Ground-borne vibrations due to

press-in piling operations

D.J. Rockhill, M.D. Bolton and D.J. White

Cambridge University Engineering Department

Abstract
The press-in method has the potential to facilitate pile driving in locations
poorly suited to traditional dynamic piling methods, since it creates less noise
and ground vibration.
A body of vibration data gathered from press-in sites in Japan and the UK is
presented, from which a semi-empirical method for the prediction of the
ground-borne vibrations associated with press-in piling is derived. Aided by this
work, designers can assess the possibility of specifying the press-in technique in
areas sensitive to vibration.

Introduction
Design codes place limits on the ground vibrations and noise created by
construction operations. These limits are intended to prevent disturbance to
humans and damage (both cosmetic and structural) to nearby buildings.
Irreparable damage caused to listed buildings is of particular concern.
Conventional dynamic piling methods, such as vibrators and drop hammers,
create large vibrations and thus their use is precluded in certain locations,
particularly densely populated urban areas.
The press-in method is a non-dynamic method for the installation of pre-
formed piles (Figure 1). The technique uses hydraulic rams to push piles into
the ground and is presented as a ‘silent’ or ‘vibration-free’ method, although
there is limited data to quantify this feature. As such, when designers are
considering the press-in method they are unable to predict the associated
ground-borne vibrations, since the field measurements of piling-induced
vibrations used in design code guidelines are from dynamic piling methods.
2 Header – book title

Figure 1 Installation of sheet piles using the press-in method

Background
For engineering purposes, ground vibrations are usually quantified in terms of
Peak Particle Velocity (ppv), which is defined as the vector sum of the
maximum velocity components of vibration, as shown in Equation 1.

ppv = v 2x,max+v 2y,max+v 2z,max (1)

PPV is a measure of the damage potential of vibration – the velocities

themselves do not cause structural damage or human disturbance. In the case of
building damage, it is the resulting dynamic strains that are of concern1. Human
distress is often linked to acceleration level1. However, the ppv parameter is
easy to measure and correlates well with the measured effects of ground-borne
vibrations1, and therefore provides a robust indicator of damage potential.
This paper reports fieldwork in which velocities in three orthogonal
directions are measured directly using a triaxial geophone set (Figure 2).
Geophones are self-exciting, giving an output voltage proportional to the
imposed velocity and with low impedance, permitting long cable runs. Two
triaxial geophones were used, permitting simultaneous measurement at different
positions on a test site, as is recommended practice2. The ppv at each location is
determined by combining the peak measured velocity components, which may
not occur simultaneously; the resulting value of ppv is referred to as the
‘simulated resultant’ ppv3. In all field tests, ppv values were calculated from the
velocity history during the installation of a single pile.
 Header – chapter title or author 3

Figure 2 Geophone, showing orthogonal directions of velocity measurement

Limits on the maximum allowable ppv caused by construction operations are

given in various design codes4-6. This paper refers only to the Eurocode 3 limits
on ground-borne vibrations6 (Figures 3 and 4).

10
Acceptable if warning is given:
Construction period < 6 days
Construction period 6-26 days
Construction period > 26 days
ppv (mm/s)

Acceptable if no warning is given:

Construction period < 6 days
Construction period 6-26 days
Construction period > 26 days

Laboratories, hospitals and libraries

0.1
1 10 100
Distance (m)

Figure 3 Eurocode 3: Maximum acceptable vibrations to prevent human

disturbance
 4 Header – book title
100

Buried services
Heavy industrial
Light commercial
ppv (mm/s)

Residential
10

Ruins and buildings of architectural merit

1
1 10 100
Distance (m)

Figure 4 Eurocode 3: Maximum acceptable transient vibrations to avoid

structural damage

Ground vibration propagation

Over the last thirty years a large body of research has been carried out on the
ground-borne vibrations created by traditional dynamic piling techniques. An
extensive database of vibrations measured at various construction sites has been
compiled and, from this, predictive methods have been developed1. There are a
number of different empirical predictors, but all take the form of a power law as
shown in Equation 2.
n
 w
ppv = C   (2)
 r 
Here w represents the energy per cycle of the piling process in Joules and r is
the distance between the source of vibration and the point of measurement in
metres. PPV is predicted in mm/s. The parameters C and n are site-specific and
depend on the soil characteristics, piling technique, pile type and ground profile.
These parameters also take account of the dimensional inconsistency of the
equation. C typically varies from 0.5 to 1.5; n varies from 0.5 to 1 in the various
standards and studies1, 7 and is specified as equal to 1 in Eurocode 36.
Previous authors state that approximately two-thirds of the energy of a
ground vibration is carried by Rayleigh waves7. Because Rayleigh waves
propagate as expanding rings, the energy per unit area of the wave decays in
inverse proportion to the distance from the source. This form of decay is known
 Header – chapter title or author 5
as geometric damping, because the damping is purely a function of the area
enclosed by the wave front as it propagates away from the source.
The other main mechanism by which the energy of the waves is dissipated is
material damping, whereby frictional losses occur during propagation. This is
purely a function of the propagating medium. There are other dissipative
mechanisms, such as reflection and refraction, which have a relatively small
effect on attenuation in the case of ground-borne vibrations, since the ground is
usually relatively homogeneous. Compared to the effects of geometric damping,
other damping mechanisms have a minimal influence on the attenuation. These
effects are largely ignored by predictive methods for ground-borne vibration1.
For the purposes of this work, only the effects of geometric damping on wave
attenuation are considered.
The assumption that wave propagation is non-dissipative allows the
application of simple elastic wave theory to find an expression for the ppv of a

wave at a given distance from a point source. The derivation of this expression
is given below.

Figure 5 Wave emanating from point source with amplitude a(r, t), peak
amplitude a0, travel velocity vr, and transverse particle velocity vp.

The wave equation of motion is:

a = a0 sin γ ( r − vr t ) (3)

Where the frequency of excitation ω = γ vr

Differentiating equation 3 to obtain vp, the transverse particle velocity:
da
vp = = −γv r a0 cos γ ( r − v r t ) (4)
dt
6 Header – book title

If the soil is assumed to be linear elastic with arbitrary stiffness k, the energy
transmitted by the source on each cycle E is:
1 2
E= ka 0 (5)
2
If the wavefront is assumed to be cylindrical in shape, then the energy of the
waves decays in inverse proportion to distance from source, due to geometric
damping.
1 A
E∝ ∴ a0 = where A is an arbitrary constant (6)
r r
Substituting a0 back into Equation 4:
Avrγ Aω
vp = − sin γ (r − vr t ) = − sin γ (r − vr t ) (7)
r r
Taking the maximum value of vp gives the ppv:
Aω
v p , peak = − (8)
r
Alternatively, if the waves are assumed to propagate as expanding spheres, then
the energy of the waves decays in inverse proportion to the square of the
distance from the source, giving:
Aω
v p , peak = − (9)
r
A is a parameter which depends on the properties of the medium and the initial
energy of the wave. The similarity between equations 8 and 9 and equation 2
should be noted. Fieldwork has been conducted to empirically establish the
value of the parameter A for the prediction of ground vibrations near press-in
piling.

Fieldwork
A database of ground vibrations caused by piling activities has been collated
from monitoring visits to sites in Japan and the UK using two triaxial
geophones and DASYLab 6.0 data acquisition software. Recordings have been
made at two types of site where the press-in method is in use:

• Test sites – where the piling is conducted for the purpose of vibration
measurement
• Construction sites – where the recording is a secondary purpose of the
piling work.
 Header – chapter title or author 7
Test sites tend to be more carefully controlled and so there is less
background noise and disturbance; a much cleaner recording is achieved.
Conversely, construction sites yield a vibration recording that generally has a
lower signal to noise ratio, yet is more representative of real conditions.
Monitoring has taken place at one test site (using two different piling machines)
and five different construction sites. Different modes of press-in operation
(including water-jetting and augering) have been monitored at the various sites.
Vibrations arising from dynamic piling operations have also been recorded in
order to make a direct (site specific) comparison with the press-in method.

Table 1: Description of test sites (continued overleaf)

Soil
Test Location Date Piler type Pile type
properties
Made
Takasu test- 0.4m x
July ground Giken Super
1 site, Kochi 6.5m sheet
2002 overlying Auto 75
Japan piles
silty sand
Made 0.1m
Takasu test-
July ground Giken NT diameter
2 site, Kochi
2002 overlying 150 8m tubular
Japan
silty sand piles
Made
Takasu test- Diesel
July ground
3 site, Kochi generator N/A
2002 overlying
Japan (1800 rpm)
silty sand
Othu Loose, 0.4m x
July Giken Super
4 Funaire, stony fill 10m sheet
2002 Auto 150
Kochi, Japan overlying piles
Type SS-40L
Othu Loose, 0.4m x
July low amp –
5 Funaire, stony fill 12m sheet
2002 high freq
Kochi, Japan overlying piles
vibrohammer
Giken Super
0.4m x
Tosashi, July Loose, Auto 100
6 14.5m
Kochi, Japan 2002 stony fill (water jetting
sheet piles
@ 7MPa)
Made
Giken Super
Atago, July ground 0.4m x 8m
7 Crush 100M
Kochi, Japan 2002 overlying sheet piles
(auger)
silty clay
Iriake, July Rocky made Giken Super 0.4m x 6m
8
Kochi, Japan 2002 ground Auto 75 sheet piles
8 Header – book title

Soil
Test Location Date Piler type Pile type
properties
Rubble fill Giken Super
Westbourne over soft Auto UP150 0.6m x
January
9 Grove, clay and (water jetting 12m sheet
2003
London London for piles
Clay lubrication)
0.6m x
Autumn Silt, sand
108 Norway Giken ZP150 15m sheet
1998 and clay
piles

Results
The large amount of data recorded in the acquisition stage required analysis in
order to extract the salient ppv information and draw accurate and useful
conclusions. In order to reduce and analyse the data, a graphical user interface
(GUI) was developed using Matlab. The GUI performs a number of operations:

• Reads in data file, plots the time series for all six channels, calculates ppv
for both geophones for any specified time interval
• Plots the frequency spectrum for all six channels.

With the aid of this GUI, a plot of ppv against distance for the various
different monitoring sites has been produced (Figure 6). A single value of ppv
has been extracted from the geophone time history of each pile installation.

1000
Test 1
Test 2

100 Test 3
Test 4
ppv / mm/s

Test 5
10
Test 6
Test 7

1 Test 8
Test 9
Test 10
0.1
0.1 1 Distance / m 10 100

Figure 6 ppv vs distance data acquired from press-in sites

 Header – chapter title or author 9
Frequency spectra of different states of operation of the press-in piler are
shown in Figures 8 to 10. Figure 8 shows the vibration spectrum over a time
period during which a number of piles were installed intermittently. Figure 9
shows the frequency spectrum when the piler is inactive. Figure 10 shows the
frequency spectrum at the moment at the end of a stroke, when the piler chuck
releases the pile. It is evident that the piling activity causes ground vibrations of
frequencies less than about 15Hz. The peak at around 30Hz relates to the
generator (which runs at approximately 1800rpm) and the peak at around 50Hz
is due to electromagnetic interference from the mains electricity supply. It
should be noted that the geophones are only accurate at frequencies above 6Hz.
Analysis shows that the major vibrations caused by the press-in method,
represented by spikes in the time series (Figure 7), are transient. Crosschecking
with a time history of the piling activity shows that these spikes correspond to
events such as the closing and releasing of the piler chuck. At these instances, as
the grip of the piler on the pile is released, any elastic compression or bending
of the pile is released, leading to the transient vibration spikes evident in Figure
7. The regulatory limits on structural damage caused by ground-borne
vibrations depend on whether the disturbance is transient or continuous. In the
case of the press-in method, the ppv occurs during the transient vibrations
associated with gripping, driving and releasing the pile, and is the vector sum of
the transient piler vibrations and the continuous (and approximately constant)
background vibrations from the generator and crane. Because the transient
velocities are so much greater than the continuous velocities, the ppv can be
assumed to be transient.
2 4.0
Pile penetration (m)

1.5 3.0

1 2.0

0.5 1.0
Velocity (mm/s)

0
0

-0.5

-1

-1.5

-2
0 50 100 150 200 250 300 350
Time (s)
Figure 7 Plot of ground velocity against time for a typical press-in piling
operation, showing increasing penetration depth of pile.
10 Header – book title

Relative amplitude

Generator frequency

Mains electricity frequency

0 10 20 30 40 50 60 70 80 90 100
Frequency / Hz

Figure 8 Frequency spectrum of the unfiltered signal during pile installation

Relative amplitude

0 10 20 30 40 50 60 70 80 90 100
Frequency / Hz

Figure 9 Frequency spectrum of the background signal (i.e. when the piler is
inactive)

Over short distances (closer than 2 m from the pile) it has been assumed that
the ground vibrations decay as per Equation 8, i.e. that the propagating ground
waves are cylindrical. By analysing the collated data, a value for the constant A
 Header – chapter title or author 11
in Equation 8 has been derived. Based on the frequency spectrum shown in
Figure 10, the value of frequency f, which is the dominant frequency of the
transient vibrations caused by the piler, was chosen as 8 Hz. This single value is
for all the data for the press-in method gathered to date and, as such, does not
account for variations in pile type, piler type and soil conditions – all of which
will affect the value of A - and therefore has a large standard deviation. The
best-fit value of A is 0.000147, with standard deviation 0.000106. This value of
A was used to plot a single predictive line for the press-in method (see Figure
11).
At greater distances from the pile it has been assumed that the ground
vibrations decay as per Equation 9, with Aω = 10.43 being an appropriate
constant. This fitting correlates well with the data collected in this paper, as well
as with the data collected by NPRA8 at greater distances. These two prediction
lines for near and far-field ground vibrations are shown as Equation 10. The
predictive line derived by White et al (2002)9 is also shown on Figure 11, along
with the predictions of ground-borne vibrations arising from dynamic piling
methods, as predicted by Eurocode 3. Tables 2 and 3 show recommended
minimum separations between piling works and people or structures which have
been calculated by combining Equation 10 with the Eurocode limits shown in
Figures 3 and 4.
7.37 10.43
For r < 2m v = ; r ≥ 2m v = (10)
r r
Relative amplitude

0 10 20 30 40 50 60 70 80 90 100
Frequency / Hz

Figure 10 Frequency spectrum associated with the transient vibrations caused

by the piling operation
12 Header – book title

1000

Euro co de 3 predictio n: 25 kJ dro p hammer

100 Euro co de 3 predictio n: 170kW, 27Hz vibro hamme

P ress-in predictio ns

10
Equat ion 8, Aw = 7.37

Ro ckhill et al (2003)
0.1 White et al (2002)9
Equat ion 9, Aw = 10.43

0.01
0.1 1 10 100
D is t a nc e ( m )

Figure 11 Predictions of ground-borne vibrations arising from various piling

techniques

Table 2: Recommended minimum separations between people and piling

Maximum construction Piling method
time / days (limits from 25 kJ drop 170 kW 27Hz
Press-in
Eurocode 3) hammer vibrohammer
<6 (3 mm/s) 3.5 m 39.5 m 18.5 m
With
6-26 (2.3 mm/s) 4.5 m 51.5 m 24.1 m
warning
>26 (1.5 mm/s) 7.0 m 79 m 37 m
<6 (1.5 mm/s) 7.0 m 79 m 37 m
Without
6-26 (1.3 mm/s) 8.0 m 91.2 m 42.7 m
warning
>26 (1.0 mm/s) 10 m >100 m 55.5 m

Table 3: Recommended minimum separations between sensitive buildings

and piling
Piling method
Building type (limits on
vibrations from Eurocode 3) 25 kJ drop 170 kW 27Hz
Press-in
hammer vibrohammer
Architectural merit 2.6 m 29.6 m 27.7 m
Residential 0.5 m 11.8 m 13.8 m
Light commercial 0.14 m 5.9 m 5.5 m
Heavy industrial 0.06 m 3.9 m 3.7 m
Buried services 0.03 m 2.9 m 2.2 m
 Header – chapter title or author 13
It should be noted that at particularly noisy construction sites, at distances
above approximately 5 metres, the vibrations from the press-in piling rig are in
the region of 1 mm/s and approach the magnitude of incidental vibrations
arising from passing traffic and other works. Above this distance the predictive
method is therefore of limited relevance at these sites, since the piling operation
is indistinguishable from the background vibrations.

Conclusions
The press-in method combines the sustainability advantages of traditional
dynamic piling techniques (in that preformed piles can be extracted and the site
reused) with the low environmental disturbance associated with bored piles.
Through field measurements of ground-borne vibrations at press-in piling
sites, a method has been developed to predict these vibrations. The reduction in
ground-borne vibrations achieved through the use of the press-in method instead
of other dynamic methods can reduce the separation between piling operations
and sensitive structures by a factor of 10-50. The separation between the piling
operations and the public can be decreased by a factor of up to 5. Equipped with
this guidance, designers can predict the level of disturbance associated with the
press-in method and thus confidently specify the technique in locations for
which displacement piling could not previously be considered.

Acknowledgements
This research was conducted with the support of Giken Seisakusho Co. Ltd. The
authors acknowledge the technical assistance provided by Mr. T Nagayama,
Ms. A.G. Yetginer and Mr. A.J. Deeks.

References
1. HEAD, J.M. and JARDINE, F.M. Ground-borne vibrations arising from piling.
CIRIA Technical Note 142, 1992.
2. BRE DIGEST 403, Damage to structures from ground-borne vibration, 1995.
3. HILLER, D.M. and HOPE, V.S. Groundborne vibration generated by mechanized
construction activities. ICE Proc. 131: 223-232, 1998.
4. BS 7385-2:1993, Evaluation and measurement for vibration in buildings – Part 2:
Guide to damage levels from groundborne vibration.
5. BS 5228-4:1992, Noise control on construction and open sites – Part 4: Code of
practice for noise and vibration control applicable to piling operations.
6. ENV 1993-5, Eurocode 3: Design of steel structures – Part 5: Piling.
7. HILLIER, D.M and CRABB, G.I. Groundborne vibration caused by mechanised
construction works. TRL Report 429, 2000.
8. NPRA. Environmental effects related to the construction of a cut and cover road
tunnel. Norwegian Road & Transport Research Vol 13, No.1 4-5, 2001
9. WHITE, D., FINLAY, T., BOLTON, M. & BEARSS, G. Press-in piling: Ground
vibration and noise during pile installation, ASCE Spec. Pub. 116 363-371. 2002.