Machar field: Unlocking the potential of North Sea chalk - Brief Article

Low, early-appraisal-well productivity slowed field development until state-of-the-art reservoir descriptions, optimum well planning, effective stimulations and improved water injection more than tripled well rates to 25,000 to 30,000 bopd

Machar field is located in the North Sea in UK Continental Shelf (UKCS) Block 23/26a. BP Amoco operates the field as part of the ETAP (Eastern Trough Area Project) development. The reservoir is mainly low-permeability (less than 1 md) fractured chalk. Production rates from appraisal wells in the 1980s and early 1990s were disappointing at 2,000 to 8,000 bopd, reducing the economic attractiveness of developing the field.

Machar is now under development, and recent production wells have delivered flowrates in excess of 30,000 bopd. These improvements have been achieved by close and careful integration of all aspects of well planning and operations which have evolved over several years.

This case-history article discusses some of the, technical areas which have contributed to unlocking the chalk potential of Machar.

Principal conclusions from the project, as will be discussed, are summarized here.

1. Use of a phased development, with the gathering of key dynamic data to reduce reserves uncertainty.

2. Optimization of well planning, fracture identification and acid stimulation to provide highly productive wells. Key components here were the use of:

* Reprocessed seismic. Better quality seismic data to identify faulted and fractured areas into which to target high-angle wells.

* Accurate wellbore fracture zone identification via mud losses and electrical wireline tools. Mud-loss monitoring provides a reliable, cost-effective way of detecting natural fracture permeability while drilling and--when combined with "Stoneley wave" chevrons from the full waveform sonic log-highlights which zones to complete and stimulate.

* Large-volume, high-rate damage removal and diverted acid stimulation. Selection of short perforation intervals targeted at fractured zones followed by large-volume, well-planned diverted-acid stimulation into these zones.

INTRODUCTION

Machar field, discovered in 1976, is located in UKCS Block 23/26a, in 312-fi (95-m) water, about 150 mi (240 km) east of Aberdeen. The field is a normally pressured oil reservoir with about 410 MMstb oil in place--135 MMstb expected (recoverable) reserves--contained in steeply dipping, fractured Cretaceous chalk and Paleocene sandstone in a high-relief structure over a salt diapir.

Production flowrates from recent development wells have increased dramatically compared with earlier appraisal wells. The methods used to deliver highly productive wells from tight, fractured chalk reservoirs are described. The large increase in productivity from recent Machar field development wells has accelerated production offtake from a lower number of development wells than previously expected. The associated accelerated revenue and lower development costs have increased project capital efficiency.

Field development setting. Machar is 100% BP Amoco owned, and is presently being developed as part of ETAP (Eastern Trough Area Project). ETAP is the simultaneous and integrated development of seven different reservoirs each of which, because of their small scale, was considered commercially "marginal" as a stand-alone development. The ETAP area is about 150 mi (240 km) east of Aberdeen in the Central North Sea. Fields involved in the development are BP Amoco-operated Marnock, Mungo, Monan and Machar; and Shell-operated Heron, Egret and Skua.

Machar is a subsea development with a water injection scheme, comprising five oil producers and two water injectors. It is tied back about 22 mi (35 km) to the ETAP Central Processing Facility (CPF), which directly overlies Marnock field, Fig. 1. All produced fluids are transported through a single multiphase line to the CPF for processing and disposal of produced water. A single water injection line from the CPF provides injection water for Machar.

[Figure 1 ILLUSTRATION OMITTED]

Machar reservoir description. The structure is a four-way dip closure over a salt diapir which pierces the structure's crest. The structure has great vertical relief from the crest at 4,264 ft subsea (1,300 mss), with closure extending down to some 3,000 mss. The crest is subdivided by faults down-throwing to the east and north. The reservoir units are tilted, with dips of up to 50 [degrees] on flanks of the diapir. Fig. 2 is a top-reservoir map, Fig. 3 an east-to-west field cross-section.

[Figures 2-3 ILLUSTRATION OMITTED]

The Machar reservoir interval is subdivided into six stratigraphic units. Reserves are found in Paleocene sandstones; Ekofisk, Tor and Hod chalks; and a Celestite zone immediately overlying the salt diapir. These units can be identified on wireline logs and are summarized in the type log section from Well 13, Fig. 4.

[Figure 4 ILLUSTRATION OMITTED]

Thicknesses of the units are controlled largely by stratigraphic thinning onto the diapir crest and by faulting. Of the estimated crude volume in place, most is in the tight chalk matrix, which produces via a naturally occurring .fracture system within the tight chalk. Key reservoir characteristics are noted below.

Reservoir parameters
  Oil gravity, [degrees] API                            41
  Gas/oil ratio, scf/bbl                               877
  Form. volume factor, rb/stb                         1.45
  Oil viscosity, cp                  0.4 @ 190 [degrees] F
  Avg. bubble point, psia                              3,160
  Reservoir temp., [degrees] F:          190 [degrees] F @
                                              1,300 mTVDss
  Crest, mss                                      ca 1,300
  Structural closure, mss                         ca 3,050
  Gas column                                          None
  Typical formation dip, [degrees]                   45-60

Typical chalk properties
  Matrix NTG                                           0.8
  Matrix porosity, %                                 10-35
  Matrix permeability, md                         0.01-5.0

PHASED DEVELOPMENT, EARLY APPRAISAL

Machar was discovered in 1976, and has been extensively appraised by data from 11 wells, with 12 sidetracks (seven for geological reasons). Two 3-D seismic surveys have been acquired. The initial appraisal was aimed at reducing uncertainty regarding oil-in-place and reserves (recoverable). Wells were drilled on northern (Well 12Z, as will be discussed) and northwestern flanks of the field (Wells 13, 13Z and 16) to establish presence or absence of oil-bearing productive reservoir on the flanks and position of the field oil-water contact, Fig. 2. The wells evaluated the fractured-chalk reservoir by comprehensive logging, coring and testing.

By the early 1990s, the appraisal program had reached the point where further drilling or testing of conventional wells could not reduce reserves uncertainty sufficiently to take Machar closer to a development decision. Only acquisition of dynamic reservoir data through long-term production testing could sufficiently reduce the uncertainty in reserves and production profiles.

A three-phased development was initiated in 1994. Phase 1 (June 1994 to May 1995) involved producing 7.7 MMstb of oil from two wells, 6Y and 18Z, under natural depletion. These wells were completed with individual SWOPS trees--but with connected flowlines and controls.

The flowlines were tied back through a riser connector package and a single riser that could be positioned on either tree to the Sedco 707 drilling rig. This rig had production facilities installed on the deck, including 35,000-bpd, two-stage separation capacity. Oil was pumped through a 1,500-m flexible flowline to the dynamically positioned Stena Savonita tanker. The prime objectives--to establish viable long-term productivity from the fractured chalk reservoir and to test natural depletion as a recovery mechanism--were achieved.

Phase 2 (August 1995 to June 1996) tested reservoir response under water injection; 6.8 MMstb oil was produced from two wells (6Y and 20Y), and water was injected at close to voidage replacement into Well 18Z. This well had been converted from its initial use as a producer in Phase 1. The Sedco 707 was used in the same way as in Phase 1, with addition of a 40,000-bpd water injection facility. Deaeration of the raw seawater was achieved in a vacuum tower, and the rig pumps were used for injection.

During Phase 2, well productivity was maintained over the period and GOR remained stable, near the solution level. There is considerable uncertainty over waterflood effectiveness in a fractured chalk system. However, no water was produced throughout the Phase 2 period from either of the two producers, and this increased confidence that significant additional reserves would be produced under a water-injection development scheme.

Both Phase 1 and 2 had significant impacts on reserves estimation, with P50 reserves increasing from 62 MMstb, under natural depletion, to 135 MMstb under waterflood. While considerable uncertainty still existed over the waterflood reserve, the P90 reserve was sufficient to support a full development. Thus, the current-full development plan was initiated (Phase 3).

The ongoing challenge was to ensure that high well productivities/injectivities were obtained to minimize development-well numbers and support a peak production rate of 42 Mbpd. The advances made in increasing well productivity are detailed in the following sections.

WELL PLANNING/DESIGN: PAST/PRESENT

Machar well design, data acquisition techniques and operational methods have evolved and been optimized over the last decade. The dramatic improvements in well production rates have been achieved by close and careful integration of well planning and well operations.

To illustrate the changes in practice and the resulting production benefits, an outline of the methods used and results obtained in an early appraisal well (23/26a-12Z) are compared with techniques now used in the most recent development wells.

Appraisal Well 23/26a-12Z. This well was drilled and tested in the Machar North area in June 1989. It was vertical and penetrated only 170 m of chalk reservoir. Core was taken over the oil-bearing interval, and a basic suite of logs was run--density, neutron, sonic and pressure measurements. A 31-m chalk interval was perforated and acidized, and a DST was carded out. Well productivity from the chalk was low, with a sustainable rate of about 1,000 bpd.

There are three possible reasons for the poor well productivity:

1. Only a few fractures intersected the well. This well was targeted from 2-D seismic data which was not able to show faulted/fractured areas. More recent 3-D has confirmed that the well intersected an area with low potential for faults and fractures. This well was vertical and, therefore, did not penetrate a thick chalk sequence. This reduced the probability of encountering fracture zones. One objective of future wells was to maximize the length of reservoir penetration to improve the chance of intersecting more fracture zones.

2. The fractures that did intersect the wellbore were not perforated. Fractures and faults have the potential to provide excellent productivity. These features are also the causes of mud loss during drilling, ranging from a few barrels to total losses. The optimal technique for identifying these fracture zones is now known to include careful monitoring of mud loss while drilling the reservoir section. In addition, logging using the Dipole Shear Image tool is now used to confirm fracture-zone intervals, that can then be targeted for perforation.

In Well 12Z, the process of recording mud loss was not carried out while drilling the reservoir section, and no specific fracture-identification logs were run. As no mud loss or fracture log data was available, the only remaining indication of open fractures was from visual core inspection. This was not carded out in a timeframe that would have provided input to assist selection of the perforation intervals.

Perforation interval selection criteria in the well were based on the intervals of highest porosity--this proved to be less than optimal. Inspection of the core for open-fracture intervals suggested that the two zones with the highest fracture density were not perforated.

3. A non-optimal diverted acid stimulation operation. During the last decade, the diverted-acid method used to enhance productivity from the chalk has been refined and optimized. The main objectives of the acid stimulation are to:

* Overcome drilling-induced formation damage within natural fractures by removing mud/lost circulation material (LCM) used to control mud loss during drilling.

* Extend/enhance fractures in the tight chalk matrix away from the near-wellbore region. This allows near- and far-field fracture permeabilities to be accessed.

The acid stimulation treatment of Well 12Z was the first attempted on a Machar well. The job was small by current field standards, with only 500 bbl of 15% HCl used. Current practice is to pump 5,000 bbl of 28% HCl in 500- to 600-bbl stages. The method of using ball sealers to divert the acid was carried out for the first time, but the ball injector failed after 37 balls were released. Thirty-one meters of chalk were perforated at 4 spf--about 400 perforation holes if all charges fired.

Acid was pumped into the well and did successfully decrease skin and increase the productivity, index. It did not, however, increase the well's kh, which suggests that the acid did not penetrate very far into the formation and only washed out the wellbore. The well was then flowed at rates in excess of its sustainable rate, resulting in high drawdowns, and it produced fluids with an increasing GOR.

Current well planning and design. Regarding well type and optimal placement to intersect as many natural fractures as possible in the chalk, all recent development wells have been drilled at high angles, 40 [degrees] to 65 [degrees], through the reservoir section. These wells follow the chalk stratigraphy down the side of the salt diapir and commonly penetrate in excess of 1,000 m of oil- bearing reservoir, Fig. 5.

[Figure 5 ILLUSTRATION OMITTED]

Until recently, it was not possible to image the steep flanks of the diapir from seismic data. This resulted in several sidetracks in appraisals due to uncertainties in reservoir positioning. Recent reprocessing of the seismic data--utilizing steep-dip algorithms that were unavailable at the time of the original processing--has significantly improved flank imaging. This new seismic data has enabled three recent development wells to be successfully planned/drilled without sidetracking.

The reprocessed data has also been useful in improving understanding of the Machar chalk fracture system. This is best visualized on dip attribute maps that highlight the presence of faults, Fig. 6. Localized steep dips at top reservoir correlate well with faulting identified in the wells.

[Figure 6 ILLUSTRATION OMITTED]

The new data shows that the Machar structure appears to be dominated by concentric faults, with the crest being less faulted than the flanks. This has resulted in development wells being planned to be radial to the structure, intersecting the concentric faults at right angles to maximize the number of fracture zones intersected for a given well length.

In addition, this data has also been used successfully to monitor progress during the recent drilling of development Wells Al, A2 and A3, and to predict occurrence of faults, which often correlate to mud-loss zones. There is a good correlation between loss zones observed during drilling of the reservoir section and zones of locally higher dips interpreted as faults. Fig. 7 shows, in more detail, the dip attribute map on top chalk around the area of Well A3. The lower part of this well, in the most faulted area, was completed as a water injector and is currently capable of injecting over 45 Mbwpd.

[Figure 7 ILLUSTRATION OMITTED]

FRACTURE DETECTION

Fracture detection is critical in Machar wells, as it is necessary to only selectively perforate and stimulate a handful of specific, short, highly fractured chalk intervals. Typically, eight intervals of less than 4 m per interval are perforated within a chalk reservoir sequence, normally in excess of 1,000 m. It is recognized that the key to detecting natural fracture permeability is an integrated approach using as many different methods as possible.

Mud loss control monitoring. Mud losses are the basic indicator for open fractures within the chalk. They have the advantage of being very cheap information to acquire and are available in real time. Mud losses must be monitored very carefully while drilling the chalk reservoir section. Losses on the scale of 1 to 5 bbl in the Machar chalk have proved to be indicative of a potential highly productive zone.

It is, therefore, very important that any "noise" be noted on the mud system's pit level charts and be taken into account. For the data to be useful and reliable, any rig activity such as moving cranes, ballasting the rig, switching mud centrifuges on or off, or mixing-in mud must be restricted to a minimum and recorded. The active pit must be heave compensated, e.g., two pit monitors on a semisubmersible. Accurate depth matching--to the wireline information--of mud loss data measured by drill pipe is vital to locate the correct perforation intervals.

During drilling of the chalk reservoir section, operational difficulties can occur unless rate of mud loss into the faults and fractures can be controlled. The aim is to control the losses, i.e., cover up or fill the fractures in such a way that they can be restored and cleaned up to allow production from these zones.

The best practice found for the chalks of Machar is to drill with a background level of ground marble and other LCMs, which are used to plug small fractures as they are drilled. If small losses ([is less than] 10 to 20 bbl) are detected, this treatment concentration is increased. For larger losses, a lost-circulation pill of this material is pumped. If this fails, an acid-soluble cement is pumped. The LCMs and cements used must be chosen so that they will be dissolved by the later acid stimulation operation.

Electric logs for fracture detection. Identification and successful stimulation of open fractures are the key to high productivity in Machar, as the fracture system provides high permeability conduits for production. The bulk of the reservoir consists of chalk of about 1 md or less. During field appraisal, numerous fracture-detection logs were run, with varying degrees of success.

In the appraisal wells, various image logs were used to detect fractures. Many of these had the advantage of determining orientation and density of fractures rather than just detecting fracture presence. Unfortunately, the interpretation from this form of log data is often not available in the timeframe for assisting perforation-interval selection. In addition, rugosity and/or spiraling in the borehole wall can destroy any detail in the image. This well was drilled with a motor and a bent sub. A spiral groove exists about the wellbore circumference, about one inch deep and with a wavelength of about a meter.

In later wells, "Stoneley wave" processing from the Dipole Shear Image (sonic) tool has proven reliable in detecting open chalk fractures. Fig. 8 shows the variable-density log display of Stoneley wave chevrons produced when an open Machar chalk fracture is encountered.

[Figure 8 ILLUSTRATION OMITTED]

Other supporting log data are used in confirming the presence of open fractures, e.g., caliper log and PEF curves from the density log. The latter log can pick up spikes over intervals where barite from the drilling mud is held within the open fractures. Fig. 9 is an example of a summary of some of the information which is then used for selecting perforation intervals.

[Figure 9 ILLUSTRATION OMITTED]

Good correlation exists between fault/fracture areas seen on the reprocessed seismic and fractures detected by mud loss during drilling and from electrical logs. Figs. 5, 7 and 8 illustrate this relationship from recent development wells.

ACID STIMULATION METHOD

Within the Machar chalk reservoir, most well productivity is via fractures in and around the wellbore; and stimulation of these fractures with acid has greatly improved productivities. The method of stimulation is a large, 5,000-bbl acid stimulation with diverter ball sealers across selected perforated zones.[1] Once suitable fracture zones have been identified, they are perforated using 2-spf, zero-degree-phasing wireline guns on the low side of the hole. Eight, 4-m zones are chosen as a compromise between maximizing the amount of acid available per perforation and maximizing the number of perforations open to flow.

After perforation, there is a short clean-up flow to remove the perforation debris. The acid stimulation then follows, with an initial pre-flush followed by an 8-to 10-stage acid stimulation--each stage terminating with a dose of diverter balls to divert the acid to unstimulated zones. The diverter balls--being denser than the injected gel--fall onto the low side of the hole and into the perforations that are currently accepting acid. This staged-diversion process is then repeated on other perforated intervals, Fig. 10. Quality of cement behind the liner is important for successful diversion of the acid.

[Figure 10 ILLUSTRATION OMITTED]

The listing below illustrates, in chronological well order, productivity improvement with time in the Machar wells. This method has been shown to be successful, with productivity increases for Wells 18Z and 20Y of 50 fold and 500 fold, respectively, after acidization.

Well                   Pre-acid PI, bpd/psi      Post-acid Ply
                                                       bpd/psi

No acid jobs,              1976 to 1989
1 RE                            11                   Not taken
6Y                               3                   Not taken
10                              12                   Not taken
10Z                              6                   Not taken
Small (S00-bbl) acid jobs, 1989 to 1993
12Z                              0.3                      4(*)
13                               0.2                  6(*)/110
18Z                              3.0                       150
Large (5,000-bbl) acid jobs, 1993 to present
20Y                              1                         500
A1                          Not taken                      590
A2                          Not taken                      130

(*) Acid diversion failed.

PHASE 3 DEVELOPMENT, FIELD START-UP

Phase 3--full-field subsea development with water injection is planned to use seven wells--five oil producers and two water injectors. Three wells (6Y, 20Y and 18Z) from Phase 1 and 2 were re-used, with four new wells drilled, of which three (A1, A2 and A3) have been completed to date. These new wells were perforated over an interval to allow an adequate stand off from any gas or water legs early in field life. This was important to minimize requirement for costly subsea water or gas shut off interventions.

Peak production from Machar was originally planned to be 42 MMbopd, but due to the excellent productivity achieved in the latest development wells, higher offtake rates were possible. This, in mm, increased the importance of early pressure support. As a consequence, a key change introduced to Phase 3 development was the acceleration of water injection.

Water injection was originally scheduled between nine and 12 months after Phase 3 first-oil production. However, it was possible to speed up construction and optimize injection installation, with start up taking place only two months after first oil. This flexibility to change Phase 3 development at a late stage allowed upside potential to be captured, i.e., accelerated oil production in early field life and increased reserves from the extended waterflood period.

ETAP commenced production on July 19, 1998, with first oil from Machar on August 12; water injection was commissioned during October. There is still a large uncertainty as to how Machar will perform under waterflood. Reservoir surveillance forms a key part of optimizing reservoir management and identifying how best to capture any upside performance. The principal surveillance data being gathered, where operationally possible, are:

* Continuous bottom- and tophole pressure measurements

* Well testing through the subsea multi phase meter

* Monitoring of GOR and watercut through: subsea multi phase meter, CPF metering and indirectly from lift correlations

* Tracer deployment in water injection, and

* PLT/TDT monitoring of gas and water movement.

To date, Machar delivery performance has been excellent, with 10 months of dry oil production from three wells, each having production potential in excess of 25 Mbpd. Field offtake rates of 60 Mbpd have been achieved, which exceeds the original design of 42 Mbpd.

ACKNOWLEDGMENT

The authors wish to acknowledge BP Amoco for its permission to publish this article. Thanks are also due to the large number of people involved over the last decade for their contributions to the success of the Machar project. Thanks also to Kathleen Brown for her help in preparing the various figures. This article was adapted from paper SPE 56974, presented by the authors at the 1999 Offshore Europe Conference, Aberdeen, Scotland, September 7-9, 1999.

LITERATURE CITED

[1] Gilchrist, J. M., A. D. Stephen and O. M. N. Lietard, "Use of high-angle, acid-fractured wells on the Machar field development" paper SPE 38917, presented at the European Petroleum Conference, London, October 25-27, 1994.

Alastair Brown, Senior Engineer/Team Leader, BP Amoco, graduated with a BSc (Hons) in applied geology from Strathclyde University in 1979. He has 20 years of experience in the industry, mostly in subsurface disciplines. Before joining BP Amoco in 1989, he was employed by Shell International for 10 years, working as a geologist and petrophysicist in the UK and Brunei.

Merv Davies, is a geophysicist presently working in BP Amoco's Eastern Trough Area Project in Aberdeen. He holds a BSc in mining geology from the University of Wales, Cardiff, and an MSc in geophysics from Leeds University. He began his career with Western Geophysical in 1978, prior to joining BP in 1981, working in the UK, Norway and the U.S.

Hugh Nicholson, field geologist for the Machar, Monan and Mungo fields, for BP Amoco, has worked as a geologist since 1990, after earning a PhD at Edinburgh University. He worked for BP from 1990 to 1994 in research and reservoir geology. He left BP in 1994 to work for Western Atlas in the Middle East and London before rejoining BP in 1996.

Brian Gene, a senior reservoir engineer with BP Amoco, graduated in 1974 from Bristol University with a BSc (Hons) in chemistry. With 25 years of experience in the industry, he is presently specializing in reservoir simulation.

COPYRIGHT 2000 Gulf Publishing Co.
COPYRIGHT 2000 Gale Group