LargeScaleOffshoreStaticPileTests-Practicality andBenefits P Barbosa IberdrolaRenovables OfshoreDeutschland Berlin Germany M Geduhn Ramboll Hamburg Germany RJ Jardine Imperial College London UK FC Schroeder Geotechnical Consulting Group London UK Abstract Iberdrola is developing the 350MW Wikinger offshore wind farm in the German Baltic Sea where ground conditions are dominated by Glacial Tills and Chalk. A review of current pile design methods highlighted significant design uncertainties that could lead to unnecessarily conservative pile dimensions for the seventy four-legged jacket wind-turbine support structures and one six-legged offshore substation. To address these concerns the project missioned in advance of final design offshore dynamic and fully autonomous static pile load tests that were pleted 10 weeks after driving on 1.37m diameter piles with penetrations of up to 31m in water depths of around 40m. This paper provides an overview of the testing and its practicalities as well as outlining the project risks and opportunities that justified the field testing programme. Lessons learned are sharedand conclusons are drawn regarding pile test feasibility design and planingCost-benefit analysi demonstrates the potential value of offshore static testing in future offshore developments that face para- bly difficult ground conditions. 1. Introduction advance dynamic and static offshore tension tests in Iberdrola is developing the Wikinger Offshore two phases. Phase I involved installing nine piles at Windfarm (OWF) in the German Baltic Sea three test locations giving a representative spread of halfway between the German island of Rugen andthe OwF's ground conditions.Pile driving the Danish island of Bornholm.Seventy Wind monitoring and Dynamic Load Testing (DLT) was Turbine Generators (WTG) and one Offshore performed on six piles at the End of Driving (EoD). Substation (OSS) have now been installed in water Intermediate DLTs were alsoincorporated after depths between 35m and 42m giving a 350MW short driving pauses necessitated by the test total installed capacity. arrangements. The WTGs are supported by four legged jackets A rest period of 10 weeks followed pile installation founded on open steel piles with outside diameters to allow excess pore pressures to dissipate of 2.7m (106") while the OSS relies on a six legged (particularly in the Tills) and soil set-up to progress layers consist of Glacial Till over Chalk which in maximum possible within the project deadlines. It terms of the Lord et al. (2002) classification system also matched the time that could be safely expected varies from Grade D (structure-less low density) to to elapse between pile driving and WTG installation. Grade A1/A2 (structured low to medium density). Further details on the range of soil conditions Phase II of the advance field trials involved Static pile locations are provided by Barbosa et al. (2015). driven at the three locations followed by a re-strike DLT on an identical adjacent test pile. The test piles' The pile loading is predominantly axial and shaft details were as follows: resistance governs design.A review of current design methods and related onshore pile tests ●outside diameter 1.37m downscaled by 50% (Barbosa et al. 2015) highlighted significant diameter from the production piles to reduce uncertainties and led to the decision to conduct load test equipment size and costs;
wall thickness of 40mm identical to the stage. This process led to a petitive tender that production piles to ensure similar remouldedallowed an extensive offshore pile test campaign at a chalk interface thicknesses (see Muir Wood cost that was significantly lower than the adjusted et al. 2015); and risk costs discussed above. The offshore testing penetrations of 16.8m to 31m that were campaign created opportunities to: representative of the production piles considering a conservative estimate of set-up optimize pile design and save pile steel; and safety factors allowable after pile testing. validate the DLT equipment and procedures 2. Risks and Opportunities to be used later during construction; and 2.1 Risks inform the selection of noise mitigation The project's interest in conducting offshore pile systems to minimise pile driving impact on tests was motivated by risks perceived around the marine mammals. design consent process.Difficulties in establishing a Optimized pile design was expected to generate parable onshore test site for another Baltic Sea significant steel savings because of the bination Chalk OWF had indicated a high risk of delays to of less stringent design factors being required and design consent and consequently the start of the possibility of improved design soil parameters construction (Knight 2012). The risk associated being justified especially for the Chalk. Site specific with delays to construction and first power were pile testing allowed the partial factors used in design quantified and resulted in a significant loss of to decrease from 1.4 to 1.1 in pression and 1.5 to project present value and other potential long term 1.15 in tension (DIN 1054 2010-12). However losses such as reduced energy tariffs. during the final design it was found that reducing the partial factors alone did not yield practical benefits Design consent assurance was critical to the as the associated benefit was counterbalanced by an project's E1.4bn Final Investment Decision (FID) increase in expected cyclic degradation during the which had to be made just before the major supply design storm. Nonetheless the test results justified and installation contracts could be signed. However considerably improved design soil parameters the 2"d and 3d releases from the German BSH allowing large reductions in pile steel. regulatory body required to start construction are generally granted at a much later stage. Therefore A further requirement by BSH (2011) is that DLTs developers have to proceed at risk in the intervening are performed during construction on at least 10% of period between FID and BSH releases. Any design the WTG locations to demonstrate sufficient pile changes or conditions imposed at the release stage capacity. This requirement is particularly onerous are likely to result in variations requiring scope for developers as the standard DLT interpretations programme andcost re-negotiations with pile do not recognize any contribution to pile capacity supply fabrication and installation contractors. from set-up after driving. In media such as Chalk Restricted vesselavailability and installation that can develop marked set-up EoD DLTs can give windows may also result in longer weather highly conservative indicators of operational pile downtimes and further project costs and delays. capacity. However if DLTs are to be conducted at later dates it is necessary to recover instrument The adjusted risk cost which considered both event cables to deck for sealing and conditioning before probabilities and the possible consequences of returning them to the seabed employing systems that problems with BSH releases was conservatively ensure safe storage easy retrieval and reliable valued in the low 10s of EM. An offshore test operation at the re-strike date. It is also necessary to programme at a cost below this risk cost was remobilise vessels equipment and personnel to carry therefore economically attractive. out the re-strikes and DLT measurements. In this respect it is important on Health & Safety grounds to 2.3 Opportunities undertake if possible the operations without divers. A tender process was launched in 2013 to evaluate the appetite for technical feasibility and costs of Additional benefits were taken from the testing offshore pile testing. Several submissions were campaign by applying experience gained from the received detailing various feasible technical Phase I driving monitoring and the Phase II testing solutions. The proposals were carefully reviewed to the procedures applied during the DLTs of the and technicaldetails discussed with potential production piling operations.Noise mitigation contractors with the aim of ensuring test quality and 1system monitoring during the trials also informed peo ue e siend euod Sueanu pue Supoe the choice of noise mitigation systems for
construction. Noise emissions during test pile greater than expected and the 15MN rig capacity installation were considerably higher than expected proved insufficient to obtain full failure at the two and very close to the threshold limit imposed by the Chalk dominated locations. Extrapolation of the pile German authorities. The noise emission prediction creep trends observed under maintained load stages models were therefore revised and it was concluded and correlation with the successful parallel dynamic that more effective mitigation systems would be re-strike tests allowed the tension capacities to be required during construction.The noise emission estimatedfor these cases withreasonable monitoring trials allowed for the proper planning confidence. avoided costly construction stops. and design of the final mitigation systems and so Three different pile testing systems were considered during the tendering process:(i) floating 3.Pile Test Campaign vessel/barge solutions (ii) jack-up rigs and (i) fully 3.1Selection and designof pile testing systems autonomous seabed systems. The successful tenderer EA-Pfahle (2014) and DIN EN 1997-1 stipulate a (Bilfinger Construction GmbH) proposed the fully number of criteria for SLTs that exclude constant autonomous seabed systems indicated in Figure 2 rate of penetration procedures (CRP) and require a essentially consisting of a central SLT pile two maintained load (ML) approach. The selection and outer reaction piles (one of which was used for design of the pile testing system was dominated by DLTs during Phase II) and a reaction beam. Even in the SLTrequirements with the three most the relatively mild Baltic Sea conditions the main appropriate and important objectives being to: technical disadvantage of any floating solution was its susceptibility to variations in the verticality and 1. provide test data to allow the pile design to magnitude of the applied loads due to actions of meet the BSH and DIN requirements; currents waves and tides. While a jack-up rig would 2. define ultimate pile resistance and load provide a more stable platform for undertaking the displacement behaviour pun slow SLTs technical disadvantages included potential monotonic loading including the piles' short-term creep characteristics; and affecting the testing campaign and the later jacket 3.investigate the cyclic stiffness under a load installation. cycle imposed around the characteristic pile load and other load levels. hollow reaction piston jacks beam In order to satisfy these objectives and maximise the value of the SLTs the key technical issues addressed during the procurement process were: nchor 1.rules to guide the specification （during bars testing) of load increments to ensure load- displacement behaviour was captured without ambiguity and clear criteria set for pile the definition of the ultimate pile resistance; trial pile for static 2.accuracy and precision of pile displacement load test and load measurement systems; and tril plle for 3.stability and control of the applied pile load (EoD and re-strike) dynamic load test in the offshore environment. Figure 2: General arrangement for sea bed testing system. Designing the SLT system to achieve the technical Detailed discussions with the Authors led to a final objectives involved assessing the ultimate pile pile testing system that worked well. One limitation resistances for each test location. Underestimates was the load carrying capacity of the connection would lead to an inability to provide unambiguous system deployed to link the test pile to the loading definitions of ultimate pile resistances whereas assembly. The potential risk of an insufficient overestimates would lead to over-dimensioned and system capacity was raised at various stages but it potentially unaffordable testing systems. The final was not feasible within the project timescales and design requirement set for the load system was to resources to raise its rating. While a higher capacity apply a maximum load of 15MN; several times the would have proved valuable the static and one-way required design load estimated by the project team. cyclic testing offered good load control and high However the set-up developed in Chalk proved far resolution load and displacement measurements.
Recognising that the field testing programme would Blowcounts/0.25m have to be limited and could not address all of the 10 20 30 scientific questions raised concerning driven pile Blows behaviour in Chalk Iberdrola Imperial College and Energy GCG made a successful research grant application to Innovate UK and launched a Joint Industry Project (JIP) that involved close analysis of the Wikinger 10 offshore tests and new experiments at the St. 15 Nicholas site at Wade in Kent (Buckley et al. 2017). 3.2 Pile driving and Phase I testing 2 The pile testing campaign (Phases I and I) was 25 2015. The custom built pre-piling template shown in 30 Figure 3 ensured accurate positioning of the three test piles driven at each test location. 35 0 200 400 600 Hammer Energy [k] Figure 4: Driving records location 1 The pile capacities interpreted at the re-start of driving at the Chalk dominated test locations were 2.5 times those before each (5 to 25 minute) driving pause although these gains degraded rapidly after re-starting (see Figure 4). No such short term capacity gains were evident after the driving pause at the Glacial Till dominated test location. Following pile installation the 100m long dynamic pile monitoring instrument cables were recovered back to deck and conditioned for storage. The Figure 3: Pre-piling template recovery was managed by an ROV and personnel on deck.The cables were pulled below the pre-piling The piles were driven with a hydraulic Menck template and returned to the vessel so that the MHU800S hammer whose output was limited to template could be recovered without damaging any 500kJ during driving as contractually stipulated. A cable. On deck the cable connections were sealed in major pile installation challenge was to control the waterproof terminations and the cables reeled onto a hammer energy output to avoid excessive set per special drum that was then lowered to the seabed. Its blow. Resistance to driving was low with an average 8 blows required per 0.25m penetration. to facilitate later recovery. Installation pauses were planned for each pile to 3.3 Post driving operations and Phase II testing allow the pile guides to open and avoid damaging Once the Phase I work was pleted and all the the DLT sensors; driving halted for between 5 and cables safely secured the pre-piling template was 25 minutes when the piles were approximately 7m retrieved and modified for later service as a static from target depth The interruptions resulted in a pile displacement measurement reference system noticeable increase in the resistance to driving at the during Phase II of the testing campaign (see Figure Chalk dominated locations with a higher hammer 5). The pre-piling template top section was removed blowcount measured at re-start and over a limited and a displacement measurement set-up added into distance of I to 5m after as shown in Figure 4. the centre SLT pile position. This consisted of a guide cone a moveable ring and a subsea Buckley et al (2017) give further details of advanced stress wave matching analyses of the DLT data head movement measurements to be made at three undertaken as part of the associated JIP. The stress points relative to a remote datum. wave matches confirmed shaft resistances between 10 and 20kPa during continuous driving in Chalk.
was maintained for at least 30min and the full SLTs took between 13 and 18 hours to plete. 3.4Phase Il:StaticLoad Test results The SLT experiments were performed to meet a failure criterion defined by the semi-logarithmic creep rates developed under maintained load steps. The critical ‘failure’ rate was chosen as 4mm/log cycle of time which was adjusted up from the 2mm/log cycle of time specified in EA-Pfahle (2014) to take account of the relatively large test pile dimensions. Figure 5: Reference frame It was fundamentally important to have a stable load The moveable ring was positioned by lowering the system and ahigh resolution displacement reference frame's extended mudmats in a stepwise measurement system in order toaccurately procedure. The moveable ring finally rested on pre- determine the creep rates developed under each load installed brackets placed on the outside of the SLT step. The displacement measuring system was stable pile.The position of the moveable ring was to around 10μm and the hydraulic loading system confirmed by the displacements measured in the was also very steady limiting load variations to subsea extensometers. After placing the reference within ±40kN in tests that applied up to 15MN. frame the test frame shown in Figure 6 was lowered Overall the system was able to perform the three to the seabed. A guide cone facilitated the placement SLTs without any malfunction. Figure 7 presents the of the connecting and loading system for the centre loadversus time trends for the Glacial Till SLT pile. The test frame was rotated until the dominated location demonstrating the excellent load catcher plates contacted the reaction piles and then control achieved. Post-failure cyclic loading was lowered onto the reaction piles. With the frame in also applied successfully at this test location. position locking plates were hydraulically activated and latched into a steel ring inside the SLT pile. The 10 latching process was monitored through subsea cameras installed in the test frame. 15:00 18:00 21:00 00:00 03:00 06:00 (w) Figure 7: SLT load vs time for Glacial Till dominated location Figure 6: Load test beam and frame Static testing was then initiated. The test loads were each test with predefined 30 minute pause periods applied in tension by hydraulic actuators fed from the surface vessel and the loads were measured by (see Figure 7). Once a 2.6MN load had been reached (which corresponds to half the WTG characteristic subsea load cells. The SLT was conducted as a maintained load or creep test in accordance with tension load so accounting for the test pile scale EA-Pfahle (2014). Loading was applied through an factor) the pile was unloaded to the first load step array of jacks on the seabed frame that were fully before being reloading back to 2.6MN. From this programmable by the test operators. Each load step point onwards the magnitudes of the load steps were determined by the creep rates measured in the
大型近海静态桩试验-实用性和益处.pdf
