Download as pdf or txt
Download as pdf or txt
You are on page 1of 15

O

OTC 2294
49

T
The Bencchmarking of the New
N ISO for the Site-Speciific Asses
ssment o
of
M
Mobile Ja
ack-Up Units
JJohn Stiff, ABSG Consultin
ng, Inc., David
d Lewis, Lewis Engineering
g Group

C
Copyright 2012, Offshore Technology Confere
ence

T
This paper was prepare
ed for presentation at the Offshore Technolog
gy Conference held in Houston, Texas, USA , 30 April–3 May 2012 .

This paper was selected for presentation by an


T a OTC program comm mittee following review
w of information containned in an abstract subm mitted by the author(s)). Contents of the papeer have not been
re
eviewed by the Offshore Technology Confere ence and are subject to o correction by the autthor(s). The material dooes not necessarily re
eflect any position of the Offshore Technologyy Conference, its
officers, or members. Electronic
E reproductionn, distribution, or stora
age of any part of this paper without the wriitten consent of the O Offshore Technology C Conference is prohibite
ed. Permission to
re
eproduce in print is res
stricted to an abstract of
o not more than 300 words;
w illustrations mayy not be copied. The abbstract must contain coonspicuous acknowled dgment of OTC copyrig ght.

AAbstract
TThe paper pressents the logic behind, and th he results of, a series of commparison analyyses between a number of jaack-up site-
sspecific assessmments using bo oth the new ISO O 19905-1 on thet site-specifiic assessment oof mobile jack--up units [Ref.. 1] and the
SSociety of Nav val Architects and
a Marine En ngineers Techn nical & Researrch Bulletin 5-5A [Ref. 2]. T The study show ws that the
ddifferences in the
t results of analyses
a betweeen those undeertaken using 55-5A and ISO 19905-1 are ggenerally small, although
oone area of po otentially larger difference, which was in ntentionally exxcluded from the comparisoon, (the norm mally lower
kkinematics redu uction factor in
n 19905) could d lead to lower wave loads in 19905 calculattions.
ISO 19905 isi largely based d on 5-5A, how wever, the ISO O development incorporated ssome major chaanges, includinng:
 Enhan ncements to callculation metho ods
 Signifficant changes in the geoteechnical calcullations, includding the methhod of calculaating acceptabble bearing
capaciity, incorporatiion of deep pen netration cases,, distinction beetween backfilll and backflow
w, etc.
 Chang ges to enhance compatibility with other ISO O documents inn the 19900 serries, particularrly ISO 19902 addressing
Fixed Offshore Strucctures [Ref. 3] (e.g. the structtural code checcks for tubular members)
 Introdduction of addittional structuraal member checcks
 Re-strructuring of 5-55A into a moree logical ISO fo ormat
 Introdduction of “ben nign” changes required for ISO compatibillity (e.g. introdduction of appparent and intrrinsic wave
periodds), although soome proved to be less benign than others
 Additiional loadcasess for jack-ups operating
o closee to resonance
Given so many
m changes, there was concern that the ISO I may not bbe useable (e.gg. there would be gaps, inconnsistencies,
aand serious traansposition errrors), and eveen if an analy ysis could be undertaken ussing it, the annalysis results would be
ssignificantly different from a similar assessm ment using 5-55A.
To ensure the compatibility and acceeptability of ISO I assessmennts results, a two part bennchmarking prrocess was
uundertaken:
Part 1 Con ntract a single consultant
c to ensure
e the commplete ISO couuld be used to pproduce an ansswer. In effecct, could an
aanalysts start an
n analysis usinng the documeent and arrive at a a solution - regardless of w whether the soolution was corrrect. This
wwork resulted inn some importtant changes.
Part 2 Con ntract four con nsultants to asssess four jack k-up designs, and compare the results to a comparablee series of
aassessments baased on SNAM ME Bulletin 5-5A. In most caases each jack-uup was assesseed by two conssultants with thhe analyses
aaligned at specific points. Alignment allow wed differencess in ISO interppretation to be identified, andd ensured that results did
nnot diverge as they
t progressed d. A detailed sample
s calculation “Go-By” w was also produuced.
The results highlighted differences
d bettween 5-5A an nd ISO 1990005, and also soome areas of interpretation difference.
TThere were a number
n of diffficulties by co
onsultants align ning at the reqquired alignmeent points. Foortunately the ddifferences
bbetween consulltants were gen nerally not thatt great that thiss caused probleems, but it hass made comparrisons betweenn the results
2 OTC 22949

considerably more complex. A number of corrections were made to the ISO prior to its final acceptance, and a considerable
number of typographical errors were discovered. The studies brought out a number of areas where it was acceptable to use a
more favorable interpretation from ISO, and where new calculation methods had not been fully incorporated into 5-5A. It
also showed that some of the required changes in the ISO were less benign than had been originally thought (e.g. the effects
of intrinsic and apparent wave period). While the consultants often found it difficult to follow the new ISO, a few E-mail
exchanges normally sufficed in correcting problems, and when comparable analysis routes were taken through 5-5A and ISO,
similar results emerged. Given the large number of changes included in the ISO, this was a successful outcome.
One final thought on the use of the ISO is that any consultant that has used 5-5A in the past will likely be able to use the
ISO without having to take on significant new technology, but it should not be assumed that the transition from 5-5A to ISO
will be easy. There are a large number of small differences, and if the analyst is not extremely careful, these can be lost in a
“scan” reading of the ISO. The ISO needs to be read carefully and used in its entirety. In almost all cases the differences
between 5-5A are intended, and sometimes they can be very subtle.

Background to the Benchmarking


The history of the development of SNAME 5-5A, and thence ISO 19905-1, and the companion Technical Report, ISO
19905-2 [Ref. 4], is long, starting in the mid-to-late 1980 and continuing through to present. It has been described in a
number of papers [Refs. 5, 6 & 7], including one in this particular OTC Session [Ref. 8], and a repeat description is beyond
the scope of this paper. What is pertinent is some of the concerns that arose soon after the original publication of 5-5A. 5-5A
had been developed, with joint industry funding, through the efforts of a Working Group representing all relevant aspects of
the jack-up community (including drilling contractors, operators, designers, classification societies, and marine warranty
surveyors) over a period of approximately 6 years. When it was published by SNAME in 1994, it came into much more
common use; up to this point use had been limited, and on a trial basis. It was after publication that some concerns about the
conservatism of the document came to light, and work started to fund additional studies to improve the document, and reduce
the conservatism, largely under the aegis of the IADC Jack-Up Committee. Because of these early issues with 5-5A, there
was concern, particularly among the drilling contractors, that the conversion of 5-5A to ISO 19905 would result in similar
increases in conservatism, particularly as there were some relatively strict requirements that 19905 be consistent with other
documents in the 19900 series, and particularly the fixed structure ISO [Ref. 3].
The idea of benchmarking the Jack-Up ISO was mooted early on in the development process, and it was even suggested
that there should not be a Draft International Standard (DIS) issued before full benchmarking had been completed. It became
clear that this would not result in the best use of funds, and could lead to some complicated delays, so it was agreed to stagger
the benchmarking. Independent of the overall benchmarking, there was concern that Clause 12, on the structural member
checks, had changed so much from the 5-5A version that there was merit in undertaking a form of benchmarking for this
clause by itself. Table 1 gives an overview of the timeline for the benchmarking process. It can be seen that there were some
significant delays during the process. In most cases these were due to arranging funding. It should not be inferred that
funding was not available, but it was difficult to get the formal commitments to funding that were necessary to commence
work - there are few consultants who will start work on an assurance of funding from someone who is commissioning the
work, but is only handling the technical, not financial end of the situation.

Item Date Comments


th
Clause 12 Structural Member Check “usability” Request 28 September 2004 No clear funding available, but hoped to obtain from
for Quote (RFQ) sent to two consultants OGP.
Need for complete ISO 19905-1 benchmarking RFQ January 2007
identified, and work started on first draft
th
First meeting of Benchmarking Panel 30 August 2007 Panel members were suggested by the Editing
Review Panel based on experience and
representation of the funding companies. First task
to review the RFQ.
th
Phase 1 RFQ sent out to seven consultants 5 September 2007 Two bids were received
th
Clause 12 “usability” work awarded to Noble Denton 27 September 2007 Funding was from Shell ($17,500)
th
Clause 12 “usability” report received 11 January 2008 Generally found to be usable, and slightly less
nd
(Rev 1 22 January) conservative than SNAME
th
Phase 1 awarded to Noble Denton 10 December 2007 Funding ($70,000)was split between IADC, Shell,
and UK HSE
th
Phase 1 Draft report received 10 December 2008
th
Phase 1 final report received 16 April 2009
th
Phase 2 RFQ sent to 13 consultants 24 June 2009 Work to be based on the DIS and SNAME Rev 3
(both yet to be published)
OTC 22949 3

Item Date Comments


th
Draft International Standard (DIS) of 19905-1 Issued 11 September 2009
th
Phase 2 consultants notified of intent that they undertake 11 September 2009 Asked not to start work as funding had not been
the work finalized
th
Phase 2 “Early Start” Consultant (MSC) asked to start 5 November 2009 There were still some funding issues, but sufficient
work were available for a start of the work.
th
Other three consultants given official request to start work 17 December 2009 Funding for this work was supplied by IADC Jack-up
(Bennett and Assoc., GL Noble Denton, Global Maritime) Committee (50%) and ExxonMobil and Shell (25%
each) - Total $320,000
th
Deep Penetration added to SNAME for use in Phase 2 12 May 2010 A number of issues had come to light and were
Benchmarking discussed at the WG7 meeting during OTC
st
Phase 2 results presented at WG7 meeting in Hamburg 21 October 2010
th
RFQ for additional geotechnical cases 26 October 2010 This was to correct a problem that was found during
the benchmarking, and discussed during the Panel 4
meeting in Hamburg.
st
Phase 2 reports required deliver date 21 November 2010
th
Additional geotechnical cases delivered 30 November 2010
th
Final Draft International Standard (FDIS) issued to ISO 8 February 2011
th
“2-week review” copy of FDIS received from ISO 7 April 2011
th
“2-week review” comments sent to ISO 13 May 2011
th
Follow-ups (mostly ignored) 30 May 2011
th
Final Draft International Standard (FDIS) issued by 7 July 2011
ISO
Additional benchmarking funding obtained from Chevron July 2011 These funds were in part used to complete updates
needed because of the changes to the standard
between the DIS and FDIS stages. Will also be
used for development of a Benchmarking Summary
Report.
Proposal sought for benchmarking of seismic clauses in September 2011 This work had not started at time of writing as
19905 funding had not been secured.

Table 1 Timeline of Events in the Benchmarking of ISO 19905-1

Structural Member Checking Benchmarking


As stated above, the structural member checks had been completely rewritten since their publication in 5-5A, as described in
[Ref. 9]. For consistency reasons, the checks on tubular members had to based on the same approach as those in the fixed
structure ISO 19902, but because jack-ups have prismatic leg members comprised of different yield strength steels, all the
structural checks in 19905-1 had to be founded on a load based capacity rather than an equivalent stress. While this was
consistent with 5-5A, it was not consistent with 19902. In addition, the structure of the clause had completely changed since
5-5A, with many additional prismatic member checks added to the ensure completeness. A proposal was developed in late
2004 to get a consultant to use the newly formatted Clause 12 and, by running through a test example, to determine if the
clause was actually usable. For various reasons this work was delayed, but a contract for a modified and expanded scope of
work was finally awarded to Noble Denton in September 2007, a full 3 years after the original scope of work was written.
The original scope had been for a simple run through of the structural checks to ensure that Clause 12 was usable, but the
expanded scope entailed a simplified comparison to a 5-5A analysis, and more parametric assessments involving 3 different
jack-up designs and chord types.
The study took 3½ months to complete, and was funded by Shell. One of the requirements for the work was that the
analyst be an experienced engineer, but with little experience analyzing jack-ups, supervised by an experienced jack-up
analyst who could quickly help answer questions as they arose. The intent was that the analyst would then follow what was
actually written in the standard rather than “reading between the lines”, as an experienced jack-up analyst would be inclined
to do. The analyst was also to keep a log of all questions that needed to be resolved by their supervisor. This log was then to
be supplied to Panel 10, responsible for the development of Clause 12, so that they would have the necessary information to
improve clarity. One other requirement was that the jack-up legs should be checked for loads close to their capacity, thereby
checking the standard at its most critical area. To achieve this, Noble Denton used locations that, from previous analyses,
were known to be close to the structural limits based on a 5-5A analysis. They then ratioed the loads until the 5-5A checks
very close to unity. These loads were then, in turn, used to assess the legs to the ISO standard. The plan was not to
undertake detailed leg strength analyses, but to use the new standard and compare, as simply as reasonably possible, the
member strength check results with those obtained from 5-5A analyses.
4 OTC 22949

The first finding, and in many respects the most important, was that not only was Clause 12 usable, but it produced the
expected results. There were some areas where it was known that 19905 would be less conservative than 5-5A, particularly
in the use of the column curve for prismatic members, see [Ref. 9]. Under certain specific combined axial and compression
conditions, the utilization checks for prismatic members using the ISO were as low as 67% of those developed through use of
5-5A. There was also a general reduction in utilizations for tubular members, but only about 5%. Some of the other findings
were:
 The screening equation included for lateral torsional buckling (LTB) was too course, leading to too many cases
where LTB had to be formally checked. This was later addressed by introducing an additional screening check that
could be used for all closed sections (as the vast majority of jack-up leg chords are)
 There were a number of places where the slenderness limits on equations resulted in step changes in strength, clearly
an unacceptable situation. Some of these were typographical errors, and in all cases they were later resolved.
 A very imaginative interpretation of a reinforcing plate provision allowed the analyst to avoid having to classify a
particular chord component as “slender”, thereby avoiding having to use the rather complex slender component
equations. The serious aspect of this “convenient” interpretation was that it clearly brought out how carefully the
document wording had to be crafted in order to avoid misinterpretation (either intentional or unintentional).
 There were a large number of cross reference errors and instances of insufficient clarity
 There were a number of inconsistencies in the slenderness limits on tubular members that needed to be resolved.
Overall, this first step into benchmarking had brought out how important it was to have an analyst run through the
standard using it for a “real” analysis, and how difficult it was to ensure that all the details in the standard are correct. It
proved a very valuable lesson, but also showed that the ISO development was on the right track.

Benchmarking Panel
Work started in early 2007 on the development of a scope of work for benchmarking the entire jack-up ISO 19905-1
standard. The outline of how the benchmarking was to be structured was presented at a number of different times (e.g. IADC
Jack-Up Committee meeting in January 2007, and at WG 7 meeting during OTC in 2007). There were also discussions as to
the expected cost of the benchmarking, and how it would be funded. By August 2007 there was a solid foundation for the
scope of work, and the funding was informally in place. As discussed below, the work was to be split into phases, with
Phase 1 funding coming from the IADC Jack-Up Committee and Shell. Phase 2 funding was to be from ExxonMobil and
Shell each paying 25% of the cost with the IADC Jack-Up Committee paying the other 50%. In August 2007 a
Benchmarking Panel was formed to represent the funding organizations, supplemented by others who had been actively
involved in writing the draft scope of work. The Benchmarking Panel consists of:

Panel Member Company Affiliation Panel Member Company Affiliation


Jim Brekke Transocean John Stiff (Chair) ABS Consulting
Rupert Hunt Shell Pao-Lin Tan ABS
Dave Lewis (no bid review) Lewis Engineering Group Ward Turner ExxonMobil
Yi Li Diamond Offshore Paul Versowsky (in 2011) Chevron

Table 2 Members of the Benchmarking Panel

The first order of business for the Panel was to formally review and finalize the developed Benchmarking scope of work,
and to decide on how to proceed. (While it will be seen from the timeline in Table 1 that the structural member checking
work was being started around this time, there was very little involvement of the Benchmarking Panel as this was work
funded directly by Shell, and most of the protocols for its completion were established before the Panel’s formation.)

Benchmarking Scope of Work


Early on it became apparent that the benchmarking would need to be carried out in phases. While there was a time pressure
to complete the work, if a formal quantified benchmarking of the standard was started too early using text that was still
evolving, it was likely to come across problems that would need to be resolved, so would neither run smoothly, nor produce a
good set of complete results. It was therefore decided to split the benchmarking into a number of discrete phases that can be
summarized as:
Phase 1 Completeness Check: A complete run through the entire standard to check that the document was complete
and adequately explained to arrive at a single solution for any given particular course through the document. (There are
different analysis methods accepted within 19905-1, but each should arrive at a unique, not multiple, solution.) This run
through was not designed to check the validity of the results, however, any inferences that could be drawn from the analysis
would be used in better defining the phase 2 scope
OTC 22949 5

Phase 2 Comparison to 5-5A: Complete run through the entire standard to ensure that the results obtained were in
reasonable compliance with results from a similar analysis following 5-5A. It was anticipated that there would be some
differences between the results from the two documents since there have been some real changes to the document, however,
it was not anticipated that these changes would cause major changes in results. However, if major differences were found,
then it would be important to identify the source, and quantify the magnitude of the changes. It was anticipated that there
could be significant differences between the results obtained by different consultants, particularly if they used different
analysis methods (albeit all contained within the ISO). Notwithstanding this, it was hoped that the final results would be
comparable after they had been aligned at the various alignment points (as discussed below).
Phase 3 “All Routes” Check: Complete run through of all acceptable routes within the standard to ensure that there are
no inconsistencies within some of the less often used aspects of the document. Many parts of the standard will be rarely
used, and may not be properly checked in Phase 2. There was a desire to formally check these prior to release of the
standard.
Each phase of the benchmarking is described in greater detail below.

Phase 1 Completeness Check


The first task was to undertake a completeness and usability check to ensure that ISO 19905-1 was a workable document.
This entailed taking a limited number of cases and running each through the entire document using the most commonly
applicable analysis methods. Only the main routes through the document were to be used, and alternate routes were
discouraged, unless there is good justification for using them. The check included investigation of:
 Truss leg unit (although any inferences that can be drawn about tubular leg units would be valuable).
 Single water depth and metocean condition combination (but one that gives representative dynamic response)
 Pinned foundation and foundations including fixity when assessing dynamics
 Sand and Clay (normally consolidated Gulf of Mexico type clay) foundations
 Simple foundation yield surface assessment
 Preliminary development of a “Go-By” document
Because the study was based on a document that was still being edited, it was important that it be aimed at completing the
analysis rather than getting too involved in detailed analysis of all stages. In effect, if an error or omission was found, then
the analysts should move forward, and not get bogged down in trying to resolve all of the issues. Conversely, if there were
simple assumptions or modifications that could be made to complete an analysis, then it was considered reasonable that these
be made in order to improve the quality of the results output. It was imperative that any assumptions made were fully
documented.
As with the structural checks, it was envisaged that the work would be undertaken by a competent engineer, but with
limited jack-up experience (junior engineer), overseen by a project manager who was experienced with jack-up site
assessment, and preferably both 5-5A and ISO documents (PM). The PM was permitted to provide instructions to the
Engineer and could clarify some of the input information, but it was extremely important that any instructions or questions
addressed by the PM were documented in a Questions Register.
Since it was important that the Standard was complete, the analysis was to follow 19905-1 through all the basic steps that
would be taken in a normal site-specific assessment. Having chosen a suitable jack-up and site, the analyst was to develop
the needed metocean data, foundation parameters, rig loading, geometry checks, calculate wind and hydrodynamic
coefficients, ascertain leg penetrations, calculate spudcan fixity, develop the loads and responses, and calculate the various
utilization factors. As much as possible the analyst was to use standard tools that are publically available, and not rely on
“in-house” software. There was also a request that the analyst assess the potential for additional settlement by arbitrarily
increasing the leg loads to overload the foundation. Again, the intent was to test the standard, not to generate realistic results.
The results were to be presented in an extremely detailed report that set out exactly what was done, all the problems
encountered, and any assumptions that had to be made because the standard was incomplete or inconsistent. There was also a
request that the consultant develop a scope of work for the Phase 2 Benchmarking. Most of this had already been done to a
high level, and was included in the Phase 1 scope of work, but it was realized by the Benchmarking Panel that they did not
have the time to develop the Phase 2 scope to the detailed level needed to allow reasonable bidding. The Phase 2 scope was
also to include empty tables to be completed during a Phase 2 assessment, thereby allowing consistent reporting between
different analysts. It was realized that this scope would not be exactly what was needed, but by having its development as
part of the Phase 1 study, there was an increased likelihood that it would be delivered to prospective consultants on time.
The Phase 1 scope of work was sent out to seven consultants, but only two decided to bid the work. The Phase 1 contract
was awarded to Noble Denton in December 2007 with funding supplied by the IADC jack-Up Committee, Shell, and a
smaller amount being supplied by the UK HSE.
6 OTC 22949

Phase 2 Comparison to 5-5A


The “meat” of the benchmarking was to be contained in Phase 2. Phase 1 was to be the precursor, designed to allow
Phase 2 to proceed as efficiently as possible. Phase 2 was to undertake a quantified assessment, and compare the results of a
5-5A site assessment with a similar assessment using ISO 19905-1. But to make the situation a little more difficult, the
Benchmarking Panel also needed to find out how different consultants would interpret the ISO Standard, and if the results of
two different consultants would be comparable (for the same basic input). It was known from previous experience in
comparing analyses from different consultants (e.g. studies undertaken prior to the development of 5-5A, and soon after its
publication) that this could be a difficult job to control and interpret. To this end, considerable care was taken in developing
the Phase 2 scope of work to ensure that all bidders knew what was expected of them.
The basic analyses required were similar to those undertaken in Phase 1, but for Phase 2 the assessment cases needed to
be chosen close to the operational limits of the specific jack-up being assessed.
The original plan was to have 4 independent consulting companies assess a total of four different jack-ups, two by each
company. Each consultant would assess each of the jack-ups to the ISO 19905-1, but only one consultant would assess each
jack-up to 5-5A. A sample assessment matrix, as originally planned, is shown in Table 3.

Consultant Phase 2 Rigs assessed


A Rig 1 (5-5A & ISO) & Rig 2 (ISO only)
B Rig 2 (5-5A & ISO) & Rig 3 (ISO only)
C Rig 3 (5-5A & ISO) & Rig 4 (ISO only)
D Rig 4 (5-5A & ISO) & Rig 1 (ISO only)

Table 3 Planned Rig Cases and Consultants

Provisions were included in the scope for having an “early start” consultant. The intent of the early start was that they
would start their analysis before the other consultants and would have the opportunity to raise flags if they found problems
with the standard. While it was envisioned that the benchmarking would be carried out using a DIS of 19905-1, the standard
was still being reviewed and edited, so it was thought likely that some problems would be encountered. By having an “early
start” consultant there would not be mass confusion when all consultants came across the same problem at the same time.
One would arrive there first, would raise the flag, and the problem could be resolved with as little impact on the other
consultants as possible. It transpired that there were not too many roadblocks, but the early start consultant did manage to
raise some flags early enough to prevent the other consultants following courses through the assessment that would have
made comparison between 5-5A and 19905-1 difficult, and significantly less meaningful.
19905-1 allows a number of different analysis approaches, and there was no restriction on the approach that could be
taken by the analyst, as long as it was within the strict letter of the standard. For example, it was assumed that most
consultants would undertake a time domain dynamic analysis to develop the responses, and hence the dynamic amplification
factors (DAFs). But there was no requirement to follow this route, and if a consultant decided to follow the single degree of
freedom (SDOF) approach in developing the DAF, then this would be acceptable assuming the strict 19905-1 methodology
was followed.
For each different jack-up being assessed there would be an assigned “lead” consultant who would undertake both an ISO
assessment, and one to 5-5A.
There were to be alignment/stop points in the assessments. These were deemed to be critical to the success of this project.
The scope contained a limited number of required alignment points, but consultants were encouraged to have additional ones
to help clarify calculations. At these alignment points the two consultants assessing the same jack-up type were to compare
their intermediate results. The intent was to ascertain the causes of any differences. The intermediate results, any
differences, and their causes were to be documented at the time of alignment for inclusion in the final report. If ambiguity or
lack of clarity in 19905-1 was determined to be the cause of the difference, the reference and suggested changes in wording
was to be included in the report. It was stated that the consultants were responsible for resolving any differences found, with
limited, if any, input from the Benchmarking Panel. The importance of this documentation and their causes was stressed; this
task was important to the objectives and success of the study, and was not be treated lightly. However, it was appreciated
that this was not intended become a research project either. Therefore, application of sound judgment and consultant
interaction to resolve differences was strongly encouraged. Once the causes had been found and documented, both
consultants were to continue the analysis using the same intermediate values. In most cases the Lead Consultant values were
to be used, unless there is a good technical reason that other values should be used. The choice of value to be used in the
continuing analysis was not intended to reflect on the competency of any analyst, but unless there was consistency, it would
not be possible to draw reasonable conclusions. It was stressed that work should not continue beyond an alignment point
with consultants using different values.
OTC 22949 7

There were a number of issues that complicated the situation slightly. Since the assessments would be using real jack-up
designs, or modifications of real designs, there was some concern for confidentiality, and it became necessary to determine if
any of the possible consultants had jack-up designs that they were actively promoting. In addition, it was imperative that the
exercise not be seen as a comparison between rig designs, or consultant’s ability to interpret the code more or less leniently.
None of these issues were insurmountable, but considerable time was taken ensuring that there were no conflicts of interest,
and that reports could not be used as “sales” for a specific unit’s capability. Indeed, to help prevent any potential for this type
of misuse, it was decided to change certain parameters for two of the jack-ups that were assessed so that it would not be
possible to draw any operational capability conclusions from the final reports.
The process of the analyses was very similar to the Phase 1 process, but in this case, the two consultants assessing a
specific jack-up had to agree on the parameters to use. Geotechnical data was supplied for a sand location and a soft clay
location, but apart from that the consultants had to choose a suitable set of location parameters that put the unit close to its
survival limits (Ultimate Limit State - ULS). There were only four alignment points specified, but consultants were
encouraged to have others if felt they would add value. The 5-5A analyses would continue beyond the alignment using the 5-
5A calculated values, but the consultants were required to align for the 19905-1 assessments. The four required points were:
1. After the spudcan penetration and soil spring stiffness had been determined.
2. Natural periods both with and without foundation fixity, with and without P-Delta effects, gravity load case
reactions, and choice of hull sag percentages.
3. Dynamic amplification factors, including specification of the method of derivation
4. Breakdown of the wind, wave & current, and inertia loads, plus responses to those loads (e.g. hull sway, leg
bending moment at the lower guide, footing reactions, etc.)
The reports were to be detailed including, among other requirements, the completed tables that were included in the scope
of work (modified if necessary), alignment results, differences for any lack of alignment, lists of areas where it was thought
the standard was inconsistent or unclear, and any questions that needed to be raised. For the consultants assessing a jack-up
to 5-5A as well as 19905-1, the report was to include a full description of differences found, and their likely consequences.
Go-By Document One of the documents produced during the development of 5-5A was a “Go-By” document; it set out a
detailed calculation following one complete path through the assessment process. This document proved to be very valuable
to people new to jack-up site-specific assessment, and to those who had not completed one for some time. The intent was not
to allow an inexperienced engineer to undertake complex jack-up analyses, but to help a good competent engineer understand
the process of assessment, and guide them through the analysis. The development of a go-by document was always planned
to be part of the Phase 2 scope of work, and if sufficient funds were to be found, one consultant would be chosen to lead its
development. The document was to set out all the steps followed by the consultant as they progressed through the analysis,
including detailed references to the relevant clauses in the standard. Where possible, sufficiently complete calculations were
to be provided such that future users of the standard could follow the go-by. The document was to be delivered in steps.
Within 2 weeks of completing each Alignment point, the chosen contractor was to deliver the go-by for all the steps up to
that point. The steps of the Go-By document were then to be given to the other consultants for comments. Comments could
then be incorporated, and a final Go-By delivered soon after all the analyses are completed

Phase 3 All Routes Check


Once the standard had been benchmarked to 5-5A, and the go-by developed, there was a plan to make sure that all the
parts of the document that are rarely used were both usable, and produced reasonable answers. An example could be the use
of the slender section assessment methods described in Clause 12 on structural assessment. It is expected that there are very
few slender section components in jack-up leg chords, but were one to be found, then the analysis of it will not be simple.
Inherently, the parts of a slender section that are effective will depend on the loading on the section, so assessment will be
complicated by the need to know what the loading is before the section properties can be determined. The plan in Phase 3
was to go through these sorts of clauses in the standard. It now appears that this complete work will not be undertaken, but
certain specific clauses will likely be tested. One such part is the clauses dealing with seismic assessment, which is discussed
in “Future Work” below.

Phase 2 Benchmarking Consultants Chosen


Bid requests were sent to 13 different consultants and quotes were received from eight, including some that had teamed up
into groups. All bids were sent to the Benchmarking Panel Chairman, and he was the only person who saw the full details of
consultant name and cost until after an initial review by the rest of the Panel. Each panel member was responsible for their
own bid assessment process, but most followed the usual assessment approach of considering cost, experience, understanding
of the work scope, etc. It was decided, based on the cost, capability, and available funds that the originally planned
arrangement of using four consultants to assess four jack-ups would be slightly modified and contracts were awarded to
Global Maritime, GL Noble Denton, Bennett and Associates, and MSC. The biggest change over the original plan was that
MSC would be the only consultant assessing the CJ 62 (an MSC designed unit), and that they would be the “early start”
8 OTC 22949

consultant. The other three consultants would each assess two jack-ups to 19905-1, and each jack-up would be assessed by
one consultant to 5-5A. GL Noble Denton was chosen to develop the go-by document based on the LeTourneau Super
Gorilla. It is important to note that certain design parameters of the Keppel B Class and the LeTourneau Super Gorilla were
changed for the Phase 2 benchmarking

Company Rig 1 SNAME? Rig 2 SNAME? Go-By? Early Start?


Global Maritime 116 C Yes Super Gorilla No No No

GL Noble Denton Keppel B Class Yes Super Gorilla Yes Yes (on SG) No

Bennett & Assoc Keppel B Class No 116 C No No No

MSC CJ 62 Yes - - No Yes

Note: SNAME? shows which consultants undertook a 5-5A assessment of the design (only one per rig)
Go-By document is similar to that developed for 5-5A, but in this case for ISO 19905
Early Start consultant started work before the other consultants, and maintained a lead start in order to help find problems
that could cause complications or delays
Table 4 Consultants Chosen for Phase 2 Benchmarking

Value of Early Start Consultant


The intent of having an early start consultant was to catch roadblocks in 19905-1 prior to the other consultants arriving at
them. It transpired that there were not many problems with actual roadblocks in the standard, but there were some changes
from 5-5A that would have had a significant impact on the usefulness of the benchmarking exercise.
Kinematics Reduction Factor: 5-5A has a simple kinematics reduction factor of 0.86 that is normally applied to the
wave height. 19905-1 incorporates the wave spreading work which determines a spreading factor that includes regional
consideration that was undertaken by Noble Denton [Ref. 10, 11, 12] and this factor is applied by reducing the wave particle
kinematics. MSC pointed out that this simple change, which was known to have a significant effect on the overall wave &
current loads, could swamp the results of an ISO to 5-5A comparison, thereby hiding other significant changes that existed.
Because of this early warning it was possible to alert the other consultants and request that they calculate the relevant
kinematics reduction factor that would be produced through use of 19905-1, and report this value in their final reports, but
then use the consistent value of 0.86.
Apparent and Intrinsic Wave Period: The ISO standard addressing metocean issues, 19901-1 [Ref. 13] requires
consideration of the effects of current on wave period. This requirement had been incorporated into 19905-1, but its full
impact had not been understood. MSC alerted the other consultants to the potential significance of this change (see also
discussion in section on results below and [Ref. 10]). Preliminary results presented at meetings in May 2010 gave a 12%
increase in the inertial load from 5-5A to 19905. This difference was later revised down when other factors were taken into
account, but it pointed out that this was an area of potential significance.
Deep Penetration in Clay: The benchmarking consultants had been requested to use 5-5A revision 3 in their
comparisons. MSC pointed out that the 5-5A Rev. 3 did not contain the deep clay penetration equations that had been
incorporated into 19905-1. This early warning gave the Benchmarking Panel time to send out an unofficial revision to 5-5A
Rev 3 that could be used in the benchmarking, thereby allowing a more realistic clay case comparison. (The omission of the
deep clay penetration cases from 5-5A Revision 3 was an oversight that had not been noticed up until this point.)
In addition to these major items, there were some typographical errors identified that allowed information to be passed to
the other consultants, and thereby save some additional effort. But fortunately there were not too many of these “errors”.

Results of Comparison between 5-5A and ISO 19905-1


Three consultants were involved in comparative analyses between 19905-1 and 5-5A, assessing a total of four separate jack-
ups. As can be seen from Table 4, MSC assessed the CJ 62; Global Maritime the LeTourneau 116; and GL Noble Denton the
Keppel B Class and the LeTourneau Super Gorilla. In all cases there were no great surprises, particularly with the
forewarnings from MSC as early start consultant, although there were some differences in the results. Notwithstanding these
changes, there was little change in the conclusions of the comparative analyses when referenced to the “approvability” of a
jack-up; some of the unity checks changed more significantly than others, but these were never the critical unity check. In a
few cases there were differences between the 5-5A analysis and the 19905-1 analysis because of errors or omissions in the
ISO document. For example, the DIS version of 19905-1 used for the benchmarking contained an error in that the drag
coefficients of leg members of circular tubular cross section exposed to wind was 0.5, lower than the intended value of 0.65.
This error was not significant, but has been corrected in the published ISO. Some of the issues discovered are discussed in
the following sections.
OTC 22949 9

Spudcan Penetration
Two cases were considered; a sand foundation and a clay foundation. The sand foundation resulted in penetrations of
approximately 2 meters for all jack-ups assessed, and there was very little difference between the 5-5A value and that
calculated for 19905-1, although generally 5-5A produced marginally larger penetrations. On the clay foundations the
penetrations were between approximately 30 m for the smaller units and 45 m for the larger but, again, there was little
difference between the 5-5A and the 19905-1 penetrations. In two cases the 5-5A penetration was smaller, and in two cases it
was larger. But the largest difference was just under 10%.

Natural Period
There was very little difference in the calculated natural periods between 5-5A and 19905-1. In two cases they were the
same. For the other two jack-ups one had a higher period for 5-5A, and the other had a higher period for 19905-1.

Wind Force
In most cases there was little or no difference in the wind force between 5-5A and 19905-1, except that one consultant
used 0.5 as the drag coefficient for tubular leg members in air, including the “tubular” direction of leg chords (as the given
version of 19905-1 incorrectly stated it should be). The only other differences were if there was different leg penetration
resulting in a different reserve of leg above the upper guide.

Wave Loads
The biggest factor affecting the difference in wave load between 5-5A and 19905-1 was the use of the kinematics
reduction factor. As mentioned above, the Benchmarking panel requested all consultants to use a consistent value of 0.86 for
the factor but, in compliance with their normal analysis methods, at least two of the consultants applied the factor differently
between the two cases: for 5-5A assessments it was applied as a wave height reduction, and for the 19905-1 case it was
applied as an actual factor on the kinematics. This resulted in slightly higher loads in the 19905-1 case, of about 4% or 5%.
There was little difference in the magnitude of the increase between wave force and wave moment. It is important to note
that in all the cases considered where the location was stated as being either North Sea or Gulf of Mexico, the use of the
19905-1 location specific wave kinematics factor would have resulted in a lower overall 19905-1 wave force. However,
there are a few locations in the world where the use of 0.86 would be non-conservative.

Dynamic Amplification Factor


The DAF was an area in the comparison that could produce significant differences between the two analysis methods,
although in the cases considered the magnitude of the differences were only up to about 5%. It is important that one of the
consultants, MSC, used the single degree of freedom (SDOF) approach to develop the DAF (i.e. used the simplified two
stage approach). Both of the other consultants doing 5-5A-19905-1 comparisons (GL Noble Denton and Global Maritime)
used a more complex two stage approach based on time domain dynamic analyses (i.e. develop the DAF as the ratio of two
time domain analyses, one run with mass, the other without). [It is of note that these DAFs will necessarily not be
comparable, even if they were being generated for exactly the same jack-up in the same conditions. The reason is that the
SDOF DAF is used to determine an inertial loadset based on the amplitude of the time varying component of the
deterministic wave. The DAF developed through time domain analysis methods uses the magnitude of the deterministic
wave force (or moment) to determine the inertial loadset.]
There are a number of factors that will cause difference between the two analysis methods, but the most significant are the
foundation fixity levels and the effect of the difference between the apparent and intrinsic wave periods.
As discussed in [Ref. 10], the calculation of the DAF in 19905-1 is based on the apparent wave period rather than the
intrinsic. [The different wave periods are caused by the effects of current. The apparent period is that which will be
observed from a facility fixed to the seabed. The intrinsic is that observed by someone travelling at the speed of the current.
Structural dynamics are driven by the apparent wave period. Wave particle kinematics, and hence wave forces, are driven by
the intrinsic wave period.] When calculating an SDOF DAF, the apparent wave period is simply used in the calculation in
place of the intrinsic. Given that in most analyses the current will be in the same direction as the wave, the apparent wave
period will be shorter than the intrinsic. Since the jack-up natural period is normally shorter than the wave period, a
reduction in the exciting period (e.g. from intrinsic to apparent) will tend to increase the DAF. This is what was found by
MSC with their preliminary results, presented during May 2010, giving a 12% increase in inertial load due to incorporation
of apparent/intrinsic effects. The final differences were not quite as large with SDOF DAFs up to 5% larger for the pinned
foundation case. In the cases with spudcan fixity, there was a much smaller difference, approximately 1%, but this was in
part be due to the two competing differences of fixity level and wave period (the deep penetration clay fixity level through
19905-1 was 10% higher than that from 5-5A, for the CJ 62, but this was in part due to an omission in 19905 that was later
corrected - final 19905 fixity is the same as 5-5A).
10 OTC 22949

When using a time domain dynamic analysis approach, the incorporation of the apparent/intrinsic difference can be more
complex. Global Maritime used an approach, allowed within 19905-1 under certain limits, of simply reducing the peak
period of the input wave parameters. It is understood that GL Noble Denton modified their software to incorporate the
effects of current on each of the random wave components. [It is of note that during the benchmarking there was
considerable ongoing discussion as to how this effect should be incorporated, so there are a number of methods that could
have been legitimately used at the time.] One of the GL Noble Denton cases had zero current, and in that case the difference
between the 5-5A and 19905-1 DAFs is very small, but in random directions. For the other case, in which there is a current
present, there is up to a 5% difference in DAF, however, it is consistently higher for the 5-5A sand case, and consistently
higher in the 19905-1 clay case. Global Maritime found the 19905-1 case consistently higher by 2%.
In these particular cases the effect of the apparent/intrinsic difference were not significant, but for shorter period wave,
this will not necessarily be the case.

Extra Loadcase Equations


19905 contains some additional loadcase equations that should be considered when a jack-up is assessed under conditions
close to resonance, see [Ref. 15]. These extra equations were not part of the original benchmarking process, but some simple
cases were run prior to finalization of all of the work. These cases demonstrated that use of the additional loadcases will
normally result in a step increase in the leg loads at the lower guide. This effect is discussed further in [Ref. 15].

Foundation Utilization and Additional Cases


Summary of Foundation differences between 5-5A, DIS 19905, and FDIS 19905. The following is a very brief
description of some of the differences between the foundation assessment in 5-5A, the DIS 19905 used for initial
benchmarking, and the final FDIS 19905. For a more detailed description, see [Ref. 16].
5-5A foundation check includes the variable VLO which is used in the yield surface equation to determine foundation
utilization. The horizontal foundation capacity for sands and the moment capacity for clays and sands are also written in
terms of VLO. 5-5A defines VLO as “maximum vertical load applied to the foundation during preload”. This would appear to
be a simple definition which is easily understood by the user. However, the structural and geotechnical engineer have
different interpretations for the definition of this term. Most structural engineers interpreted VLO to be the sum of the
lightship, variable load, and preload ballast, minus the leg and spudcan buoyancy. Geotechnical engineers interpret VLO to
include the same actions as the structural engineer, but adjusted for the soil column due to the backflow and infill, minus the
buoyancy of the soil displaced by the spudcan. DIS 19905 addressed these two interpretations by the introduction of the
concept of gross ultimate vertical bearing capacity and net ultimate vertical bearing capacity (see [Ref. 16]).
When factoring the yield surface from 5-5A, some assessors used the curve origin (0,0) while others used the still water
level (SWL) reaction. To reduce the confusion, when the DIS 19905 was developed, the yield surface was specified to be
factored from the (0,0) origin, not the SWL reaction. To determine the foundation utilization, the vertical reaction VLO/(2*
was defined for a horizontal force of 0 as the starting point. A figure was then added to show the intent of the standard.
DIS 19905 to FDIS 19905: Gross and net ultimate bearing capacities were carefully defined in DIS 19905, however, the
new definitions resulted in unexpected consequences [Ref. 17]. 5-5A permitted the assessor to use a simple foundation
check, without having to perform a higher level analysis. The simplified check remained in the DIS 19905 but, with the
introduction of the gross and net terms, it became apparent that the simplified check produced lower utilizations than the
more correct higher order, more sophisticated, check. This is not a desirable outcome. The study also found for friction
angles above 25° in sand, the sliding capacity did not intersect the yield surface, also not a desirable outcome [Ref. 17]. The
geotechnical engineers found that by using the same partial factor for fully penetrated and partially penetrated spudcans, and
adjusting the origin for deep penetrations, the foundation checks produced more realistic results without the problems
described above. To assist the assessor, improved “Vertical-horizontal foundation capacity envelopes” figures were
introduced in the FDIS 19905. The differences from the original 5-5A through the DIS and FDIS ISO 19905 standards used
in the benchmark study was succinctly presented by one of the consultants in a table and is adopted below.
OTC 22949 11

Footing Reactions
Bearing Capacity Origin used for
Practice or Utilization Origin Resistance Factor, R,VH construction of
Standard (FH, FV)ORG Partial Fully factored V-H
Penetration Penetrated envelope
1 2
SNAME (0, SWL ) 1.11 1.18 (0,0)
ISO DIS (0, Qv/2R,VH) 1.1 1.15 (0,0)
ISO FDIS (0, Qv/2R,VH) 1.1 1.1 (0, WBF,O-Bs)
1
SWL = Stillwater footing reaction
2
This origin was used by at least one consultant and other consultants used the SWL or other origin
Table 5 Comparison of partial factors and assumptions for origin in the construction of the yield surface in foundation utilization
checks

Modified Foundation Analysis Towards the end of the Phase 2 benchmarking process when it was realized that there
were inconsistencies in the way the clay foundation was assessed, changes were made to 19905, and the consultants
requested to reassess the clay cases using a consistent set of footing reactions for the revised procedure.
Comparison of Results between 5-5A, DIS 19905, and FDIS 19905: The consultants were requested to reassess the
foundations, using a consistent set of reactions, to 5-5A, the original DIS version of 19905 that they had been supplied for
benchmarking, and the new modified 19905. The calculated utilization results are presented in Table 6.

New New ISO / Old ISO / New ISO /


Case Old ISO 5-5A
ISO 5-5A 5-5A Old ISO
GM 116C 0.96 0.84 0.86 0.97 1.12 0.87
BASS 116C 0.78 0.69 - - - 0.90
GM Super Gorilla 1.48 1.25 - - - 0.85
GLND Super Gorilla 1.45 1.31 1.72 0.76 0.84 0.90
GLND B-Class 1.24 1.11 1.52 0.73 0.82 0.90
BASS B-Class 1.12 1.03 - - - 0.91
MSC CJ 62 1.06 0.90 0.89 1.01 1.19 0.85
Note: GM is Global Maritime, BASS is Bennett and Associates, GLND is GL Noble Denton
Not all consultants assessed foundations to 5-5A, as marked by a “-“.
Table 6 Clay Foundation: Utilizations from different Consultants for different Jack-Up designs to original DIS 19905, 5-5A, and
Modified 19905

It can be seen that all consultants found the difference between the old and new versions of 19905 to be approximately
0.9, with very little variation (right column). It is interesting to note that GL Noble Denton (GLND) found 5-5A to be more
conservative than the original DIS version of 19905 for the two cases they assessed (second column from the right) while
both Global Maritime (GM) and MSC found it the other way around. Not all of the reasons for this are clear, but part of the
reason could be a different interpretation of the utilization in 5-5A. GLND develop the factored 5-5A yield surface by
factoring from the (0,0) origin (similar to the original DIS 19905) while GM and MSC factored it from around the still water
vertical load (similar to the modified 19905). In addition, there appear to be a differences between the consultants in the
horizontal capacity. While additional work is needed to ensure that the interpretation of 19905 is consistent, it appears from
an assessment of the reports associated with the results given in Table 6, the majority of the differences between consultants
are in the interpretation of 5-5A foundation assessments (not specifically assessed in this study).

Conclusions and Final Utilization Comparison between 5-5A and 19905


In general, the 5-5A and 19905 utilizations were very similar (except for the original foundation utilizations, as discussed
above) and were nearly always within 10% of each other. In some cases the 19905 overturning moment was slightly larger
than the 5-5A equivalent, and this lead to some marginally higher 19905 utilizations. The leg member capacities in 19905
tend to be higher than those in 5-5A, thereby offsetting the increase in overturning (see also more detailed discussion in
[Ref. 9]). This is particularly true for the leg chords. To quote MSC in their 5-5A / 19905 comparison report
“...chord has approx 15-20% higher factored axial strength due to which the resulting chord strength utilization in
ISO is approx 10% smaller (with some 10% higher axial loading)”.
The increase in chord strength from 5-5A to 19905 is significant, and makes the chord much less likely to be the critical
component. The obverse to this was that MSC found higher utilizations for the leg fixation system in the same case (due to
the higher chord loads).
12 OTC 22949

The differences in the 116C assessment resulted in all values (excluding foundation) being within 4% of each other,
except for the chord utilizations which were 6% lower for the 19905 case due to the increased chord capacity.
On the Super Gorilla the chock holding utilization was 10% higher for the 19905 clay case, but 4% lower for the 19905
sand case. Conversely, on the KFELS B-Class, the 19905 sand case chock utilization was 5% higher and the clay case 9%
lower.
The CJ 62 tended to show a more consistently higher set of utilizations for the 19905 cases than the 5-5A cases, while the
Super Gorilla and KFELS B-Class tended to show more consistently lower 19905 utilizations. But the differences were still
less than 10% in almost all cases.
The following quote from one of the GL Noble Denton reports almost perfectly describes the differences between 5-5A
and 19905 in that while one part may be a little more onerous, another part makes up for it.
The greater predicted penetration depth calculated using SNAME results in a shear modulus that is 8% greater
than that calculated for the ISO case. The stiffness depth factors in the ISO document, however, are greater
than those in SNAME for the case of a clay foundation (with Poisson ratio = 0.5). Consequently the overall
stiffnesses calculated using the ISO document are very similar to those using SNAME (SAGE) .... As different
depth factors are used for the vertical, horizontal and rotational stiffnesses, the ratio between the SNAME and
ISO stiffnesses are different for each component stiffness. The loading condition is similar; ISO results are
slightly less onerous for site 1 (sand), and near-identical, or slightly more onerous for site 2 (clay) when
compared to SNAME ...;(the difference in) lower guide reaction loads ... is up to 12% on axial load, 16% on
moment (for the maximum hull weight case) for the non critical legs, but shows close agreement on the critical
(most highly loaded) legs.
In conclusion, apart from the prismatic member strength checks, there were no significant trends found by the assessments,
and the scatter between results within an analysis technique are going to be as large as the differences between the different
techniques. It is likely an analyst could create a greater difference in results within an assessment, to either 5-5A or 19905,
by “sharpening their pencil” and undertaking a more detailed analysis than they would find between the different
assessments. Notwithstanding the above conclusions, in all the cases assessed the foundation utilizations were critical. It is
possible that were a different foundation to be assessed that was not critical, then the less conservative structural checks in
19905 could have more of an impact. Also, for most locations, the 5-5A kinematics factor of 0.86 will tend to be
conservative, and the 19905 value will produce a less conservative result. These effects have not come out in the current
study.

Results of Parallel Analyses by Consultants


This was an extremely difficult study for all the consultants that were having to align their results at the specified points, and
this must be kept in mind when assessing the results of the studies. [N.B. MSC were not required to align with another
consultant as they were the only one assessing the CJ 62.] There were certainly times when one consultant was wanting to
move forward but could not because of lack of alignment. At other times it would be the other way around. Plus the onus
was more heavily on the non-lead consultant as they were the one having to align; the lead consultant could normally move
on and discuss alignment when the non-lead caught up. Notwithstanding all these complications, it was frustrating to find
that the absolute requirement of aligning before moving forward was not consistently followed.
It is beyond the scope of this paper to go into the details of the alignment issues, but some general points can be made
about the results:
 There was a danger of the consultants not fully and carefully reading 19905-1, but applying what they expected to
be present rather than what was actually present. This tended to be an issue early on that was resolved in some
cases, but it is possible that it could have been the cause of some apparently unaccountable misalignment later in the
calculations
 The difference between DAFs needs to be carefully considered. The DAF is used to calculate an inertial loadset,
and the inertial loadset is comprised of the product of the deterministic wave force/moment with (DAF-1).
Consequently, when comparing DAFs from two separate analyses it may be preferable to compare (1-DAF). Hence
the difference between a DAF of 1.2 and another of 1.1 is not 1.2/1.1 = 1.09 but (1.2-1)/(1.1-1) = 2.0, equivalent to a
100% increase rather than a 9% increase in the inertial loadset.
 There were some very large difference in calculated DAFs when using time domain dynamic analyses approaches.
It is not clear why this occurred, although one consultant was consistently lower than the other two. It is also of note
that two consultants used both the Winterstein and Drag/Inertia methods to calculate the DAFs. The results were
significantly different between the two methods for a particular series of sand locations with Winterstein being
between 6% and 17% higher than the Drag Inertia method (average of 12% higher). In this case the ratio of natural
period to wave period was 0.37. For the equivalent clay case the differences between Winterstein and Drag/Inertia
method were very small, with Winterstein being the smaller DAF for one assessment. For this clay case the ratio of
natural period to wave period was 48%. It is possible that the sand case represented an anomaly because it was the
OTC 22949 13

case with the largest difference in DAFs between the consultants, had Winterstein DAFs that were similar to or
above the clay case DAFs, and yet had the smallest ratio of natural period to wave period.
 There were a number of cases in which consultants made numerical errors. It is likely that many of these would
have been caught by the consultants own internal QA process, but in this particular study they were found during
results comparison.
 Leg buoyancy was not consistently included in reporting the results, and it was often difficult to know when it had
been included, and when not.
 The use of P-Y curves increased the lateral foundation springs in one case making comparison of the standard case
difficult.
 In one case of poorly following the alignment requirements, there appeared to be a divergence of results as the
analyses progressed, but surprisingly, when the total loadings were calculated, including wind, wave & current,
inertia, P-Delta, etc., the final results were remarkably similar (less than 10% difference).
 There were a number of “false starts” where there was erroneous inclusion of one parameter or another, only to be
later corrected (e.g. inclusion of too much or too little leg rack in leg stiffness calculations). As with the numerical
errors, it is likely that many of these issues would have been resolved internally by the consultant without any
outside “prodding”.
 There are some apparent anomalies in the member stress checks. It appears that one consultant was consistently
getting higher member unity checks for relatively similar leg loads. It is possible that they were not accounting for
some of the enhancements in the column curve allowed in 19905-1, but not in 5-5A.
 One consultant used simple single step inclusion of P-Delta effects in their final assessment rather than an iterative
approach, or use of a negative spring. While unlikely to produce significantly different results for the case
considered, it was a misinterpretation of the standard.
 Wind areas were different between two consultants due to scaling differences off a drawing. The difference was
small, but noticeable.
 Simple differences between the way consultants report their results can make comparisons considerably more
complicated. For example, altering table column heading orders, or metocean approach directions can add
significantly to the time required for a comparison.
 Not all comparison results were given to other consultants in time for publication in the final reports, again making
comparisons more complex.
 Some consultants produced long tables with multiple values, while others produced simple single value equivalent
tables, e.g. leg hydrodynamic coefficients were given for multiple sections by one consultant, and only one section
by another.
 It is not always clear how hull sag was included, factored down, or even excluded by the consultants.

Future Work
There are some future studies planned that will make extremely good use of the additional funds made available by Chevron
contributing to the Benchmarking fund. The first item is to create a summary report of all the submitted benchmarking
reports from the four consultants. This work had always been planned, but no funding was available for it. Some of the
necessary work was undertaken in preparing this paper, but there is still scope for a more detailed assessment of the
differences, particularly those between consultants. It appears that the differences between the 5-5A and 19905 analyses are
not too great and, while there is scope for demonstrably showing there are no significant trends, there is little point in
spending significant funds on quantifying differences between assessments that vary mostly within the analysis rather than
between analyses. As an example, if the calculated DAFs are all very similar and within the expected scatter of a Winterstein
MPME assessment, why try to chase down the differences? The differences between consultants tended to be larger, and
could be more significant. While these can be difficult to isolate, it is important to try to find trends that could point to areas
of misinterpretation or ambiguous interpretation.
There has been a proposal request issued to assess the seismic parts of 19905, and to determine if they make sense. There
was no seismic assessment required within 5-5A, so all the material in 19905 is new, although it relies heavily on the seismic
standard, ISO 19901-2 [Ref. 14]. The purpose of this proposed work is to ascertain how often it is likely that some level of
seismic assessment would be required, whether the screening technique contained in 19905 will adequately resolve most of
those cases, and if the screening technique is fully workable. It is hoped that this work will be underway by the time this
paper is published.
There are some other possible areas for additional study based on the results of the submitted reports, but these need to be
more clearly identified before they can be discussed.
14 OTC 22949

Conclusions
The benchmarking exercise was a massive undertaking that involved difficult communications between companies that,
while used to communicating technical details, may have been uncomfortable comparing actual calculations. However, the
results have been very valuable and act as a permanent record of an important exercise.
There were some differences in the results of analyses between those undertaken using SNAME T&R Bulletin 5-5A and
ISO 19905-1, but they are not large. It is possible that at certain specific locations an analysis to 19905-1 could show a unit
is unacceptable while one to 5-5A shows it to be acceptable, or vice versa, but it will be rare. The member strength checks in
19905-1 are more lenient than those in 5-5A, and this is particularly true for most leg chords, but this factor was not
significant in the comparisons since in all cases the foundation was found to be critical. It is also important to realize that
although the geotechnical unity checks are different in 19905-1, the effective capacities are comparable, although the factored
capacities will be different, depending on the initial interpretation of 5-5A. Another area that was intentionally not assessed
in this study was the effect of kinematics factor. For most areas of the world, the 19905 factor will produce lower loads, and
hence lower utilizations, than 5-5A.
There are certainly differences in the way consultants undertake analyses. In certain cases the differences in results were
large, and could be very significant, but it is likely that were a particular jack-up location close to its operating limits, a more
detailed assessment would help to tighten up the alignment, particularly with a more careful reading of the standard.
All the consultants were experienced with performing 5-5A analysis and a few had experience with sections of 19905.
The similarity between 5-5A and 19905 can be misleading and using 5-5A interpretations of 19905 can result in an incorrect
answer. It also appeared from the work that familiarity with the development of 19905 helped some of the consultants
progress more easily than others. This difference should become less noticeable with increased use of the standard.
ISO 19905-1 is not a simple document to use, and considerable experience is needed to get through it without error. Even
the experienced analysts undertaking this benchmarking exercise had problems occasionally. It is hoped that the
development of the “Go-By” document, included in ISO 19905-2, will help even experienced analysts to follow the standard
error free.
In summary, the study has shown that the differences in the results of analyses between those undertaken using 5-5A and
ISO 19905-1 are generally small.

Acknowledgement
The authors would like to thank the IADC Jack-Up Committee, ExxonMobil, Shell, and Chevron for their generous funding
of the Benchmarking effort, and to all the members of the Benchmarking Panel for their efforts in giving direction to the
study, and in reviewing proposals and reports. The Benchmarking Panel members, apart from the authors of this paper, are:
Jim Brekke (Transocean), Rupert Hunt (Shell), Yi Li (Diamond Offshore), Pao-Lin Tan (ABS), Ward Turner (ExxonMobil),
and Paul Versowky (Chevron). Thanks are also due to Steve Newell of ABS Americas for his assistance with the
Benchmarking Project. Finally, we would like to thank the consultants who undertook the benchmarking studies. This was
not an easy job, and they all worked extremely hard to complete the work while trying hard to meet the difficult requirements
of the scope.

References
1 ISO 19905-1 “Petroleum and natural gas industries - Site-specific assessment of mobile offshore units - Part 1: Jack-ups” International
Organization for Standardization, 2012
2 SNAME, 2008, The Society of Naval Architects and Marine Engineers, “SNAME Technical & Research Bulletin 5-5A”, Rev. 3
3 ISO 19902 “Petroleum and natural gas industries - Fixed steel offshore structures” International Organization for Standardization,
2006
4 ISO 19905-2 “Petroleum and natural gas industries - Site-specific assessment of mobile offshore units - Part 2: Jack-ups commentary
and detailed sample calculation” International Organization for Standardization, 2012
5 “Jack-Up Assessment - Past, Present and ISO”, Mike Hoyle, John Stiff, Rupert Hunt, Alberto Morandi, ISOPE 2006-PM-06
6 “Jack-Up Site Assessment - The Voyage to an ISO”, Mike Hoyle, John Stiff, Rupert Hunt OMAE 2011-50056
7 “The role of SNAME OC 7 in the development of the ISO for Site Assessment of Jack-ups” John Stiff, Mike Hoyle, David Lewis,
17th SNAME Texas Offshore Symposium
8 “Background to the ISO 19905-Series and an Overview of the New ISO 19905-1 for the Site-Specific Assessment of Mobile Jack-Up
Units” Mike Hoyle, John Stiff, Rupert Hunt, Offshore Technology Conference, April 30 - May 3 2012 (OTC 23047)
9 “Structural Acceptance Criteria in the New ISO for the Site-Specific Assessment of Mobile Jack-Up Units” Paul Frieze, Dave Lewis,
John Stiff; OTC 23071, 2012
10 “Environmental Actions in the New ISO for the Site-Specific Assessment of Mobile Jack-Up Units” M. J. Dowdy, M. Hoyle, J. J.
Stiff, Offshore Technology Conference, April 30 - May 3 2012 (OTC 23342)
11 “3D Nonlinear Wave Spreading on Jackup Loading and Response and its Impact on Current Assessment Practice”, Smith SF, Hoyle
MJR, Ahilan RV, Hunt RJ, Marcom MR, Offshore Technology Conference, May 1-4 2006 (OTC 18266).
12 “Further Work on the Effects of Nonlinear Wave Spreading and its Impact on Current Jack-up Assessment Practice”, Hoyle MJR,
Smith SF, Ahilan RV, Hunt RJ, Marcom MR, Offshore Technology Conference, May 4-7 2009 (OTC 20297).
OTC 22949 15

13 ISO 19901-1 “Petroleum and natural gas industries - Specific requirements for offshore structures — Part 1: Metocean design and
operating conditions” International Organization for Standardization, 2005
14 ISO 19901-2 “Petroleum and natural gas industries - Specific requirements for offshore structures — Part 2: Seismic design
procedures and criteria” International Organization for Standardization
15 “Structural Modelling and Response Analysis in the New ISO Standard for the Site-Specific Assessment of Mobile Jack-Up Units””
P. Tan, J. Stiff, D. Stock, M. Perry, B. Mobbs, Offshore Technology Conference, April 30 - May 3 2012 (OTC 23040)
16 “Foundation Modelling and Assessment in the New ISO Standard for the Site-Specific Assessment of Mobile Jack-Up Units” P. C.
Wong, O. A. Purwana, M. S. Hussain, H. Hofstede, C. Martin, M. J. Cassidy, Offshore Technology Conference, April 30 - May 3
2012 (OTC 23521)
17 “Review of ISO 19905-1 Foundation Check and Proposed Amendments”, KeppleFELS and MSC, June 2010 (report to ISO19905
Panel 4)

You might also like