This paper presents a case study from Onshore wells in Sharjah, UAE on investigating liquid loading in 5 multilateral gas wells having various trajectories ranging from toe-up, toe-down and hybrid openhole legs. These wells are subjected to wellhead pressure reduction to maximize production rates. The main objective of the study was to evaluate the production performance for different completion designs with respect to liquid loading onset and overall production assessment with declining reservoir pressure. Dynamic multiphase flow simulator was used to conduct this study to accurately capture the details of the multilaterals system and its complex trajectories. The first step involved validating the well model with reasonable history match between the simulation and actual production data. The validated model then was used as a basis for predicting the liquid loading onset point for a given reservoir pressure decline. Multiple cases were investigated to evaluate various completion options (i.e. with or without tubing) to determine how and when the liquid loading occurs at different laterals with varying lateral trajectory. This study has showed that in such complex multi-lateral wells, laterals load up at different points in time and reservoir pressures, being affected mainly by the geometry and orientation of lateral and the production contribution. Moreover, installing tubing in these wells had the opposite anticipated effect on liquid loading by accelerating the liquid loading onset in the laterals due to the imposed additional restriction. Generally, toe-down trajectory tends to have thicker liquid film and a potential for reduced flow contribution due to liquid accumulation at the toe. These wells have a fishbone openhole multilateral network with comingled flow in the vertical section. It is observed that production tubing in the vertical section provides friction that accelerates the onset of liquid loading and hence results in decreased production for wells operating in very low reservoir pressure range. Based on overall production assessment ‘no tubing’ scenario would be more beneficial. Further, the timing of implementation of the tubing restriction later in the field life can be selected based on dynamic simulations (also evaluating economic constraints vs production gain). Transient mechanistic flow model captures the liquid loading phenomena by film reversal which usually occurs before the critical rate limit based on droplet drag forces assessment. Further, liquid loading onset occurs in the laterals first rather than the tubing section which reduces the applicability of conventional nodal analysis tools. Evaluating liquid loading behaviour in such multilateral wells with proper dynamic simulation is critical for understanding the laterals behaviour and therefore optimizing the production performance to maximize the wells uptime and ultimately the overall gas recovery as well as optimal usage of CAPEX.

SNOC embarked on an ambitious project to extract maximum value out of its mature fractured carbonate fields by converting them to underground gas storage facilities. This required the integration of a multitude of new and legacy data including new seismic acquisition, advanced processing and interpretation along with geophysical modelling, PVT, petrophysical and injection/pressure profiles to develop a robust reservoir model for the Moveyeid gas-condensate field. Enabling the operator to use this tool for advanced gas storage simulation predictions and to quantify risk in field development strategy to maximize condensate recovery while minimizing cushion gas volumes. Various seismic frequency cubes and survey vintages were used to delineate the Moveyeid structure and map top Shuaiba formation within the Thamama carbonates of Onshore Sharjah. Legacy wireline logs were reinterpreted using new multi-log workflows, providing inputs for static model development. As a mature field, in production for over 35+ years, cumulative produced volumes were used as an additional control, creating several iterations until achieving a match that honoured the geology. PTA, RFT and PLT were integrated in the model to quantify flow change over time. Advanced imaging revealed a north-South trending normal fault that segmented the field in to two, with all existing wells located on the up-thrown eastern block. Volumetric determination revealed that the eastern block alone was not sufficient to match hydrocarbons produced to date in any static scenario modelled. This was achieved with the inclusion of the western block and an unchanged gas-water contact. The new static model developed is more robust, with an enhanced layering configuration and property arrays that better reflect input data. Benefits were also seen during dynamic simulation where lower property multipliers were applied during history matching. Optimizing the layering and using block parallel computing power enabled the team to considerably reduce runtimes and produce an array of scenarios. The model was put to the test when the gas injection pilot project was commissioned in 2017 with low-pressure injection up to 2020, yielding a well performance accuracy within 5% of actual rates. Optimizing the strategy was essential in reducing planned cushion gas requirements by up to 20% for meeting the production target in the depleted reservoir, improving the project's capex. Utilising a combination of new and legacy data, a depleted gas-condensate carbonate reservoir has been successfully modelled. The model is being used as a tool to formulate the strategy and effectively define the field's suitability for gas storage and enhanced condensate recovery. This paper provides a case study for how these strategies can be implemented in other Middle Eastern analogues where gas storage in mature fields can act as a strategic tool for energy security.

Sharjah National Oil Corporation (SNOC) currently operates 3 fractured carbonate mature gas-condensate fields with some 35 years of production history. Until recent years these fields were operated by leading International Oil Companies (IOCs) which utilised some of the then latest technologies, such as underbalanced coil tubing drilling in order to maximise the production rate. The reservoir development and management scheme, however, did not involve gas re-injection to maintain reservoir pressure above the dew point. This led to production by simply blowing-down the field. Since there was negligible aquifer support the reservoir pressure declined rapidly and the dew point pressure was reached within 3 years, resulting in condensate drop-out in the reservoir. It is estimated that more than half of the original condensate in place still remains in the reservoir, although more than 97% of the gas in place has already been produced and the reservoir pressure have declined to around 10% of initial pressure. In order to determine the location and quantity of condensate remaining in the field, dual porosity reservoir models were created with legacy data which replicated the naturally fractured reservoir. These models were history matched and gas injection simulation runs were performed in order to estimate the injection rates, reservoir pressure increase, field communication and potential for condensate re-vaporization and mobilisation theory at a variety of pressures. This theory was put to test and confirmed when SNOC recently performed a pilot gas injection project in one of its fields. A mixture of processed gas from the gas plant was injected and allowed to stabilise. The new mixture of injected and reservoir gas was reproduced to estimate the deliverability and ability of dry gas to vaporise the in-situ condensate. A fundamental challenge with SNOC was to determine the PVT property of the initial reservoir fluid from a surface recombined sample which made it extremely difficult to decipher the original fluid properties and history matching the reservoir model. Utilising the field for gas storage can help elevate the reservoir pressure and increase the vaporisation of condensate, however since the field is naturally fractured it is susceptible to the injected gas fingering into a producing well. SNOC now plans to continue the next phase of the project to mature the modelling work, evaluate various sources of injection gas, understand the project uncertainties and establish the conditions required for the ECR project to be economically viable. This paper discusses the challenges, observations and its conclusion through the pilot gas injection project and its impact on the decision making for large scale implementation of enhanced condensate recoveries in the Middle East. Maximizing field development objectives by combining various opportunities is the key to determining sustainability in the lower oil price environment. This paper demonstrates how new technology combined with large volumes of legacy data can provide the perfect platform to evaluate the potential for enhanced condensate recovery (ECR) projects and take informed decisions for operators.

Sharjah National Oil Corporation (SNOC) operates three onshore reservoirs in the Emirate of Sharjah. The reservoir simulation models use compositional modelling to capture the fluid dynamics in mature, low porosity highly fractured gas condensate fields. The scope of this project was to improve the reservoir characterization by investigating and overcoming lack of water production in compositional models for effective EOR and gas storage strategies. Water cut of 30%+ comprised of a combination of produced and condensed water in a reservoir with no active aquifer, thus posing a modelling challenge combined with a lack of comprehensive historical PVT data. All existing PVT reports in the database were retrieved and a comprehensive quality check was performed. The best possible PVT results for each field were short-listed and taken as reference datasets for validating the compositional EoS in a depleted field. A new EOS was generated for these fields based on legacy PVT data combined with 38+ years of production data. A shortfall of this new EOS was the inability to produce condensed water as observed in the field with Chloride counts less than 1500 ppm. To rectify this low water production mismatch, a blind test was conducted introducing water as a component in the EoS in the simulation model to see the effect. Moreover, extensive scale problems in any of the wells of 30-year-old mature assets leading to regular interventions never occurred in the asset's operational history. As expected, mobility of the fluids in the system had changed and low salinity condensed water was seen to have a good match. Liberated water was traced at the surface to confirm water production rate of the same order of magnitude as observed in production data. Due to overwhelming water production rates from the trial test, SNOC decided to perform a comprehensive extended PVT study. The naturally fractured carbonates were subjected to geological and material balance study and the data indicated an absence of active aquifers, which made it difficult to match observed water production in simulation models. To effectively plan future EOR projects like gas storage, it was necessary to model the effects of water and its interaction with injected fluids in the reservoir while honouring low water movement in the subsurface. The paper provides a novel workflow for generation of the compositional equation of state with water as a component in retrograde condensate fields. The workflow followed the lumping of hydrocarbon components to minimise runtime and capture maximum possible fluid dynamics in the reservoir without compromising the fluid properties observed in the PVT lab. It was also vital for the simulation model to honour the production history spanning over three decades. It also highlights the ability and importance of including water as an EOS component to effectively capture the condensed water in the reservoirs that many works of literature and simulators are unable to provide insight on.

Objectives/Scope This study aims to use modern techniques to re-characterise the diagenetically altered Thamama Group reservoir units of multiple gas-condensate fields in Sharjah, UAE and determine robust rock-typing framework from the full dataset and recent core analysis program. This would be used to reduce mismatches observed in static and dynamic properties and demonstrate that a matched-outcome can be achieved with less model manipulation by focusing on textural variances within the units. Methods, Procedures, Process Results, Observations, Conclusions Four petrophysical rock types were identified and found good equivalence to the identified petrographical rock types; the algorithm separated mono-modal micritic packstones from highly diagenetically altered grainstone-wackestone rudstone facies, with the rock-type clusters also being defined by Winland-r35 and Lucia poro-perm threshold lines. A single rock-typing framework, suitable for all studied fields with observed differences being explained by variability in the rock type proportions. When compared to the previous rock typing framework and reservoir models, better matches were achieved between predicted properties and core data in QC wells. Static model property distributions were more realistic in achieving a volumetric match with produced gas. Better saturation distribution with realistic Swcr and Socr were observed by using the new Rock Typing Sw equations. Rel-perm modification for increasing water production to match the observed data was negligible due to presence of more water saturation in the crest of the reservoir. Multipliers for permeability and porosity were significantly reduced to match the well productivities and tubing head pressure estimations were improved due to less mismatch with liquid production rates. Novel/Additive Information This work represents the first time petrophysical and petrological rock typing was conducted for several gas-condensate fields in Sharjah, UAE. Newly acquired core data, petrographical information and core descriptions were integrated in the study. The previous workflow, established in 1993, was updated using modern machine-learning techniques incorporating new data and a wider range of data than the previous rock typing model that was based solely on porosity measurements, remaining consistent to pore-scale and textural changes.