TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractWell-test model identification and, subsequently, model parameters determination are more complex in horizontal wells than in vertical wells. This is due to the increase in number of flow regimes occurring during a flow period and the fact that strong correlation exists between model parameters.This study presents a new approach for automatic model identification and computer-aided well-test interpretation in horizontal wells. The new approach is based on using neural network to (1) identify the well-test interpretation model; (2) identify flow regimes; and (3) mark the position of identified flow regions on the derivative plot of well test data.This work consists of first generating common model signatures using Ozkan and Ragavan analytical solutions for horizontal wells in various reservoir and inner boundary conditions assuming laterally boundless reservoirs. Next, these signatures are used to train neural networks for three identification stages; model identification, flow regime identification, and position of flow regime identification. Separate networks were trained, then tested and validated using synthetic as well as field data. Once the three identification stages are completed, specialized plots for data points falling into each flow regime are used to determine initial model parameters. Final model parameters are determined using nonlinear regression.A comparative study was carried out using different network architectures. Modular approach with direct data utilization is found to be most suitable for implementation of our approach.

Fluid identification is an important step in reservoir characterization and hydrocarbon volume estimation of an exploratory target. Volatile oil / Gas condensate reservoirs are well known for complex behavior due to near critical fluid nature. Objective of this study was to identify fluid nature in a low permeability reservoir where a single phase PVT fluid could not be obtained. This study describes a carefully designed analytical and interpretive workflow that utilized analyses of the separator samples and integration of the data from G&G sources to discern fluid nature. This paper describes the methodology of identifying fluid nature of collected separator samples from ADCO's exploratory reservoir in a field that has potential to add significant reserves to its portfolio. Due to high pressure encountered during drilling of the exploration target reservoir, Wireline Formation Tester (WFT) pressure gradient and bottom-hole samples were not acquired. Wireline logs were acquired after several weeks and consequently did not help in distinguishing the hydrocarbon type. Production test conducted across the 6 ft thick top layer resulted in 49° API oil with GOR varying from 3000 – 4000 scf/bbl. Three (03) sets of oil and gas separator samples were acquired and physical recombination was performed in the PVT laboratory at reservoir conditions using two different GOR values. Composition analysis of separator liquid was also performed and compared to the same reservoir in analogue fields at comparable depth. Physical recombination at GOR values of 3000 and 3759 scf/stb exhibited oil and gas-condensate behavior at respective GORs. Composition analysis of separator liquid also indicated significant amount of heavy components (C7+: 65.5% and C12+: 23.6%). Reservoir compositional data was plotted and compared to similar oil / gas-condensate reservoirs from nearby fields. Fluid was identified to be of volatile nature based on PVT study as bubble and dew points pressures were very close in recombined samples. The mud log data showed high total gas reading of 17% in the zone of interest. An assessment of hydrocarbon phase at this high total gas reading based on Wetness (Wh) and Balance (Bh) ratios already indicates presence of light oil. Further PVT work including an attempt to acquire WFT sampling of a single phase fluid is recommended in appraisal wells to thoroughly understand the near critical hydrocarbon phase in the reservoir of interest.

Kadanwari field in Middle Indus Basin (Pakistan) was discovered in 1989 and brought on stream in 1995. The producing reservoirs are Cretaceous Lower Goru sands D-E-F-G. The gas production started from better quality E and F sands; after 2004 layer G started to drain western block of the field, with the first hydraulic fracture job made in Pakistan (well A). Layer G represents a complex target for petrophysical characterization; reservoir sandstones are micro-porosity rich, with variable presence of Chlorite affecting flow properties. Positive results encouraged the operator to drill & frac well B and to consider possibility to extend gas production throughout western block, including sand reservoirs of variable quality, from moderate to tight. The paper describes how reservoir study faced layer G complexity and how production data of wells A and B allowed a post fracjob evaluation integrating well-test data and frac-job interpretations into 3D dynamic model. After history match, the computed GOIP suggested an infilling program in G sand reservoir, with side-tracks of existing wells and new wells, all hydraulically fractured. So far, one sidetrack and one new well have been drilled; results fully confirmed the complexity of local geological setting. The sidetrack revealed rock quality slightly better than expected (frac not necessary). Pilot well C targeted G-Sand in a sweet seismic anomaly in western area, a gas flare was observed during DST pre-frac. Mini-Fall Off was conducted to estimate closure pressure and effective mobility, but permeability computed from MFO was not conclusive due to important filtrate invasion. DST post hydraulic fracture job confirmed commercial gas rate production higher than 1 MMscfd with a peak of 3.5 MMscfd. The successful pilot well results open new horizon to improve reserve from tight sand of Lower Goru formation.

Generally, appraisal wells are drilled to reduce uncertainty. However, occasionally reserves uncertainties may increase in a heterogeneous carbonate reservoirs specially challenging stratigraphic limit of reservoir facies. Under such circumstance, sometime operators rethink of further investment in the field development when in-place volumes are marginal. The objective of the study is to present how we achieved well design modification and test strategy in a dynamic environment. Optimal well test design, execution and analysis can help mitigate major uncertainties, which were not considered during initial planning phase The subject appraisal well was drilled as a vertical hole in an up dip direction to the first appraisal well. However, Open Hole (OH) and mud log data indicated the reservoir to be tight and in some portion dominated by water flow during sampling even though clear hydrocarbon presence observed in core chips and cuttings analysis. After detailed studies of the available data, a decision was taken to horizontalize the well towards first appraisal well. While drilling, geological barriers were encountered as indicated by the presence of different fluids in the horizontal section. Variable fluid presence (water and oil) posed a challenge with respect to well completion and testing. This paper describes the process of completing the well in an evolving complicating situation and how successful well test design and execution helped to mitigate the uncertainties. OH Logs, Wire Line Formation Tester (WFT) and test data from the studied and existing wells in the area were used to design the well test and interference with first appraisal well in an evolving situation, which is not typically faced in well operations. Hence, the results obtained provide an additional information that helped to conclude variable fluid distribution and its dynamic connectivity to the first appraisal well. Well was completed followed by test as designed and Production Logging Testing (PLT) was conducted to define reservoir contribution. Post well test analysis and comparison with existing WFT and test data from existing well helped to conclude results and address the uncertainties. This paper summarizes the design process, challenges faced in an unexpected variable fluid distribution in the horizontal section and accordingly how well test analysis was performed to conclude the results that helped to take optimal investment decision for the development of this marginal reservoir.

One of the main challenges in exploration / appraisal phase is to measure the reservoir pressures and sample down-hole single phase fluids to identify the fluid types in tight layers (permeability <1md). This challenge escalates when multiple reservoirs are exposed in same open-hole with varying differential pressures across them. A novel wireline formation tester (WFT) technique was applied in an open-hole condition to collect in-situ representative fluid samples faster and cost effectively. It also significantly reduces the risks associated while sampling across multiple reservoirs in comparison to the conventional techniques. Conventional approach of using dual packer module requires longer set-unset time as well as pumping out the trapped volume between the packers before receiving fluid from the formation. On the other hand, the probe module is more effective for sampling in high permeability layers. In tight reservoirs, the fluid sampling is very challenging due to low formation withdrawal rates, high drawdown and elongated station times. These challenges were overcome with a radically designed WFT that utilizes a 3D elliptical probe module with an optimum reservoir contact area. These probes enable circumferential flow from the formation with a faster cleanup process and hydrocarbon breakthrough. The new sampling technique showed an improvement, over the conventional methods in several aspects: (1) Tight zone pressures are obtained faster with lower supercharging effects. (2) In-Situ representative fluid samples are acquired in tight reservoirs above saturation pressure enabling proper fluid characterization PVT studies. (3) Stationary times during sampling are reduced due to negligible trapped volume. (4) Set and retract timing is shorter so that a new sampling method of retract-move-reset is developed to minimize the mud invasion between multiple settings. (5) Reduces the differential sticking risk in longer operations with multiple reservoirs exposed in the same open-hole. (6) Minimizes the sampling runs resulting in significant cost saving. Logging operation proved to be successful with the new sampling method by acquiring representative fluid samples in tight formations, which was not possible earlier. This paper describes a case history and recent achievements made in acquiring representative single phase fluid samples in tight reservoirs, overcoming the challenges and risks associated with conventional sampling techniques.