2025 Gilbert H. Cady Award

Presented to Joan S. Esterle

Citation by Jim Hower

Joan Esterle, born in Louisville, Kentucky, and educated at Duke and Kentucky; made her way to New Zealand for a post-doc at Canterbury; and then to Australia for positions in CSIRO, in industry, and finally at the University of Queensland. Along the way, she made outstanding contributions to the understanding of peat depositional systems, coal bed gas evolution, coal exploration, and the industrial uses of coal, among other aspects of coal science.

Joan has had an outstanding career as a coal petrographer and geologist. Her dedication is reflected in the interval between getting her MS degree in 1984 and starting her PhD studies in 1987, both at Kentucky, when she held five positions, each dedicated to furthering her peat and coal science education. The journey was inspired by a sojourn made by John Ferm, her MS and PhD advisor, during his graduate career, but the path was pure Joan. She returned to Kentucky ready to plunge into a project that greatly enhanced our appreciation of coal-forming environments.

Her years with CSIRO provided her with an appreciation of the importance of applications of coal geology and petrology to industry. No doubt, she also educated the coal industry on the applications that would be important to them. Industry interactions taught her to prepare her students for research in applied coal science. Preparation was a common theme throughout her graduate school, CSIRO, and University of Queensland careers. In graduate school, she interacted with a steady stream of coal industry geologists and engineers visiting John Ferm at Kentucky and was made aware of the applied engineering research at the University of Kentucky Center for Applied Energy Research. Coal geology is an economic geology discipline; esoteric studies may be part of our output but, ultimately, the most valuable feedback does not come from fellow academicians, but from coal-company geologists reading our work and applying the studies to their real-world situations.

Just as Joan is recognized by this award, her unique experiences and contributions enrich the Energy Geology Division and enhance the prestige of the Gilbert Cady Award, as has been true for all of her predecessors at this podium.

Response by Joan Esterle

My brush with brightness

I am humbled by the laudation for the Geological Society of America’s Gilbert Cady Award written by Jim Hower and supported by Steve Greb and Tim Moore. I have spent my professional career abroad, but my intellectual backbone was formed at the University of Kentucky (UK) from 1982 to 1990. There I learned about the connections between organic petrology, palaeobotany, organic geochemistry, sedimentary environments and modern peat mires as predictive models for ancient coal-bearing strata- and brightness profiles. Eventually I was forced to leave the nest, but my mentors at UK—Jim Hower, John Ferm, Jerry Weisenfluh, and my fellow Fermites—formed the foundation for my professional career in Australia.

Jim Hower introduced me to lithotypes and coal macerals, the microscopic organic equivalent to minerals (Stopes, 1935) derived from the varying degrees of degradation of plant communities that show successions within modern mires and have evolved over geological time scales (O’Keefe et al, 2013). While we argue about how some of these macerals formed, and how best to interpret stratigraphic trends in maceral composition and mire types, we forget an important feature—texture as used by Cady (1939) and Schopf (1960). Modern mires are defined by the source of water and nutrients, and whether they host communities dominated by big trees, little trees, shrubs, ferns or herbaceous plants- so the peats should reflect this.

Way back in the early 1980s, Jerry Weisenfluh pointed out to me that bituminous coal seams were often brighter at the bottom, and dulled upwards towards the roof, but didn’t know how this varied spatially. I spent my MSc scuttling in a 36” coal seam split like a crab in cross section to find out. Brighter coals were thought associated with the presence of big trees and occurred at the base and margins of peat deposits where nutrients were supplied by adjacent water courses. In contrast, the centre of the coal was duller and derived from more easily decomposed smaller or more herbaceous plants, similar to modern bogs (Esterle and Ferm, 1984). For the Hance seam, this helped to predict seam splits in advance-if the coal turns bright, turn right. I spent the next 6 or so years in cold, temperate and tropical mires mentored by Cornelia Cameron of the US Geological Survey, to learn that our “dome hypothesis” for coal type didn’t always work (Esterle, Ferm and Tie, 1990). Something else had caused the intense decomposition prior to or during burial. Dull coal could be derived from drowning or from drying (i.e. the wet and dry durains of A.H.V. Smith [1962]). Steve Greb nicely illustrated the complexity of different mires, their demise and regeneration, leading to multiple cycles stacked through a coal seam (Greb et al., 2002).

In parallel, Tim Moore was working in thick Cenozoic coals in Indonesia, which were so bright that it was difficult to discern banding at sub- to bituminous ranks. Ferm wouldn’t let Tim crush the coal thereby preserving the association of the macerals with their botanical sources or phyterals (sensu Cady, 1942, cited in Hower et al., 2021) with the aid of etching (Moore and Ferm, 1992). To capture this information, a very cool computer program was written in HypercardTM (Moore and Orrell, 1991), that really needs resurrection.

Importantly, the texture of phyterals, in addition to their maceral composition, was directly related to mechanical grindability and behaviour in crushers (Moore 1990). Similar coal types and behaviours were observed for Cenozoic coals in New Zealand, where I spent a short time with Jane Newman at the University of Canterbury before heading across the ditch to Australia to work with the Commonwealth Scientific and Industrial Organisation (CSIRO) in 1992. In 2004 I moved to the University of Queensland, first part time and then as the Chair of the Vale-UQ Coal Geoscience Program, from 2010-2023.

I continued my fascination with coal lithotypes but needed to convince stakeholders of their value. I had no experience with Australian coals. I flew to Sydney, rented a car and took a packed lunch to an outcrop near Wollongong and applied Tim’s technique to the Permian Bulli Seam. Surprisingly, the phyteral size distribution was similar to the Cenozoic coals and to the modern tropical peats, even though the plants comprising them were very different (Esterle et al., 1992). Australian Permian coals had higher inertinite in their non banded dull coal, and that was often attributed to high latitude and permafrost freeze and thaw (Hunt and Smyth, 1989), or to tectonic setting and subsidence (Hunt, 1989). As a failed palaeobotanist I just started using "big trees, little trees, compost”, and looked for industrial applications.

My break came when the Australian coal industry was transitioning from open cut to underground through highwall mining. One not only had to predict ground conditions in unsupported roof by integrating high resolution sedimentology and structural analysis with geomechanical modelling, but to understand variation in coal strength for barrier pillar design and the impact of selective mining on fines generation. Smyth and Cook (1976) had analysed over 3000 brightness profiles for Australian Permian seams, with the majority exhibiting “dulling up”. Preparation plants were designed for full seam extraction of thick coals, and selectively mining the bright banded metallurgical plies (benches) had potential to overload the fines circuits and hold moisture in the product, resulting in penalties.

Enter a reason to revitalise the use of brightness profiles to predict mechanical behaviour for coals at different ranks.

Funded through the Australian Coal Association Research Program (ACARP), we (actually Graham O’Brien) broke over 60,000 cubes of coal of different sizes and orientations for lithotypes across different ranks of Australian coal. The breakage data underpinned mathematical models developed at the Julius Kruttschnit Minerals Research Centre (JKMRC) that could predict the daughter particle size distribution after an energy event (e.g. blasting, digging, handling, crushing, milling). Liberated dull and bright bands reflect their in situ thickness distribution (Esterle et al., 2002). Follow-on work (Kojovic et al, 2002) developed predictive models for washability as a function of size and O’Brien later automated microlithtype analysis using Coal Grain Analysis (O’Brien et al., 2011). Sion “Max” Pretorius, who is 2m tall, conducted sequential drop shatter-CT scan experiments on cores with different lithotypes that provided input to fragmentation and washability modelling (Liu et al., 2017). Since manual profiling can be subjective, projects were trialled along the way to automate lithotype profiling on core (Yu et al., 1997), from geophysical logs (Roslin and Esterle, 2015) and most recently using hyperspectral imaging to capture both lithotype and rank (Rodrigues et al., 2023).

Access to company brightness profiles pulled me back to looking for crabs i.e. a pod of thick coal that splits in multiple directions. In the late Permian Bowen Basin, individual coal seams run for hundreds of kilometres. Mining leases, gas leases and mine plans often stop or change hands on the splits. With support from ACARP, integrated with industry studies, we were able to merge detailed mine-site data into a regional “supermodel” for the Moranbah-German Creek coal measures (Esterle, Sliwa et al., 2002), and later the Rangal Coal Measures (Sliwa et al., 2017). These built on seminal work from Chris Fielding (Fielding et al., 2000) that continues today.

For the Moranbah Coal Measures, correlative super-seams defined a clastic wedge that thinned southwards as net coal declines. Where thick, meandering split lines defined southerly trending palaeochannel and overbank deposits; where thin, split lines aligned parallel to the palaeoshoreline. Upriver, thick (8-10m) coal seams dulled upwards overall, but increased in brighter lithotypes as they split and thinned, especially proximal to the palaeocoast. The lateral variation in interburden character, coal thickness and quality was known to the government and industry geologists, but I needed the maps to see it. The maps also stimulated more detailed mine hazard plans conveying geology into engineering design.

I found my crab(s), but they were huge and more like a stacked crabfest often over areas of stable basement, pointed out by Guy LeBlanc Smith. Detailed profiling found cycles of bright bands ranging from > 10 mm to < 1 mm in thickness occurring over stacked decimetre to metre intervals. A simplistic decompaction of units might suggest stacked “big tree-little tree” cycles on the order of 5,000-10,000 years over a 100,000 year duration for a 10 m coal seam, punctuated by flooding cycles and volcanic ash falls. The mid volatile bituminous rank of these coals precluded detailed palynology, but coal etching studies suggested the bright bands were dominated by Glossopterid-wood, at least the thick ones, or no trees at all (Creech, 2002).

Combining phyterals, macerals and δ13C data through the late Permian, van der Wetering et al. (2013) decoupled shorter duration autogenic cycles in coal seams from longer scale allogenic cycles between measures. Although seams tended to “dull up then drown” they became increasingly duller and more inertinite-rich with increased fluvial influence in the Rangal Coal Measures as the Bowen Basin foreland developed. The story is of course more complicated (e.g. Fielding et al., 2022 and references therein), but the integration of the coal facies with the basin development and changing palaeoclimate is still a challenge for me, best tackled by others (e.g. Ayaz et al., 2016; Wheeler et al., 2018; Troup, 2025).

The burgeoning coal seam gas industry spawned a variety of studies. Dawson and Esterle (2010) developed a direct relationship between cleat spacing and bright or dull band thickness in bituminous rank coals. This highlighted bright banded plies of increased permeability but compromised strength leading to borehole instability. Potential differences in reservoir stimulation between “brighter” and “duller” coal measures were demonstrated by hardness and fracture propagation tests (Klawitter et al., 2015). Some measures contain a lot more mineral matter than others (e.g. Rodrigues, unpublished; Davis et al., 2021; Sun, et al., 2020) and stimulation requires additional chemical methods (e.g. Jing et al., 2021) and an understanding of structure and stress (Sliwa et al., 2018; Rajabi et al., 2024).

Jurassic coals of the Surat Basin became a major gas play so we explored fracture permeability (e.g. Mukherjee et al, 2021) and fracture toughness (Barbosa et al., 2019) for these sub-bituminous coals lithotypes. A problem we had in the Jurassic Walloon coal measures was the increased heterogeneity and a lack of regional stratigraphic markers (discussed in Hamilton et al., 2014; Andrade et al, 2025, and many papers in between). Although there weren’t any crabs we did find support for our regional correlations in the consistent stratigraphic trends in maceral, phyteral analysis, and δ¹³C signatures (Hentschel et al., 2016). They also highlighted when doming just might take place in the sequence.

I could go on and talk about finding marine signatures in fluvial systems (Bianchi et al., 2018), but that’s a story for Fermites. In closing this ramble, one can tell a lot about how a coal will behave by looking at it (e.g. Congo et al., 2024). It starts with a simple question and ends up with a career.

References