The end-product desired of a 1-D cycle analysis is a well-bore divided into depositionally and petrophysically significant packages (cycles and their component lithofacies). The cycle framework will form the basis of a 2-D stratigraphic correlation framework and 3-D model.

In contrast to siliciclastic successions where wireline log responses to most lithofacies are fairly well defined and there is a good match between grain-size and gamma counts, log-lithofacies interpretation in carbonate reservoir strata requires a hands-on examination of lithofacies in core to allow a high-resolution stratigraphic framework to be constructed. The reason for this dependency on core information is that in most carbonate environments the component of siliciclastic-bearing fines is minimal, and thus the correlation between API gamma ray (still the most common log in many reservoirs and therefore the main lithology indicator) and grain size/depositional energy, while useful, can be misleading, particularly if uranium is present in dolomite or other minerals. Once core material from carbonate reservoir sections has been examined in detail and compared visually and numerically with wireline logs, it is common that a fair relationship between core-based lithofacies and log response can be determined.

A standard interpretation of a wireline log from a Permian (Guadalupian) Grayburg reservoir succession (below) illustrates four major "upward-cleaning" packages, which represent candidate cycles. This log will be examined later to illustrate how core description can enhance - and, in fact, significantly change - the stratigraphic framework interpretation

The interpretation of lithofacies in core is an essential skill for construction of carbonate reservoir frameworks. Unfortunately, this skill is more and more being relegated to technical "specialists" in company technology centers or being out-sourced. Whereas this division of labor may be part of a currently-accepted work flow, it carries with it the probable pitfall that the core description will not be well integrated into the wireline log framework from logs alone.

We have encountered numerous examples of wireline log correlations that do not follow key surfaces defined by the core description, and a few examples of the opposite. The important thing to avoid in establishing a stratigraphic framework is the tendency to correlate gamma-ray events around the field, not knowing what they represent. This process generally leads to a layer-cake (lithostratigraphic), meaningless framework, with the rare exception of deep-water and tidal flat settings. In other words, a correlation based on log signature matching is, by definition, a lithostratigraphic ("layer-cake") correlation. The carbonate reservoir stratigrapher must understand the lithologic signature of wireline logs and the cycle hierarchy based on lithofacies description in order to interpret the stratigraphic framework. Lithofacies description and cycle interpretation are the first steps in an integrated and iterative process where data from the different tasks must be understood and utilized in order to be incorporated appropriately into the next successive phase.

The remainder of this section deals with the interpretation of 1-D vertical lithofacies successions, mainly from core, to form a cycle hierarchy useful in 2-D correlation and rock-fabric identification. Basic skills used are lithofacies identification and interpretation, interpretation of cycles, bundling of cycles into different hierarchical levels, and keying of lithofacies into rock-fabric or other petrophysically-significant lithofacies types.

Step 1: Core Overview

Scan the entire length of the cored interval to gain an appreciation for the total range of lithofacies, water depths, environments, and stratigraphic packaging. This should be done with wireline log(s) in hand to get a qualitative "feel" for the log response (e.g. Gamma Ray, CGR, Spectral Gamma Ray, Caliper, Density, Neutron, Sonic, Resistivity) in the cored interval. This step, which should proceed quickly, requires a preliminary interpretation of lithofacies using widely accepted indicators of depositional environment such as tidal flat laminites, cross-bedded grainstones, or graded cyclic turbidites. An understanding of the invertebrate fauna, ichnofauna, and its paleoecology are very useful, particularly in mud-rich low energy lithofacies that may represent either very shallow restricted environments or deeper water environments, both being settings where mud can accumulate.

Step 2: Depositional Environment

The goal of this step is to describe the major depositional environments in the core, based on the lithofacies present. It is useful to mark the most dramatic environmental shifts as seen in core and examine the associated wireline log response. Hopefully there will be some representation of these events, as they may represent the most important well log correlation ties in the data set.
In carbonate reservoir systems dominated by physical sedimentary processes, instead of reef-building or other biologically-dominated processes, the reservoir zones of key interest are those within fair-weather wave base down to a position between fair-weather and storm wave-base (lower foreshore and shoreface). These high-energy deposits have a large proportion of interparticle porosity between wave-winnowed grains, and good permeability. Tidal flat deposits and sub-storm-wave-base sediments are generally low-energy, mud-rich sites of sediment accumulation that have poorer reservoir quality. In specific circumstances where diagenesis is limited and abundant fine or chalky porosity is preserved, such as in the Cretaceous chalky carbonate fields of the Middle East, production from sub-storm-wave-base lithofacies is significant, but this is only because of high total porosity values.
In these shallow shelf deposits, the three most important environmental thresholds are the shoreline (sea level), fair-weather wave base, and storm wave base. These zones include tidal flat, beach foreshore, upper shoreface, and lower shoreface-slope (McCubbin 1982; Clifton et al. 1971) environments. Fair-weather wave base and storm wave base are not simple water depth indicators but instead are positioned on the basis of depositional energy, and vary depending on the nature of wave fetch and shoreline orientation relative to prevailing winds and storm tracks. Recognizing the potential for significant variation, a typical depth range for these environmental thresholds are 20-40 feet for fair-weather wave base, and 100-250 feet for storm wave base. Protected lagoonal settings and backreef or back-barrier settings may display significantly less wave energy, and consequently bioturbation may dominate over physical sedimentary structures at depths less than a few feet. For example, when tidal flats (sea level), cross-bedded grainstones (fair-weather wave base), and fusulinid packstone (storm wave base) lithofacies data from core are added to the log in the Grayburg example above (compare step 2 to step 3), the interpretation is enhanced. The tops of the tidal flat-capped cycles are marked as candidates for potential sequence boundaries, and fusulinid-bearing intervals are marked as candidate maximum flooding surfaces.

These three physical environmental zones control particle size and sorting of carbonate sediments, which in turn strongly affect permeability in carbonates (Lucia 1972, 1995). The following is a brief list of some of the lithofacies and sedimentary structures characteristic of the three main depositional environments and their petrophysical affinities.

Shoreline/foreshore/tidal flats
•· smooth algal laminite in lower intertidal
• fenestral cryptalgal laminite in intertidal-supratidal mudstones, wackestones and packstones
•· bubble fenestral or keystone vug porosity in grainstones
• swash stratification, or seaward-shingled wedge-set strata
•· beachrock breccias or incipient breccias
•· microkarst surfaces (rare)
•· marine vadose cements
Reservoir properties
• tidal flat rock fabrics: mud-rich packstones or wackestones with low interparticle porosity with potential abundance of vuggy fenestral porosity
• foreshore rock fabrics: high interparticle porosity grainstones

Lithofacies at or above fair-weather wave base (upper shoreface)
• general a dearth of bioturbation and domination by cross-stratification
• grain-dominated fabric with high percentage of interparticle porosity
•· trough or tabular-planar cross bedded grainstone
• sheet-stratified grainstone
• vertically migrating habitation burrows that keep up with rapid sediment accumulation

Lithofacies below fair-weather wave base (lower shoreface)
• high ratio of bioturbation/burrowing relative to cross stratification
• thin sheet stratified or hummocky (rare) stratified very fine to medium grained grainstone
• alternating massive bioturbated wacke-packstones and thinner packstone units representing periodic • physical reworking of shelf sediments followed by bioturbation
• wave ripple cross-lamination
• shell lags or intraclastic breccia beds representing storm activity

Lithofacies assemblages - Within this simple three-fold bracketing scheme there are an infinite number of vertical lithofacies assemblages, that vary depending on geologic age, basin setting, climate, proximity to siliciclastic source, and several other variables. The complete range of carbonate types is the subject of several excellent texts, for example Wilson (1975) and numerous papers (Bebout et al. 1986; Grotzinger 1986; Bova and Read 1987; Koershner and Read 1989; Goldhammer et al. 1990 1993a, b; Kerans and Nance 1991; Elrich and Read 1991; Osleger and Read 1991; Lucia et al. 1992; Holtz et al. 1992; Cowan and James 1993; Sonnenfeld and Cross 1993; Kerans and Fitchen 1995). Unfortunately only a few of these illustrate comparable log signatures. This is either because the data are from outcrop, or because there was a poor correspondence between depositional and log lithofacies. Using the simple rule of separating lithofacies into the water-depth/energy regime packages discussed above will form an important link to both interpretation of high-frequency cycle content and to the identification of petrophysical lithofacies on logs.

Step 3: Vertical Lithofacies Successions

This stage of core description is the most important, and commonly the most tedious. The detailed description is an essential step in the interpretation of the vertical lithofacies succession, which will then form the cornerstone of the subsurface 3-D model. It is easy for this step to become a somewhat mindless task, but this would be a major error.

The goal of this step is to capture and quantify as much information as necessary to describe the vertical lithofacies successions (cyclicity) and the key attributes (variables) that will be used to calibrate wireline logs. The first task is to design a graphic logging form. Most wireline logs take a reading every half foot, which represents a one to two vertical foot volume. We have found that five vertical feet per inch is typically plenty of detail for core description, and ten feet per inch is often adequate.

No two reservoirs require the same description format. The choice of what you describe must be some combination of necessary information versus cost/benefit. An understanding of the range of lithofacies and depositional environments present in the total reservoir interval (Step 2) is useful going into this step, because it gives a feel for the level of detail required for adequate reservoir description. Are there large unconformable breaks that need to be identified in terms of diagenetic overprints, truncation of strata, or onlap above. Are there systematic cycle changes preceding the major flooding surfaces? Is there an abrupt out-of-sync lithofacies shift?

General rules of thumb are to keep the number of lithofacies in any one analysis down to 8 or less, and to attempt to have lithofacies linked to a modified Dunham classification as proposed by Lucia (1995). Also, on the first-pass description do not over- or under-interpret the lithofacies. Some thin beds, such as 0.5 ft dense mudstones or 0.5 ft tidal flat laminites could be both petrophysically and depositionally far more important to reservoir performance than the surrounding lithofacies, and should be noted.

Lithofacies arrangements - Once an initial analysis of water-depth/energy regime lithofacies has been accomplished (Step 2), it is possible to scan the core to assess the general lithofacies arrangement trends. For example, are lithofacies arranged in a single upward shallowing trend, as multiple alternations of sub-fair-weather-wave-base (FWB) to above-FWB packages succeeded by above-FWB to intertidal packages, as a set of entirely intertidal lithofacies, or as some other pattern of alternation of environments is observed.

Keep in mind the age of the section to determine the icehouse-greenhouse affinity. Major lithofacies offsets and surfaces of missing time should be expected in an icehouse setting, and more systematic stacking of lithofacies in a greenhouse setting.

The "Ideal Cycle" - With the vertical lithofacies log in hand, it is necessary to extract an ideal high-frequency cycle from the data to use as a preliminary working model. This critical step is highly complex, and will almost certainly be misleading, as there is no single "ideal cycle". In fact, depending on which lithofacies tract or tracts the core intersects, it is possible that two or three typical cycles representing inner ramp, ramp crest and outer ramp, exist. However, because at this stage of the analysis there is no complete sequence or paleogeographic picture, a single universal cycle should be generated. The first attempt will almost certainly be refined later in the study, but the remainder of the 1-D stratigraphic analysis must proceed from the basis of this cycle definition.

The initial interpretation of vertical lithofacies successions (cycles) provides a template for identification of stratigraphic trends and higher-order surfaces. A simple log with a foot-by-foot description of lithofacies for 300 feet, with no grouping into units, cycles, and cycle sets, is nowhere near as useful as one that has been interpreted, if only in a preliminary way. The initial cycle picks using the ideal cycle should, if anything, err on the side of being "over interpreted" because subtle candidate cycles in one cored well may turn up as major, clearly defined packages in the next well. If these subtle changes are not marked in both wells then logical correlations may be missed. It is easy to discard cycle tops retroactively.

The cored interval should be interpreted completely in terms of candidate high-frequency cycles (HFCs). Once this is done, cycle hierarchies of lower order can be explored for, followed by possible candidate composite sequence boundaries and maximum flooding surfaces.

None of the interpretative picks in this initial analysis are set in stone. Only through correlation with adjacent wells will it be determined which vertical lithofacies successions mark significant cycle and sequence boundaries, and which merely represent local variations. These preliminary picks are a record of what the interpreter considered important surfaces at the time the data was collected, and are analogous to the cycle-picking process done with outcrop data. Too many times this preliminary interpretation stage is conducted "back in the office" where it is not possible to recheck observations.
The core description and preliminary cycle interpretations should show trends in accommodation (stacking patterns) that may correspond with the major flooding surfaces or exposure surfaces identified in core. These surfaces providing a first-pass evaluation of the cycle framework. This preliminary interpretive step, when carried out using multiple wells/cores, is the most useful technique for error-checking and validating the stratigraphic framework as it evolves.

An example of a finalized initial core description with detailed vertical lithofacies profile, and initial cycle picks interpreted from a Walther's law ideal cycle is shown for the Grayburg example. Cycle stacking patterns from this initial pass seem consistent with the major lithofacies successions picked initially, with thick cycles corresponding to fusulinid facies flooding surfaces, and thin cycles corresponding to tidal flats.

Step 4: Additional Analyses

Not all of the necessary information can be gleaned from macro core description alone. It is often necessary to examine thin sections or acetate peels in order to enhance or confirm macro description, especially of pore types, grain types, and diagenetic features.

Sample the core for petrographic analysis and any other special analysis, such as SEM, XRD, CT-Scan or SCAL. For petrographic work, is recommended for two or three samples for each major lithofacies, in order to confirm the lithofacies classification scheme.