Offshore Northern Europe: Strontium isotope analysis can help define compartmentalization

Aug. 27, 2001
The Infield Systems database shows that, in summer 2001, there were 2,918 offshore fields in production worldwide.

This article describes how strontium isotope (87Sr/86) ratios in formation waters are used to evaluate compartmentalization of hydrocarbon reservoirs.

Strontium Isotope Residual Salt Analysis (SrRSA) of core samples provides a means of measuring the 87Sr/86 ratios in formation water from hydrocarbon columns and aquifers. Smooth SrRSA profiles suggest progressive, uninterrupted filling and the absence of sealed barriers, while a step change in a profile normally suggests a barrier sealed updip from the well penetration.

Inferences about lateral connectivity are made by comparing SrRSA profiles from neighboring wells at true vertical depth (TVD). Profiles that are superimposed when plotted at TVD suggest the well sections share a common filling history and lie in the same flow unit. Neighboring SrRSA profiles that are not superimposed normally suggest segmented compartmentalization of the reservoir.

Introduction

Acquiring information on reservoir compartmentalization during the early stages of field evaluation and development is important to cost-effective exploitation of hydrocarbon resources.

It is common for hydrocarbon reserves to be overestimated in fields that are compartmentalized.1 In such cases, the negative impact compartmentalization has on recovery factors and reserve estimates only emerges after production has started.

Unrecognized compartmentalization may constitute a substantial development risk for operating companies. Ideally, therefore, as complete a picture as possible of the degree of compartmentalization should be determined during the earliest stages of field evaluation and development.

Quantitative information on reservoir fluid connectivity is normally obtained by static pressure measurements (e.g., repeat formation test tool or equivalent) or by the production testing of appraisal wells. However, pressure testing may not always be possible and, in certain circumstances, preproduction static pressure data might not reflect the full degree of compartmentalization.2

Another approach to assessing reservoir fluid connectivity is to measure physical and geochemical properties of the hydrocarbons and formation waters ("pseudo-dynamic" data of Smalley;3 e.g. England,4 Smalley & England,5 Larter and Aplin,6 and Kaufman et al.7).

Recovering a sufficient number of formation water samples that are not contaminated by drilling mud filtrate is the main obstacle to the routine acquisition of water chemistry data. However, the advent of the SrRSA methodology has enabled the study of variations in the 87Sr/86r isotope ratio in formation waters at a level of detail limited only by the availability of conventional core.8-15

Click here to enlarge image

In SrRSA, formation water is reconstituted from small 2-5 cc sandstone or limestone core samples by adding 10 ml of deionized water to the disaggregated sample. The deionized water dissolves or finds equilibrium with the water-soluble Sr salts, which are then extracted and the 87Sr/86 ratio analyzed using a high performance thermal ionization mass spectrometer. A series of analyses conducted throughout a core, normally at 3-5 m spacing, results in a profile that reflects 87Sr/86 ratio variations in the formation water throughout the cored reservoir interval (Fig. 1).

SrRSA has been conducted on over 350 wells and is now applied on a semi-routine basis by several companies. The method is not widely applied outside of the North Sea and Europe although use in the US Gulf Coast, West and North Africa, and the Middle East has grown in recent years. This article is an abridged version of Mearns and McBride16 and aims to provide a background to the SrRSA method with examples from the North Sea.

Background

All subsurface waters contain the alkaline earth element strontium (Sr) in trace quantities. Fresh and meteoric waters may contain only a few parts per million, while deep, evolved, saline basinal brines may contain > 1,000 ppm Sr.

Strontium is one of the most soluble elements in water and has four stable isotopes 84Sr, 86Sr, 87Sr, and 88Sr. The isotope 87Sr is formed continuously in nature by the natural radioactive decay of 87Rb. Thus strontium, which has different age and-or origin, acquires variable 87Sr/86 ratios.

Three factors need to be considered in the interpretation of 87Sr/86 isotope ratios of subsurface water.

Click here to enlarge image

First the initial depositional water composition (e.g. seawater or meteoric water) is modified by water-rock interaction (mineral dissolution), fluid flow (compaction, overpressure release, hydrodynamism), and diffusion with time (Fig. 2).

In simple terms, in a system that has water as the continuous phase it is expected that diffusion will cause homogenization of chemical and isotopic variations resulting in uniform 87Sr/86 ratios throughout the connected volume of water. If the rock quality is poor or if it contains continuous seals to vertical homogenization (e.g. mudstones or cement bands) or lateral seals (stratigraphic pinchouts or faults) that segment the system, then this may give rise to water segments that have distinctly different 87Sr/86Sr ratios to each other.

Second, introducing hydrocarbons to the system has a major impact upon the behavior of the 87Sr/86Sr water system. The effective porosity and permeability of the system for water are greatly reduced by the presence of the hydrocarbons. Depending upon saturation levels and wettability the water may cease to be a continuous phase.

These processes effectively eliminate advective flow of water, greatly reduce or eliminate diffusion owing to increased tortuosity of the water phase, and dramatically decrease or halt mineral dissolution within a hydrocarbon column. This has led to the concept that the composition (chemical and isotopic) of irreducible water is effectively "fossilized" at the time it became trapped as a result of hydrocarbons entering a system. In essence, the 87Sr/86 ratio of irreducible water should represent the composition the aquifer had at the time the water became trapped at a dynamic contact during reservoir filling (Fig. 2).

The third factor that needs to be taken into account is that chemical and isotopic compositions of aquifer water may change with time. During the time it takes to fill a reservoir the 87Sr/86 ratio may evolve in the aquifer below the dynamic contact. Thus, successive "layers" of hydrocarbon may trap water with progressively different 87Sr/86 ratios which are preserved as water saturation levels fall within the hydrocarbon column (Fig. 2).

In summary, three main concepts are used to interpret 87Sr/86 variations in subsurface waters:

  1. Connected aquifer should have uniform 87Sr/86 ratios.
  2. The 87Sr/86 ratio in irreducible water is fossilized at the time the water was trapped by hydrocarbons at a dynamic contact.
  3. The 87Sr/86 ratio of an aquifer below a dynamic contact may change with time as a reservoir fills with oil or gas.
Click here to enlarge image

These concepts are used to draw conclusions about reservoir connectivity (Fig. 3).

In the interpretation of SrRSA data a major distinction is drawn between baffles and barriers. Baffles are regarded as low permeability features of finite extent that do not impact the reservoir fill process (Fig. 3A), while barriers are regarded as features that have both seal capacity and geometry to influence the filling of the reservoir (Figs. 3B, 3C).

A vertically and laterally connected reservoir is shown schematically in Fig. 3A. At any point in time the aquifer below the dynamic contact has a uniform 87Sr/86 ratio, but this ratio changes with time as the reservoir fills.

This gives rise to an uninterrupted trend in the 87Sr/86 ratios throughout the hydrocarbon column. Neighboring wells that have shared the same filling history will have similar absolute 87Sr/86 ratios at any given depth. This is illustrated by the SrRSA profiles for the blue and green wells and which are superimposed at true vertical depth (TVD) (Fig. 3A).

In the case of a compartmentalized reservoir (Figs. 3B, 3C), the presence of barriers causes a fill and spill process. On opposing sides of a barrier, the trapping time and 87Sr/86 ratios of irreducible water may be quite different, giving rise to step changes in the 87Sr/86 ratio profile when a barrier is crossed. In this way SrRSA data may be used to discriminate between baffles and updip sealed barrier features within a reservoir.

There may also be differences in trapping times of irreducible water across paleo contacts which may be identified from a step change in an SrRSA profile within a sequence that lacks obvious seals (Fig. 3C). In compartmentalized reservoirs the SrRSA profiles of neighboring wells do not match when plotted at TVD (Figs. 3B, 3C).

Mud contamination

Contamination of core fluids by drilling mud filtrate is identified as the main technical limitation of the SrRSA method and is discussed in some detail by Mearns and McBride.16

Click here to enlarge image

Because of drilling mud invasion, SrRSA is not universally applicable and in certain cores it may prove impossible to acquire meaningful results. If all of the main factors controlling mud invasion of core are positive (Table 1) then obtaining good quality results will normally be assured, while if all are negative poor quality data will normally be obtained.

The relative concentrations of Sr in the formation water and aqueous mud filtrates also play an important role in determining the effect of mud contamination. Favorable conditions exist when the formation water contains a lot of Sr (>> 100 ppm) and the aqueous mud filtrate contains little Sr (<< 5 ppm).

In reality, the majority of cases represents a mixture of positive and negative factors and technical success is often determined by the expertise of the coring company combined with careful sampling to avoid drilling mud invaded zones. For example, it has proven to be possible to obtain good quality SrRSA data from unconsolidated sediments in cores cut using oil-based mud and low invasion coring with rig site freezing of core samples.

Example reservoirs

Click here to enlarge image

The SrRSA profiles for two wells from Arbroath field, UK Central North Sea, are shown in Fig. 4 alongside gamma-ray log and the stratigraphic subdivison. The reservoir is Paleocene Forties formation turbidite sandstone.

Well A tends to have zones of clean sand interbedded with more muddy sand, while Well B has clean sands separated by well-defined but thin mudstone layers. The SrRSA profiles of each well display trends of progressively increasing 87Sr/86 ratio with depth down to the contact and the profiles lack significant step changes. This suggests a continuous fill sequence and the absence of updip seals. Plotted together at TVD, the two profiles have a near perfect match, suggesting simultaneous filling at each well locality in contact with the same isotopically evolving aquifer. The SrRSA data suggest a vertically and laterally connected reservoir analogous to that shown in Fig. 3A.

Vertical compartments

Click here to enlarge image

An SrRSA profile together with gamma-ray log and static pressure data are shown for a cored interval from the Upper Jurassic Brae formation conglomeratic sandstones of Thelma field, South Viking graben, UK North Sea (Fig. 5).

In this well, the static pressure data show a significant step change across the argillaceous interval towards the top of the section. This demonstrates that the oil above the mudstone has a different contact to the oil below, indicating that the mudstone forms a laterally continuous seal with geometry to trap hydrocarbons analogous to that shown in Fig. 3B.

The SrRSA profile also exhibits a significant step change across the mudstone interval that is consistent with the mudstone forming an updip sealed barrier. In this particular case, the pressure data alone are sufficient to confirm the mudstone is a sealed barrier and sophisticated geochemical data are not required to reach this conclusion.

However, static pressure data may not always be a reliable guide to connectivity.2 The schematic compartmentalized reservoir in Fig. 3C has filled by stages of fill and spill so that the spill points of the barriers have been bypassed, resulting in a single OWC for the accumulation. In this example, uniform pressure gradients should be expected across the field and the presence of updip sealed barriers may not emerge until after production has started.

Click here to enlarge image

Gamma ray log, stratigraphy, pressure, and SrRSA data from the Brent Group reservoir in Dunbar field, northern North Sea, are shown in Fig. 6. The Broom, Rannoch, and Etive formations are predominantly sandstones deposited in foreshore-shallow marine environments. The Ness formation comprises interbedded fluvial channel sandstones, mudstones, and coals deposited in a deltaic environment. The Ness formation is overlain by massive shallow marine sandstones of the Tarbert formation and Upper Massive Sandstone (UMS) member. The static pressure data show more or less a single gradient and do not provide good evidence for vertical compartmentalization of the reservoir (Fig. 6).

The SrRSA profile in this well shows two unequivocal step changes at the Ness B-Tarbert boundary and within the Ness B zone, both occurring across thin coal horizons. These significant step changes in the SrRSA profile suggest that these coal beds form updip sealed barriers within the reservoir, an interpretation that has since been supported by production data. The uniform pressure gradient combined with the step changes in the SrRSA profile suggest, therefore, a reservoir geometry analogous to that schematically shown in Fig. 3C.

Lateral compartments

Click here to enlarge image

Lateral connectivity was discussed previously for the turbidite sandstone reservoir example illustrated in Fig. 4 where it was concluded that the reservoir was vertically and laterally connected and comprised a single flow unit. A more complex example is now introduced which shows SrRSA and pressure data from two Dunbar field wells (Fig. 1, 6) superimposed at TVD (Fig. 7).

The SrRSA profiles for the UMS and Tarbert formation zones are concordant, consistent with this interval comprising a single continuous flow unit between these two wells (analogous to the example shown in Fig. 4). However, the SrRSA profiles for the underlying Ness formation zones to a large extent do not match each other at TVD, suggesting different filling histories and some form of barrier to lateral flow between the two wells within the Ness formation zones.

To a certain extent the dipping coal bed sealed barriers contribute to lateral compartmentalization of the reservoir. Lateral facies variations within the Ness combined with faulting may also contribute to lateral segmentation of the Ness reservoir zones. The interpretations that the UMS unit forms a single continuous flow unit between these wells, and that the Ness formation is compartmentalized both vertically and laterally, have since been verified by production data.

Interpretive complexities

The concepts outlined above provide a framework for the interpretation of SrRSA in formation waters based on reservoir filling history concepts.

The assumptions attached to this framework are known on occasions to be false. Furthermore, the default practice of comparing results at current TVD has the inherent assumption that a reservoir today is in the same structural configuration as it was when it filled.

Many SrRSA data sets bear evidence to post-fill structural adjustments where the SrRSA profile in one well may be raised in a vertical sense relative to a neighboring well. These geological complexities, discussed in more detail by Mearns and McBride,16 do not invalidate application of the SrRSA method, but an awareness of geological variables is required in order to avoid erroneous interpretation of results.

Static pressure measurements can be an unreliable guide to reservoir connectivity. Also, high cost limits the use of well testing. SrRSA provides an additional means of acquiring information on three-dimensional reservoir connectivity during field appraisal, ahead of the design, drilling, and completion of production and injection wells.

Acknowledgments

The ideas presented here have been developed over many years and have benefited from numerous discussions with geologists and engineers from many companies. ENI, BP, and TotalFinaElf are gratefully acknowledged for allowing us to use their data. Without their consent this article would not have been possible.

References

  1. Dromgoole, P., and Speers, R., "Field Reserves Uncertainty-the effect of compartmentation," AAPG-EAGE Research Symposium; Compartmentalized Reservoirs: Their Detection, Characterization and Management (abs.), October 1996, The Woodlands, Tex.
  2. Baillie, J., Coombes, T., and Rae, S., "A multidisciplinary approach to update Brent reservoir description and modelling," SPE 35528, European 3-D Reservoir Modelling Conference, Stavanger, 1996.
  3. Smalley, P.C., "Integration of static, dynamic and pseudo-dynamic data to detect reservoir compartmentalization" (abs.)., AAPG Bull., Vol. 81, 1997, p. 1,413.
  4. England, W.A., "The organic geochemistry of petroleum reservoirs," Organic Geochemistry, Vol. 16, 1990, pp. 415-425.
  5. Smalley, P.C., & England, W.A., "Reservoir compartmentalization assessed with fluid compositional data," SPE Reservoir Engineering, August 1994, pp. 175-180.
  6. Larter, S.R., and Aplin, A.C., "Reservoir geochemistry: methods, applications and opportunities," in Cubitt, J.M., and England, W.A., eds., "The Geochemistry of Reservoirs," Geological Society, London, Special Publication 86, 1995, pp. 5-32.
  7. Kaufman, R.L., Fitzmorris, R.E., and Eisenberg, L.I., "The integration of geochemical, geological and engineering data to determine reservoir continuity in the Iagifu-Hedinia field, Papua New Guinea" (abs.), AAPG Bull., Vol. 81, p. 1,387.
  8. Gibbons, K., "Use of variations in strontium isotope ratios for mapping barriers: An example from the Troll field, Norwegian Continental Shelf" (extended abs.), in proceedings of the 6th European Symposium on Improved Oil Recovery, Stavanger, May 21-23, 1991. Norwegian Petroleum Directorate, Vol. 1, pp. 205-211.
  9. Smalley, P.C., and England, W.A., "Assessing reservoir compartmentalisation during field appraisal: How geochemistry can help," proceedings of the European Petroleum Conference, 1992, pp. 423-431 (SPE 25005).
  10. Smalley, P.C., and Oxtoby, N.H., "Spatial and temporal variations in formation water composition during diagenesis and petroleum charging of a chalk oilfield," in Kharaka, Y.K., and Maest, A.S., eds., "Water-Rock Interaction," Balkema, Rotterdam, 1992, pp. 1,201-04.
  11. Smalley, P.C., Lonoy, A., and Raheim, A., "Spatial 87Sr/86 variations in formation water and calcite from the Ekofisk chalk oil field; implications for reservoir connectivity and fluid composition," Applied Geochemistry, Vol. 7, 1992, pp. 341-350.
  12. Stolum, H-H., Smalley, P.C., and Hanken, N-M., "Prediction of large-scale communication in the Sm rbukk fields from strontium fingerprinting," in Parker, J.R., ed., "Petroleum Geology of Northwest Europe," proceedings of the 4th Conference, Geological Society, London, 1993, pp. 1,421-32.
  13. McBride, J.J., Feroul, J-M, and Mearns, E.W., "A Sr Residual Salt Analysis (SrRSA) study to investigate the connectivity of the Brent Group reservoir of the Dunbar field, UK North Sea" (abs.)., AAPG Bull., Vol. 79, 1995, p. 1,235.
  14. Mearns, E.W., Bramwell, M., and McBride, J.J., "Application of 87Sr/86, d18O and d13C Isotopes to Diagenesis, Correlation and Connectivity of a Fractured Chalk Reservoir, the Sidi El Kilani field, Tunisia" (extended abs.), AAPG Bull, Vol. 79, 1995, p. 1,235.
  15. Smalley, P.C., Dodd, T.A., Stockden, I.L., Raheim, A., and Mearns, E.W., "Compositional heterogeneities in oilfield formation waters: identifying them, using them," in Cubitt, J.M., and England, W.A., eds., "The Geochemistry of Reservoirs," Geological Society, London, Special Publication 86, 1995, pp. 59-69.
  16. Mearns, E.W., and McBride, J.J., "Hydrocarbon filling history and reservoir continuity of oil fields evaluated using 87Sr/86 isotope ratio variations in formation water, with examples from the North Sea," Petroleum Geoscience, Vol. 5, 1999, pp. 17-27.

The authors

Click here to enlarge image

Euan Mearns is managing director of Isotopic Analytical Services Ltd. in Aberdeen. For more than 15 years he has worked on developing applications of isotope data to problems within petroleum geology focusing on refined models for fluid connectivity and stratigraphic layering. He has BSc and PhD degrees from Aberdeen University and was a researcher at the University of Oslo for 6 years. He established IAS in 1991. E-mail: [email protected]

Click here to enlarge image

John McBride is geology supervisor at IAS with particular responsibility for developing integrated core fluid geochemistry applications. His career began with Robertson Research. After completing a PhD at Aberdeen University in 1992 he was with the Geochem Group briefly before joining IAS in 1992. Since joining IAS, he has worked on the application and integration of Strontium Residual Salt Analysis data with log and pressure data on more than 100 fields worldwide.