BP Macondo Estimate 2

Exhibit 1
(Part 2 of 2)
Case 2:10-md-02179-CJB-SS Document 10533-3 Filed 06/27/13 Page 1 of 106
MJBluntExpertReport
104
AppendixD. Additionalpressureanalysis
This Appendix provides additional results and the equations used in the analysis of the measured
pressuredatadiscussedinSection4.3.
D.1. Parametermatches,sensitivitiesandmodelcomparisons
D.1.1Definitionsandbestmatchvalues.TableD.1summarizesthepropertiesthatwecanderivefrom
the pressure analysis. They are all introduced in Appendix C. Table D.2 provides my values of these
propertiesusingthemethodspresentedinAppendicesCandEandSection4.3.
Parameter Definingequation Meaning Foundfrom
q
p =
p
0
4nKb
(C.11)
Drawdowninradialflow.Usesheight
hatthewell(93ft),andfinalflow
rate.
Valueofpressure
derivative
t
W
w
=
ow
2
16

(C.15)
Timeforpressuretoseechannel
width
Endofradialflow
regime
t
1

1
= oI
1
2
(C.59)
Timeforpressuretoreachchannel
end
Matchtochannelflow
pressurebuildup
t
2

2
= oI
2
2
(C.59)
Timeforpressuretoreachtheother
channelend
Matchtochannelflow
pressurebuildup
|
[ =
2p
o
Kbwno
=
p
o
2n
w
Kb
(C.38)
Drawdowninlinearflow Slopeofchannelflow
pressureregime
Ap p
=
n
2
[
(
1
+
2
)
t
p
(C.66)
Pressuredecline Finalstabilized
pressure
Table D.1. Table of parameters that will be determined from fitting the analytical model in Appendix C to the measured
pressureresponse.
Parameter
t
1
t
2
| Ap p
f
p
csf
Pressure
error
Derivative
error
CoreLabsdata 11,300,000 115,000 0.8011 1,423 10,433 7,040 2.03 12.00
Schlumbergerdata 12,000,000 87,900 0.7589 1,325 10,531 7,066 1.84 11.78
Intertekdata 11,800,000 93,700 0.7713 1,354 10,502 7,041 1.89 11.85
Unit s s psi.s
1/2
psi psi psi psi psi
Table D.2. Parameters determined from the pressure match to the linear flow model. Table D.1 provides definitions and
descriptions of each of these terms. The errors are the rootmeansquare errors Eqs. (D.2), (D.3). p
csf
is the final capping
stackpressure,computedusingthemethodsinAppendixBoncethetemperatureinthewellboreisthegeothermalgradient
seeFigureD.6.

MJBluntExpertReport
105
Parameter q t
W
t
1
t
2
| Ap p
f
Pressure
error
Derivative
error
CoreLabs
data
55.6 60,873 10,226,054 190,423 0.79939 1,444 10,412 1.37 11.14
Schlumberger
data
53.4 61,791 11,020,620 165,194 0.76131 1,341 10,515 1.13 11.09
Intertekdata 54.4 62,177 10,809,080 172,072 0.77329 1,371 10,485 1.37 11.14
Unit psi s s s psi.s
1/2
psi psi psi psi
TableD.3.Parametersdeterminedfromthepressurematchtoarectangularflowmodel.TableD.1providesdefinitionsand
descriptionsofeachoftheseterms.TheerrorsaretherootmeansquareerrorsEqs.(D.2),(D.3).
ThepressuredataispresentedinFigure4.10as79distinctvalues.
217
Itisthepressuremeasuredatthe
capping stack from shutting in the well to just before the well was cemented.
218
The pressure in the
reservoirdeepundergroundisestimatedusingtheconversionpresentedinAppendixB:Iwilltreat
this as the measured reservoir pressure in my analysis. I have three conversions giving three sets of
pressurematchesforthethreesetsoffluidpropertiesintroducedinAppendixA.
The values shown in Table D.2 are best matches to the pressure data and the pressure derivative
219
usingthelinearflowmodelpresentedinAppendixC.Thismodelisusedforallthequantitativeanalysis
tofindthelengthofthereservoirandthepressuredecline.
Table D.3 shows the pressure matches to an analytical rectangular flow model, which better captures
the transition from radial to channel flow.
220
This model is useful to determine q, used to compute
permeabilityinSection4.3andt
W
,usedtofindtheconnectedwidthofthereservoirinSection5.Italso
providesacheckthatthelinearmodelarrivesatsimilarvaluesoft
1
andthefinalpressure(within21psi,
orapressuredropdifferenceof1.5%),whichiscontrolledbythelatetimebehaviour.
Note that we match the pressure to around 2 psi or better, which is well within the precision of the
pressuregaugeitself(around5psi).
221
D.1.2Objectivefunctionformatching.Thematchisfoundfromminimizingtheobjectivefunction,O,
222
217
Ihavetakenpressurereadingswiththesamevalueandaveragedthetimesatwhichtheywererecorded.I
haveignoredthefinalfivereadings,whichareperturbedbytheinjectivitytestpriortocementingthewell.
218
IusedtherevisedRatzelreport[3]toassumethatthereisnoflowfromthecompletionofthechokeclosureat
14:22on15
th
July.Itaketheendofthebuildupperiodwhentheinjectivityteststartedat13:00on3
rd
August.At
is the time since choke closure, while t
p
is the production time. From the BP Accident Investigation Report (BP,
2010),Iassumethatthewellstartedflowingatapproximately21:00on20
th
April2010(thisisalsoconsistentwith
theEmilsenReport[27]).Thisgivest
p
(untilchokeclosure)of85days,17hoursand22minutesor7,406,520s.
219
Thederivativeiscomputedasadifferencebetweenthepreviousandthesubsequentpressurereadingsdivided
bythedifferenceinthelogarithmoftime.Thereisaperiodlateon31
st
Julywhenthepressurefallsslightly[T]
thisnegativederivativeisignoredintheanalysis.
220
TheequationsarenotgivenherebutcanbefoundinEarlougheretal.(1968).
221
Itispossibletofitthedatawithanaverageaccuracybetterthanthepressuregaugesensitivity,sincewematch
toadiscretepressureattheaveragetimethatthispressureisrecorded:atthattime,therealpressureisalmost
certainlyveryclosetowhatisrecordedmuchcloserthanthesensitivityofthegauge.
222
TheminimizationhasbeenperformedusingtheSimplexalgorithm.Ihavealsocheckedtheresultsbyhandto
confirmthatagoodmatchhasbeenobtained.
MJBluntExpertReport
106
0 = w
(p
- p
t
(t
))
2
n
=1
+ w
_
p
+1
- p
-1
ln(t
+1
) - ln(t
+1
)
-
Jp
t
Jln(t)
_
t
i
_
2 n-1
=2

(D.1)
where we have n pressure points in the match (n=60 for the linear model, beyond At = 1 day; for the
rectangular model we have 79 values after At = 3,600 s one hour), p
i
at times t
i
, and the theoretical
pressure p
t
and its derivative dp
t
/dln(t) are computed from Eq. (C.64). The function O is minimized by
varyingtheparameters|,t
1
andt
2
inthecalculationofp
t
forthelinearmodel.Theweightingfactorwis
thenormalizedlengthoftimeforwhichagivenpressurevaluewasrecordedthistendstoweightthe
latetime behaviour more heavily. Using no weighting (w=1) has a negligible effect (less than 0.5%)
differencetomypredictionsofpressuredrop.
Figure D.1. The rootmeansquare mismatch in pressure and pressure derivative for different assumed values of the final
reservoir pressure. Note that our theoretical model can predict the measured pressure to, on average, around 2 psi. The
verticallineisthebestmatchpresentedinTableD.2thatusesthebestfittobothpressureanditsderivative.Herewe
useCoreLabsfluidproperties.
TablesD.2andD.3showtherootmeansquareerrorsinthematch,computedasfollows:
Calculatedfinalpressureusingthe
bestmatchtobothpressureand
pressurederivative
Pressurederivative
mismatch
Pressuremismatch
MJBluntExpertReport
107
Root mcon squorc crror in prcssurc = _

1
n
w
(p
-p
t
(t
))
2
n
=1

(D.2)
Root mcon squorc crror in Jcri:oti:c =
_
1
n -2
w
_
p
+1
- p
-1
ln(t
+1
) - ln(t
-1
)
-
Jp
t
Jln(t)
_
t
i
_
2 n-1
=2

(D.3)
Figure D.2. The average absolute mismatch in pressure and pressure derivative for different assumed values of the final
reservoirpressure.TheverticallineisthebestmatchpresentedinTableD.2thatusesthebestfittobothpressureand
itsderivative.HereweuseSchlumbergerfluidproperties.
D.1.3 Sensitivitytothe final reservoir pressure.FiguresD.1toD.3showthemismatchtothepressure
dataifweassumeafinalreservoirpressureandthenfindthebestfittotheotherparameters.So,Apis
fixed,whilethetimest
1
andt
2
tohittheboundariesareadjustedtoobtainthebestmatchtothedata.
pressurederivative
Pressurederivative
mismatch
Pressuremismatch
MJBluntExpertReport
108
However, if we assume a pressure drop that is not optimal, the data match will always be poor,
regardlessofwhereweplacethereservoirboundaries.Weusethelinearmodel.
My best match for the final reservoir pressure is the one that has the best combined fit to both the
pressure and pressure derivative.
We can shift the final pressure by only around 10 psi before a

noticeablynonidealfitisseen.
Figure D.3. The average absolute mismatch in pressure and pressure derivative for different assumed values of the final
reservoirpressure.TheverticallineisthebestmatchpresentedinTableD.2thatusesthebestfittobothpressureand
itsderivative.HereweuseIntertekfluidproperties.
D.1.4 Finding lengths from times. In Section 5 we use the times to reach boundaries presented in
TablesD.1andD.2toestimatethesizeoftheconnectedreservoir.Thedistancesarefoundasfollows.
Width,w = 4
w
o; length to the North (or NorthWest),I
1
=
1
o; and length to the South (or
SouthEast),I
2
=
2
o. o is the inverse diffusivity defined by Eq. (C.4):o =
qc
t
K
. The viscosity,
=0.2197mPa.s,istakenatthefinalreservoirpressure(seeTableA.1).|c
t
istabulatedinTableA.8and
Itaketheentirerangeofvalues.The(connected)areaA
c
=W(L
1
+L
2
).Itaketheaverageareaforagiven
permeability K for the three sets of fluid data. I consider the upper bound K = 329 mD from the
Gringartenreport.
pressurederivative
Pressurederivative
mismatch
Pressuremismatch
MJBluntExpertReport
109
D.1.5Findingsinuosity.InSection5wealsoderiveasinuositytodeducetheflowpathinthereservoir.
The sine wave shown in Figure 5.1 has a dimensionless amplitude times frequency c=0.75.
Mathematically, the path (arc) length is the straightline path multiplied a factor
2
n
,
1 +
2
E(e
2
(1 +
2
)) whosevalueisaround1.14.Eisthecompleteellipticintegralofthesecond
kind (Abramovich and Stegun, 1970). The tortuosity I find is at the lower end of values computed for
otherchannelturbidites(DykstraandKneller,2008).
D.1.6Findingtheconnectivity.IhavefoundtheconnectedareaA
c
fromthepressureanalysis(Section
D.1.4 above). Ialsoknow thegrossrock volumeV=Ah fromthe BPseismic analysis(seeTableA.6). I
assumethatthedisconnectedareaisresolvedatthelimitoftheseismicinterpretationwithanaverage
thicknessh
L
=10ft(seeSection5).Thentheconnectedvolumeis:V
c
=Ah(AA
c
)h
L
.Theconnectivity
isV
c
/Ah=1(1A/A
c
)(h
L
/h).IfindthevaluesshowninTableD.4.
Thisisanupperboundonconnectivity:Itaketheupperboundonpermeabilityandthelowerpossible
boundonthicknessattheperipheriesofthefield.Theaveragethicknessintheconnectedregionvaries
between 69 and 89 ft, see Table D.5, close to that encountered at the well (93 ft). Technically, the
pressure analysis finds a mobility, or a permeability times thickness. In my analysis and that of Dr.
Gringarten,thisisassumedtobetheheightnearthewell(93ft)intheradialflowperiod.However,as
evident from the seismic interpretation, Figure 4.5, the height away from the well is lower. In our
connected domain, this means that the mobility is slightly lower away from the well. However, this
value still exceeds the base case mobility determined in the Gringarten report, using a midrange
permeabilityof238mD.
The connected heightsarenow moreconsistent withthe expertreinterpretationoftheseismicsurvey
by Prof.TorresVerdin atthe Universityof Texas, which placestheoil ina morelaterallyconfined area
withalargeraveragethicknessofaround70ft.
223
Connectivity(%) Fluidproperties
High
(Intertek)
Middle
(Schlumberger)
Low
(CoreLabs)
Highcaserockcompressibility 88 88 87
Midcaserockcompressibility 89 89 88
Lowcaserockcompressibility 90 90 89
Table D.4. The connectivity for the full range of fluid and rock compressibilities. Note that a high compressibility leads to
lowerconnectivity,sincethepressureresponsemovesslower.
Connectedthickness(ft) Fluidproperties
High
(CoreLabs)
Middle
(Schlumberger)
Low
(Intertek)
Highcaserockcompressibility 89 84 84
Midcaserockcompressibility 81 76 76
Lowcaserockcompressibility 74 69 69
TableD.5.Theconnectedthicknessforthedifferentrangesoffluidcompressibilities.
223
TheTorresVerdinexpertreport[TV].
MJBluntExpertReport
110
D.1.7 Discussionofthelinearmodel.Theprincipalreasonforemployingthelinearflowmodelformy
quantitative analysis is that it decouples the uncertainties of flow rate and well cooling in the first day
fromthekeyassessmentofpressuredeclineandtheoverallextentofthefield.
For example, consider the radial flow analysis in Section 4.3 used to determine permeability. Here, as
evidentfromtheequationinTableD.1theconstantderivativevaluefound(around55psi)isaratioof
flow rate to a product of permeability and thickness. The flow rate is the flow rate just before well
closureinthefinaldayofthespill.Ihaveassumedavalueof45,000stb/day.Thethicknessistypically
assumed to be the thickness encountered at the well, 93 ft. The permeability is the permeability
averaged over the distance the pressure wave encounters in the radial flow period up to a radius of
around750m(halfthereservoirwidth).
Whataboutchannelorlinearflow?Heretheoveralldrawdownisgivenby|(TableD.1).Herewefind
again a ratioof flowrateto permeabilitytimesthickness. Buttheflow rate isnow theaverage during
the spill period, since this is the latetime behaviour, the thickness we assume is lower (see Table D.5)
andthepermeabilityistheflowbasedaverageacrossthewholefield.Myassessmentisthattheratiois
thesameasintheradialflowperiod:thethicknessislower,asistheaverageflowrate(comparedtothe
finalvalue)andthetwothingsbroadlycancel.Theratioofthewellthicknesstoanaveragevalueof78
ft(TableD.5)isaround1.2,equaltotheratioofthefinalflowratetotheaverage(takentobearound
37,000stb/day).Thepermeabilityismoreorlessthesameasdeducedintheradialflowregime.
Thisexplainswhyaconstantraterectangularmodelmatchesthedata verywellseeTableD.3.Even
throughtheflowratesfortheradialandlinearflowregimesarelikelytobedifferentandthereservoir
thicknesslikelydecreasesawayfromthewell,consistentwiththeseismicinterpretation(Figure4.5),a
modelthatassumesconstantthickness,rateandpermeabilityfitsthepressureresponseaccurately.
However, using a rectangular flow model for all the analysis involves some complexities, as we would
need to make assumptions on flow rate history and thickness (which varies) to make a quantitative
assessmentforthewholebuildupperiod.Theuseofthelinearmodelavoidstheseproblems:allItake
from radial flow is the value of the derivative and the time to see channel flow, both of which are
evidentfromthedataitself(Figure4.12).

MJBluntExpertReport
111
D.2. Pressurederivatives
HerewepresenttheequationsforthepressurederivativesmentionedinSection4.3.
D.2.1.Radialflow.FromEq.(C.20),forradialflowandforAt<<t
p
wefindthederivative:
Jp
w
Jlnt
=
Jp
w
Jt
= p _1 -
1
1 + t
p
t
_ = p
(D.4)
Thepressurederivativeisconstantforradialflow.
D.2.2 Channel flow. From (C.33) the pressure derivative for channel flow (at early time before the
presenceofanyboundariesisobserved)isgivenby:
t
Jp
w
Jt
=
[t
2
_1 -
1
1 +t
p
t
_ =
[t
2

(D.5)
forAt<<t
p
.
D.3 Pressurematchforchannelflow.
AdirecttestofchannelflowisfromEq.(C.53)toplotthepressureasafunctionofsocalledchannel
time:
t +t
p
- t. For an infinite channel, this plot should be a straight line with an intercept at
the initial reservoir pressure. Figure D.4 presents this plot: there is a convincing and unambiguous
straight line. There is no doubt that the pressure response in Macondo indicates channel flow. This is
alsoconsistentwiththereservoirgeology(Section5)andprovidesconfidencethatwehaveinterpreted
thepressurebehaviourcorrectly.
Thebestlinearfitthroughthedatagivesagradient|0.5psis
1/2
withp
i
11,100psi(seeTableD.1for
anexplanationofterms).Butthemeasuredp
i
=11,856psi;thefittedvalueistoolowbymorethan700
psi: this indicates some additional drawdown in the system and demonstrates that despite the good
linearfit,Macondocannotbemodelledasaninfinitechannel:thepressuremusthaveseenoneormore
barrierstoflowattheendsofthechannel.
Shown in Figure D.5 is the best fit to the data using the linear and rectangular models.
224
The
parameters used to obtain this prediction are provided in Tables D.2 and D.3. The match is excellent:
the pressure signal detects one end of the channel during both drawdown and buildup, but the
influenceoftheother,farther,endissomewhatweaker.t
1
isthetimetakenforthepressuretoreach
the farthest reservoir boundary, while t
2
is the time to reach the nearer end of the channel. t
p
is the
productiontime(thelengthofthespillperiod).Wefindt
1
>t
p
> t
2
.Wecantellthepresenceofthefar
224
MathematicallythelinearmodelpredictionismadeusingEq.(C.64)seeAppendixCforthederivation.
MJBluntExpertReport
112
boundary, since it makes a detectable perturbationto the pressure response. On the other hand, t
1
is
larger than t
p
, the flow period, meaning that the influence of the far boundary does not dominate the
pressure: if t
p
were greater than t
1
, we would see a noticeable and sharp decrease in the pressure
derivativeduringthebuildup;thisisnotobserved(Section4.3).
Figure D.4. The reservoir pressure as a function of channel timet +t
p
- t. Note the almost exact straight line
indicatingchannelflow.TheredlineisthebestfitusingthelinearflowmodelpresentedinAppendixCandtheparameters
listedinTableD.2,whiletheblacklineistherectangularmodelfit(TableD,3):theyalmostexactlyreproducethemeasured
pressures(crossescorrectedtodownholeconditions).Theinsetshowsthepredictionforlatertime(channeltimecloseto
zero)showingthatthepressurereachesfinalvalueof10,433psi.
225
225
ThisisthefinalreservoirpressurecomputedusingCoreLabsfluidproperties.Ihaveperformedthesame
analysisforallthreesetsoffluidproperties.
MJBluntExpertReport
113
D.4 Horneranalysisandcappingstackpressure
A standard method to study radial flow once it has been identified on the pressure derivative is
throughtheuseoftheHornerplot.TheHornerplotwasalsoemployedintheworkofDr.Hsieh(2010,
2011).Forcompleteness,thisisshownhere.
FigureD.5showsthepressureasafunctionofHornertime:fromEq.(C.19)thepressurebuildupshould
beastraightlineforradialflowwhenplottedasafunctionof(t
p
+At)/At.Theonsetofchannelflowa
deviationfromapproximatestraightlinebehaviourisataround60,000s.
Figure D.6 shows the predicted and measured capping stack pressures for reference. Notice that I
predict that the pressure reaches a maximum and then declines: this is at very late times (around 3
months after choke closure) when the cooling of the wellbore has a more significant impact on the
pressurethantheriseinpressureinthereservoir.
FigureD.5.Hornerplotofthepressureresponse:herepressureisplottedagainstthelogarithmof(t
p
+At)/At:astraightline
indicatesradialflow.Theredlineusesalinearmodel,whichdeviatesfromthedataatearlytimeslargeHornertime,while
theblacklineusesarectangularmodel,whichprovidesanexcellentfittothedata(crosses)atalltimes.
MJBluntExpertReport
114
FigureD.6.Predictedandmeasuredcappingstackpressure.Theredlineusesalinearmodel,theblacklinetherectangular
model,whilethedataisshownascrosses.NotethatIpredictthatthecappingstackpressurereachesamaximumandthen
declinesatlatetimewellbeyondwhenthewellwascementedtoafinalvaluearound7,040psi:theeffectsofwellbore
coolingovertaketheeffectsofthepressureincreaseinthereservoir.

MJBluntExpertReport
115
AppendixE. Considerationofvariableflowrates
Mypressureanalysishasassumedaconstantflowrate,followedbyaninstantstoppingoftheflow.In
contrast,in Macondo, therewas a complex sequence ofchanges in flowrateasthechokewas closed.
Furthermore, the flow rate most likely varied throughout the whole spill period as a consequence of
alterationsintheflowresistancenearthewellboreandinthesurfaceequipment.
InthisSectionweconsideragainusingstandardpetroleumengineeringtechniquestheimpactofa
variable flow rate. We do this for two reasons. Firstly, we study the flow rate at the time of choke
closure. This should, in theory, give us a more refined estimate of permeability. We find in this case
though that it has no impact on our calculations. Secondly, we will examine how variable flow rates
couldaffectourlatertimepressureanalysisandestimatesofoilreleased.ThenIprovideananalysisof
flowresistance(skin)betweenthereservoirandthewellboretostudytheevidenceforaflowratethat
increasedduringthespill.
E.1 Superpositionmethodtofindpermeability
InthisSectionwestudythepressurebehaviourshortlyafterchokeclosure.Toperformthisanalysiswe
do need an estimate of flow rate particularly flow rate as the well was closed. I have digitized the
figuresintheRatzelreport[3;Figure9]thatshowflowrateasa functionoftime,andasafunctionof
choke closure, using the LANL data. While this gives slightly lower flow rates than the other values
considered,itappearstobemostconsistentwiththeGriffithsreport[2;SKG].Iamnotsuggestingthat
thisiscorrectitissimplyusedformycalculationsasIdonothaveanotherindependentassessmentof
flowratehistory:indeedIconsideritscumulativethetotalflowtobeincorrect.
ForcompletenesswerepeattheradialflowanalysisfromAppendixCforavariableflowrate.The
pressureresponseforunitrateisgivenbyEq.(C.17):
p
(t) =
p
4nKb
ln _
4t
oyr
w
2
]
(E.1)
ThenusingDuhamelsprinciplethewellpressurecanbewrittenfromEq.(C.18):
p
w
= p
- o(t)(t) + _
o()
o
t
0
p
(t - )J
(E.2)
Nowconsidertheflowratetobeaseriesofconstantflowperiodsseparatedbyjumpsinrate.Wetake
Q(t=0)=0andthenattimest
i
thereisachangeinflowrateAQ
i
,leading,eventually,toarateofzeroat
time t=t
p
; At is the time since flow ceased. AQ
i
is positive if the flow rate increases and negative if it
decreases.FromEqs.(E.1)and(E.2)inthebuildupperiod:
MJBluntExpertReport
116
p
w
= p
-
p
4nKb
ln_
4(t
p
+At - t
)
oyr
w
2
_ = p
-
p
4nKb
ln(t
p
+ At -t
)
(E.3)
since

= u, thesumofalltheratechangesiszero(theflowratestartsandendsatzero).
IfwedefineareferenceflowrateQ
o
,wecanwriteEq.(E.3)asfollows:
p
w
= p
-
p
o
4nKb
ln(t
p
+ At - t
) = p
+ pln
(E.4)
whereq isdefinedinTableD.1andwehavedefinedasuperpositiontimet:
ln() = -
A
o
ln(t
p
+ At - t
)
]
=1

(E.5)
Thereferenceflowratecanbeeitherthefinalflowrate(justbeforewellclosure)orsomeaverageflow
rate.InthisanalysisItakeQ
o
=53,000stb/day,whichistheflowrateestimatedbeforethebeginningof
chokeclosurefortheflowratehistorywehaveassumed[2,3].
E.2.1Pressurederivative.Thepressurederivative,Eq.(D.1)isfoundasafunctionoftratherthantime
t:

Jp
w
J
= p
(E.6)
Verysimplyvariableratesintheradialflowregimebecomeaconstantderivative.
Ifwedonothaveradialflow,wefollowthesameprocedure:wesimplytreatthederivativecomputed
usingthesuperpositiontimeasanormalderivativeforinterpretation.
Figure E.1 shows the reservoir pressure derivative, but using the superposition time t rather than
normal time t. The results are very similar, giving essentially identical assessments of permeability and
channel flow. Hence, for simplicity, in the main text (Section 4.3) I only present a constant flow rate
case. We see a radial flow stabilization at around q = 55 psi and a transition to channel flow t
W
at
60,000 s, consistent with thebestmatch rectangular flow models presented in Appendix D. Using the
equationin TableD.1 wefind,as inSection 4.3, a permeabilityofaround300 mD. This isclosetothe
average of the core measurements (Appendix A.5), while representing a flowbased average over
around 750 m from the well. However, before a day, the conversion from capping stack to reservoir
pressuresisuncertain(AppendixB)andsoIdonotconsiderthisestimateofpermeabilitytobeasrobust
as the determination of 238 mD by Dr. Gringarten. He also finds a flowbased average, but from
pressuremeasurementsmadedownholeshortlyafterdrillingthewell,whichavoidstheproblemofthe
pressureconversion.
MJBluntExpertReport
117
E.2 Effectofvariableflowratesonthelinearflowregime
Myanalysisofthelatetimepressureresponse,usedtodeterminethefinalreservoirpressure,assumes
aconstantflowrate.ThisrepresentstheaverageflowrateduringtheincidentandIhaveimpliedthat
theresultsofmycalculationsareinsensitivetotheflowratehistory.
However, the flow rate historydoes have some impact on thepressure response. Inthis section I will
matchmylinearmodeltodifferenthypotheticalflowratehistories.Imustemphasizethoughthatthe
flowratesIhaveusedareforillustrativepurposesonly;Iamnotclaimingthattheyarecorrect,oreven
possible. In particular I am not necessarily suggesting that the initial flow rate was zero (see below)
myconclusionsdependonlyontheaverageflowrateintheearlierhalfofthespillrelativetothatinthe
latterhalf.
Figure E.1. Reservoir pressure derivative using the superposition time, Eq. (E.5) (blue crosses) compared to the data
uncorrected for flow rate changes (black crosses). Accounting for flow rate makes little difference to the derivative after
10,000 s and sohasno impact on the analysis.There isevidence of radial flow witha bestmatchq of around55 psi and a
transitiontochannelflowatt
W
=60,000s.
TableE.1presentstheresultsofapressurematchusingdifferentpurelyhypotheticalflowratehistories.
Weassumethatatearlytimestheflowrateincreasesfromzerotoamaximumvalue,causedbysome
unspecified erosion in the surface equipment or downhole. When this erosion ceases, we apply flow
Latetimechannelflow
Correctedforflowrate
MJBluntExpertReport
118
conditionsconsistentwiththeGovernmentreportswheretheflowratedeclinesslowlyovertimeasthe
reservoir pressure falls. We assume a 6% decrease in flow rate when the capping stack was installed,
anda4%increasewhentheriserwascut.
The pressure response is computed using Eq. (C.67) the linear flow model. We find the reservoir
parameters that give the best fit to the measured capping stack pressure and pressure derivative. We
haveusedtheCoreLabsfluidproperties.
Relativeoilreleased t
1
(s) t
2
(s) Ap(psi)
1(Constant) 11,270,444 115,079 1,433
1.099(Govt.) 11,585,238 90,289 1,423
1.081 11,426,063 96,378 1,423
1.059 11,244,226 103,939 1,424
1.039 11,069,093 111,934 1,425
1.018 10,899,775 120,447 1,426
0.997 10,736,090 129,546 1,427
0.976 10,578,211 139,301 1,427
0.956 10,426,536 149,781 1,428
0.935 10,281,669 161,062 1,429
0.915 10,144,414 173,217 1,429
0.895 10,018,331 186,038 1,429
0.875 9,903,928 199,569 1,430
0.855 9,800,620 214,040 1,430
0.835 9,711,066 229,298 1,430
0.815 9,638,577 244,985 1,430
0.794 9,587,229 260,369 1,429
0.774 9,561,033 273,907 1,429
Table E.1. The results of matching the analytical linear reservoir model to the pressure data, assuming different flow rate
historiesandcumulativeflowsrelativetoaconstantratecase.Thefirstrowinredhasaconstantflowrate(presentedinthe
mainbodyofthereport),whilethesecondrow(black)usestheGovernmentsassumedflowratehistory.Ingreenisacase
consistentwithafinalflowratesof45,00055,000stb/dayandmymaterialbalancecalculations.Therelativeoilreleasedis
thecumulativeflowrelativetoaconstantflowratecasewiththesamefinalflowrate.
In all cases we have an acceptable pressure match: this analysis cannot discriminate between possible
flow histories. Figure E.2 shows these hypothetical flow rate histories, illustrating one that matches
broadlythatassumedintheRatzelandGriffithsreports,aconstantrate,andanothercasewherethe
cumulativereleaseisapproximately80%oftheconstantratecase.FigureE.3showsthecorresponding
predictionsofpressureincreaseandderivativebyeyethedifferentcasesareidenticalatlatetime:in
all cases the predicted pressure drop varies by less than 10 psi from the constant rate case (less than
1%),whilethepressureismatchedto2.7psiorbetter.
MJBluntExpertReport
119
FigureE.2.Differentputativeflowratehistories.Therateisshownnormalizedbytherateattheendofthespill.Inredis
thebasecaseinthisreportaconstantreservoirrate.
226
Inblackisaratehistorythatcorrespondstothatapproximately
assumedintheGovernmentreports.Ingreenisanincreasingflowrateatearlytime,whichwillgiveacumulativerelease
consistentwithmymaterialbalancecalculationsandafinalflowratearound45,00055,000stb/day.
ThecaseshowningreeninFiguresE.2andE.3givesacumulativereleaseofbetween3.1and3.6MMstb
for final flow rates between 45,000 and 55,000 stb/day. This is now consistent with all the reservoir
data, the estimated final flow rates from Government investigators, while providing a better pressure
match than all of Dr. PooladiDarvishs simulation runs (Appendix F.4). The estimated locations of the
reservoirboundaries(Section5)arealsoconsistentwiththegeologyindeedaratherbetteragreement
is found, since the pressure analysis now allows the connected reservoir to be slightly longer to the
South (t
2
is larger than the constantrate case), closer to the location of the boundaries of the field
identifiedintheseismicsurvey.t
1
isslightlysmaller,althoughstillconsistentwiththeboundariesofthe
fieldtotheNorthWest.
226
Technically,thebasecaseisaconstantreservoirrate(AppendixC).Sincethereservoirpressurevariesover
time,sodoestheoilformationvolumefactor(measuredatthewellborepressure,ignoringskin),andhencethe
surfacerate(measuredinstb/day)changesslightly.Ihavemadethiscorrectiononthegraph,showingaslight
decreaseinsurfaceratewithtimeforthebasecase.Theaveragenormalizedrateisdefinedtobe1.
Constantreservoirrate(basecase)
Governmentestimatedrates
Hypotheticalflowratehistory
consistentwithafinalflow
ratearound45,00055,000
stb/dayandmymaterial
balanceanalysis
MJBluntExpertReport
120
This is an important analysis, as it demonstrates that it is possible to honour the final flow rates
proffered by the Government investigators, while being consistent with all the reservoir data and my
materialbalancecalculation.
Figure E.3. The capping stack pressure and pressure derivative matches using the linear flow model with the flow rate
historiesshowninFigureE.2.Crossesarethedata,theblacklineassumesadecreasing(Government)flowrate,theredlines
a constant rate (the base case) while green is an increasing rate at early time. In all cases the predictions are
indistinguishableatlatetime.
E.3 Analysisofskinanddownholepressure
Anincreasingflowraterequiressomeopeningofflowpaths,eitherdownholeorthrougherosionofthe
blowout preventer and/or other surface equipment. My expertise is in reservoir engineering and so I
will ignore for the sake of argument any erosion in the surface equipment. In this Section I will
demonstratethatitispossiblethatchangesintheflowresistancefromthereservoirtothewellboreat
reservoirdepthalonearesufficienttoallowanincreasingflowrate.
It is usual in normal oilfield operations for the flow path of oil to be impeded in the near wellbore
region. This extra resistance to flow is accounted empirically by a socalled skin factor, introduced
mathematically in Appendix C. In Macondo this skin has three components: partial penetration skin,
because the oil entered the wellbore over a restricted interval of the reservoir; reduced permeability
Pressure
Derivative
MJBluntExpertReport
121
near the well, since the rock may be clogged with drilling mud and cuttings (sand) forced into the
formationduringdrilling;andanyadditionalresistancetoflowduetocementandothermaterialinor
nearthewellboreatreservoirdepth.
E.3.1. Estimate of the skin at the time of the blowout. The Emilsen report in its analysis of the
accident,
227
estimates that the oil flow was initially coming from between only 13 to 16.5 ft of the
reservoirthickness.Wecanusethisinformationtoestimatetheskinfactor.
228
S = _
b
b
]
- 1_ln _
b
r
w
_
K
h
K
-2_
(E.7)
wherehisthethicknessofthereservoir(Iwilluse93ft
229
);h
f
isthethicknessoftheflowinginterval(I
takeamidrangevalueof15ft);r
w
isthewellradius(4.25inches)whileK
h
andK
v
arethehorizontaland
verticalpermeabilitiesrespectively.TypicallyK
h
isgreaterthanK
v
duetothepresenceofthin,horizontal
layers of shale or other low permeability rock that restricts vertical flow: I will assume K
h
/K
v
=10, taken
fromthepressureanalysisintheGringartenreport.ThenfromEq.(E.7)IfindthanS=35.
Inotethatthepermeabilityofcoresamplestakendirectlyfromthewellhadanaveragepermeabilityto
oil of 75 mD (Appendix A.5); in contrast Dr. Gringartens estimate of the fieldaverage permeability is
around 240 mD. This suggests that the rock near the wellbore could have been clogged with drilling
mud and cuttings (sand) dislodged during drilling. On the other hand, the pressure analysis in the
Gringarten report sees no skin due to permeability reductions in the near wellbore region before
cementing.
Last,itisalsodifficulttoquantifytheimpactofotherrestrictionstoflowfromthereservoirtothewell
bore,throughthecementandothermaterialpresentdownhole.
Overall,theestimatedskinfactorof35mustbeconsideredalowerbound:Iwillconsideralikelyrange
of 3550 in the discussion that follows the upper end being used (as a fixed value) by Dr. Pooladi
Darvish.
230
The skinintroduces anadditionalpressuredrop fromthe reservoirtothe wellbore that isgiven, from

Eq.(C.16)as:
p = 2pS =
p
0
2nKb
S
(E.8)
227
FinalEmilsonreport[27],Section2,page7.
228
Adeterminationofthepartialpenetrationskinrequiresknowledgeofthereservoirgeometryandseveral
differentcorrelationshavebeenproposedintheliterature.HereIuseastandardandsimpleexpressionderived
bySaidikowski(1979).
229
TheEmilsonreportuses86ft[27].
230
ThePooladiDarvishreportconsidersarangeof050forskin;[PD],AppendixIII,slide8.However,theskin
factorisassumedtobeconstantduringthespillperiod.
MJBluntExpertReport
122
From Section E.1, a representative value of q is around 55 psi for a given final flow rate final rate Q
f
.
ThisenablesmetowritethefollowingequationtofindApinpsi:
p = 11u S

]
S
(E.9)
whereQ
i
istheinitialflowrate,oratleasttheflowrateintheearlypartofthespillperiod,andQ
f
isthe
finalflowrate.
Let us now estimate the magnitude of this pressure drop. If we assume that the initial flow rate is
aroundhalfthatattheend(seethepreviousSectionE.2)andaskinof35,thepressuredropis1,925psi;
with a skin of 50 it is 2,759 psi. Note that this is larger than the overall decline in reservoir pressure
(around1,400psi Section4.3). We canuseEq. (E.9) to putindifferent hypothetical skin factorsand
flowratesneitherofwhichweknowwithcertaintyatthebeginningofthespillperiodbutitislikely
thattheskinintroducedapressuredroplargerthanthedeclineinaveragereservoirpressure.
E.3.2Theskinattheendofthespill.Wehavetwopiecesofevidencetohelpestimatethefinalskin:
thepressureincrease measured atthecappingstackwhen thewellwas closed, and theflow test with
baseoilpreformedjustbeforethewellwascemented.Iwillconsidereachinturn.
Theskinfactorisnormallyestimatedfromtherapidriseinpressurewhenawellisshutin:themoreor
lessinstantaneouspressureriseallowsustoestimatetheskinfromEq.(E.9).
231
The pressure rose by around 4,000 psi during choke closure. The Government investigators
232
have
modelled this successfully assuming no skin; they assigned no additional resistance between the well
boreandthereservoir:allthispressureriseisaccountedforbytheendingofflowinthewellboreand
surfaceequipment.Thiswouldsuggestthatonlysmall,orevennegligible,portionofthispressureriseis
duetothereservoirskin.
The pressure increase due to skin alone, from Eq. (E.9), is 110S. If S is in the range 3550, this gives
valuesbetween3,850and5,500psi.Ifonlyasmallfractionofthe4,000psipressureriseisduetoskin,
thenthissuggeststhattheskinattheendofthespillismuchlowerandmayindeedbezerobutitis
impossibletobemoreprecise.
The other, more definitive, determination of skin comes from the injection of base oil into the capped
well prior to mud injection and cementing. This oil was injected through the capping stack at rates
between 1 and 7 barrels per minute.
233
The pressure increase on oil injection was very low, around 15
231
Instandardpetroleumpractice,whenthedownholepressureismeasured,theskinisdeterminedfrom
pressuretransientanalysisseeSection4.3.
232
TherevisedRatzel[3]andrevisedGriffiths[2]reportsandtheworkofHsieh(2010,2011)[11,44].
233
http://www.bp.com/liveassets/bp_internet/globalbp/globalbp_uk_english/incident_response/STAGING/local_a
ssets/downloads_pdfs/BP_technical_audio_08032010.pdf
MJBluntExpertReport
123
psi on average bearing in mind the gauge precision of only 5 psi, it is difficult to say more than the
pressureincreasedslightly.
234
Inthistest,wehavemuchslowerflowsthanduringthespilltheseinjectionratesare1,440to10,080
stb/dayandthefluidisasingle(oil)phaseinthewellbore.Wecanassumethat,unlikeduringchoke
closure, the principal pressure drop is due to skin and flow in the reservoir itself. In this case, the
pressureincreasecanbecalculatedassumingthatfortheminutestohourlongdurationofthetest,we
areintheradialflowperiod.Then,fromEq.(C.16):
p = p _ln_
4t
oyr
w
2
] + 2S]
(E.10)
Letusfirstestimatethemagnitudeofthelogarithmicterm.oyr
w
2
hasavalueofapproximately0.0038
s.
235
The duration of the injectivity test was around 1 hour, while the pressure reached 6,976 psi its
highest consistent reading after only 10 minutes (600 s): this is a Ap of 24 psi. Since we find a
logarithm,themagnitudeofthetermisrelativelyinsensitivetotheprecisetimeused,buttoavoidover
estimatingthereservoirpressurechange,Iwilltakethelowertimet=600s:thelogarithmictermthen
hasavalueof13.4.
Tofindthepressuredrop,Ineedtocorrectqfordifferentflowratesitsvalueis55psitimestheratio
of the mud injection rate to the estimated final flow rate during the spill, measured at reservoir
conditions. This is an approximate calculation, so lets assume that the highest mud injection rate is
aroundonetenthofthefinalflowrate.
Hadtheskinfactorbeenaround3550,pressurechangesofover250350psiwouldhavebeenseen(Eq.
(E.10)):thisisanorderofmagnitudehigherthanwhatwasobserved.Evenwithnoskin,themaximum
pressure rise is predicted to be around 70 psi, so if anything there is a negative skin, implying
enhancedflownearthewellattheendofthespill.Thissuggestslimitedresistancetoflowbetweenthe
reservoirandthewellboredownholeattheendoftheincident
I have ignored any pressure drop in the capping stack and wellbore, I have not accounted for the
changing gravitational head between the capping stack and the reservoir during injection, while the
measured the capping stack pressures only have a precision of 5 psi and I have taken one relatively
highpressuredropformycalculations.Theskinattheendofthespillislowerthanitslikelyvalueat
thebeginning.ItisreasonabletoassumeasdidalltheGovernmentinvestigatorsthattheskinwas
zeroattheendofthespill.
234
Truslerreport[T],fromastudyofthecappingstackpressures(PT3K2)from13:00onAugust32010(line
120,951onthespreadsheetprovided).Thepressurebeforebaseoilinjectionisbetween6,952psi,risingtoat
most6,976psioninjection,exceptforonereadingthatspikesjustover7,000psi.
235
o=0.5633s/m
2
,usingCoreLabspropertiesandthemidrangecompressibility,theEulerconstant is0.57722,
andthewelldiameteris8.5inches,givingr
w
=0.108m.
MJBluntExpertReport
124
E.3.3Likelychangeinskinduringthespillperiodandimplicationsforflowrate.Wecannotdrawany
definitiveconclusionsconcerningtheskin,orflowresistancefromthereservoirtowellbore,orhowit
changed over time. However, the available evidence does suggest that the skin was initially high and
declinedduringtheflowperiod.Thischangewouldincreasethepressureinthewellboreatreservoir
depthsbyanamountthatisofthesamemagnitude,orgreaterthan,thereservoirpressuredecline.The
neteffectisadownholepressurethatmayhaveincreasedovertime,leadingtoanincreasingflowrate,
evenifweignoreanyerosionofthesurfacefacilities.TheworkofOldenburgetal.(2012)thatstudied
flow through the wellbore and blowout preventer, implies that small changes in this pressure could
havealargeimpactonflowrate.
Theimprecisionofthisanalysis,however,demonstratesthatitisverydifficulttodetermineflowratesat
thebeginningoftheflowperiod.Ratherthanmakeunsupportedassertionsconcerningpressuredrops
andflowresistance,itispreferabletoemployanapproachthatavoidsthisproblemcompletely,namely
amaterialbalanceanalysisthatdeterminesthecumulativeoilflowdirectly.
MJBluntExpertReport
125
AppendixF. CritiqueofGovernmentexpertreports
ThisAppendixprovidesadetailedcritiqueofthethreeGovernmentexpertreportswhichincludeastudy
of reservoir flow, or fluid properties. These are the reports prepared by Dr. Zick (the Zick report [Z]),
Drs. Kelkar & Raghavan (the Kelkar & Raghavan report [KR]), and Dr. PooladiDarvish (the Pooladi
Darvish report [PD]). Dr. Zick developed an equationofstate model to predict fluid properties. Drs.
Kelkar & Raghavan presented a material balance calculation of oil released, following, conceptually, a
similar approach to mine. As we show below, however, the details and the final answers are
different.Dr.PooladiDarvishlinkedamodelofoutflowperformancetoareservoirsimulatortodeduce
cumulative release. His approach is most similar to the Sandia researchers Drs. Griffiths and Ratzel,
albeitwithamoresophisticatedreservoirmodel.
F.1 Howthecalculationswereperformedandwhatparameterstheyused
My analysis multiplies three quantities together to determine the volume of oil released: (1) the oil
volumeconnectedtothewell;(2)thecompressibilityoftherockandfluids;and(3)thepressuredrop.
This is simply a statement of conservation of mass: any estimate of oil released however it is
presentedmustobeymaterialbalance.Therefore,regardlessofthecomplexityofareservoiranalysis,
thetotalamountofoilreleasedcanbedeterminedsimplyfromthesethreequantities.InthisSection,
expanding on the discussion in Sections 2 and 3, I discuss how these quantities were determined, to
assess the quality of the analysis, and what values were assigned, to determine the effect on the
estimateofoilreleased.
F.1.1Methodology,orhowkeyparametersweredetermined.
1. Connected oil volume. Neither the Kelkar & Raghavan nor the PooladiDarvish reports
mentionedthepossibilitythatnotalltheoilinthefieldisconnectedtothewell.Therewasno
independentdiscussionofgeologyandthe seismicsurveyaspartof a quantitativeconsistency
checkwiththeresultsofthepressureanalysis(mySection5).Theconversionfromreservoirto
surfacevolumeswasperformedinconsistentlyandwasnotbasedonmeasureddata.
2. Rock and fluid compressibility. The Kelkar & Raghavan report did not take a value of pore
volume compressibilityfrom directmeasurements,relying insteadonan incorrect inference as
towhatBPconsideredamostlikelyvalue.However,Dr.Kelkar,whenheconsultedfortheUS
GovernmentFlowRateTechnicalGroup,didusevaluesbasedonthemeasurements.Andwhen
Kelkar & Raghavan used a rock compressibility closer to the measured value, they calculate a
cumulativeflowconsistentwithmyown.
3. Pressure drop. The PooladiDarvish reports base case numerical match to the pressure data
hadapressuredifferencebetweenthecappingstackandthereservoirthatwasunfeasiblylow:
during the 19 days that the oil remained motionless in the capping stack, surrounded by the
cold, deep ocean, it was assumed that this oil maintained a nearreservoircondition
temperatureof220
o
F.
MJBluntExpertReport
126
F.1.2 Results, or what values were assumed. At the end of this Appendix, I provide a table that
quantifiestheimpactofdifferentassumptionsonthecalculatedvolumeofoilreleased,andshowshow
bycorrectingtheseerrorsmyvaluesarerecovered.HereImentionthekeypoints.
1. Connectedoilvolume.TheZickreportprovidedvaluesofoilformationvolumefactorthat
lie outside the range of direct measurements.
236
An inappropriate value of formation
volumefactorledtoabiasedoverstatementofoilvolumemeasuredatsurfaceconditions
in the Raghavan & Kelkar report. Furthermore, they assumed 100% connectivity of the
reservoir, or at least never discussed connectivity explicitly. The result was an over
statementofoilvolume,comparedtomydetermination,ofupto30%ormore.
2. Rock and fluid compressibility. The Raghavan & Kelkar report assumed a pore volume
compressibility of 12 microsips a value that is about twice the average of direct
measurements.TheonlyanalyticalpressuretransientanalysispresentedbyPooladiDarvish
usedanoilcompressibilitythatisalsotwicethemeasuredvalue.
3. Pressuredrop.TheRaghavan&Kelkarreportusesaverydifferentapproachtominetofind
the final reservoir pressure and it understates the weight of the oil between the capping
stackandthereservoir.ThePooladiDarvishreportunderstatesthefinalreservoirpressure
byatleast200psi,againdueprincipallytotheunderstatementofthepressuredifference
betweenthecappingstackandthereservoir(seeAppendixB.4).
TheKelkar&RaghavanandPooladiDarvishreports,despiteusingdifferentassumptionsandvaluesfor
thesekeyquantities,reachthesameconclusion:thattheoilreleasedwasapproximately5MMstb.My
report demonstrates that it is not possible to arrive at this volume while also honouring the pressure
data, the measured fluid and rock properties, and the mostlikely estimates of reservoir size. If you
assert that 5 MMstb was released, then you have to demonstrate why the data is wrong, or
unrepresentativeofthefield.Whilethiswasnotadmittedintheexpertreports,itisevidentoncareful
readingandcloseexaminationoftheassumptionsmade.Theauthorsofthesereportscannotarriveat
acumulativereleaseof5MMstbwithoutignoringthedata.
F.2 TheZickreport
Dr.Zicktookthemeasuredfluidpropertiesandcompositiontoconstructanequationofstatemodelof
theMacondohydrocarbons.Heusedthisequationofstatetoconstructfluidpropertytables.Thisisa
standard approach in the oil industry, and allows fluid properties such as density, viscosity and the
ratioofoiltogastobecomputedatanytemperatureandpressureandforanycombinationofoiland
gas phases. Dr. Zick provided his model to the other experts providing analyses of flow rate for the
Government. As stated in the Zick report, it is reasonable to construct a model that is thoroughly
checkedforconsistencyagainsttheexperimentaldata.
F.2.1Conversionfromreservoirtosurfacevolumes.Inpetroleumengineeringitisstandardpracticeto
refer to oil volumes in stock tank barrels. This is the volume of oil measured at surface conditions of
236
Dr.Zick[Z]alsousedadifferentdefinitionforthisconversion,whichIdiscusslater.HereIsimplynotethathis
modelgivespredictionsthatlieoutsidetherangeofthemeasurementsinamannerthatwillconsistentlyover
statethevolumeofoilreleased.
MJBluntExpertReport
127
60
o
Fand1atmospherepressure(15psi).Theanalysesofflowratepresentedinthepublicdomain,
237
andintheexpertreportsdiscussedinthisSection,havefollowedthisconvention,ashaveI.
The oil formation volume factor, B
o
, is the ratio of reservoir volume to surface volume. My analysis
determinesthereservoirvolumeofoilthatwasreleased.Idividebytheoilformationvolumefactorto
findthesurfacevolumeat60
o
Fand1atmpressure.Arelatedquantityisthegas/oilratio,R
s
,whichis
theratioofthesurfacevolumetothesurfacevolumeofoil.
F.2.2BiasintheZickmodelthatwilloverstatetheoilvolume.Unfortunately,Dr.Zicksmodelpredicts
oil formation volume factors and gas/oil ratios that lie outside the range of the measurements,
introducingabiasthatwilloverstatethevolumeofoilreleasedbybetween3and4%.Thiscomparison
ispresentedinTableF.1.AswiththeBPtables,Iwillnotusethesevaluesbutreplyinsteadonusingthe
measuredvaluesinmyanalysis.
Quantity Corelabs
238
Schlumberger
239
Intertek
240
Averageofthe
measurements
Zick
report
value
%deviation
betweenZick
andmeasured
average
Formationvolumefactor
(singlestage
separation)
241
2.618 2.539 2.5104 2.556 2.473 3.2

Formationvolumefactor
(multistageseparation)
242
2.339 2.262 2.3875 2.330 2.227 4.4

Gas/oilratio,scf/stb
(singlestageseparation)
2,906 2,945 2,831 2,894 2,826 2.3
Gas/oilratio,scf/stb
(multistageseparation)
2,485 2,442 2,747 2,558 2,418 5.5
Table F.1. Table comparing measured formation volume factors and gas/oil ratios to the values presented in the Zick
report.
243
The values in the Zick report lie outside the range of the measured data and lead to an overstatement of oil
volumeatsurfaceconditions.
F.2.3 Singlestage separation. There is a subtlety in how the conversion from the reservoir to the
surface is defined; this explains why two sets of formation volume factors and gas/oil ratios are
presentedinTableF.1.
The singlestage flash or separation occurs when oil from the reservoir is taken to the surface while
remainingincontactwiththeexsolvedgas.InthiscaseB
o
andR
s
donotdependontheexactsequence
oftemperatureandpressurechangesastheoilwasbroughttothesurface.Thesinglestageseparation
occursforflowupthewellboretocollectionatthesurface,asdiscussedinAppendixA.2.
237
See,forinstance,McNuttetal.(2012b).
238
BPHZN2179MDL00063084(CoreLabsfluidpropertyreport)[20].
239
BPHZN2179MDL01608973(SchlumbergerfluidanalysisonMacondoSamples)[34].
240
BPHZN2179MDL04440732(Intertekfluidpropertyreport)[18].
241
Thisistheformationvolumefactormeasuredatthesaturationpressureitsmaximumvalue.Heretakenfora
singlestageseparationtosurfaceconditions.
242
Thisistheformationvolumefactormeasuredatthesaturationpressuremeasuredforamultistageseparation
tosurfaceconditions.
243
Zickreport[Z],Table3,page15.
MJBluntExpertReport
128
F.2.4 Multistage separation.TheZick reportsuggeststhatitismorerepresentativetousevaluesofB

o
and R
s
from a multistage separation. Let me explain the reasoning in my own words, and I will then
providemyownassessment.Oilismorevaluablethangas,andsocompaniesdesigntheirfacilitiesto
capture as much oil as possible. While most of the methane in the reservoir oil forms a gas at the
surface, some of the other, lighter hydrocarbon components (socalled intermediates, such as ethane,
butane and propane) can be present in significant quantities in either the gas or liquid phases. To
produce as much valuable liquid as possible, oil companies want to ensure that as most of the
intermediatecomponentsremainintheliquid(oil)phaseatthesurface.Thisisachievedbyseparating
the gas from the oil at high pressure. The remaining hydrocarbon then evolves less gas, containing
fewerintermediatesandgivingalargeroilvolumethanforasinglestageseparation.Thehydrocarbon
maygothroughseveralstagesofseparationatprogressivelylowerpressuresuntilstocktankconditions
arereached.
Thefluidpropertyreportscontaindetailsofafourstageseparation.TheZickreportassumedthatthis
represents the separator conditions had the accident not occurred and BP had produced the oil
normally.So,inahypotheticalscenario,withnoaccidentandwithBPapplyingtheseparatorconditions
studied in the fluid property reports, the conversion from surface to stock tank conditions would be
performed using B
o
from this fourstage separation. This conversion produces more oil for the same
volumereleasedfromthereservoircomparedtoasinglephaseflash,bydesign.TheZickreportstates
thattheoilvolumeis11%higherifthemultistagevaluesareemployed.
TheZickreportdescribessinglestageseparation
244
asthoughtheoilthatspilledintotheGulfofMexico
wentimmediatelyfromreservoirtemperaturesandpressurestoambientsurfaceconditions.Thiscould
bemisinterpreted: singlestageseparation isindependentof thespeedoftheprocessorthesequence
of temperature and pressure changes, as long as the oil and gas remain in contact (the gas is not
separated from the oil). The report then recommends using values from a multistage separation
processthatwouldmoreaccuratelyreflectthewaystocktankoilisnormallyproduced.
F.2.5Ambiguitiesinthedefinitionofmultistageseparation.Theuseofamultistageseparationprocess
introducesambiguities,astheexactvalueofB
o
dependsonthetemperatureandpressureofeachstage,
aswellasthenumberofstages:thereareaninfinitenumberofpossibilitiestochoosefrom.Itiscorrect
that one particular sequence was studied in the fluid property reports. However, armed with a good
equationofstate model, BP engineersmayhave been able to designabetter process to captureeven
moreoil,oroncostgroundschoseninsteadtousefewerstages,orhavebeenforcedtocombinethe
production with that from other wells and use existing separators that were not designed for the
Macondofluids,andsowerelessefficient.
ItspremiserestsondeterminingwhatBPwouldhaveproduced(andsold)hadagivenreservoirvolume
of oil been extracted. This involves a consideration of BPs production plans, or design criteria for the
field,whichhavenodirectbearingonacalculationofoilreleased.WhiletheZickreportassertsthatthe
244
Zickreport[Z],page16.
MJBluntExpertReport
129
fourstageseparationprocessrepresentstheselikelyconditions,theremaybeevidencetothecontrary,
addingalayerofunnecessarycomplexityandpotentialuncertaintytoanalreadyinvolvedcalculation.
Iusevaluesfromasinglestageseparationinmyreport,basedontheexperimentalmeasurements.The
oilthatwasreleasedtotheoceanalsounderwentaseparation,butatthetemperaturesandpressures
attheseabed,followedbyfurtherseparationsasitrosetothesurface.
F.3 TheKelkar&Raghavanreport
TheKelkar&Raghavanreporthasthreedistinctcomponents:(1)acalculationoftheflowratenearthe
endofthespill;(2)ananalysisofthecappingstackpressuresafterchokeclosuretodeterminethefinal
reservoirpressure;and(3)theuseofamaterialbalanceanalysistoestimatethecumulativevolumeof
oilreleased.Iwilldiscussthelattertwoanalyses.
F.3.1 Pressure buildup analysis. Kelkar & Raghavan state:
245
We use the data gathered during the
shutin to assess certain characteristics of the reservoir, including the shape and well location. They
thenpresentthestandardpressurebuildupandderivativeplots
246
andsay:
247
Ouranalysisofthedata
also suggests that the Macondo well is in the corner of a reservoir with a rectangular shape. Both of
these observations are consistent with the BP predrill report, which suggests that the well is indeed
locatedatthecornerofareservoirwitharectangleshapeTheythenshowafigure
248
aseismicdepth
mapthatpurportstosupportthisconclusion.
TheseismicdepthmapisofsuchpoorqualitythatIcannotdecipherthewelllocationunambiguouslyon
it,nordoesitappeartoindicatetheextentofthefield.Whileindeeditiscorrecttomodelthereservoir
as a rectangle for the pressure analysis, the well was not drilled at one corner of the field (see, for
instance, Figure 4.5). No direct evidence to support this assertion is given; it contradicts the pressure
analysisofDr.Hsieh,BPengineersandthePooladiDarvishreport(seelater).
The standard practice in pressure transient analysis is to compare the measured pressure and its
derivative with the output of an analytical model of the pressure response for linear flow (as I have
done) or for a rectangle (as Dr. PooladiDarvish and I have done as well). Instead, Drs, Kelkar &
Raghavan simply fit a curve to the pressure buildup.
249
The authors state that this curve is a
rectangularhyperbolawhichmayleadthenonexperttoconsiderthisisanappropriatemethodology
for a rectangular reservoir. It is not: the functional form for the pressure buildup is inconsistent with
therigorouslyderivedequationsforflowinachannel(seeAppendixC).
TheKelkar&Raghavanreportprovidesareference(Hauglandetal.,1984)tojustifythisapproach.This
paperstates:Ifnothingisknownaboutthereservoir,thenthehyperbolamethodcanbeusedtogeta
rough estimate of the pressure. But Drs. Kelkar & Raghavan and everyone else who has studied
245
KelkarRaghavanreport[KR],page21.
246
KelkarRaghavanreport[KR],figure7,page22.
247
248
249
MJBluntExpertReport
130
Macondodoknowsomethingaboutthereservoirithasanapproximatelyrectangularshape,while
theseismicsurveyindicatesitslikelysize.
Drs.Kelkar&Raghavanthenmentionaseriesofothercurvefitapproachestofindtheaveragepressure
thatprovidelowernumbers.Againthesearemethodsofpoorreliabilityroughestimatesthatare
normallyonlybeappliedtocaseswherethereservoirgeometryisunknown.
In the Appendices they discuss standard petroleum engineering methods. For pressure transient
analysistheysaythatitcanbeusedtofind
250
thelocationofboundariesorbarriers.However,theydid
notdothisforMacondo.
F.3.2 Conversion from capping stack to reservoir pressures. Drs. Kelkar & Raghavan used a constant
head of 3,220 psi to convert the capping stack pressure to downhole conditions.
251
As discussed in
AppendixB,thisapproachisincorrect,asitignoresthecompressibilityoftheoil:thepressuredifference
changeswithpressureandtemperature,andishigherthantheyassumed.ThedatareportedinFigure8
show a nonmonotonic trend (one point above the others) at late time this is not seen in the data I
haveusedandisnotevidentintheirFigure7either.
Inconclusion,theauthorsreportaseriesoffinalpressuresbetween10,235and10,396psi.Theytook
the highest number for their material balance calculation. My range of final pressure using rigorous
methods is between 10,433 and 10,531 psi with the last recorded capping stack pressure
corresponding to a downhole value between 10,276 and 10,371 psi, dependent on what set of fluid
measurementsisused.
252
IconsiderDrs.Kelkar&Raghavansfinalvalue10,396psitoolow.
F.3.3 Material balance analysis. This is similar, conceptually, to the work presented in this report.
Kelkar & Raghavan correctly stated that they are
253
using an industry standard material balance
methodology.Iagreewiththemthatmaterialbalanceisindeedthecorrectapproachtodeterminethe
totalvolumeofoilreleased.
They followed the same approach as I do in Section 4: determining the original oil in place, the total
compressibilityandthepressuredrop,beforemultiplyingthesethreequantitiestogethertoobtainthe
oilreleased.
Thereisadifference,though,inhowwedotheanalysis.Drs.Kelkar&Raghavandidnotatanystage
usefluidorrockdatatakenfromacriticalandindependentreviewofthedirectmeasurements,despite
this data being made available to them. Instead, theypickeddata from secondary sources, with scant
justification.
F.3.4Inappropriatevalueoftheoilformationvolumefactor.TheKelkar&Raghavanreportestimated
theSTOIIP(stocktankoilinitiallyinplace)as137MMstb.Thisishigherthanmyestimatesofbetween
250
251
252
TableD.2.
253
MJBluntExpertReport
131
122 and 127 MMstb (with 100% connectivity). However, this is a consequence of the use of an oil
formationvolumefactorof2.14.TheauthorsappeartohavetakenavaluefromaBPpropertytable.
254
The value of B
o
is inconsistent with the data, as I discuss in Section 4.1. They assumed a reservoir
volumeofoilthatis293MMrb,whichisveryclosetomyvalueof291MMrb(seeTableA.3).Hence,the
overstatementofoilvolumeissimplyaconsequenceofapoorvalueofB
o
employedsomethingthat
wouldhavebeenevidenthadtheauthorsstudiedthefluidpropertyreports.
Drs.Kelkar&Raghavanpresentvarioussensitivities,includingalowerSTOIIPof110MMstb,avaluethat
theysaywasreportedbyBP.Intheend,therangeofSTOIIPisnotlinkedtoaproperassessmentofthe
oil volume underground, but arises from a selection of various incorrect conversions from reservoir to
surfacevolumes.
Drs.Kelkar&Raghavanusedanoilcompressibilityof14.3microsips.Thisisinmyrangeofvaluessee
Table A.5. However, they did not review the fluid data directly, but again relied on an average value
fromBPsfluidpropertytables.
F.3.5 Exaggerated pore volume compressibility of 12 microsips. The report states:
255
Our calculation
usesaformationcompressibilityis12x10
6
psi
1
.AccordingtoBP,thisisthemostlikelyvalueofformation
compressibility.Thisisthesinglemostsignificanterrorintheiranalysis.Thereisnoscientificevidence
tosupporttheuseofthisvalue.ItisnotthemostlikelyvalueaccordingtoBP.Thisvaluewasemployed
during the well integritytest as a worst case scenario: BP considered 6 microsips to be the most likely
formation (or pore volume) compressibility, and this number was used after choke closure in their
analysis of the pressure response.
256
I consider the fact that Drs. Kelkar & Raghavan have ignored the
directlymeasured data, which gives an average compressibility of approximately 6 microsips
presentedinSection4.2tobeaseriousshortcomingintheiranalysis.
F.3.6 Dr. Kelkar had earlier used 5.61 microsips for his work on Macondo. The insistence on 12
microsips is all the more baffling, when compared to the analysis that Dr. Kelkar performed as a
consultant for the US Governments Flow Rate Technical Group. During the spill his team estimated
mostlikelyflowratesofbetween27,000and32,000stb/day.
257
However,itisnottheflowratesthatI
will comment on I agree with Drs. Kelkar & Raghavan that the appropriate method to determine
cumulativereleaseisamaterialbalanceanalysisbutonthedatathatwereemployed.
Dr. Kelkar used a base case pore volume compressibility of 5.61 microsips
258
and a high case of 8.29
microsips.
259
These values are taken directly from Weatherfords reports of core measurements.
260
254
SeethevalueofB
oi
fortheBPtablesquotedinTableA.4:2.1431.
255
256
See,forinstance,PinkyVinsonsdeposition[47],page300,line15.AlsoDr.Merrillsdeposition[54],page192,
line5;page214,line23;andpage216,line15onwards.
257
SeeMcNuttetal.(2012b),Table3.
258
SeeDonMaclaysdeposition[62],page393,lines1013;;Exhibit9859(KelkarFRTGReport).
259
SeeDonMaclaysdeposition[62],page413,line25andpage414,line1;;Exhibit9859(KelkarFRTGReport).
260
BPHZN2179MDL02393883(Weatherfordsummaryofporevolumecompressibility,UPVCtab,cellsB17and
D17[26].
MJBluntExpertReport
132
Looking at the measured data, is, of course, the appropriate approach which I have followed.
261
The
Kelkar&Raghavanreport,however,didnotdiscussthemeasureddataatall.
F.3.7Calculatedoilreleasedof3.4MMstbifacompressibilityof6microsipsisused.InAppendixC,
262
the Kelkar & Raghavan report considered a formation compressibility of 6 microsips (close to the
measuredvalue),forwhichthecalculatedoilreleasedhadalowerrangeof3.4MMstb.Thisissimilarto
the midrange of my calculations; this demonstrates that when Drs. Kelkar & Raghavan admitted the
measuredvaluestheyarrivedatasimilardeterminationofoilreleasedtomine.
However,theyconcludedthemaintextwitharangeofoilreleasedbetween4.5and5.5MMstb,usinga
rangeofSTOIIPbetween110and137MMstb.Theydidnotdiscussthelikelyconnectivityofthefield,
nordidtheyprovideanyanalysisofthereservoirgeology,whichisavitalandstandardconsistency
checkinanyreservoirengineeringcalculation.
F.4 ThePooladiDarvishreport
Dr. PooladiDarvish also simulated flow through the wellbore, capping stack and other surface
equipment to estimate the flow rate near the endof the spill. Assuming fixed outflow conditions, the
model was run for the entire period of the spill to find the cumulative volume of oil released. In a
secondstageofthework,awellflowmodelwascoupledtoareservoirsimulationmodelandrunwith
differentsetsofassumedparameters.Themodelsmatchedthepressureresponseatthecappingstack
afterchokeclosurewithanaverageerrorofbetterthan0.1%andpredictedthecollectedvolumesofoil
within25%.Thesegoodmatchmodelshadacumulativeoilreleasedofbetween5and5.3MMstb.
The report contains a long series of Appendices prepared as slides, rather than text. However, it is
possibletodisentangletheanalysisandexplainhowdifferentconclusionsarereachedfrommyanalysis.
Onething I will also emphasize is that althoughthe finalnumbers appear consistentwith theKelkar &
Raghavanreport,theinputdataaresignificantlydifferent.
F.4.1Analyticalmethodwithfixedoutflowperformance.ThePooladiDarvishreportfirstpresentsan
analyticalestimateofoilreleased.ThisworkisakintotheRatzelandGriffithsSandiareports,usingan
estimateofthefinalflowrateandfixedoutflowconditionstoextrapolatetheflowthroughoutthespill
period.Thismethodmadeunsupportedassumptionsconcerningtheflowresistanceintheearlypartof
thespillandessentiallyfixedthecumulativeoncethefinalflowratewasdetermined.
F.4.2Welltestanalysisandanexaggeratedoilcompressibility.Intheanalyticalmodel,PooladiDarvish
coupled his wellbore flow model to a prediction of the pressure buildup. I presume that this is a
conventionalwelltestanalysis.TheresultsareshowninAppendixII.
263
Asisstandardinthepetroleum
community,thedataandamodelmatchtothepressureandpressurederivativeareshown.
261
Thesevaluesaretheporevolumecompressibilitiesatafluidpressureof11,000psi.Theporevolume
compressibilitydecreasesasthepressuredrops:asexplainedinAppendixA,Iusevaluesthatexactlyreproduce
thechangeinporevolumeforapressuredropfrominitialtofinalreservoirpressures.
262
263
PooladiDarvishreport[PD],AppendixII,slides30and31.
MJBluntExpertReport
133
TheinputstoPooladiDarvishsmodellackfoundation.First,thepressurederivativedataareverynoisy
muchmorenoisythanthedataIpresentinSection4.3.Itisnotclearhowthedatawereselectedor
thederivativescomputed.Thederivativedatashowsarapiddroptoclosetozeroneartheendofthe
buildupperiod.Thisisnotevidentinmyanalysis,orinthedatashownbyDrs.Kelkar&RaghavanorDr.
Hsieh. As a result, the match is unlikely to be precise (as the data are noisy) and underestimate the
finalpressure(becausethefinalderivativevaluesaretoolow).Theestimatedfinalpressureis10,336psi
indeed this is too low. A fixed pressure difference of 3,282 psi
264
was applied between the capping
stack and reservoir. Using a fixed value introduced a systematic error, as discussed in Appendix B.
However,atleastthevalueitselfisintherangeIconsider.
Moresignificantthoughistheassumedoilcompressibilityinonecase:28.5microsips,whichistwicethe
measuredvalue(seeTableA.5).Theporevolumecompressibilityisunderestimatedat2microsips,but
the total compressibility, Eq. (A.7), is exaggerated at 26.6 microsips. The total compressibility times
porosityis5.85microsips:thisisfaroutsidethemeasuredrange3.44.5microsips(seeTableA.8).
WhydidDr.PooladiDarvishdothis?Hecannotmatchthepressuredataotherwise.Hehashookedhis
modeltoawellboresimulatorandassumedconstantoutflowconditions.Hence,ifheestimatesafinal
flowrateof51,00054,000stb/day,hehastohaveacumulativereleaseofaround5MMstb.Hismodel
isnotpredictive:oncethefinalflowrateisdetermined,soisthecumulativetowithintightbounds.This
is why all his good match models have moreorless the same total flow they can do nothing else,
unless he relaxes the constraintof a fixed outflow. But, he also has to obey material balance. He has
fixedhisreservoirvolumetoaround109MMstb.His(poor)matchtothepressuredatagivesapressure
drop of1,515psi. Hiscumulativeis forthiscase 5.20MMstb. Hence,from Eqs.(A5)and (4.2),he
has to have a total compressibility of around 26 microsips. If he assumed a compressibility that was
consistent with the measurements, he would not have been able to match the pressure. This would
have indicated that some of his underlying assumptions specifically the assumption of a cumulative
flowaround5MMstbwerewrong.
There is a fundamental point here that lies at the heart of my analysis, and which is ignored in
differentwaysbyalltheGovernmentinvestigators:
It is not possible to match the pressure buildup with a cumulative release of 5 MMstb while
honouringthemeasuredreservoirrockandfluidproperties.
Either5MMstbiswrong,orthemeasuredreservoirpropertiesarewrong.Youcannothaveboth.Drs.
Kelkar &Raghavan(andDr.Hsieh)avoidedthisproblembyusingaporevolumecompressibilitythatis
morethantwicethemeasuredvalue,whileDr.PooladiDarvishadjustedtheoilcompressibilitytoobtain
afit.LateraswediscussDr.PooladiDarvishdidtakerepresentativereservoirproperties,butuseda
newtricktohidethediscrepancybyallowingunfeasiblylowfinalpressuresandgrowinghisreservoir.
264
PooladiDarvishreport[PD],AppendixII,slide15.
MJBluntExpertReport
134
F.4.3Confusion over fluid properties.ThePooladiDarvishreportstatesthatituses:

265
Themeasured
Pencor data obtained from the single stage flash test. Again, multistage values are not used,
contradictingtheassertionsintheZickreport.However,aswithDrs.Kelkar&Raghavan,thevalueofB
o
isnotrepresentativeofthemeasurements.AgraphofB
o
asafunctionofpressureisshown,displaying
valuesthatabovethebubblepointareallgreaterthan2.3.However,inthewelltestanalysis(see
above),B
o
isassumedtobe2.131.Thislowervaluewouldappeartobefromtheseparatortest.
266
But
thenlater,itisstatedthatsinglestagevaluesareusedfortheoutflowperformancesimulationwhich
iscorrect,asthesearetheconditionsthatpertaininthewellbore.
267
But,Idontseehowdifferentsets
ofdatacanbeusedindifferentpartsofthesimulationyouhavetobeconsistent.Thisisoneofseveral
examples,wheredifferentpropertiesareusedfordifferentcalculations,withasignificantshiftbetween
theanalyticalandnumericalmodels.
F.4.4 Assumed oil volume is too large. Dr. PooladiDarvish then coupled the outflow simulation to a
reservoirsimulator.Abasecasereservoirmodelisdescribed.Thishasarectangulargeometrywithan
oilvolumeof137MMstb.ThisvalueisthesameasreportedbyDrs.Kelkar&Raghavan,buthigherthan
usedintheanalyticalmodel.ThejustificationappearstocomefromBPsestimatesbeforedrilling.
268
It
is not clear why the reservoir size has increased from the analytical model, and if any independent
assessmentofreservoirvolumewasundertaken
F.4.5Correctassignmentofporevolumecompressibility.ThePooladiDarvishreportthenpresents25
differentsimulationswithparameterspresentedinTable2.
269
Iwillnoteonesignificantdifferencewith
the work of Drs. Kelkar & Raghavan: the base case pore volume compressibility is 6 microsips, as
assumed by BP and close to my midrange value. This value is based as it should be on the
measurements.
F.4.6 Exaggerated range of data. Dr. PooladiDarvish presented tables giving ranges of parameter
values.
270
In many cases the range was overstated and not supported by the available data. He then
allowed his reservoir model to have properties within these ranges, outside the values from direct
measurement.
- Permeability.Therangeis170850mDwithabestestimateof550mD.Thebestcasedoes
not correspond to the average of the core measurements or from pressure analysis (Section
4.3),whilethehighcaseliesoutsidetherangeofmeasuredvalues(AppendixA.5).Overallthere
is a bias to overestimate permeability. The average permeability is twice as high as the most
robustdeterminationfrompressureanalysis(238mDintheGringartenreport).Thisleadstoa
doublingofpredictedflowrate.
265
SeePooladiDarvishreport[PD],AppendixII,slide45.
266
SeePooladiDarvishreport[PD],AppendixIII,slide8.
267
268
269
SeePooladiDarvishreport[PD],page14.
270
SeePooladiDarvishreport[PD],AppendixIII,slides8and9.
MJBluntExpertReport
135
- Porosity. The base case value is 0.23 with a range 0.170.28 the values are taken from BPs
assessmentbeforedrillinganddirectdatawasavailable.Onceporosityhadbeenmeasured,the
best estimate of the average value is lower 0.217 and the range of values considerably
narrower(seeSection4.1).
- Oilcompressibility.Abasecasevalueof12.3microsipsisgiven.Thisisbelowtherangeofthe
measurements: 13.714.8 microsips (see Table A.5). However, Dr. PooladiDarvish allows a
rangefrom6to32.3microsips.Thisrangeisfartoowide:hishighvalueisthecompressibilityat
thebubblepoint.
271
However,thereservoirpressureisatalltimesmuchhigherandsothese
values are unrepresentative and have no relevance for calculating oil expansion. In the
simulationruns,itappearsthattheoilcompressibilitywasadjustedtomatchthedata:thevalue
heusedwasaround15.3microsips,abovethemeasuredvalues.
272
- Pore volume compressibility. As mentioned above, Dr. PooladiDarvish took a representative

base estimate of 6 microsips. However, his high estimate is the alltoofamiliar value of 12
microsips. He states that both values are taken from the test conducted by Weatherford.
While 6 microsips can indeed be justified, 12 microsips is not a value that was measured (see
Section4.2).
- Oilvolume.Dr.PooladiDarvishquotesarangefrom75200MMstbwithabaseestimateof137
MMstb.HebasedthisonBPspredrillrangeof138239MMstb.However,thevalueofB
o
used
byBPintheircalculationswasfarbelowwhatwaslatermeasured.Heneedstocorrectforthis.
Itisnotcleariforhowthiswasdone.Moreover,asIstateinSection5,theconsistencyofthe
predictedandmeasuredheightofoilbearingsandstoneatthewell,combinedwiththeresults
ofthepressureanalysis,considerablyreducetheuncertaintyinconnectedoilvolume. Instead
Dr. PooladiDarvish suggests a wider range, even though we now have additional information:
his range is approximately 39%, as opposed to BPs 27%. In any event, the base case is too
highasmentionedabove.
- Oil formation volume factor, B
oi
. The range of B
oi
from 2.02.35 with a base case of 2.1 has a
bias that leads to an overstatement of oil released. The measured range (Table A.4) is from
2.272.36.
Dr.PooladiDarvishsconsideredarangeofvalues,anchoringtherangeonmeasurements.However,in
most cases, the justification for his choices was somewhat flimsy with a tendency to overstate the
uncertaintyandinsomecasestobiastheestimates.
F.4.7Poorpressurematchusinganumericalsimulator.Figure7
273
ofthePooladiDarvishreportshows
the match to the pressure after choke closure and the associated error. By eye the match appears
impressive and similar to the matches presented in this report and by Dr. Hsieh. However, as I have
271
Asatechnicalpoint,thefluidcompressibilityatthecriticalpointofafluidisinfinite.Inanearcriticalfluid,such
astheMacondohydrocarbon,thebubblepointisclosetothecriticalpointandhenceverylargevaluesof
compressibilityareexpected.Thisisfascinatingthermodynamicallybuthasnobearingonthecalculationofoil
released.
272
Ihavededucedthisvaluetobeconsistentwithmaterialbalance.WhileDr.PooladiDarvishhasmadeallhis
computerfilesavailable,Ihavenotbeenabletodeducetheoilcompressibilitydirectlyfromhisdata.
273
SeePooladiDarvishreport[PD],page16.
MJBluntExpertReport
136
emphasized, close scrutiny is required. I have matched the reservoir pressure with an average (root
meansquare) error of 2 psi or less.
274
This is a fractional error of 0.02%. This is superior to any of the
pressure matches achieved by Dr. PooladiDarvish.
275
Furthermore, Figure 7 shows that at late time
just the period when we need to match the pressure as precisely as possible to find the correct final
value the mismatch increases to almost 0.1%, or 10 psi.
276
This is significant by the standards of
pressure transient analysis. I do note, however, that some of the other matches shown are better at
late time, although the overall error is always 0.03% (3 psi) or greater. The problem here is that
numericalsimulationisnotthebestwaytomatchpressuretheuseofanalyticalwelltestanalysistools
providesasuperiordeterminationofreservoirproperties.
F.4.8Incorrectpressuredifferencebetweenthecappingstackandthereservoir.ThePooladiDarvish
reportpresentsadetailedanalysisofthispressuredifference,quantifyingitsvaluewithalinearfunction
of temperature and pressure.
277
This approach mirrors that in Appendix B of this report. He
demonstrates,correctly,thatvariationsinpressureandtemperatureleadtoachangeof150psiinthe
pressuredifference.However,Dr.PooladiDarvishkeptthisdifferencefixedduringthesimulationofthe
pressure buildup: this is incorrect, it increases with the rising capping stack pressure and as the well
bore cools. This is important, as it affects the shape of the pressure buildup in the reservoir and the
derivative.
Forthesimulations,heassumedapressuredifferenceof3,157psi.
278
Thisissignificantlylowerthanthe
valueheusedintheanalyticalmodel(3,282psi)andiswellbelowtherangeofmy values. Thisledto
finalreservoirpressuresthatwerefartoolow,andanexaggeratedpressuredrop.Withthenumerical
simulation, mismatches in the final pressure were hidden, allowing a cumulative release of 5 MMstb,
obeyingmaterialbalanceandusingotherwiserepresentativefluidandrockproperties.
Iquotefromthereport:
279
Thefluidatwellheadisexpectedtobebetween243and40F(withthehigh
andlowlimitscorrespondingtofluidvelocitiesofinfinityandzero).Here,theuncertaintyinhydrostatic
pressure is evaluated if the wellhead pressure was uncertain between 140 and 220 F. (Note that the
BaseModelconsideredinthisstudyassumesawellheadtemperatureof220F.But,duringthebuild
up period when the pressure is matched there is no flow. So, the temperature should be nearer
40
o
F,accordingtoDr.PooladiDarvish.Later,adetailedsensitivityanalysisispresented:everyvaluefor
the pressure drop that is calculated using measured fluid properties lies above his base case.
280
This
introducesabiasintothemodel.Thepressurechangeissignificantlyunderstatedandisthekeyerror
inthisanalysis,givingfinalreservoirpressuresthataretypicallymorethan200psilowerthanthevalues
Idetermine.
274
SeeAppendixD,TablesD.2andD.3.
275
SeePooladiDarvishreport[PD],Table1,page3.
276
Notethatthefigureusesalogarithmictimeaxis,soasmallincrementinthelogarithmoftimeattheend
representsasignificantfractionofthebuildupperiod.
277
278
PooladiDarvishreportSee[PD],AppendixIII,slide5;seealsoAppendixIV,slide6.
279
PooladiDarvishreport[PD],AppendixIII,slide18.
280
PooladiDarvishreport[PD],AppendixIII,slide23.
MJBluntExpertReport
137
Asanexample,ItakeDr.PooladiDarvishsbasecasemodelthathasafinalreservoirpressureof10,202
psi.
281
I consider this value implausible: as stated in Section F.3, at the end of the buildup period the
downhole pressure lies between 10,276 and 10,371 psi dependent on the fluid properties used. The
final pressure cannot be below this. Indeed, using rigorous methods, I find a final pressure between
10,433 and 10,531 psi (Section 4.3). Since from material balance the cumulative release is
proportional to the pressure drop the difference between the initial and final pressures this
introducesasignificantbiasintheanalysis,consistentlyoverstatingtheoilreleased.
For the good match cases the final reservoir pressures lie between 10,053 and 10,239 psi.
282
I
consider all these values implausible. They all lie below the likely downhole pressure when the well
wascementedandthewelltestanalysisfitbyDr.PooladiDarvishof10,336psi.
F.4.9Anincreasinginitialflowrateallowsalowerestimateofcumulativerelease.Dr.PooladiDarvish
also considered a case where the flow rate increases initially before reaching a maximum and then
declining. This scenario isnot dissimilar to the onesI explore inAppendix E.3. The cumulative flowin
thiscaseis4.7MMstb,withapossible4.5MMstbdiscussedinafootnote.
283
Dr.PooladiDarvishdoes
notexplore,however,thefullrangeofpossibleflowratehistories,someofwhichmayhavemuchlower
cumulative releases of oil. The quoted range cumulative release 5.05.3 MMstb was based on the
unsupportedassumptionthattheoutflowconditionsremainconstantduringthespillperiod.
Imagine that Dr. PooladiDarvish done the following: considered a wider range of flow rate histories;
keptthe(correct)fluidandrockcompressibility;adjustedtheoilvolumetobeconsistentwithBPsmid
rangeestimateandthelikelyconnectivityofthefield,whileusingthecorrectconversionfromreservoir
tosurfacevolumes;andusedamorerepresentativevalueofthepressuredifferencefromthecapping
stacktothereservoir(ashepresentsinhisAppendixIII).Ifhehaddoneallthis,thenIsuggestthathe
wouldhaveobtainedsimilarvaluesofoilreleasedtomine.
F.4.10 Final discussion taking a fixed outflow predetermines the cumulative. The PooladiDarvish
report suffers from the same problem as the Ratzel and Griffiths reports, even though the reservoir
models employed are more sophisticated. Dr. PooladiDarvish linked a reservoir model to the outflow
performance, found a final flow rate of around 50,000 stb/day and then assumed a fixed outflow
performanceovertheperiodofthespill.Thispredeterminedwithinnarrowbandsthecumulative
flow.Thisexplainsthatevenwithverydifferentassumedreservoirproperties,thetotalflowremained
between5.0and5.3MMstbinallcases.Onlyalimitedinvestigationofchangingoutflowperformance
wasprovided,andsothefullrangeofpossiblecumulativereleasewasnotconsidered.
In order to match the capping stack pressure buildup, while allowing the reservoir model to release 5
MMstb,the reportmadetwo errors. First, theconnectedreservoirvolume inthe simulations was too
large 137 MMstb or more. The possibility that the reservoir is not completely connected was not
considered. As in the Kelkar & Raghavan report, there was no discussion of reservoir geology at all.
281
PooladiDarvishreport[PD],AppendixIV,slide17.
282
PooladiDarvishreport[PD],AppendixV,slides4and5.
283
PooladiDarvishreport[PD],page26n.10.
MJBluntExpertReport
138
Second, the pressure difference between the capping stack and the reservoir was underestimated
duringthebuildupperiod,givingreservoirpressuresthatweretoolow,overstatingthepressuredrop
andtheoilreleased.
F.5 Overviewofestimatesofcumulativeoilreleased
I conclude this Appendix with an overview of the estimates provided by Drs. Kelkar & Raghavan, Dr.
PooladiDarvishandDr.Hsieh(seeAppendixG).InTableF.2Ishowhow,bycorrectingforthepressure
drop,compressibilityandconnectedoilvolume,mycalculationofoilreleasedcanberecovered.While
the methodology is in places correct, such as the use of pressure transient analysis and material
balance,theyallcontainflawsintheapplicationofthesemethodsandintheuseofmeasureddata.
Investigator Estimated
oilreleased
Usingmy
connected
volume
Correctingthe
compressibility
Correcting
thepressure
drop
Corrected
calculationof
oilreleased
Drs.Kelkar&
Raghavan
5MMstb
284
0.48MMstb 1.04MMstb 0.22MMstb 3.26MMstb
Dr.Hsieh 4.92MMstb +0.07MMstb 1.22MMstb 0.51MMstb 3.26MMstb
Dr.Pooladi
Darvish
5.03
MMstb
285
0.93MMstb 0.16MMstb 0.68MMstb 3.26MMstb

Table F.2. Table illustrating the impact of the different assumptions in the analyses presented by Dr. Hsieh, Drs. Kelkar &
Raghavan,andDr.PooladiDarvish.Thetableshowsbyhowmuchtheestimatespresentedbytheseinvestigatorschangesif
we use the connected oil volume, compressibility and pressure drop, compressibility values presented in this report. The
final column provides my determinations of oil released. I have taken the midcase pore volume compressibility of 6.35
microsipstocorrectthecompressibilityandhavetakentheaverageofmycalculationforthethreesetsoffluiddata.These
resultsarealsoshowninthesummarytableinSection2.

284
Theirmidrangevalue,whichrequiresanassumptionofaconnectedoilvolumeof123.5MMstb.Kelkar&
Raghavanreport[KR],page28.
285
ThebasecasePooladiDarvishsimulationmodel[PD],AppendixIV,slide14.
MJBluntExpertReport
139
AppendixG. FurthercritiqueoftheHsiehanalysis
This Appendix provides a detailed critique and analysis of the reservoir modelling work of Dr. Hsieh
reportedintheopenliterature(Hsieh,2010,2011),andinreportsandpresentations[11,12].Iwillalso
refertohisdepositioninSeptember2012[42,43]andusedataprovidedtomeinspreadsheetform[44].
FurtherinformationisalsocontainedinMcNuttetal.(2012a,b)andHickmanetal.(2012).
G.1 Comparisonofpressuredataandanalysis
G.1.1ComparisonofthedatausedbymeandDr.Hsieh.BothDr.HsiehandImatchthepressuredata,
sothiscritiquestartswithapresentationofthedatathatweused.FigureG.1showsthecappingstack
pressurethatformedthebasisofbothouranalyses.
286
AsmentionedinSection4.3,Ihaveusedallthe
distinct pressure readings and taken the average time for consecutive readings of the same pressure
value. Dr. Hsieh has stated that he selected data values that had an approximately even spacing in
Hornertime.
287
Sincethepressuregaugeonlyhadasensitivityof5psi,selectingdataatpredetermined
timesimmediatelyintroducesanerror,whereasaveragingthetimeforagivenreadingasIhavedone
preservesalltheinformation.Itisevident,moreover,thatthedataweusearequitedifferent.While
at late time (after around 1 day or 100,000 s) the trend is similar, my values are lower by an
approximatelyconstantoffsetofalmost40psi.NotethatIhavemorepressurepointsatlatetime.This
means that in matching my model to the data, I give more weight to later times. This allows me to
represent better the end of the buildup period and hence the final reservoir pressure. At early time,
not onlythepressure values, buttheirtrend is verydifferent. Dr. Hsiehusesdata that display a much
steeperriseinpressureataround20,000s.
I also show the match to the pressure provided by Dr. Hsiehs rectangular simulation model and my
analytical rectangular model using a constant flow rate. In my rectangular model I have applied the
temperaturedependentconversionfromcappingstacktoreservoirpressurespresentedinAppendixB
using Core Labs fluidproperties.Dr. Hsieh appears toobtainoverall agoodmatch to thepressure.
However, it is the latetime behaviour that is most significant in determining reservoir boundaries and
thefinalpressure,aswewillseelater:weseeaslightundershootofthefinalpressuremeasurement.
G.1.2 Comparison of pressure derivatives. Figure G.2 shows the pressure derivative for the two
datasets.Atlatetime,weseeroughlysimilarvalueswithasignatureofchannelflow.AtearlytimeDr.
Hsiehsdataindicateahigherderivativeofaroughlyconstantvalue,suggestingradialflow.
Thesamplingofthedatadoeshaveasignificantbearingonthepressurematch.Thepresenceofthefar
reservoirboundary,whichdeterminesthefinalreservoirpressure,isonlyseenatthelatesttimes.Ifyou
have few data points to match in this time period, then, when you fit the data, you may allow a
mismatch of these few points in favour of a better fit to an earlier time period, where there are more
286
ThedatafromDr.Hsiehcomesdirectlyfromaspreadsheetprovided(seeExhibit8617(Hsiehpressureanalysis))
[44].ThisispresentedaspressureasafunctionofHornertime.Ihaveconvertedtotimesincechokeclosure
usinghisassumedflowperiodt
p
of86days.
287
Dr.Hsiehsdeposition[43],page363,line13.
MJBluntExpertReport
140
data.However,thisearliertimeperiodisunaffectedbythepresenceofthefarboundaryandsodoes
nothelpspecifythefinalreservoirpressure.ThisiswhereDr.Hsiehintroducesasourceofinaccuracyin
hisanalysis:hehasfewdatapointsatlatetimeandsodoesnotsteerhispressurematchtofitthisvital
latetimebehaviour.
FigureG.1.Comparisonofthepressuredatausedbyme(crosses)andDr.Hsieh[44](circles).Theinsetshowstheearlytime
behaviour in more detail. Note an offset at late time, due to an expert reassessment of the raw pressure recordings [T]
madeafterDr.Hsiehpublishedhisanalysis.AlsoshownareDr.Hsiehsmatchtothepressure(theredline)andmymatchto
hisdatausingaconstantflowraterectangularmodel(blackline).
Dr.Hsiehusedanoptimizationpackage,linkedtoanumericalreservoirsimulator,tofindthereservoir
shape,welllocationandpermeabilitythatledtothebestmatchofpressure.LikeDr.Hsieh,Itoohave
used an optimization routine to find the bestfit parameters. However, in the optimization as
described in Appendix D I use an analytical model whose results can be evaluated to arbitrary
accuracy. Table G.1 provides my bestmatch parameters for the three sets of fluid data studied: this
Table conveys the same information as Table D.2, except that I have matched to Dr. Hsiehs pressure
datainthisSection.
MJBluntExpertReport
141
FigureG.2.Comparisonofthepressurederivativecomputedfromthedatausedbyme(crosses)andDr.Hsieh(redcircles).
At late time the derivatives arebroadly similar, althoughmy final fewvalues lie above thoseof Dr.Hsieh. The bestmatch
models from Dr. Hsieh (red line) and my analysis (black line) are also shown. Note that Dr. Hsieh tends to underestimate
thederivativeatlatetime,leadingtotoolowafinalreservoirpressure.
G.1.3Comparison of pressurematches.Dr.Hsiehsmodelunderstatesthepressurederivativeatlate
time, just before the well was cemented. He predicts a decrease in the derivative when this is not
apparentinthedata.Mybestmatchestohisdata,showninTableG.1,indicateahigherfinalreservoir
pressurethanwhenIusemyowndata(comparewithTableD.2)andsohadIusedhispressuredata
Iwouldpredictlessoilreleasedthanreportedhere,notmore.SinceDr.Hsiehonlyconsideredpressure
and the readings were only accurate to only 5 psi, his matches are somewhat insensitive to the final
estimatedreservoirpressure,asdiscussedinAppendixD.Overall,Dr.Hsiehmatcheshispressuredata
witharootmeansquareerrorof2.2psi;Idobetterwithanerrorofonly1.5psi.However,Ialsomatch
tothepressurederivativewitharootmeansquareerrorof8.5psi;Dr.Hsiehserroris16.7psi.Thisis
outsidetherangeofacceptablematchesshowningraphsD.1toD.3.

MJBluntExpertReport
142
Parameter t
1
t
2
Ap p
f
Pressure
error
Derivative
error
Usingproperconversionfrom
cappingstacktoreservoir
288
10,734,170 209,361 1,392 10,458 1.46 8.53

Fixedpressuredifference
betweencappingstackand
reservoir
11,300,000 775,000 1,499 10,351 1.81 6.47
Dr.Hsiehsanalysis 1,583 10,267 2.22 16.72
Unit s s psi psi psi psi
Table G.1. Parameters determined from the pressure match to Dr. Hsiehs data. Table D.1 provides definitions and
descriptionsofeachoftheseterms.Ihaveusedaninitialreservoirpressureof11,850psitofindp
f
,sincethiswasthevalue
usedbyDr.Hsieh.Dr.Hsiehsestimatesoffinalpressureandpressuredroparealsoshownforcomparison.
Dr.Hsiehselectedtoofewpressurepointsatlatetimeandsotoleratedanerrorintheirtrendinfavour
of a better fit to the earlytime data. This makes his analysis insensitive to his assumed final reservoir
pressure. Instead, I have focussed my attention on the important parameter in this analysis: the final
reservoirpressure.Todeterminethisprecisely,IhaveensuredthatImatchthemeasuredpressuresand
their trend accurately at the end of the buildup period. I place less emphasis on the earlytime
transient(althoughIalsomatchthiswell),sincethishaslessimpactonthematerialbalancecalculation.
G.1.4Fixedpressureconversionfromcappingstacktoreservoir.Thereisanothersignificantdifference
between my analysis and that of Dr. Hsieh. Dr. Hsieh assumed that the reservoir pressure was a fixed
3,199 psi larger than the capping stack pressure.
289
In contrast, since the oil is compressible, the
pressure difference is itself a function of temperature and pressure, as described in Appendix B: my
pressuredifferencesarelargerandvaryfromaround3,200to3,400psidependentonthecappingstack
pressure and fluid properties used. This discrepancy also leads Dr. Hsieh to estimate lower reservoir
pressures than in my analysis: for the same capping stack pressure, he underestimates the reservoir
pressureandthereservoirpressurederivative.InTableG.1,IalsoshowthebestmatchtopressureifI
retain Dr. Hsiehs fixed pressure difference: this isolates the impact of the poor match to the pressure
derivativefromtheeffectofhissimplificationofthepressuredifferencebetweenthecappingstackand
thereservoir.Ifindafinalreservoirpressureof10,351psiifImatchDr.Hsiehsdatausinghisassumed
pressuredropfromreservoirtocappingstack.
Dr. Hsieh predicted a final reservoir pressure of 10,267 psi,
290
while I obtain 10,458 psi using his data.
Thisdiscrepancycanbebrokendownintotwocontributions.First,correctingDr.Hsiehspoormatchto
the latetime pressure derivative increases the predicted final reservoir pressure by 84 psi (the
differencebetweenhisestimatedfinalpressureof10,267psiandthebestmatchusinghispropertiesof
10,351psi).Second,introducingmycalculatedpressurechangefromcappingstacktoreservoir,based
onthemeasuredfluidproperties,leadstoafurtherincreaseof107psiinthepredictedfinalpressure.
288
ComputedusingtheconversionfromcappingstacktoreservoirpressuresforCoreLabspropertiespresentedin
AppendixB,arectangularflowmodelandaconstantflowrate.Thecasewithafixedpressureconversionusesa
linearflowmodel.
289
Thisistakendirectlyfromthespreadsheetvalues(seeExhibit8617(Hsiehpressureanalysis))[44].
290
Thisvalueistakendirectlyfromhisspreadsheet(seeExhibit8617(Hsiehpressureanalysis))[44];inhisreport,
IGS642000215(10/13/2010DraftHsiehreport)[11],heroundsthisnumbertoquoteavalueof10,300psi.
MJBluntExpertReport
143
Isuggestthatthroughneglectingthelatetimepressurederivative,combinedwiththeassumptionofa
fixed pressure drop from capping stack to the reservoir, Dr. Hsieh obtained a final reservoir pressure
thatwassignificantlytoolow.
G.1.5 Radial flow period to find permeability. The final part of the pressure analysis is to study
separatelytheearlytimeradialflow.Theconstantvalueofthederivativeindicatesradialflow,from
which the permeability can be calculated (Table D.1). The transition to channel flow occurs at around
60,000 s from my model match. From Figure G.3, the constant derivative value in the first day (times
from10,00080,000s)isaround40psi.Beforeproceeding,Iwillnowassumethatthisrepresentsthe
reservoir pressure derivative. As discussed in Section 4.3, this is an error, due to the neglect of oil
compressibilityandleads,inanyevent,toanoverestimateofpermeability.Letusignorethisfornow,
and simplyanalyseDr. Hsiehsdata usinghisassumptions. Iemployhisviscosity of 0.168mPa.s and a
reservoir height of 90 ft (see Table G.2 for a full listing and comparison of reservoir properties). I also
needtousethefinalflowrate.Tobestrictlyconsistent,IwilltakethevaluethatDr.Hsiehcomputesfor
thefinaldayofthespill[44]:52,603stb/dayandconverttoreservoirconditionsusinghisassumedvalue
ofB
o
,2.35.
I obtain a permeability of 407 mD. Dr. Hsiehs best matched permeability value is 593 mD.
291
In his
deposition he explains how his estimates of bestmatched parameters varied as he used smaller time
steps and a more refined grid in his numerical model.
292
His estimate of permeability increased from
515mDto570mDandthen593mDandhisfinalpressuredecreasedfrom10,400psitobelow10,300
psi as he refined his model. His estimate of oil released changed from 4.6 MMstb to 4.76 MMstb and
thenfinallyto4.92MMstb.
Simulation models solve the governing equations for flow those presented in Appendix C
numerically.Itiswellknownthatthesimulationresultsaresensitivetothetimestepandgridsize,asis
thecaseforDr.Hsiehsmodel.IhavenotseenanyevidencethatthefinalresultspresentedbyDr.Hsieh
are converged that is, that they represent the true solutions of the governing equations where the
numericalerrorsareinsignificant.
Toavoidnumericalproblems,itisstandardpracticeintheoilindustrytomatchthepressureresponse
to analytical mathematical expressions, where computational errors are eliminated, and to use
optimizationtofindthebestfitparametersinthesemathematicalmodels.ThisistheapproachIfollow
inthisreport.
Dr.Hsiehsmodelifcorrectmustgiveanexactlyconstantderivativeatearlytime.Insteadweseea
high initial value (see Figure F.2) followed by a decline to a roughly fixed value. Analytically this value
must be consistent with the permeability, but it is not. He is in effect matching the data with
numericalerrorsinhissimulation;theresultsdonotprovideavalidrepresentationofthereservoir.
291
Exhibit8615(10/22/2010HsiehDraftReport,Table2)[67].
292
Dr.Hsiehsdeposition[43],pages410420.
MJBluntExpertReport
144
Dr.Hsiehdidnotfollowstandardpracticeintheoilindustryinhisanalysis.Thedatathatheconsidered
were different from mine; he neglected to examine the match to the derivative; he assumed a fixed
pressure difference between the capping stack and the reservoir; and his results were affected by
numerical errors, leading to an inconsistent and overstated value of permeability, even in the context
ofhisotherassumptions.
G.2 Comparisonofreservoirproperties
Dr. Hsieh assumed a fixed initial oil in place, and full connectivity, and assumed values for the
compressibilitiesoftherockandfluids.Oncethefinalreservoirpressurewasdetermined,itwouldbe
possibletocalculatetheoilreleasedusingmaterialbalance,Eq.(3.1).ThiswasnotdonebyDr.Hsieh
insteadheusedhissimulationmodelwithafixedoutflowperformancetopredictflowrateasafunction
oftime.However,sincehismodeldespitetheproblemsoutlinedabovedoesstrictlyconservemass,
theresultsareconsistentwithamaterialbalanceanalysis.
I start from Tables 1 and 2 of Dr. Hsiehs draft report.
293
Table G.2 compares his data with my
calculations based on the analysis I have presented.
294
Those properties emphasized in italics display
significantdiscrepancies,whichIwilldiscussbelow.
ManyoftheassumptionsmadeinDr.Hsiehsmodellingworkseemperfectlyreasonableinthelightof
the detailed analysis presented in this report. However, there is no evidence that he performed a
careful scrutiny of all the available rock and fluid property data, or compared his reservoir model with
thestructureinferredfromseismicsurveys.
There are, however, some significant discrepancies that help explain why Dr. Hsiehs estimated oil
released,around4.9MMstb,differfrommyvalues(2.83.8MMstb).
1. His assumed rock compressibility of 12 microsips lies outside the measured range on core
samples. There is no direct evidence to support this value. My midrange value is 6.35
microsips.Thisistheprincipalreasonwhyhefindsalargervolumeofoilreleased.
2. Asmentionedabove,heunderstatesthefinalreservoirpressure.
3. He assumes a constant reservoir permeability of 593 mD. As discussed above, this is
inconsistent with his own data. The most robust determination of permeability from Dr.
Gringartenis238mD.
4. The reservoir geometry proposed by Dr. Hsieh is shorter, less wide and with a greater height
thanimpliedfromthegeology,indicatingincomparisonwiththeseismicsurveyincomplete
connectivity of the field. He assumed that BPs calculated gross rock volume was correct.
However, the main problem is that the reservoir geometry used by Dr. Hsieh is not consistent
293
IGS642000215(10/13/2010DraftHsiehreport)[11]:thisdraftreportcontainsallthedatausedtoconstructDr.
Hsiehssimulationmodel,whilethematerialthatwaspublishedinthepublicdomainhassomeofthisinformation
missing.Myunderstandingisthatsomeinformationwasremovedfromthepublicversionssinceitwasbasedon
proprietaryBPdata(Hsieh,2011).
294
ForsimplicityIonlycomparethesetofpropertiesdescribedinIGS642000215(10/13/2010DraftHsiehreport)
[11].IdonotethatrangesofpossiblereservoirparameterswerepresentedinadditionalworkbyDr.Hsieh(see,
forinstanceIGS629003048(HsiehModelingPresentation)[12];Hsieh,2010;thisisdiscussedlater.
MJBluntExpertReport
145
with the geological interpretation (Section 5). This is illustrated in Figure G.3, which compares
BPs seismic map with the reservoir dimensions in Dr. Hsiehs work. He assumes a fixed
thicknessof90ftclosetothevaluemeasuredatthewell.However,asmentionedinSection
5, the reservoir is longer, wider and, on average, less thick that the total height of sandstone
encounteredatthewell.Thepressureanalysisidentifiestheboundariesofthefieldthesedo
not encompass the entire volume of oilbearing sandstone identified from the seismic survey.
Hence,Dr.Hsiehsconnectedoilvolumecouldbetoohigh.
Property ValuefromDr.Hsiehswork Valuecalculatedinthisreport
Oilvolume 110MMstb Connected volume 109114
MMstb
Oilformationvolumefactor 2.35 Valueatfinalreservoirpressure
liesbetween2.3and2.4
Porosity 0.21 0.217
Porevolumecompressibility 12microsips 6.35microsips
Oilviscosity 0.168mPa.s 0.205mPa.s
Oilsaturation 0.9 0.878
Watercompressibility 3microsips 3microsips
Oilcompressibility 14.6microsips 13.714.8microsips
Permeability 593mD 300mDfrompressureanalysis
Totallength,L 6,788m 8,178m
Table G.2. Comparison of values used in the reservoir modelling of Dr. Hsieh taken from [11, Tables 1 and 2 and Figure 1]
comparedtocalculations presentedinthisreport.Thepropertiesindicatedinbolditalicsaresignificantdiscrepanciesthat
arediscussedinthetext.
These discrepancies help explain the differences in our values of oil released. Using the data in Table
G.2,IfindthatDr.Hsiehassumedaneffectivecompressibilityof28.27microsips.Notethatthisismuch
larger than the values I take (19 to 24 microsips see Section 4.2). The oil in place is 110 MMstb
(coincidentally in my range of calculated connected oil volume). Then, for the final simulation, the
pressuredropis1,583psi.
295
Multiplyingthesenumberstogetherusingthematerialbalanceequation
Eq.(3.1)
296
givesatotaloilreleasedof4.92MMstbexactlyDr.Hsiehsnumber,asithastobe.The
same exercise can be performed for his other simulation runs using different parameters. Instead of
multiple optimizations, heonlyneededtoperformone bestmatch to findthefinal pressure,andthen
the cumulative oil released could be computed from his assumed reservoir properties directly. Hsieh
(2010) quotes an uncertainty of +/ 10% in the estimates of oil released. This is based on assuming a
25% variation in his assumed reservoir properties. In his Table 4, the sensitivities can be determined
immediately from their impact on the material balance equation. A 25% change in STOIIP leads,
obviously, to a 25% change in oil released, as does a 25% change in total compressibility (obtained by
addingthecontributionsfrompore,oilandwatercompressibilitytogetherinhisTable4).Thisisnota
true sensitivity analysis, since with arbitrary reservoir properties, any value of oil released could be
295
UsingDr.Hsiehsassumedinitialpressureof11,850psiandafinalreservoirpressureof10,267psi.
296
WhileDr.Hsiehdidnotexplicitlyperformamaterialbalancecalculation,materialbalanceissimplyastatement
ofconservationofvolumeandsoDr.Hsiehsresultsareconsistentindeedhavetobeconsistentwithamaterial
balanceanalysis.
MJBluntExpertReport
147
Dr.Hsiehestimatedoil
released:4.92MMstb
Usingmy
connected
volume
Correctingfor
compressibility
Correctingforthe
pressuredrop
Corrected
calculationofoil
released
CoreLabsdata 0.04MMstb 1.08MMstb 0.39 MMstb 3.41MMstb
Schlumbergerdata +0.10MMstb 1.26MMstb 0.61 MMstb 3.15MMstb
Intertekdata +0.16MMstb 1.31MMstb 0.55 MMstb 3.22MMstb
Table G.3. Table quantifying the impact of the different assumptions in Dr. Hsiehs analysis on the calculated oil released,
using the material balance equation. Dr.Hsieh estimated an oil release of 4.92MMstb: thetable shows byhow much this
value changes if we use the connected oil volume, compressibility and pressure drop values presented in this report. The
final column provides my determinations of oil released. I have used the midcase pore volume compressibility of 6.35
microsipstocorrectthecompressibility.ForcompletenessIpresentthecalculationsforallthreesetsofmeasuredfluiddata.
ThecorrectionsaveragedforallthreesetsoffluiddataareshowninthesummarytableinSection2andF.2.

MJBluntExpertReport
148
AppendixH.Critiqueofpublishedflowrateestimates
H.1 Overview
The Proceedings of the National Academy of Sciences (PNAS) published a special feature on December
11
th
2012 entitled Science applications in the Deepwater Horizon oil spill. This is a collection of
researcharticlesoutliningthescientificendeavoursofGovernment,academiaandindustryinresponse
tothisaccident.ThisspecialissuewasmentionedbySecretaryChuinhisdepositionascontainingthe
most recent, peerreviewed, published estimates of flow rate.
297
In this Appendix, I will briefly review
thepapersrelevanttomywork.
Threeindependentapproachestoestimateflowratearepresented.
1. Insitumeasurementsofoilflowintotheocean.
2. Theuseofpressuredifferencesacrossthecappingstacktocalculatetheflowrateneartheend
ofthespill.
298
3. Reservoiranalysis,usingthecappingstackpressurebuildupandestimatedreservoirproperties
to determine flow rates and oil released. This is the work of Dr. Hsieh at the USGS and is,
conceptually,mostsimilartothecalculationpresentedinthisreport.
Inbriefmycritiqueofthethreemethodsaboveisasfollows.
1. Overestimateoftheflowofoilrelativetogas.Theinsitumeasurementsfoundthecombined
flow oil, gas (or gas hydrate) andentrained sea water. To convert this to an oil volume alone,
the relative proportions of gas and oil in the flow need to be known. This can be determined
from the gas/oil ratio, a standard quantity in petroleum engineering (see Appendix J); it is the
ratio of the volume of gas to the volume of oil of the produced fluid, measured at surface
conditions. This was done for Macondo: the value, measured independently at three
laboratories, lies between 2,831 and 2,945 scf/stb, with an average of 2,894 scf/stb (Table
A.7).
299
Downhole the fluid is a single (oil) phase making it possible to take a representative
sample;atthesurfacewehavegasandoilflowingtogetheranditisextremelydifficulttocollect
the correct overall proportion of the two phases.
300
Many of the papers contain flow rate
estimatesusedagas/oilratioof1,600scf/stb.
301
Thiswasmeasuredonamixtureofoilandgas
collected in the deep ocean directly above the Macondo well from a remotelycontrolled
297
SecretaryStevenChusdeposition[58];page293,line17onwards.
298
DescribedinfurtherdetailintheSandiareports(revisedGriffithsreport[2]andrevisedRatzelreport[3]).
299
BPHZN2179MDL03290054(BPPostWellSubsurfaceTechnicalMemorandum)alsoquotesgas/oilratios
measuredonadditionalsamplestakenatdifferentdepthsinthereservoir:thesevaluesare2,840,2,909and3,017
scf/stbforsamplestestedoffshoreasaqualitycontrol([6],page40);andservicecompanymeasurements
between2,890and3,096scf/stb([6],page41).
300
Toquotefromastandardreservoirengineeringtextbookreferringtothecollectionofoilandgas,asopposed
toasinglephasesample(Dake,2001;page35):Nomatterhowcarefullythesamplingisconducted,however,itis
oftenverydifficulttoobtaintheoilandgasmixedintherequiredratio.
301
See,e.g.,Ryersonetal.(2012)ChemicaldataquantifyDeepwaterHorizonhydrocarbonflowrateand
environmentaldistribution.
MJBluntExpertReport
149
vehicle.WhileIdonotknowwhythereissuchadiscrepancy,theuseofdownholesamplesis
theindustrystandardbenchmarkandshouldbeacceptedasbeingaccurate.Thepapersunder
statethegas/oilratio,leadingtoanoverestimateofoilflow.
2. Unsupported assumption of a declining flow rate. McNutt et al. (2012a) state the most
definitive measurement of the flow occurred just before the well was shut in, referring to the
use of differential pressure measurements across the capping stack to determine flow rate.
However, even if we accept the determination of final flow rate, this does not allow a
quantification of total oil released: to know this, we need to know the flow rates during the
entire period of the spill. It has been assumed by all the Government investigators that the
pressure difference between the wellbore (at reservoir depth) and the blowout preventer
decreasedovertime,astheaveragereservoirpressuredeclined.Thisleadstoaflowratethat
also declines with time, if we neglect the impact of any possible erosion of the blowout
preventerandothersurfaceequipment.Thepressuredeclineisconsistentlyoverstated(point
3below),butthisisnotthefundamentalproblem.Toreachthewellbore,theoilhastofinda
tortuous pathway to enter the damaged well; it is implausible to suggest that for the entire
period of the spill the oil flow was completely unimpeded over the full 93 ft height of the
reservoir. The Emilsen report states that at the beginning of the incident, the likely length of
reservoiropentoflowwasbetween13and16.5ft.
302
Furthermoretherocknearthewellbore
could have been clogged with cuttings from drilling: core samples taken from the well had
permeabilities around five times lower than measured after cleaning (Appendix A.5). These
effects,plusanyfurtherresistancetoflowinthewellboreitselfatreservoirdepths,leadtoan
additionalpressuredropbetweenthereservoirandthewell.AsdiscussedinAppendixE.3,this
pressure drop is likely to be larger than the reservoir depletion. At the end of the spill, the
injectivitytestindicated,however,thatthisflowresistancewasnegligible.Thisimpliesthatthe
flowbecamelessrestrictedduringthespillperiod,leadingtoapressuredifferencebetweenthe
welldownholeandtheblowoutpreventerthatincreasedovertime,givinganincreasingflow
rate. It is impossible to quantify this effect with any certainty, but this discussion serves to
illustratehowitisproblematictoquantifytheflowrateduringthespillfromameasurementat
anyonetime.Itisnecessarytofindamethodologythataddressesthecumulativeflowdirectly
namelythematerialbalanceanalysispresentedinthisreport.
3. Flawsin thereservoir analysis.ThereservoiranalysisofDr.Hsiehusedproperties,specifically
pore volume compressibility, that lay far outside the range of the measured data and did not
use the best approach to analyse the capping stack pressure data and so overestimated the
pressure decline. As a consequence he overestimated the volume of oil released. Since this
work is most aligned to my own calculations, it requires a separate and more detailed critique
whichisprovidedinAppendixG.
302
SeeAppendixE.6:takenfromEmilsenreportonthewellfailure[27],Section2,page7.
MJBluntExpertReport
150
H.2 Detailsofselectedpapers
IwillnowbrieflyreviewseveralpapersinthePNASspecialissue,relevanttomywork.
H.2.1. Lubchenco et al. (2012) Science in support of the Deepwater Horizon response. This is the
introduction to the special issue and presents the context for science during the spill response. The
paper states .. a final estimate derived from pressure measurements of 62,000 bpd at the outset and
53,000bpdatwellshutinforatotalof4.910%millionbarrels.Thepapermentionsthatthe(capping
stack) pressure measurements allow the most reliable determination of final flow rate. However, the
flow rate may have increased during the earlier part of the spill period (see Appendix E.3); to quantify
thetotalflow,areservoiranalysisisrequireditcannotbeinferredfromaflowrateattheendofthe
spill.
H.2.2. McNutt et al. (2012a) Applications of science and engineering to quantify and control the
Deepwater Horizon oil spill. This paper provides an overview of thedifferent scientific and technical
assessments made by different agencies in support of the oil spill response. The paper quotes a final
flow rate of 53,000 stb/day 10%. Note that the uncertainty in final flow rate is the same as the
uncertaintyintotalflow:noadditionaluncertaintyisassociatedwiththeextrapolationoftheflowrate
atonetimetothewholespillperiod.Thisisanunsupportedassumptionforthereasonsgivenabove.
ThepaperalsoaddressesthereservoirmodellingworkofDr.Hsieh.ItstatesBothHsiehandBPplotted
thetimedependenceofthepressureinaHornerplot,whichisprovidedinFigure3ofthepaper,while
thediagnosticsofradialandlinear(channel)flowarediscussedinthetext.MyunderstandingisthatBP
initiallyusedaHornerplotsimplybecauseitcouldbeprovidedrapidlywithouthavingtospecifyaflow
rate in their proprietary software.
303
However, the industry standard to identify flow regimes is the
pressure derivative, as used later by BPs engineers who studied the capping stack pressure.
304
The
presentation of the plot indicates a poor understanding of pressure transient analysis. A flow rate of
50,000stb/dayismentionedeventhoughtheplotdoesnotrequireaflowratetobespecified.Second,
differentmatchestothedataareshownforvaryingreservoirwidths.Fundamentallythisisnotcorrect:
thechangingshapeofthecurveisdeterminedfromthepressurewavehittingafarboundary,indicating
thelengthofthefield.Thewidthisdeterminedfromtheearliertimeresponsewhenthepressurewave
hitsthesideofthechannel.Theonlyreasonwhywidthappearstovaryistheunnecessaryconstraintin
Dr.Hsiehsmodelthatthetotalreservoirvolumeisfixed;thereforealongerreservoirhastobethinner.
The paper mentions a 1,800 psi pressure decline. The reservoir pressure decline is in fact around
1,400psi(seeSection4.3);Dr.Hsiehfoundavaluecloserto1,600psi(AppendixG).Thisisanexample
ofthepersistentoverstatementofpressuredeclineinGovernmentreportsandpapers.Moreover,the
use of this value outside an explicit reservoir engineering calculation is to estimate the downhole
pressure drivingflow up the wellbore andthroughthesurface equipment. Thispressure insteadmay
have increased, because of the changing flow resistance in the nearwell region, as mentioned above.
Hence,quotingthisvalueofpressuredeclineismisleadinginthewidercontextofflowrateestimates.
303
Dr.Merrillsdeposition[54],page34,line1.
304
SeeSection4.3andDr.Merrillsdeposition[55],page344,line18.
MJBluntExpertReport
151
The paper also says On August 3 (day 106), base oil was injected into the well as the first part of the
injectivitytest.Thepressurerisewasonly35psi,hundredsofpsibelowthemostoptimisticexpectations
ofthelowestpressurerise neededtostopand thenreverse the hydrocarbonflow.. The impedance to
flowatthewellreservoirinterfaceorelsewhereinthelowerrockformationwasverysmall.Whywas
thepressureriseexpectedtobehundredsofpsi,andwhataretheimplicationsofthistest?AsIdiscuss
in Section E.3, analysis of thepressure at thetimeof the blowout indicates highly restricted flow into
thewellborefromthereservoir.Hadthisflowresistancepersistedthroughouttheincident,then,when
oil was injected back into the formation, a pressure rise of at least 250350 psi would be anticipated.
Instead, the pressure rise was an order of magnitude lower. This demonstrates that the resistance to
flowprobablydecreasedduringtheincident,andthepressureinthewellboreatreservoirdepthscould
haverisen,givinganincreasingflowrate,evenifanyerosionofthesurfaceequipmentisignored.This
isavitalpieceofevidenceindicatinganincreasingflowrate,contradictingGovernmentassertionstothe
contrary.
H.2.3.McNuttetal.(2012b)ReviewofflowrateestimatesoftheDeepwaterHorizonoilspill.This
paperpresentsanoverviewofthedifferentflowrateestimates.Thepaperreviewsseveralmethodsfor
calculatingflow,includingthereservoiranalysisofDr.Hsieh.Dr.HsiehsworkisdiscussedinAppendix
G:itisnotbasedonanindependentscrutinyofthemeasuredfluidandrockdata,andoverstatesthe
overall pressure drop. In the supplementary material, Figure S4 shows the Horner plot again: close
scrutinysuggeststhatabettermatchtothefinalpressuremeasurementswouldgiveahigherestimated
finalpressuresomethingthatismoreevidentifthebestdiagnosticmethodthepressurederivative
isemployed.FigureS5showsthesimulatedflowrateasafunctionoftime:thismakestheunsupported
assumptionthatthereisnoresistancetoflowinthenearwellboreregionandfixedoutflowconditions.
The flowrate estimatesusing reservoir modellingby consultantsto theFlowRateTechnical Groupare
presentedinTable3ofthepaper.Theseshowawiderangeconsistentwithvaluesaboveandbelowmy
averageflowrates.However,theseestimatesweremadebeforethewellwasclosed,andthepressure
data was available. Hence, these somewhat unconstrained calculations will be much less certain than
thoseinformedbythereservoirpressurebehaviouraftershutin.
H.2.4.Hickmanetal.(2012)ScientificbasisforsafelyshuttingintheMacondowellaftertheApril20,
2010 Deepwater Horizon blowout. This paper describes the analysis used to conclude that the well
couldremainsafelyclosedaftershutin.Figure3showstheHornerplotofthecappingstackpressure
thethirdtimeinthespecialissuethisplotisshown.Thedirectlymeasuredpressure(ratherthanpublic
domain reports of the pressure) is shown. The repeated presentation of this plot implies that is a
particularly useful tool in this pressure analysis: it is not. As discussed in Section 4.3, no competent
reservoir engineer would present a pressure analysis without showing the pressure derivative; the
Hornerplothas very poordiagnostic ability. Hadthe standardderivativeplot been shown, theunder
estimateofthefinalreservoirpressure,leadingtoanoverstatementofoilreleased,wouldhavebeen
apparent.
ThepaperstatesThereservoirpermeabilityandformationcompressibilitywerealsoadjusted,butthese
measurements remained within ranges typical of reservoir sands. What is not reported is that the
MJBluntExpertReport
152
authors did not perform any independent analysis of rock and fluid data for Macondo; their values lay
outsidetherangeofthesemeasurements.
ThefinalcommentofthepapermentionsExcellentaccesstocompanydataforusebygovernment
oversightteamsandtheircollaborators.Thismayhavebeenthecase,buttheselectiveuseofthisdata
isconcerning. Torecap: muchoftheanalysisusedagas/oilratioreliedondatafrom asubseasample
withtoolowavaluecomparedtomeasurementsonadownholesample;Dr.Hsiehsreservoiranalysis
did not use reservoir properties consistent with the measured data; and last, there was no holistic
approachtoallthefactsforinstance,thesignificanceofthelowpressuresneededtoflowoiloncethe
wellwascapped,comparedtothelikelyflowresistanceatthetimeoftheaccident,wasignored.

MJBluntExpertReport
153
AppendixI. Noteonunits
Traditionally,intheUSoilindustry,quantitiesaremeasuredinfieldunits.ThedataIuseinthisreportis
generally given in these units. However, most scientific calculations are performed in SI (metric)
units,atleastoutsidetheUSandtheoilindustry.TheadvantageofusingSIunitsisthatquantitiescan
becomputeddirectlyfromequationswithouthavingtoaccountforunitconversionfactorstomakethe
calculationsconsistent.
To allow easy comparisons with other work on Macondo, I will present my main results in field units.
However,IwillalsoquoteSIunitswhereappropriate,andallscientificcomputationswillbeperformed
usingSIunitsforconsistency.
Unfortunately,thisdoesintroducesomeconfusion,asIwilltendtoquotequantitiesformycalculations
inSIunits(m,ands)whileusingtraditionalunitsinmanyplacesft,psi,stbandmDforreadability.
Asaguide,belowIprovideatableofunitsandconversionfactors.
Quantity Length Area Oil
volume
Permeability Pressure Time
Fieldor
customary
unit
ft(feet) Acres bbl
(barrels);
stb,rb
mD(milliDarcy) psi(poundsper
squareinch)
Days
SIunit m
(metres)
m
2
m
3
m
2
Pa s
(seconds)
Conversion:
fieldtoSI
0.3048
4,046.85642 6.2898
9.86923310
16
6,895 86,400
Asisevidentintheanalysis,itiscrucialinanydiscussionofoilvolumestoclarifyiftheoilisatreservoir
orsurfaceconditions.Thisisdisplayedintheunitused.So,forinstance,stb(stocktankbarrels)refers
to a barrel of oil at surface conditions whereas rb (reservoir barrel) refers to a barrel at reservoir
conditions. Now, the volume of thisbarrel is the same in both cases the apparently different unit is
giventoclarifytheconditionatwhichthevolumeismeasured.

MJBluntExpertReport
154
AppendixJ. Nomenclature
Symbol Meaning Units(field) Units(SI)

A Area Acres m
2
B
o
Oilformationvolumefactor rb/stb Dimensionless
c Compressibility psi
1
Pa
1
c
t
Totalcompressibility psi
1
Pa
1
C Heatcapacity Jkg
1
K
1
g Accelerationduetogravity ms
2
h Reservoirthickness ft m
H Wellboreheight ft m
K Permeability mD m
2
L Lengthofchannel ft m
N STOIIP,oilinitiallyinthereservoir stb m
3
N
p
Oilproducedatsurface stb m
3
p Pressure psi Pa
Q Flowrate rb/day m
3
/s
r Radialcoordinate ft m
R
s
Solutiongas/oilratio scf/stb dimensionless
S Skin dimensionless dimensionless
S
o
Oilsaturation dimensionless dimensionless
S
w
Watersaturation dimensionless dimensionless
t Time days s
t
p
Durationwellwasproducing days s
T Temperature
o
F K
V Volume stborrb m
3
W Widthfrompressureanalysis ft m
W
max
Totalreservoirwidth ft m
x Distancecoordinate ft m
y Coordinateforradialflow dimensionless dimensionless
z Depthcoordinate ft m
Greeksymbols
o Inversediffusivity s/m
2
| Drawdowninchannelflow psi.day
1/2
Pa.s
1/2
Eulersconstant dimensionless dimensionless

o Drawdownfromskin psi.day/rb Pa.s/m
3
c Strain dimensionless dimensionless

Ap Pressuredrop psi Pa
At Timesincewellclosure day s
MJBluntExpertReport
155
| Porosity dimensionless dimensionless

k Turbulentresistancefactor psi.day
2
/rb
2
Pa.s
2
/m
6
Thermalconductivity(AppendixB) Wm
1
K
1
Viscosity cp Pa.s
q Drawdowninradialflow psi Pa
Density kg.m
3
Weightoffluid psi Pa
t TimeinDuhamelintegration day s
t
W
Timeforpressuretotraversethe day s
widthofthereservoir
t
L
Timeforpressuretoreachthe day s
Endofthechannel
t
1
Timeforpressuretoreachoneend day s
ofthechannel
t
2
Timeforpressuretoreachtheother day s
endofthechannel
Subscripts
cs Cappingstack
c Cement(AppendixB),connected(AppendixD)
D Forunitflowrate
f Finalorformation(rock)
o Oil
p Produced
i Initial
L Length
r Reservoir
s Surface(orsedimentAppendixB)
t Theoretical(model)
T Thermal
w Waterorwell
W Width
0 Atthewellbore
1 Channelononesideofwell
2 Channelonothersideofwell
MJBluntExpertReport
156
AppendixK. Glossaryofoilfieldterms
Every scientific field has its own unique jargon and terminology. Below, I explain some of the terms
used in my report. The petroleum industry, unlike many other disciplines, such as medicine, tends to
useshort,evocativewordsfortechnicalterms,ratherthanelaborateLatinateexpressions.
Blackoil While crude oil is indeed black in colour, this term is used to refer to how we
describe the properties of oil as a function of temperature and pressure. In
particular,itdescribesanoilcontainingdissolvedgasthatcomesoutofsolution
as the pressure drops. BP prepared black oil tables that presented the
properties of the oil as a function of temperature and pressure, matched to
measureddata.
Buildup This refers to the period once a well is closed. The pressure buildup is the
increaseinpressuremeasuredatthewellafterflowhasceased.Afterverylong
times a final equilibrium value is reached the final reservoir pressure. The
natureofthebuildupprovidesvaluableinformationonthereservoirstructure
andproperties(Section4.3).
Compressibility Thisisthefractionalchangeinvolumewithachangeinpressure.Weusethisto
describe thecompression(shrinkage)of therock. It is alsousedin reverse
todescribetheexpansionoffluidaspressuredrops.
Downhole Thisreferstoconditionsinthewellwhereitcontactsthereservoir.Sometimes
thisconditioniscalledbottomhole.Downholeorbottomholepressurerefers
to the pressure in the well at reservoir depth. I have not used the phrase
bottomholepressureasthiscanbeshortenedtoBHP,whichinevitablycauses
confusion,evenforexperts,withtheblowoutpreventer(BOP).
Gas/oilratio Thisistheratioofthevolumeofgastothevolumeofoilforareservoirsample
of hydrocarbon that is brought to the surface. It is measured in units of scf
(standard cubic feet) per stb (stock tank barrels). The use of the words
standardandstocktankemphasizethatthevolumesaremeasuredatsurface
conditionsof60
o
Fand1atmospherepressure.Inthisreport,thevaluesIquote
areforasinglestageflash,wherethegasandoilarekeptinequilibriumasthe
pressureandtemperaturearedroppedfromreservoirtosurfaceconditions.In
the oil industry it is standard practice to measure the gas/oil ratio on fluid
samplescollecteddownholeandtakentothelaboratoriesofservicecompanies
whospecializeinmakingthesemeasurements,aswasdoneforMacondo.
Logs These are measurements of electrical resistance and rock density taken using
sophisticatedinstrumentsinthewellborebeforethewellcasingandcementis
MJBluntExpertReport
157
putin place. Theselogsareused toinfer water(and hence oil) saturationand

porosity.
Outflowperformance This is the relationship between flow rate and pressure drop through the well
bore, blowout preventer and other equipment. A full quantitative analysis of
oilfield flow rates couples the reservoir behaviour (the inflow performance or
relationship between flow rate and pressure drop in the reservoir) to the
outflowperformance:theflowintothewellfromthereservoirmustbeequalto
theflowthroughthewell.
Skin This is an additional pressure drop near the well, because of a restricted flow
path or a lowered permeability. During drilling, rock fragments can be forced
into the rock itself around the wellbore clogging the pore space and reducing
permeability. It is very difficult to predict the magnitude of this effect. It is
modelled as an additional pressure drop with a skin factor that is adjusted to
matchtheobservedbehaviour.InMacondo,itislikelythatoilinitiallycouldnot
flow into the well over the whole 93 ft section of oilbearing sandstone, but
followed a tortuous path to reach the damaged sections of the well. This also
resultsinanadditionalpressuredropcalledthepartialpenetrationskin.Atthe
end of the spill, the flow rate tests with oil before the well was cemented
indicated that the skin was very low, suggesting opening of the flow paths
duringtheincidentandaskinthatdecreasedovertime.
STOIIP Thisstandsforstocktankoilinitiallyinplaceandisthevolumeofoilinitiallyin
thereservoirmeasuredatsurface,orstocktank,conditions.

MJBluntExpertReport
158
AppendixL. BiographyofMJBlunt
MartinJ.Blunt
ProfessorofPetroleumEngineering
DepartmentofEarthScience&Engineering,ImperialCollegeLondonSW72AZ,UK
Tel.+44(0)2075946500,Email:m.blunt@imperial.ac.uk
Education
1985 BANaturalSciences,CambridgeUniversity(FirstClassHonours)
1988 PhD,TheoreticalPhysics,CambridgeUniversity.
TheGrowthandPropertiesofFractalBoundaries.
Employment
19881992 ResearchPhysicist,BPResearch,SunburyonThames
19921999 Facultymember,DepartmentofPetroleumEngineering,StanfordUniversity:Assistant
Professor19921995;AssociateProfessor19951999
1999date ProfessorofPetroleumEngineeringImperialCollegeLondon:HeadofthePetroleum
EngineeringandRockMechanicsresearchgroup(PERM)19992006;Headofthe
DepartmentofEarthScienceandEngineering20062011;visitingprofessor,Politecnico
diMilano,Italy,2012date
HonoursandAwards
1985 ResearchScholarship,TrinityCollegeCambridge
1985 ClerkMaxwellandverHeydendeLanceyPrizes,CambridgeUniversity
1991 TallowChandlersPrize,BP
1996 Teachingaward,SchoolofEarthSciences,StanfordUniversity
1996 CedricFergusonMedal,SocietyofPetroleumEngineers
2001 DistinguishedLecturer,SocietyofPetroleumEngineers
2011 LesterUrenAwardandDistinguishedMember,SocietyofPetroleumEngineers
2012 DarcyAwardforlifetimeachievement,SocietyofCoreAnalysts
ProfessorBlunt'sresearchinterestsareinmultiphaseflowinporousmediawithapplicationstooiland
gas recovery, contaminant transport and cleanup in polluted aquifers, and geological carbon storage.
Heperformsexperimental,theoreticalandnumericalresearchintomanyaspectsofflowandtransport
inporoussystems,includingporescalemodellingofdisplacementprocesses,andlargescalesimulation
usingstreamlinebasedmethods.Hehaswrittenover150scientificjournalpapersandisEditorinChief
of the journal Transport in Porous Media. He is a director of two startup companies Streamsim
Technologies and iRock Technologies founded by his former PhD students. In 2011 he was awarded
the Uren Award from the Society of Petroleum Engineers for outstanding contributions to the
technology of petroleum engineering made before the age of 45. In 2012 he was awarded the Darcy
AwardfromtheSocietyofCoreAnalystsforlifetimeachievement.Prof.BluntisaFellowoftheEnergy
InstituteandaCharteredPetroleumEngineer.
MJBluntExpertReport
159
AppendixM. MJBluntsPublicationslist

M.1 Journalpapers
1. MJBluntandRCBall,AfamilyofexponentsfromafractalmodelofviscousfingeringandDLA,JPhysA:MathGen,
2059615969(1987).
2. MBluntandPKing,Scalingstructureofviscousfingering,PhysicalReviewA37(10)39353941(1988).
3. RBallandMBlunt,Afractalmodelfortheimpedanceofaroughsurface,JPhysA:MathGen,21197204(1988).
4. MBlunt,Theimpedanceofaselfaffinesurface,JPhysA:MathGen,2211791192(1989).
5. RCBall,MBluntandWBarford,Cansurfacestatesbeinducedbyinterfacialroughness?JPhysA:MathGen,22
25872595(1989).
6. RCBall,MJBluntandORathSpivak,Diffusioncontrolledgrowth,ProceedingsoftheRoyalSocietyofLondonA,
423123132(1989).
7. MBlunt,Hydrodynamicforcedistributiononafractalcluster,PhysicalReviewA,3958015806(1989).
8. MBlunt,WBarfordandRBall,PolymerAdsorptionandElectronBindingonRoughandFractalSurfaces,
Macromolecules,2214581466(1989).
9. RBallandMBlunt,Screeninginmultifractalgrowth,PhysicalReviewA,39(7)35913596(1989).
10. MBlunt,Geometryofmultifractalsystems,PhysicalReviewA39(5)27802782.(1989).
11. RBallandMBlunt,Dynamicsofscreeninginmultifractalgrowth,PhysicalReviewA,41(2)582589(1990).
12. MBluntandPKing,Macroscopicparametersfromsimulationsofporescaleflow,PhysicalReviewA,42(8)4780
4787(1990).
13. FJFayers,MJBluntandMAChristie,Accuratecalibrationofempiricalviscousfingeringmodels,RevuedeLInstitut
FrancaisduPetrole,46(3)311324(1991).
14. MBluntandPKing,RelativePermeabilitiesfromTwoandThreeDimensionalPoreScaleNetworkModelling,
TransportinPorousMedia,6407433(1991).
15. MJBlunt,PRKingandJAGoshawk,SimulationsofViscousFingeringinaRandomNetwork,inTheMathematics
ofOilRecovery,Ed.PRKing,TheInstituteofMathematicsanditsApplicationsConferenceSeries,no31,ClarendonPres,
Oxford(1992).
16. SBryantandMBlunt,Predictionofrelativepermeabilityinsimpleporousmedia,PhysicalReviewA,4620042011
(1992).
17. MBluntandBRubin,ImplicitFluxLimitingSchemesforPetroleumReservoirSimulation,JournalofComputational
Physics,102(1)194210(1992).
18. MBlunt,MJKingandHScher,Simulationandtheoryoftwophaseflowinporousmedia,PhysicalReviewA,46(12)
76807699(1992).
19. FJFayers,MJBluntandMAChristie,ComparisonsofEmpiricalViscousFingeringModelsandTheirCalibrationfor
HeterogeneousProblems,SPEReservoirEngineering,7(May)195203(1992).
20. MBlunt,FJFayersandFMOrr,CarbonDioxideinEnhancedOilRecovery,EnergyConversionManagement,34
11971204(1993).
21. MBluntandMChristie,HowtoPredictViscousFingeringinThreeComponentFlow,TransportinPorousMedia,12
207236(1993).
22. MJBlunt,JWBarker,BRubin,MMansfield,IDCulverwellandMAChristie,PredictiveTheoryforViscousFingering
inCompositionalDisplacement,SPEReservoirEngineering,9(1)7380(1994).
23. MBluntandMChristie,TheoryofViscousFingeringinTwoPhase,ThreeComponentFlow,SPEAdvanced
TechnologySeries,2(2),5260(1994).
24. MJBlunt,DZhouandDHFenwick,ThreePhaseFlowandGravityDrainageinPorousMedia,TransportinPorous
Media2077103(1995).
25. MRThiele,MJBluntandFMOrr,Modelingflowinheterogeneousmediausingstreamtubes:Imiscibleand
immiscibledisplacements,InSitu,19(3),299339(1995).
26. MRThiele,MJBluntandFMOrr,Modelingflowinheterogeneousmediausingstreamtubes:IIcompositional
displacements,InSitu,19(4),367391(1995).
27. MJBluntandHScher,PoreLevelModelingofWetting,PhysicalReviewE,5263876403(1995).
28. MRThiele,RPBatycky,MJBluntandFMOrr,"SimulatingFlowinHeterogeneousSystemsUsingStreamtubesand
Streamlines"SPEReservoirEngineering,11512February(1996).
29. MJBlunt,KLiuandMRThiele,"AGeneralizedStreamlineMethodtoPredictReservoirFlow"PetroleumGeoscience,
2259269(1996).
30. MRThiele,SERaoandMJBlunt,"QuantifyingUncertaintyinReservoirPerformanceUsingStreamtubes"
MathematicalGeology,28843856(1996).
MJBluntExpertReport
160
31. DZhouandMJBlunt,"EffectofSpreadingCoefficientontheDistributionofLightNonAqueousPhaseLiquidinthe
Subsurface"JContaminantHydrology,25119(1997).
32. MJBlunt,"EffectsofHeterogeneityandWettingonRelativePermeabilityusingPoreLevelModeling"SPEJournal,2
7087March(1997).
33. AAKeller,MJBluntandPVRoberts,"MicromodelObservationoftheRoleofOilLayersinThreePhaseFlow"
TransportinPorousMedia,26277297(1997).
34. DZhou,MJBluntandFMOrr,"HydrocarbonDrainageAlongCornersofNoncircularCapillaries"JColloidInterface
Sci,1871121(1997).
35. DRonen,HScherandMJBlunt,"OntheStructureandTransportPropertiesintheCapillaryFringeofPhreatic
Aquifers"TransportinPorousMedia,28159180(1997).
36. RPBatycky,MJBluntandMRThiele,"A3DFieldScaleStreamlineBasedReservoirSimulator"SPEReservoir
Engineering,12246254November(1997).
37. MJBlunt,"PoreLevelModelingoftheEffectsofWettability"SPEJournal,2494510December(1997).
38. DHFenwickandMJBlunt,"Threedimensional,ThreePhaseNetworkModelingofDrainageandImbibition"
AdvancesinWaterResources,21(2)121143(1998).
39. DHFenwickandMJBlunt,"NetworkModelingofThreePhaseFlowinPorousMedia"SPEJournal,38697March
(1998).
40. MJBlunt,"PhysicallyBasedNetworkModelingofMultiphaseFlowinIntermediateWetMedia"JournalofPetroleum
ScienceandEngineering,20117125June(1998).
41. DZhouandMJBlunt,"WettabilityEffectsinThreePhaseGravityDrainage"JournalofPetroleumScienceand
Engineering,20203211June(1998).
42. AAKeller,PVRoberts,andMJBlunt,"EffectofFractureApertureVariationsontheDispersionofContaminants"
WaterResourcesResearch,35(1)5563(1999).
43. TFirincioglu,MJBluntandDZhou,"ThreePhaseFlowandWettabilityEffectsinTriangularCapillaries"Colloidsand
SurfacesA,155259276(1999).
44. CESchaefer,PVRobertsandMJBlunt,"MeasurementandPredictionofEffectiveDiffusivitiesthroughSpreading
andNonspreadingOilsinUnsaturatedPorousMedia"EnvironmentalScienceandTechnology,33(17)28792884(1999).
45. MJCraneandMJBlunt,"StreamlinebasedSimulationofSoluteTransport"WaterResourcesResearch,35(10)3061
3078(1999).
46. YGautier,MJBluntandMAChristie,"Nestedgriddingandstreamlinebasedsimulationforfastreservoir
performanceprediction"ComputationalGeosciences3295320(1999).
47. AAKeller,MJBluntandPVRoberts,"BehaviorofNonaqueousPhaseLiquidsinFracturedPorousMediaunderTwo
PhaseFlowConditions"TransportinPorousMedia,38(1/2)189203,(2000).
48. LADillardandMJBlunt,"Developmentofaporenetworksimulationmodeltostudynonaqueousphaseliquid
dissolution"WaterResourcesResearch36(2)439454February(2000).
49. DADiCarlo,ASahniandMJBlunt,"ThreePhaseRelativepermeabilityofWaterWet,OilWetandMixedWet
Sandpacks"SPEJournal58291March(2000).
50. DZhou,LADillardandMJBlunt,"APhysicallyBasedModelofDissolutionofNonaqueousPhaseLiquidsinthe
SaturatedZone"TransportinPorousMedia39(2)227255May(2000).
51. DADiCarlo,ASahniandMJBlunt,TheEffectofWettabilityonThreePhaseRelativePermeability,Transportin
PorousMedia,39347366(2000).
52. CESchaefer,DADiCarloandMJBlunt,Determinationofwateroilinterfacialareaduring3phasegravitydrainage
inporousmedia,JournalofColloidandInterfaceScience,221308312(2000).
53. MHHuiandMJBlunt,EffectsofWettabilityonThreePhaseFlowinPorousMedia,J.Phys.Chem.B,1043833
3845(2000).
54. CESchaefer,DADiCarlo,andMJBlunt,ExperimentalMeasurementofAirWaterInterfacialAreaDuringGravity
DrainageandSecondaryImbibitioninPorousMedia,WaterResourcesResearch,36885890,(2000).
55. RGHughesandMJBlunt,PoreScaleModelingofRateEffectsinImbibition,TransportinPorousMedia,40(3),295
322(2000).
56. DRonen,HScherandMBlunt,Fieldobservationsofacapillaryfringebeforeandafterarainyseason,Journalof
ContaminantHydrology,44103118(2000).
57. DADiCarloandMJBlunt,Determinationoffingershapeusingthedynamiccapillarypressure,WaterResources
Research,3627812785(2000).
58. MJBlunt,AnEmpiricalModelforThreePhaseRelativePermeabilitySPEJournal5435445December(2000).
59. RGHughesandMJBlunt,Networkmodelingofmultiphaseflowinfractures,AdvancesinWaterResources24409
421(2001).
60. LADillard,HIEssaidandMJBlunt,Afunctionalrelationshipforfieldscalenonaqueousphaseliquiddissolution
developedusingaporenetworkmodel,JournalofContaminantHydrology48189199(2001).
MJBluntExpertReport
161
61. SPBertels,DADiCarloandMJBlunt,Measurementofaperturedistribution,capillarypressure,relative
permeability,andinsitusaturationinarockfractureusingcomputedtomographyscanning,WaterResourcesResearch,37(3)
649662(2001).
62. MJBlunt,Flowinporousmediaporenetworkmodelsandmultiphaseflow,CurrentOpinioninColloidand
InterfaceScience,6(3)197207(2001).
63. MJBlunt,Constraintsoncontactanglesformultiplephasesinthermodynamicequilibrium,JournalofColloidand
InterfaceScience,239281282(2001).
64. HSAlHadhramiandMJBlunt,"ThermallyInducedWettabilityAlterationtoImproveOilRecoveryinFractured
Reservoirs"SPEReservoirEngineeringandEvaluation,4179186June(2001).
65. RGHughesandMJBlunt,PoreScaleModelingofMultiphaseFlowinFracturesandMatrix/FractureTransferSPE
Journal,6(2)126136June(2001).
66. MAChristieandMJBlunt,TenthSPEComparativeSolutionProject:AComparisonofUpscalingTechniques,SPE
ReservoirEngineeringandEvaluation,4(4)308317August(2001).
67. MDJacksonandMJBlunt,EllipticRegionsandStableSolutionsforThreePhaseflowinPorousMediaTransportin
PorousMedia,48(3)249269,September(2002).
68. MJBlunt,MDJackson,MPiriandPHValvatne,Detailedphysics,predictivecapabilitiesandmacroscopic
consequencesforporenetworkmodelsofmultiphaseflow,
AdvancesinWaterResources,2510691089,(2002).
69. MPrvost,MGEdwardsandMJBlunt,StreamlineTracingonCurvilinearStructuredandUnstructuredGrids,SPE
Journal7(2)139148,June(2002).
70. JJHastings,AHMuggeridgeandMJBlunt,ANewStreamlineMethodforEvaluatingUncertaintyinSmallScale,
TwoPhaseFlowProperties,SPEJournal8(1)3240,March(2003).
71. BBijeljic,AHMuggeridgeandMJBlunt,MulticomponentMassTransferacrossWaterFilmsDuringHydrocarbonGas
Injection,Chem.Eng.Sci.58(11)23772388,June(2003).
72. GDiDonato,EIObiandMJBlunt,Anomaloustransportinheterogeneousmediademonstratedbystreamlinebased
simulation,GeophysicsResearchLetters30(12)16081612,June(2003).
73. BAgarwalandMJBlunt,StreamlineBasedMethodwithFullPhysicsForwardSimulationforHistoryMatching
PerformanceDataofaNorthSeaField,SPEJournal8(2)171180,June(2003).
74. XLopez,PHValvatne,andMJBlunt,PredictivenetworkmodelingofsinglephasenonNewtonianflowinporous
media,JournalofColloidandInterfaceScience,264(1)256265,August(2003).
75. MDJackson,PHValvatneandMJBlunt,Predictionofwettabilityvariationanditsimpactonflowusingporeto
reservoirscalesimulations,JournalofPetroleumScienceandEngineering,39231246,(2003).
76. PAudiganeandMJBlunt,DualMeshMethodforUpscalinginWaterfloodSimulation,TransportinPorousMedia,
557189,(2004).
77. GDiDonatoandMJBlunt,Streamlinebaseddualporositysimulationofreactivetransportandflowinfractured
reservoirs,WaterResourcesResearch,40,W04203,doi:10.1029/2003WR002772(2004).
78. WHuang,GDiDonatoandMJBlunt,Comparisonofstreamlinebasedandgridbaseddualporositysimulation,
JournalofPetroleumScienceandEngineering,43,129137(2004).
79. PHValvatneandMJBlunt,Predictiveporescalemodelingoftwophaseflowinmixedwetmedia,WaterResources
Research,40,W07406,doi:10.1029/2003WR002627(2004).
80. EOObiandMJBlunt,Streamlinebasedsimulationofadvectivedispersivesolutetransport,AdvancesinWater
Resources,27913924,(2004).
81. BBijeljic,AHMuggeridgeandMJBlunt,Porescalemodelingoflongitudinaldispersion,WaterResourcesResearch,
40,W11501,doi:10.1029/2004WR003567(2004).
82. MPiriandMJBlunt,Threephasethresholdcapillarypressuresinnoncircularcapillarytubeswithdifferent
wettabilitiesincludingcontactanglehysteresis,PhysicalReviewE70,061603(2004).
83. HOkabeandMJBlunt,Predictionofpermeabilityforporousmediareconstructedusingmultiplepointstatistics,
PhysicalReviewE70,066135(2004).
84. BAgarwalandMJBlunt,AStreamlineBasedMethodforAssistedHistoryMatchingAppliedtoanArabianGulf
Field,SPEJournal9,437449,December(2004).
85. MSAlGharbiandMJBlunt,Dynamicnetworkmodelingoftwophasedrainageinporousmedia,PhysicalReviewE
71,016308(2005).
86. HOkabeandMJBlunt,Porespacereconstructionusingmultiplepointstatistics,JournalofPetroleumScienceand
Engineering46121137(2005).
87. MPiriandMJBlunt,Threedimensionalmixedwetrandomporescalenetworkmodelingoftwoandthreephase
flowinporousmedia.I.Modeldescription,PhysicalReviewE71,026301(2005).
88. MPiriandMJBlunt,Threedimensionalmixedwetrandomporescalenetworkmodelingoftwoandthreephase
flowinporousmedia.II.Results,PhysicalReviewE71,026302(2005).
MJBluntExpertReport
162
89. PHValvatne,MPiri,XLopezandMJBlunt,PredictivePoreScaleModelingofSingleandMultiphaseFlow,
TransportinPorousMedia58,2341,doi:10.1007/s1124200454682(2005).
90. ZTavassoli,RWZimmermanandMJBlunt,AnalyticAnalysisforOilRecoveryDuringCounterCurrentImbibitionin
StronglyWaterWetSystems,TransportinPorousMedia58,173189,doi:10.1007/s1124200454744(2005).
91. MDJackson,PHValvatneandMJBlunt,PredictionofWettabilityVariationWithinanOil/WaterTransitionZone
andItsImpactonProduction,SPEJournal10(2),184195,June(2005).
92. ZTavassoli,RWZimmermanandMJBlunt,Analysisofcountercurrentimbibitionwithgravityinweaklywaterwet
Systems,JournalofPetroleumScienceandEngineering48,94104(2005).
93. HBehbahaniandMJBlunt,AnalysisofImbibitioninMixedWetRocksUsingPoreScaleModeling,SPEJournal,
10(4)466474,December(2005).
94. HBehbahani,GDiDonatoandMJBlunt,Simulationofcountercurrentimbibitioninwaterwetfractured
reservoirs,JournalofPetroleumScienceandEngineering50,2139(2006).
95. BBijeljicandMJBlunt,Porescalemodelingandcontinuoustimerandomwalkanalysisofdispersioninporous
media,WaterResourcesResearch42,W01202,doi:10.1029/2005WR004578(2006).
96. EIObiandMJBlunt,StreamlinebasedsimulationofcarbondioxidestorageinaNorthSeaaquifer,Water
ResourcesResearch42,W03414,doi:10.1029/2004WR003347(2006).
97. MERhodesandMJBlunt,Anexactparticletrackingalgorithmforadvectivedispersivetransportinnetworkswith
completemixingatnodes,WaterResourcesResearch42,W04501,doi:10.1029/2005WR004504(2006).
98. RJuanesandMJBlunt,AnalyticalSolutionstoMultiphaseFirstContactMiscibleModelswithViscousFingering,
TransportinPorousMedia64(3),339373,doi:10.1007/s112420055049z(2006).
99. GDiDonato,ZTavassoliandMJBlunt,Analyticalandnumericalanalysisofoilrecoverybygravitydrainage,Journal
ofPetroleumScienceandEngineering54,5569(2006).
100. RJuanes,EJSpiteri,FMOrr,Jr.andMJBlunt,ImpactofrelativepermeabilityhysteresisongeologicalCO2storage,
WaterResourcesResearch,42,W12418,doi:10.1029/2005WR004806(2006).
101. VSSuicmez,MPiriandMJBlunt,PorescaleSimulationofWaterAlternateGasInjection,TransportinPorous
Media,66,259286doi:10.1007/s1124200600179(2007).
102. MERhodesandMJBlunt,Advectivetransportinpercolationclusters,PhysicalReviewE,75,011124,
doi:10.1103/PhysRevE.75.011124(2007).
103. GDiDonato,HLu,ZTavassoliandMJBlunt,Multiratetransferdualporositymodellingofgravitydrainageand
imbibition,SPEJournal12,7788(2007).
104. ASAlKharusiandMJBlunt,Networkextractionfromsandstoneandcarbonateporespaceimages,Journalof
PetroleumScienceandEngineering56,219231(2007).
105. BBijeljicandMJBlunt,Porescalemodelingoftransversedispersioninporousmedia,WaterResourcesResearch,
43,W12S11,doi:10.1029/2006WR005700(2007).
106. HOkabeandMJBluntPorespacereconstructionofvuggycarbonatesusingmicrotomographyandmultiplepoint
statistics,WaterResourcesResearch,43,W12S02,doi:10.1029/2006WR005680(2007).
107. MIJvanDijke,MPiri,JOHelland,KSSorbie,MJBlunt,andSMSkjveland,Criteriaforthreefluidconfigurations
includinglayersinaporewithnonuniformwettability,WaterResourcesResearch,43,W12S05,doi:10.1029/2006WR005761
(2007).
108. RJuanesandMJBlunt,ImpactofViscousFingeringonthePredictionofOptimumWAGRatio,SPEJournal12(4),
486495(2007).
109. TSochiandMJBlunt,PorescalenetworkmodelingofEllisandHerschelBulkleyfluids,JournalofPetroleum
ScienceandEngineering60,105124(2008).
110. VSSuicmez,MPiriandMJBlunt,Effectsofwettabilityandporeleveldisplacementonhydrocarbontrapping,
AdvancesinWaterResources31,503512(2008).
111. EJSpiteri,RJuanes,MJBluntandFMOrr,Jr.,ANewModelofTrappingandRelativePermeabilityHysteresisforAll
WettabilityCharacteristics,SPEJournal13(3),277288(2008).
112. MERhodes,BBijeljicandMJBlunt,Poretofieldsimulationofsinglephasetransportusingcontinuoustime
randomwalks,AdvancesinWaterResources31(2)15271539(2008).
113. HLu,GDiDonatoandMJBlunt,GeneralTransferFunctionsforMultiphaseFlowinFracturedReservoirs,SPE
Journal13(3)289297(2008).
114. ASAlKharusiandMJBlunt,Multiphaseflowpredictionsfromcarbonateporespaceimagesusingextracted
networkmodels,WaterResourcesResearch44,W06S01,doi:10.1029/2006WR005695(2008).
115. OTalabi,SAlSayari,SIglauerandMJBlunt,PorescalesimulationofNMRresponse,JournalofPetroleumScienceand
Engineering67(34),168178(2009).
116. RQi,TCLaForceandMJBlunt,Designofcarbondioxidestorageinaquifers,InternationalJournalofGreenhouse
GasControl3195205(2009).
MJBluntExpertReport
163
117. MERhodes,BBijeljicandMJBlunt,Arigorousporetofieldscalesimulationmethodologyforsinglephaseflow
basedoncontinuoustimerandomwalks,SPEJournal14(1)8894(2009).
118. GFrenkel,RBlumenfled,PRKingandMJBlunt,TopologicalAnalysisofFoamsandTetrahedralStructures,
AdvancedEngineeringMaterials11(3)169176(2009).
119. CJXie,ZLGuan,MBluntandHZhou,Numericalsimulationofoilrecoveryaftercrosslinkedpolymerflooding,
JournalofCanadianPetroleumTechnology48(4)3741(2009).
120. XGarcia,LTAkanji,MJBlunt,SKMatthaiandJPLatham,Numericalstudyoftheeffectsofparticleshapeand
polydispersityonpermeability,PhysicalReviewE80,021304,doi:10.1103/PhysRevE.80.021304(2009).
121. NAlBulushi.PRKing,MJBluntandMKraaijveld,Developmentofartificialneuralnetworkmodelsforpredicting
watersaturationandfluiddistribution,JournalofPetroleumScienceandEngineering68197208(2009).
122. HDongandMJBlunt,Porenetworkextractionfrommicrocomputerizedtomographyimages,PhysicalReviewE
80,036307,doi:10.1103/PhysRevE.80.036307(2009).
123. SKAlMansoori,SIglauer,CHPentlandandMJBlunt,Threephasemeasurementsofoilandgastrappinginsand
packs,AdvancesinWaterResources32,15351542(2009).
124. MWBelayneh,SKMatthi,MJ.BluntandSFRogers,Comparisonofdeterministicwithstochasticfracturemodels
inwaterfloodingnumericalsimulations,AAPGBulletin93(11)16331648(2009).
125. RQi,TCLaForceandMJBlunt,Athreephasefourcomponentstreamlinebasedsimulatortostudycarbondioxide
storage,ComputationalGeoscience13493509,doi:10.1007/s1059600991399(2009).
126. VTBeraldo,MJBluntandDJSchiozer,CompressibleStreamlineBasedSimulationWithChangesinOil
Composition,SPEReservoirEngineeringandEvaluation12(6)963973(2009).
127. NMuller,RQi,EMackie,KPruessandMJBlunt,CO
2
injectionimpairmentduetohaliteprecipitation,Energy
Procedia135073514(2009).
128. SKAlMansoori,EItsekiri,SIglauer,CHPentland,BBijeljicandMJBlunt,Measurementsofnonwettingphase
trappingappliedtocarbondioxidestorage,InternationalJournalofGreenhouseGasControl4283288(2010).
129. NAlBulushi,PRKing,MJBluntandMKraaijveld,Generatingacapillarysaturationheightfunctiontopredict
hydrocarbonsaturationusingartificialneuralnetworks,PetroleumGeoscience167785(2010).
130. NAIdowuandMJBlunt,PoreScaleModellingofRateEffectsinWaterflooding,TransportinPorousMedia83151
169(2010).
131. XZhao,MJBluntandJYao,Porescalemodeling:Effectsofwettabilityonwaterfloodoilrecovery,Journalof
PetroleumScienceandEngineering71169178(2010).
132. C H Pentland, E Itsekiri, S AlMansoori, S Iglauer, B Bijeljic and M J Blunt, Measurement of NonWetting Phase
TrappinginSandpacks,SPEJournal15274281(2010).
133. O Talabi and M J Blunt, Porescale network simulation of NMR response in twophase flow, Journal of Petroleum
ScienceandEngineering7219(2010).
134. EUnsal,SKMatthiandMJBlunt,Simulationofmultiphaseflowinfracturedreservoirsusingafractureonlymodel
withtransferfunctions,ComputationalGeosciences14527538(2010).
135. PGittins,SIglauer,CHPentland,SAlMansoori,SAlSayari,BBijeljic,andMJBlunt,Nonwettingphaseresidual
saturationinsandpacks,JournalofPorousMedia13(7),591599(2010).
136. AJCavanagh,RSHaszeldineandMJBlunt,Openorclosed?Adiscussionofthemistakenassumptionsinthe
Economidespressureanalysisofcarbonsequestration,JournalofPetroleumScienceandEngineering74107110(2010).
137. SIglauer,SFavretto,GSpinelli,GSchenaandMJBlunt,Xraytomographymeasurementsofpowerlawclustersize
distributionsforthenonwettingphaseinsandstones,PhysicalReviewE82056315(2010).
138. AMAlSofiandMJBlunt,StreamlineBasedSimulationofNonNewtonianPolymerFlooding,SPEJournal15(4)901
911(2010).
139. CHPentland,RElMaghraby,SIglauerandMJBlunt,Measurementsofthecapillarytrappingofsupercriticalcarbon
dioxideinBereasandstone,GeophysicsResearchLetters,38,L06401,doi:10.1029/2011GL046683(2011).
140. HMNick,APaluszny,MJBluntandSKMatthai,Roleofgeomechanicallygrownfracturesondispersivetransportin
heterogeneousgeologicalformations,PhysicalReviewE84056301(2011)
141. BBijeljic,PMostaghimiandMJBlunt,ThesignatureofNonFickianSoluteTransportinComplexHeterogeneous
PorousMedia,PhysicalReviewLetters107204502(2011).
142. SIglauer,APalusznyCHPentlandandMJBlunt,ResidualCO
2
imagedwithXraymicrotomography,Geophysical
ResearchLetters38L21403(2011).
143. SIglauer,WWlling,CHPentland,SKAlMansooriandMJBlunt,CapillaryTrappingCapacityofRocksand
Sandpacks,SPEJournal16(4)778783(2011).
144. RMElMaghraby,CHPentland,S.IglauerandMJBlunt,Afastmethodtoequilibratecarbondioxidewithbrineat
highpressureandelevatedtemperatureincludingsolubilitymeasurements,J.ofSupercriticalFluids625559(2012).
145. AMAlSofiandMJBlunt,Asegregatedflowschemetocontrolnumericaldispersionformulticomponentflow
simulations,ComputationalGeoscience,doi:10.1007/s1059601292782(2012).
MJBluntExpertReport
164
146. SIglauer,MAFern,PShearingandMJBlunt,Comparisonofresidualoilclustersizedistribution,morphologyand
saturationinoilwetandwaterwetsandstone,JournalofColloidandInterfaceScience375187192(2012).
147. NIAlBulushi,PRKing,MJBluntandMKraaijveld,Artificialneuralnetworksanditsapplicationinthepetroleum
industry,NeuralComputingandApplications21(3)409421(2012).
148. AQRaeini,MJBluntandBBijeljic,Modellingtwophaseflowinporousmediaattheporescaleusingthevolumeof
fluidmethod.JournalofComputationalPhysics231(17)56535668(2012).
149. YTaninoandMJBlunt,Capillarytrappinginsandstonesandcarbonates:Dependenceonporestructure,Water
ResourcesResearch48W08525,doi:10.1029/2011WR011712(2012).
150. JDPaulandMJBlunt,Wastewaterfiltrationandreuse:AnalternativewatersourceforLondon.SciTotal
Environ43717318415(2012).
151. PMostaghimi,BBijeljicandMJBlunt,SimulationofFlowandDispersiononPoreSpaceImages,SPEJournal
1711311141(2012).
152. PMostaghimi,MJBluntandBBijeljic,ComputationsofAbsolutePermeabilityonMicroCTImages,
MathematicalGeosciences,doi:10.1007/s1100401294314(2012).
153. OGharbiandMJBlunt,Theimpactofwettabilityandconnectivityonrelativepermeabilityincarbonates:A
porenetworkmodelinganalysis,WaterResourcesResearch48W12513,doi:10.1029/2012WR011877(2012).
154. RMElMaghrabyandMJBlunt,ResidualCO
2
TrappinginIndianaLimestone,EnvironmentalScienceand
Technology47227233,doi:10.1021/es304166u(2013).
155. JPLatham,JXiang,MBelayneh,HMNick,CFTsangandMJBlunt,Modellingstressdependentpermeability
infracturedrockincludingeffectsofpropagatingandbendingfractures,InternationalJournalofRockMechanics&
MiningSciences57100112(2013).
156. SIglauer,APalusznyandMJBlunt,Simultaneousoilrecoveryandresidualgasstorage:Aporelevelanalysis
usinginsituXraymicrotomography,Fuel103905914(2013).
157. BBijeljic,ARaeini,PMostaghimiandMJBlunt,PredictionsofnonFickiansolutetransportindifferentclassesof
porousmediausingdirectsimulationonporescaleimages,PhysicalReviewE87,013011(2013).
158. MJBlunt,BBijeljic,HDong,OGharbi,SIglauer,PMostaghimi,APalusznyandCPentland,Porescaleimagingand
Modelling.AdvancesinWaterResources,51197216(2013).
M.2 Conferenceproceedingsandbookchapters
159. MBluntandBRubin,ImplicitFluxLimitingSchemesforPetroleumReservoirSimulation,SPE22192,proceedingsof
the2
nd
EuropeanConferenceontheMathematicsofOilRecovery,Arles,France,September1990,EditionsTechnip,Paris
(1990).
160. BRubinandMJBlunt,HigherOrderImplicitFluxLimitingschemesforBlackOilSimulation,SPE21222,proceedings
ofthe11
th
SPESymposiumonReservoirSimulation,Anaheim,California,February1720(1991).
161. BRubin,J.W.Barker,MJBlunt,MAChristie,IDCulverwellandMMansfield,CompositionalReservoirSimulation
WithaPredictiveModelforViscousFingering,SPE25234,proceedingsofthe12
th
SPESymposiumonReservoirSimulation,
NewOrleans,February28March3(1993).
162. MJKing,MJBlunt,MMansfieldandMAChristie,RapidEvaluationoftheImpactofHeterogeneityonMiscibleGas
Injection,SPE26079,proceedingsoftheSPEWesternRegionalMeeting,Anchorage,Alaska,May(1993).AlsoinNew
DevelopmentsinImprovedOilRecovery,ed.HJdeHaan,GeologicalSocietySpecialPublicationNo.84.
163. MJBlunt,DHFenwickandDZhou,WhatDeterminesResidualOilSaturationinThreePhaseFlow?SPE27816,
proceedingsoftheSPE?DOENinthSymposiumonImprovedOilRecovery,Tulsa,1720April(1994).
164. MRThiele,MJBluntandFMOrr,PredictingMulticomponent,MultiphaseFlowinHeterogeneousSystemsusing
Streamtubes,proceedingsofthe4
th
EuropeanConferenceontheMathematicsofOilRecovery,Roros,Norway,June710
(1994).
165. DHFenwickandMJBlunt,PorelevelModelingofThreePhaseFlowinPorousMedia,proceedingsofthe8
th
EuropeanSymposiumonImprovedOilRecovery,Vienna,May1517(1995).
166. RPBatycky,MRThieleandMJBlunt,AStreamlineSimulatortoModelFieldScaleThreeDimensionalFlow
proceedingsofthe5thEuropeanConferenceontheMathematicsofOilRecovery,Leoben,Austria,September36(1996).
167. DHFenwickandMJBlunt,CalculatingThreePhaseRelativePermeabilitiesUsingNetworkModelingproceedingsof
the5thEuropeanConferenceontheMathematicsofOilRecovery,Leoben,Austria,September36(1996).
168. DZhou,MJBluntandFMOrr,EffectofWettabilityandSpreadingCoefficientonThreePhaseDistributions
proceedingsofthe17thIEACollaborativeProjectonEnhancedOilRecoveryWorkshopandSymposium,Sydney,Australia,
September30October2(1996).
169. RPBatycky,MJBluntandMRThiele,A3DMultipurposeStreamlineSimulatorwithGravityandChangingWell
Conditionsproceedingsofthe17thIEACollaborativeProjectonEnhancedOilRecoveryWorkshopandSymposium,Sydney,
Australia,September30October2(1996).
MJBluntExpertReport
165
170. ASahni,RGuzmanandMJBlunt,TheoreticalAnalysisofThreePhaseFlowExperimentsinPorousMediaSPE
36664,proceedingsoftheSPEAnnualTechnicalConferenceandExhibition,Denver,CO,October69(1996).
171. MRThiele,RPBatyckyandMJBlunt,AStreamlineBased3DFieldScaleCompositionalReservoirSimulatorSPE
38889,proceedingsoftheSPEAnnualTechnicalConferenceandExhibition,SanAntonio,TX,October58(1997).
172. BMPeters,DZhouandMJBlunt,ExperimentalInvestigationofScalingFactorsthatDescribeMiscibleFloodsin
LayeredSystemsSPE39624,proceedingsoftheSPE/DOEImprovedOilRecoverySymposium,Tulsa,OK,April(1998).
173. ASahni,JBurgerandMJBlunt,MeasurementofThreePhaseRelativePermeabilityduringGravityDrainageUsingCT
ScanningSPE39655,proceedingsoftheSPE/DOEImprovedOilRecoverySymposium,Tulsa,OK,April(1998).
174. DADiCarlo,ASahniandMJBlunt,ThreePhaseRelativePermeabilityforWaterWetandOilWetSandpacks
proceedingsoftheconferenceontheEvaluationofReservoirWettabilityanditsEffectonOilRecovery,Trondheim,Norway,
June(1998).
175. CESchaefer,DSKosson,MJBluntandPVRoberts,DiffusiveReleaseofAromaticHydrocarbonsinthePresenceof
ImmobileNAPLproceedingsofEnviroMEET'98,UniversityofCalifornia,Irvine,July2123(1998).
176. AAKeller,MJBluntandPVRoberts,BehaviorofDenseNonAqueousPhaseLiquidinFracturedMediaunderTwo
PhaseFlowConditionsproceedingsofEnviroMEET'98,UniversityofCalifornia,Irvine,July2123(1998).
177. RGHughesandMJBlunt,NetworkModelingofMultiphaseFlowinFracturedPorousMediaproceedingsofthe6th
EuropeanConferenceontheMathematicsofOilRecovery,Peebles,Scotland,September811(1998).
178. TKawanishi,YHayashi,PVRobertsandMJBlunt,FluidFluidInterfacialAreaduringTwoandThreephaseFluid
DisplacementinPorousMedia:ANetworkModelStudyproceedingsoftheInternationalConferenceandSpecialSeminaron
GroundwaterQuality:RemediationandProtection,Tubingen,Germany,SeptemberIAHS2508996(1998).
179. YGautier,MJBluntandMAChristie,NestedGriddingandStreamlineBasedSimulationforFastReservoir
PerformancePredictionSPE51931,proceedingsoftheSPEReservoirSimulationSymposium,Houston,February(1999).
180. MHuiandMJBlunt,PoreScaleModelingofThreePhaseFlowandtheEffectsofWettabilitySPE59309,
proceedingsoftheSPE/DOEImprovedOilRecoverySymposium,Tulsa,OK,April(2000).
181. MDJacksonandMJBlunt,DynamicUpscalingfromthePoretotheReservoirScale,proceedingsofthe21
st
AnnualInternationalEnergyAgencyWorkshop,Edinburgh,UK,1922September(2000).
182. BRBijeljic,AHMuggeridgeandMJBlunt,EffectofCompositiononWaterblockingforMulticomponentGasfloods,
SPE77697,proceedingsoftheSPEAnnualMeeting,SanAntonio,Texas,29September2October(2002).
183. MPiriandMJBlunt,Porescalemodelingofthreephaseflowinmixedwetsystems,SPE77726,proceedingsofthe
SPEAnnualMeeting,SanAntonio,Texas,29September2October(2002).
184. PAudiganeandMJBlunt,DualMeshMethodinUpscaling,SPE79681,proceedingsoftheSPEReservoirSimulation
Symposium,Houston,Texas,35February2003.
185. MAlGharbiandMJBlunt,A2Ddynamicporescalemodelformodelingprimarydrainage,proceedingsoftheESF
WorkshoponRecentAdvancesinMultiphaseFlowandTransportinPorousMedia,EuropeanScienceFoundation,Delft,2003.
186. HOkabeandMJBlunt,Multiplepointstatisticstogenerategeologicallyrealisticporespacerepresentations,
SCA2003A33,proceedingsoftheSocietyofCoreAnalystsAnnualMeeting,Pau,France,September(2003).
187. HOkabeandMJBlunt,MultiplePointStatisticstoGeneratePoreSpaceRepresentationsofSandstones,
proceedingsoftheInternationalWorkshoponXrayCTforGeomaterialsGeoX2003,Kumamoto,JapanandpublishedinX
RAYCTForGEOMATERIALSsoil,concrete,rocks,Balkema.p.229237(2004).
188. GDiDonato,WHuangandMJBlunt,StreamlineBasedDualPorositySimulationofFracturedReservoirs,SPE
84036,proceedingsoftheSPEAnnualMeeting,Denver,Colorado,58October(2003).
189. PHValvatneandMJBlunt,PredictivePoreScaleNetworkModeling,SPE84550,proceedingsoftheSPEAnnual
Meeting,Denver,Colorado,58October(2003).
190. MRThiele,RPBatycky,MIdingandMJBlunt,Extensionofstreamlinebaseddualporosityflowsimulationtorealistic
geology,proceedingsoftheninthEuropeanConferenceontheMathematicsofOilRecovery(ECMOR),Cannes,France,
September(2004).
191. PRNurafza,MJBluntandMRFassihi,Evaluationofwaterandgasinjectioninacarbonatereservoir,SCA2004A42,
proceedingsoftheSocietyofCoreAnalystsAnnualMeeting,AbuDhabi,UAE,September(2004).
192. HOkabeandMJBlunt,Predictingpermeabilitythrough3Dporespaceimagesreconstructedusingmultiplepoint
statistics,SCA2004A43,proceedingsoftheSocietyofCoreAnalystsAnnualMeeting,AbuDhabi,UAE,September(2004).
193. HOkabeandMJBlunt,Multiplepointstatisticstogenerateporespaceimages,proceedingsoftheGeostatistics
Congress,Banff,Canada,September(2004).
194. XLopezandMJBlunt,PredictingtheImpactofNonNewtonianRheologyonRelativePermeabilityUsingPoreScale
Modeling,SPE89981,proceedingsoftheSPEAnnualMeeting,Houston,Texas,2629September(2004).
195. JBronchalo,RVenkataramanandMJBlunt,AMultiscaleMethodologyforSimulatingMiscibleGasInjectionProjects
AppliedtotheRODFieldinAlgeria,SPE90247,proceedingsoftheSPEAnnualMeeting,Houston,Texas,2629September
(2004).
MJBluntExpertReport
166
196. MBarthelemy,DHFenwick,YGautierandMJBlunt,AFrameworkforHistoryMatchingUsingLocalOptimizationin
StreamlineDefinedRegions,SPE90137,proceedingsoftheSPEAnnualMeeting,Houston,Texas,2629September(2004).
197. M Calabrese, F Masserano and M J Blunt, Simulation of PhysicalChemical Processes during Carbon Dioxide
Sequestration in Geological Structures, SPE 95820, proceedings of the SPE Annual Meeting, Dallas, Texas, 9 12 October
(2005).
198. EJSpiteri,RJuanes,MJBluntandFMOrr,Jr.,RelativePermeabilityHysteresis:TrappingModelsandApplicationto
GeologicalCO2Sequestration,SPE96448,proceedingsoftheSPEAnnualMeeting,Dallas,Texas,912October(2005).
199. V S Suicmez, M Piri and M J Blunt, Pore Scale Modeling of ThreePhase WAG Injection: Prediction of Relative
Permeabilities and Trapping for Different Displacement Cycles, SPE 95594, proceedings of the SPE/DOE Symposium on
ImprovedOilRecoveryheldinTulsa,Oklahoma,2226April(2006).
200. ASAlKharusiandMJBlunt,PermeabilityPredictionandNetworkExtractionfromPoreSpaceImages,proceedings
ofCMWRXVI,Copenhagen,June(2006).
201. BBijeljicandMJBlunt,LongitudinalandTransverseDispersioninPorousMediabyPoreScaleModelingand
ContinuousTimeRandomWalk(CTRW)Theory,proceedingsofCMWRXVI,Copenhagen,June(2006).
202. VSSuicmez,MPiriandMJBlunt,Porescalesimulationofwateralternategasinjection,proceedingsofCMWRXVI,
Copenhagen,June(2006).
203. HOkabeandMJBlunt,Reconstructionofporespaceimagesusingmicrotomographyandmultiplepointstatistics,
proceedingsofCMWRXVI,Copenhagen,June(2006).
204. MIJvanDijke,MPiri,KSSorbie
andMJBlunt,Criterionforthreefluidconfigurationsincludinglayersinaporewith
nonuniformwettability,proceedingsofCMWRXVI,Copenhagen,June(2006).
205. MRotondi,GNicotra,AGodi,FMContento,MJBluntandMAChristie,HydrocarbonProductionForecastand
UncertaintyQuantification:AFieldApplication,SPE102135,proceedingsoftheSPEAnnualMeeting,SanAntonio,Texas,24
27September(2006).
206. BBijeljicandMJBlunt,APhysicallyBasedDescriptionofDispersioninPorousMedia,SPE102869,proceedingsof
theSPEAnnualMeeting,SanAntonio,Texas,2427September(2006).
207. HDong,MTouatiandMJBlunt,PoreNetworkModeling:AnalysisofPoreSizeDistributionofArabianCoreSamples,
SPE105156,proceedingsofthe15
th
SPEMiddleEastOil&GasShowandConference,Bahrain,1114March(2007).
208. FVerre,MJBlunt,AMorrison,TMcGarva,ApplicabilityofWaterShutOffTreatmentforHorizontalWellsinHeavy
OilReservoirs,SPE106908,proceedingsoftheSPEEuropec/EAGEAnnualConferenceandExhibitioninLondon,UK,1114
June(2007).
209. HLuandMJBlunt,GeneralFracture/MatrixTransferFunctionsforMixedWetSystems,SPE107007,proceedings
oftheSPEEuropec/EAGEAnnualConferenceandExhibitioninLondon,UK,1114June(2007).
210. VTBeraldo,MJBlunt,DJSchiozerandRQi,StreamlineSimulationwithanAPITrackingOption,SPE107498,
proceedingsoftheSPEEuropec/EAGEAnnualConferenceandExhibitioninLondon,UK,1114June(2007).
211. GFrenkel,RBlumenfeld,PRKingandMJBlunt,TopologicalAnalysisofFoamsandtetrahedralstructures,
proceedingsofMetFoam2007,Montreal,Canada,57September(2007).
212. RQi,VBeraldo,TLaForceandMJBlunt,DesignofCarbonDioxideStorageinaNorthSeaAquiferUsingStreamline
BasedSimulation,SPE109905,proceedingsoftheSPEAnnualMeeting,Anaheim,California,1114November(2007).
213. ASAlRabaani,MJBluntandAHMuggeridge,CalculationofaCriticalSteamInjectionRateforThermallyAssisted
GasOilGravityDrainage,SPE113351,proceedingsoftheSPEImprovedOilRecoverySymposium,Tulsa,Oklahoma,1923April
(2008).
214. OTalabi,SAlSayari,MJBlunt,HDongandXZhao,PredictivePoreScaleModeling:FromThreeDimensionalImages
toMultiphaseFlowSimulations,SPE115535,proceedingsoftheSPEAnnualMeeting,Denver,Colorado,USA,2124
September(2008).
215. RQi,TCLaForceandMJBlunt,DesignofCarbonDioxideStorageinOilfields,SPE115663,proceedingsoftheSPE
AnnualMeeting,Denver,Colorado,USA,2124September(2008).
216. NAIdowuandMJBlunt,Porescalemodelingofrateeffectsinwaterflooding,IPTC12292,proceedingsofthe
InternationalPetroleumTechnologyConference,KualaLumpur,Malaysia,35December(2008).
217. AMAlSofi,TCLaForceandMJBlunt,SweepImpairmentduetoPolymersShearThinning,SPE120321,
proceedingsofthe.SPEMiddleEastOilandGasShowandConference(2009).
218. WSifuentes,MJBluntandMAGiddins,ModelingCO
2
StorageinAquifers:AssessingtheKeyContributorsto
Uncertainty,SPE123582,proceedingsoftheSPEOffshoreEuropeOil&GasConference&Exhibition,Aberdeen,UK,811
September(2009).
219. SKAlMansoori,SIglauer,CHPentland,BBijeljic,andMJBlunt,MeasurementsofNonWettingPhaseTrapping
AppliedtoCarbonDioxideStorage,ProceedingsoftheGreenhouseGasControlsTechnologyConferenceWashingtonD.C.,
U.S.A(2009).
MJBluntExpertReport
167
220. SKAlMansoori,CHPentland,SIglauerandMJBlunt,ThreePhaseMeasurementsofNonWettingPhaseTrapping
inUnconsolidatedSandPacks,SPE123994,proceedingsoftheSPEAnnualMeeting,NewOrleans,Louisiana,USA,47October
(2009).
221. MJBlunt,RQiandTCLaForce,"Carbondioxide(CO2)injectiondesigntomaximiseundergroundreservoirstorage
andenhancedoilrecovery(EOR)"In:DevelopmentsandInnovationincarbondioxide(CO2)captureandstoragetechnology.
Volume2:Carbondioxide(CO2)storageandutilisation.EditedbyM.MercedesMarotoValer,WoodheadPublishingSeriesin
Energy.Pages169184(2010).
222. CSLamy,SIglauer,CHPentland,MJBluntandGMaitland,CapillaryTrappinginCarbonateRocks,SPE130720,
proceedingsofSPEEUROPEC/EAGEAnnualConferenceandExhibition,Paper,Barcelona,Spain(2010).
223. HOkabe,YTsuchiya,CHPentland,SIglauer,andMJBlunt,ResidualCO
2
saturationdistributionsinrocksamples
measuredbyXrayCT,proceedingsofthe3rdInternationalWorkshoponXRayCTforGeomaterials,NewOrleans,Louisiana,
USA.(2010).
224. AMAlSofiandMJBlunt,TheSelfflatteningNatureofTrailingShocksinAugmentedWaterfloodingSegregation
inflowReestablishSelfsharpness,A004,proceedingsofECMORXII,Oxford(2010).
225. CHPentland,YTanino,SIglauerandMJBlunt,CapillaryTrappinginWaterWetSandstones:Coreflooding
ExperimentsandPoreNetworkModeling,SPE133798,proceedingsoftheSPEAnnualTechnicalConferenceandExhibition,
Florence,Italy.(2010).
226. CHPentland,R.ElMaghraby,A.Georgiadis,S.Iglauer,M.J.Blunt,ImmiscibleDisplacementsandCapillaryTrapping
inCO2Storage,proceedingsoftheGreenhouseGasControlsTechnologyConference,Amsterdam,TheNetherlands,September
(2010).
227. CHPentland,SIglauer,RElMaghraby,YTsuchiya,HOkabeandMJBlunt,MeasurementofCarbonDioxideCapillary
TrappinginCoreAnalysis,SPE138476,proceedingsoftheSPEInternationalConferenceonCO2Capture,Storageand
Utilization,NewOrleans,Lousiana,USA,1012November(2010).
228 JKalunka,TCLaForceandMJBlunt,EffectsofCO2stroageinSalineAquifersonGroundwaterSupplies,SPE
139665,proceedingsoftheSPEInternationalConferenceonCO2Capture,StorageandUtilization,NewOrleans,Lousiana,USA,
1012November(2010).
229. MJBlunt,CarbonDioxideStorage,GranthamBriefingPaper,No4,ImperialCollegeLondon(2010).
230. MRThiele,MGerritsenandMBlunt(Eds.),StreamlineSimulation,SocietyofPetroleumEngineers,RichardsonTX,
ISBN:9781613990650(2011).
231. MJBlunt,Twohundredbarrelsleft:ananalysisofpopulationgrowth,oilreservesandcarbondioxideemissions
http://www.decc.gov.uk/en/content/cms/meeting_energy/int_energy/global_oil/cfe_crude_oil/cfe_crude_oil.aspx
232. AMAlSofiandMJBlunt,PolymerFloodingDesignandOptimizationUnderUncertainty,SPE145110,proceedings
oftheSPEEnhancedOilRecoveryConferenceheldinKualaLumpur,Malaysia,1921July(2011).
233. LESobers,MJBluntandTCLaForce,Designofsimultaneousenhancedoilrecoveryandcarbondioxidestorage
appliedtoaheavyoilfieldoffshoreTrinidad,SPE147241,proceedingsoftheSPEAnnualTechnicalConferenceandExhibition,
Denver,Colorado,USA.(2011).
234. RMElMaghraby.CHPentlandandMJBlunt,CorefloodmeasurementsofCO
2
trapping,SPE147373,proceedings
oftheSPEAnnualTechnicalConferenceandExhibition,Denver,Colorado,USA.(2011).
235. ASAbushaikha,MJBlunt,TCLaForceandORGosselin,ImprovedMobilityCalculationforFiniteElement
Simulation,SPE154480,proceedingsoftheEAGEAnnualConferenceandExhibitionincorporatingSPEEuropec,Copenhagen,
Denmark,47June(2012).
236. BBijeljic,MJBlunt,HDong,PMostaghimi,ARaeini,OGharbiandMAndrew,PorescaleImagingandModellingof
TransportandFlowinComplexPorousMedia,ProceedingsoftheEnergyandStorageMaterialsWorkshop,DiamondLight
Source(DLS)SynchrotronUsersMeeting,Didcot,UK,56September(2012).
237. OGharbi,BBijeljic,EBoekEandMJBlunt,ChangesinPoreStructureandConnectivityInducedbyCO
2
Injectionin
Carbonates:aCombinedPorescaleApproach,ProceedingsoftheInternationalConferenceonGreenhouseGasTechnologies
(GHGT)11,EnergyProcedia,Kyoto,Japan,1822November(2012).
238. AAlQattan,MJBlunt,OGharbi,ABadamchizadeh,JMAlKanderi,MAlJadi,HHDashti,VChimmalgi,DJBondand
FSkoreyko,EvaluationoftheEffectofAsphalteneDepositionintheReservoirfortheDevelopmentoftheMagwaMarrat
Reservoir,SPE163331,proceedingsoftheSPEKuwaitInternationalPetroleumConferenceandExhibition,KuwaitCity,Kuwait,
1012December(2012).
239. MDJackson,JLMAGomes,PMostaghimi,JRPercival,BSTollit,DPavlidis,CCPain,AHElSheikh,AHMuggeridge
andMJBlunt,ReservoirModelingforFlowSimulationUsingSurfaces,AdaptiveUnstructuredMeshesandControlVolume
FiniteElementMethods,SPE163633,proceedingsoftheSPEReservoirSimulationSymposium,TheWoodlands,TexasUSA,
1820February(2013).

MJBluntExpertReport
168
AppendixN. Factsanddataconsideredinformingmyopinion
N.1 USGovernmentandBPexpertreports
Letter Dateposted Name Filetype Descriptionofcontents
Z 20130322 Expertreport,Equationofstatefluid
characterizationandanalysisoftheMacondo
reservoirfluids,preparedbyAaronA.Zick
Pdf TheZickreport
KR 20130322 Expertreport,Ratepredictionfromthe
MacondoWell,preparedbyMohanKelkar
andRajagopalRaghavan
Pdf TheKelkar&Raghavanreport
PD 20130322 Expertreport,Estimateofthecumulative
volumeofoilreleasedfromtheMC252
Macondowell,preparedbyMehranPooladi
Darvish
Pdf ThePooladiDarvishreport
SKG 20130322 Expertreport,OilReleasefromtheMC252
MacondoWell,preparedbyStuartGriffiths
Pdf TheGriffithsreport
ACG 20130105 ExpertreportpreparedbyAlainGringarten Pdf TheGringartenreport
T 20130105 ExpertreportpreparedbyMartinTrusler Pdf/Excel TheTruslerreportandasscociated
pressuredata
W 20130105 ExpertreportpreparedbyCurtisWhitson Pdf/Excel TheWhitsonreportandassociated
fluidpropertytables
RWZ 20130105 ExpertreportpreparedbyRobert
Zimmerman
Pdf TheZimmermanreport
TV 20130105 ExpertreportpreparedbyDr.TorresVerdin Pdf TheTorresVerdinreport
N.2 Confidentialreports,depositiontranscriptsandothermaterial
Number Dateposted Name Filetype Descriptionofcontents
1 20120224 WeatherfordSummaryofEffective
PermeabilitytoOilMeasurements
BPHZN2179MDL05223139
Excel Weatherfordpermeability
measurementstooil
2 20111114 Griffiths_Revised_6_2011 Pdf RevisedGriffithsReport
3 20111114 Revised_Ratzel_9_2011 Pdf RevisedRatzelReport
4 20111111 BPHZN2179MDL05223139 Pdf BPinternalemailconfirmingwater
depthanddepthofcappingstackHC
connector
5 20111018 (6d16a1d913363e181102bbfc06f0b9a0)
BlackOilTablesfromEoSforAllTemps
11June2010;BPHZN2179MDL04578104
Excel BPfluidpropertytables
6 20111012 PostWellSubsurfaceTechnical
Memorandum
BPHZN2179MDL03290054
Word BPsseismicandloginterpretation
7 20111006 BPHZN2179MDL04549798 Excel Descriptionofcappingstackclosure
8 20110927 MacondoOOIP
(ff9a70a171043e7d458cae0765f5aa5c);BP
HZN2179MDL04440267
Excel SummaryofMacondooilinplace
9 20110927 TAMChapter1(BPHZN2179MDL02900640) Excel Technicalassurancememorandum:
drillingplanandreservoirdescription
predrill
10 20110927 TechnicalAssuranceMemorandumSections
2,3and4(BPHZN2179MDL02107723)
Excel Asabovesections24
11 20120306 10132010Predecisionaldraft,Hsieh,
ComputersimulationofReservoirDepletion
andOilFlow(IGS642000215)
Pdf DraftofHsiehreportwithall
reservoirparameterslisted
12 20120306 HsiehModelingPresentation(IGS629
003048)
Powerpoint Dr.Hsiehpresentationwith
sensitivities
13 20110920 DataDump_MC252_K_303.csv Excel Cappingstackpressure
MJBluntExpertReport
169
BPHZN2179MDL04440614andothers;BP
HZN2179MDL04440614;BPHZN
2179MDL04440804;BPHZN
2179MDL04440775;BPHZN
2179MDL04440733;BPHZN
2179MDL04440614;BPHZN
2179MDL04440613;BPHZN
2179MDL04440584;BPHZN
2179MDL04440466;BPHZN
2179MDL04440382;BPHZN
2179MDL04440263;BPHZN
2179MDL04440262;BPHZN
2179MDL04440249;BPHZN
2179MDL04440192;BPHZN
2179MDL04440168;BPHZN
2179MDL04440100
measurements
14 20110920 MacondoTechnicalNote(PreliminaryDraft)
WellIntegrityTest
(601555a84d8d77a9bf7fbd119604ca45);BP
HZN2179MDL04440249
Pdf Discussionofcappingstackpressure
measurementswithlocationof
transducer(Figure4)
15 20110901 BPHZN2179MDL04440691 Powerpoint BPsummaryofloganalysis
16 20110901 BPHZNBLY00120160 Excel BPsummaryofpermeability,
porosityandsaturation
17 20110901 BPHZN2179MDL04440238 Powerpoint BPseismicsummary
18 20110823 BPHZN2179MDL04440732;seealsoExhibits
9734,10423
Excel Intertekfluidpropertyreport
19 20110823 BPHZN2179MDL00063016 Excel CoreLabsfluidpropertyreport
8583,8584,8585
Excel CoreLabsfluidpropertyreportwith
coverpage
21 20110823 BPHZN2179MDL00470598 Excel Weatherfordsummaryofcores
collected
22 20110823 BPHZN2179MDL00470599 Excel Weatherfordcoresummary
23 20110823 BPHZN2179MDL02394182;seealsoExhibit
9070
Excel Weatherfordpermeability
measurements
9067
Excel Weatherfordporevolume
compressibilityrawdata
9066
Excel Weatherfordgrainsizeanalysis
9053
Excel Weatherfordsummaryofpore
volumecompressibility
27 20120306 FinalEmilsenReport Pdf Emilsenreportonthewellfailure
28 20110825 BPHZN2179MDL00477088 Excel BPdrillingreportandwell
description
29 20120511 BPHZN2179MDL00059145 Powerpoint BPShallowHazardspresentation
30 20120511 BPHZN2179MDL05173765 Powerpoint BPgrossrockvolumeassessment
31 20120511 BPHZN2179MDL06566208 Powerpoint BPPredrillreview
32 20110819 BPHZN2179MDL04440977 Excel Intertekmultistageseparatortests
andcompositions
33 20110819 BPHZN2179MDL04440978 Pdf Intertekmultistageseparatortest
results
10452,10459
Pdf SchlumbergerFluidAnalysison
MacondoSamples
35 20110802 BPHZN2179MDL01872218 Excel CoreLabsfluidpropertiesreport
36 20110523 June222010MacondoReview
(BPHZN2179MDL05181294)
Powerpoint Macondoreviewwithdatafrom
analoguefields
9068
Excel Weatherfordrockcomposition
analysis
MJBluntExpertReport
170
38 20110802 BPHZNBLY00000526 Pdf BPDeepwaterHorizonAccident

InvestigationReport,AppendixW
DynamicSimulations
39 20120703 UnconfinedLowReliefSlopeChannelSystem
BPHZN2179MDL06604338 52
Powerpoint BPpresentationofreservoirgeology
inchannelsystems
40 20120716 08MC252_PT_3K_2CORR2Calculated1(BP
HZN2179MDL07279441)and08
MC252_PT_3K_2CORR2Calculated2(BP
HZN2179MDL07279439)
Excel Correctedcappingstackpressure
measurementsusedinthisreport
41 20120716 Equation_Corr(BPHZN2179MDL07279440) Excel Equationusedtocomputepressure
fromtherawdata
42 20120911 30(b)(6)depositionofPaulAnthonyHsieh,
PhD,USGS,takenatthePanAmerican
Building,601PoydrasStreet,11thFloor,New
Orleans,Louisiana,70130,onthe11thdayof
September,2012.
Pdf TranscriptofDr.Hsiehs(USGS)
deposition:firstday
43 20120911 Continuationofthe30(b)(6)deposition
ofPaulAnthonyHsieh,Ph.D.,USGS,takenat
thePanAmericanBuilding,601Poydras
Street,11thFloor,NewOrleans,Louisiana,
70130,onthe12thdayofSeptember,2012.
Pdf TranscriptofDr.Hsiehs(USGS)
deposition:secondday
44 Hsiehpressureanalysis;EXHIBIT8617,BATES
IGS770000026
Excel DatausedinDr.Hsiehsanalysis.
ProvidedbyemailSeptember2012
45 20121004 30(b)(6)depositionofBryanDavidRitchie,
BP,PhD,takenatthePanAmericanBuilding,
601PoydrasStreet,11thFloor,NewOrleans,
Louisiana,70130,onthe4thdayofOctober,
2012.
Pdf TranscriptofDr.Ritchies(BP)
deposition:firstday
ofBryanDavidRitchie,BP,takenatthePan
AmericanBuilding,601PoydrasStreet,11th
Floor,NewOrleans,Louisiana,70130,onthe
5thdayofOctober,2012.
Pdf TranscriptofDr.Ritchies(BP)
47 20120918 30(b)(6)depositionofGrahamPinkyVinson,
BP,PhD,takenatthePanAmericanBuilding,
Louisiana,70130,onthe18thdayof
September,2012.
Pdf TranscriptofPinkyVinsons(BP)
deposition:firstday
ofGrahamPinkyVinson,BP,takenatthe
PanAmericanBuilding,601PoydrasStreet,
11thFloor,NewOrleans,Louisiana,70130,
onthe19thdayofSeptember,2012.
Pdf TranscriptofPinkyVinsons(BP)
49 20120920 30(b)(6)depositionofJaimeLoos,takenat
thePanAmericanBuilding,601Poydras
Street,11thFloor,NewOrleans,Louisiana,
70130,onthe20
th
dayofSeptember,2012.
Pdf TranscriptofJaimeLoos
(Weatherford)deposition
50 20101023 PresCommBPCommentstoOct6Working
PaperNo3AmountandFateofOil
Pdf BPsresponsetoflowrateestimates
51 20121112 WFTMDL00039271;WFTMDL00039272;
WFTMDL00039273;WFTMDL00039274;
JPEG Weatherfordscorephotographs
MJBluntExpertReport
171
MJBluntExpertReport
172
WFTMDL00039841
52 20121112 WFTMDL00039232;WFTMDL00039234;
WFTMDL00054283;seealsoExhibits9063,
9064,9065
JPEG WeatherfordsCTscansofcore
samples.
53 20121130 30(b)(6)DepositionofEdmondShtepani,
Ph.D.,P.Eng,Intertek,takenatthePan
30thdayofNovember,2012.
Pdf TranscriptofEdmondShtepanis
(Intertek)deposition
54 2013115 30(b)(6)DepositionofRobertC.Merrill,Jr.,
takenatPanAmericanBuilding,
Louisiana,70130,onthe15thdayofJanuary,
2013.
Pdf TranscriptofDr.Merrills(BP)
deposition:firstday
55 2013116 Continuationofthe30(b)(6)Depositionof
RobertC.Merrill,Jr.,takenatPanAmerican
Orleans,Louisiana,70130,onthe16thdayof
January,2013.
Pdf TranscriptofDr.Merrills(BP)
56 2013130 DepositionofMichaelM.Levitan,Ph.D.,
takenatthePanAmericanBuilding,601
PoydrasStreet,11thFloor,NewOrleans,
Louisiana,70130,onthe30thdayofJanuary,
2013.
Pdf TranscriptofDr.Levitans(BP)
deposition:firstday
57 2013131 ContinuationofthedepositionofMichaelM.
Levitan,Ph.D.,takenatthePanAmerican
Orleans,Louisiana,70130,onthe31stdayof
January,2013.
Pdf TranscriptofDr.Levitans(BP)
58 2013124 DepositionofSECRETARYOFENERGY,DR.
STEVENCHU,takenatDepartmentofEnergy
Headquarters,1000IndependenceAvenue,
Pdf TranscriptofSecretaryStevenChus
deposition
MJBluntExpertReport
173
SW,Washington,D.C.,20508,onthe24th
dayofJanuary,2013.
59 20121024 30(b)(6)DepositionofMarciaKemper
McNutt,Ph.D.,UnitedStatesofAmerica,
takenatthePanAmericanBuilding,601
Louisiana,70130,onthe24thdayofOctober,
2012.
Pdf TranscriptofProf.McNutts(USGS)
deposition
60 20120912 DepositionofAnadarko,byandthroughBrian
ONeill,takenatPanAmericanBuilding,
September,2012.
Pdf TranscriptofBrianONeills
(Anadarko)deposition
61 20121031 DepositionofDonMaclay,takenatPan
31stOctober,2012.
Pdf TranscriptofDonMaclays
deposition:firstday
62 20121101 ContinuationofthedepositionofDon
Maclay,takenatPanAmericanBuilding,601
Louisiana,70130,onthe1stofNovember,
2012.
Pdf TranscriptofDonMaclays
63 20111018 BPHZN2179MDL03742328 Pdf MDTreportusedtofindthedatum
depth
64 20130411 Exhibit8624,Geologicalevidenceforan
elongate,heterogeneousreservoir
Pdf USGSdiscussionofconnectivity
65 20130421 Exhibit8635 Powerpoint GovernmentsFlowAnalysis
ActivitiesPresentation
66 20120910 DepositionofPencor,byandthroughJason
LeBlanc,takenatPanAmericanBuilding,601
September,2012.
Pdf TranscriptofJasonLeBlancs
(Pencor)depositionandassociated
exhibits8583and8584.
67 20120917 Exhibit8615 Pdf Dr.HsiehOctober22
nd
2010draft
report

N.3 Books,papersandreportsinthepublicdomain
AdetailedandfrequentlyupdatedbibliographyofpublicdomainpapersandreportsontheDeepwater
HorizonIncidentcanbeaccessedat:
http://www.lib.noaa.gov/researchtools/subjectguides/dwh_bibliography.pdf
MAbramovichandIAStegun,HandbookofMathematicalFunctions,DoverPublications(1970).
VAbreu,MSullivan,CPirmezandDMohrig,Lateralaccretionpackages(LAPs):animportantreservoirelementindeepwater
sinuouschannels,MarineandPetroleumGeology20631648(2003).
FOAlpak,MDBarton,FFvanderVlugt,CPirmez,BEPratherandSHTennant,SimplifiedModelingofTurbiditeChannel
Reservoirs,SPEJournal15480494(2010).
FOAlpak,MDBartonandSJNaruk,Theimpactoffinescaleturbiditechannelarchitectureondeepwaterreservoir
performance,AAPGBulletin9251286(2013).
ACAzi,AGbi,TWhittleandACGringarten,Evaluationofconfidenceintervalsinwelltestinterpretationresults,SPE113888,
proceedingsoftheSPEEuropec/EAGEAnnualConferenceandExhibition,Rome,Italy912June(2008).
MJBluntExpertReport
174
Bijeljic,HDong,OGharbi,SIglauer,PMostaghimi,APalusznyandCPentland,PorescaleimagingandModelling.Advancesin
WaterResources,51197216(2013).
DBourdet,WellTestAnalysis:TheUseofAdvancedInterpretationModels,Elsevier,Amsterdam,ISBN0444509682(2002).
DBourdet,TMWhittle,AADouglasandYMPirard,ANewSetofTypeCurvesSimplifiesWellTestAnalysis,WorldOil695
106(1983).
DBourdet,JAAyoubandYMPirard,UseofPressureDerivativeinWellTestInterpretation,SPEFormationEvaluation4293
302June(1989).
MJBourgeois,FHDaviauandJLBoutauddelaCombe,PressureBehaviorinFiniteChannelLeveeComplexes,SPEFormation
Evaluation11262268September(1996).
GEPBox,EmpiricalModelBuildingandResponseSurfaces,coauthoredwithNormanR.Draper,p.424,ISBN0471810339
(1987).
BP,DeepwaterHorizonAccidentInvestigationReport,Houston,TX.
http://www.bp.com/liveassets/bp_internet/globalbp/globalbp_uk_english/incident_response/STAGING/local_assets/downloa
ds_pdfs/Deepwater_Horizon_Accident_Investigation_Report.pdf(2010).AlsoAppendixW,ReportDynamicSimulations
DeppwaterHorizonIncidentBP(fromaeaddenenry).
RCamilli,DDiIorio,ABowen,CMReddy,AHTechet,DRYoerger,LLWhitcomb,JSSeewald,SPSylvaandJFenwick,
AcousticmeasurementoftheDeepwaterHorizonMacondowellflowrate,ProceedingsoftheNationalAcademyofSciences
109(50)2023520239(2012).
HSCarslawandJCJaeger,ConductionofHeatinSolids,2
nd
Edition,OxfordSciencePublications,ClarendonPress,Oxford
(1946).
TJCroneandMTolstoy,Magnitudeofthe2010GulfofMexicoOilLeak,Science330624(2011)plussupportingmaterial.
LPDake,FundamentalsofReservoirEngineering,ElsevierSciencePublishers,Amsterdam(1978).
LPDake,ThePracticeofReservoirEngineering,RevisedEdition,ElsevierSciencePublishers,Amsterdam(2001).
MDykstraandBKneller,Lateralaccretioninadeepmarinechannelcomplex:implicationsforchannellizedflowprocessesin
turbiditycurrents,Sedimentology,doi:10.1111/j.13653091.2008.01040.x(2008).
HDarcy,ThePublicFountainsoftheCityofDijon,Kendall/HuntPublishingCompany,Dubuque,Iowa,Englishtranslation2004
byPBobeck,originallypublishedinFrenchbyVictorDalmont,Paris(1856).
RCEarlougher,Jr.,HJRamey,Jr.,FGMillerandTDMueller,PressureDistributionsinRectangularReservoirs,Journalof
PetroleumTechnologyFebruary199208;Trans.AIME243(1968).
RCEarlougher,Jr.,AdvancesinWellTestAnalysis,HenryL.DoughertyMonographSeries,Volume5,SocietyofPetroleum
Engineers,Dallas,USA,ISBN0895202042(1977).
JForrest,EMarcucciandPScott,GeothermalGradientsandSubsurfaceTemperaturesintheNorthernGulfofMexico,GCAGS
Transactions55233248(2005).
AGovan,TPrimmer,CDouglas,NMoodie,MDaviesandFNieuland,ReservoirManagementinaDeepwaterSubseaField
TheSchiehallionExperience,SPEReservoirEngineeringandEvaluation,SPE96610PA,9382390(2006).
TGowers,Mathematics:AVeryShortIntroduction,OxfordUniversityPress,OxfordUK(2002).
ACGringarten,FromStraightLinestoDeconvolution:TheEvolutionoftheStateoftheArtinWellTestAnalysis,SPE
ReservoirEngineeringandEvaluation,SPE102079PA,11(1)4162(2008).
ACGringarten,MScWelltestanalysislecturenotes,ImperialCollegeLondon(2012).
SKGriffiths,OilReleasefromMacondoWellMC252FollowingtheDeepwaterHorizonAccident,EnvironmentalScienceand
Technology4656165622(2012).
HNHall,CompressibilityofReservoirRocks,JournalofPetroleumTechnology5(1)1719(1953).
THaugland,LLarsenandSMSkaeveland,AnalyzingPressureBuilupDatabytheRectangularHyperbolaApproach,SPE13079,
proceedingsoftheSPEAnnualTechnicalConferenceandExhibition,Houston,Texas,USA,September1619(1984).
MJBluntExpertReport
175
SHHickman,PAHsieh,WD.Mooney,CBEnomoto,PHNelson,LAMayer,TC.Weber,KMoran,PBFlemings,andMK
McNutt,ScientificbasisforsafelyshuttingintheMacondoWellaftertheApril20,2010DeepwaterHorizonblowout,
ProceedingsoftheNationalAcademyofSciences109(50)2026820273(2012).
RNHorne,Modernwelltestanalysis:Acomputeraidedapproach,Petroway,PaloAlto,CaliforniaISBN:0962699217(1997).
BZHsieh,TLChen,CCTseng,GVChilingarandZSLin,Casestudyofestimatinggaslossfromaproducingwellblowout,
JournalofPetroleumScienceandEngineering70(34)327333(2010).
PAHsieh,SimulationofReservoirDepletionandOilFlowfromtheMacondoWellFollowingtheDeepwaterHorizonBlowout.
Reston,VA:USGSOpenFileReport20101266http://pubs.usgs.gov/of/2010/1266/(2010).
PAHsieh,ApplicationofMODFLOWforOilReservoirSimulationDuringtheDeepwaterHorizonCrisis,GroundWater49(3)
MayJune319323(2011).
FPIncroperaandDPDeWitt,IntroductiontoHeatTransfer,JohnWileyandSonsInc.,NewYork,NY(1985).
MMKamal(Ed.),TransientWellTesting,SocietyofPetroleumEngineers,Richardson,Texas,USA,ISBN9781555631413
(2009).
JLee,JBRollinsandJPSpivey,PressureTransientTesting,SocietyofPetroleumEngineersTextbookSeries,Volume9,
Richardson,Texas,USA,ISBN9781555630997(2003).
BLiu,RDessenberger,KMcMillen,JLachandMJKelkar,WaterFloodingIncrementalOilRecoveryStudyinMiddleMiocene
toPaleoceneReservoirs,DeepGulfofMexico,SPE115669,proceedingsoftheAsiaPacificOil&GasConferenceand
Exhibition,PerthAustralia,2022October(2008).
Lubchenko,MKMcNutt,GDreyfus,SAMurawski,DMKennedy,PTAnastus,SChuandTHunter,Scienceinsupportofthe
DeepwaterHorizonresponse,ProceedingsoftheNationalAcademyofSciences109(50)2021220221(2012).
TMcHargue,MJPyrcz,MDSullivan,JDClark,AFildani,BWRomans,JACovault,MLevy,HWPosamentierandNJ
Drinkwater,Architectureofturbiditechannelsystemsonthecontinentalslope:Patternsandpredictions.Marineand
PetroleumGeology28728743(2011).
AJLove,CTSimmonsandDANield,Doublediffusiveconvectioningroundwaterwells,WaterResourcesResearch43
W08428,doi:10.1029/2007WR006001(2007).
MMcNutt,RCamilli,GGuthrie,PHsieh,VLabson,BLehr,DMaclay,ARatzelandMSogge,AssessmentofFlowRateEstimates
fortheDeepwaterHorizon/MacondoWellOilSpill,NationalIncidentCommand,InteragencySolutionsGroup,FlowRate
TechnicalGroup.http://www.doi.gov/deepwaterhorizon/loader.cfm?csModule=security/getfile&PageID=237763(2011).
MKMcNutt,SChu,JLubchenco,THunter,GDreyfus,SAMurawskiandDMKennedy,Applicationsofscienceandengineering
toquantifyandcontroltheDeepwaterHorizonoilspill,ProceedingsoftheNationalAcademyofSciences109(50)20222
20228(2012a).
MKMcNutt,RCamilli,TCCronec,GDGuthried,PAHsieh,TBRyerson,OSavasandFShaffer,Reviewofflowrateestimates
oftheDeepwaterHorizonoilspill,ProceedingsoftheNationalAcademyofSciences109(50)2026020267(2012b).
CSMatthewsandDGRussell,PressureBuildupandFlowTestsinWells,HenryL.DoughertyMonographSeries,Volume1,
SocietyofPetroleumEngineers,Dallas,USA(1967).
MMuskat,PhysicalPrinciplesofOilProduction,IHRDCPublishers,Boston,USA,ISBN:09346340706(1981).
GHNewman,PoreVolumeCompressibilityofConsolidated,Friable,andUnconsolidatedReservoirRocksUnderHydrostatic
Loading,JournalofPetroleumTechnology25(2)129134(1973).
TANewsonandPBrunning,ThermalConductivityofDeepwaterOffshoreSediments,InternationalJournalofOffshoreand
PolarEngineering14(4)NoTM31(2004).
HEHuppertandJSTurner,Doublediffusiveconvection,J.FluidMechanics106299329(1981).
CMOldenburg,BMFreifeld,KPruess,LPanL,SFinsterleandGJMoridisGJ,NumericalsimulationsoftheMacondowell
blowoutrevealstrongcontrolofoilflowbyreservoirpermeabilityandexsolutionofgas,ProceedingsoftheNationalAcademy
ofSciences109(50)2025420259(2012).
TLOsif,TheEffectsofSalt,Gas,Temperature,andPressureontheCompressibilityofWater,SPEReservoirEngineering3
February175181(1988).
MJBluntExpertReport
176
RMOstermeier,StressedOilPermeabilityofDeepwaterGulfofMexicoSands:MeasurementsandTheory,SPEReservoir
FormationEvaluation11226235(1996).
RMOstermeier,CompactionEffectsonPorosityandPermeability:DeepwaterGulfofMexicoTurbidites,Journalof
PetroleumTechnologyFebruary6874(2001).
HWPosamentier,Depositionalelementsassociatedwithabasinfloorchannelleveesystem:casestudyfromtheGulfof
Mexico,MarineandPetroleumGeology20677690(2003).
HWPosamentierandVKolla,Seismicgeomorphologyandstratigraphyofdepositionalelementsindeepwatersettings,
JournalofSedimentaryResearch73(3)367388(2003).
SEPringleandRJGlass,Doublediffusivefingerconvectioninasolutesystematfixedbuoyancyratio:Structuralintricacyand
lengthscalegrowthinRayleighspace,JournalofFluidMechanics462161183(2002).
RRaghavan,Welltestanalysis,PrenticeHall,EnglewoodCliffs,NewJersey,USA,ISBN:0139533656(1993).
GNRagagninandMASMoraes,SeismicGeomorphologyandConnectivityofDeepwaterReservoirs,SPEReservoir
EvaluationandEngineering11(4)686695(2008).
HJRamey,WellboreHeatTransmission,JournalofPetroleumTechnology,April427435(1962).
CMReddy,JSArey,JSSeewald,SPSylva,KLLemkau,RKNelson,CACarmichael,CPMcIntyre,JFenwick,GToddVentura,B
ASVanMooyandRCamilli,CompositionandfateofgasandoilreleasedtothewatercolumnduringtheDeepwaterHorizon
oilspill,ProceedingsoftheNationalAcademyofSciences109(50)2022920234(2012).
TBRyerson,RCamilli,JDKessler,EBKujawinski,CMReddy,DLValentine,EAtlas,DRBlake,JdeGouw,SMeinardi,DD
Parish,JPeischi,JSSeewaldandCWarneke,ChemicaldataquantifyDeepwaterHorizonhydrocarbonflowrateand
environmentaldistribution,ProceedingsoftheNationalAcademyofSciences109(50)2024620253(2012).
RMSaidikowski,NumericalSimulationsoftheCombinedEffectsofWellboreDamageandPartialPenetration,SPE8204,
proceedingsoftheSPEAnnualTechnicalConferenceandExhibition,2326September,LasVegas,Nevada(1979).
MHSharqawy,JHLienhardVandSMZubair."Thethermophysicalpropertiesofseawater:Areviewofexistingcorrelations
anddata,"DesalinationandWaterTreatment,16354380(2010). http://dx.doi.org/10.5004/dwt.2010.1079
PSkalle,DrillingFluidEngineering,bookboon.com,ISBN:9788776819293(2011).
SVStankus,IVSavchenko,AVBaginskii,OIVerba,AMProkopevandRAKhairulin,Thermalconductivityandthermal
diffusivitycoefficientsof12Kh18N10Tstainlesssteelinawidetemperaturerange,HighTemperature46731733(2008).
TCWebera,ADeRobertis,SFGreenaway,SSmith,LMayeraandGRice,Estimatingoilconcentrationandflowratewith
calibratedvesselmountedacousticechosounders,ProceedingsoftheNationalAcademyofSciences109(50)2024020245
(2012).
JWhelan,LEglinton,LCathlesIII,SLoshandHRoberts,SurfaceandsubsurfacemanifestationsofgasmovementthroughaN
StransectoftheGulfofMexico,MarineandPetroleumGeology22479497(2005).
CWolfe,CRussell,NLuiseandRChhajlani,LogBasedPoreVolumeCompressibilityPredictionaDeepwaterGoMCase
Study,SPE95545,proceedingsoftheSPEAnnualMeeting,Dallas,TexasOctober(2005).
GRWooley,ComputingDownholeTemperaturesinCirculation,InjectionandProductionWells,JournalofPetroleum
Technology15091522(September)(1980).
RWZimmerman,CompressibilityofSandstones,ElsevierSciencePublishers,NewYork,NY,USA,ISBN:0444883258(1991).
http://www.bp.com/liveassets/bp_internet/globalbp/globalbp_uk_english/incident_response/STAGING/local_assets/downloa
ds_pdfs/BP_technical_audio_08032010.pdf
http://www.engineeringtoolbox.com/specificheatfluidsd_151.html
http://www.engineeringtoolbox.com/specificheatsolidsd_154.html
http://www.engineeringtoolbox.com/overallheattransfercoefficientsd_284.html
http://fluidflowvisualization.sandia.gov/double_diffusion.html

BP Macondo Estimate 2

Enviado por

Dados do documento

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

BP Macondo Estimate 2

Enviado por

Direitos autorais:

Formatos disponíveis

Exhibit 1

Case 2:10-md-02179-CJB-SS Document 10533-3 Filed 06/27/13 Page 2 of 106

Root mcon squorc crror in prcssurc = _

We can shift the final pressure by only around 10 psi before a

The skinintroduces anadditionalpressuredrop fromthe reservoirtothe wellbore that isgiven, from

2.618 2.539 2.5104 2.556 2.473 3.2

2.339 2.262 2.3875 2.330 2.227 4.4

F.2.4 Multistage separation.TheZick reportsuggeststhatitismorerepresentativetousevaluesofB

F.4.3Confusion over fluid properties.ThePooladiDarvishreportstatesthatituses:

- Pore volume compressibility. As mentioned above, Dr. PooladiDarvish took a representative

0.93MMstb 0.16MMstb 0.68MMstb 3.26MMstb

Case 2:10-md-02179-CJB-SS Document 10533-3 Filed 06/27/13 Page 39 of 106

10,734,170 209,361 1,392 10,458 1.46 8.53

Symbol Meaning Units(field) Units(SI)

Eulersconstant dimensionless dimensionless

c Strain dimensionless dimensionless

| Porosity dimensionless dimensionless

putin place. Theselogsareused toinfer water(and hence oil) saturationand

38 20110802 BPHZNBLY00000526 Pdf BPDeepwaterHorizonAccident

Você também pode gostar