�ݺ�ߣ

Dr.C.R. Golightly, BSc, MSc, PhD, MICE, FGS.
Geotechnicaland Engineering Geology
Consultant
Rue Marc Brison 10G, 1300 Limal, Belgium
Tel. +32 10 41 95 25 Mobile: +44 755 4612888
Email: chris.golightly@hotmail.com
skype: chrisgolightly; Linked In:
“You Pay for a Site Investigation - Whether You do One or
Not”– Cole et al, 1991.
Golightly Monopiles 14th
April 2011 Page 1
Monopile Grouted Connection Failures
In late 2009 there were initial reports that the grouted connection joints between
large diameter driven tubular steel foundation piles and the connecting tubular steel
members placed over these piles were starting to fail, some years after completion of
construction and installation. The designs adopted were approved and certified. Shell
first discovered this systemic design error in October 2009 at their Egmond aan Zee
wind farm and the industry was made aware in January 2010, but it was only widely
publicised from April 2010, onwards. Grouted joints for wind turbine monopiles have
been found to be seriously high risk and several have failed on at least 18 projects at
the last count. Figures were published last year regarding the likely cost of fixing this
problem. The figures quoted were high, but they are probably nowhere near the real
costs. There are several projects still assessing the impact of this problem and what
to do about it.
Since then, it has been widely perceived that grout failures have primarily been due
to the widespread non-use of shear keys. The grouted connections which have not
failed did include for provision of shear keys by their designers. However the grout
failures also related to manufacturing and installation tolerances leading to piles and
guide sleeves being out of shape and the grout cement layers therefore being
unknown in thickness. The grouts used also behave as strong very brittle rocks
which have a tendency to crack and crush leading to progressive failures at the top
and base of the connection layer as the piles bend back and forth over long periods
of wind and wave conditions in the North Sea.
In the oil and gas sector it is common practice to use shear keys in grouted pile to
sleeve connections in the design of oil and gas platforms. The DNV J101 2007
design code left it open for designers not to use them, which is what they did,
because it is the cheaper, quicker option.
The original practice of not using shear keys, using a well-tested appropriate
underwater grout was developed as a means of grouting piles into bedrock at Danish
sites. However, the technique and the grouts used were subsequently adopted for
the grouting together of structural steel piles and transition pieces to wind turbine
towers. This was basically to save time and money and allow adjustment of pile out-
of-verticality using triple hydraulic jacking to level the turbine tower prior to grouting.
By contrast, the first offshore UK Round 1 Blyth and Scroby Sands projects used
bolted pre-fitted welded flange connections.. In deeper water, the Beatrice
demonstrator project (the DownVInD Project) selected a tried and tested oil and gas
technique known as pile swaging (Hydra-Lok® system) for the jackets. This secures
the structures by expanding the driven pile radially into a surrounding sleeve built into
the substructure, The swaging method is easily monitored and has the advantage of
being much quicker than alternative methods and does not involve grouting. A
solution not yet taken up by the industry is the much stiffer more robust “sandwich
tower”, involving the use of two steel tubes with concrete infilling, researched by
Prof. Peter Schaumann at the Leibniz University of Hannover, through an extensive
German federal government scheme known as “GROW”

Consultant
Tel. +32 10 41 95 25 Mobile: +44 755 4612888
April 2011 Page 2
While these are more expensive techniques, these solutions (Hydralok expansion,
sandwich tower and bolted flange connectors) are all structurally sound in the long-
term and not at risk of structural failure, if properly sized and designed by
experienced consultants.
The original DNV 2007 code did not explicitly require the use of shear keys led to the
decisions by designers on many monopile projects in the North Sea to go-ahead with
grouted connection designs which did not include shear keys, with a small number of
exceptions. These projects have now all exhibited grout cracking/crushing and
settlement problems necessitating expensive repairs.
Unlike oil and gas platforms that sit on the seabed due to their sheer dead weight,
wind turbines are relatively light and experience a lot of seabed cyclic bending loads
or moments over their design life in excess of 20 years. This constant variable
bending has been shown to result in crushing of the grout at the top and bottom of
the connection. Subsequent progressive failure or “unzipping” and shear failure of
the grout annulus may have occurred, which causes the structure to drop and
probably tilt slightly. However, the full story of why and precisely how these joints
have failed has unfortunately not been made public. The grout is failing, but
ultimately, the failure is down to the design.
The fundamental philosophy of the use of unreinforced grout in this fashion must be
questioned and several outstanding technical question marks as to their long-term
integrity and behaviour remain. There was a mistaken attitude that this type of grout
would act as a sort of long term rigid “glue” sticking two overlapping tubular steel
members together. Alternative structural solutions should be investigated
immediately.
The fundamental problem is that very high compression strength grouts, when cured
after a short time period, set to the strength of a strong intact brittle fine-grained rock
(compression strength of 100 to 200 MPa). This is much stronger than normal
concrete.
In addition, during offshore installation of the transition pieces, there can be little
accurate knowledge of the actual thicknesses of the grout sheaths subsequently
injected into the pile-sleeve gap, or indeed of the variation in thickness of the grout
annulus itself.
This is a fairly critical parameter required in analysing the capacity of any grouted
connection, with or without shear keys. A wind turbine subject to cyclic wave and
wind loads over a 20-25 year design life, being held by a hollow cylinder of brittle
rock of unknown and variable thickness sandwiched between two steel tubes, is not
a good situation.
The only monopile foundations that appear not to have settled/failed are those that
used shear keys. The industry should be looking at steel-to-steel connections using
flanges and bolts, the old-fashioned way. But it’s all been down to cheapness and
speed thus far.
The “spin” being applied over the last 12 months to the effect that this is an
“unforeseen” design problem which could not have been foreseen is somewhat

Consultant
Tel. +32 10 41 95 25 Mobile: +44 755 4612888
April 2011 Page 3
disingenuous. In addition, the statements being made to the effect that there is little
background information available to assess grouted connection design with shear
keys for large diameter piles is dubious.
There is a wealth of information available published from past oil and gas industry
research going back to the 1970’s and a DNV 2004 publication on fatigue testing of
grouted connections is freely available in the public domain.
DNV's latest solution as reported in January may fall short of solving the problem
with monopile foundations.The problems with the conical design are that the
philosophy behind it seems to be to allow or even expect grout failure, and for the
conical configuration geometry to then permit the joint to recompress to some sort of
stable condition. This is fundamentally poor engineering and would probably not be
permitted say, in the reinforced concrete design of multi-storey structures or major
bridges for example.
It is accepted that these large diameter monopoles behave essentially as thin-walled
shells that can distort substantially compared to smaller diameter thicker walled
rotationally and laterally stiffer offshore oil and gas platform piles.
In addition, whereas the load on offshore oil and gas platform foundation piles is
predominantly axial, wind turbine towers are lightweight but exert much larger
seabed level shear force and bending moment upon large diameter thin walled
monopiles.
A brittle grout connection is unlikely to behave well in a long term fatigue context
under this loading regime, and it is almost certainly the case that in hindsight grouted
pile to TP connections should never have been adopted and certified in the first
place. It is a systemic error, which needs to be stopped, regardless of the supposed
short term cost savings being touted around.
The actual geometry of any grout annulus, due to uncertainties about pile installation
and the structural fabrication tolerances of the piles and sleeves themselves presents
an issue. Piles cannot be consistently installed in the seabed completely vertically,
and are usually allowed a tolerance of up to 1 degree. This, combined with the
uncertainty on installation in offshore conditions of the TP over the pile, suggests that
the actual final thickness variation and average width of each and every individual
grout sheath has been and will be unknown in reality.
There are a considerable number of other uncertainties yet to be fully addressed in
my opinion, which are implicit within the recent DNV January 2011 Report which
aims at providing a solution to these problems and whose recommendations (conical
configuration) have been taken up by several projects, including London Array,
Walney 2, including:
 Geochemical behaviour of grout in the long term in seawater conditions.
 The shrinkage properties of the grout and the long term fatigue capacity and
coherence under water saturated conditions.

Consultant
Tel. +32 10 41 95 25 Mobile: +44 755 4612888
April 2011 Page 4
 The degree to which provision of a conical section reduces the likelihood of
bending ovalisation on thin walled monopiles due to variations in bending
moment magnitude and direction leading to loss of bond contact.
 How it is that provision of a conical section improves the degree of “tolerance
variation” (surface irregularities and steel roughness), observed as being
critical to the capacity of non shear-keyed connections.
 How conical configuration reduces the likelihood of grout crushing and
progressive unzipping failure along the connection at the top and bottom of
the connection itself.
 Uncertainty about the degree of abrasive wear along the contact surfaces.
 The durability of grout seals in terms of long-term sealing and water-tightness.
The grout seals often adopted are those originally designed for onshore bored
piles and are known to have been unreliable (numerous blow outs during
grouting on Greater Gabbard)
 The fact that DNV and other authors have previously stated that Finite
Element Numerical Modeling is not suitable in modeling the long term cyclic
load fatigue behavior of such high strength brittle materials. The behaviour is
essentially a rock mechanics problem.
 The fact that DNV recommend continuing monitoring of settlement to check
against friction and wear, indicating that potential movement and settlement
of such joints is more or less expected. This “acceptance of future failure”
design method should not normally be acceptable, in view of the uncertainties
on how the tower will behave when cracking/crushing and progressive
disintegration of the joint commences and the tower settles and shifts
downwards. How the conical section recompresses the grout annulus in these
conditions and therefore retains some sort of stability for the long term seems
to be completely unknown.
The best solution for monopiles may be to abandon grouted pile to TP connections
completely and devise a working bolted flange connection or similar, possibly in
conjunction with “in-tower” seismic style dampers, the hydralok system adapted to
larger diameter piles, or the “Sandwich” type composite pile.
In lieu of this, a solution could be the adoption of a configuration where the structural
installation and fabrication (ovaility chacks, out-of roundness, welding) tolerances are
sufficiently well defined and tightened up, the annulus thickness is better known and
shear keys are shown to function properly at correct spacings, possibly also adopting
vertical reinforcement orthogonal to the circumferential weld beads or stiffeners.
In addition there could be further development of an accepted and proven spring-
loaded support system, as used at Statoil’s Sheringham Shoal site, or the pinning
solution currently being proposed by Proserv Offshore.
Above all, data and information from each project where there have been problems
to date should be made freely available to all parties, instead of the current situation
where only limited information is available. This may require national government or

Consultant
Tel. +32 10 41 95 25 Mobile: +44 755 4612888
April 2011 Page 5
EU intervention. What is required is an independent assessment driven forward by
national governments in the UK (Crown Estate), Germany (BSH) and Denmark, as
the current major national European players, possibly with support from the EC and
the EWEA.
The large diameter monopile foundations currently proposed for the London Array,
Walney and Cape Wind US projects are still likely to incur similar problems due to
the absence of shear keys (London Array and Walney) and the use of the same type
and strength of grout and methods of installation, despite the change to a slight
inclination between the pile and the overlapping transition sleeve (the so-called
conical section).
The consequences of further grout cracking/crushing, settlement and tower tilt
failures on wind farms entering or already in a construction phase, where conical
tapered grouted connections have been or are being adopted, some with shear
keysand some without, are too serious to contemplate.
Monopiles drilled and grouted into bedrock and bolted flange connected towers could
still be acceptable in up to 20 to 25 m water depth (WD). However above that WD or
where soil cover is deep (> 15 to 20 m),monotowers on a single suction caisson,
which do not require a grouted connection, could be acceptable to perhaps 30 mWD
or more under the right geotechnical conditions.
Beyond 30m WD the optimal solution could be a tripod (Alpha Ventus style) founded
on either drilled and grouted mini-piles or tension anchors into shallow bedrock, or
suction caissons in a group of three, either pre-installed with a template or installed
after fixing to the tripod legs in port and floated out. Tripods are preferable to 4-
legged jackets. Concrete gravity base structures (GBS) may be possible in some
cases if environmental, scour and cost conditions permit. The above solutions
(monotower, tripod on suction caissons and GBS) have all been extensively studied
as part of the UK Carbon Trust Wind Accelerator Project. The use of mono and triple
suction caissons is now viable in most soil conditions, provided critical scour
protection in the form of frond mats or mattresses is installed over the correct area of
influence.

�ݺ�ߣ

Golightly_Monopiles

More Related Content

Golightly_Monopiles