Scribd Logo
Upload

Welcome to Scribd!

  • Alm Hamre 2001 PDF

    Uploaded by

    David White
    7 views6 pages

    Original Title

    alm_hamre_2001.pdf

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    7 views6 pages

    Alm Hamre 2001 PDF

    Uploaded by

    David White

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    of 6
    Soil model for pile driveability predictions based on CPT interpretations Modellisation des sols pour la prediction du battage des piles basee sur I'interpretation des CPT “TAlm ~ Aker Technoegy, Co, Nonray Hamre - NOTEBY A'S. Cab, Nornay ABSTRACT: Through the later years, the authors have been involved with pile driveability predictions for lenge, open-ended ples in- stalled in different soil conditions in the North Sea, and a prediction model was presented in 1998, Since then, the database of back calculated records have been continuously updated, and in particular, information from installations of long ples in normally consoli- dated clays has been gathered. Based on this complete database, a new and improved soil model is presented in the paper. This soil model is directly corelated to CPT measurements hence avoiding biased individual interpretation of measured data. Comparisons between the back calculated and the post predicted soil resistances are shown for a large number of locations asa verification ofthe new model. RESUME: Durant ces démitres années les ingenicurs ont és confronts au probleme du batage des piles dans diftérents types de sol en mer du Nord. Un model pour prédie le comporemert des piles durant le battage a e® présente en 1998, Depuis, analyse des batiages efféctues a donné une base nouvelle pour en affine a prédction. En particulier les informations de Tnstalation de Tongues piles dans un sol agileux normalement consoldé ont &é assembles, Basé sur cette masse de données un nouvead model de Feprésentation des sols € développé et et présenté dans ex ance. Ce model ext directement corolé ad mesure CPT et par I Gite toute interpréation inividvelle des données. Le model a ei verifié en comparant In resistance des sols caleulée vee celle des batagesrelement ffectué et ela pour un grand nombre de oes. 1 INTRODUCTION In the continuous search for an optimum pile design in terms of ‘minimum cost, the offshore industry is designing piles with in- creasing diameters and penetrations in order to arrive at ma ‘mum pile capacities. Documentation of achievable penetration is performed through driveability predictions, and models of good reliability is thus of vital importance. Fatigue damage due to driving will, for many cases, govern the pile design, and reliable ‘models for driveability predictions are a requirement for realistic Fatigue calculations as wel, ‘Alm & Hamre (1998) have previously presented model for eiced to approximately 30 %. In the SRD predictions, the full pile tip area, not reduced for wedging, should be used in combi- nation withthe above relation. snc earaee oc ay p40 and actual resort 180 . g 2 ; i ; z o 02 o4 06 08 1 Friction Degradation Figure 2 Examples of degradation shapes for diferent sis “The total stati resistance is then calculated similarly to pile ‘pearing capacity principle, and is contributed by pile tip resis: raree and side faetion along the pile. For all cases, unplugged piles have been assumed during driving. 6. RESULTS OF POSTPREDICTIONS. “The results of the post-predictions are shown together with the back-calculated SRD profiles on figures 4 through 19 for each of the 18 diferent sets of installation data considered, These post- predictions are based on an interpreted average of a series of re~ Poaded CPT profiles at each site. In addition, a suggested upper ‘bound resistance is included, taken as a 25% increase of the pre dicted best estimate. in general it can be seen thet the described model gives very good correlation to the back-callated data. In average the pre~ rote best estimate curve lies very close to, or slightly above the Dack-caloulted ones. Also, it can be seen thatthe upper bound turve generally covers the range of the back-calculations, with Some exceptions. The majority of such exceptions are however felated to resistances occurring immediately after stops in the Grving sequence, which are due to hammer breakdowns or tops vhen followers are inserted, The following comments are made To the results where significant deviations to the model predic- tions seems to occur ‘At the typical oft clay sits, like Brage,Vesleikk and Ose- beng Out the model i seen to predict very wel. At Oseberg Ost fone pile is seen to lie above the upper bound model prediction. ‘This is believed t0 be caused by the sand layer being Locally ‘nore dense than used as basis. Jn this sand layer, it may also be ‘seen thatthe driving resistances vary considerably ‘At the clay sites with moderate over-consoliation tke Em- bia, and Sleipner Riser and Flare, and the upper 40 meters of Sleipner B, the model predicts generally very well. At Sieipner Riser and Flare however, the model slightly under-predits the resistance from 25 to 50 meters. Tis is believed tobe due to un- reliable hammer performance for some piles, as reported for the Sleipner Riser installation, At Sleipner B, CPT data was scarce in he upper part. and the actual data showed relatively large dis- Ccepancies 10 the laboratory test results. It is belioved that the CPT design profiles are t00 low, leading to the under-prediction shown, Heavy over-consolidated soils with very dense clay layers are encountered at Oseberg B and Gass, and at Foy, Huldra and Heimaal. For all these cases the model predicts excellent, except for one pile at shallow depths at Fray, and for two piles with very low resistance in one comer of the Heimeal Jacket, The Uiserepancies are at Fray is likely to be caused by low hammer performance, while at Heimdal, tis likely that the soil consi- tions in this comer of the platform is different than used in the basis forthe predictions. For the Oseberg B conductors, the model does to some ex tent over-predict the resistance throughout the profile to 50 me- ters dopth, This is however believed to be caused by the external Giving shoe mounted at the conductor tip, leading to reduced uiside ftion. After driving to 50 meters penetration, the inside Sol columa was dled out, and further dril-out using an under~ Tamer was performed to about 105 meters depth. In the mode] for prediction of SRD below 50 meters, initial friction values ‘was assumed for all layers above 5O meters, and a gradual fic~ tion degradation was included according to the presented mode ‘At Varg and Jota, thick layers of medium dense to dense sands are encountered, and for both locations, the model predicts very well. One may however observe that resistances generally ‘ary more rapidly than in typical clay sites, which is due tothe nature of variability in density normally found in sand layers One pile at Varg shows however significantly higher resistance than any model would predict, and this anomaly can not be €X- plained by locally high sand density. Hitting of @ boulder or par- tial plugging due to other reasons may however be plausible ex- planations. Mixed layers of over-consolidated clays and dense sands are encountered at Ekofisk and Oseberg Spr. Good correlation is found at both sites, but at Oseberg Ser, the variability is found to bo high, leading to under-pediction in some depth ranges. This js caused by variable haramer performance that occurred during the first face ofthe pile driving 7 RECOMMENDATIONS FOR USE. In summary, the model is shown to predict resistances lying clove to or slightly above the average of the back-calcuted val- ties, Additional conservatism is also included since optim hhammer performance has been assumed, leading to maximum backccalculated resistances. The model will therefore produce best estimate resistance profiles lying slightly on the conserva- tive side. The soil profiles to be used in predictions should be interpreted as characteristic profile, ic. average values on the conservative sie. For prediction of the upper bound resistances, the effect of soil variability should be taken into account. Ideally, a factor representing the local variability of the actual layer should be tatblished, but sufficient data to evaluate this are seldom avail- Sle. Based on the back-calculations, a factor of 1.25 is sug- tested, which has been earlier shown to cover the variability for ost cases within the average plus 2 standard deviations. “However, for some cases, and in particular for thick sand deposits at large depths, a higher factor may be found appropri- tate, and for critical cases, sensitivity studies with factors ebove 1.25 are thus recommended 1298 Veslefrikk Jacket so, an . 7 © © © Ww Oseborg Ost Jacket RO, MW ecenancn oop imp 0 s ent pm) Oseberg B Jacket Pil Sleipner Riser and Flare n © © 2 Sleipnor B Jacket aezeeagy Oseberg Gass Jacket om mo © © mw 3.8 & s 1300 8 Figures 411. Results of back- calculations and post-predictions Froy Jacket Hulda Jacket Hoimdal Jacket Varg Jackot Jotun Jacket ‘Ru ee ee) ‘0, i f° t $ [: ® 10 on x Jct nd Bae Oseberg Sar Jacket ‘Soper ‘sro. om E> @ 10 o 2 9 @ » Ww i eT oe i fA = i \ ; Figures 12-19 Results of back- : caleulaions and pos predictions ease 101 marmot peg) [me | te | fom feta |e [tae | a smn am | os fase] + | smnam | te |srsentestinwtiiertn so meme Joon] ow [as | a [sma | ve Saepbete [BB] Sf | MERE | S| ee mmsere ne io ae sipanenimeaas Scone |e] | SS | | "mum | Yo SEHGiss [Be] 2] 8 | | MER. | S| epee nentenie Sona see eee remraeraen |x| 4 | aio | + | snmnnnen| se | riemntepa tent mene [wx we fam] | sam | ve | semen font | on | a te | Songun nt Sms [sip] a]: se | Sumner ste «fale * woh fa | « ve | att seam |e] xe | w | te | cranntionint sa ole 2s Te3 Daina fe Nath Se pi ina f SUMMARY AND CONCLUSIONS ERENCES ‘The new soil model developed is proven to give very reliable predictions of SRD profiles for a variety of typical North Sea Soils. The SRD profile can be used to predict Blowcount curves bby means of the wave equation method, provided that damping coefficients and quake values are included as recommended ear lier by the authors. On average, the predicted resistances are slightly above the back calculated and thus slightly on the con- servative side. The model, without additional factors, is thus considered appropriate for prediction of best estimate curves. For prediction of upper bound resistance, the effect of soil variability should be taken into account. A factor of 1.25 is nor- ‘mally sufficient to cater for this. For critical cases, however, en- sitivity studies with factors above 1.25 are recommended. 9 NOMENCLATURE ile sce friction (kNin?) Ina ple side frtion, (Nin?) = Residal pile side fiction, Nin") Depth to actual cay layer, (m) Pile tip penetration (m) ‘Shape factor for degradstion, Unt pile tp resistance, Nin “Total cone ip resistane fom CPT, (kN?) Elective overburden pressure, QNim) © Reference pressure = 100 kN ‘Constant volume fiction ange (degrees) “Hlorizomtal stress ratio after deving (-) eS 10 ACKNOWLEDGMENTS . 6 4 o a P k are or Pr a 3 K "The installation data and relevant sil information forthe diffe ent locations included in this study have been released by the di {erent Oil Companies operating at each locaton. The authors ac- knowledges permission to publish these data from Statoil as ‘operator at Vesefrikk, Sleipner, and Huldra, Norsk Hydro Pro- uksjon a. as operator at Brage, Oscberg, Heimedal and Varg, Phillips Petroleum Company Norway, as operaior at Ekofisk and [Embla, and Elf petroleum Norge AS, as operator at Fray. ‘Alm, 7, Bye, A & Kvalstd, T. 1989. New Inexpretaton of Soil Resis- tance for Pile Driveabliy Analysis. Proceedings of 12th, ICSMFE, Rio de Jane. ‘Aim, 7 & Hamre, L. 1998, Soil mode for dxiveailiy predictions. Off ‘nore Technology Conference, OTC, Houston, OTC paper no. 8835. Collin, 1, Vergobbi,P. & Puech A.'1993, Friction Degradation and Sei-Up Bllects in Hard Clays Ofshore Congo and Angola. 25 th Oihore Technology Conference, OTC, OTC paper no. 7192. Cola, J, BoisardP, Puech, A. & Vergobi,P. 1996. An improved ile Driveabilty Model for Soft and Hard Clays fom Offshore Pile Diving Case Histories. Fifth Inemational Conference onthe Appi cation of Stress-Wave theory o Pies, Orlando, Goble, Rausche, Likins, & Associates, Inc. 1997. GRLWEAP - Wave ‘Equation Analysis of Pile Diving. Progeam Manual and updates. ‘Hecrema, EP. 1979. Relationships between wall fiction, displacement ‘velocity and horizontal sues in lay and in sand, fr ple deveabi ity analysis. Ground Engineering. Heerema, EP. 1980. Predicting pile driveability: Heather 22 an il: ‘ation of the “eton fatigue” theory. Ground Engineedng. ‘eerema, EP. 1981, Dynami point resistance in sand and in clay, foe pile driveability analysis. Ground Engineering. Jardine, KJ. & Chow, FC. 1996. New dosign methods for offshore piles, Report to Marie Technology Discorte, MTD, Publication 0, 96/103, Semple, R.M. & Gemeinhardt, JP. 1981. Suess history approach t0 ‘analysis of sol esstance to ple driving, Offshore Technology Con- ference. OTC pape no. 3969. Smith, EAL. 1960. Pile Diving Analysis by she Wave Equation. Jour nal of the Soil Mechtnic and Foundation Division, ASCE Vol. 86, Stevens, RS, Wilsi, EA. & Turton, TH. 1982. Evaluating ple drive- ability for hard clay, very dense sand and rock. Offshore Technology Conference, OTC paper no. £205. ‘Toolan, FE, and Fox, .A. 1977. Gotechnical planning of pled foun-

    You might also like