Production Optimization Projects
A summary of some of the production optimization projects undertaken by Weatherford consultants over the last few years are listed below. They comprise a vast scope, from simple modeling and diagnostic exercises, entire field production models for artificial lift, right up to well and network modeling for field automation.
Case #1 PRAP
PRAP (Proyecto de Automatizacion de Produccion) is a major project of PDVSA, the national oil and gas producing company of Venezuela. The objective of the project is to use optimization software, linked with automation through SCADA to control and optimize the production from three production districts in Lake Maracaibo in Venezuela. The production facilities consist of gas lifted and naturally producing wells producing to flow stations where gas and liquids are separated. The gas is then sent to compression stations which supply the gas lift network, compression fuel, and third party users. The ReO® software application is used to optimize the entire system subject to the practical constraints which apply.
Results from using ReO software in 4 separate parts of different oilfields have been measured and validated. The gains in oil production are between 2.3% and 7%, averaging 5%, and the savings in lift gas also average 5%. The total oil production from the 4 field sections was 114,000 barrels/day, and this has increased by about 5,000. At $20 per barrel this represents $100,000 per day extra cash, or $36 million per year impact to the bottom line. Scale up to the whole of Lake Maracaibo is now underway as PDVSA rolls this proven technology out to the other production areas.
Back to Top
Case #2 Egypt - Gas Lift
A WellFlo® model of this Middle East field has been used for optimization of the gas lift production and for assessing the impact of additional wells, pipelines, and platforms. The field is located in the Gulf of Suez and produces a total of 225 Mbpd of fluid and 165 Mbpd of oil and the volume of gas injected is 110 MMscf/day. The total number of active wells is 55, of which 52 are gas lifted, 2 jet pump lifted and one producing naturally.
The field has 9 satellite producing platforms, one platform with combination gas lift and fluid separation (Complex Platform) and one for gas compression. The production network comprises 7 production pipelines of outer diameters 24", 20" 18" and 12".
The wells have been modeled with WellFlo software. The production network has been modeled in FieldFlo™ software, with seven modules which represent the seven pipeline that connect the 9 satellite platforms to the complex platform. This model has been tuned to the pressure observed at the manifold header of the satellite and complex platforms.
The WellFlo model was run for the following cases:
- Case-1: Optimum gas lift allocation of available lift gas volume (i.e. 110 MMscf/d). For this case optimization resulted in an oil production increase of 4.3%
- Case-2: Optimum gas lift allocation of 20% extra lift gas volume. Optimization resulted in an oil production increase of 7.6%.
- Case-3: Optimum gas lift allocation of unlimited lift gas volume. For the maximum lift gas volume (around 260 MMscf/d) optimization resulted in an oil production increase of 14.4%.
The optimized WellFlo model was then used by the client to study the impact on operations of the addition of new wells, new pipelines, and a new platform. The result showed that increased oil production was possible with reduced lift gas.
Back to Top
Case #3 Venezuela - Field Automation
PRAP (Proyecto de Automatizacion de Produccion) was a large on-going project in Venezuela to automate process control and optimize production from three production districts. The production facilities consist of gas lifted and naturally producing wells producing to flow stations where gas and liquid are separated. The gas is then sent to compression stations which supply the lift-gas network, compression fuel and third-party users.
Weatherford was the main sub-contractor in the area of optimization and our scope of work was to provide:
- Petroleum engineering consulting services
- Supply of software tools
- Software design and development
- Training services
The above services are tailored to meet the project objectives in the following way:
- Provision of consulting petroleum engineering services to construct 250 well models using WellFlo software and to integrate these models into the gas lift optimization system. This has been done by constructing flow station models which are produced gas and oil collection points and which are each supplied with lift gas from multiple gas lift manifolds. Each well model includes modeling of fluid PVT properties from one of a number of produced reservoirs and inflow performance based on flowing gradient surveys or calculated from production horizon geological and wellbore skin parameters. Each well has a gas lift completion and once the vertical lift performance is tuned to match measured performance it is included in a FieldFlo software flow station model. As part of the gas lift design module WellFlo software also has facilities to model the performance of gas lift and orifice valves, ensuring injection pressures and valve mandrel positioning are optimal based on well inflow performance.
- The oil production facilities are modeled with FieldFlo software and include all wells, chokes and surface production flow lines from wellheads to production or test separators. These models allow the gas allocated to each flow station by ReO software to be optimally distributed amongst wells. This is done by generating well performance curves based on produced oil. FieldFlo software will be included as part of an automated well test and well performance update system which drives the Dynamic Gas Optimizer to optimize the field gas allocation, production, and profit.
- Provision of WellFlo software software for modeling and problem diagnosis of wells and surface facilities and gas lift design. The software incorporates flow correlations which have been customized to model the vertical lift performance of wells, based on flowing gradient survey data collected over the last three years. This involves accurate modeling of the reservoir fluid PVT properties and of the flow regimes and associated static and frictional pressure gradient components in the well completions. The resulting correlation is a mechanistic model based on published correlations and tuned to match the client's field conditions.
- The client's existing expert system for diagnosing problems with gas lifted wells was replaced with a new system using the WellFlo software models to calculate pressures and temperatures in the tubing, casing and gas lift valves at various depths in the well. The expert system reads surface pressures and gas injection rate automatically from the SCADA system and combines these with the well models and recent production test data to determine whether the gas lift design is performing optimally. Diagnoses and recommendations for remedial action are automatically generated and routed to the production engineers.
- Customization of Weatherford's ReO software gas network software to allow optimization of the lift gas allocation. ReO software uses a compositional gas model to calculate the pressures in the non-hierarchical gas network and optimizes the distribution of available gas to gas lift manifolds based on flow station performance curves generated by FieldFlo software. A detailed compressor model has been developed so that the performance of the real machines is considered in the solution generated. Since the Dynamic Gas Optimizer is integrated with SCADA, well gas injection rate set-points can be sent automatically to be implemented by the automatic rate controllers on the lift gas manifolds.
- Implementation of Weatherford's ReO software multiphase network software to model a complicated network of naturally flowing wells. The network consists of several production manifolds where production from wells is gathered and then sent to the central separation plant along 10 parallel flow lines This geometry requires ReO software's non-hierarchical network solution capability. The model allows production engineers to determine the optimal line-up of flow lines and wells to high and intermediate pressure separation systems.
- Development and delivery of training courses to ensure technology transfer to client personnel.
- Development of software interfaces to integrate the Weatherford components with the SCADA system and common databases shared by all other applications. The project has required a virtual organization to be created consisting of Weatherford in Edinburgh and Houston, the client in Venezuela and the project headquarters in Venezuela and Houston. The key to the success of this virtual organization has been the use of sophisticated IT systems.
Back to Top
Case#4 UK North Sea Gas Lift Diagnostics
The operator's gas lifted horizontal well was producing liquid in an unstable manner unless it was run "flat-out" in terms of injection gas rate and FWHP. The operator had made the decision to work over the well, but wanted some quantification of the problem itself and of the likely effects of various remedial actions (e.g. valve change outs, velocity string insertion).
Models were built in WellFlo and DynaLift™ software to investigate the gas lift regime for the well and to see if there were any indicators which would confirm/refute the possibility of phase separation. Both models were tuned to current operating conditions and the DynaLift model was used to explore the nature of the observed production instability. The DynaLift model was then used to investigate the potential effects of changes to gas lift valve properties and the introduction of velocity strings. Any steady-state solutions were checked against WellFlo software predictions for the same conditions. Indicators of phase instability, such as unfavorable flow regimes, could be identified.
The platform engineers for the well agreed that the model responses to variation in gas injection rate and FWHP were qualitatively (and largely quantitatively) close to observed behavior. The dynamic gas lift model was able to show the various degrees of instability of the well, and also to mimic conditions of near-stability when the well was being run "flat-out".
The conclusions reached were that the orifice valve (24/64")was oversized for the conditions. Raising the injection rate and CHP were only possible in the short term, due to operational difficulties (Xmas tree vibration), so downsizing the valve (16/64") was the preferred route. The well would operate in a stable manner and production would be increased by up to 22% in the previously unstable operating region. The stable operating envelope of the well would also be greatly improved.
The second element possibly causing the well instability, phase separation, was not directly modeled, but the indications were that it could be a contributing factor under certain flow conditions. The probable effects of velocity strings with and without a change in well PI value were determined. In each case, production levels would drop relative to current levels, and the change out of the orifice valve would still be necessary.
Back to Top
Case #5 Australia - Gas Lift
This large Australian field consists of over 400 online producers, the majority of which require some form of artificial lift, with rod pumping the predominant mechanism. There were some 90 gas lifted wells, 80 of which were continuously gas lifted; the remainder being intermittent or gas lift assisted plunger lift. Total oil production was approximately 15000 stb/day with a total lift.
A WellFlo model was constructed for each continuously gas lifted and naturally flowing well and the model performance was matched to flowing gradient survey data. The performance of the rod-pumped and intermittent gas lift wells was modeled in FieldFlo software using manually constructed performance curves.
Total production was matched to within 0.1% of the reported value and optimization of the current lift gas capacity resulted in an uplift of 3% of the total oil produced from the gas lifted wells. The results also indicated that it may be possible to maintain current production with a lower total gas lift rate.
Back to Top
Case #6 Venezuela - Choke Modeling
An optimization study of this onshore field has been done for one of the pipelines of the field which connects 7 naturally producing wells to a main separator station. This field is located in South-East Venezuela producing around 50 Mbopd. Production is controlled through the well production chokes according to an established depletion policy for the field by the client. Critical flow through the chokes is required in order to avoid severe interference problems between the wells and the optimum choke size for each well was determined such that the wells produce the maximum oil rate maintaining critical flow through the production chokes.
WellFlo software was used to model the 7 wells and FieldFlo software to model the production pipeline and the chokes. The maximum choke size was determined using WellFlo software and the optimum choke size for critical flow was determined using the FieldFlo software network model.
This study was a pilot project carried out for the client to evaluate the feasibility of using WellFlo software to optimize field-wide oil production.
Back to Top
Case #7 U.A.E - Gas Lift
The objective of this project was to build a field model for an offshore Middle East field using WellFlo software that matched well test and production data to within 5%. The field consists of 48 wells on five jackets draining seven production zones and feeding to four first stage separators.
PVT data was characterized and well deviation, completion and reservoir data included in the models. A best match flow correlation was used to match the latest test data using straight line IPRs, and tuning of reservoir pressure. Each well was modeled and then integrated into the field model using FieldFlo software, reproducing test results to within 1.8%.
The gas lift allocation for the field was then optimized which gave an increase of 4.8% in production from the gas lifted wells. The top ten candidate wells for gas lift optimization required a change of from 6% to 75% in the gas lift injection rates. Recommendations were also made to improve the model by obtaining flowing gradient surveys to improve the selection of the vertical flow correlations. Further optimization opportunities were identified during the study, including reallocation of wells to different separators, reducing separator pressures, choke sizing, tubing sizing, tubing versus annular flow, and ESP well optimization.
Back to Top
Case #8 Austria - Gas Condensate
A gas condensate field in Austria drained by six wells feeding to a common separator was modeled in order to optimize and increase production required to match an upgrade in gas treatment facilities planned by the operator. The project was divided into phases to allow assessment of results as the analysis progressed. Firstly an analysis of PVT data was made to allow modeling of condensate flow. Secondly a model of the complete field production system was built using FieldFlo software and sensitivity studies done to identify the optimal tubing sizes to increase production without causing condensate drop out.
A third phase modeled a new gathering system delivery performance to the separator as reservoir pressure declined for a fixed or variable rate production profile.
Back to Top
Case #9 Argentina - Well Modeling
Nine wells in the Neuquen province of Argentina, six gas lifted wells and three with ESPs were modeled from the reservoir to separator using WellFlo software. The wells produced from three sandstone horizons and production was fed through flow lines of 500m to 2000m to a manifold/separator. The project combined training of the operator's personnel in the use of WellFlo software and the modeling of the wells followed by sensitivity studies. Once the well models were built and tuned to recent production data, sensitivities were run on gas injection rates for gas lifted wells and for pump speed on ESPs. A comparison study was made on replacement of gas lift with ESPs and vice versa. This showed significant oil production gains from two wells by installing ESPs and excluded other candidate wells due to low flow rates and high GOR.
The injection gas rate of one well was found to be above optimum and could be reduced without loss of production. Three well performances could only be matched by reducing the effective tubing internal diameter, giving an indication of paraffin build up in the production string of these wells, which was confirmed by client engineers.
This short study indicated that a full field model would be required to optimize gas lift and a set of gas lift injection rate performance curves would be required. Orkiszewski and BAX flow correlations were identified as suitable for the production rates studied, but a wider study would be needed to determine the flow correlations required for each set of possible flow conditions over the fields remaining life.
A further study was done on 9 condensate gas wells feeding one separator unit. Improved matching of IPR curves was done followed by evaluation of re-completion of these wells after the onset of depletion. The study identified the key parameters that would be required to be monitored and tested to ensure optimal field production. Liquid loading was identified as a key problem in field decline which would require further data gathering. Recommendations were made for future well tests to improve data quality and reservoir modeling
Back to Top
Case #10 U.K. North Sea - Mature Field
This project was performed on a mature water-injected North Sea oil field with 23 oil producers, some of which require artificial lift due to decreasing reservoir pressure and increasing water cut. Pressure maintenance is being provided by 9 water injection wells. Injection of some 100,000 bbl/d was less than required due largely to the unaccountably high pressure losses in the surface injection network. Also the lack of opportunity for full testing of the wells, due to operational constraints, meant that the injection rates and layer pressures for these injection wells were not known.
The surface flow lines were fitted with both chokes and non-return valves. Examination of the available data suggested that there were appreciable pressure drops across the valves as well as the chokes. Subsequent re-calibration of the gauges yielded data that indicated the pressure drop to be across the choke section of the flow line only. Investigating both the theoretical choke performances and the available historical data the pressure losses were some 10 bar above that expected, with the likely cause being debris build-up in the flow lines. From the historical data it was possible to determine the injectivity indices for each well and the layer pressure. This provided a guide both to the current conditions in the reservoirs and also pressure history over the preceding 6 months for which data was available.
WellFlo software was then used to model the wells to match the current injection with the determined injectivity indices and pressures and performance curves of the wellhead pressure and rate were generated. From these, possible future injection rates could be determined from the wellhead pressures likely to be available. This field is now out-performing its production forecasts by some 10%.
Back to Top
Case #11 Australia - Field Modeling
Weatherford modeled the production system of six oil and two gas wells on four monopod platforms in Australia for two base case oil production rates and four gas production rates. These wells produce from two separate fields so PVT data for both fields and appropriate vertical and horizontal flow correlations were used to match available test data to within 1.2%.
The model was used to predict system performance to the separator for 32 different production scenarios and although the oil producers were naturally flowing the model included gas lift valves for future production simulations. FieldFlo model provided required choke sizes for each scenario and confirmed critical flow in each case. The FieldFlo model allows daily management of the field by predicting production rates, GOR, and pressures at pipeline tie in points.
Back to Top
Case #12 China - Offshore ESP Design
A field model was required for a subsea completion of ten ESP completed wells in China producing via two flexible riser to surface, allowing a prediction of wellhead pressures for a range of operating conditions. As the system consisted of two identical five well manifolds, only one of these was modeled to save cost and time. The model includes the well deviation and completion, flow lines from each well, the 500m horizontal section of the riser, the flexible section to vertical terminating at the separator
Wellhead pressures were modeled for naturally flowing wells at low water-cut and then for ESP completions as water-cut increased. FieldFlo software confirmed that the wells would flow naturally for the first year and predicted operating pressures, rates and temperatures for this period. The original design was done using a fixed PI but as water cut increased the PI was expected to change. Although no relative permeability data was available the reduction in viscosity as water-cut increased made significant changes in operating conditions and therefore the selection of ESPs.
Back to Top
Case #13 Venezuela - Field Model
A gathering system in Venezuela was modeled in tandem with training of the operators personnel in the use of WellFlo and FieldFlo software. This highlighted to the operator what data would be required to build such a model and to optimize production. The model included individual well models and a gathering system model involving vertical and horizontal multiphase flow correlations. The system was matched to within 0.2% then the options for optimization were highlighted and a dynamic tool for economic evaluation of various field management strategies was implemented.
A optimization on the allocation of injection gas yielded significant changes in injection rates of most wells and predicted a 4.4% oil production increase, leading to a gross revenue increase of US $1.6 million, assuming a pessimistic decline in production rates of 30% per year.
Back to Top
Case #14 U.K. North Sea - Water Injection
Two well models were built in order to assist in diagnosis of well problems on a North Sea water injected field. The wells modeled had suffered a significant drop in productivity. No PVT data was available, so a nearby well and best fit correlation data was used to build a model to match the flowing gradient survey at a measured liquid rate, GOR, FTHP, water cut, and gas lift injection rate. This model was then used to predict FBHP which, combined with recent estimates of reservoir pressure, was used to define the inflow performance relationship, using the Vogel model. This then allowed a prediction of skin factors for various drainage geometries to be calculated, showing the build up of scale damage over time. For the first well it was shown that fracturing the well and scale inhibitor treatment could have significant effects on productivity but for the second well the predicted small improvement in well productivity did not justify a second scale treatment.
Back to Top
Case #15 Argentina - Gas Condensate
A model of part of the production system of a gas condensate field located in the south of Argentina was done with WellFlo software. Four gas condensate wells connected to common manifold were used for this pilot study to demonstrate the feasibility of using FieldFlo software to model gas condensate production and optimize the operation of the field.
A detailed study of the inflow performance of the well was done using the comprehensive reservoir description module of WellFlo software. This included the analysis of the impact of non-Darcy flow determined with modified Isochronal tests. The main objective of this study was to determine deliverability as a function of separator pressure.
A similar of study was done for Perez-Companc on a gas condensate field located in the Neuquen Province of Argentina.
Back to Top
Case #16 Colombia - Jet Pump/ESP Evaluation
Weatherford was requested to select suitable candidates for conversion to ESP from a 17 well field producing approximately 4000 STB/D. The first objective was to model the performance of 10 jet-pumped wells, so that optimum lift capacities could be determined and used as a basis of comparison with the performance after the installation of an ESP. The second stage consisted in selecting, from the 10 jet pump lifted wells and 3 naturally flowing wells, the best three candidates for ESP pumping.
The jet pumped wells were modeled using spreadsheet software developed in-house by Weatherford, incorporating current industry best practice and models. Input friction pressure data, often poorly estimated from simplified charts and assumptions, was accurately determined using WellFlo software for both the power fluid and produced fluid in the jet pumped wells. The performance of the naturally flowing wells and potential ESP candidates were modeled using WellFlo software.
Modeling of jet pump performance showed that minor modifications to nozzle/throat combinations in several wells could lift production in the field by 5%. Further modeling indicated that if scale removal treatments were performed that production of the field could be increased between 50-75%, depending on the actual PI increases resulting from the treatments.
Of the thirteen well models modeled with an ESP, the top three candidates for conversion were selected on the basis of additional oil produced compared with the natural flow/jet pump cases. The predicted total gain after conversion to ESP was approximately 10,500 STB/D, again subject to effective removal of scale during workover. An additional 5000 STB/D could be expected from converting an additional two jet pumped wells to ESP.
Back to Top
Case #17 Indonesia Network Model for GL Platform
As part of a software purchase agreement with its client, Weatherford was requested to provide training, consulting and analysis services for its FieldFlo software applied to one of the client's production platforms located in the Indonesian sector of the Natuna Sea.
The outlined objectives of the project were achieved:
- Eight individual well models (WellFlo software) were reviewed and updated with respect to layer properties, PVT, and depth information. Appropriate flow correlations were identified for VLP and gas lift performance curves derived for the gas lifted wells.
- A working FieldFlo model has been constructed using the updated well models and tuned to observed production test data. The history match was within 4%. The modeling process identified serious shortcomings in platform data acquisition (particularly water cut determination).
- The tuned model was used to identify possible opportunities for increased production. A potential gain of 900 stb/d of oil (3.5% of total production) was identified. Also the best candidate well for gas lift was determined to have an inadequate design to achieve its potential.
Back to Top
Case #18 Indonesia Artificial Lift Analysis
Weatherford was asked to analyze production from a mature offshore oilfield producing with a high water cut and evaluate the possibilities of improving current gas lift potential or switching to other forms of artificial lift. The decline in field production coupled with increasing water cut presented the client with two major technical issues relating to increasing well production:
- The optimization of production from existing wells based on current completions and gas lift. The technical requirement was to find ways of improving current well productivity.
- The feasibility (technical & economic) of replacing current gas lift systems with alternative forms of artificial lift in order to increase production. This was particularly relevant to wells located in certain regions of the reservoir where considerable reserves remain to be produced, provided that sufficient drawdowns (higher than those currently achieved) could be applied.
Individual models for each producing well were constructed using Weatherford's WellFlo software, and tuned to flowing gradient survey data obtained in 1997. These models were then used to evaluate the performance of each well under different artificial lift options: increased lift gas, decreased tubing head pressure and downhole pumps.
The major conclusion from the study was that downhole pumping would provide a potentially significant increase in oil production compared with increased lift gas injection rate or tubing head pressure reduction. On the basis that downhole pumps located in every string will achieve sustainable drawdowns of 75% of AOF, compared with current levels of » 30%, oil production would increase by » 2500 stb/d. This was an optimistic forecast, however even at lower drawdowns the incremental production would be greater than in the case of modified gas lift.
For most of the strings studied total liquid flow rates would still be below 1000 stb/d, indicating that PCP pumps would be more appropriate technically and economically compared with ESP pumps. At the same time three of the wells were identified as possible ESP candidates, based on potential liquid rates greater than 1000 stb/d.
A comparative analysis of monobore completions in some of the wells indicated that deeper access for gas lift valves in a conventional completion would be expected to increase oil production by around 30% if the original monobore well was producing in stable flow. In stable wells with a significant fluid layer above the production packer, the use of downhole pumps in current monobore completions would still outperform conventional, deeper gas lift completions.
Back to Top
Case #19 Malaysia Gas Lift Audit and Production Optimization
Weatherford was requested to undertake an audit of the gas lift system in place on a seven platform offshore field and to construct WellFlo and FieldFlo models of the active wells. This was then used to optimize individual gas lift allocation on a field-wide basis.
Phase one of the project consisted of a physical, on-site audit of the field's gas lift system and infrastructure, data gathering, and validation. The audit identified several areas of concern and recommendations were made for improvements to the system. These related mainly to establish stabilization and control of the injection gas, as well as the identification of poorly performing valves. The recommendations provided the basis of an on-going work program by the client.
Phase two consisted of the modeling of 60 individual producing strings using WellFlo software and the field as a whole using FieldFlo software. The second phase also involved a field trial on a selected jacket, where the improvements/recommendations identified during the WellFlo software modeling were implemented and the results monitored. The results showed a saving of 25% in the total lift gas used for the field trial wells and the released lift gas was reallocated to other wells, giving an additional 10% oil. Slugging/heading and multi-pointing wells were stabilized and control of the lift gas at surface was possible with the single string wells.
Phase three of the project consisted of installation of the entire field network model on the client's computer system and training of the client's engineers to run, maintain, and update the model.
The project was completed on schedule and within budget and the operator is now continuing the work and following recommendations made by Weatherford, has assigned a dedicated team of engineers to the project to continue the gas lift optimization on a field-wide and individual well basis.
Back to Top
Case #20 West Africa ESP Design
Weatherford was initially approached by a client to perform some modeling work to size a new pump being proposed by an East European supplier. The work consisted of modeling ESP performance for a generic well in an unspecified field. Reservoir, well, and ESP performance data provided by the ESP supplier were loaded into Weatherford's WellFlo software, ESP production modeling software and a number of sensitivity studies performed to various production scenarios, including changes in reservoir pressure. Both fixed and variable frequency options were modeled prior to final pump and motor sizing.
Back to Top
Case #21 Qatar Field Production Model
Weatherford was requested to build a WellFlo model for the production network of an offshore gas lifted oilfield. The main objective was to provide an engineering tool for the back-allocation of the total fluid production back to individual wells and also for general production optimization studies. Other objectives were validation of IPR, tubing performance and production test data.
A total of 64 wells were modeled for this study: 38 by Weatherford and 26 by the client. The FieldFlo model was tuned and history-matched to production data for May/June 1996. Some of the major results of the project were:
- The model predicted that variations in the separator pressure of +/- 25% would not have a significant impact on production because most of the wellhead production chokes are in critical flow.
- FieldFlo software was successfully used to identify flow lines and chokes restrictions with high pressure drops.
- The flowing gradient surveys carried out expressly for the study resolved several of the uncertainties in the test data and were used to update both the tubing performance and well IPR relationships.
- The static pressures of several crestal producers showed large variations which could be attributed to the presence of conductive faults in the reservoir.
Back to Top
Case #22 India Field Production Model
The operator of a small, but prolific, oil field offshore India was considering the feasibility of doubling current production through an already existing field infrastructure. The field network consisted of six oil/gas well platforms (A, B, C, D, E, and F) and one free gas platform (G). During the field development stage a number of new pipelines had been added to increase handling capacity.
Since all the platform facilities and pipelines had been designed and installed based on initial estimates of reserves and expected plateau rates (with some cushion), it was believed that there could be significant bottlenecks in the system which could prevent the field from producing efficiently at the then current rates, or at higher rates planned.
Weatherford was contracted to develop and install a WellFlo model of the field, and to use the model to investigate the performance of the field for a number of possible future production scenarios. WellFlo models of 16 producing strings were constructed and tuned to match the most recent available production data at the time.
These well models were then incorporated into a FieldFlo network model of the field. The model was first tuned to history match the reported test data for a selected day, when a relatively large number of wells were flowing and had been doing so for several days (indicating a stable flow environment and representative data). The model was tuned to match total field oil production to within +0.35%, which is well within the tolerance considered acceptable and generally this match was achieved with reasonable errors in the wellhead and manifold pressures (compared with reported data).
The tuned model was then used to predict field performance for four production scenarios selected by the operator where the effects of declining reservoir pressure and increasing water cut were considered. With the exception of one well (A), for each of the four cases selected by Cairn, the model predicted that the target rate for each well could be achieved. In three of the scenarios, well A was predicted to be unable to flow at the high calculated manifold pressure caused by the 8" flow line from platform D to E.
The main bottle-necks in the system were identified as the 10" flow line from A to T and the 8" flow line from D to E. The flow lines from E to T and B to T also were shown to result in significant pressure losses.
The results of this study and the working field model were presented to operator management in their India headquarters. At the same time operator production engineers were also trained to build, run, and maintain the well and network models. Nearly one year after delivery, the model is still used to monitor field production and diagnose potential bottlenecks.
Back to Top
Case #23 U.A.E - Offshore Oil Field
Four Middle East offshore fields, comprising a total of 38 platforms and 240 gas lifted wells were modeled. Total oil production from each field was as follows:
- Field 1 29600 stb/day
- Field 2 1855 stb/day
- Field 3 170400 stb/day
- Field 4 118000 stb/day
The combined oil rate was 320000 stb/day. Total lift gas rate for all fields was 1047 MMscf/day.
Each WellFlo model was tuned to match flowing gradient survey and test separator data using vertical flow correlation, L-factor, reservoir pressure, and productivity index as tuning parameters. The outflow performance of the well was first tuned by selecting the vertical flow correlation giving the closest match to the flowing gradient survey and then the match was fine tuned by adjusting the L-factor. Often the flowing gradient survey was 1-2 years old so to bring the model up to date the inflow performance of the well was adjusted (by changing reservoir pressure and productivity index) to match the current test separator data.
Production from each outlying platform flows to central facilities platforms for processing via subsea flow lines The FieldFlo model included details of the flow lines from each well to the platform header, the risers and flow lines from the outlying platforms to central facilities as well as flow lines on central facilities. The reported wellhead pressures were matched to within ± 10 psi by adjusting the pipeline L-factors.
The history match models matched total oil production to within -0.7% to +1.4% of the reported sum of the individual well test rates and since total field production with the wells flowing to the production separators is not reliably measured, this was considered to be a good match. Optimization of the currently available lift gas resulted in the following uplift:
- Fields 1 & 2 2140 stb/day +7%
- Field 3 4235 stb/day +2.5%
- Field 4 1280 stb/day +1.1%
- The average uplift was 2.4% corresponding to a total increase of 7655 stb/day.
Back to Top