service capacity may be 50–200% higher. PDA restrike tests after 24 hours measure this directly.

The combined effect is that ENR-predicted capacities have a coefficient of variation of roughly 50% when compared against static load tests on the same piles. Even the best-tuned formulas (Gates, Janbu) bring this down only to about 25–35%. PDA + CAPWAP gets to 10–20%. Static load testing — the absolute reference — is the only method that defines capacity by direct measurement, but it is too slow and expensive for production use.

Real-world driven-pile projects

• Confederation Bridge (Canada, 1997). 12.9 km bridge between Prince Edward Island and New Brunswick. 1.8 m diameter steel pipe piles driven 24 m into seabed sandstone using a Menck MHU 1700 hydraulic hammer. Capacity verified by PDA on every pile.
• Burj Khalifa (UAE, 2010). 192 bored piles, but the auxiliary construction crane platforms used driven steel H-piles with ENR-based driving criteria.
• Tappan Zee replacement / Mario Cuomo Bridge (USA, 2017). 6 km Hudson River crossing. 1,200 driven steel pipe piles, 1.4 m diameter, 60–110 m long. Capacity controlled by GRLWEAP wave-equation modelling and verified by PDA on 30% of piles.
• Hornsea offshore wind farm (UK, 2022). 174 monopile foundations driven into the North Sea bed. Each monopile is 8.1 m diameter, 70 m long, 1,100 tonnes; hammer is the IHC IQIP S-3500, 3,500 kJ per blow. All piles dynamically monitored.
• Salford Quays (Manchester, UK). Hundreds of timber and concrete piles driven 1880s–1920s using the original Wellington-era ENR formula. Many still in service after a century.

Common pitfalls in driven-pile capacity

• Treating ENR as a design tool. ENR is a field-control rule, not a design formula. Design capacity comes from static analysis (Q = Q_skin + Q_tip) using soil parameters from boreholes; ENR or PDA verifies that the design assumption held.
• Ignoring soil setup. Driving a clay-shaft pile, measuring set, declaring capacity, and loading it 6 hours later: the clay around the shaft has not re-consolidated yet and the actual capacity is much lower than it will be in a week. PDA restrike after 24 hours is the standard cure.
• Switching hammers without re-analysing. A faster hammer with the same rated energy may transfer less per blow (efficiency change), or may damage a pile that the slower hammer would not. GRLWEAP re-runs are mandatory after any hammer change.
• Counting blows over too short an interval. A 1-inch (25 mm) interval is too short — a single soft soil layer can dominate. Standard practice is to count over 12 inches (300 mm) for averaging.
• Driving past structural limit. Hard driving generates compression and tension stresses in the pile that can exceed yield (steel) or cracking (concrete) limits. The wave-equation analysis predicts these stresses; the contract specifies a maximum allowable blow count to prevent structural damage.
• Using ENR with a vibratory hammer. Vibratory hammers do not produce discrete blows; they shake the pile down by harmonic vibration. ENR is irrelevant; capacity must be verified by static load test or by switching to impact hammers for the final 1–2 m.