—

.-. —— ——

——-——
:—.
,..
SPE 39-926

Monte Carlo Techniques for Evaluating Producing Properties

J. R. Gilman, R. T. Brickey, M. M. Reddi Marathon Oil Company

.:
@@9hf l=, S=iefy of Petroleum Engirraars. Tnc
,,
W paper was preparad Tor pfesentafi~ =_99S SPE Rtiy Mountain Ragiona17Low- result of an unsuccessful workover. Also, a workover could
Parmeability Reservoirs Symposium ati Exhbition held in Denver, Colorado. H April 1998.
have some chance of failure resulting in no change to the
This paper was seleclad Tor presentation by en SPE Program Committee foIlowing review of
information mtiad in an abatract subinifted by fhe atihor(s]. Con!enls of the papar, as expected reserves.
Prasankd, han not been reviAwed by the Society of Patrolaum Enginaars and ara subjact 10
~ by me aulhws) me material, ae presented, does not necessarily reflect any
The parameters which were risked for each well were
@sitiin of fhe S&7aty of \ atroleum Enginaars, ifs oficars, or members. Papers presented at downtime, welI life, weI1faiIure, workover reserves, workover
SPE maatings ara subject to publication reviaw by Editorial Committees of the Sociely of
Petroleum Engineers, Electronic raprcducticm, distribution, or atoraga of any pafl of this paper incremental rate, workover failure, and workover expense. For
for mmemial purposes wlhouf the written mnaent of the Sociely of Petroleum Engineera is
prohibited. Permission to rapmduce in prinl is rastricled 10 an abstract of not more lhan 300
the field, the “risked” variables were hydrocarbon prices and
&s: titrations mav not ba woid l’he abstract must contain mnsoicuous operating and abandonment expenses. All the uncertainty
adm&ladgmant of Whar; and by whom ~ha paper was presanled. Write Librarian, SPE, P 0.
~ -%, Richardsm. TX 750S3-3836, U~S.A,
fax01 -972-952-~35. described here could not have been properly assessed without
the use of Monte Carlo techniques.
Abstract
When purchasing property, selling property or estimating Introduction
fbture Ctih-fliw from a project, forecasts of future production’ Monte CarIo methodologies are at least a half century old and
rates are required. Whether those forecasts are made with even in the first half of this century, many statistical sampling
detailed reservoir :imuIation or decline curve techniques, methods could be considered Monte Carlo techniques’. me
there will be uncertainty in the forecasts. Even if the reservoir recent emphasis in the petroleum literature in regard to the
description was exact and an accurate reservoir simulation use of Monte Carlo sampling appears to be a result of
could be made, the field. development plans would be improvements in computer hardware and sotiare which has
uncertain because of unce,fia~ntyin future product prices and simplified the use of the technique. The Mon~e Carlo method
expenses. This paper demonstrates how Monte Carlo peforms random sampling from probability functions which
techniques were used to generate a range of production profile describe the uncertainty of various input parameters In” a
distributions to estimate the economic value of a producing mathematical model. The di~culty of estimating the
prop-. probability functions and the accuracy of the underlying
The basic methodology used decline curve equations for mathematical model is still a major Imitation in the use of the
each weI1 in the field to describe the future production rate of technique. The ‘probability finctions we chose in this paper
the wells. Hyperbolic decline equations were used for aIl were very subjective based on our experience and so-me
wells. Gas-oil ratio was modeled by another empirical limited historical ‘data.
equation. ”The standard decIfi”e equations were modified to Within the petroleum industry, Monte Carlo techniques
properly account for downtime while still honoring ultimate have been used quite extensively in the exploration business3
recovery. Downtime thus resulted in delayed production. but to a much lesser degree in production forecasting. ~is
However, a weII did not always reach its expected ultimate paper demonstrates how the Monte Carlo technique was used
recovq if it “failed” for mechanical reasons or if the field for a risk analysis approach to production forecasting, in a
was abandoned for economic reasons. Well failure was single field. In this field there is much uncertainty in wethore
modeled as a discrete distribution (i.e., the producing rate was integrity and workover success. In our production risk
multiplied by a value of zero or one). analysis we were trying to choose among a number of fiture
Antither unique =pect of the spreadsheet presented here development schemes which would maximize our revenue.
was the incorporation of mechanical and workover faiIure This paper does not outline the various comparisons, ‘but
probabilities. The decline equations were modified to allow illustrates the methodology
for changes in reserves and rate as a result of workovers. The
expeded change in reserves could have some probability of
being negative if there was risk of damaging a well as a
 .“

2 J. R. GILMAN, R. T. BRICKEY, M, M. REDD S~E 39926

.— .-—
,“
Description of Methodology second type, the new completions were commingled with the
me basic mefiodology used decline curve equations4 for each current completions. To incorporate increased rate and
weII in ti field to deicr~e the future production rate of the reserves into the decline equations, the initial rate and
wells. Commercial spreadsfieet and add-in sotiare programs cumulative production following a workover was increased by
were used to allow an input probability distribution (rather the Monte Carlo sampled values. Eq. 2 was then use~ to soIve
Wan a single value) for certain spreadsheet cells. In the for a new decline constant. Workover fa;ltire was a~so
commercial sotiare, Monte Carlo techniques ran the modeled as a discrete distribution with a “failure” value of O
spreadsheet for a specified number of iterations. At each or 1. The workover failure probability depended on the
iteration, the sotiare sampled the input probability completion type and the mechanical condition’ of. :..the “weII
distributions (e.g. normal, Iog normal, triangular) and gathered FaiIure probabilities ranged from 20 to 75%.
statistics for specified output cells2’s’6.The Latin Hypercube Rather than assuming a constant gas-ofi ratio,
method of sampling’ was used in this analysis to ensure cumulative gas production (G~) was assumed to be a-finction
repres~ sampling for a reduced number of iterations of cumulative oil production ~NP) according,to Eq. 3s.
(e.g. 500]. Hyperbolic decline equations were used.
Hyperbolic decline assumes that decline is proportional to a
fictional power (O<b<I) of the production rate. The rate- (3)
thne relationship is given by Eq. 1.
,..
where

fl=n-(n-1)+-+
where qi’is the initial rate and Di is the initial decline factor.

Cumulative production (,VP)is given by Eq. 2. Eq. 3 requires input of initial gas-oiI ratio (Ri), final gas-oiI
—~::!
i-atio(Rfl, average well life gas-oil ratio R and a curvature
()
.Np.-=.---q: parameter (n). Average gas-oil ratio for this study was set to
(l-b)Di(q:-’-q’-’)
(2)
be 1.5 times the initial value, a curvature parameter of 5 was
assumed and final g&-oil ratio was calculated from Eq.-Z;
The above equations approximate an exponential decline ,..:...=, -,
as ~ approaches zero and in a harmonic decline as b
approaches 1.” ~ese equations were “modified in the R,= 0.75(n -(n -2):)$ (4)
spreadsheets to properIy account ‘for downtime in order’ that .. . . ,+
cumulative oiI production was always given by Eq. 2 and
Input Assumptions
downtime resulted in delayed production. For a given time
Following is a summary of the spreadsheet input data based
rnterval, Eq. I was used to calculate the ending rate from the
on the above equations. The parameters which were aIIowed
initial rate assuming no’ downtime. Eq. 2 was then used to
to vary (risked) for each well were downtime, well life (which
calculate cumulative production for the period assuming no
alters decline rate and reserves), well failure, workover
downtime. Cumulative production for the period was then
reserves, workover incremental rate, workover failure, and
reduced by downtime and Eq. 2 was re&anged to solve for
workover expense. For the field and well batteries (platforms),
ahl -ffial rate ~- “a function of the adjusted cumulative
production. A welI wou~< not always reach its expected the risked variables were hydrocarbon prices and operating
and abandonment expenses.
cumu~ve production if ~t‘failer for mechanical reasons or
For each well the following information was required
if a well battery (piatforin) was abandoned for economic
(see Table I). This information is based on past historical
reasom. Failure was modeled as a discrete distribution (i.e. the
performance of this field and other fields in the area.
“failure+”value ,w& O ~[ I) which multiplied the producing ... . ..,.,
rate. For the field under study, the probability for well faiIure
1. Economic Oii Rate, STB/mo -- This is the economic
was s~t to be 1‘A for the first year and was increased by a
rate limit for the well (i.e. to shutin the weII). The
factor of 1.I -.———
each year thereafter.
. ...-. ,,, economic limit affects the expected reserves according
Equations I M“d2 were also modified to allow for step
changes in reserves and rate as a result of workovers. Tw”o to Eq. 2.
2. Downtime, ‘A -- ~is is expected well downtime. A
types of workovers were considered. In the first type, the
normal distribution with a mean value of 5.33°/0 and
current completions are nebng their economic life and were
standard deviation of 1.07°Awas applied to this value
shut-in when the new completions were opened. In the
SPE 39926 MONTE CARLO TECHNIQUES FOR EVALUATING PRODUCING PROPERTIES 3

for the Monte Carlo sampling. Downtime reduced to sample workover reserves for each Monte Carlo
cmntive oiI production for each year effectively simulation from input values of minimum, expected
deIaying production to a fiture time. and maximum workover reserves. Minimum reserves
3, Initial Oil Rate, ST31mo -- The initial oil rate is a were 750/. of the expected reserves. Maximum reserves
known input value. For subsequent times, the oil rate were 125°/0of the expected reserves.
was a calcubted vaIti based on Eqs. 1 and 2 adjusted 11. Workover Average GOR, Mscf/STB -- This value
h downtime and workovers. was used to caIcuhte workover gas reserves which
4. DecLine Factor, I/yr -- me initial decline factor is a were then added to original gas reserves following a
known input value, Subsequent decline factors were workover. This resulted in a new value for ultimate gas
~ by the hyperbolic constant and a risked range reserves (Gpa) in Eq. 3.
of reserve Iife, and step changes in reserves and rates 12. Workover Incremental Rate, STB/mo -- This is the
due to workovers. The decline factor was also adjusted incremental rate added to well rate as a result of a
for downtime. In other words, if downtime was 100% workover. A triangular distribution was used to sample
for a year, there wouId be no decline in rate. This workover rate for each Monte Carlo simulation from
ensures that the fmaI unrisked cumulative recovery input values of minimum, expected and maximum
agrees with the initiaI decIine estimate plus workover workover rate. Minimum rate was 75°Aof the expected
reserves @rovided there were no well or workover value. Maximum rate was 1250/.of the expected value.
fiIures). 13. Recompletion Expense, $ -- This represents possible
5. Initial GO~ Mscm -- The initial gas-oil ratio costs of a workover. A triangular distribution was used
(GOR) of the weII is a known input vaIue.-Subsequent to sample recompletion expenses for each Monte Carlo
values were calculated Eom Eq. 3. simulation from input values of minimum, expected
6. Average GO% Msc~~ -- Eq, 3 is used to calculate and maximum workover expenses. Minimum expense
the cumtiative gas production and thus current gas-oil was 900/. of the expected value. Maximum expense was
ratio of the welI. ~ equation requires initial, average 150°Aof the expected value.
and fti GOR’S as well as a curvature parameter (n), 14. Workover Failure Probability, V. -- This is the
The curvature parameter is equal to 5.0 and the final probability that a workover will ‘fail (i.e. it will add no
GOR is function of the other three parameters (Eq. 4). incremental rate or reserves). Failure probabilities
The average GOR is 1.5 times the initial GOR. varied from 250/. to 75°/0 depending on the condition of
7. Hyperbolic Constant, dimensionless -- A constant the well (type of completion and mechanical
valu of O.I was used for all wells. condition). If a well “failed” during a Monte Carlo
8. WelI Life, Years -- Eq. 1 was used to calculate an simulation, then no reserves or rates were added to the
expected well life (assuming no downtime). Input of a current well values; however recompletion expenses
minimum and maximum well life, allowed variability were still added to total field costs. .,,
in the welI life (and ~us decIine factor and reserves) for
Monte CarIo analysis. Miimum well life was 75% of To arrive at a total rate profile and net present value for the
the expected well Iife. Maximum well life was 150% field and each well battery (or platform), the individual rates
of the expected welI Iife. A triangular distribution was and expenses of all the wells were added together. me
nsed to sample welI life for each simulation based on following additional data was required for each battery and
the minimum, expected and maximum well life values. the total field (see Table II). No capital expenses (other than
9. WelI FaiIure Probability, V. -- This is the probability workovers) were used in the results presented here.
of a mechanical weI1 faiIure that would cause a well to
be shut-in prior to reaching economic limit. A failure 15. Base Oil Price, $ -- ‘An input base price of ““oil’ was
probability of IY. was assumed for the first year and entered for the first year. This base price was then
was hicreased by a factor of 1.1 for each subsequent escalated each year ( 1.5°/0in this example). The actual
year. Rates calculated from the decline equations were Monte Carlo sampled price was based on a normal
multiplied by a discrete distribution (a value O or 1) as distribution with a standard deviation equal to 10% of
defined by the failure probability. the base price.
16. *e Gas Price, $-- An input base price of g& was
For each scheduled workover, the following information was entered for first year . This base price was then allowed
required. to escalate each year ( 1.5% in this example). The actual
Monte Carlo sampled price was based on a normal
10. ~orkover Reserves; STB -- This represents the range distribution with a standard deviation equal to 10% of
of 03 reserves to be added to total welI reserves as a the base price.
result of a workover. A triangular distribution was used 17. Base Plant Liquids Price, $-- An input base price of
5.-
,-.
~—
,..
..=
,.
4 J. R. GILMAN, R. T. BRICKEY, M. M. REDD SPE”39926
-. .—

plant liquids was entered for first year. This base price distribution. Workover or mechanical well failures were
w-en aI1owed to escalate each year ( 1.5°Afor this sampIed using a discrete distribution. “’ “
extunple). The actua~ Monte Carlo sampled price was
based on a normal distribution with a standard Monte Carlo Results
deviation equal to 10VOof the base price. Once all tle input data was entered, a Monte Carlo technique
18. Operating Expenses, $ -- These are total field and was used to sample all the input distributions and run a
battery operating expenses, These values were input number of spreadsheet “iterations”. The commercial software
f6r each year. A 3% escalation was used for these program ‘then generated statistical information from the
..... -—- iterations, Five hundred iterations were used “to generate the
expmses. A triangular distribution was used to sample
these expenses for each simulation from the input resuIts shown here. This required about 1.5 CPU hours on a
vaIues of minimum, expected and maximum costs. The 200MHz Pentium processor. For each Monte Carlo iteration,
minimum expenses were 90°A of the most likely the software had to evaluate over 4000 distributions in this
expenms. The maximum expenses were 150°A of the spreadsheet. Figures I through 8 summarize results from the
‘“rniw-’likely expenses, Operating expenses were set Monte Carlo simulation. Figure 1 shows the cumulative
‘e~al to zero once’ the rate of a battery reached the probability distribution function (Colt) for final net-present
Gcottomjc limit (revenues were less then expenses). value. The figure shows that the current net present value
Thus the battery economic limit varied for each Monte (NPV) ranges from about 75% to 125% of the mean value.
CarIo simulation. Figures 2 through 8 show expected values for NPV, cash flow,
19. Abandonment Cost, $ -- This is field abandonment oil rate, cumulative oil, gas-oil ratio, iurn-ulative gas and
costs. A 3°A escalation was used for these cost. The active number of wells as a function of time (alI plots are
abandonment cost was not applied until the field relative to a mean vaIue). The active well plot (Fig. SJ shows
became uneconomic (revenues were less than operating the decline in number of wells as a result of failure, well
tx-p=s~r~he final year. A triangular distribution economic limit or battery economic Iimit,
was used to sampIe abandonment cost for each Six lines are shown on the plots -- unrisked value, mean
simulation from the input values of minimum, expected value, and 10/0,10O/o, 90°Aand 99°h confidence ranges. A 10°/0
and maximum costs. The minimum costs were 90°A of confidence range means that there is only a 10Ohchance that a
the most Iikely costs. The maximum costs were 150°A Monte Carlo iteration would faII below the disp~ayed value.
of the most likely costs. The 90% confidence range implies there is only a 10O?chance
20. Discount Rate, % -: This is the constant discount rate a simulation would fall above the displayed vaIues.-Note that
used to estimate net present value. the unrisked oil and gas production values are generaIly
higher than everi the 99% confidence interval. The unrisked
- Usirig ‘-tie above ~nforrnation, the total field (and case used mean input values for aI1 analysis except that
battery) oil, gas and plant liquid revenues were calculated by workover and well failures were not allowed. This-indicates
year. Fuel gas (a constant input value for each battery) and the power of Monte Carlo techniques to account for discrete
gas shrinkage was accounfed for in the gas revenue stream. A fiilure probabilities.
con-” Iiqtiid yield was assumed for calculation of plant This actual field evaluation case demonstrates the
liquids. power of Monte Carlo techniques for allowing uncetiainty in
ExpenSes-and royalties were subtracted from revenues many variables which affect future field performance. The
to ‘d@ermrne cash flow. Cash flow was set to zero once a mean (expected value) is not the same value calculated from a
battery (platform] was abandoned (except for the main batte~ shnpIe ‘spreadsheet using all the mean values for ‘input.”T’hisis
expenses which continue to the f’maI year). Batteries were especially true when incorporating discrete distributions (e.g.
abandoned when current cash flow was unabIe to meet well and workover) failure. Thus such an analysis can be used
opera- @ses. ‘The ‘a~andonment costs were not applied to better ektimate the expected value of an expensive
until the field became uneconomic (revenues were less than workover or the value of accelerating the workover schedule
o~erating ~msj or in ‘the final year, The cash flow was to avoid an increasing degree of weIl and workover fa~ures.
discounted “using a ‘constarit input discount rate to determine ., ...-,
net present value ~Pv. Table I shows an example Conclusions
spreadsheet for a single well and Table II shows the A technique is demonstrated for evaluating producing
spretisheet calculations for a single battery. properties using commercial spreadsheets and Monte Carlo
In the spreadsheet, inputs which require a minimum, sotiare. An actual field case demonstrated the usefulness of
most likely (expected) , and maximum value were sampled Monte Carlo techniques for incorporating workover and well
ass~ing”a trimgular distifiution (e.g. well life, and workover failure in the economic analysis. Such an analysis allowed an
expense). nose with a base value and standard deviation (e.g. unbiased comparison among various field development
down time and oil price)”were sampled assum-ing a normal scenarios.

224
SPE 39926 MONTE CARLO TECHNIQUES FOR EVALUATING PRODUCING PROPERTIES 5

References
I. Meyer, H, A.; Symposium on Monte Carlo Methoa!s, John Wiley
& SO% Inc. New York, (March 15-16, 1954).
2. MurthL J. A.: “Monte Carlo Simulation: Its Status and Future,”
.JPT(ApriI 1997), p. 361-373.
3. Steinmeti R.: fie Business oJPetroIeum fiploration, American
Association of Petroleum Geologists, Tulsa Oklahoma ( 1992).
4. Arps, J. J. and Smith, A. E,: “Practical Use of Bottom-hole
Pressure Build-up Curves.’’ API Drill. Prod. Practice (1949), p.
155.
5. Connolly, M. R,: “Probabilistic Model Better Defines
Development Well Risks’”,~il&Gas J. (Ott. 14, 1996),pp.78-
84.
6. M- J. A.: “Incorporating Historical Data Into Monte Carlo
Simutio&” SPECA (April 1994), p. 1I- 17,
7. S~ R. A. Ad Wattenberger, R. A. “An Improved
Computation Procedure for Risk AnaIysis Problems With
Unusual Probability Functions,” paper 13722 SPE Hydrocarbon -—
Economics and Evaluation Symposium, Dallas, TX (March 14-
15, 1985).
8. Teague, N.: “Program Projects Realistic Solution Gas Drive
GOR”, Ofl-&Gas J. (Feb. 21, 1994), pp. 51-53. -

Nomenclature
P= gas-oil ratio equation coefficient in Eq. 3
(dimensionless)
b= hyperbolic decline constant (dimensionless)
Di = initial decline factor (1/day)
~p . cumulative gas production to date (Mscf)
Gpa ultimate gas recovery (Msc~
R= average gas-oil ratio (Mscf/STB)
Rf = finaI gas-oil ratio (Mscff/STB)
Ri = initial gas-oil ratio Mscf/STB)
= Eq. 3 curvature parameter (dimensionless)
:P . cumulative oil production to date (STB)
Npa = ultimate oil recovery (STB)
9= current oil rate (STB/day)
9i = initial oil rate (STB/day)
t= thne, days
 ,—-
~:,,.
,_
A. .=...
6 J. R. GILMAN, R: T. BRICKEY, M. M. REDD .SPE 39926
. ..— .=
,- .-
.:...=+
.=.
‘..

TabIe I. Spreadsheet for Well EX-1

.. .,. :.. . . . .
Start Date 1-Jan-97 ..
Date (End of Pefiod) 31-Dec-97 31-Dec-98 31-Qec-99 30-Dec-00 31-Dee-01 ~31.-Dec-O2, -
,,’,

Economic Oil Rate (STB/mo) 900 900 900 900 900 900 900
Expected Downtime % 5.33% s.sz~o 5.25% 4.14% 5.31% 5.27% 5.50%
lnital Oil Rate. STBfmo 3990 3990 2999 2270 1724 10334 6031
Decline Factor (i/yr) 0.3120 0.3096 0.3009 0.2926 0.2847 0.5948 0.5636
Initial GOR MSCF/STB 3.05 ..= ..—
Average GOR, MSCF/STB 4.58
Hyperbolic Factor 0.10
Minimum Well Life, Years 4 ,.
Maximum Well Life, Years 8
Well Failure Probability, % 1Y. 1% 1% 1% 1% 2% ,2% ‘- -&_...
Well Success Probability 99% 99% 99% 99% 99% 98% 98%
Final Oil Rate (WODT) STB/mo 2944 2230 1701 1302 5799 3485
Hnal Oil Rate (actual)STB/mo 2999 2270 1724 1324 6031 3622
Avg. Dil Rate, STBfmo 3255 2457 1889 1423 7415 4376
Total Oil Production(wo DT) STB 41116 31122 23642 18027 93932 55570
Total Oil Production(actual) STB 38931 29487 22664 17070 88984 52512
Total Oil to Date, STB 38931 68418 91082 108152 197136 249648 <
Remaining Resewes, STB 126845 87914 58427 35763 18693 116917 64405
Calculated Final GOR 5.15
Calculated GP, MSCF 580315
GOR, MSCF/S~ 3.17 4.11 5.40 6.11 0.00 1.76
Period Gas Production, MSCF 123578 121267 122439 104360 0 92255
Cumulative Gas Production, Mscf 123578 244844 367284 471644 471644 563899
Remaining Well Life, Years 5.15 5.19 4.25 3.31 2.36 4.65 3.72
Minimum Decline Factor 0.42
Minimum Oil Reserves, STB 94414
Maximum Decline Factor 0.21
Maximum Well Reserves, STB 188828

Note: Workovers Are Applied to the Year Following the Workover: ..

Min. Workover Reserves, STB o 0 0 127500 0 .0.
Most Likelv Workover Reserves. STB o 0 0 170000 0 0
Max. Wok-over Reserves, STB o 0 0 212500 0 0
Workover Average GOR, MSCF/STB 0 0 0 1 0. 0
Mn. WO Inc. Wte, STB/mo 0 0 0 5625 0 0=
Most Likely WO Inc. Rate, 0 0 0 7500 0 .,, 0
STB/mo
Maximum WO inc. Rate, STB/mo 9375
Mirr Recompletion Expense, $ $:
$0
$:
$0
$: $52,093 $: $:
Most Likely Recompletion Expense, $ $0 $57,881 $0 $0
Maximum Re@mpletion Expenses, $ so $0 $0 $86,822 $0 $0
Workover Failure Probability, % 50~o 50% 50% 20% 50% 50%
Workover Suc~ss Probability, Y. 5070 50% 50% 80% so~o 50%
Risked Workover Reserves, STB 0 o o 187209 o o
Risked Workover Rate, STB/mo 0 0 0 9010 0 0
Workover Gas Reserves, MSCF 0 0 0 187209 0 0.

,-
. .=.

. .
.,. .
SPE 39926 MONTE CARLO TECHNIQUES FOR EVALUATING PRODUCING PROPERTIES 7
. ..— ——

._
TabIe ~. Spreadsheet for Well Battery C

Start Date 1-Jan-97

Date (End of Period) 31-Dee-97 31-Dee-98 31-Dec-99 30-Dec-00 31-Dec-01 31-Dee-02

Battery Economic Oil Rate 0 3556 2977 2854 2780 3871 3506
Battery ‘C’ Oil Rate, STB/mo 13682 11667 9637 8362 7165 12649 9086 .
Period Oil Production, STB 139527 115650 100582 85982 151786 109029
Battery ‘C’ Oil Production, STB : 139527 255177 355759 441741 593528 702557
Battery ‘C’ GOR, MSCFISTB 3.69 3.74 4.27 5.01 5.48 . 2.23 3.48
Battery ‘C’ Gas Rate, MSCF/mo 50450 43591 41196 41986 39274 28226 31634
Period Gas Produdlon, MSCF 0 521300 494356 503826 471294 338777 379607
Battery ‘O Gas Production, Mscf 0 521300 1015656 1519482 1990776 2329493 2709100
Fuel Gas, Mscf (Col. 1 = Rate/Day) 260 94640 94965 94965 94965 94965 94965
Gas Plant Liquids, STB (Col. ~ =Yield) 34.11 14553 13623 13946 12837 8314 9709 :
NPV Rate (Discount Factor), V. 15% 15% 15% 15% 15% 15~o 15% -.
Base Oil Price, $/STBjCol: 1=Esc.) 1.50% $24.07 $24.43 $24.80 $25.18 $25.56 $25.94
BaseGasPrice, $/MSCF (C~l.,~ =Esc.) 1.50% $1.87 $1.90 $1.93 $1.96 $1.99 $2.02 .
Plant Liq. Price, ,&STB (Col. 1=Esc.) 1.50% $15.70 $15.94 $16.18 $16.42 $16.67 $16.92
Minimum OpeTating Expense, $/yr $1,079,268 $1,079,268 $1,079,268 $1,112,136 $1,146,004 $1,180,903
Most b~ely operating Expense, $/yr 3% $1,199,187 $1,199,187 $1,199,187 $1,235,706 $1,273,337 $1,312,115
Maximum Operating Expense, $/yr $1,798,780 $1,798,780 $1,798,780 $1,853,559 $1,910,006 $1,968,172
M~nimum Capital Expense, $/yr $0 $0 $0 $0 $0 $0
Expected Capital Expense, $/yr 3% $0 $0 $0 $0
Maximum Capital Expense, $/yf : $0 E
Minimum Abandonment Cost, $ $3,729,673 $0 : g $0 :
Expected Abandonment Cost, $ $4,144,081 : $0 $0 $0 . $0
Maximum Abandonment Cost $ $6,216,122 E $0 $0
Minimum Recompletion Expense, $ 0 $0 $0 $0 $52,0;! : $36,4~
Most Likely Recompletion Expense, $ 0 $0 $0 $0 $57,881 $0 $40,517
Maximum Recompletion Expense, $ 0 $86,822 $60,776
Expenses, $ $1,377,6: $1,260,3~ $1 ,156,0ti $1,343,6~2 $1 ,536,3;; $1,591,395
Oil Revenue, $ O% $2,948,768 $2,888,856 $2,486,346 $2,377,497 $3,605,826 $2,745,733
Gas Revenue, $ (Col.1 =Shrinkage) 7.73yo $746,649 $766,310 $824,792 $582,078 $410,917 $491,327
Plant Liquids Revenue, $ o% $238,678 $226,830 $211,051 $225,496 $143,604 $163.676
Cash Flow, $ (Col. f = Royalty) 16.67% $1,900,714 $1,974,675 $1,779,157 $1,310,615 $1,930,628 $1,242,552
Battery ‘C’ Cash Flow (NPV), $ $1,900,714 $1,699,671 $1,318,113 $835,762 $1,059,682 $587,030
Battery ‘C’ NPV, $ $7,359,965 $5,459,252 $3,759,581 $2,441,468 $1,605,706 $546,024
Active Number of Wells 8 8 7 7 7 7 7.
>.
...’
.-
,.. ,..
,.
.,.
,“.’

.>: -

.
,.’ ,.
.—.
.. ,- .,.
.’

. --

..
.,. .,
8 J. R: .GILMAN, R. T.. BRICKEY, M. M. REDD ~PE 3992
.,

.—
j. 1
I
Fig, ‘1.~stribution tiTotal Field W I
Fig. 2 Total Field Wt Present Valw
1 i’ I 1.40_.
0.9 1

1.20 ----------------------- __~

0.8 i
I I.cxl
0.7
1
0.6 I O.m
I
a5
O.m
0.4
0.40
a3

0.2

al
0.00
o
a5 1.0 -o.a
123456789107712 13
I

I
I
i.—..
..-...,,,,
. ,:.
--.,--- ;
., =...,
I
Fig.4 Field Oil@, .
1.40 ~: :-.
. – __:_u___
.. --........... . ...-...7 /. I.m q-- t
. .. $ I I

la . . . . .. . . O.m

I.al \~ 0.60
1

+--- .,
-----

0 1

-—.

Uw ~-.
, 0.20
a;.
0.10
---- ~
-a40 - -. -------- -— -- —--—-–—— a am —–.—.— :
1234567891011 1213 23456789101112 1314
-. —..-— ,.....—
-a--~ ~~o —M?an -

~~ —-9-= . . ..~
...—.
—..— — . . ___ .—-~ —
 SPE 39926 MONTE CARLO TECHNIQUES FOR EVALUATING PRODUCING PROPERTIES 9
,.
.

Fig. 5. Field Oil Htion Fig. 6. Field Ga~l Ratio

la .._. ._ 3.a _..—_— __
.- ,!

-,
I .
3.00 –1
““-A--”--
~
1

jo.m . . . . –.. —... ..—

. . . . . .

0.=:. . . .._. ..-____..____ .-_____”_j

-,..-
, am f . ......–-—.—--- ––.– .—----- O.m ,. .......-=, 4
12345678910111213 14 123456789101112 1314

L_ —

Fig. 8. Wve Number of Well

I la ,.–
I

am: ----- . ._~

am- ..- . k
Cla:. -.....J
,U. a..

- -+

-- ----- \
am# .--— —.——. ..—
12345678910111213 14
0.0. . ... —.——— .——.—_
123456789101112 1314

_.g_Pl ~plo -h