Petroleum Geoscience - Volume 7, Issue 1, 2001

It is evident that all or most North Sea Chalk fields have dipping oil–water contacts and that some also have dipping free water levels. Furthermore, there are Chalk fields, such as South Arne (Denmark) and Joanne (UK) with steeply dipping oil–water contacts and a large element of non-structural trapping, while other discoveries, such as the recent Halfdan discovery northwest of the Danish Dan Field, have no element of structural trapping and appear to lie on a Tertiary migration pathway. It is also evident that the presence of North Sea Chalk oil, both as fields and shows, is localized and, generally, confined to areas of early source maturation – although there are exceptions, such as the UK Fife and Flora Chalk oil accumulations, which are distant from a source basin.

We believe that these attributes reflect the same cause, that is the very low lateral and vertical permeability (typically 0.1 to 1 mD air permeability) of the Chalk system. Our hypothesis is that oil enters the Chalk from the Jurassic at a limited number of entry points within the early mature basin and, only if the oil reaches Chalk of relatively high permeabilities, normally the Tor Formation, does the oil migrate significant distances laterally. This oil migration occurs at geological timescales, i.e. kilometres over millions of years and we, therefore, have to consider the configuration of the Chalk – with respect to migration routes and traps – at the time the oil was injected into the Chalk and began to migrate through it, which was normally prior to the North Sea Miocene inversion and later tilting. There is considerable evidence for trapping in Chalk palaeostructural traps; oil migrated into a palaeohigh, segregated vertically and is trapped by the low oil relative permeabilities within the transition zone when the trap later tilts. Another trapping mechanism is the migration trap, as described for the Halfdan discovery in Denmark, where, in an area mature for Chalk exploration, oil has been discovered in continuity with the Dan Field but with a free water level 700 ft lower. This type of trap may contain significant reserves of oil in chalks of 1 mD permeability and, in recent years, it has become possible to exploit chalks of this nature commercially. It is also possible to develop these chalks when no effective natural fracture system exists.

Projected future increases in Algerian gas production will, in part, come from more complex reservoirs within the Ahnet-Timimoun Basin; here, conventional quality reservoirs (>1 mD) are interbedded with volumetrically significant, low permeability sandstones (<1 mD) – the ‘tight gas’ sandstones. The challenge has been to develop a programme of work which will establish reserves in this sub-millidarcy resource; on balance, the results of this evaluation programme are positive.

The main tight gas sections are of Lower Devonian age. Grain rimming chlorite cement within thin intervals locally inhibited later pervasive quartz cementation; the quartz is the cause of reservoir degradation in the bulk of the sandstone resulting in permeabilities of <1 mD, and often in the microdarcy range. Generally, the Lower Devonian has a low density of faulting and fractures; open and closed fractures are observed in core, whilst mud losses suggest some fractures are conductive in the subsurface.

One of the main structures, the Teguentour Field, has been evaluated through the drilling of two new wells. The first of these wells (Teg-14) evaluated interbedded tight and conventional sandstones, whilst the second (Teg-15) was a dedicated tight sandstone completion involving an induced fracture programme.

The permeability range in the tight gas sandstones extends below the resolution of conventional porosity/permeability measurement; determination has been improved through mercury injection derived permeabilities. Water saturations determined from core and log suggest gas may be present in the low permeability rock. The presence of gas could not be confirmed by formation tester samples and thus dynamic data were required from the Teg-15 well which was completed within the tight sandstone interval; only the top 5 m was perforated to limit the possibility of a communication pathway down to conventional quality sandstones. After nitrogen lift, the well flowed gas to surface at c. 5 ×104 SCFD for several hours. The low leak-off coeffcient determined suggests there was no connection to a pervasive conductive fracture network, although fractures intersected by the wellbore may have contributed to flow. Also, large poro-elastic back stresses indicate that a connection to high matrix or fracture permeability was unlikely. On balance, the test is a positive indicator of gas presence in very low permeability/porosity rocks which may recharge associated conventional layers following suffcient drawdown.

A new methodology to describe fractured reservoirs by construction of geological models calibrated to well test data is evaluated by case application. The aim of the approach is to construct a three-dimensional (3D) fracture/matrix model that represents the main components of the reservoir architecture. The model allows a close approximation of complex geological features. The geological model is calibrated against well tests to determine the geometrical and single phase flow properties required to reproduce the well test. This process may allow discrimination between different equally plausible geological scenarios and the output of the approach has implications for input to reservoir simulators.

Input to the model includes the seismic-scale reservoir structures, the geological layering and observed intersections with flowing fractures in wells. A structural model is built within this framework in accordance with the geologist’s understanding of the fracture system and geomechanical principles. The structural model is comprised of surfaces, which may represent individual fractures or, alternatively, may represent more complex geological features including fractured layers or fault zones. The matrix is represented either implicitly or explicitly.

A case application is described where three well tests in the same reservoir volume are modelled. All three well tests were matched within the same internally consistent geological model, which previously had not been achieved using standard analytical well test analysis tools. This may illustrate the advantage of utilizing a more realistic geological model. Having established a series of models consistent with both the geological and flow data it becomes possible to identify the key components of the model that must be represented in a reservoir simulation model.

Sub-vertical natural fracturing is critical for ensuring economic gas production from tight sand reservoirs in many provinces. Over the last seven years, the US Department of Energy (DoE) has sponsored a series of programmes in the continental United States to develop cost-effective technologies for investigating naturally fractured reservoirs. The results from a study of a Laramide-age faulted anticline reservoir in the Lower Fort Union Formation of the Wind River Basin, Wyoming are presented. At the field site a multi-azimuth, multi-offset 3D compressional wave seismic survey was acquired and processed using azimuth-dependent processing followed by multi-azimuth attribute analysis. Geophysical attributes such as velocity, frequency, reflectivity and amplitude variation with offset, together with geological attributes such as depth below closure and distance from faults, were correlated to estimated ultimate recovery of gas for established wells within the field. The correlations were made using linear statistical methods, non-linear rank methods and neural networks to combine attribute results. The geophysical attributes and other geological information were combined into prospectivity maps for areas of highest fracture density. Based on the geophysical and geological results an 80% success rate was achieved at prospecting well locations.

A major aspect of the DoE programmes has been the investigation of costeffective protocols for fracture detection. From this study the following procedure is recommended for cost-effective mapping of fractures and associated gas: (1) perform field and remote-sensing reconnaissance of structural trends; (2) acquire 3D P-wave surveys with offsets equal to or greater than target depth in all azimuths; (3) process in at least two dominant structural azimuths; (4) combine best correlated geophysical and geological attributes to fractures for future fracture zone mapping. The implications of the results from this survey have important consequences for the design and implementation of multi-azimuth, multi-offset data acquisition on land and for marine surveys where multi-component ocean bottom cable surveying is used.

The authors have previously reported that gas condensate relative permeability will increase with increasing velocity when conducting steady-state measurements. The increase in relative permeability was referred to as ‘positive rate sensitivity’ or the ‘positive coupling effect’. A systematic series of core tests has since been conducted, generating data at IFT values ranging from 0.015 to 0.78 mN m−1, using core types ranging in permeability from 11 to 350 mD. The results confirmed that the positive coupling effect existed in low permeability cores and different lithologies at low and high IFT. The maximum tested velocity was in the region of 75 m per day, which was estimated to be at the boundary above which inertia would be significant.

To investigate the competition between inertia and the positive coupling effect, subsequent core tests have been conducted at velocities ranging from 7 to 700 m per day, with the higher velocity being in the region of fractions of a metre from the wellbore. The maximum test velocity was one order of magnitude above the velocity boundary with significant inertia. The test programme has generated previously unpublished data, showing that the positive coupling effect initially reduces the effect of inertia before becoming dominant in the core tests conducted using gas condensate fluids. Such unique data provide valuable information on the forces governing production at the wellbore to reservoir engineers.

This paper introduces the Cretaceous Etjo Formation of NW Namibia as a sedimentological analogue for the Leman Sandstone, the principal gas reservoir of the UK Southern North Sea. Special reference is made to the documentation of hard knowledge from the analogue for input into very fine-scale reservoir models suitable for upscaling. The data presented address the sedimentary aspects of reservoir heterogeneity in aeolian sandstones providing size populations for grain flow laminae which typically exhibit permeability of an order of magnitude better than the grain fall and wind-rippled strata which encase them. The results also have significant and more general implications for existing models of aeolian sequence stratigraphy. The data collected are used to demonstrate that there is no link between preserved bedform thickness and original dune height. This has implications for the role of subsidence in controlling preserved bedform thickness.

Petrophysics and geophysics have not always worked together as closely as one might expect. This paper discusses current advances and sketches future developments. The major objective of geophysical research is to obtain sharper pictures. With the advent of 3D seismic, there is a strong trend to make displays both in time and depth and to convert seismic attributes into rock properties. Conversions based on seismic data alone give non-unique results due to limited resolution. Further progress hinges on calibration of acoustic attributes with parameters measured on core. For time-lapse seismic, the situation is even more complex, due to changing effects of temperature, pressure and compaction. All these effects have to be properly quantified if 4D seismic interpretation is to be integrated with dynamic reservoir models to realize the situation where all geoscientists work on one ‘unified’ 3D earth model. Biot–Gassman fluid substitution algorithms were successfully applied for decades, but suffer, like all mixing laws, from non-uniqueness. Recent investigations demonstrate that dispersion in the low frequency band could have strong implications for seismic attribute analysis. It is important to find out which relations measured on core at high frequencies can be used at seismic frequencies and which require additional laboratory work. The shock-tube is well suited to this task, because it handles large rock samples, and obtains responses below 1 kHz.

As part of an effort to try to halt the decline in onshore oil production and accelerate production from onshore fields in Azerbaijan, a World Bank-sponsored study of four fields was carried out: the Bibi-Eybat, Kala, Buzovni-Mashtagi and Zirya fields. The study integrated geoscience, engineering, economics and environmental issues. This paper concentrates on the determination of methods for increasing production rates and recovery. The characteristics of the subject fields are summarized and their current status described. Methods for improving oil production rates in the fields include: mechanical work-overs, artificial lift optimization, sand-control and changes to well completion practices. Our work indicates that production rates could be raised between two-and ten-fold through the application of modern oil-field practices, but financial constraints have prevented their widespread use. Well intervention techniques form the basis of several outline field development plans, described with the associated production forecasts. A brief review is made of possible exploration plays which exist in and around the subject fields.

This paper describes the application of horizontal well geochemistry for the first time. Closely spaced samples, taken along a horizontal or deviated well path, allow the identification of reservoir compartments. Significant variations in geochemical parameters are related to the field filling and, if they persist over geological time and over tens of metres, should reflect barriers to diffusion and, by inference, possibly also to fluid flow.

Two different approaches have been chosen: (a) closely spaced head space gas samples; (b) wet cuttings-based techniques. Case studies are presented documenting the detection of sealing faults between different reservoir compartments at the time of sampling. This novel approach has several advantages: (i) the need for costly pressure measurements is reduced; (ii) high sample density allows better spacial definition of barriers compared to both seismic and pressure measurements; (iii) detection of sub-seismic barriers; (iv) it allows detection of compartments even in cases where no pressure differences exist; (v) the technique works in wells drilled with oil-based mud systems. The strong variations in the geochemical signal over short distances has implications for the filling of reservoirs. The data suggest that reservoir filling is far more complex than previously envisaged.

Blenheim is a small Palaeocene oil field (c. 53 × 106 BBL oil in-place), in which significant seismic uncertainty complicated a development decision. Initial appraisal plans required a conventional deviated well sited to minimize the uncertainty, gather further reservoir and fluid data, and, if successful, provide a production well.

Later studies indicated improved economics if horizontal wells were successful. Several well designs, with and without pilot holes, were analysed to discover the optimum balance of data gathering, risk reduction and development cost reduction. The optimum well required accurate well steering in an area of seismic uncertainty. A high-resolution biostratigraphic study was made and provided detailed correlation of intra-reservoir mudstones, enabling wellsite analysis to aid steering decisions.

The pilot hole proved the most likely seismic interpretation to be correct and permitted the horizontal well to be drilled as planned. Wellsite biostratigraphic analysis and real-time MWD aided steering decisions and contributed to a successful horizontal well being drilled. The well flowed in excess of the minimum economic rates required and allowed the field to be successfully developed.