What is shale? Clarifying the muddy terminological waters

Every student in SYGN 101 learns that “shale” refers to a clay-rich, fine-grained, laminated (thinly-bedded) sedimentary rock that breaks apart in sheets. Any geologist could quote this definition. But, as with most words, “shale” has come to mean many things over the years which have little to do with the original definition.

Dr Bruce Hart, a researcher from Statoil Houston and a former AAPG (American Association of Petroleum Geologists) Distinguished Lecturer, began his talk with a multiple choice question: is a shale a rock type, a stratigraphic designation, or a play (ie. oil/gas reservoir) type? The answer, he revealed, depends on who is being asked to define it. The industry, he observed, uses the term “pretty much indiscriminately”, leading to confusion as to its precise meaning.

First, he took a lithological approach. Even in academic literature, the exact meaning of a shale is unclear. In some places, the “clay” in “clay-rich” means clay minerals, such as illite and kaolinite. In other places, “clay” means clay-size particles, without any consideration for the composition of those particles. Furthermore, while laminations are easy to define, fissility, or the tendency of a rock to break into thin plates, is difficult to see in a drill core, for instance, because fissility is often the product of weathering processes and a rock that has never been exposed to the surface will be unweathered. In other words, a fissile rock at the surface might, at depth, be solid and apparently unlaminated. Because oil and gas researchers, in the few instances when they see a rock and not a wire-line log, deal mainly with cores rather than outcrops, this presents a problem.

The disjoint between academia and the industry only becomes wider when the “clay” component of the definition is considered. Grain size is not used in the oil industry; rather, rock samples are ground up and put through x-ray mass spectrometry (XRD) analysis, which gives the bulk elemental composition of the rock, and, from there, the mineralogy. This gives a highly simplified picture of what may in fact be a very complex rock. For this reason, the industry lexicon tends to assume that any formation is homogeneous, when in truth there is heterogeneity at every scale.

To illustrate this, Hart gave an example of a rock with XRD results indicating a mixed carbonate/clay mineralogy. Citing different well-known shale plays, such as the Marcellus in Pennsylvania, the Barnett in Texas, and the Haynesville in Louisiana, Hart listed various forms the carbonate minerals might take in hand specimen. In the Marcellus, much of the carbonates are present as fossils, including shells and corals, some quite large. The Barnett’s carbonates include turbidites, or high-speed underwater landslide deposits, and replacement textures, where the carbonates have taken the place of a different mineral. The Haynesville has carbonate concretions and fracture fills, where a calcite-rich fluid came in after the formation had lithified and filled in the cracks. Other formations have alternating layers of clay-rich and carbonate-rich rocks, chalks, or even flattened, fossilized fecal pellets. XRD might give good chemical analysis, but it captures none of this detail.

The problems with heterogeneity in the rock being ignored or smoothed over in the oil industry do not end with rock textures. On the outcrop scale – usually below the resolution of seismic data – a bed is rarely without variation. Sedimentary rocks often go through cycles, seen by repeated stratigraphic units. There might be layers of alternating mineralogy: clay-rich (or muddy) to sand-rich (or clean), for instance. There might be alternating modes of deposition: turbidites overlain by background sedimentation. The alternating units might be organic-rich and organic-poor (of prime importance to the oil industry), or they might differ in competence, blocky rocks overlain by fissile ones. These small cycles then stack into larger-scale cycles, called bedsets, which might represent a fall and rise in sea level. The layered variations may also be found at very small scales in the form of microscopic laminations. When the wealth of detail available in the rocks is discounted by a single blanket term – like “shale” – the industry loses potentially valuable information. Thus, Hart recommends a new classification scheme for rocks, which could accommodate both mineralogy (from XRD) and texture (from observation).

The confusion does not end here, however. Once again, the question of grain size vs. mineralogy arises. When a driller uses the term “shale” or “clay”, it indicates a fine-grained rock, disregarding lithology entirely. The epithet then gets grandfathered in, so that a unit like the Eagle Ford formation of Texas is called a shale when it is, in truth, a carbonate. Indeed, using a mineralogical cut-off of fifty percent clay to be a shale and any less to be some other type of rock, none of the famous shale plays in the United States would be called a shale at all. Because a gas shale is generally only productive when hydro-fractured, clay-rich rocks tend to scare operators away, as too much clay means the rock will not fracture well.
This led Hart to his final discussion: what constitutes a shale play? The first productive shale, ironically, was a true textbook-definition shale which had natural gas seeps flowing from it; after some village children accidentally set a nearby creek on fire, the townsfolk dug a twenty-five-foot-deep well and began tapping the gas seep to light their lamps. In modern times, however, the wells can be several miles long and must be hydro-fractured in order to draw gas; this involves breaking open the rock at depth to create a space where the gas can flow. A shale play is a tight, brittle rock which is both the source and the reservoir for the oil and gas that is being produced. It must be drilled into horizontally in order to be productive, as the units involved cannot be too thick or they will not fracture well.

Since the media discovered hydrofracturing (or “fracing”), however, any oil/gas play that needs to be fraced to be economical is being called a shale, even by the AAPG. This is both erroneous and misleading. Hart cited multiple examples of play types that have been lumped under the umbrella of “shale gas play”. The Bakken formation of Montana is neither a shale nor does it produce from its source rock. Instead, it is a “sandwich” play, where the reservoir units – where the gas is stored – are bracketed by source units – where the gas is produced – and the wells are drilled into the reservoirs. Other mislabelled fields include “halo” plays, conventional reservoirs surrounded by tight rock; and stacked or “hybrid” plays, where thin alternating conventional and unconventional reservoirs are produced from a vertical well.
To better illustrate his point, Hart drew a Venn diagram. On the outside was a circle containing all mudstones (that is, all clay-rich, fine-grained sedimentary rocks). Within this were three overlapping circles representing thermally mature mudstones (meaning they have the right burial history to produce oil/gas), organic-rich mudstones (meaning the material from which oil and gas is derived is present in abundance), and brittle mudstones (meaning ones which can be effectively fractured). Where the three fields overlap lies the true shale play, as it can be both a good source rock and a frac-able reservoir. Anything else should be called something different.

Hart concluded that better communication between the disciplines and better, more precise vocabulary will allow better prediction of sub seismic heterogeneity in the rock, and thus more efficient oil and gas production. He ended with a quote from the book Alice in Wonderland: paraphrased, a word means what the person using it wants it to mean, nothing more and nothing less.

Ironically, one of the attendees at the talk pointed out that in many other languages, including French and Spanish, there is no word for shale at all. The word for schist, a metamorphic rock completely distinct from shale, is used instead. Clearly, when it comes to clarifying the muddy waters of shale terminology, Dr Hart has his work cut out for him.