Technology Today Series

Case History: Field Development

Design Using A Risk-Based
Integrated Production Model
C.V. Chow, ProDuce Consulting; J.C. Mothershead, DeGolyer and MacNaughton; and
D.L. Cox, New Territory Petroleum

Abstract The RIPM process combines the benefits of an integrated

A risk-based integrated production model (RIPM) was systems analysis and quantitative risk analysis to guide
used to guide development of an offshore volatile-oil field strategic decision-making in the design, implementation,
and a retrograde-condensate-gas field. Strategic decisions and operation phases of reservoir management. Details of
regarding facility alternatives and well completion design the concepts and benefits of the RIPM process and tools1
were analyzed in light of uncertainties in reservoir size and and high-level application examples2 have been presented.
performance by use of a systems approach. Key parameters For this case, the integrated production model (IPM) tool
and business drivers were identified to focus the technical was key for quickly evaluating the effect of different sub-
work on important development decisions. The RIPM system options on the value delivered by the overall system.
enabled making decisions and evaluating a large number of A risk-based IPM (i.e., RIPM) tool was needed to quantify
options under a “fast-track” mandate in which relatively the effects of a number of uncertainties on the project’s
little data existed. The RIPM analysis was facilitated with overall value to facilitate strategic decision-making.
predictions from technical tools for reservoir performance The RIPM methodology was used throughout the asset’s
and completion design. Field results to date have validated development. For the purposes of this paper, discussions
this decision-making approach. will be directed at two phases of the project—the front-end
strategic design/decision phase and the development oper-
Introduction ation phase—in which initial data was acquired and ana-
This paper presents a case history in which the RIPM lyzed to validate the approach and decisions made.
approach1,2 provided the ability to make key early decisions
with relatively little data while evaluating a large number of Front-End Design and Strategic Decisions
options in light of reservoir uncertainties. The process con- Initially, the field was identified as amplitude events from
tinued through the entire project timeline from strategic 3D seismic analysis. Two primary structures exist on two
front-end infrastructure design, development execution, and offshore blocks separated by three miles. Reservoir depths
initial operation of a hydrocarbon asset. The asset is a recent- range from 10,000 to 13,000 ft. Initial reservoir pressure
ly discovered, relatively small offshore volatile-oil and retro- was roughly 10,200 psi, and formation temperature was
grade-condensate-gas field. The small reservoir size dictated 260°F. The seismic amplitude maps indicate three likely
that project economic success was crucially linked to the reservoir compartments on the western structure and one
proper integration and optimization of each subsystem on the eastern structure. Considerable uncertainty existed
(reservoir, wells, and facility) component. in delineating actual reservoir boundaries and thickness
With a traditional approach to development planning, from seismic, as well as extent of reservoir connectivity
the reservoir engineer would commonly make estimations within an amplitude event.
of the reserves in place and formation parameters (e.g., As a result of the desire to fast-track project develop-
pressure and permeability), then hand them off to the com- ment, many important, consequential decisions had to be
pletions engineer to size production tubing and determine made with very limited data. At the point that facility deci-
flow rates. Then, the facilities engineers would be given the sions were made, two successful exploration wells were
assumed production profiles to design the facilities. This drilled and tested on two stratigraphic accumulations. The
approach severely limits any ability to optimize the project fluid system was characterized as a volatile oil on the east-
interactively between the engineering disciplines or to ern structure and a retrograde condensate on the western
investigate the effect of changes in any one part of the sys- structure based on data from the exploration wells. Excel-
tem with corresponding effects elsewhere in the system lent productivity resulted from permeability on the order
(e.g., reserve size, tubing size, completion technique, or of 10 to 100 md. Good-quality seismic data in the area and
various facility options). nearby discoveries led to very high confidence in finding
hydrocarbons. The more difficult question concerned
whether reserves were sufficient to justify setting the facil-
Copyright 2003 Society of Petroleum Engineers
ities needed for development.
This is paper SPE 69085. Technology Today Series articles are general, descriptive repre- Economic sensitivity analysis was conducted in the
sentations that summarize the state of the art in an area of technology by describing recent
developments for readers who are not specialists in the topics discussed. Written by individu- scoping stage of the project. In the order of effect on the
als recognized as experts in the area, these articles provide key references to more definitive
work and present specific details only to illustrate the technology. Purpose: to inform the gen-
overall project’s net present value (NPV), key business dri-
eral readership of recent advances in various areas of petroleum engineering. vers were determined to be reservoir size, platform costs,

61

JANUARY 2003 61
 number of wells to efficiently develop the reservoirs, and hydrocarbon recovery. The gas/condensate material-bal-
well deliverability. During the front-end design phase, the ance model used Corey function exponents from laborato-
decision with the largest effect regarded the facility infra- ry core tests to describe relative permeability for the retro-
structure. For the western structure, which appeared to grade condensate. A numerical simulation tool with a
hold more reserves, the primary consideration was pro- compositional constituent model also was used to investi-
cessing capacity. For the smaller eastern accumulation, a gate condensate dropout behavior by matching it to con-
host of alternatives were considered, including selling the stant-volume-depletion laboratory measurements. Because
reserves, a standalone development, flowing the produc- of differences in the predictions from the modeling tools,
tion to a facility owned by another operator for a fee, and RIPM scenarios were run by use of both simulation and
installing a wellhead platform with production going to material-balance models to predict retrograde-condensate
the western platform for processing. performance. It was later determined that the material-bal-
Considering the uncertainty in many of the important ance model had a flaw in predicting gas/condensate behav-
variables and the number of infrastructure alternatives, a ior and, hence, its predictions were not technically sound.
risk-based integrated systems analysis appeared to be an Condensate-dropout predictions based on simulation
ideal methodology to fully investigate options under vari- analysis were included in the IPM by reducing the well
ous reservoir scenarios. Uncertainties were defined and productivity index by 50% below the dew point. The
entered for the different subcomponent models (i.e., reser- “base” case used in the economic evaluation consisted of
voirs, wells, and facilities). The reservoir models addressed four wells in gas/condensate reservoirs and two wells in
uncertainties of reservoir size, connectivity, and fluid volatile-oil reservoirs. Future recompletions were planned
behavior. The well models addressed the uncertainty of in each of the two oil wells.
well productivity and longevity in a traditional sand-pro- The final RIPM assumed a depletion-drive mechanism.
ducing region. The facility model addressed uncertainties This assumption was made because the reservoirs are
of processing capacity and processing alternatives. The stratigraphic and highly overpressured, and analogous
goal of the risk-based integrated systems analysis was to wells in the area appeared to exhibit depletion drive. A
facilitate decision-making on the design and operation of depletion-drive assumption was conservative in estimating
the system’s infrastructure (i.e., facility and wells) to prop- recovery for the oil reservoirs and potentially optimistic for
erly balance risks from uncertainty in reservoir geometry the gas/condensate reservoirs. The first two wells each
and performance. The quantitative approach allowed the penetrated a minimum of three gas-pay zones at shallower
risks associated with uncertainties in the different subsys- depths. These zones were not modeled in the production
tems to be compared on a common (NPV) basis for a host profile and economics, and they were used instead to com-
of scenarios. This provided flexibility in the design of the pensate for the possibility of an optimistic depletion drive
infrastructure to cover the range of “likely scenarios” gas production profile.
based on reservoir uncertainties.
Wellbore Model. A key driver of performance for the well
Reservoir Model. The reservoir models for the development subsystem was well productivity because it was affected by
were derived using data from cased-hole drillstem tests on completion design. Previous wells in analogous reservoirs
the two exploration discovery wells integrated with log and were completed with sand control. A well-failure model
seismic data to estimate reservoir size and continuity. indicated that sand control was not needed for the explo-
The first drillstem test provided reservoir fluid that was ration wells’ drillstem test (DST.) The wells were completed
analyzed for PVT behavior. The test also was used to inter- without sand control and were tested at high rates without
pret reservoir boundaries and permeability from a buildup sand. A more detailed core study was conducted to assess
analysis. The indicated boundaries then were used to tune the risk of sand production more accurately. The issue was
the seismic amplitude-extraction map to obtain amplitude critical because a nongravel-packed completion resulted in
intensity cutoff limits consistent with the well’s test data. lower initial costs and higher productivity, but higher failure
This seismic image provided the outline used to estimate risk (with potentially catastrophic results.) Conversely, a
reservoir extent. Log data was used to interpret a net pay sand control (frac-pack) completion was expected to result
thickness and to calculate water saturation and porosity. in higher upfront costs and lower initial productivity, but
Whole-core data from one of the appraisal wells were used lower long-term risk of failure over the well life.
to check the porosity and effective permeability calculated Laboratory tests were conducted on cores from an
from log analysis and the drillstem tests. Formation com- appraisal well to provide additional quantification of failure
pressibility was measured from a special core-analysis risk. Rock mechanical properties, hollow cylinder strength
study on portions of the whole core. The results were used measurements, and perforated-core flow experiments
to prepare material-balance tank models to predict pro- helped assess the likelihood of sand production. The tests
duction profiles. To mitigate the risk associated with limit- indicated that the wells would not fail at the initial targeted
ed reservoir connectivity, the RIPM was constructed with rates and drawdown pressures. The risk of failure was pre-
individual reservoirs (tank models) for each well. For the dicted to increase significantly after depletion.
base case, the ultimate per-well recovery was limited to Well models were constructed for both nongravel-
less than the average recovery for analogous wells nearby packed and frac-packed well completion designs. Histori-
in a similar depositional environment. cal data were used to estimate the differences in cost and
A difficult aspect in forecasting reservoir performance productivity of the two completion types for the perforat-
was behavior of the retrograde-condensate fluid. Of signif- ed interval(s) considered. Conventional nodal analysis
icance was whether condensate dropout in the reservoir examined the effects of tubing size, completion efficiency,
would limit well rates, condensate recovery, and ultimate and separator pressure.

62 JANUARY 2003
 tor, or paying a processing fee
2.0
for flowing back to the nearby
1.8
operator.
(Nongravel packed) The complete set of options
1.6 had a wide range of parame-
ters and variability. Numerous
Mean differential
constraints on the production
Relative NPV over 35 months

1.4
profiles further increased the
1.2 complexity of the decision.
The IPM allowed rapid evalu-
1.0
ation of all options. The pro-
No workover for analogous offset (gravel packed duction profiles generated
.8
were entered into economic
.6
models to quantify and evalu-
ate the associated NPV and
.4 risk on a consistent basis. The
standalone development
Non GP Oilwell
.2 option was selected on the
Analogous Offset
(gravel packed) basis of these evaluations
0 because it provided the best
0 5 10 15 20 25 30 35 40
Workover Time, month risk-based economics.
Typically, the model’s pro-
Fig. 1—Lifetime NPV for different failure scenarios in nongravel-packed vs. frac- duction predictions displayed
packed oil wells.
an initial total field rate equal
to the facility oil production
Several failure scenarios were examined to compare the capacity (i.e., it constrained the oil rate). Facility design
nongravel-packed well with the frac-packed well. Eco- engineers indicated that break-over points existed at
nomic analyses quantified the value created in the different 10,000, 15,000, and 30,000 BOPD, driven primarily by the
scenarios from forecasts of the production profile and asso- costs associated with power-generation requirements. Fur-
ciated costs. In the nongravel-packed well, failure costs thermore, gas capacity was fairly flexible and not as expen-
included risked estimates for failure consequences, the sive to upgrade as the oil capacity.
cost of a workover to recomplete the well, and lost pro-
duction. The analyses indicated that the nongravel-packed Results of IPM. This development is unique in at least two
well generated a higher risked NPV than the frac-packed respects. One is the high deliverability of the wells. The
well (Fig. 1.) In this graph, the time for failure of the non- discovery well was tested at approximately 1,000 BOPD
gravel-packed well is treated as a variable on the horizon- (actually condensate, as described below) with very low
tal axis. The plot shows that the NPV of the nongravel- drawdown. Preparations were made to test the first
packed well is higher if it produces without a failure for 4 appraisal well at much higher rates of 5,500 BOPD. The
months or more (i.e., even if a failure occurs after 4 models show that these wells are capable of initial rates in
months, the nongravel-packed completion was the correct excess of 8,000 BOPD. The second unique characteristic of
choice). By use of this unique approach, the decision was the system is the retrograde-condensate fluid system that
made to complete without sand control. exists in the reservoir. This fluid has an unusually high
The reservoir and well models then were combined to condensate yield on the order of 270 bbls/MMscf. The
generate production profiles for different scenarios in both result is an uncommon production profile, in which gas
the oil and gas/condensate reservoirs. Flow characteristics flow rates actually increase slightly (because of wellbore
measured during the drillstem test were keys for accuracy. hydraulics) for a period of time when the dewpoint is
In particular, initial welltest data points (consisting of rates reached in the reservoir until returning to a natural
and flowing tubing and bottomhole pressures) were used decline. It is unlikely that the larger oil or gas rates would
to match inflow and vertical-flow performance, and were have been included in a less rigorous, estimated approach.
validated through analyses of DST and log data. Tubing lift A matrix of more than 40 possible scenarios was run on
curves and inflow performance relationship curves then the IPM, varying reservoir size, number of wells and tim-
were calculated over the range of conditions expected in ing, and facility capacity. Production profiles were generat-
the producing life of the reservoir. ed for each scenario and entered into an economic model.
The field produced on plateau at the initial oil rate (and
Facilities Model. Preliminary volumetric calculations of corresponding gas rate) for a time period until the oil rate
hydrocarbons in place in the separate stratigraphic accu- fell below capacity and began declining. At this point,
mulations showed that neither structure was large enough because the maximum oil rate is no longer a limiting fac-
on its own to support setting a complete processing facili- tor, the gas rate increased to either the maximum deliver-
ty. Hence, a full range of development options was evalu- ability or facility capacity. A peak gas rate occurred; there-
ated to determine which option provided the best return. after, the gas rate declined in typical depletion fashion.
The options included selling the reserves, setting facilities The base-case scenario, when run against a 15,000
for a standalone development, unitizing and flowing pro- BOPD constraint, stayed on plateau for more than 1 year.
duction back to a nearby facility owned by another opera- The wells were capable of delivering more than this rate,

63
JANUARY 2003 63
 collectively, but a scheme was developed to stage the wells motivation was to determine early in the reservoir life
over time to maintain the plateau for this period. A whether additional wells were needed to drain the ampli-
30,000 BOPD capacity could conceivably be met if the tude. The gauges also were desired for surveillance of the
wells were drilled successively; however, this rate was flowing bottomhole pressure to optimize production at a
maintainable for a relatively short period of time before desired drawdown. This optimized production was espe-
falling on decline. For the base case, production had essen- cially important to mitigate sand-production risk.
tially reached the economic limit after 5 to 6 years. This After 3 weeks of production, the reservoir was thought
fairly rapid depletion was the result of the high well deliv- to be in pseudosteady-state flow. During the third week, no
erability and relatively limited reservoir size. In essence, changes in choke settings were made, and the daily reduc-
the short field life under either scenario does not provide tion in flowing-bottomhole pressure was recorded. The
great incentive to accelerate production by investing in average drop in flowing-bottomhole pressure was
larger capacity at higher cost. 8.5 psi/day in both wells. A material-balance calculation
Substantial consideration was given to larger reserve on the combined production resulted in an original-oil-in-
sizes. If reserves were actually larger than originally esti- place value within 3% of the volumetrically calculated
mated, the production profile would be extended and per- value, validating the reservoir connectivity as modeled in
haps a larger facility would be desirable. However, for this the IPM.
exploration-type prospect, this would be known only after
some production history had been obtained. To account Oil Reservoir Results. The decision to complete the oil
for this risk, some flexibility (at minor cost) was included wells without sand control was validated by comparison
in the facility design to allow for a future capacity upgrade with an analogous well that was completed with a frac-
to 30,000 BOPD. This upgrade feature provides an oppor- pack in the same sand in an adjoining block. One of the oil
tunity to capitalize on potential reserves upsides while lim- wells, completed without sand control, produced at an ini-
iting initial investment to only the amount required with tial DST rate of 5,545 BOPD at a flowing tubing pressure
fair certainty. There is a fair probability that the actual gas of 5,200 psi and a drawdown of less than 200 psi. The
deliverability would exceed the base-case gas capacity. In drawdown in this well was monitored daily and remained
this case, the incremental investment to increase gas at 530 psi at a sustained rate greater than 7,000 BOPD for
capacity to 150 MMscf/D is easily offset by the additional 15 months. The other oil well logged an unexpected water
revenue that could be gained. Some runs with the best- contact and started producing at a high water cut 2 weeks
case scenario for reserves, wells, and timing reflected gas after initial production. The productivity and drawdown
rates as high as 200 MMscf/D. on this well were almost identical to the first well. The per-
manent bottomhole pressure gauges allowed validation of
Risk/Monte Carlo Economic Model. The results of the the well models, showing them to be accurate to within a
IPM runs were entered into a stochastic (probabilistic) eco- few percent. The frac-packed well on an offset lease had an
nomic model. The cost distributions for the different facili- initial DST rate of 5,100 BOPD at a flowing tubing pressure
ty scenarios were also entered, and an NPV distribution was of 5,100 psi and a drawdown of 600 psi. The well pro-
calculated (i.e., incorporating uncertainty because of reser- duced at a peak average rate of 4,800 BOPD for 8 months.
voir size, connectivity, and well performance) for each facil- The drawdown increased to 1,500 psi, presumably because
ity option. The results of the stochastic economic modeling of plugging in the frac-pack. The rate subsequently
showed that the NPV of the different scenarios ranged from dropped to 800 BOPD over a period of 4 weeks. Three
less than zero to four times that for the most likely scenario. stimulation treatments were required to return the well to
This provided management with a clear perspective of the a sustainable rate of 3,000 BOPD.
potential economic outcomes for each facility option and Actual production rates from the nongravel-packed and
the effects of uncertainty in key parameters. The develop- frac-packed wells are shown in Fig. 2. The nongravel-packed
ment option selected on the basis of a risk/reward analysis well produced approximately 1 million BO more than the
consisted of a main processing platform with six slots and frac-packed well over the first 10 months of production.
a satellite platform. The processing facility was chosen with Fig. 3 shows the average daily oil rates during the first
capacities of 15,000 BOPD, 10,000 BWPD, and 20 months of production, compared with predictions from
150 Mscf/D, with additional space to increase capacity to material balance with depletion drive and a compositional
capture upside opportunity at a minimal cost (resulting in simulation assuming waterdrive (aquifer size approxi-
a potential four-fold increase in NPV.) mately equal to oil reservoir size.) Actual performance was
The oil reservoirs were determined to have the longest better than that predicted because of a larger aquifer.
producing life; therefore, the decision was made to drill the Cumulative oil recovery to date has exceeded projections.
oil wells first and complete them without sand control to The good oil producer has maintained its peak rate longer
have them online at startup. A stand-alone development than predicted in the RIPM, and the recovery factor likely
was determined to have the highest NPV of the full range will be higher than the 27% assumed in the material-bal-
of options considered. A staged development was planned ance model.
to drill three additional wells after setting the facilities. In hindsight, the completion design was a critical factor
Startup production came from the first three wells drilled on the project economics. Taking greater risk to complete
during the exploration and appraisal phase. the wells without sand control allowed higher completion
efficiency and well productivity. The additional deliverabil-
Development Operation Phase ity was required when one of the oil wells began producing
The two wells in the largest reservoir were equipped with with a high water cut. This history helps validate the risk-
permanent bottomhole pressure gauges. The primary based approach by showing the interdependence of uncer-

64 JANUARY 2003
 sure at least as low as 7,000 psi.
10000
This well was worked over, 2500

9000
frac-packed, and brought back
online at a slightly lower rate of

Cumulative Oil Production, thousand BO

8000 2000
16 to 18 MMscf/D. It is felt that
7000
a frac-pack will help recover
gas reserves in a low resistivity
(laminated) portion of the
Oil Rate, BOPD

6000 1500

5000
reservoir. Another factor com-
plicating well-performance
4000 1000
analysis is that the bottom 27 ft
3000
of the 71 ft of perforations were
blocked by the perforating
2000 500
guns, which failed to drop dur-
1000
ing the initial completion.
Additional gas wells in the
0 0 development also have not per-
100

118

136

154

172

190

208

226

244

262

280

298
10

28

46

64

82

109

127

145

163

181

199

217

235

253

271

289

307
19

37

55

73

91
1

Day
formed as well as predicted.
Although the seismic ampli-
Gravel Packed Well BOPD Non Gravel Packed Well BOPD Gravel Packed Well Oil Cum Non Gravel Packed Well Oil Cum
tudes associated with the pres-
ence of gas were considered
Fig. 2—Actual oil rates in gravel-packed vs. nonpacked oil wells.
continuous, performance has
indicated that they are not con-
tainties in reservoir geometry and well performance and, nected hydraulically. Although they were drilled and pro-
hence, the need to properly balance the associated risks. duced at different times, the four gas wells each exhibited
original reservoir pressure, indicating pressure isolation. A
Gas Reservoir Results. Initial concerns of reservoir con- lack of reservoir continuity is believed to be the reason for
nectivity were supported by a rapid decline in reservoir lower-than-expected cumulative recoveries. Also, the rapid
pressure of the first producing gas well. The well produced decline in reservoir pressures is attributed to lower-than-
at a sustained rate of 18 to 20 MMscf/D from an initial predicted original gas-in-place volumes.
reservoir pressure of 10,200 psi, down to approximately Fig. 4 shows cumulative condensate production vs.
6,500 psi, with an estimated drawdown of 1,000 psi. At this cumulative gas production for the first gas well drilled com-
time, the well began producing large amounts of sand and pared with a prediction from the compositional simulation
the rate was reduced. Several unsuccessful attempts were model. The results show that the simulation prediction was
made to attain a sand-free rate greater than 2 MMscf/D. The accurate during early production. Later predictions were off
failure analysis had predicted that a nongravel-packed com- by more than 100% because of assumptions that conden-
pletion would remain sand-free down to a reservoir pres- sate would not flow after dewpoint was reached.

8000

7000

Actual with Aquifer

6000

5000
Oil Rate, BOPD

Compositional Forecast with Aquifer

4000

3000
Material Balance Forecast
(no Aquifer)
2000

1000

0
11

13

15

17

19

21

23

25

27

29

31

33

35

37

39

41

43

45

47

49

51
1

Months

Material Balance Forecast Compositional Simulation Forecast ACTUAL AVERAGE (Including Downtime)

Fig. 3—Single well material balance vs. simulation vs. actual oil rates.

JANUARY 2003 65
 3000

Cumulative Gas Production, MMscf

2500
ACTUAL

2000

1500

1000 SIMULATION

500

0
0 50000 100000 150000 200000 250000 300000
Cumulative Condensate Production, bbl

Fig. 4—Single gas well compositional simulation vs. actual rates.

The facility sizing decision was validated because oil select the highest NPV alternative. These options were eval-
rates for processing were 9,000 to 14,500 BOPD for the uated in a timely fashion with minimal data and manpow-
first 24 months of operation. Oil processing capacity of er. A team of three to four people thoroughly evaluated the
10,000 BOPD would have constrained production, and development options and economics in approximately 6
capacity above 15,000 BOPD would have been excessive. months. The development option and well performance
Overall, investment payout on a net cash basis was evaluated with RIPM was validated through early field
achieved in approximately 2 years. The project NPV esti- performance and independent reservoir-simulation pro-
mate has been met, even with delays in development jects completed before and just after field startup. JPT
drilling. The poorer-than-predicted gas reservoir perform-
ance was offset by better-than-predicted oil reservoir References
performance and higher-than-predicted commodity 1. Chow, C.V. and Arnondin, M.C.: “Managing Risks Using
prices. Gas volume projections have not been met, but oil Integrated Production Models: Process Description,”
volume projections have been exceeded. The decision to JPT (March 2000) 54.
complete the oil wells without sand control, combined 2. Chow, C.V. et al.: “Managing Risks Using Integrated Pro-
with excellent oil reservoir performance, are the largest duction Models: Applications,” JPT (April 2000) 94.
factors in the commercial success of the development,
despite poorer-than-predicted gas reservoir performance.
The risk-based integrated systems analysis facilitated Chris V. Chow, SPE, is President of ProDuce Consulting. He
management’s approval for funding this marginal (i.e., has 13 years research and consulting experience with Mobil
small reservoir size and large investment required) fast- Oil Co. specializing in sand management, perforating,
tracked development with minimal data. It is speculated completion efficiency, integrated modeling, and risk and
that, under a traditional approach for development design, design analysis. Chow holds a PhD degree in chemical
the project would not have been funded because of a lack engineering from the California Inst. of Technology. Jim C.
of quantification and/or communication of subsystem Mothershead, SPE, is a petroleum engineer in the Inter-
uncertainties, an imbalance of associated subsystem risk, national Division of DeGolyer and MacNaughton, where he
and time delays because of a sequential design approach. performs reservoir studies and prepares reserve estimates
for client companies. Previously, he was a reservoir engineer
Summary with ExxonMobil. Mothershead holds an MS degree in
A risk-based integrated production model proved to be a petroleum engineering from Louisiana State U. and an MBA
useful tool in the planning and operation stages of an off- degree from Southern Methodist U. David L. Cox, SPE, is
shore field development just after an exploration discovery. an independent oil and gas producer. Before forming New
The model was used for decisions on facility sizing, capital Territory Petroleum, he held various engineering and man-
expenditure timing, and well-completion design. The full agerial positions with Mobil Oil Co. His experience is in the
range of monetization options, including selling reserves, Gulf of Mexico, Permian Basin of west Texas, and the Rocky
joint development, third-party processing, and standalone Mountains. Cox holds a BS degree in petroleum engineer-
development, were analyzed quickly and accurately to ing from the Colorado School of Mines.

66 JANUARY 2003