A thermal maturity map based on vitrinite reflectance of British coals

A compilation of new and previously published vitrinite reflectance (R0) data from Carboniferous coals constitutes the most comprehensive map of reflectance across Great Britain. Values of R0 range from 0.38 to 3.29%, recording an ambient thermal maturity in the early oil window (standard reference point for reflectance studies), modified by elevated heat flow in northern England and along the Variscan orogenic front. The map provides a context for other geological datasets. Supplementary material: A statistical summary of vitrinite reflectance data is available at https://doi.org/10.6084/m9.figshare.c.4529969


Use of vitrinite reflectance data
Vitrinite reflectance is a measure of the cumulative thermal maturity of coalified plant remains. The use of vitrinite reflectance data is diverse, and includes the following examples: (1) estimation of remaining potential to yield hydrocarbons, including conventional oil and gas, and shale gas (e.g. Hackley & Cardott 2016); (2) determination of palaeogeothermal gradients, by measuring changes in vitrinite reflectance with depth (e.g. Marshall et al. 1994); (3) assessment of broad regional variations in maturity owing to metamorphism and/or tectonic history (e.g. Creaney 1980); (4) assessment of local variations in maturity owing to igneous intrusions (e.g. Bishop & Abbott 1995); (5) measurement of displacement across major faults through contrast in maturity across the fault (e.g. Cavailhes et al. 2018); (6) assessment of provenance of clasts containing coaly matter, in sedimentary successions (e.g. Vandenberghe 1976); (7) identification of potential sites for exploitation of geothermal energy (e.g. Manning et al. 2007); (8) identification of palaeogeothermal anomalies to guide mineral exploration (e.g. Maynard et al. 2001); (9) combination with apatite fission-track data to deduce the thermal histories of basins (e.g. Bray et al. 1992); (10) assessment of provenance of coal fragments encountered in archaeological sites (e.g. Smith 1996).
The wide-ranging applicability of vitrinite reflectance data makes a database of compositions valuable.

Methods
There are 14 new analyses in this study, from Inninmore Bay, East Trodigal, High Tirfergus, Uddingston, Arran, Ascog (Bute), Kello Water, Jenkin Beck, Tan Hill, Rowanburn, Wrexham, Hanwood, Pembroke and Midsomer Norton. Samples for new vitrinite reflectance analyses were crushed and then mounted in epoxy resin and polished, according to the method of Bustin et al. (1990). Samples were analysed in reflected, non-polarized, monochromatic light (λ = 546 nm) under oil immersion (ν = 1.518) using a Zeiss Axioskop MPM400 microscope equipped with MPS 200 system by J&M Analytik AG. The standard materials used to calibrate the microscope depend on the coal rank and are spinel with relative reflectance (R r ) of 0.426%, sapphire with R r of 0.585%, YAG (yttrium-aluminium-garnet) with R r of 0.905%, GGG (gadolinium-gallium-garnet) with R r of 1.72 and cubic zirconia with R r of 3.09. About 50 measurements for each sample were made to statistically constrain heterogeneities in the analysed kerogen.

Data
The vitrinite reflectance (R 0 ) values range from 0.38 to 3.29% (Figs 3 and 4; Table 1 and supplementary material Appendix I). The distribution of values (Fig. 3) shows that a majority of samples have R 0 values in the range up to 0.80%. A smaller group are in the range 0.80-1.0%, and a third group in the range from 1.0% upwards has a distinct geographical distribution (Fig. 4). Depths given for samples in Table 1 are estimated based on information (where available) on coal seam depth, depth of extraction or depth of the colliery from where the samples were taken. In general, samples from boreholes are stratigraphically well constrained, whereas spoil and ex situ samples are less accurate. Borehole sampling also provides a better opportunity to sample at greater depths than what may be achievable for samples collected through conventional shaft sinking excavation. Depths presented are also subject to available data in references or obtainable colliery or site information. The depths presented do not account for any uplift, burial or erosion, and therefore may not fully represent the maximum burial experienced by the samples whilst in situ. No correlation is observed between estimated sample depth and R 0 %.   Table 1.

Discussion
Reflectance data presented here are constrained to the highest possible accuracy, in terms of considering R 0 % standard deviation of each sample for newly acquired data (see Appendix I), and assessing the adequacy of methods used in published data. However, it is important to note that when interpreting the values presented, there are inherent expected variations of reflectance with vertical depth in a Coal Measures sequence and in the source data referenced. Data presented and assumptions made here represent conservative interpretations, which can be more rigorously tested with additional data (e.g. more samples for areas where one or few samples have been analysed) and studies on a given sample, region or stratigraphic section. The values below 1.0% are typical of sedimentary basins in western Europe, where Carboniferous sediments are in the window of oil generation. The oil generation window extends from 0.5 to 1.3%, equivalent to coals that are classified as bituminous (Tissot & Welte 1984;Petersen 2006), typically formed at 2-6 km depth and 50-150°C (Bjørlykke 2015;Mani et al. 2017).
The values above 1.0% are from regions that have experienced anomalous localized heating. The samples from South Wales, the Bristol region and Kent form a linear belt in the vicinity of the Variscan (Hercynian) orogenic front. During orogenesis, hot fluids    were expelled northwards, causing mineralization in South Wales, and contributing to the maturation of coal Alderton et al. 2004). The Variscan thrust system propagated northwards into the coalfield even as Pennsylvanian sedimentation was continuing (Gayer et al. 1998). The pattern of elevated reflectance in coals adjacent to the orogenic front extends west across Ireland (Clayton et al. 1989) and east in continental Europe (Koch 1997). In the north of England, coals on the Alston Block were subject to anomalous heat flow from the Devonian Weardale Granite (Creaney 1980;Manning et al. 2007). There is a marked change in reflectance over just a few kilometres from high values in the Pennines to low values in the Ingleton Coalfield, across the Craven Faults. The coal on Arran was metamorphosed by the intrusion of Paleogene granite (Fyfe et al. 1993). Contact metamorphism associated with laterally extensive Tertiary basaltic dykes may have also affected coals across Ayrshire (southern Scotland), Cleveland (north Yorkshire), Anglesey (north Wales) and central England.
The data imply that the potential for definitive discrimination of source regions in provenance studies is limited. Coals from along the Variscan orogenic front can be identified, and a regional distinction can be made between northern England and regions to the north and south. Consequently, where reflectance has been used for provenance in archaeological studies (Smith 1996(Smith , 2005Erskine et al. 2008), the results are confirmatory rather than definitive.
Contact metamorphism can play a pivotal role in organic matter thermal maturity (Bishop & Abbott 1993, 1995. Some coals may have experienced short-term additional heating. For example, samples from Bute and Fife exhibit veining by calcite owing to hydrothermal activity. However, reflectance is a kinetically controlled parameter, and so is influenced by long-term heat flow rather than short-term events, as observed elsewhere (Parnell et al. 2005).

Conclusions
The reported database is the first comprehensive collation of vitrinite reflectance for British coals. In addition to providing a frame of reference for thermal maturity in sedimentary sections of Carboniferous age, the database of vitrinite reflectance has applications for diverse purposes that involve national surveys. For example, the data can be used in assessments of coal bed methane (DECC 2013), shale gas potential (Smith et al. 2011), geothermal energy (Gluyas et al. 2018) and the interpretation of trace element data in coals (Bullock et al. 2018 (2013) Lat., latitude; Long., longitude; R 0 , mean vitrinite reflectance; n, number of analyses; Inferred Fm, inferred host formation; CM, Coal Measures (Upper, Middle, Lower); WG, Warwickshire Group; CG, Clackmannan Group; LCG, Limestone Coal Group; LCF, Lower Limestone Formation; YG, Yoredale Group; n.g., number of analyses not given or not deciphered from source. Depths given estimated based on seam stratigraphy, shaft excavation depth or borehole data (where available).