Preview
Filed 09 J uly 24 P12:20
Loren J ackson - District Clerk|
Harris Coun!
ED101) 015465939
By: Charleta J ohnson
CAUSE NO. 2007-75537
BILLY JACK McDANIEL, A’7LETA IN THE DISTRICT COURT
McDANIEL INDIVIDUALLY AND AS
NEXT FRIEND OF CARNEY
McDANIEL, A MINOR
VS. 6157 JUDICIAL DISTRICT
SMITH INTERNATIONAL, INC., SONERRA
RESOURCES CORPORATION, HARBOR
PRODUCTS, INC., KIRKHILL MANUFACTURING §
COMPANY A/K/A KMC RUBBER PRODUCTS, §
ADVANCED POLYMER TECHNOLOGIES, LLC § HARRIS COUNTY, TEXAS
PLAINTIFFS’ RESPONSE TO SMITH
INTERNATIONAL, INC’S MOTION TO EXCLUDE THE
OPINIONS AND TESTIMONYOE STEVEN J. ROENSCH
TO THE HONORABLE JUDGE OF SAID COURT:
NOW COME Plaintiffs and submit this Response and Brief in Opposition to SMITH
INTERNATIONAL, INC’S (hereinafter Smith) Motion to Exclude the Opinions and Testimony
of Steven J. Roensch and in opposition thereto would demonstrate the following to this
honorable Court:
I
SMITH RDH 2500
A. The Product
The product in question is the Smith RDH 2500 rotating control device and the required
component, Smith high-pressure natural rubber stripper clement. The rotating control device is
used in a number of oilfield applications where the well bore pressure must be controlled and/or
diverted during various drilling operations. The rotating control device and the component
stripper element are used to seal the annulus between the drill pipe and the casing. The stripper
element, which is bolted into place in this design, acts as a seal around the drill pipe during the
operation/movement of that pipe and as a diverter of hydrocarbons and drilling fluids, when
necessary. The stripper element is conical in shape and contains a metal insert to which the
rubber is bonded through the use of the Chemlok 220/205 bonding system. The stripper element
is sold by Smith to drilling operators for use with the rotating control devices. In the instant case,
Smith leased the rotating control device to Sonerra Resources, Inc, which was the operator at this
drilling sight.
B DESIGN
The stripper element was originally designed by Grant Tool company in the early 1980s
as detailed by various engineering drawings. Smith obtained Grant Tool Company and the
rotating control device product line in 1994. Since that time, there have been numerous revisions
ending with change M (2000), which was iteration in effect at the time of the manufacture of the
incident stripper clement in November of 2005. While many of the revisions did not relate to
function, several of the revisions were clearly directed at the problem of inversion. The stripper
elements were molded by various vendors including Harbor Products, Inc. beginning in the
1990s. The stripper elements were to have been molded in conformity with the specifications,
requirements and instructions provided by Smith. During 1997-1998, the Smith rotating control
device line was re-badged to remain competitive with the rating systems offered by its
competitors. The subject Smith RDH 2500 rotating control device was rated by Smith in static
and dynamic test environments. There does not appear to have been any finite element analysis
or other evaluation of the design of the product from the time of the Grant purchase until after
the subject incident. There is no evidence that Smith conducted any confirmation and/or
validation testing of the stripper elements at any point. There is no evidence that Smith
conducted any destructive testing regarding the stripper elements and the safety critical rubber to
metal bond.
On November 8, 2005, the high pressure, natural rubber stripper element was molded by
Harbor Products, Inc. The stripper clement was then shipped to Smith where it was allegedly
inspected for gross abnormalities. The Smith employees then inspected the stripper element with
a feeler gauge’. The stripper clement was then sold to Sonerra Resources and bolted into the
Smith 2500 RCD on March 2, 2006. The involved Smith high pressure, natural rubber stripper
element failed after thirty nine hours of service as the crew was tripping pipe at approximately
1100 psi. With the sealing element partially extruded through the rotating control device, the
well bore contents were allowed to escape and ignite. The subsequent conflagration engulfed
Billy Jack McDaniel inflicting third degree burns over 85% of his body. The appropriateness of
the design, marketing, and manufacture of the subject Smith high pressure, natural rubber
stripper element is a contested issue in this litigation and necessarily requires expert testimony
for explanation.
C. Steven J. Roench
The Plaintiffs retained Bison Engineering’s forensic engineering services in this matter
on or about August 1, 2006. The Plaintiffs specifically retained Randall K. King, Ph.D., PE
(hereinafter “King”) to evaluate the mechanical engineering and design aspects of the stripper
element involved in the event made the subject of this litigation. King retained the services of
Steven J. Roensch to create a finite element analysis as a part of King’s evaluation of the incident
stripper element and comparable designs. Specifically, Roensch was asked to model the stripper
rubber element containing the metal insert in various scenarios. Ultimately, Roensch created 32
‘ Smith had utilized a compression test more commonly known as the “squeeze test” to assess the safety
critical rubber to metal bonds but that test was discontinued inexplicably by Smith prior to the
manufacture of the subject stripper element.
cases which depicted the original stripper element with both a good bond and bad bond which
equates to a point where there was no bond or there was a weak bond which failed at loads far
below the load expected for that rubber. The FEA also permitted comparisons done with the
Smith combo stripper element with a good bond and a bad bond or an unbond to determine the
relative propriety of the bond location in the subject stripper element. There were also cases
where the drill pipe was moving through the stripper element. This stress analysis utilized
certain material properties based upon the stress versus strain data of rubber from APT which
indicated the tensile strength of the rubber from the batch of rubber utilized in the subject
stripper element. The goal of the analysis was to examine the design of the rubber from the
batch of the subject stripper element and compare that to the natural rubber combo stripper
element. The goal also incorporated a well bore pressure range which included a representative
pressure of 1100 psi since that was the pressure on the involved stripper element at the moment
of failure on March 3, 2006. The pressure ranged from 0 psi to slightly over 4000 psi which
equated to over 5005 of the Engineering Strain. The goal of the stress analysis was to predict the
stresses and resultant deformations for specific regions of the rubber in order to compare the
design concepts where the subject products were subjected to stresses comparable to the well
bore pressure monitored at the time of the failure of the incident stripper element. Drs. King and
Kasner provided the necessary inputs for the material properties as well as the mathematical
model which best matched the test data from ADT while also addressing the need predict the
response of the rubber at the higher stresses. His work was contained in the Bison Enginecring
expert report dated January 22, 2009 as Exhibit “1”. This exhibit includes the graphic depictions
of the finite elemental analysis cases as well as the APT data which was utilized in making
certain determinations discussed below. The Plaintiffs have attached the April 21, 2009
deposition of Roensch hereto as Exhibit “2”. This report, which was provided in compliance
with the Court’s Second Amended Docket Control Order, set out King’s analysis and the
underlying data which included the Roensch finite elemental analysis cases. On May 26, 2009,
King provided his supplemental report addressing certain issues and questions raised by the
depositions of David McStravick, Ph.D., PE and Ronald Frishmuth, Ph.D., PE. That
supplemental report also included information from Steve Roensch, The supplemental report is
attached hereto as Exhibit “3”.
Ill.
Finite Element Analysis
A Work performed
The evaluation of the engineering and design analysis of any product follows certain
readily ascertainable rules regardless of the product, In the instant case, Dr. King decided that he
wanted to utilize a finite element analysis to compare certain specific regions of the subject
stripper element to the comparable Smith combo stripper element. The finite element analysis
computer models can demonstrate the expected deformation of materials like rubber and/or metal
under readily ascertainable variables. The finite element analysis was not meant to replicate the
events of March 3, 2006 but it was meant to determine the comparative stresses and resultant
deformations at critical areas of the stripper element as determined by a thorough review of the
photographs, microscopic evaluation of the subject metal insert and disassembly of the metal
insert. The Roensch analysis” began with the selection of the material properties for the subject
rubber, Dr. King provided him with data from APT regarding the tensile strength of rubber from
the batch from which the subject stripper element was molded. Dr. Kasner also provided the
? Attached hereto as Exhibit "6" are excerpts from the publication Finite Element Procedures by Klaus-
Jurgen Bathe, which is a revision of the 1982 publication: Finite element procedures in engineering
analysis. The attached 1996 Prentice Hall publication addresses the finite element procedures for
elastomers.
tensile modulus and Poisson’s tatio for the rubber. Roench approximated the rubber as being
linear with a 700 psi tensile modulus because the goal of the investigation was to determine the
relative deformation of rubber at certain specific locations at specific pressures where the
literature and experience demonstrated that the rubber acted in a linear fashion. From those data
plots, Roensch determined that a simple curve fit would be best for the stress strain curve. Based
upon his experience and the literature, Roensch, King and Kasner determined that this was
appropriate for the comparison of the stripper elements. In his experience, Roensch has utilized
linear modeling on rubber where he was attempting to model the behavior of the rubber at the
high stress and strain regions. This linear modeling allowed Dr. King to determine the maximal
principal stresses for the rubber during at specific areas of the rubber within the areas identified
in the photographs and disassembly of the subject stripper element.
B Comparison not accident reconstruction
The Roensch FEA consisted of 32 cases where a number of different scenarios were
analyzed utilizing the standard and accepted analysis techniques and methodologies, King
utilized the Roensch FEA to compare the design parameters for the subject stripper element to
the Smith combo stripper element. The FEA provided information on the following queries
under four conditions®:
(a) Was the observed faulty/metal bond responsible for the failure of the RCD?
(b) Is there an alternative design which does not rely as heavily on the rubber/metal bond
and which would perform more safely in the instance of a bond failure?
While the results of these queries was important, King determined that the analysis must include
both pressure and movement of the representative drill pipe sizes in order to be more
representative of those aspects of the environment of use at the time of the failure. Roensch
3 The four conditions were as follows: original design with good bond; original design with bad/weak bond;
alternative design with good bond; and alternative design with bad/weak bond.
added the pipe to the analysis. Each of these conditions was analyzed for the following load
scenarios:
(a) Wellbore pressure - stripper element sealing on 3 1%” drill pipe — no movement;
(b) Wellbore pressure — stripper element sealing on 3 %” drill pip — pipe moving upward
(pipe well lubricated versus pipe poorly lubricated);
(c) Wellbore pressure — 4 % “tool joint being pulled upward through stripper element (pipe
well lubricated versus pipe poorly lubricated in rubber element), and
(d) Wellbore pressure ~- stripper clement sealing on 4 %4” tool joint — pipe moving upward
(pipe well lubricated versus pipe poorly lubricated).
King then compared the maximum principal stresses in the rubber elements from the various
scenarios and rendered certain opinions on the various opinions. The analysis focused on the
region around the tip of the metal insert which was clearly the most problematic area in the
photographs and disassembly of the subject stripper clement. The Roensch FEA revealed that in
the specific area studied, it took much less load (from the well bore pressure} to raise the stresses
to the point of failure. King summarized the finite element analysis results as follows:
The alternative design has much lower stress on the rubber in a worst case where
the critical location has a bad bond. Thus, the alternative RCD design does not rely
on the existence of a sound bond for safe operation and is much more robust and
much less likely to experience a catastrophic failure as a result of a poor bond.
Cc Frishmuth FEA
In defense of this litigation, Smith ordered finite element analyses to be performed on the
subject stripper element by Ronald Frishmuth, Ph.D and/or Tejas Laboratories. Frishmuth and/or
Tejas Laboratories have conducted several cases utilizing a nonlinear analysis. That analysis,
although different from the linear analysis utilized by Roensch, revealed that the highest
expected area of strain is at the distal end of the metal insert which is the same place that was
found in finite element analysis. While the values vary, the location and the underlying principal
is the same for this particular area of the stripper element. Attached hereto as Exhibit ‘4’ is a
graphical interpretation of the Stress versus Strain of Rubber; APT Test Data of Incident Batch
with Roensch & Frishmuth Models. The Frishmuth plot is similar to the Roensch model and
both compare favorably to the Test Data of the Incident Batch. It is interesting to note that the
Frishmuth plot takes on a linear and/or near linear form for the regions of stress where the parties
each focused their analysis on, The deposition of Ronald Frishmuth, Ph.D. is attached hereto as
Exhibit ‘5°.
Tk.
RULES GOVERNING THE ADMISSIBILITY OF EXPERT TESTIMONY
The Texas Rules of Evidence assign to the trial judge the task of ensuring that an expert’s
reasoning and methodology is scientifically reliable and relevant to the facts in issue. Daubert v.
Merrill-Dow Pharmaceuticals, Inc., 309 U.S. 579, 592-93 (1993). The Supreme Court is clear
that judges must undertake this gate-keeping obligation with diligence, even ‘though it “will
sometimes ask judges to make subtle and sophisticated determinations about scientific
methodology and its relation to the conclusions an expert witness seeks to offer.” General
Electric Co. v. Joiner, 522 U.S. 136, 147 (1997) (Breyer, J., concurring). The Court’s gate
keeping function applies to al/ expert testimony, whether based on scientific, technical or other
specialized knowledge. See Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 142 (1999). As
the Supreme Court explained:
[W]hether the specific expert testimony focuses upon specialized observations,
the specialized translation of those observations into theory, a specialized theory
itself, or the application of such a theory in a particular case, the expert's
testimony often will rest "upon an experience confessedly foreign in kind to [the
jury's] own." The trial judge's effort to assure that the specialized testimony is
reliable and relevant can help the jury evaluate that foreign experience, whether
the testimony reflects scientific, technical, or other specialized knowledge.
Kumho Tire, 526 U.S. at 149 (quoting Hand, L., Historical and Practical Considerations
Regarding Expert Testimony, 15 HARV. L. REV. 40, 54 (1901).
The party offering the expert testimony bears the burden to demonstrate that the
testimony is reliable by a preponderance of the evidence. Moore v, Ashland Chem., Inc., 151
F.3d 269, 276 on Cir. 1998), cert. denied, 526 U.S. 1064 (1999); see also Dart v. Kitchens Bros.
Mfg. Co., 253 Fed. Appx. 395, 398 (" Cir. 2007) (citing Moore). In the instant case, Roensch
has provided his finite element analysis with the underlying material properties to the parties in
the context of the King reports and mechanical engineering opinions. The disagreements over the
material properties are due to Smith’s basic misunderstanding of the purpose of the Roench FEA
and the King expert opinions.
RULE 702. TESTIMONY BY EXPERTS
{f scientific, technical, or other specialized knowledge will assist the trier of fact
to understand the evidence or to determine a fact in issue, a witness qualified as
an expert by knowledge, skill, experience, training, or education, may testify
thereto in the form of an opinion or otherwise.
RULE 703. BASES OF OPINION TESTIMONY
The facts or data in the particular case upon which an expert bases an opinion or
inference may be those perceived by, reviewed by, or made known to the expert at
or before the hearing. If of a type reasonably relied upon by experts in the
particular field in forming opinions or inferences upon the subject, the facts or
data need not be admissible in evidence.
RULE 705(c). DISCLOSURE OF FACTS OR DATA UNDERLYING
EXPERT OPINION
(c) Admissibility of Opinion. If the court determines that the underlying facts or
data do not provide a sufficient basis for the expert’s opinion under Rule 702 or
703, the opinion is inadmissible.
A The Material Properties are reliable and appropriate for comparison
The selection of the material properties by Drs. King and Kasner were appropriate for the
Roensch finite element analysis because it was not an effort to reconstruct the exact events which
lead to the failure of the stripper element on March 3, 2006 but, instead, was meant to determine
the specific region of highest stress and predict the deformation of the rubber in that area while
evaluation the presence of the bond in that location and whether there were more appropriate
locations for that bond in this safety critical tool. In order to be admissible, expert testimony
must be “not only relevant, but reliable.” The tensile modulus of 700 psi was selected where
Roensch wanted to have a range of stresses which rose to 1100 psi and above. Specifically,
Roensch testified that he wanted to look at a higher stress range of 300% - 600%, which
calculates to 2100 psi to 4200 psi. The tensile strength of the rubber in this batch varied widely
but the upper range for that rubber was approximately 3,700 psi. This analysis demonstrated
where the highest loads would be found and helped determine the appropriateness of the bond
location. It further elucidated the issues of whether there was a good bond or a weakened or
absent bond, The linear modeling was selected because it correlated best with the data at the
higher stresses and strains and allowed those ranges and the resultant deformation to be
examined and evaluated.
B. Not an Accident Reconstruction
The Roensch FEA was not intended to be an accident reconstruction. In order to conduct
an accident reconstruction, the analysis would have to take into account each and every variable
which was involved in the subject stripper element failure and explosion. That would
necessarily include the following: temperature; well bore pressure; deterioration of the rubber
after 39.5 hours; alignment variances; and the affect of the well bore fluids. This list is not
exhaustive but it is illustrative. The Roensch FEA was utilized to evaluate the designs of the
subject and the alternative design embodied in the combo stripper element. The criticism of the
Roensch FEA regarding the fact that it did not include cach and every variable is not a valid
criticism where the Plaintiffs readily admit that the purpose was not to conduct such an
investigation. Smith’s objection to the Roensch FEA is stated at page 8 of the pleading thusly,
“The FEA conducted by Roensch should be excluded because (sic) does not model the facts in
this case.” It was never meant to be and neither was the Frishmuth FEA, In fact, Frishmuth who
initially tried to model an Ogden’ 3 parameter three dimensional analysis was unable to complete
the analysis due to the number of variables and complexity of the equation. Roensch sought to
simplify the equation such that the values were useful. When he determined that the values were
appropriate and predictive, he utilized those values.
Cc. Verification Testing Standard is inapplicable, unnecessary and confusing
Verification/confirmation testing is an appropriate step in the design process but it is not
appropriate or required where the goal of the analysis was to determine and evaluate the relative
deformations of the subject stripper element and comparison. The finite element analysis had
limited parameters and those parameters were satisfied. Smith’s insistence on
verification/confirmation testing is interesting because of the efforts of their own retained FEA
expert, Dr. Frishmuth. Frishmuth determined that testing was necessary but then chose not to test
the products utilizing the pressures involved at the drill site. In fact, there is no
verification/testing by either party to confirm whether each and every movement depicted in the
graphical representations of the stripper elements are accurate. In the Roensch FEA such
accuracy at every point on the rubber was unnecessary for the analysis. If the Frishmuth FEA
was designed to replicate the failure sequence then one would expect the graphical representation
4 A non-linear equation involving the use of three parameters. Frishmuth was able to complete the Ogden
one parameter model.
to be precise and accurate but it is not nor could it be. See Exhibit “5”. The FEA is just a tool
that the designer and mechanical engineers use to predict the behavior of the material under
various scenarios but it alone cannot predict the events which took place on March 3, 2006 nor
did the Plaintiffs attempt to utilize it for that purpose. The goal of the Plaintiffs FEA was
achieved such that no further analysis and/or testing were necessary for the opinions regarding
the analysis of the design of the stripper element.
Iv.
The general and specific objections to the Roensch FEA are based upon the
misconception that the analysis should be an accident reconstruction rather than a design
analysis. The Roensch analysis relied upon recognized authorities and material properties to
investigate the relative deformation of certain specific portions of the stripper element where
those specific parts have been previously identified due to the photographs and disassembly.
The solutions fit the direct problem presented and should not be criticized by Smith for failing to
do what the analysis was never supposed to do. While Smith may disagree with the use of the
linear analysis, it is clear that the rationale behind performing a lincar analysis is appropriate for
the limited purposes of the analysis. For the limited investigations performed by King, the linear
finite element analysis provided reliable and predictive values. Smith criticizes King and
Roensch for failing to conduct confirmation or verification tests of alternative design stripper
elements or those stripper elements which would incorporate a different system than the involved
stripper element. King determined that the finite element analysis as well as his experience in
mechanical design and engineering made the testing unnecessary’. Additionally, the testing
5 Please note that Ronald Frishmuth, Ph.D. was retained by Smith to create a finite element analysis.
While Dr. Frishmuth attempted to perform confirmation testing, that testing never achieved or attempted
to achieve a pressure of over 1000 psi which was of critical importance in the failure analysis of the
proposed by Smith’s own expert to “validate” the design of the subject stripper element has
never been conducted according to Frishmuth’s testimony. The Frishmuth testing protocol reads
in pertinent part beginning at page 142 of his deposition:
Q. Do you think that an FEA could investigate the issues he references?
A. I believe if done properly, they probably would help. I’m not sure that’s the end-all and be-all,
but it’s another method—or what should I say, confirmation, or another item to consider.
Let’s put it that way.
Q. In not FEA, then what?
A. Well, I think laboratory testing.
Q. Utilizing what?
A. Utilizing some actual RCDs, and putting on pressure, and pulling pipe through, and
instruments, and measuring, and looking at how stuff really behaves.
Q. How many RCDs would be sufficient, in your mind for the laboratory testing?
A. I would say a fair number. Probably six, at least.
Q. Nondestructive testing or destructive testing?
A. I think you would start off with nondestructive testing. Maybe some ultrasonic—I’m not quite
sure what works in rubber with respect to inspection, but 1 would think every piece that you
wanted to test in the laboratory, you’d want to do a very good pedigree on what it is and what
you're testing. And then I think you would proceed with a destructive test, where you'd
intentionally try to pull this apart in a manner that’s similar to the real failure.
There is no evidence that this is an industry standard or even a practice that has been adopted by
Smith in its defense of this litigation. This Smith criticism in light of its own conduct is
confusing and duplicitous.
WHEREFORE, PREMISES CONSIDERED, Plaintiffs request that the Court deny the
Smith International, Inc.’s Motion to Exclude the Testimony of Steven J. Roensch, and for such
other and further relief to which the Plaintiffs may be entitled under law and equity.
involved stripper element. A test, which leaves out such a critical component of the analysis, does not
test the design.
Respectfully submitted,
Hortman Harlow Bassi Robinson
& McDaniel
P.O. Drawer 1409
Laurel, MS 39441-1409
Ph: (601) 649-8611
f
Fax: (601) 649-606
By: Zo,
CHRISTOPHER BSMCDANIEL
MS State Bar No. 10711
EUGENE M. HARLOW
MS State Bar No. 3086
APRIL C, LADNER
MS State Bar No. 101479
&
KETTERMAN ROWLAND &
WESTLUND
165006 San Pedro, Suite 302
San Antonio, Texas 78232
(210) 490-7402; Telephone
(210) 490-8372; Facgffnile
By: C0
DOUGLAS D. KETFERMAN
State bar No. 11362950
BRIAN C. STEWARD
State Bar No. 19201100
ATTORNEYS FOR PLAINTIFFS
CERTIFICATE OF SERVICE
Thereby certify that a true and correct copy of the above and foregoing has been served
l
via facsimile, on this the ft day of July 2009 to the following:
Mr. J. Chad Parker Mr. Andrew McKinney, Esq.
The Parker Firm, P.C. McKinney & Cooper, L.L.P.
3808 Old Jacksonville Road Three Riverway, Suite 500
Tyler, Texas 75701 Houston, Texas 77056
Mr. Ernest P. Geiger, Jr.
Mr. John E.W. Baay, II
Geiger, Laborde & Laperouse
117 West Loop South, Suite 750
Houston, Texas 77027
ATTORNEY FOR DEFENDANT, SONERRA RESOURCES CORPORATION
John C. Kilpatrick
Law Offices of Kilpatrick & White
2777 Allen Parkway, Suite 622
Houston, Texas 77019
ATTORNEY FOR DEEFNDANT, KIRKHILL MANUFACTURING COMPANY aka KMC
RUBBER PRODUCTS
Ronald Max Raydon
Law Offices of Ronald Max Raydon
1718 Fry Road, Suite 450
Houston, Texas 77084
ATTORNEY FOR DEFENDANT, ADVANCED POLYMER TECHNOLOGIES
Joe Grady Tuck Derek R. Van Gilder
Law Office of Joe Grady Tuck Law Office of Derek R. Van Gilder
1404 Pine Street 916 Main Street
Bastrop, Texas 78602 Bastrop, Texas 78602
ATTORNEYS FOR DEFENDANT, HARBOR PRODUCTS
Mr, Jeffrey S, Davis
Mr. T. Mike Wall
Gardere Wynne Sewell LLP
1000 Louisiana, Suite 3400
Houston, Texas 77002
ATTORNEY FOR DEFENDANT, SMITH INTERNATIONAL, INC.
Mr. E. Wayne Shuffield
Adami, Shuffield, Scheihing & Burns
9311 San Pedro, Suite 900
San Antonio, Texas 78216
ATTORNEY FOR INTERVENOR, NEW HAMPSHIRE INS. CO.
Mr, Chris Evans
Mr. Michael N. Mire
Adams and Reese
One Houston Center
1221 McKinney, Suite 4400
Houston, Texas 77010
ATTORNEY FOR HELMERICH & PAYNE INTERNATIONAL DRILLING CO.
(.
BRIAN C. STEWARD