Gulf of Suez Petroleum Company

GUPCO

October Field PROSPER & GAP Model instructions

Prepared by:

Kareem Basha
Challenge PE – North team

July 2014
Table of contents:

1. Introduction …………………………………………………………………………………… 2
2. PROSPER Modelling ………………………………………………………………………….. 2
2.1. Using QuickLook tool to estimate POI ……………………………………………………. 5
2.2. Important notes while using PROSPER …………………………………………………... 7
2.3. Model uncertainties …………………………………………………………………………... 7
2.4. PROSPER outcomes ………………………………………………………………………….. 8
3. GAP Modelling ………………………………………………………………………………… 9
3.1. Matching GAP Model ………………………………………………………………………… 11
3.2. GAP Model Optimization ……………………………………………………………………. 12
3.3. Copying the model …………………………………………………………………………… 16

October PROSPER & GAP Model instructions 2

 1. Introduction

Project: October field production optimization using PROSPER & GAP Models.

Start date: March -14

Completion date: Jun – 14

Team members:

 Mohamed Atteya (SPA for GoS Gas-lift optimization)

 Kareem Basha (Tag for October field GL optimization)
 Mohamed Okasha
 Fady Farag
 Aly Elbassyoni
 Mohamed Said Ibrahim
 Shakeel Ahmad
 Mohamed Sayed Abdellatif

Project phases:

1. PROSPER Modelling.
2. GAP Modelling.

2. PROSPER Modelling

October team worked together to do the modelling for 42 open wells in October, there were
5 groups with 5 SPA’s working on updating the existing PROSPER models. Modelling
procedures for each well were as follows:
a) Reviewing well history (E-well book) to best understand your well performance
history.
b) Updating old PROSPER model, here. In case of new wells or Sidetracks, a new
model is constructed.
c) For the old/ new PROSPER model, do the following to update and match the model:

Check producing zone from Bar chart and choose the suitable PVT
available here where you can find the PVT database in excel formats or you
can use the (.txt) format to import directly to PROSPER, here.
 Importing method: from PROSPER do the following:

PVT DATA > Match Data > Import > Import file (select the file) > Continue > File
format (Tab Delimited) > Continue > Define the columns > Continue > Done.

 If there is a well head sample from a well measured oil gravity API, it
is better to update the PVT with the actual API as it highly affects the
model.
 Select suitable PVT correlations that best matches your parameters.

 Deviation survey, deviation surveys are available here. If you can’t

find a deviation survey, ask the drilling engineer or geologist to send
you the survey of the subject well as they have a complete set of
surveys for each well.
 Do not model surface equipment; keep it unchanged unless you
have production choke which is not the case in October field.
 Downhole equipment: review the current existing wellbore sketch
and best describe it in this section. The last casing point should be at
mid-perfs (THF).

October PROSPER & GAP Model instructions 3

  Geothermal gradient: enter the temperature profile; two points are
required; at the surface and at mid-perfs.
 Average heat capacities, use default values.

 Enter Gas-lift valves depths from existing wellbore sketch.

 Enter gas-lift details. (Injection depth, Gas-lift rate, Gas-lift gas gravity
assumed 0.75)

 To construct IPR and VLP curves, try to find the latest point at which
there is sufficient data (Well test / PLT, GLS, and S.S), this point
should be after any workover, WSO, Add-perf or Re-perf that could
change well productivity.
 Well test data, here. PLT data here. Or from EC web here.
 GLS data, here.
 Reservoir pressure data, here, Reservoir pressure should be
converted to mid-perf depth (TVD THF).

 In case there is only PLT or well test available at some date and
no GLS or S.S at the same date:

 Use well test data from the PLT; double check WC from WC
samples.
 Try to find reservoir pressure from an offset well at that
date, ask RE’s for advice to best estimate Pr. Compare this
pressure with the pressure from the shut-in pass in the PLT,
they should be reasonably close.
 Find well head parameter (WHP, WHT, Gas-lift injection rate,
downstream pressure) during PLT from the analysis report or
from the automation at the PLT date, automation sheet is
available here.
 Make sensitivity on the POI (Point of injection) using
correlation comparison. (Use Petroleum Experts 2 if
deviation <60° and Petroleum Experts 4 if deviation > 60°),
double check using QuickLook tool by setting the POI and
orifice diameter and try to get best match.
 Use all these data to construct system IPR/ VLP and adjust
productivity index (PI) value, QC the adjusted value from
offset wells or the range of PI in the field, if accepted, OK, if
not, then re-visit your input data.
 Update the model from sorting list, update WC and Gas-lift
injection rate, update reservoir pressure and then re-adjust PI
value, this step to match the model to the sorting list, our
baseline.

 In case there is no any data available after the last intervention

in the well we use the following data:
 Production data from virtual test (from sorting list).
 Well head parameters (WHP, Gas-lift, Casing pressure) from
automation system.
 Reservoir pressure from an offset converted to mid-perf
depth (TVD THF).
 Use QuickLook tool to best estimate the POI based on the
data available. (Match on casing pressure)
 Use the data above with the POI to adjust PI and construct
IPR/VLP.

October PROSPER & GAP Model instructions 4

 2.1. Using QuickLook tool to estimate the POI:

 Enter surface measurements

 Casing head pressure is D/S pressure
 From EC (test data)
 Form Automation (instantaneous)
 Enter gas injection data
 Orifice diameter (from GLD sheet)
 Default design: 1/4 , 5/16, 3/8 …., 1/2.
 Vertical flow correlation: Petroleum Experts 2 (Petroleum Experts 4 for highly
deviated wells > 60°)
 Calculate and plot.
 Make sensitivity on injection depth till you get match (see figures below).

The following figures are showing an example of using QuickLook tool to estimate gas-lift
injection depth:

Click once finished

Matching parameter,
Adjust orifice diameter downstream pressure
to model valve leaking
(bigger orifice) or scale
Make sensitivity on POI,
(smaller orifice)
start from the deepest
valve
Make sure you are
using the right
correlation

October PROSPER & GAP Model instructions 5

 Case 1 Case 2
POI V#6 @ 9132 POI V#5 @ 8544

Injection not possible at

Injection not possible at
this valve
this valve

Case 3 Case 4
POI V#4 @ 7945’ THF POI V#3 @ 7100’ THF

POI V# 4 @ 7945’ THF Best representative case

Orifice diameter = 50!!
Multiple POI?!
No dip across valve which is not logic

POI V# 3 @ 7100’ THF

Orifice diameter = 33!!
Multiple POI?!

October PROSPER & GAP Model instructions 6

2.2. Important notes while using PROSPER:

 Make sure you keep a record for each input in PROSPER in the PROSPER Data
sheet here , to make it easier for subsequent users to use the model and understand
where the numbers are coming from. It is better to do the documentation from the
beginning while the data is fresh in your mind and not leave it to the last moment.
 Reference depth for casing point, reservoir pressure, flowing pressure, reservoir
temperature is at mid-perfs. To convert the pressures into mid-pers use the Depth
converter here, or can be done in the PROSPER Data Sheet as well in the same
folder.
 Make sure the reservoir pressure and water cut values are the same in the IPR and
VLP sections.
 Make sure the depth of injection and gas-lift gas rate values are the same in the gas
lift data and VLP sections.
 Before constructing the IPR/VLP, make sure to estimate U value (Heat transfer
coefficient) first and transfer it to your model if reasonable. It should be something
around 8 BTU/h/ft2/F for oil wells and 3 BTU/h/ft2/F for gas wells.
 Before adjusting the productivity index (PI), or using QuickLook tool, make sure you
are using the proper vertical correlation (for example: Petroleum experts 2 not
petroleum experts 4 )
 If there are multiple tests in the VLP section, make sure you select the test of
interest while adjusting the productivity index.
 Make sure that the Rate method under System (3 variables) screen is automatic –
Geometric, this give more accurate results.
 If you have done any changes to an existing model, make sure to write your name in
the summary screen with the date of this change, and a brief description of that
change. You can access this screen by double clicking on OPTIONS SUMMARY.

2.3. Model uncertainties:

The general rule says that model results will be as good as the input data, so the better
quality of the input data, the most accurate and more reliable the model. Main challenge in
October field or in GoS was the lack of data and data quality; accordingly, many assumptions
have been done to complete the model. There are many assumptions made while doing the
PROSPER modelling for October field as follows:

1. PVT data are accurate and representative for each well assuming the reservoir fluid
properties are the same across the entire reservoir.
2. In wells producing from two different zones, the well is modelled as being producing
from the zone with the most contribution. For example, a well that is producing from
Nubia and Raha is modelled as Nubia producer.
3. Estimated reservoir pressures from offset wells are assumed to be accurate
assuming there is good reservoir communication between offsets.
4. Gas-lift injection rate measurements are assumed to be accurate as this is was the
only source to calculate the gas-lift injection rate, it worth highlighting that there are
some uncertainties in the gas-lift injection rates as per the gas balance study
performed in Feb. 2014 and showed around 20% losses from automation sensors on
wells.
5. In case there is no GLS, the injection depth estimated from QuickLook tool is
assumed to be the right POI.
6. For wells with old well tests that still have the same BFPD in the sorting list with
only updated WC and Gas-lift injection rate, it is assumed that these are still
producing the same rate as there is no recent well tests.
7. For wells where there are no well test data available, virtual test based on well head
parameters is used to match the model.

October PROSPER & GAP Model instructions 7

 2.4. PROSPER outcomes:

PROSPER modelling showed a bunch of opportunities either directly from the modelling itself or
indirectly while reviewing wells history where some other opportunities have been identified such as
re-perforating some intervals in some wells and checking chemical injection systems in other wells.
Total estimated gain from these opportunities is ± 1600 BOPD and can be summarized in the
following table.

Estimated gain,
P/F Well Recommendations
BOPD

POI V#7 , Partial plugging reflected in orifice 20.5/64" instead of 24/64".

OCT_C-2B 100
GLS, based on which GLVC may be performed to go one deeper valve (V#8, RDO)

C
POI V#8.
OCT_C-3 150
GLS, based on which GLVC may be performed to go one deeper valve (V#9, RDO)

Well matched to IGOR data. POI V#3, well authorized for GLS, Based on PROSPER,
OCT_C-6S T1 500
V#7 may be changed to RDO.
OCT_D-8 Well test.
D POI from RDO, Orifice 19/64" (Severe plugging) well has scale tendency from
OCT_D-9B
history, evaluate tbg cleaning.
Well has communication … there is leak in date after WSO 2009 so we neglect the
OCT_G-7S T1
quick look match as it is not applicable. Recommend to perform GLS.
W/O due to severe T/A communication (in plan with Comet Rig).
OCT_G-11S T1
GLS as a baseline & optimization test after the W/O
Check efficiency of the current asphaltene inj. + update WHS analysis. Stop
G OCT_G-12 chemical injection, collect sample, send it to lap & Nalco for analysis.
Run GLS once it is possible
• GLS for possible GLVC (Last WxW outcome dated June-2014)
• PLT (Last WxW outcome dated June-2014)
OCT_G-13 60
• As per the model results, GLVC for V#6 (now RDO) with new active valve will
add +/- 60 BOPD
H OCT_H-9 POI is V#3, GLS for possible GLVC to inject from a deeper valve V#6 RDO) 250
PLT log (Nov. 13) showing high GR in top of the upper perfs. Indicating scal
OCT_J-2
accumulation, Acid stimulation may be recommended.
POI @ 8928’ V#5 , with orifice size 27, indicating erosion, or valve leakage.
Well has TA communication.
OCT_J-4B Well has an obstruction @ 9865 RKB (3 valves below) 50
GLS is recommended, based on which GLVC may be recommended , V#6 to be
converted to RDO (Maximum depth of injection).
Test the well (New well)
OCT_J-6C S T4
GLS to check GLS (POI)
POI V#6.
GLS for possible GLVC to go one deeper valve (RDO V#7).
OCT_J-9A 70
Well has scale tendency, Tubing cleaning, Perf. Acid wash may be considered to
improve well performance.
OCT_J-11S T1 Need to be tested. (WO Jun. 13)
• Change gas lift leaking valve #6 (RDO) GLS Aug-2006 with normal valve
OCT_K-1 • Reperf exsiting interval to improve PI 200
K
• Expected gain 200 BOPD
OCT_K-4 Well Authorized for WO
POI V#5.
ET_A-1 50
GLS, Based on it , GLVC and change V#6 to RDO. (Maximum depth of injection)
Flowing survey, evaluate Pwf, and PI, based on PI Acid stimulation job may be
ET_A-2
recommended, the well productivity declined by 50% in 2 years.
POI V#6.
NO159-5A 100
ET-A GLS for possible GLVC to change V#8 to RDO. (Maximum depth of injection)
POI V#7.
NO159-6S T1 70
GLS for possible GLVC to inject from deeper valve (V#8, RDO)

Acid job is recommended, well PI decreased from 12 to 3 bbl/day/psi in 4 years.

GS 160-6S T1
GLS recommended after Acid job to evaluate Pwf.

October PROSPER & GAP Model instructions 8

3. GAP Modelling

The next step in this project was to build a GAP model where all system data can be integrated
together to be able to optimize the whole system; the model was built into steps as follows:

1. Model skeleton: the model structure was first built including all wells (Open and shut-in
wells), production pipelines between platforms, all connected to the separator at October
complex. Flow lines were not modelled for the sake of simplicity as long as they are very
short joints that will not add significant pressure drop.
2. Pipelines data: after building the structure of the model , next step was to start loading the
data into the model, starting with the pipelines data;
a. Upstream and downstream of the pipe.
b. Size of the pipe.
c. Length.
d. Pipeline profile (elevation).
October field pipeline network data sheet can be found here.
3. Wells data: after loading pipelines data, next step was to load wells data from PROSPER,
this step is subdivided into two steps:

a. IPR generation: used to transfer the IPRs from the PROSPER files to the GAP well
models, the result is that the IPR is imported in the well IPR section as shown
below:

GAP uses its current type of IPR model for the well regardless of the IPR model used in the
generation by PROSPER. GAP takes the reservoir pressure and PVT information from PROSPER
along with three IPR data points. These points become the Match points in the GAP model, and
GAP fits its IPR coefficients to these points.

October PROSPER & GAP Model instructions 9

b. VLP generation: This section describes the process of batch generating VLPs in
PROSPER and transferring the information to GAP.
To generate the VLPs click " "

Check the “ ” option and Click the button. Once all the well files
have been read the screen shown below is presented.

To check and edit the values, the buttons shown above can be used to fill in
the table with numbers.

October PROSPER & GAP Model instructions 10

 The values used for generating the VLP curves are dependent upon the conditions of
the field. The objective is to generate the VLP curves with a range of values that
encompass the conditions that will be experienced in the well/field in future.

The reason for using a high range of values is so that GAP will always interpolate
between the values entered. If the range of data points entered is sparse, then GAP
will have to extrapolate on the values which may give incorrect results.

The following values can be used as possible intervals to enter the data for the VLP
curves. These values are only provided as guidance:

No. of Distribution
From To Used Parameters
points of points
Low value
AOF (Absolute Geometric
Liquid rate (Say 100 20 100 – 10000 BFPD
open flow) of well spacing
stb/d)
Separator
Top Node Linear 50 – 500 Psi (Separator pressure
pressure Reservoir pressure 10
pressure spacing 67 psi)
value
Linear
Water cut 0% 100 % 10 0 – 100 %
spacing
A value
less than Geometric 250 – 5000 SCF/STB
GOR 25000 scf/stb 10
solution spacing
GOR
GLR Linear
0 25600 scf/stb 10 0 – 5000 SCF/STB
injected spacing
Maximum value
that could be
Gas lift Linear
0 injected in the well 10 0 – 6 MMSCFD
Gas Rate spacing
(for example 6
MMSCF/D)

3.1. Matching GAP Model

After loading all wells and pipelines data, generating IPR and VLP curves, run the model to check
the results. Check data at each node and compare it to the actual data, you have to check the
following:

 Each well data (BFPD, BOPD, IG, WHP).

 Each riser pressure & compare it to actual pressure, should be close enough.
 Flow directions in the pipelines.
 Vertical correlation used to construct the VLP.

When everything is ok, the model should give you good results close enough to the actual data.
Our model showed very good matching.
Note: There were some troubles with pipelines from H platform to the complex, it sometimes
showed back flow, i.e. flow into wells, to overcome this point, the pipelines from H to Complex
were bypassed given that it is only a short joint ( ≈200 ft.) and there is no significant pressure
drop on it.

October PROSPER & GAP Model instructions 11

In the following figure we can see oil rate matching results, showing very small error in most of the
wells (< 10 %), and the overall error was about 1.5% which is accepted.

Oil Rate matching results

Sorting BOPD GAP BOPD Diff. %
2000 100%

1800 80%

1600 60%

1400 40%

% Difference
1200 20%
BOPD

1000 0%

800 -20%

600 -40%

400 -60%

200 -80%

0 -100%

OCT_H-3

OCT_H-7

OCT_H-9
OCT_A-2B
OCT_A-4A

OCT_B-7

OCT_C-6ST1
OCT_C-8ST1

OCT_J-1A

OCT_J-4B

OCT_J-9A

OCT_K-3A
OCT_F-3C

OCT_G-7ST1
ET_A-1
ET_A-2

OCT_G-11ST1

OCT_G-5

OCT_H-4A
NO159-5A

OCT_A-1ST1

OCT_A-8ST1

OCT_C-3

OCT_D-8

OCT_H-1AST1

OCT_H-8AST1
OCT_A-6
OCT_A-7

OCT_E-1B

OCT_G-12
OCT_G-13

OCT_J-11ST1

OCT_J-2

OCT_J-6C ST4

OCT_K-1

OCT_K-4
OCT_K-5
OCT_C-2B

OCT_D-9B
GS160-6ST1

NO159-6ST1

OCT_C-10

3.2. GAP Model Optimization

Accepting GAP model matching results, five optimization scenarios have been assumed.
Optimization method in all scenarios was Oil rate only as the objective is to maximize the oil
production.

October PROSPER & GAP Model instructions 12

Different optimization scenarios can be described as follows:

a) First scenario: Given the available amount of gas (Total IG) according to the sorting list
(117.2 MMSCFD), optimize the gas-lift injection rate in all wells by giving the GAP a range of
values of IG in each well, from 0 to 6 MMSCFD and let the model redistribute the same
amount of gas-lift over all wells in a way that will maximize the oil production.

No. of optimized wells 42

Optimization Oil gain 773 BOPD
Saved Gas-lift 0 MMSCFD
Net oil gain 0 BOPD

b) Second Scenario: Reviewing the results from the 1st scenario, selected wells with high
value changes, either gas saving or oil gain, to limit the optimization to some wells rather
than all wells, using new IG rates from 1st scenario in the selected wells and keeping IG rates
in other wells as original case and running the model with fixed IG rate in all wells.

No. of optimized wells 22

Optimization oil gain 411 BOPD
Saved Gas-lift 11 MMSCFD
Net oil gain 800 BOPD

Optimization summary

Actual IG New IG Gain/Loss Saved Gas

Well
MMSCFD MMSCFD BOPD (MMSCFD)
ET_A-1 3.0 2.3 -13.8 0.7
GS 160-6S T1 0.8 3.4 363.5 -2.6
NO159-5A 3.5 1.5 -15.1 2.0
OCT_A-4A 3.5 2.5 -8.7 1.0
OCT_A-6 1.5 2.2 31.8 -0.7
OCT_B-7 3.2 2.0 -21.5 1.2
OCT_C-6S T1 1.5 3.1 123.7 -1.6
OCT_C-8S T1 3.0 2.6 -4.8 0.4
OCT_D-8 2.3 1.8 -3.4 0.5
OCT_G-12 2.5 2.1 -3.9 0.4
OCT_G-5 3.0 4.4 62.0 -1.4
OCT_G-7S T1 5.6 2.6 -33.7 3.0
OCT_H-3 4.0 3.2 -10.4 0.8
OCT_H-4A 5.0 3.9 -9.2 1.1
OCT_H-7 4.0 2.2 -18.4 1.8
OCT_H-8AS T1 3.0 2.1 -5.7 0.9
OCT_J-2 2.0 4.0 73.9 -2.0
OCT_J-4B 4.0 3.1 -16.5 0.9
OCT_J-9A 3.0 2.6 -4.6 0.4
OCT_K-1 3.0 2.2 -17.0 0.8
OCT_K-3A 3.0 1.4 -9.4 1.6
OCT_K-5 4.5 2.7 -24.0 1.8
Total 68.9 57.9 434.7 11.0
S aved Gas 11.0 MMSCFD
Oil gain 434.7 BOPD

October PROSPER & GAP Model instructions 13

c) Third Scenario: Running the original matched model with only changing the POI (Point of
injection) in 8 wells as per PROSPER recommendations to maximize oil production, it was
necessary to generate new VLP curves for the subject wells in this scenario.

No. of optimized wells 8

Optimization Oil gain 1465 BOPD
Saved Gas-lift 0 MMSCFD
Net oil gain 0 BOPD

Optimization summary
GAP GAP
Sorting Gain/Loss
Well Actual Optimized
actual BOPD BOPD
BOPD BOPD
ET_A-1 826 864.271 965.81819 102
NO159-5A 279 290.782 353.82489 63
NO159-6S T1 562 587.686 641.23533 54
OCT_C-2B 750 761.935 861.48363 100
OCT_C-3 1680 1699.225 1824.4305 125
OCT_C-6S T1 352 375.773 1081.0041 705
OCT_H-9 781 795.626 1038.5354 243
OCT_J-9A 459 487.609 561.91929 74
Net oil gain (BOPD) 1420

d) Fourth Scenario: Repeating the 1st scenario, redistributing the gas-lift injection gas on all
wells, but this time with the new generated VLP curves for the 8 wells with new POI.

No. of optimized wells 42

Optimization Oil gain 2376 BOPD
Saved Gas-lift 0 MMSCFD
Net oil gain 0 BOPD

e) Fifth Scenario: Repeating the 2nd scenario, by changing the IG rate in wells of good value
along with wells with new POI and keeping the IG rates constant in the other wells as
original case.

No. of optimized wells 26

Optimization Oil gain 1977 BOPD
Saved Gas-lift 7 MMSCFD
Net oil gain 2200 BOPD

October PROSPER & GAP Model instructions 14

 Optimization summary
Old IG New IG Gain/loss
Well Notes
MMSCFD MMSCFD BOPD
ET_A-1 3.0 3.00 136 POI V#6 instead of V#5
GS 160-6S T1 0.8 3.40 287
NO159-5A 3.5 1.98 64 POI V#8 instead of V#6
NO159-6S T1 2.0 2.00 77 POI V#8 istead of V#7
OCT_A-4A 3.5 2.41 -19
OCT_A-8S T1 2.0 1.64 -10
OCT_B-7 3.2 2.00 -23
OCT_C-2B 2.4 2.40 98 POI V#8 instead of V#7
OCT_C-3 3.8 3.80 123 POI V#9 instead of V#8
OCT_C-6S T1 1.5 3.58 900 POI V#6 instead of V#3
OCT_D-8 2.3 1.75 -21
OCT_D-9B 2.3 1.53 -18
OCT_E-1B 1.9 2.87 -17
OCT_G-12 2.5 1.99 -19
OCT_G-5 3.0 4.62 44
OCT_H-3 4.0 2.88 -14
OCT_H-4A 5.0 4.04 53
OCT_H-7 4.0 2.12 -25
OCT_H-8AS T1 3.0 1.92 -6
OCT_H-9 2.8 2.80 258 POI V#6 instead of V#3
OCT_J-2 2.0 4.05 77
OCT_J-4B 4.0 3.01 -3
OCT_J-9A 3.0 3.00 103 POI V#7 instead of V#6
OCT_K-1 3.0 2.21 -19
OCT_K-3A 3.0 1.30 -15
OCT_K-5 4.5 2.70 -32
Total 76.0 69.0 1977.1
Saved Gas (MMSCFD) 7.0
Oil gain (BOPD) 1977.00
Net oil gain (including prize
2200
from additional gas, BOPD)

The following figure is a comparison between the five scenarios:

No. of optimized wells Incremental oil gain (BOPD)

1950 from optimized wells

+ 250 from additional gas
2376
2200

1465
400 from optimized wells
+ 400 from additional gas

773 800

42 22 8 42 26

Case#1 Case#2 Case#3 Case#4 Case#5

It is obvious from the above results that both 2nd and 5th scenarios are most reasonable as
optimization included limited number of wells which makes the optimization job easier. However, the
2nd scenario seems to be simpler than 5th scenario as it only includes changing the Gas-lift rate in

October PROSPER & GAP Model instructions 15

some wells that can be done remotely using the automation system in October field to check the
effect of changing the gas-lift rate on well performance that can be evaluated from wellhead
parameters (wellhead temperature), also it can help measure the accuracy of the model, on the other
hand, the 5th scenario may require additional surveillance such as GLS before doing any intervention.
The complete set of results could be found here.

Given that both PROSPER and GAP models have some uncertainties due to lack of data and
old well tests, these results are not that accurate but can be used as a guide, however these
results showed some facts:

1. There is a room for optimization which can achieve around 1000 BOPD (Risk weighted) by
doing some wellwork and optimizing IG rate in some wells. (Case #5)
2. Regular surveillance data and well testing are necessary to update the model and reduce
uncertainties.
3. District engineers are doing a good job in optimizing wells and gas-lift rates; this was obvious
from 1st scenario results after applying a risk factor and considering model uncertainty.

3.3. Copying the model

In case you want to copy the model to check the effect of any specific change to the system, or
changes to VLP ranges, Do Not copy the file (.gap) to any other location or even to the same location
of the original file as any changes in IPR or VLP curves will be applied to all versions using the same
(.vlp) file and this may damage your original model, the best way is to copy the model with all files to
make sure any change will not affect the original model, to do so; use Archive option, this option
offers the capability of creating an Archive of the whole project. It is in a sense a ZIP file that includes
all the associated files required when running the model.

a. Open the file you want to copy.

b. File >> Archive >> create

c. Choose the new location to which you want to copy the file, then click save.

October PROSPER & GAP Model instructions 16

d. Select all files, then click write, this will create an archive of all files into the new location.

e. Go to the new location and open the file (.gar), the following screen will appear:

October PROSPER & GAP Model instructions 17

f. Select all baggage and then click on Extract. This will extract all files into the new location,
and a dialogue box will appear, click open.

g. GAP will open the new file but will require you to edit file paths as appears in the dialogue
box below:

October PROSPER & GAP Model instructions 18

h. Click Yes, Select all files, then click on Alter paths, after that select file location and click OK.

October PROSPER & GAP Model instructions 19

i. All files paths will be corrected as shown in the screen below, then click OK, and now your
copy of the model is ready to be used without affecting any other versions.

October PROSPER & GAP Model instructions 20

Finally, all data can be found on the operations share folder here.
\\gupmafv03\Opr_Share\NGOS_Share\BASE MANAGEMENT\PROSPER_2014

October PROSPER & GAP Model instructions 21