Professional Documents
Culture Documents
Mugie Ginanjar*
Hani Mohede*
Martin Mlacnik*
RESERVOIR MODELING AND configurations. This suggests that the total OGIP
FORECASTING estimation from material balance model is robust.
Multi-Tank MBAL Model Generation Well Deliverability Modeling, IPR Limitation and
Sensitivity, Rate and THP Matching
A multi-tank material balance modeling approach
has been implemented to capture long-term forecast Well deliverability model is generated by surface
uncertainty due to reservoir connectivity behavior performance matching (Rate-THP match). To do
and well deliverability (IPR shapes) at low pressure. this, in MBAL, “Manifold Pressure per Well Model”
This was done using MBAL software from must be selected in prediction options to constraint
Petroleum Experts. A mixture PVT which already the production on well by well basis. If this is not
account for impurities, condensed water and selected, the constraints of prediction will be on field
condensate is fed into the model. The approach to level. Thus, we cannot perform wells matching in
model PVT mixture is using Sutton correlation for rate and THP on each well. Trial and error analysis
hydrocarbon gas mixtures, Kay’s mixing rule for then conducted on IPR parameters to get THP match.
adjusting nonhydrocarbon content, Beggs & Brill for Figure 10 is the last 2 years of THP matching
Z-factor and Lee et al for gas viscosity. Production example of several wells in Suban. These results
history input using well by well basis will improve suggest that the well deliverability model for Suban
the workflow of analyzing various number of tanks had been well generated and the material balance
in the reservoir which are suitable to historical model is in good shape to forecast future gas
pressure data. Pressure data are primarily coming production. The other wells are having similar
from SGS data converted to certain datum depth. matching quality to the wells presented in the figure.
P/Z History Matching, OGIP ranges However, due to high deliverability wells in Suban
field, the true shape of IPR especially at low
Suban material balance model is divided into 2 big pressure, frequently cannot be estimated accurately.
areas, West-Central area and East area. This division This happens due to rate limitation in well testing
is based on interference test which suggest West and equipment and pressure limitation in the facility.
Central areas are connected but separated from East Figure 11 is the example of actual data if plotted in
area. Therefore, the P/z analysis of West-Central and the IPR chart. The data only covers small percentage
East area can be interpreted separately. Trial and of potential absolute open flow (AOF). As
error approach was done to get suitable tank consequences, 3 shapes of IPR curves can be drawn
configuration and transmissibility to match historic (as example low, mid, high) with similar high-
data. Figure 7 and Figure 8 show comparison of P/z pressure regime rate-THP matching quality. The
vs Gp matching from West-Central areas as well as impact of this uncertainty will be addressed in the
East area using 2 and 3 tanks model. From the following section.
figures, it is suggested that 3 tanks model from both
Long Term Reservoir Forecast: OGIP and IPR
areas have superior matching. Therefore 3 tanks
sensitivity Impact to Forecast
model is selected as the basis for further analysis.
Single-tank model is not suggested as the data clearly One of the simplest method to understand the
shows multi tanks behavior. On the other hand, there significance of the uncertainty is to use it to predict
is currently no strong justification to suggest more the production forecast. Both OGIP and well
than 3 tanks. This analysis, however, can be changed deliverability uncertainties which are mentioned
if the future data suggests differently. Better data above will be used to predict production forecast
acquisition strategy will help us to reduce the using various development scenarios. There are
uncertainty in the future. currently four (4) development scenarios planned for
Suban field. The scenarios are:
Sensitivity analysis was performed on the OGIP
value by changing the configuration of tank OGIP 1. As is, no additional project.
and transmissibility between tanks, to see whether
the simulation P/z can still match the historical data. 2. Install Compressor
After numerous iterations, we observed that we could
have similar P/z history match using different sets of 3. Install Compressor and 1 future well
tanks OGIP (Figure 9) even though only small
variations of total OGIP were observed from these 4. Install Compressor and 2 future wells
Figure 12 shows production forecast result of 4 able to test high deliverability wells and will not
development scenarios above based on OGIP create any potential hazard to the gas processing
uncertainties while Figure 13 is showing the results system in the plant.
based on well deliverability uncertainties. Stronger
variances of decline period and decline rate are SUMMARY AND CONCLUSION
observed from well deliverability uncertainty. Figure
14 is summarizing the range of EUR impact. The Multi-tank material balance model has been
impact of the well deliverability uncertainty is almost generated to analyze the range of original gas-in-
double the impact of OGIP uncertainty based on raw place, internal connectivity and well deliverability of
gas recovery. Both uncertainties, however, have Suban, which strongly affects the long-term
significant effect to the field especially to the production behavior from the field. The model
fulfillment of the current demand as well as the performed well to capture more complex
ability to capture potential future demand. Therefore, connectivity than the originally assumed single-tank
future data acquisition strategy must be refined to get behavior thus reducing the uncertainty of OGIP
better understanding of the long-term behavior and range. On the other hand, well deliverability model
reduces uncertainties. has been matched to actual rate-THP profiles but
limited to small portion of AOF due to system
FUTURE DATA ACQUISITION STRATEGY constraints. Both connectivity and well deliverability
uncertainties have significant impact to the recovery
SGS and PLT Data Acquisition Strategy and thus the future development plan of Suban field.
Historic data acquisitions are concentrated on To further reduce the uncertainty of connectivity and
peripheral wells due to aquifer concern. With well deliverability, newly refined data acquisition
diminishing needs to monitor aquifer (because no strategy and production test beyond current facility
evidence of water influx until recently), and limitation have already been planned. This will
increasing needs to better characterize reservoir enhance the accuracy to predict the long-term
tanks, it is important to acquire data not only in the production behavior and boost our confidence to
periphery wells. Newly refined strategy of SGS and fulfill current and future market demand.
PLT data acquisition for future years has been
communicated to management. Over the years, the ACKNOWLEDGEMENTS
data will be more concentrated on area of interest to
anticipate the interpretation of additional tanks in The authors would like to thank ConocoPhillips
material balance model and to support the analysis of Grissik Ltd., Corridor PSC partners (Talisman
future development wells. (Corridor) Ltd., PT Pertamina Hulu Energi Corridor,
PT Pertamina EP) and MIGAS for permission to
Production Test Beyond Current Limitation publish this technical paper. We also would like to
thank all ConocoPhillips colleagues which names
Current facility pressure will limit the ability to cannot be mentioned one by one who support the
estimate the full shape of the well’s IPR especially at authors in many technical discussions to makes this
low pressure. As mentioned before, the impact from paper complete.
this uncertainty is significant to the long-term
production behavior. This behavior will dictate our
REFERENCES
ability to fulfill current demand as well as capture
potential new market. Therefore, it is critical to
Brill, J.P. and Beggs, H.D., 1974, Two-Phase Flow
acquire the production data beyond current facility
in Pipes, paper presented at the U. Tulsa
limit.
INTERCOMP Course, The Hague (1974)
The project to test the wells beyond facility limit is
Kay, W.B., 1936, Density of Hydrocarbon Gases and
ongoing and had been discussed with production
Vapors at High Temperature and Pressure, Ind. Eng.
engineer, operations and facility engineer. Flare
Chem. (1936) No. 28, 1014
system in the plant will be used to test the well with
much lower THP without distracting current
processing facility. Slight modification is necessary Lee, A.L., Gonzales, M.H., and Eakin, B.E., 1966,
to bypass the system and connect separator directly The Viscosity of Natural Gases, JPT (August 1966)
to the flare. This modification will make the project 997, Trans, AIME, 237
McKetta, J.J.Jr., and Wehe, A.H., 1962, Sutton, R.P., 1985, Compressibility Factors for
Hydrocarbon/Water and Formation Water High-Molecular Weight Reservoir Gases, paper SPE
Correlations, Petroleum Production Handbook, T.C. 14265, SPE Annual Technical Conference and
Frick and R.W. Taylor (eds.), SPE, Richardson, Exhibition, Las Vegas, Sep 1985
Texas (1962)) II, 22.
Whitson, C.H., & Brule, M.R., 2000, Phase Behavior
Mohede, H., Malick, K., and Tyberoe, G, 2014, (Vol.20, SPE Monograph Series), Richardson,
Suban-South Sumatra Giant Fractured Gas Reservoir Texas: Henry L. Doherty Memorial Fund of AIME,
Development and Challenges, Proceedings of the Society of Petroleum Engineers,
Indonesian Petroleum Association Thirty-Eight
Annual Convention and Exhibition, May 2014
Figure 1 - Suban top Pre-Tertiary depth structure map. Generally, Suban is divided into West, Central and
East Area.
Figure 2 - Suban gas production history. Most contributors of production are the well in West area due to
higher deliverability. Several projects increase the capacity of the plant overtime. Decline of
production is due to demand driven not reservoir driven.
Figure 3 - Water production and water gas ratio of Suban. Increasing water gas ratio fits with the line from
correlation of condensed water prediction from reservoir.
Figure 4 - Static Gradient Survey (SGS) location and count bubble map. Peripheral wells have more data
acquisition than wells in the middle.
Figure 5 - P/z plot of West-Central and East areas. Recent data in West-Central and East areas suggest
compartmentalization or internal connectivity might be more complex than previously assumed.
Figure 6 - Interference test summary. West-Central areas are in pressure communication but separated from
East area.
Figure 7 - West-Central Areas P/z match using 2 and 3 tank model. 3 tank model suggested to have better match
Figure 8 - East area p/z match using 2 and 3 tank model. Similar with West-Central areas, 3 tank model suggested to have better match because it honors the most
recent data
.
Figure 9 - Sensitivity of OGIP with different configuration of West and central OGIP. With different value
of transmissibility between tanks, all 9 cases could be matched to historic pressure data. However,
the total OGIP remains similar for all cases.
Figure 10 - Good THP matching of several wells in Suban. Other wells having similar quality of matching. Note that this matching is the results of forcing the rate of
the well in the model to the actual rate.
Figure 11 - IPR derived from actual data for Suban-15 well (converted to bottom-hole). Due to pressure limitation, only small percentages of AOF is captured by
actual data. Therefore, the uncertainty of actual shape of IPR is quite significant. Low-mid-high IPR having similar match to actual data.
Figure 12 - Resulted forecast from 9 cases of OGIP configuration using 4 development scenarios. Similar production behavior is observed for all 9 cases.
Figure 13 - Resulted forecast from low mid high well deliverability (IPR) estimation. Observed different production profile with different decline period and decline
trend of produced gas.
Figure 14 - Gas recovery impact derived from max – min total gas recovery. Observed both impacts are
significant with stronger impact coming from well deliverability uncertainty.
Scenario 1: Suban as is, Scenario 2: Suban as is + Compressor, Scenario 3: Suban as is +
Compressor + 1 well, Scenario 4: Suban as is + Compressor + 2 wells.