Petroleum Development Oman (PDO), with the government of Oman (60%) and Shell (34%) as primary shareholders, selected Oman’s Fahud field for a pilot implementation of integrated production system modeling, which combines the entire production process from reservoir up to the hydrocarbon export point into one model. Fahud was chosen because of its complex mixed gas-oil gravity drainage (GOGD) and waterflood recovery mechanisms and large number of active wells. Success there could provide the confidence for further roll out within the company.

PDO decided to use a software package from Petroleum Experts, IPM, to construct a model for the pilot to evaluate tool functionality — given the complex nature of Fahud’s recovery mechanisms — and to establish benefits in production system optimization and IPM’s forecasting capabilities.

Fahud production challenges
The Fahud field has been on stream since 1967 and is producing 31° API oil from the Natih carbonate reservoir rock, which has a varying degree of fracturing. The Fahud field is a highly complex reservoir with seven units, two areas, and two development mechanisms, gas oil gravity drainage (GOGD) and waterflood. The waterflood is concentrated on layers with relatively low degree of fracturing (EFG, Shuaiba), while the GOGD process targets the more fractured reservoir rock. The waterflood developments in Fahud are relatively immature and average water cut is relatively low at 45%. The total gas/oil ratio (GOR) in the Fahud field is high at 600 cm/cm. Produced gas is compressed and re-used for gas lift or injected at the crest of the reservoir to maintain and control the position of the GOGD oil rim.

Gas compression capacity is the main facility constraint, and managing the well stock within this compression constraint is one of PDO’s key objectives for IPM. The Fahud field is produced through approximately 180 active producer wells supported by 25 water injectors and four gas injectors. Approximately 50 producers are on intermittent, stop-cock, production cycles. With intermittent production behavior in the GOGD process, wells come in at a high oil rate and low-formation GOR after a closed-in period. The oil rim is rapidly produced, and with gas coning, the GOR goes up while the oil rate drops. When a well’s GOR reaches a cut-off level (set by facilities constraints), the well is closed in and the oil rim will be recharged for the next production cycle. Evaluation demonstrated significant scope for optimizing the production management of 50 stop-cock producers within the gas-constrained production facilities.

IPM overview
PDO began using the Petroleum Experts (Petex) IPM suite of software in 2004. The objective was to integrate reservoir engineering, production technology, facilities engineering, and operations workflows through a single tool and a new way of working. Traditionally, each discipline used its own application tools. Reservoir performance modeling in the Petex software uses material balance (MBAL) techniques based on the MBAL application. Completion design and modeling is carried out in PROSPER.

Facilities modeling and integration with MBAL reservoir models and PROSPER completion models is done within GAP. The GAP tool optimizes the production system by maximizing a pre-defined objective function (oil, gas, net present value) within pre-defined constraints. The IPM model has the capability of generating production forecasts while predicting reservoir performance and pressure and temperature throughout the production system. Petex IPM is a very open system that has the capability to link into other applications, such as dynamic MoReS simulation and HYSYS process modeling, as well as being capable of communicating with MS Excel using Visual Basic command language.

A working IPM model has three main components:
•    A description of the surface network infrastructure, including definition of constraints.
•    A subsurface potential of pressure, GOR, and water cut, in principle delivered through an MBAL material balance model. Alternatively, a link to dynamic simulation or production correlations may be used, as in the case for Fahud.
•    Suitable vertical lift and inflow correlations developed with PROSPER, that translate subsurface potential into surface rates.

Surface network
The Fahud surface infrastructure consists of four production stations, each with limited local gas compression capacity. Excess gas is transported from the four production stations to
a central gas compression facility. Compressed gas is either re-injected into the reservoir or used for gas lift.

All wells are connected to one of the production stations, and all wells are connected to one of the reservoir compartments. A subsurface potential is translated into a surface potential using the lift correlations, subsequently the surface network makes an assessment of available production potential and maximizes oil production within the gas compression constraints by choking back or closing in high GOR wells. At the same time, gas lift, mainly targeted at waterflood area wells that are on high basic sediment and water (BSW), is also adjusted to honor total gas constraints as it shares common compression facilities.

For derivation of the subsurface potential, the material balance techniques applied through MBAL are not suitable for a full description of the gas-oil gravity drainage process. It is especially difficult to provide a description of the erratic intermittent production behavior due to gas coning. With the assistance of Petex experts, an alternative route was developed, which demonstrated the flexibility of the software. An Excel-based prediction tool was set up that provides correlations of BSW and GOR as a function of cumulative oil production. For regular producers, these can be based on straightforward decline analysis or derived from dynamic simulation models. For the intermittent stop-cock producers, an analytical description was developed that describes the GOR development as a function of cumulative oil during
a production cycle. Guidelines were defined for minimum closed-in periods after gassing out. Although material balance will not provide GOR and BSW for Fahud GOGD, the intention is to use MBAL for pressure prediction and voidage management.

Lift models

Vertical lift models have been developed for the approximately 180 producer wells. The large well population was split into low GOR wells, high GOR wells, and gas-lifted wells.
Representative candidates with flowing surveys from each category have been used to select the most appropriate vertical lift correlations. Common correlations could be drawn, but the exercise highlighted a significant data gap in either gas measurement or historical THP/choke data, which were also confirmed by choke correlation matching. Ongoing work shows that, with accurate measurements, a correlation between the productivity index and GOR can be derived. Once the correlations had been selected, the construction and matching of the vertical lift models for the large number of wells was facilitated by automated macro spreadsheets.

Data review

For the Fahud project, an export routine has been developed that allows IPM forecast data to be exported to the OFM production database, where the forecast data can be compared
to the historical data. Besides providing suitable displays of production streams, OFM is also powerful in summing production in-well selections, such as by layer or by area.

Implementation results

In its roll out to the field operations, the project initially focused on management of the intermittent stop-cock producers. Based on the analog models, weekly guidelines have been issued to operations indicating which wells are ready to be opened, the wells that require to be closed, and wells that need testing. As a result, the average production of the target stop-cock well population has increased by more than 1,258 b/d (200 cm/d) compared to previous years.

The large production variations give significant measurement uncertainty, but a clear positive impact is observed. The management of the stop-cock well population had been largely carried by a small group of experienced production operations staff, and consequently the field performance depended strongly on availability of these staff. Implementation of the model has introduced structure and clear guidelines to the well management. The implementation required significant effort in communication and integration between petroleum engineering and production operations.

Future work will concentrate on making the model ready for use in generating short-term production forecasts. In order to achieve this, more work will have to be carried out in improving the calibration between historical data and model forecasts. The IPM model shows substantial reduction in gas-lift requirements and significant scope is expected in gas lift optimization, which needs to be followed up by field trials. A HYSYS model for the surface facilities will be constructed and the value of linking IPM with HYSYS and the impact on production optimization and debottlenecking will be established. A decision on gas compression expansion is pending and the IPM tool will be used to evaluate the impact of varying the gas compression constraint.

The implementation of Petex IPM on the Fahud field has been successful so far. It has increased Fahud production through improved management of the intermittent stop-cock producers. The IPM workflow stimulates integration between the disciplines and enforces good practices. Although the IPM model has not yet been used for forecasting, it is considered suitable for short-term forecasting application. The Petex software is very open and flexible, and several automation routines assisted in managing 180+ wells in the Fahud field. Total-time investment to date is estimated at approximately 120 man-days, demonstrating that an IPM model for a large field can be set up in a relatively short time frame. PDO recommends application of the Petex IPM tool; consideration should be given especially for facility-constrained fields.