Pile Run Protection

A pile run is the sudden drop of a pile when the soil below it gives way during driving. What it does to the crane, what stops it, and what a POLARIS shock absorber did in a matched 1,500 t pile-run study.

By Tord Martinsen, CEO · March 2026 · Updated September 2026 · 10 min read

POLARIS, an offshore pile hammer and naturally hanging slack arrest slings.

In brief

A pile run is an uncontrolled drop of a pile, with the hammer on it, when the pile tip reaches a weak soil layer whose resistance is less than the weight of the pile and hammer. They can drop several metres in a second or two.
If the crane still holds the hammer, the line has to stop it, and the pile too if the two are clamped together. With a rigid connection, only the stretch of the crane, wire and rigging stops them, so the force spikes.
In our matched CONSTELLATION study of a 1,500 t pile and hammer at 5.0 m/s, the hook load reached 37.7 MN on a rigid link, 2.28 times the static hook load (a dynamic amplification factor, DAF, of 2.28) and 154 % of the crane’s safe working load (SWL). On a 2,000 t POLARIS it reached 23.6 MN (DAF 1.43, 96 % of SWL), which used 4.98 m of its 6.0 m stroke.
The absorber is sized for the worst credible run: the energy of the moving mass, the peak force the crane can take, and the stroke the arrest needs.

During driving, the hammer stands on the pile and the crane only follows it, on slack slings. If the pile runs, the slack is taken up in a moment and the crane line has to arrest what it holds: the hammer, and the pile too if the two are clamped together.

What a pile run is

Offshore piles are set in a guide frame or pile gripper and driven into the seabed with a hydraulic hammer. Between blows, the friction along the pile shaft and the resistance at its tip hold it up. Where the tip passes from a stronger layer into a weak one, such as loose sand or soft clay under a stiffer layer, that resistance can fall below the weight of the pile and hammer, and the pile accelerates under its own weight and the hammer’s. Soil borings and cone penetration tests (CPTs) show where such layers are; a pile-run analysis compares the soil’s resistance with that weight and works out how far and how fast a run could go.

Figure 1. During driving: the hammer stands on the pile, and POLARIS hangs in line in the crane fall with its rod retracted and slack soft slings to the hammer, so the hammer blows never pass through it. In a pile run: the pile tip reaches the weak layer and the pile runs on; the hammer drops until the slings come taut, and POLARIS strokes, holding the force at its valve setting until the hammer stops. The hammer’s drop is the slack taken up plus the stroke. Schematic, not to scale.

The absorber’s stroke is not the distance the pile runs through the soil. It is the travel between the absorber’s own connections while the pile and hammer are brought to rest, and the lift model relates the two.

What it does to the crane

A moving mass $m$ at speed $v$ carries the energy $\tfrac{1}{2} m v^2$ . The 1,500 t pile and hammer of the study, at 5.0 m/s, carry 18.75 MJ. Stopped by a connection of stiffness $k$ that comes taut at that speed, the kinetic energy, and the work the weight does while the connection stretches, go into stretching it, and the peak force is

F_{\text{peak}} = W + \sqrt{W^2 + k\,m\,v^2}

where $W$ is the weight of the moving mass. The stiffer the load path, the higher the peak. A crane hoist, its wire and the rigging are stiff, so a rigid arrest asks the crane for far more than the static load (a high DAF), and the line can then go slack and snap taut again as the load rings down. A shock absorber puts a soft element in that path: its gas spring builds the force up to a set level, and its valve holds the force there while the load is stopped.

With and without POLARIS

In our matched CONSTELLATION study CS-2026-005, a 1,200 t pile with a 300 t hammer clamped to its head moves at 5.0 m/s, with the line already carrying their weight. The model runs twice from one parameter set: once with a rigid link, once with a 2,000 t POLARIS with 6.0 m of stroke.

Figure 2. The same pile run, twice: a 1,200 t pile and 300 t hammer at 5.0 m/s, once on a rigid link and once on a 2,000 t POLARIS with 6.0 m of stroke. Study CS-2026-005, modelled in CONSTELLATION.

Result	Rigid link	With POLARIS
Peak hook load, against the 24.5 MN crane SWL	37.7 MN (154 %)	23.6 MN (96 %)
Hook DAF, on the 16,553 kN static load	2.28	1.43
DAF below the absorber, on the 14,710 kN payload	2.41	1.31
Full-slack rebounds after the arrest	2	0
Stroke used	no absorber	4.98 m of 6.0 m (83 %)

The compensated run was cross-checked against an independent solver. Because the line already carries the weight, the closed form for this case is $W + v\sqrt{k m}$ ; it brackets the rigid-link peak between 34.2 MN, with the crane’s own stretch, and 44.8 MN, with a rigid crane. From slack slings the peak would be higher, and with a hammer standing free on the pile the line would catch only the hammer. Read the pile-run arrest study →

Sizing the absorber

Three requirements have to agree, for the worst credible run rather than the expected one:

Energy. The kinetic energy of the moving mass when the line takes it: 18.75 MJ in the study. The weight keeps pulling while the absorber strokes, so only the force above that weight slows the load.
Peak force. The force the crane, wire, rigging and hammer can take. The valve setting sits between the weight and that limit: in the study, 18.0 MN, between 14.7 MN and the crane’s 24.5 MN SWL.
Stroke. The travel the force above the weight needs to take up the kinetic energy, with margin at the end. An absorber that bottoms out stops protecting the crane.

Over the 4.98 m the study used, the absorber took the 18.75 MJ of kinetic energy plus about 73 MJ of work done by the weight. Our crane shock absorption article gives a first screen for the stroke from the energy and the allowed peak; at the DAF of 1.43 the study reached, it asks for 5.85 m, and the 6.0 m unit used 4.98 m. A time-domain model of the crane, rigging, absorber, hammer and pile then confirms the peak.

Planning for pile-run risk

Pile-run risk is assessed while the piling is planned. A pile-run analysis of the soil data gives the credible run scenarios, each with a run distance and speed. The pile and hammer masses turn them into energy, and the crane’s chart sets the force it can take. Together they give the absorber’s energy, force and stroke.

The rigging matters as much as the absorber. POLARIS hangs in line in the crane fall, and soft slings run from its rod end to the hammer. During driving they hang slack and the rod stays retracted, so the hammer’s weight and blows never pass through the absorber and its full stroke is ready when the pile runs.

The absorber’s own weight also counts against the crane chart for the lift. POLARIS keeps a small gas volume and CFD-designed valves inside one hydraulic cylinder, with no external accumulators or needle valves, so it stays light for its capacity: 75 to 4,000 t, with 1.0 to 8.0 m of stroke.

Sources and assumptions

The study is our CONSTELLATION study CS-2026-005 (Revision A, 4 August 2026): a 1,200 t pile with a 300 t hammer clamped to its head at 5.0 m/s, with the line already carrying their weight, a crane with a 24,517 kN SWL, and a 2,000 t POLARIS with 6.0 m of stroke and an 18,000 kN valve setting. The hook DAF is on the 16,553 kN static hook load. It is a modelled study, not a certified project analysis.
The peak-force relation is the energy balance of a falling mass stopped by a linear spring that comes taut at the run speed. Real load paths are not linear, and the time-domain result governs.
Run distances and speeds come from each site’s own soil data; the study’s 5.0 m/s is one case, not a typical value.

Frequently asked questions

What is a pile run?

An uncontrolled drop of a pile during driving, when its tip reaches a weak soil layer whose resistance is less than the weight of the pile and hammer. They can drop several metres in a second or two, and the crane line then has to stop the hammer, and the pile too if the two are clamped together.

How dangerous is an unprotected pile run?

In our matched study of a 1,200 t pile and 300 t hammer at 5.0 m/s, a rigid link put 37.7 MN on the hook (DAF 2.28), 154 % of the crane’s SWL, and the line went fully slack twice as the load rang down. The peak in another case depends on its moving mass, its speed and how stiff the load path is.

How does a shock absorber protect against a pile run?

It hangs in line between the crane fall and the hammer. When the pile runs and the slings come taut, it strokes: its gas spring builds the force up to the valve setting, and the valve holds it there while the load is brought to rest. In the study, POLARIS kept the hook at 23.6 MN (DAF 1.43), inside the crane’s SWL.

How is pile-run protection sized?

From the worst credible run. Soil borings and CPTs feed a pile-run analysis, which gives the run distance and speed; the moving mass and speed give the energy; the crane’s chart gives the force it can take; and the stroke has to let the force above the weight take up that energy, with margin at the end. Our crane shock absorption article has a first screen for the stroke.

Why are the slings slack before a pile run?

During driving the hammer stands on the pile and the crane only follows it. Slack soft slings keep the hammer’s weight and blows out of the absorber, so its rod stays retracted and its full stroke is ready when the pile runs.

Is pile-run protection relevant for offshore wind?

Yes. Monopiles are driven through layered soils, and a run can happen where a weak layer under a stronger one cannot carry the weight of the pile and hammer. The absorber’s own weight also counts against the crane chart, which is where a light unit helps.

Protecting a crane against a pile run?

Send us:

the pile and hammer masses
the soil data, or the run scenarios with distance and speed
the crane, its chart and the rigging arrangement
the peak load or DAF you can accept

We’ll return a POLARIS screening and the inputs needed to verify it.

Send your lift case →