BP's Khazzan-Makarem (KM) appraisal project is located in Central Oman. BP committed to appraise four deep tight gas reservoirs and has drilled seven wells to date. An Extended Well Test (EWT) facility is designed to provide a long term (multi-month) flow test of those wells.Basic information relating to trap, seal, areal extent, and Gas-in-Place (GIP) is not the most significant problem in this appraisal project; the most important issue is long term well performance. The tight gas nature of these reservoirs results in considerable uncertainty in prediction of future deliverability. It is not uncommon for tight gas wells to achieve excellent flow rates after fracture stimulation, yet decline precipitously over a short period of time (typically months) once on production. Prediction of well performance based on this early flow rate is a highly unreliable indicator of long term performance. This can result in drastic miscalculations for Full Field Development (FFD) planning and economics.Extensive surveillance activities are critical to evaluate this uncertain behavior. An examination of stimulation, well testing and Pressure-Build-Up (PBU) analysis in these wells provides the starting point for understanding well deliverability.BP's KM appraisal wells were planned to selectively target reservoirs addressing variations in reservoir quality and continuity. Completions were designed from the start, to facilitate extensive stimulation programs. Surveillance programs were constructed with permanent down hole gauges (PDHG) to provide critical real time information about the stimulation efforts, well testing, fluid composition analysis, and extreme long term (~ 1 year) PBU. All of this effort provides guidance in planning the EWT facility for an extended flow testing campaign which will feed into FFD planning. Achieving commercial rates in tight gas reservoirs in Oman has proved to be challenging. However this comprehensive appraisal and surveillance program is providing a growing level of confidence in the development potential of the Khazzan-Makarem project. Surveillance IssueValue Solution Implemented Areal extent of hydrocarbons -establish minimum commercial volume area Resource magnitude, development area for FFD, possible future well spacing, processing plant location Drill in a pattern that maximizes value issues; in this case, focus on pay quality rather than GIP.

This work presents the use of pseudoisation to account for faults using practical representations of the multiphase behaviour of fault rocks within production simulation models The work starts by revisiting the conventional representation of fault sealing capacities in simulation models using transmissibility multipliers (TMs). TMs, as proposed by Manzocchi et al. (1999) are absolute and phase-independent values that are used to take into account the presence of fault-rocks within production simulation models and are calculated based on the fault thickness and absolute permeability, as well as the size and properties of the undeformed grid-blocks adjacent to faults. The crucial limitation of TMs is that they do not take into account the multiphase behaviour of fault rocks, whose relative-permeability and capillary-pressure curves can have very different characteristics to their juxtaposed host-rock curves. This paper presents and tests a practical representation of the multiphase behaviour of fault rocks within production simulation models using in-situ-generated pseudofunctions. The generated dynamic pseudofunction curves are attached to the upstream cells of the faults using an analogous method to that proposed by Manzocchi et al. (2002). This method takes into account in-situ fractional flows and compartmentalised phase pressures across fault faces between non-neighbour connections. The proposed method gives very promising results when applied to a 3-D, two-phase model with fault-rock thicknesses ranging between 0.5 and 3 ft, and typical fault rock permeabilities (0.1–0.01) mD Introduction Previous work, see Manzocchi et al. (2002); Fisher and Knipe (1998; 2001), has shown that the way that transmissibility multipliers, TMs, are conventionally used in simulating faulted reservoirs does not take into account the multiphase behaviour of fault rock. This static treatment of faults can lead to serious errors when predicting some parameters in a simulation study. This limitation in fault treatment can, however, be improved by using ‘fault-rock's pseudo-curves in cells adjacent to faults in simulation models. One of the least known parameters that is required for pseudo-curves generation is the across-fault Darcy velocity. Tests have shown that using a simple two-cell model, separated by fault material, where this parameter and other various parameters that affect pseudo-curves can be simply tested within reasonable limits. Such tests can provide a reference library of fault pseudo-curves for reservoir engineers to routinely model the two-phase effects of fault rocks in conventional simulation models. However, due to the large number of variables, considering both fault and host fluid and rock properties and their inter-dependence, this library can be huge, yet feasible to construct (Manzocchi et al., 2002). Furthermore, this library requires an external program, that is directly linked to the simulator output to automatically assign the correct, or approximately the correct, fault pseudo-curve to each cell adjacent to faults based on the rock properties, fluid properties, and geometries of both the fault and host rocks. Although not completely covered under the scope of this contribution. This paper presents a more practical approach of fault pseudo-curve generation in a faulted simulation model. As is the case in any study, measured data is a key element in this approach. In recent years far more data has been collected on the structure of faults (e.g. Harris et al., 2003) and the fluid flow properties of fault rocks (Fisher and Knipe, 1998; 2001). Such information allows a more sophisticated, geologically reasonable, treatment of faults in production simulation models.

For many years it has been common practice to adjust fault transmissibility multipliers within production simulation models to achieve a history match without any scientific justification. In effect, this often means that faults are made 'scapegoats' to compensate for inadequacies in reservoir characterisation. In recent years it has become increasingly popular to calculate geologically-realistic transmissibility multipliers based upon measurements of absolute fault permeability and fault rock thickness. A key problem with this method is that it does not take into account the multiphase flow properties (relative permeability and capillary pressure) of fault rocks. This is hardly surprising as the multiphase flow properties of fault rocks are still largely unknown. Here we present measurements that show that under reservoir conditions cataclastic fault rocks may often have maximum gas relative permeabilities that are over two orders of magnitude lower than the undeformed reservoir sandstone adjacent to the fault. Incorporating the multiphase flow properties of faults into production simulation models is still challenging as their static and dynamic properties vary significantly compared with the undeformed reservoir. We review different existing methods for incorporating the multiphase flow properties into simulation models, and we recommend some possible approaches for treating faults that improve on the existing knowledge and software. Introduction Fault rocks often have a significant impact on fluid flow within petroleum reservoirs. Until recently, only the absolute permeability values of fault rocks had been measured. Recently, it has been suggested that in some reservoirs it may be beneficial to take into account the multiphase fluid flow properties of fault rocks (i.e. capillary pressure and relative permeability) in simulation models (Fisher and Knipe, 2001; Manzocchi et al., 2002; Al-Busafi et al., 2005; Fisher, 2005; Al-Hinai et al., 2006). The extent to which this can be undertaken is, however, limited by the fact that there have not been any robust studies of the relative permeability of fault rocks. Here we attempt to fill this knowledge gap by presenting new relative permeability and capillary pressure measurements for fault rocks. We also discuss how these results can be incorporated into a production simulation model to dramatically improve the history match of the production data. We begin by describing the fault specimens used in the analysis as well as the experimental techniques employed. The experimental results obtained for these cataclasitc fault rock samples are then presented. A generic study of how to incorporate these results into production simulation is then presented to demonstrate the importance of accounting for the multiphase flow properties in production simulation models. The existing methods for treating faults in production simulation models are then reviewed and two alternative approaches are described and evaluated. Fault Location and Geology The extensional Lossiemouth fault zone, which lies on the southern margin of the Moray Firth, cuts through the Late Permian/Early Triassic Hopeman Sandstone exposed in the Clashach Quarry near Burghead in north-east Scotland, Figure 1. The main Lossiemouth fault slip plane trends E/W to WSW/ENE and dips to the south. The main phase of faulting probably occurred during the Late Jurassic development of the Inner Moray Firth rift. The Hopeman Sandstone is a clean, high-porosity, yellow-brown sandstone of predominantly aeolian origin, which lies unconformably on Devonian sediments of the Orcadian Basin. It is approximately 70 m thick in this area, and in general dips at a shallow angle to the north.