Abstract
The Ensign Gas Field is located in the Sole Pit basin in the Southern North Sea. The reservoir is the Rotliegend Group Leman Sandstone Formation of Lower Permian age and comprises sediments deposited in an arid continental environment. The main gas-bearing interval in the field consists of sabkha, waterlain and minor aeolian sands, reflecting deposition in an erg margin/lake margin setting. The poor primary reservoir quality of these sands has been severely reduced by extensive illite cementation resulting in average air permeabilities of <1 mD. Attempts to develop the field economically utilizing fracture‐stimulated vertical wells has met with mixed results, with flow rates of 14 MMscf per day and lower being measured during testing. The most recent appraisal well drilled on the field was a long horizontal well that was stimulated with five hydraulic fractures resulting in an economic flow rate of 44 MMscf per day. Analysis of the core and log data acquired during the appraisal of the field has shown that the reservoir contains a heterogeneous distribution of fractures, faults and micro-faults. The fracture population is dominated by conductive north–south striking fractures, with subordinate NNW–SSE resistive fractures and NE–SW mixed fractures that are arranged in clusters, with zones of high and low fracture density. Well results to date suggest that the NE–SW open and partially open fractures observed in core do not improve reservoir productivity, but those orientated north–south that are conductive appear to improve well deliverability.
The Ensign Field is located 90 km NE of the Norfolk coast in the Sole Pit area of the Southern North Sea (Fig. 1). The field is located 10 km east of the Shell-operated Barque Field and 15 km west of the Venture-operated Audrey Field. Ensign is an elongated NW–SE trending field 2 km wide and 8 km in length (Fig. 2). The water depth over the field is 90 ft, and the top reservoir is c. 8450 ft TVDSS at its crest.
Map showing the location of the Ensign Field and Sole Pit area.
Ensign Field top Leman Sandstone depth map based upon 3D data showing major faults and NNE–SSW lineaments. Line A–B shows the location of the seismic section in Fig. 4.
The reservoir is the Leman Sandstone of the Lower Permian Rotliegend Group, and is c. 800 ft thick, but only partially gas-filled, owing to the depth of the spill point. The gas is very dry, containing 96% methane and small amounts of CO2 (0.5%) and N2 (1%). The gas in place volume for the Ensign field is c. 350 BCF.
The Ensign Field name was assigned while under the ownership of Shell and Esso, and reflects the practice of naming gas fields in the SNS after vessels and nautical terms from maritime history. The ensign is the flag or colour of a ship and came into use on the ships of Queen Elizabeth I in the 1570s, and imitated regimental flags, which at that time were called ensigns.
Field history
The Ensign Field is located in Blocks 48/14a and 48/15a that are covered by the P008 and P130 Licences respectively. The P008 licence was awarded to Shell and Esso in 1964 as part of the first licensing round, while P130 was a fourth round award to Conoco in 1972. Venture acquired 100% equity in the Ensign Field as a consequence of two separate transactions involving Shell UK Limited, Esso Exploration and Centrica Resources Limited in 2005.
The Ensign Field wells are listed in Table 1 along with information on well operator and well type. The field was discovered in 1986 by the Shell-operated 48/14-2 vertical well that encountered a 542 ft gas column in the poor quality Lower Permian Rotliegend sandstone reservoir (Fig. 2). Although the reservoir failed to flow during initial testing, after stimulation by hydraulic fracturing the well flowed at a sustained rate of c. 14 MMscf per day. This rate was not economic at the prevailing gas prices and consequently the field was not developed. In 1988 Conoco drilled a well 48/15a-6 in the eastern portion of the field, but the reservoir was encountered deep to prognosis predominantly in the water leg and was consequently not tested. The next appraisal well to be drilled was the Venture-operated well 48/14-5 that spudded in September 2006. This well was planned to appraise the reservoir and consequently had a full logging suite including borehole image logs and was cored. After fracture stimulation the well tested at rates of up to 20 MMscf per day, but this declined to 12 MMscf per day and the well was suspended as a potential future production well. The last well drilled in the field was spudded in September 2007 and was as a horizontal production well. The well encountered over 4500 ft of net sand, was stimulated by five hydraulic fractures and flowed at a facilities-constrained rate of 44 MMscf per day. The well has been suspended as a future production well.
Ensign well data
Structure and tectonic setting
The Ensign Field is situated in the Southern Permian Basin in the NW–SE trending Sole Pit Basin (Fig. 1). The area has been subjected to a number of tectonic events, with the Hercynian orogeny in the Late Carboniferous resulting in the pervasive NW–SE structural grain (Glennie & Underhill 1998). The main phase of development of the Sole Pit Basin occurred in the Permian, when thermal subsidence occurred. During this period the Lower Permian Rotliegend continental sediments were deposited that form the reservoir of the Ensign Field. Basin subsidence continued from the Triassic to Late Cretaceous times by which time the reservoir of the Ensign Field would have been buried to a depth of >13 000 ft TVDSS. The next stage of the structural history was successive periods of uplift during the early Cenozoic with a major phase of inversion in the Miocene related to the Alpine orogeny. The inversion in the Sole Pit during this period removed approximately 5000 ft of post-Liassic sediments and has resulted in the Ensign reservoir being at its current depth of c. 8500 ft TVDSS.
Stratigraphy
The stratigraphic sequence encountered in the Ensign Field is shown in Fig. 3. The oldest rocks penetrated by any of the Ensign Field wells are sands, silts and shales of Carboniferous age.
Ensign Field stratigraphic sequence.
The Lower Permian Leman Sandstone rests unconformably on the Carboniferous and consists of a variety of facies associations, which have been interpreted as the products of deposition in fluvial (wadi), aeolian, sabkha and lacustrine environments (Glennie 1972). The Leman sandstones are overlain by a c. 550 ft thick Zechstein sequence comprising cyclic evaporites and carbonates that acts as the seal for the Ensign Field. The Zechstein deposits reflect the marine transgression into the Southern Permian Basin (Glennie & Buller 1983). Overlying the Zechstein are the three units of the Triassic Bacton Group: the Brockelschiefer, Bunter Shale and Bunter Sandstone that represent a return to continental sedimentation. In total the Bacton Group is c. 2000 ft thick over the Ensign Field. Subsequent deposition of the Middle and Upper Triassic Haisborough Group marks a return to marginal and fully marine environments with the occurrence of c. 2800 ft of evaporites and mudstones. Overlying the Triassic is a c. 2500 ft section of Lower Jurassic Lias shales with a thin veneer of Quaternary–Recent deposits beneath the sea bed. The absent Jurassic–Cenozoic section was removed during the Cenozoic inversion of the Sole Pit area.
Reservoir sedimentology
The Leman Sandstone reservoir of the Ensign Field was deposited in a continental desert environment (Glennie et al. 1978) and has been divided by Venture into six reservoir zones (A–F) that can be correlated across the field.
The Upper A unit has been further subdivided into the A000 unit that corresponds to the Weissliegend at the very top of the Rotliegend. This unit comprises massive sands that were reworked as a result of the Zechstein marine transgression (Glennie & Buller 1983). In the Ensign Field this unit is extensively cemented and of very poor reservoir quality. The A100 lies below the Weissliegend and is dominated by sabkha, sandsheet and waterlain sandstones that were deposited in an erg margin–lake margin environment.
The B unit sandstones are of slightly better reservoir quality, reflecting the presence of more fluvial deposits, less sabkha sands and occasional aeolian units. The B unit sandstones represent the deposits of a less humid environment with less lacustrine influence.
The C unit underlies the B unit and is composed of 10–20 ft thick muddy sabkha sandstones and siltstones, and playa lake claystones. This unit is correlatable across the Ensign Field and over several kilometres in the Barque Field (Sarginson 2003). The C unit represents an expansion of the Silverpit lake, with the muddy sabkha deposits reflecting a more humid climate. The C unit is an important unit in the Ensign Field as it occurs just above the GWC in the central part of the field and acts as a natural barrier to the propagation of any hydraulic fractures owing to its high clay content.
The D and E units are characterized by large-scale cross-bedded aeolian sandstone units thought to have been deposited in a drier erg interior setting. These units have the best reservoir properties, with average porosities of 15–20%, although in the Ensign Field these units are predominantly located in the water leg.
The lowermost F unit comprises sandsheet, sabkha and waterlain sandstones deposited in an erg margin setting. This unit is entirely in the water leg.
Although the reservoir quality of the Leman Sandstone units can be related to their primary environment of deposition, with the aeolian units having higher porosities than the sabkha sands, extensive illite cementation has greatly reduced the permeability of the sandstones. In the A and B units, the main gas-bearing intervals in the Ensign Field, the average net pay core porosity is 9.8% and the average permeability is 0.15 mD.
Reservoir structure
The Ensign Field is situated at the crest of a NW–SE trending dip-closed anticline that is fault-bound to the SW (Fig. 2). The anticline was formed by transpressive tectonic stresses during the Cenozoic, which resulted in the widespread inversion of the basin (Glennie & Underhill 1998). The field is cross-cut by NNE–SSW trending lineaments that may be faults, but have not been cored and show no offsets on seismic data. These lineaments have been recognized in the nearby Barque Field (Sarginson 2003) and by analogy it is likely that cataclasis and mineralization may have resulted in these becoming permeability barriers, potentially compartmentalizing the reservoir.
Seismic interpretation
The original seismic volume used for the seismic interpretation of the Ensign Field area was the Galleon 3D that had been acquired by Shell in 1992. Analysis of the data revealed that the quality of this 3D data was not suitable for the planned appraisal and development of the field. It was therefore decided to undertake a PreSDM reprocessing exercise over the field and surrounding area utilizing the input from five different seismic volumes. The resulting 3D volume allowed a more detailed and coherent structural interpretation to be undertaken.
The overburden interval over the Ensign Field is relatively simple (Fig. 4). The Sole Pit inversion has removed the Cenozoic and Chalk units, resulting in the Lias being the first pre-Quaternary interval encountered below the sea bed. The Triassic section is very continuous and devoid of any large faults that occur in other parts of the Ensign–Audrey Field area to the NE. The Zechstein section is significantly thinner than the regional average which has been caused by halite withdrawal that fed the Audrey salt wall to the east. Consequently, the Zechstein over Ensign is dominated by a dolomite and anhydrite rich sequence that creates a higher than normal interval velocity.
Seismic section showing relatively simple overburden above the Ensign Field.
Utilizing the initial discovery and appraisal wells to constrain the depth conversion, and the benign nature of the overburden, the subsequent appraisal wells have encountered the reservoir on prognosis.
Field appraisal Phase I
After acquiring the Ensign Field, Venture began an extensive data gathering and interpretation exercise in order to determine the best method of appraising and developing the field as quickly as possible. As mentioned above, part of this exercise involved the re-processing of the 3D seismic over the field that was part of a larger project that covered most of the acreage in the Venture-operated Audrey Field area. It was realized that the seismic reprocessing would take a significant amount of time and that this should not hold up the appraisal of the field. Consequently the results of the previous wells were reviewed in order to arrive at a better understanding of the controls on reservoir quality and deliverability. Key to unlocking the Ensign Field were the results of the Shell-operated discovery well 48/14-2 drilled 20 years previously. Although the well failed to flow when first tested, it flowed at a sustained rate of c. 14 MMscf per day after stimulation by a single hydraulic fracture. It was deemed that, with the prevailing gas price (2006), this type of well would form the basis of an economic development. In comparison it was felt that the Conoco well 48/15a-6 had less impact on the field appraisal as this well appeared to have been drilled into a complex fractured zone, was extensively fracture-cemented, and occurred in a portion of the field with a thinner gas column that would not form a part of the Ensign Field core development area.
48/14-2 well evaluation
The Ensign Discovery well encountered a 542 ft gas column predominantly in the upper A and B units of the Rotliegend sequence (Fig. 5). Unlike the neighbouring Shell-operated Barque gas field (Sarginson 2003), the lower units, which typically comprise better quality aeolian units, were below the GWC. The entire Rotliegend interval was cored (c. 800 ft), which provided a detailed facies and reservoir database.
Correlation panel of the Leman Sandstone in the discovery and appraisal wells of the Ensign Field.
Examination of the core acquired shows that the A–C units are dominated by sands interpreted as having been deposited in a damp sand flat (44%) and fluvial environment (23%) with very limited aeolian sands (5% dune facies and 2% dry sand flat). The core-derived porosity and permeability data (Fig. 6) show that these upper units of the reservoir are of very poor quality (<1 mD permeability) and scanning electron microscope (SEM) analysis confirms that this was primarily as a consequence of pore-lining illite that reduced the permeability (Fig. 7a). Although the aeolian units have slightly better reservoir quality, the average permeability of dune facies is 0.27 mD compared with the average damp sand flat permeability of 0.13 mD; it should be noted that the highest overall permeabilities were measured in the damp sand flat facies (0.82 mD). The low permeabilities measured in the aeolian sands, despite their larger grain size, better sorting and low detrital clay component, confirms the detrimental impact of the development of the authigenic fibrous illite on reservoir permeability.
Porosity–permeability cross plot of core recovered from the A–C units in wells 48/14-2 and 48/14-5.
(a) Scanning electron photomicrograph of fibrous illite from reservoir sample in well 48/14-2. (b) Scanning electron photomicrograph plate of fibrous illite from reservoir sample in well 48/14a-5.
Detailed examination of the core also showed that it was almost devoid of any fractures over its entire 800 ft. Analysis of the core photographs, core descriptions and the core itself revealed only four fractures in the c. 350 ft upper gas-bearing interval above the C unit. The fractures were either filled with anhydrite cement (2) or were small cataclastic shear fractures (2). Only one rubble zone was encountered at the base of one cored interval that appears not to be fracture-related. Core recovery of the A–C units was very good, with 100% recovery in five of the six cores and the one core with 97% recovery. The lack of any open or partially open fractures suggested that the productivity of the discovery well after fracture stimulation had not been enhanced by a naturally occurring fracture network.
The hydraulic fracture stimulation parameters for this well have been reviewed and interpreted in order to determine the size and effectiveness of the fracture induced. Table 2 shows the type and amount of proppant and fluids utilized. The initial fracture modelling results, whilst not unique, point to a fracture with a height of 160 ft, half length of 600 ft (i.e. the fracture extends 600 ft on either side of the well bore) and a width of c. 0.25 in. Utilizing these parameters and the resultant flow rates on test supports a productivity index (PI) of 11.5 Mscf day−1 psi−1 and a permeability height (Kh) of 51.2 mDft. It should be noted that the Kh of the perforated interval was significantly lower than this (c. 20 mDft), and consequently it would appear that the hydraulic fracture encountered a better quality reservoir away from the well bore.
Hydraulic fracture proppant size and volumes utilized in the Ensign Field
Although it had previously been thought that the hydraulic fracture had connected into a naturally occurring open fracture network system, evaluation of the data acquired during and after the fracture stimulation did not support this view. Analysis of the well test data supported the view that there was only weak support for a fracture network connected to the well. The relatively low flow rates of c. 14 MMscf per day are in line with the overall poor quality reservoir.
Collation of all the available data (core, log, fracture stimulation, test results) supported the view that the Ensign Field could be developed utilizing fracture-stimulated sub-vertical wells. In order to test this theory and move the field towards development, an appraisal well (48/14a-5) was drilled into the fault block adjacent to the discovery well (Fig. 2). This well was designed with an extensive data acquisition program, including image log and core, as well as a fracture stimulation programme that was designed to create a larger, more effective fracture than that created in the discovery well 20 years previously.
The appraisal well was drilled in 2006 and targeted the upper reservoir intervals (units A and B) in order to minimize the risk of the planned hydraulic fracture extending into the water leg below the C unit. The well encountered the reservoir on prognosis and was successfully fracture-stimulated. In comparison to the 48/14-2 well, 730 000 lb of proppant were emplaced, 40% more than in the discovery well (Table 2). Despite the larger fracture developed, the well failed to flow at the sustained rates of the discovery well. The initial gas rates of 20 MMscf per day declined to below 12 MMscf per day, and failed to stabilize; that is, the gas rate was still falling when the well test was concluded. The poor well deliverability was unexpected and resulted in a detailed examination of the data acquired in the 48/14-5 well in order to understand the reason for the underperformance.
48/14a-5 well evaluation
Analysis of the core showed that the interval comprised similar facies to the 48/14-2 well, but with a higher proportion of waterlain sands. Aeolian dry sand flat deposits were more common than in the discovery well (14 v. 2%) but dune sands were less prevalent (1 v. 2%). However, the reservoir quality was very poor with low permeabilities (Fig. 6). Initially there were concerns that the permeabilities were lower than the discovery well, but this proved to be an analytical artefact as a consequence of the 48/14a-5 samples having been analysed at 800 psi confining pressure compared with the discovery well samples that were analysed at 230 psi confining pressure. It is believed that the lower measured permeabilities at the higher confining pressures are a consequence of the reduction in gas slippage and bypass during the permeability measurement. To confirm this to be the case, samples from the discovery well were re-analysed at 800 psi confining pressure, resulting in permeabilities similar to the 48/14a-5 well. SEM analysis confirmed that the cause of the very low permeabilities was extensive illite cementation (Fig. 7b) that was blocking the pore throats. Although no age dating was undertaken on the illites from this well, K–Ar dating of the illite fraction from the 48/14-2 and 48/15a-6 wells produced a Late Jurassic age range of 147–167 Ma. This age of illite growth is comparable to other illite age dates in the Sole Pit Basin (Robinson et al. 1993).
In comparison to the core from well 48/14-2 that was almost devoid of fractures, the 48/14a-5 core contained a number of open and partially open fractures (Fig. 8), as well as a major drilling-induced axial tension fracture that ran for several metres along the core. The open and partially open fractures were observed in the core as high-angle features with minor amounts of vertical displacement, typically 4–5 mm, that were partially cemented by dolomite and anhydrite. Fracture widths were 0.5–6 mm, but predominantly 2–4 mm. No cataclastic shear zones were observed in the c. 193 ft of core recovered. In total, excluding drilling-induced fractures, 22 open, partially open and cemented fractures were observed in the c. 193 ft of core, that is, a fracture density of 8.7 ft per fracture, compared with the 48/14-2 fracture density of 87 ft per fracture. It was hoped that the open and partially open fractures could contribute to flow after fracture stimulation, but the well test results suggest that these fractures had no impact on well deliverability.
Photograph of core taken from 48/14a-5 showing an open and partially open fracture network.
The acquisition of Schlumberger Full-bore Micro-Imager log data (FMI) over the reservoir interval allowed calibration of the fractures observed in core and the identification of additional fractures in the uncored intervals. It should be noted that, while fractures observed in the core can be described as open, partially open or closed and these fractures can be correlated directly with the FMI log, in the uncored intervals the fracture types can only be inferred from the FMI log data. It is assumed that as the FMI log produces resistivity images of the borehole, fractures that are electrically conductive are open, whilst those that are electrically resistive are closed or cemented (Serra 1989). The images showed the well to be extensively naturally fractured, with 83 fractures identified over an interval of 391 ft. The dominant orientation of the open and partially open fractures was NE–SW with dips of c. 65° towards the NW (Fig. 9), while the drilling-induced axial tension fractures were orientated NW–SE, reflecting the current stress environment in the SNS (World Stress Map 2008). It should be noted that any hydraulic fracture stimulation in the 48/14a-5 well would be perpendicular to the direction of minimum stress and hence be aligned parallel to the drilling-induced fractures. Consequently it is assumed that the hydraulic fracture in the 48/14a-5 well will have cross-cut the open and partially open fractures observed in the core and FMI. The poor performance of this well suggested that there was not a significant amount of flow from the naturally occurring fractures and that this orientation of fractures may not have any impact on reservoir deliverability. Owing to the poor and declining flow rates, no production logging was undertaken on this well, so it is not possible to determine where exactly in the perforated and hydraulically fractured interval any flow was coming from.
Rose diagram showing strike orientation of (a) open and partially open fractures (n=53) and (b) drilling-induced axial tension fractures (n=22) in the FMI log acquired in well 48/14-5.
Appraisal Phase II
The results of the first appraisal well drilled by Venture suggested that the preferred option of developing the field utilizing fracture-stimulated near vertical wells would not be economic given the flow rates measured. Consequently the next appraisal well was planned as a long horizontal well that would be fracture stimulated with five hydraulic fractures. The successful development of the Venture-operated Chiswick Field utilizing multiple fractured horizontal wells proved this to be a viable approach to improving recovery in low productivity reservoirs (Nesbit & Overshott 2010).
The second Venture appraisal well (48/14a-6) targeted the NE portion of the field and had a trajectory such that it passed within 330 ft of the successfully fracture-stimulated discovery well (Fig. 2). The long horizontal length of the well (c. 5900 ft) was designed to allow the gas in several fault blocks to be accessed, as well as increasing the chance of encountering areas of better reservoir quality with the planned hydraulic fractures.
The well was successfully drilled in 2007 to a total measured depth of 15 275 ft, and encountered over 4800 ft of net sand (Fig. 10). The well was fracture stimulated over five intervals after the production liner was run. Each fracture was designed to emplace 500 000 lb of 12/18 proppant, that is, five times the volume of the original fracture stimulation of the 48/14-2 discovery well. When the well was tested it flowed at a rate of 44 MMscf per day that was limited by facilities constraints. The absolute open flow (AOF) of the well was calculated at 80 MMscf per day, at the high end of expectations. The well has been suspended and is planned for use as a production well.
Geological cross section showing the 48/14a-6 well path through the upper Leman Sandstone interval and the location of the hydraulic fracture-stimulated intervals.
48/14a-6 well evaluation
As the 48/14a-6 well was drilled within 330 ft of the fully cored discovery well 48/14-2, no core was taken. The reservoir interval was logged both using real-time MWD data and subsequent pipe-conveyed open hole logging, including an FMI. Analysis of the log data showed that the well penetrated the A100 and B units and had reservoir properties similar to the other wells drilled in the central portion of the field (Table 3). Analysis of the FMI data showed that the well had encountered a variable set of fractures that included faults, drilling-induced fractures as well as conductive, resistive and ‘mixed’ fractures with a distinct range of orientations. These findings were in marked contrast to previous observations that included the low number of fractures observed in the core from the discovery well 48/14-2, and the dominance of a single orientation of open and partially open fractures observed both in the core and FMI acquired in the 48/14a-5 appraisal well. It was hoped that, by studying the results of the FMI in the horizontal well, a unified theory could be derived to account for the well deliverability observed in the Ensign Field, which could aid the placement of future wells and optimize the field development.
Average reservoir properties of the Upper Leman interval by unit
Drilling induced fractures
Five drilling induced axial tension fractures were observed in the reservoir interval, with a NW–SE strike. These fractures are parallel to the direction of maximum horizontal stress, and are of the same orientation as the drilling-induced fractures observed in the 48/14-5 appraisal well.
Faults
18 faults were observed in the FMI data, with the majority dipping to the east. A large conductive, north–south orientated fault was observed on the FMI with an associated conductive fracture network (Fig. 11).
FMI image of conductive fault and associated conductive fractures observed in the FMI log from well 48/14a-6. The fault is the dark (conductive) sinusoid at 10 092 ft on the FMI image.
Fractures
A total of 806 fractures were identified on the FMI log of which 665 were conductive, 115 resistive and 42 mixed, that is, partially conductive and partially resistive. By analogy to the 48/14a-5 well it is assumed that the electrically conductive fractures are open, and those that are resistive are closed or cemented. The different types of fracture also have a preferred orientation, with the conductive fractures striking north–south (dip east and west), the resistive fractures NNW–SSE (dip ENE and WSW), and the mixed fractures striking NE–SW (dip NW and SE) (Fig. 12). It was also noticed that the fractures are not spaced homogeneously along the well path, but appear in clusters, with zones of high fracture density and low fracture density. Given the well trajectory, the likely laterally extensive nature of the sand bodies and the lack of any obvious facies control on the fracture distribution observed in the 48/14a-5 core, it is unlikely that the zones of high fracture density observed in the 48/14a-6 well are related to variations in lithology.
Upper hemisphere Schmidt stereoplots showing the orientation of conductive, resistive and mixed fractures interpreted from the FMI acquired in the 48/14a-6 horizontal well.
It should be noted that a horizontal well provides a much denser sampling of rock heterogeneity than a vertical well, especially with respect to fractures that are typically high angle, and which may be undersampled by a vertical well. In addition, although an FMI image acquired in a vertical well may intersect fewer fractures, it does not impart any directional bias. In comparison, in a horizontal well the identification of fractures is strongly affected by borehole orientation and consequent sampling bias. The probability that a fracture is identified on an FMI image in a horizontal well is dependent on the angle of intersection between the fracture and the well orientation. Fractures that are parallel to the well bore will be strongly undersampled, while those that are perpendicular to the well will be well sampled. This sampling bias can be corrected for and weighted fracture density distributions derived (Terzaghi 1965). The main impact with regards to the fracture distribution observed in the 48/14a-6 is that a Terzaghi weighted fracture distribution suggests that there are more resistive NNW–SSE fractures than imaged on the FMI log.
As well as the fracture type and orientation recorded in the FMI data, additional reservoir data was gathered after the well had been fracture stimulated. During the well flow testing a temperature log was acquired to observe whether the zones that had been fracture stimulated were actually contributing to flow. The observed temperature profile in the vicinity of the fracture-stimulated zones confirmed that they are contributing to flow, although it was not possible to quantify the relative flow contributions from each zone. Owing to operational constraints it was not possible to run a production log that may have quantified the contributions to flow from each zone.
Discussion: the potential role of naturally occurring fractures on productivity in tight gas reservoirs
The Ensign Field has proven more enigmatic than expected. The successful development of the central and northern portion of the field utilizing a multiple fracture-stimulated horizontal well has unlocked its potential, but a number of questions remain. For example: (1) why did the discovery well contain few fractures throughout its cored interval, although the FMI in this fault block shows a number of fractures and faults? (2) Why did the discovery well flow better than anticipated given its very poor reservoir quality? (3) Why did the 48/14a-5 well not flow better given that it encountered open and partially open fractures? Given the data acquired during Venture's appraisal campaign on Ensign, it is hoped that we can answer the critical question which is ‘Is there an overriding control on reservoir deliverability that will help optimize the future development of the field?’
One conclusion that can be drawn from the observed fracture distribution is that different fracture types appear to have different effects on well deliverability. The open and partially open NE–SW striking fractures encountered in the 48/14a-5 appraisal well do not appear to have any impact on well deliverability. Although these fractures are observed to be open in core, they do not appear to enhance reservoir productivity, possibly because in the subsurface under the present stress regime they are closed, although partial mineralization of the fractures would tend to inhibit this. The only interval in the 48/14a-6 horizontal well that contained a significant proportion of these NE–SW striking mixed fractures was in the most northern fault block near the toe of the well. Unfortunately this interval was not accessed by the temperature log, and consequently may or may not be contributing to flow. It should be noted, however, that this interval also had the smallest induced fracture based upon the amount of proppant emplaced (Table 3).
Analysis of the fracture orientation in the fault block containing the discovery well shows a preponderance of the north–south striking conductive, potentially open, fractures. In addition, a large conductive north–south striking fault with associated conductive fractures was observed in the FMI log (Fig. 11) in the portion of the well that passed close to the discovery well. A plot of the 48/14a-6 well path showing the location and orientation of the fractures and the location of the 48/14-2 well bore at top reservoir (Fig. 13) suggests that the discovery well is located in a relatively fracture-free zone, between a series of fracture corridors. This may account for why the 48/14-2 well core contains few fractures and why on fracture stimulation the well performed better than expected given the very poor quality of the reservoir. It is likely that the induced hydraulic fracture would have extended across the north–south trending fracture corridors and ‘tapped into’ the natural fracture network. It should be noted, however, that the potential connection of the well bore with a network of natural fractures in both the 48/14-2 and 48/14a-6 wells does not appear to result in the response typical of a classic open fracture network. In both wells there does not appear to have been a sudden loss of fluids and proppant into the fracture system during the fracture stimulation process. It is more as if the natural fractures are acting as an improved matrix rather than a large-scale open fracture system.
Diagram showing the 48/14a-6 well path in the vicinity of the 48/14-2 discovery well annotated with the interpreted fractures and likely extent and orientation of the hydraulic fracture.
Based on the observations and well results to date on Ensign it appears that: (1) the NE–SW striking open and partially open fractures do not appear to improve well deliverability; (2) the north–south striking conductive fractures appear to improve well deliverability; and (3) it is assumed that the resistive fractures that strike NW–SE have no impact upon well delivery.
The varying style and orientation of the fracture networks suggest that they may have formed under different stress and tectonic regimes during the structural development of the Ensign Field. At present the unifying theory of fracture formation, orientation and improvement in well deliverability has not been tested further in the Ensign Field. Ongoing technical work will hopefully improve our understanding on this matter. The model proposed for fracture style and orientation accounts for the well results to date and will be evaluated as the Ensign Field is developed further.
Conclusions
The Leman Sandstone reservoir of the Ensign Field comprises sabkha, fluvial and aeolian sandstones deposited in an arid continental environment. The sediments were deposited in a variety of erg margin–lake margin settings that have resulted in units that are correlatable across the field. In the Ensign Field the gas-bearing interval is dominated by sabkha and waterlain deposits with only minor aeolian sandstones.
Reservoir quality in the gas-bearing interval is universally very poor, in part owing to the poor primary depositional facies, but predominantly as a consequence of extensive illite cementation. The average core-derived permeability in the gas leg is <1 mD.
Core and FMI analysis shows that the Ensign Field has a heterogeneous distribution of fractures and faults. The fracture population is dominated by conductive (probably open) north–south striking fractures, with subordinate NNW–SSE resistive (probably closed) fractures and NE–SW mixed fractures. Analysis of the FMI data from the long horizontal well 48/14a-6 shows that the fractures are arranged in clusters, with zones of high and low fracture density.
Well results to date suggest that the NE–SW open and partially open fracture observed in core do not improve reservoir productivity, but those orientated north–south that are conductive appear to improve well deliverability.
The uncertainty in fracture type and distribution in the very tight gas reservoir of the Ensign Field has been overcome by utilizing a multiple fracture-stimulated long horizontal well. Although this technology has moved the field towards economic development, further technical work will be undertaken to reduce the uncertainty in the controls on well deliverability and optimize the field's future development.
The authors would like to thank the SNS gas asset team for their work that has resulted in the successful development of the Ensign Field. Sami Haidar at Fracture Technologies is thanked for early input into this work, and Schlumberger and Task Geoscience are acknowledged for the original FMI interpretations undertaken. The authors would also like to thank constructive reviews by Graham Goffey and Jon Gutmanis.
Delegate questions
Delegate: Ian Lisseter, Challenger Minerals Inc.
Question: Is there a density of FMI identified fracturing below which a multi-frac might not be attempted?
Answer: Owing to the uncertainty in the contribution of the fractures in the Ensign Field and the very low matrix permeability, horizontal wells that are stimulated by a number of hydraulic fractures is the base plan, irrespective of natural fracture density.
Question: What direction do the induced fractures have?
Answer: The induced fractures have the same orientation as the drilling-induced axial tension fractures that are parallel to Shmax, and consequently are orientated approximately NW–SE.
- © Petroleum Geology Conferences Ltd. Published by the Geological Society, London
References
The Ensign enigma: improving well deliverability in a tight gas reservoir
- Article
- Abstract
- Field history
- Structure and tectonic setting
- Stratigraphy
- Reservoir sedimentology
- Reservoir structure
- Seismic interpretation
- Field appraisal Phase I
- Appraisal Phase II
- Discussion: the potential role of naturally occurring fractures on productivity in tight gas reservoirs
- Conclusions
- Delegate questions
- References
- Figures & Data
- Info & Metrics