Burst v. Shell Oil Company et al

Filing 127

ORDER AND REASONS - regarding Defendants Shell Oil Company, Chevron U.S.A. Inc., and Texaco,Inc. MOTION 87 to exclude plaintiff's expert, Richard Miller. Becausethe Court finds that Miller's methodology for calculating Mr.Burst's benzene exposure is unreliable, the Court GRANTSdefendants' motion and excludes the testimony of Miller.. Signed by Chief Judge Sarah S. Vance on 5/14/15.(jjs)

UNITED STATES DISTRICT COURT
EASTERN DISTRICT OF LOUISIANA
YOLANDE BURST, individually
and as the legal
representative of BERNARD
ERNEST BURST, JR.

CIVIL ACTION

VERSUS

NO: 14-109

SHELL OIL COMPANY, ET AL.

SECTION: R

ORDER AND REASONS
Defendants Shell Oil Company, Chevron U.S.A. Inc., and Texaco,
Inc. move to exclude plaintiff's expert, Richard Miller.1

Because

the Court finds that Miller's methodology for calculating Mr.
Burst's

benzene

exposure

is

unreliable,

the

Court

GRANTS

defendants' motion and excludes the testimony of Miller.

I. BACKGROUND
Plaintiff Yolande Burst filed this products liability action
against

defendants

Shell,

Corporation), and Texaco.2

Chevron

(as

successor

to

Gulf

Oil

She alleges that her late husband,

Bernard Burst, Jr., worked at various Shell Oil, Gulf Oil, and
Texaco gas stations from 1958 through 1971, during which time he
regularly used products manufactured, supplied, distributed, and

1

R. Doc. 87.

2

R. Doc. 1.

sold by defendants.3

Specifically, she alleges that he would

regularly come into contact with gasoline containing benzene.
On June 20, 2013, physicians diagnosed Mr. Burst with acute
myeloid leukemia (AML).4

He was 71 years old.

He passed away as

a result of the leukemia on December 21, 2013.5
Plaintiff alleges that her husband's regular exposure to
gasoline containing benzene during the years he worked as a gas
station attendant and mechanic caused his leukemia.6
that

defendants

containing

negligently

benzene

and

that

manufactured
they

and

negligently

She claims

sold
failed

products
to

warn

foreseeable users about the health hazards associated with these
products.7

She also alleges strict products liability.8

As evidence of Mr. Burst's exposure to benzene through his use
of defendants' products, plaintiff relies on an expert report from
Richard Miller, an industrial hygienist.

In the report, Miller

provides an estimate of Mr. Burst's cumulative exposure to benzene
from gasoline while working at a Gulf Oil gas station over a oneyear period between 1966 and 1968.

3

Id. at 3.

4

R. Doc. 28-5 at 18.

5

R. Doc. 28-6.

6

R. Doc. 1 at 5.

7

Id. at 9.

8

Id. at 10.
2

Defendants now move to exclude

Miller's opinions on the ground that they are unreliable and
irrelevant.

II. LEGAL STANDARD
A district court has considerable discretion to admit or
exclude expert testimony under Federal Rule of Evidence 702.

See

General Elec. Co. v. Joiner, 522 U.S. 136, 138-39 (1997); Seatrax,
Inc. v. Sonbeck Int'l, Inc., 200 F.3d 358, 371 (5th Cir. 2000).
Federal Rule of Evidence 702, which governs the admissibility of
expert witness testimony, provides:
A witness who is qualified as an expert by knowledge,
skill, experience, training, or education may testify in
the form of an opinion or otherwise if: (a) the expert's
scientific, technical, or other specialized knowledge
will help the trier of fact to understand the evidence or
to determine a fact in issue; (b) the testimony is based
on sufficient facts or data; (c) the testimony is the
product of reliable principles and methods; and (d) the
expert has reliably applied the principles and methods to
the facts of the case.
Fed. R. Evid. 702.
In Daubert v. Merrell Dow Pharmaceuticals, Inc., the Supreme
Court held that Rule 702 requires the district court to act as a
gatekeeper to ensure that "any and all scientific testimony or
evidence admitted is not only relevant, but reliable." 509 U.S. at
589; see also Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 147
(1999) (clarifying that the Daubert gatekeeping function applies to
all forms of expert testimony).

The Court's gatekeeping function
3

thus involves a two-part inquiry into reliability and relevance.
First, the Court must determine whether the proffered expert
testimony is reliable.

The party offering the testimony bears the

burden of establishing its reliability by a preponderance of the
evidence.

See Moore v. Ashland Chem. Inc., 151 F.3d 269, 276 (5th

Cir. 1998).

The reliability inquiry requires the Court to assess

whether

reasoning

the

testimony is valid.

or

methodology

underlying

the

See Daubert, 509 U.S. at 592-93.

expert's

The aim is

to exclude expert testimony based merely on subjective belief or
unsupported speculation.

See id. at 590.

The Court in Daubert

articulated a flexible, non-exhaustive, five-factor test to assess
the reliability of an expert's methodology. These factors include:
(1) whether the expert's theory can be or has been tested; (2)
whether the theory has been subject to peer review and publication;
(3) the known or potential rate of error of a technique or theory
when applied; (4) the existence and maintenance of standards and
controls; and (5) the degree to which the technique or theory has
been generally accepted in the scientific community.
95.

Id. at 593-

The Supreme Court has emphasized, however, that these factors

"do not constitute a 'definitive checklist or test.'"
U.S. at 150 (quoting Daubert, 509 U.S. at 593).

Kumho, 526

Rather, district

courts "must have considerable leeway in deciding in a particular
case

how

to

go

about

testimony is reliable."

determining

whether

Id. at 152.

Courts have also considered

4

particular

expert

whether experts are "proposing to testify about matters growing
naturally

and

directly

out

of

research

they

have

conducted

independent of the litigation, or whether they have developed their
opinions expressly for purposes of testifying," Daubert v. Merrell
Down Pharms., Inc., 43 F.3d 1311, 1317 (9th Cir. 1995), whether the
expert

has

adequately

accounted

for

obvious

alternative

explanations, see Claar v. Burlington N.R.R., 29 F.3d 499 (9th Cir.
1994), and whether the expert "is being as careful as he would be
in his regular professional work outside his paid litigation
consulting," Sheehan v. Daily Racing Form, Inc., 104 F.3d 940, 942
(7th Cir. 1997).
A district court's gatekeeper function does not replace the
traditional adversary system or the role of the jury within this
system.

See Daubert, 509 U.S. at 596.

As the Supreme Court noted

in Daubert: "Vigorous cross-examination, presentation of contrary
evidence, and careful instruction on the burden of proof are the
traditional and appropriate means of attacking shaky but admissible
evidence."

Id.

The Fifth Circuit has held that, in determining

the admissibility of expert testimony, district courts must accord
proper deference to "the jury's role as the proper arbiter of
disputes

between

conflicting

opinions.

As

a

general

rule,

questions relating to the bases and sources of an expert's opinion
affect the weight to be assigned that opinion rather than its
admissibility and should be left for the jury's consideration."

5

United States v. 14.38 Acres of Land, More or Less Situated in
Leflore Cnty., Miss., 80 F.3d 1074, 1077 (5th Cir. 1996) (quoting
Viterbo v. Dow Chem. Co., 826 F.2d 420, 422 (5th Cir. 1987))
(internal quotation marks omitted).

Nonetheless, expert testimony

"must

every

be

reliable

inadmissible.

at

each

and

step

or

else

it

is

The reliability analysis applies to all aspects of

an expert's testimony: the methodology, the facts underlying the
expert's opinion, the link between the facts and the conclusion, et
alia."

Knight v. Kirby Inland Marine Inc., 482 F.3d 347, 355 (5th

Cir. 2007) (internal quotation marks omitted). "Where the expert's
opinion is based on insufficient information, the analysis is
unreliable."

Paz v. Brush Engineered Materials, Inc., 555 F.3d

383, 388 (5th Cir. 2009).
In Joiner, the Supreme Court explained that "nothing in either
Daubert or the Federal Rules of Evidence requires a district court
to admit opinion evidence that is connected to existing data only
by the ipse dixit of the expert."

522 U.S. at 146.

Rather, "[a]

court may conclude that there is simply too great an analytical gap
between the data and the opinion proffered." Id.; see also LeBlanc
v. Chevron USA, Inc., 396 F. App'x 94, 98 (5th Cir. 2010).
Second,

the

Court

must

determine

whether

the

expert's

reasoning or methodology is relevant. The question here is whether
the reasoning or methodology "fits" the facts of the case and will
thereby assist the trier of fact to understand the evidence.
6

See

Daubert, 509 U.S. at 591.

III. DISCUSSION
Richard
plaintiff.

Miller

is

an

industrial

hygienist

retained

by

Miller attempts to reconstruct Mr. Burst's work duties

during a one-year period during which Mr. Burst worked at a Gulf
Oil gas station between 1966 and 1968, almost 50 years ago, and
then to estimate, based on various models, Mr. Burst's exposure to
benzene as a component of gasoline.

Notably, because Mr. Burst

passed away before the filing of this case, Miller's reconstruction
relies on the testimony of other witnesses.

Yolande Burst, Mr.

Burst's wife, Frank Simpson, a co-worker at the Gulf Oil gas
station, and Charles Bernard, the former owner of the Gulf Oil gas
station, provided testimony on which Miller relies in establishing
Mr. Burst's work hours, work duties, and, ultimately, his exposure
to benzene as a component of gasoline.
The witness testimony indicates that Mr. Burst refueled cars
and worked as a mechanic at a Gulf Oil gas station.

Both Mr.

Simpson and Mr. Bernard testified that, during refueling, gasoline
spills exposed attendants to gasoline.

Mr. Bernard stated that

attendants could even be "bathed" in gasoline as a result of
spills.

Ms. Burst, Mr. Simpson, and Mr. Bernard also testified

that Mr. Burst, as part of his mechanic duties, frequently washed
his hands and parts in a bucket of gasoline located inside the

7

station's garage. This caused Mr. Burst's hands to become wet with
gasoline.

On occasions when Ms. Burst observed Mr. Burst washing

his hands and parts inside the garage, she testified that the fumes
caused her to become lightheaded within 15 minutes.

According to

Ms. Burst, the garage smelled strongly of gasoline in the winter
months when the garage's bay doors remained closed. Ms. Burst also
testified that Mr. Burst smelled strongly of gasoline when he
returned home from work.
Relying

on

the

witness

testimony,

Miller

calculated

Mr.

Burst's exposure to benzene from gasoline from four separate
sources.

First, Miller calculated Mr. Burst's cumulative one-year

exposure to benzene from inhaling vapors that evaporated from the
parts-washing bucket during the winter months when the garage's bay
doors remained closed.

Second, Miller calculated Mr. Burst's

cumulative one-year dermal exposure to benzene from washing parts
in gasoline.

Third, Miller calculated Mr. Burst's cumulative one-

year dermal exposure to benzene resulting from gasoline that soaked
parts of Mr. Burst's clothing.

Finally, Miller calculated Mr.

Burst's cumulative one-year exposure to benzene from inhaling
vapors while washing parts in gasoline.
Miller calculates that Mr. Burst was exposed to 5.499 ppmyears9 benzene through inhaling gasoline vapors that evaporated

9

"Expressing dose in ppm-years is easily understood. If a
person was exposed to 1 ppm of a compound every year for 20
years, his or her total exposure to the compound would be 20
8

from the parts-washing bucket during the winter months, 37.367 ppmyears benzene through dermal exposure while washing parts in
gasoline, 14.723 ppm-years benzene through dermal exposure from
gasoline on clothing, and 4.089 ppm-years benzene through inhaling
gasoline vapors while washing parts in gasoline.10 In total, Miller
calculates that Mr. Burst was exposed to a cumulative dose of
61.678 ppm-years benzene while working for a single year at the
Gulf Oil gas station.11
Defendants now attack both Miller's inhalation and dermal
exposure assessments.

Defendants contend that Miller's estimate

for background inhalation exposure from the parts-washing bucket is
unreliable because the gasoline vapor levels required to expose a
person to that much benzene would be so high that it would be
lethal within a matter of minutes.

Defendants also contend that

Miller's estimate for inhalation exposure during parts washing is
unreliable because Miller, in reaching his conclusion, ignored
relevant data from over 14 peer-reviewed publications and relied
solely on the self-reported symptoms of Ms. Burst.
estimate

of

Mr.

Burst's

dermal

exposure

from

As to Miller's
parts

washing,

defendants contend that the estimate is unreliable because Miller

ppm-years." Baker v. Chevron USA, Inc., 680 F. Supp. 2d 865, 870
n.3 (S.D. Ohio 2010).
10

R. Doc. 87, Ex. A at 37.

11

Id.
9

failed to account for the evaporation of gasoline from Mr. Burst's
hands, and because he did not consider the acute damage to Mr.
Burst's hands that would have likely occurred had his hands been
wet with gasoline for a substantial part of the day.

Defendants

also point out that Miller has never performed a dermal exposure
assessment outside the context of litigation.
A. Miller's Estimate of Mr. Burst's Benzene Exposure from
Inhaling Gasoline Evaporated from the Parts-Washing Bucket
Miller's

estimate

of

Mr.

Burst's

benzene

exposure

from

inhaling gasoline evaporated from the parts-washing bucket is
unreliable because he failed to validate his results against
studies, including at least one cited in his own report, showing
that the corresponding gasoline vapor levels required to expose a
person to that much benzene would be so high as to be lethal within
a matter of minutes.
The record includes conflicting testimony as to the size and
amount of gasoline contained in the parts-washing bucket.

Frank

Simpson testified that he and Mr. Burst used a two-gallon bucket,
about halfway full of gasoline, to clean various parts.12 According
to Simpson, they kept the bucket on a bench inside of the station's
garage13 and needed to refill the bucket two or three times per

12

R. Doc. 105, Ex. 4 at 35.

13

R. Doc. 87, Ex. A at 23.
10

day.14

There was also testimony from Simpson that the bucket was

30% full with gasoline.15

Mr. Bernard testified that they used a

pail with the diameter of a five-gallon bucket that was six inches
high with three or four inches of liquid in it.16

Ms. Burst

testified that the bucket was either two or five gallons and was
filled halfway with gas.17

Based on this testimony, Miller sought

to estimate how much benzene would have evaporated from the bucket
into the surrounding air, and how much benzene Mr. Burst would have
inhaled.

Because Ms. Burst testified that the garage's bay doors

were closed only in the winter months,18 Miller chose to perform
this calculation only for those months in which the garage was
closed.

Also, because Miller separately calculated Mr. Burst's

dermal and inhalation exposure while washing parts in the partswashing bucket, Miller limited this calculation to instances when
Mr. Burst remained two feet or more away from the bucket and was
not cleaning parts.
Citing monthly average high and low temperatures in New
Orleans, Louisiana, Miller assumed that the bay doors remained

14

Id. at 21.

15

R. Doc. 105, Ex. 4 at 35.

16

R. Doc. 87, Ex. A at 20.

17

R. Doc. 105, Ex. 15 at 2.

18

R. Doc. 87, Ex. A at 24.
11

closed throughout December and January.19

From this, and based on

the assumption that Mr. Burst worked 10.5 hours per day, Miller
calculated that there were approximately 1123.50 work hours in
which the bay doors were closed due to cooler weather over a twoyear period.20

Assuming that Mr. Burst spent 40% of those hours in

the garage, Miller calculated that Mr. Burst would have been inside
the closed garage for 449.4 hours over a two-year period.
Assuming that the bucket contained one gallon of gasoline
during the workday, and that the gasoline contained 1% benzene,
Miller calculated that evaporation of all benzene available for
evaporation

into

the

closed

bay

of

the

garage

produced

a

concentration of 50.9 ppm benzene.21 Assuming Mr. Burst was exposed
to this concentration of benzene for 449.4 hours over a two-year
period or 224.7 hours over a one-year period, Miller estimated a
cumulative benzene inhalation exposure of 10.997 ppm-years over two
years or 5.499 ppm-years over one year.22
Defendants

contend

that

Miller's

estimate

is

unreliable

because it is based on Miller's calculation that the garage
contained 50.9 ppm benzene, which could have been possible only

19

Id.

20

Id.

21

R. Doc. 87, Ex. A at 25.

22

Plaintiff has since withdrawn this estimate.
105 at 4.
12

See R. Doc.

with lethal levels of gasoline vapor. Defendants cite Miller's own
report, in which he refers to a study showing that the total
gasoline vapor level for gasoline containing 1% benzene is 166.67
times greater than the benzene vapor level.

Using this ratio,

defendants calculate that a concentration of 50.9 ppm benzene would
require a gasoline concentration of 8449 ppm.

At his deposition,

Miller described defendants' calculation as "absolutely fair."23
Defendants next point to a study, again cited by Miller in his own
report, showing that exposure to as little as 5000 ppm gasoline for
six minutes is lethal.

Yet, Miller's calculation assumes that Mr.

Burst was exposed to a concentration of 8449 ppm gasoline for
hundreds of hours over the course of a year.

There is no evidence

that Mr. Burst or any of his co-workers suffered from acute
overexposure to gasoline.

When confronted with these facts at his

deposition, Miller conceded that his calculation "was a mistake,"
and that such high levels of gasoline vapor would have been
"immediately

dangerous

to

life

and

health."24

Plaintiff

has

withdrawn this calculation and states that Miller will not include
it in his cumulative estimate at trial.25
Even though Miller has now withdrawn his calculation, that he
chose this methodology and included the findings in his report

23

R. Doc. 87, Ex. B at 108.

24

Id. at 111.

25

Id.
13

remains relevant because it informs the Court about Miller’s
overall approach to choosing and analyzing data.

See Knight, 482

F.3d at 355 (The expert's report "must be reliable at each and
every step or else it is inadmissible."); see also Castellow v.
Chevron USA, 97 F. Supp. 2d 780, 788, 791 (S.D. Tex. 2000)
(excluding expert's exposure assessment opinion in part because one
of his calculations necessitated a lethal level of gasoline even
though the expert later withdrew the opinion).

Miller made his

calculations in a vacuum without any attempt to validate his
results against reality.
pointed

out

his

error,

Specifically, until defense counsel
Miller

failed

to

recognize

that

the

evidence, which provided no indication Mr. Burst or his co-workers
sustained acute overexposure to gasoline, was not consistent with
his result.

Moreover, Miller also failed to validate his results

against existing scientific literature, including a study on which
he relied elsewhere in his report, that demonstrated why his result
had to be wrong.

These failings speak to a lack of intellectual

rigor in Miller’s approach to the issues under discussion.
Miller now tries to rationalize his approach by stating that
three components of gasoline would have evaporated faster than
benzene, but that benzene evaporates faster than “most” other
components of gasoline, which he does not identify.

But Miller

cannot deny that the other components of gasoline would have

14

evaporated at some rate.26

He now assumes that everything that

evaporated, except the benzene, escaped the garage through leaks.
Yet, in his report, he assumes that the room was airtight with “no
significant ventilation” that would have permitted the benzene to
escape over the course of “subsequent work days.”27
to

explain

how

he

can

have

it

both

ways.

Miller fails
His

post

hoc

rationalization does not salvage his methodology on this issue.
B. Miller's Estimate of Mr. Burst's Benzene Exposure from
Dermal Exposure during Parts and Hands Washing
Several witnesses have testified that Mr. Burst frequently
washed his hands and parts in a bucket of gasoline during the
workday.

For example, Ms. Burst testified that Mr. Burst would

"remove a part that was soaking in gasoline from a bucket with his
bare hands, and then wipe the part with a shop rag to remove the
grease from the part."28 Mr. Simpson also testified that he and Mr.
Burst cleaned parts in gasoline while performing mechanic work.
Specifically, Mr. Simpson stated that he spent approximately 15% to
20% of his day cleaning parts in a two-gallon bucket approximately
halfway full of gasoline.29

Mr. Bernard testified that the bucket

26

In his report, Miller assumes that “two or possibly three
gallons of gasoline-benzene mixture may have evaporated [from the
parts-washing bucket] over the course of a day.” R. Doc. 87, Ex.
A at 25.
27

Id.

28

Id. at 18.

29

Id. at 27.
15

contained "three or four inches" of gasoline.30

Mr. Simpson also

testified that he and Mr. Burst washed their hands in gasoline
after performing mechanic work.31 Mr. Simpson, however, stated that
he "probably cleaned [his] hands a lot more than Mr. Burst," and he
did not expressly state what percentage of time Mr. Burst spent
washing parts.32

Still, Mr. Simpson stated that cleaning his hands

was "a constant process" because he needed to alternate between
mechanic work and pumping gas.33

Charles Bernard also testified

that on most days Mr. Burst would have performed mechanic work
necessitating that he wash his hands in gasoline multiple times per
day.34 From this testimony, Miller concluded that "it is clear that
the attendants were constantly washing their hands in gasoline."35
Miller sought to calculate Mr. Burst's cumulative dermal
exposure--how much benzene Mr. Burst absorbed through his skin-from washing parts and his hands in gasoline over the course of one
year.

At his deposition, Miller conceded that he has never

30

Id. at 20. Miller noted that Mr. Simpson and Ms. Burst
testified that the bucket contained one gallon of gasoline. Id.
31

Id.

32

Id.

33

Id.

34

Id. at 19.

35

Id. at 28.
16

calculated a dermal dose outside of the context of litigation.36
In making his assessment, Miller relied on studies describing the
flux rate of pure benzene through human skin.37

Miller stated that

he increased the flux rate described in the literature he cited by
a magnitude of five because Ms. Burst testified that Mr. Burst had
nicks and scratches on his hands from mechanic work.38

But without

stating why, Miller did not incorporate the increased flux rate
into his final calculation.
Despite the divergent testimony as to the size and amount of
gasoline in the parts-washing bucket,39 Miller assumed that Mr.
Burst submerged both his hands and forearms, up to his elbows, in

36

R. Doc. 87, Ex. B at 8.

37

R. Doc. 87, Ex. A at 28. It is unclear whether the flux
rate of pure benzene is comparable to the flux rate of benzene as
a component of gasoline.
38

In support of his decision to increase the flux rate,
Miller relied on a paper, Maibach, HI, et al., Percutaneous
Penetration of Benzene and Benzene Contained in Solvents Used in
the Rubber Industry, 36 ARCH ENVIRON HEALTH 256 (1981), showing
that the flux rate of pure benzene through the artificially
damaged skin of Rhesus monkeys increases significantly. In this
study, the researchers stripped away the stratum corneum from the
monkeys' skin before applying benzene and measuring its flux
rate.
39

R. Doc. 105, Ex. 4 at 35 (deposition of Mr. Simpson)
(stating that they used a two-gallon bucket halfway or 30% full
of gasoline); R. Doc. 87, Ex. A at 20 (testimony of Mr. Bernard)
(stating the they used a pail with the diameter of a five-gallon
bucket that was six inches high with three or four inches of
gasoline in it); R. Doc. 105, Ex. 15 at 2 (testimony of Ms.
Burst) (stating that the bucket was either two or five gallons
and was filled halfway with gas).
17

gasoline "[b]ecause of the depth of the bucket and the amount of
gasoline (1 gallon) involved."40

This assumption is significant as

Miller's calculation used the surface area of exposed skin as a
variable, and this assumption resulted in a significantly larger
surface area of exposed skin than if Miller assumed Mr. Burst
submerged only his hands in gasoline.41
With these figures, Miller ran two "Monte Carlo" simulations.
A Monte Carlo simulation is "a risk assessment model that accounts
for variability and uncertainty in risk factors," such as the
variation in the time of Mr. Burst's dermal exposure to gasoline.
Schultz v. Akzo Nobel Paints, LLC, 721 F.3d 426, 428 (7th Cir.
2013).

The

methodology

Environmental

for

exposures.

evaluating

Protection
risk

Agency

arising

from

endorses

this

environmental

See id. (highlighting "the EPA's position that such

probabilistic analysis techniques as Monte Carlo analysis, given
adequate supporting data and credible assumptions, can be viable
statistical tools for analyzing variability and uncertainty in risk
assessments"
Guiding

(quoting

EPA,

Principles

Office
for

of

the

Monte

Scientific
Carlo

Advisor,

Analysis,

http://www.epa.gov/raf/pub lications/guiding-monte-carloanalysis.htm)).
estimates

by

The simulation "creates a large number of model
selecting

alternative

40

R. Doc. 87, Ex. A at 29.

41

See id.
18

values

for

the

model's

assumptions."

Hammes v. Yamaha Motor Corp. U.S.A., No. 03-

6456(MJD/JSM), 2006 WL 1195907, at *9 (D. Minn. May 4, 2006). "The
assumption values are selected from distributions of likely values
which are specified by the analyst."

Id.

The assumption values

take the form of a range using all possibilities between a minimum
and a maximum value for whatever variables are uncertain.42

The

completed simulation produces a range of results based on the
random input values, each with a corresponding likelihood.

For

example, if the model generated a particular result during only 30%
of the simulations, there is only a 30% chance that that result
will occur in an individual trial.

The model “is particularly

useful when reaching an exact numerical result is impossible or
infeasible and the data provide a known range--a minimum and a
maximum,

for

example--but

leave

the

exact

answer

uncertain.”

Lyondell Chem. Co. v. Occidental Chem. Corp., 608 F.3d 284, 293
(5th Cir. 2010).
Miller's first simulation included two uncertain variables,
and thus included two ranges: (1) the number of hours Mr. Burst's
forearms and hands experienced dermal exposure to benzene from
parts washing, and (2) the flux rate of benzene.43

42

For hours

For example, if an analyst were to use a Monte Carlo
analysis to estimate the range of possibilities of the time of
completion of a project, the analyst would establish a range
using the predicted minimum and maximum amount of time it will
take to complete the project.
43

R. Doc. 87, Ex. A at 30.
19

exposed--that is, hours in which Mr. Burst's hands and forearms
were actually wet with gasoline--Miller assumed a minimum contact
time of 375 hours (1.25 hours per day for 300 days per year) and a
maximum contact time of 3150 hours (10.5 hours per day for 300 days
per year).44 Miller based the minimum value on his assumption that,
at a minimum, Mr. Burst cleaned parts five times per day in 15minute intervals.

This assumption is contrary to Mr. Bernard's

testimony that Mr. Burst performed mechanic work and washed his
hands and parts in gasoline on most days, not all days.

Miller

based the maximum value on his assumption that, at most, Mr.
Burst's hands and forearms were wet with gasoline 10.5 hours per
day, the equivalent of an entire workday.45

This assumption is

based on Mr. Simpson's testimony that the attendants engaged in a
"constant

process"

of

alternating

completing mechanic work.

between

refueling

cars

and

For the benzene flux rate, Miller used

a range of 0.099 mg/cm2/hr to 1.85 mg/cm2/hr,46 which represented the
low and high flux rates observed in the scientific literature cited
by Miller on the flux rate of pure benzene.

44

In a second Monte

Id.

45

Q: You assumed that Mr. Burst's hands, on the high end
of your Monte Carlo, were wet on average from the
second he walked in the door until the second he left?
A: Yes.
Q: All 10.5 hours?
A: Yes.
R. Doc. 105, Ex. 2 at 60 (deposition of Richard Miller).
46

R. Doc. 87, Ex. A at 30.
20

Carlo simulation, Miller kept the hours and flux rate ranges the
same, and added another uncertain variable: the concentration of
benzene in the gasoline. Miller used a 1% benzene concentration as
the minimum of the range and a 2% benzene concentration as the
maximum of the range.47
percentile

(more

likely

The first simulation yielded a 45th
than

not)

cumulative

dermal

benzene

exposure for parts washing of 37.367 ppm-years and the second
simulation yielded a 45th percentile dermal benzene exposure of
53.861 ppm-years.48
The Court finds Miller's estimate of Mr. Burst's dermal
exposure to benzene from washing parts unreliable because Miller’s
assumption that Mr. Burst’s hands and forearms were wet with
gasoline for a minimum of 1.25 hours per day and a maximum of 10.5
hours per day is inconsistent with the factual record and failed to
account for evaporation.
First, the Court finds Miller's methodology unreliable because
Miller's assumption regarding the minimum and maximum time that Mr.
Burst's hands and forearms could have been wet with gasoline is not
supported by the factual record. See Moore v. International Paint,
L.L.C., 574 Fed. Appx. 513, 515 (5th Cir. 2013) ("When an expert's
testimony is 'not based upon the facts in the record but on altered

47

Id. at 30.

48

Id. at 31. This means that in an individual trial, it is
expected that these results will occur 55% of the time.
21

facts and speculation designed to bolster [a party's] position,'
the trial court should exclude it.") (quoting Guillory v. Domtar
Indus., Inc., 95 F.3d 1320, 1331 (5th Cir. 1996)).

Contrary to

Miller's assumption, the witness testimony demonstrates that Mr.
Burst's hands were wet with gasoline far less than 10.5 hours per
day, and, on some days, were not exposed to gasoline through parts
washing at all.

Mr. Simpson testified that he personally spent

only 15% to 20% of his day washing parts.

Similarly, Mr. Bernard

testified that Mr. Burst would have needed to utilize the partswashing bucket to wash his hands on most days, not on all days.
There is no testimony indicating how long Mr. Burst submerged his
hands and forearms in gasoline while cleaning parts.

The only

piece of evidence that could possibly support the inference that
Mr. Burst's hands were wet with gasoline all day is Mr. Simpson's
testimony that, for him, hand washing was a "constant process"
because he had to run in and out of the mechanic's area between
refueling cars, although he stated that he probably washed his
hands more than Mr. Burst.

In his unsworn affidavit, Miller

explained that he interpreted this testimony to mean that the "skin
exposure was intermittent, but repeated each time a car entered the
pump area--making the exposure essentially continuous."49 Likewise,
in his report, Miller stated: "This amounted to a skin exposure

49

R. Doc. 105, Ex. 3 at 5.
22

that was essentially constant."50 Miller's assumption of continuous
exposure from a factual record that shows that the exposure was
intermittent is problematic, especially as Mr. Simpson testified he
spent only 15% to 20% of the day washing parts.

Miller's reliance

on Simpson's use of the word "constant," which Simpson did not use
to describe the length of time Mr. Burst's hands were actually wet
with gasoline, to reach a conclusion with no other factual support
suggests a methodology that is result driven.
By the same token, Miller’s selection of 1.25 hours as Mr.
Burst’s minimum daily exposure is suspect.

Mr. Bernard testified

that Mr. Burst would have experienced this exposure on most days,
not on all days.

Given that a Monte Carlo simulation attempts to

gauge all possibilities within a range of all potential periods of
exposure, Miller inexplicably failed to account for zero hours as
the minimum of the range.
Second, Miller’s opinion is unreliable because his report does
not account for the effect of evaporation in assuming that Mr.
Burst's hands could have been wet with gasoline for a minimum of
1.25 hours per day and a maximum of 10.5 hours per day.

Jennifer

Sahmel, defendants' expert industrial hygienist, opines that, as a
result of evaporation, Mr. Burst would have had much shorter
lengths of dermal exposure to gasoline than estimated by Miller.51

50

R. Doc. 87, Ex. A at 26.

51

R. Doc. 87, Ex. C at 88.
23

In support, Sahmel points to studies showing that "far less than 1%
of

benzene

that

absorption."52

is

applied

to

the

skin

is

available

for

Although not in his report, Miller stated at his

deposition that he separately calculated that it would "take about
a minute or a little longer for the [gasoline] to evaporate" from
Mr. Burst's hands.53

Explaining why he did not account for the

effect of evaporation in this section of his report, Miller
speculated that the perspiration likely present on Mr. Burst's
hands or the humidity likely negated the effect of evaporation.
Unfortunately,

Miller cites no specific facts or data to support

these factual assumptions.

There is no evidence that Mr. Burst's

hands and forearms perspired constantly or that humidity always
prevented evaporation of gasoline from his skin.

Moreover, Miller

assumes elsewhere in his report that for two months a year, it was
too cold every single day to leave the garage doors open.

This

assumption is inconsistent with the assumption that Mr. Burst's
hands and forearms were covered in perspiration those same days.
That Miller never dealt with the effects of evaporation in his
report and then engages in this type of speculation to explain away
a potentially inconvenient phenomenon contributes to the overall
unreliability of his analysis.
Miller's failure to treat evaporation in this segment of his

52

Id.

53

R. Doc. 87, Ex. B at 34.
24

report is telling because Miller discusses and even accounts for
evaporation elsewhere in his report.

For example, Miller stated

that "both gasoline and benzene evaporate into the surrounding
environment."54 Miller also opined that "it would be common to have
gasoline evaporate from the pavement," which could be inhaled by
gasoline station attendants.55

Miller even calculated how much

gasoline would have evaporated from the parts-washing bucket over
the course of a day.56 This inconsistency is especially problematic
from

a

reliability

standpoint

because

Miller

accounted

for

evaporation when it increased his cumulative exposure estimate as
with his inhalation exposure estimate for the parts-washing bucket,
but did not account for evaporation in this section of his report
when it would likely have decreased his estimate significantly.57
The Court has considered the explanation that Miller supplied
in his deposition that the Monte Carlo simulation accounts for
uncertainties like evaporation.

As the Court understands Miller's

explanation, Miller asserts that because he used a range (1.25
hours/day to 10.5 hours/day) and because the Monte Carlo simulation

54

R. Doc. 87, Ex. A at 13.

55

Id. at 17.

56

Id. at 22.

57

See R. Doc. 87, Ex. C at 89 where Sahmel notes that "it
is contradictory to assume that all benzene evaporated when
considering inhalation exposure, but to completely ignore
evaporation when considering dermal exposure."
25

randomly chooses a point within this range during each trial, the
Monte Carlo simulation accounted for days when exposure would have
been less, such as when evaporation caused Mr. Burst's time of
exposure to be lower.
Miller

did

not

Nevertheless, it cannot be denied that

consider

the

effects

of

evaporation

when

establishing the range, and that the scope of the range impacts the
final result.

For example, Miller’s assumption that Mr. Burst’s

hands and forearms were wet with gasoline for a minimum of 1.25
hours

per

day

did

not

account

for

evaporation.

Had

Miller

considered the effect of evaporation, he likely would have selected
a lower minimum input, which would have reduced the final exposure
estimate.58

Given that the Monte Carlo simulation will yield

different results as a result of differences in chosen inputs, "it
stands to reason that if the data from which [Miller's] modeling
assumptions arise is invalid, or non-existent, then there is no
hope that his technique, much less his results, is going to be
reliable."

Castellow, 97 F. Supp. 2d at 792.

See also Lyondell

Chem. Co., 608 F.3d at 294 (citing cases that “stand for the
proposition that Monte Carlo analysis is unreliable when injected
with faulty inputs” and noting that courts can “gauge reliability
by

examining

input

values

and

requiring

transparency

from

testifying experts”); In re Application of Erie Blvd. Hydropower

58

Likewise, had Miller considered Mr. Bernard’s testimony
that Mr. Burst cleaned his hands and parts on most days, not on
all days, he likely would have selected a lower minimum input.
26

L.P. v. Town of Ephratah Bd. of Assessors, No. 17-1-2000-0331, 2003
WL 21172636, at *4 (N.Y. Sup. Ct. Apr. 11, 2003) (“[A]ll you are
doing in a Monte Carlo simulation is coming back to your own
assumptions, so whatever went in comes out. Stated differently, if
you make bad assumptions, you obtain bad outputs.”).
C. Miller Estimate for Inhalation Exposure from Washing Parts
Miller's

estimate

for

Mr.

Burst's

benzene

exposure

from

inhaling gasoline vapors while washing parts is unreliable because
Miller relied solely on the self-reported symptoms from Ms. Burst
from almost 50 years ago while failing to show that this is a
reliable methodology, and failed to validate his results against
scientific literature measuring actual exposure levels.
To calculate Mr. Burst's inhalation exposure from washing
parts, Miller started by determining the concentration of gasoline
vapors present during this activity.

Miller relied on Ms. Burst's

testimony that she watched Mr. Burst wash parts on approximately
two days per week, and that she became "lightheaded from the
gasoline vapor in the room" after fifteen minutes.59

Miller then

cited a study showing that exposure to a concentration of 3000 ppm
gasoline vapor for fifteen minutes can cause "slight dizziness and
irritation of the eyes, nose and throat."60
testified

that

she

experienced

59

R. Doc. 87, Ex. A at 18.

60

Id.
27

Because Ms. Burst

lightheadedness

after

fifteen

minutes, Miller concluded that "it is reasonable to assume that
both she and her husband were exposed to 3000 parts per million
gasoline vapor."61 Yet, there is no evidence in the record that Mr.
Burst

ever

experienced

any

performing any of his duties.

symptoms

while

washing

parts

or

Using this figure, Miller relied on

a study showing that 300 ppm gasoline containing 1% benzene is
associated with 1.8 ppm benzene in the air.62

Because Miller

assumed that Mr. Burst was exposed to 3000 ppm of gasoline, based
on his wife's self-reported symptoms recalled almost 50 years after
the fact, Miller simply multiplied the results of that study by ten
to reach an estimate of 18 ppm benzene.63
Miller then calculated how often Mr. Burst experienced this
exposure. Mr. Simpson testified that he and Mr. Burst spent 40% of
their time performing mechanic work, and that he spent 15% to 20%
of the time cleaning parts.64
may

have

61

performed

mechanic

Mr. Bernard testified that Mr. Burst
work

necessitating

parts

washing

Id.

62

Id. at 19. This translates to a 166.67:1 ratio between
gasoline and benzene.
63

Id. Because 3000 divided by 300 is 10, Miller multiplied
the study's figure, 1.8 ppm, by 10 to reach 18 ppm. At his
deposition, Miller explained that he could extrapolate the data
from this study because the ratio of gasoline to benzene when the
benzene concentration is 1% is linear. In doing so, Miller
assumed that the gasoline to which Mr. Burst was exposed
contained 1% benzene.
64

Id. at 19-20.
28

multiple times per day.65

From this testimony, Miller assumed that

Mr. Burst was exposed to 18 ppm benzene at least five times during
each day for durations of 15 minutes each.66

Assuming a 10.5 hour

workday, a benzene concentration of 1% in gasoline, and that Mr.
Burst spent 15% of his day cleaning parts, Miller estimated that
Mr. Burst's cumulative inhalation exposure from parts washing was
4.089 ppm-years.67
Defendants assert that this estimate is unreliable because
Miller ignored the relevant data from the scientific literature,
and instead relied on the self-reported symptoms of Ms. Burst to
calculate exposure.

In her critique of Miller, Sahmel states that

the "effect threshold of gasoline," especially as reported by Ms.
Burst, a secondary source, not Mr. Burst himself, is not a reliable
indicator of the airborne concentration of gasoline/benzene.68

She

explains:
Mr. Miller's use of a second party’s reported symptoms of
lightheadedness or dizziness recalled from an exposure
that occurred decades earlier to the spouse of the worker
to estimate the inhalation exposure to benzene during
washing parts with gasoline is not an appropriate
exposure assessment methodology.
Odor and irritation
thresholds are also not recognized in the field of

65

Id. at 19.

66

Id. at 20.

67

Id. at 21. Assuming Mr. Burst cleaned parts for 20% of
the day, Miller estimated a cumulative inhalation exposure from
parts washing of 5.45 ppm-years. Id.
68

R. Doc. 87, Ex. C at 80.
29

industrial hygiene or risk assessment as reliable methods
of estimating exposure to airborne chemicals in the
workplace.
This type of approach is considered
unreliable because there is significant intra- and
inter-human variability in detection and perception of
odors at different concentration levels (AIHA 1989;
Keller 2007). Research has shown that the presence of an
odor can result in self-reported symptoms unrelated to
the chemical itself, such as reports of irritation
following exposure to phenylethyl alcohol (PEA), which
has a detectable odor, but no ability to cause irritation
(Dalton 2001). Williams and Lees-Haley (1997) have shown
in a volunteer survey that the stated presence or absence
of a gasoline odor is likely to influence people’s
assumptions about causality. The authors concluded that
when an odor is present, “headache, sore throats,
sleeplessness, and other minor preexisting conditions
could be attributed erroneously to toxic exposure,
especially if uncertainty surrounded the initial
consideration of etiology (p. 416).” (Williams 1997).69
She asserts that Miller should have instead relied on collection of
air concentration data, an evaluation of published literature on
measured

airborne

chemical

exposure

chemical

concentrations,

concentrations

based

on

or

estimation

accepted

models

of
or

calculations.70
In response, Miller does not cite any source indicating that
his

methodology

methodology.

is

accepted

or

any

study

that

utilized

his

When asked whether OSHA endorses his methodology,

Miller stated that "OSHA has nothing to say about it one way or the
other."71

Miller also seemed to indicate that OSHA would not make

69

Id. at 81.

70

Id. at 82.

71

R. Doc. 105, Ex. 2 at 119.
30

any exposure assessment conclusions from self-reported symptoms
alone.72

At the same time, Miller asserted that an industrial

hygienist must pay attention to this information.73 Notably, Miller
could not cite any literature identifying his methodology as one
that is reliable, and he even admitted that he did not conduct any
research prior to completing his report to determine whether his
methodology had been criticized in the published literature.74 Even
after reading defendants’ expert’s report, Miller stated that he
had not reviewed the studies she cited showing why his methodology
is unreliable.75

Miller insists that he relied on scientific

literature--the studies examining symptoms resulting from varying
concentrations

of

gasoline

vapor,

but

these

studies

observed

symptoms resulting from known quantities of gasoline; they did not
calculate gasoline vapor levels from symptoms.

Ultimately, there

is no evidence in the record besides Miller's own assurances to
suggest

that

non-contemporaneous

self-reported

symptoms

of

a

secondary source can form the sole basis for an exposure assessment
opinion.

In the face of Sahmel's criticism, which is supported by

72

"And OSHA does not cite [] an estimated exposure by
somebody who said I have watery eyes in styrene; it's 300 parts
per million. OSHA is not going to cite that, but as an
industrial hygienist, you have to pay attention to it." Id. at
119-20 (deposition of Richard Miller).
73

Id.

74

Id. at 122.

75

Id. at 123.
31

multiple studies, and nothing but the ipse dixit of Miller, Miller
fails to show that his methodology is reliable.

See Joiner, 522

U.S. at 146.
Plaintiff cites Curtis v. M&S Petroleum, Inc., 174 F.3d 661
(5th Cir. 1999), for the proposition that reliance on self-reported
symptoms is an accepted methodology, but this case is inapposite.
There, the Fifth Circuit endorsed an expert's methodology when the
expert relied on not only contemporaneously self-reported symptoms
of

the

workers,

but

also

on

actual

contemporaneous

exposure

measurements, in addition to several other factors. Id. at 671-72.
Unlike the methodology examined in Curtis, Miller's methodology is
unreliable because he relied solely on self-reported symptoms to
formulate his estimate.

Moreover, Miller relies on the self-

reported symptoms from a secondary source from almost 50 years ago.
This is not to suggest that reasonably contemporaneous selfreported symptoms are irrelevant or that they cannot form part of
the basis for an expert exposure opinion. But, here, Miller relies
solely on self-reported symptoms from a secondary source from
almost 50 years ago and there is no other evidence in the record
that would buttress the reliability of this methodology.76

76

Plaintiff also cites Arabie v. CITGO Petroleum Corp., 89
So. 3d 307, 321 (La. 2012), to support her argument that Miller's
sole reliance on Ms. Burst's self-reported symptoms is reliable.
In Arabie, however, the issue of whether an expert's reliance on
self-reported symptoms is a reliable methodology was not before
that court.
32

Miller's methodology is also unreliable because he failed to
validate his result against any study that measured actual exposure
levels.

While validation against such studies is likely not

necessary in every case, other courts have recognized that in the
presence of comparable scientific data measuring actual exposures,
an

expert,

at

the

very

least,

should

validate

an

exposure

assessment based on modeling against the scientific literature.
See, e.g., Castellow, 97 F. Supp. at 791 (excluding an expert's
exposure assessment partially because he failed to validate his
modeling

assessment

with

scientific literature).
papers

and

articles

comparable

monitoring

data

from

the

Miller himself states that "peer reviewed

should

augment

and

elucidate

eyewitness

testimony and data associated with specific occupational activity
. . . ."77

Here, it appears that such validation would have been

especially important because Miller relied solely on the selfreported symptoms of a single individual, a secondary source, from
almost 50 years ago, and there is no evidence that Mr. Burst
experienced the same symptoms.

Moreover, as cited by Sahmel,

numerous studies exist that measured actual exposure levels of
gasoline station service attendants performing the same tasks as
Mr. Burst.
demonstrate

In fact, one of the studies cited by Miller to
what

symptoms

result

from

varying

gasoline

vapor

concentrations, Runion (1975), separately found, as explained by

77

R. Doc. 105, Ex. 3 at 4.
33

Sahmel,

that

inhalation

exposure

from

parts

washing

was

substantially less than Miller's estimate (0.37 ppm compared to
Miller's estimated 18 ppm).78

Despite citing this study for a

separate reason multiple times in his report, Miller offers no
explanation for why he failed to acknowledge its measurement of
actual

exposure

performed.

levels

during

the

exact

activity

Mr.

Burst

Just as Miller failed to validate his other inhalation

estimate against studies showing that gasoline concentrations above
5000 ppm are lethal after only five minutes of exposure, Miller
failed

to

validate

this

inhalation

78

exposure

estimate

against

See R. Doc. 87, Ex. C at 83 where Sahmel states that
Miller failed "to consider full data sets cited in his own report
indicating that the benzene exposure concentration during
gasoline parts washing is most likely less than 1 ppm. The
Runion (1975) study that he cited stated that 'Limited tests in
an enclosed room indicated that when such work was being done,
100 ppm total hydrocarbons would be reached frequently in the
breathing zone in the absence of positive ventilation p. 340.'
Specifically, Gulf No-Nox gasoline was found to contain 1.10%
benzene by volume (Runion 1975). It was also found that the
benzene volume percentage in the vapor phase for Gulf No-Nox
gasoline was 0.37%. Assuming that 100 ppm was the total gasoline
vapor concentration during parts cleaning, the benzene
concentration during parts cleaning would be approximately 0.37
ppm, not 18 ppm as estimated by Mr. Miller based on reported
symptoms of dizziness from Mrs. Burst when she visited the garage
at the end of the work day." Neither Miller nor plaintiff
challenge this opinion.
Plaintiff contends that Sahmel ignored the results of Runion
(1982), which reported gasoline concentration ranges of 150-240
ppm for brushing parts in gasoline, 300-425 ppm for gear removed
to work table to dry, and 500-700 ppm for doors open creating a
draft. See R. Doc. 105, Ex. 5 at 44. Applying Sahmel’s analysis
to these results, however, still demonstrates benzene
concentration levels significantly less than that estimated by
Miller.
34

studies

measuring

actual

exposures.

Instead,

Miller

accepts

witness testimony outright without any attempt to validate his
results against scientific literature.
Again, Miller asserts that "published literature was used:
Tironi--which in turn referenced Runion's papers which in turn
referenced various exposure scenarios involving dizziness within 15
minutes."79

As stated, however, Miller relied on these studies for

their observation of symptoms associated with known concentrations
of gasoline vapors.

Miller never compared his result to any

measurements of actual exposure levels, including that provided in
Runion.
E. Conclusion
After a review of Miller's report, the witness testimony, and
the parties' exhibits and briefing, the Court concludes that
Miller's opinions on the nature of Mr. Burst's exposures are
undermined by their reliance on speculation. Miller's opinions are
not based on adequate data and instead demonstrate an effort to
produce particular results and support a causation opinion without
a reliable basis.

While Miller relies on witness testimony, in

significant instances he does so unreasonably and in a manner
intended to raise his exposure assessment.

Moreover, Miller makes

significant assumptions in some instances without any factual basis
for doing so.

79

He ignores evaporation when it is harmful to his

R. Doc. 105, Ex. 3 at 6.
35

assessment, but accounts for it when it is helpful.

Compounding

Miller's unreasonable and often unfounded assumptions is Miller’s
failure to engage in any critical evaluation of his modeling
results against empirical scientific literature measuring actual
exposure

levels.

Instead,

Miller

accepts

witness

testimony

outright and only selectively chooses when to rely on scientific
literature.

Cumulatively,

Miller's

methodology

produces

an

exposure assessment that is likely artificially high and that is
not

reasonably

speculation.

based

on

the

factual

record

but

instead

on

For all of these reasons, the Court finds that

Miller's opinion as to Mr. Burst's exposure to benzene from
defendants' products is unreliable and is therefore inadmissible.

IV. CONCLUSION
For the foregoing reasons, the Court GRANTS defendants' motion
to exclude the report and the testimony of Richard Miller.

New Orleans, Louisiana, this 14th day of May, 2015.
___

____________________________________
SARAH S. VANCE
UNITED STATES DISTRICT JUDGE

36