OR/14/050 Model datasets: Difference between revisions

From MediaWiki
Jump to navigation Jump to search
m (1 revision imported)
No edit summary
 
Line 11: Line 11:


==Map data==
==Map data==
Outline and outcrop curves were extracted for the modelled horizons from DiGMapGB at 1:50&nbsp;000 scale (BGS, 2013) and from previous studies in the Firth of Forth (Monaghan, 2012<ref name="Monaghan 2012">MONAGHAN, A A. 2012. Model metadata report for the CASSEM Forth model. ''British Geological Survey Internal Report'', IR/12/002. 31pp.</ref>).
Outline and outcrop curves were extracted for the modelled horizons from DiGMapGB at 1:50&nbsp;000 scale (BGS, 2013) and from previous studies in the Firth of Forth (Monaghan, 2012<ref name="Monaghan 2012"></ref>).


The outline curves were filtered and densified to an appropriate resolution (e.g. filter 150&nbsp;m, densified 750&nbsp;m on base West Lothian Oil-Shale unit) in GOCAD<sup>®</sup> and edited manually where necessary to lie on the outer side of faulted boundaries (to ensure a continuous faulted contact in GOCAD<sup>®</sup>). As only deeply buried strata were of interest to the DECC shale study, only the  major extent polygons have been included, such that smaller isolated extents mapped at 1:50&nbsp;000 scale are missing.
The outline curves were filtered and densified to an appropriate resolution (e.g. filter 150&nbsp;m, densified 750&nbsp;m on base West Lothian Oil-Shale unit) in GOCAD<sup>®</sup> and edited manually where necessary to lie on the outer side of faulted boundaries (to ensure a continuous faulted contact in GOCAD<sup>®</sup>). As only deeply buried strata were of interest to the DECC shale study, only the  major extent polygons have been included, such that smaller isolated extents mapped at 1:50&nbsp;000 scale are missing.
Line 89: Line 89:
A total of 1,325&nbsp;km (823&nbsp;miles) of onshore seismic data was interpreted along with 478&nbsp;km (297&nbsp;miles) of previously interpreted offshore 2D seismic data (Figure&nbsp;2). The seismic data, dating from 1977–88, are of variable quality, ranging from poor in the Midlothian area and around the Rashiehill borehole and Bathgate Hills, to moderate-good in the Firth of Forth. An iterative approach was employed, finding seismic lines with the good evidence for horizon mapping and well ties, then circling back through the poorer quality lines, with an interpretation that was consistent with the BGS outcrop mapping and with nearby wells. Some areas of seismic data such as over the Burntisland Anticline and Kilsyth Basin had no well ties and time-converted borehole and mining data were used as a guide.
A total of 1,325&nbsp;km (823&nbsp;miles) of onshore seismic data was interpreted along with 478&nbsp;km (297&nbsp;miles) of previously interpreted offshore 2D seismic data (Figure&nbsp;2). The seismic data, dating from 1977–88, are of variable quality, ranging from poor in the Midlothian area and around the Rashiehill borehole and Bathgate Hills, to moderate-good in the Firth of Forth. An iterative approach was employed, finding seismic lines with the good evidence for horizon mapping and well ties, then circling back through the poorer quality lines, with an interpretation that was consistent with the BGS outcrop mapping and with nearby wells. Some areas of seismic data such as over the Burntisland Anticline and Kilsyth Basin had no well ties and time-converted borehole and mining data were used as a guide.


Seismic data from the National Coal Board were not utilised as it is focused at shallow depths and some Geological Survey lines (e.g. Line IGS-1982, Forsyth et al., 1996<ref name="Forsyth 1996">FORSYTH, I H, HALL, I H S, and MCMILLAN, A A. 1996. ''Geology of the Airdrie district''. Memoir of the British Geological Survey, Sheet 31W (Scotland).</ref>) were not available  in digital format. In the Firth of Forth, existing BGS interpretations made on reprocessed seismic data tied to the Firth of Forth&nbsp;1 well were used (Monaghan et al., 2012<ref name="Monaghan 2012">MONAGHAN, A A. 2012. Model metadata report for the CASSEM Forth model. ''British Geological Survey Internal Report'', IR/12/002. 31pp.</ref>).
Seismic data from the National Coal Board were not utilised as it is focused at shallow depths and some Geological Survey lines (e.g. Line IGS-1982, Forsyth et al., 1996<ref name="Forsyth 1996">FORSYTH, I H, HALL, I H S, and MCMILLAN, A A. 1996. ''Geology of the Airdrie district''. Memoir of the British Geological Survey, Sheet 31W (Scotland).</ref>) were not available  in digital format. In the Firth of Forth, existing BGS interpretations made on reprocessed seismic data tied to the Firth of Forth&nbsp;1 well were used (Monaghan et al., 2012<ref name="Monaghan 2012"></ref>).


The seismic horizon and fault interpretations were depth converted using a linear correlation derived from velocity-time data measured from 8&nbsp;wells (this was all the available velocity data).
The seismic horizon and fault interpretations were depth converted using a linear correlation derived from velocity-time data measured from 8&nbsp;wells (this was all the available velocity data).

Latest revision as of 13:12, 29 November 2019

Monaghan, A A. 2014. Model metadata report for BGS - DECC shale study model, central and eastern Midland Valley of Scotland. British Geological Survey Internal Report, OR/14/050.

DTM/Bathymetry

An existing combined 500 m resolution DTM and bathymetry surface derived for the Lithoframe 250k 2008 (Monaghan and Pouliquen, 2012[1]) model was used as the capping surface to the model. For a model at this scale, improvements to the DTM/bathymetry since 2008 should make negligible difference. The DTM/bathymetry is used as an approximation to the rockhead surface. In some areas of the Midland Valley with a thick covering of superficial deposits this approximation is not ideal and future work should use a rockhead surface which incorporates recent superficial deposits modelling.

Borehole and well data

Borehole data for base stratigraphic units were recalled from the BGS.Borehole_Geology database using the ‘Magpie’ Access application in April 2014 for the stratigraphic codes given in Table 1 and for boreholes greater than 20 m drilled depth. During the recall, interpreters were prioritised in the order AAMI, other interpreters entering ‘DV’ or ‘OV’ content codes, TMCM, then DJLO. The recalled data was edited manually and saved in a new file to give records reaching only the base of the stratigraphic horizons.

Base stratigraphic surface depths from DECC wells not entered into BGS.Borehole_Geology, or where a new interpretation to that in BGS.Borehole_Geology was made during the work, were manually added to each data file.

Map data

Outline and outcrop curves were extracted for the modelled horizons from DiGMapGB at 1:50 000 scale (BGS, 2013) and from previous studies in the Firth of Forth (Monaghan, 2012[1]).

The outline curves were filtered and densified to an appropriate resolution (e.g. filter 150 m, densified 750 m on base West Lothian Oil-Shale unit) in GOCAD® and edited manually where necessary to lie on the outer side of faulted boundaries (to ensure a continuous faulted contact in GOCAD®). As only deeply buried strata were of interest to the DECC shale study, only the major extent polygons have been included, such that smaller isolated extents mapped at 1:50 000 scale are missing.

Mine plan data

Two sources of mining data were incorporated into the model. BGS compilations of mine abandonment plan data for coals within the Limestone Coal and Upper Limestone formation were used as a guide to the base of depth of these formations by addition of a standard thickness (derived from generalised vertical sections on BGS 1:10 000 and 1:50 000 maps). Table 3 shows the seams used and assumed thickness to the base of the Limestone Coal Formation. Where there was more than one seam covering an area, the lowermost seam was used. These thicknesses are a simplification, as thicknesses are known to vary laterally.

Table 3    Coal seam abandonment plan data in the Limestone Coal Formation that was used to guide the depth of the base of that unit in the model. The thickness from the coal seam to the base Limestone Coal Formation is given in brackets (in metres) and the approximate stratigraphic position is indicated by the position in the table
Glasgow Lanarkshire Clackmannan West Fife Lothian
Meiklehill Main MEM (231)
Kelty Main KYMA (171)
Great Seam GSC (116)
Wilstontown Main WNMA (124) Bannockburn Upper Main Coal BNUMA (240)
Bannockburn Main Coal BNMA (210)
Knightswood Gas KDG (183)
Kilsyth Coking KHCC (102) and (62) on NW side central coalfield Dunfermline Splint DESP (60) Kailblades/Corbie Craig KCG (43)
Arniston Parrot ARP (7)

The standard thickness used from the Upper Hirst coal seam to the base of the Upper Limestone Formation was 200 m.

A second mining-data constraint was given by the ‘mining data for all seams greater than 500 m’ licensed from The Coal Authority. This dataset was used to inform a depth cut-off for the shale resource estimation (Figure 63 of Monaghan, 2014[2]) but as the deepest worked coals are in the Limestone Coal Formation, it also provided a ‘deeper than’ constraint for the base of that stratigraphic horizon.

Seismic data

Figure 2    Location of 2D seismic profile interpretations used to constrain the DECC model. Image after Monaghan, 2014[2] © DECC.

A total of 1,325 km (823 miles) of onshore seismic data was interpreted along with 478 km (297 miles) of previously interpreted offshore 2D seismic data (Figure 2). The seismic data, dating from 1977–88, are of variable quality, ranging from poor in the Midlothian area and around the Rashiehill borehole and Bathgate Hills, to moderate-good in the Firth of Forth. An iterative approach was employed, finding seismic lines with the good evidence for horizon mapping and well ties, then circling back through the poorer quality lines, with an interpretation that was consistent with the BGS outcrop mapping and with nearby wells. Some areas of seismic data such as over the Burntisland Anticline and Kilsyth Basin had no well ties and time-converted borehole and mining data were used as a guide.

Seismic data from the National Coal Board were not utilised as it is focused at shallow depths and some Geological Survey lines (e.g. Line IGS-1982, Forsyth et al., 1996[3]) were not available in digital format. In the Firth of Forth, existing BGS interpretations made on reprocessed seismic data tied to the Firth of Forth 1 well were used (Monaghan et al., 2012[1]).

The seismic horizon and fault interpretations were depth converted using a linear correlation derived from velocity-time data measured from 8 wells (this was all the available velocity data).

The velocity data from several of the wells plotted some way off the linear correlation (on both the high and low side) with the result that the depth-converted seismic interpretation did not fit exactly the observed well depths. After import of XYZ data points interpreted for each horizon and faults from the 2D seismic data, several steps were undertaken in GOCAD® to remedy the depth-data mismatch:

  1. The base Limestone Coal Formation (bLSC) surface was used as a reference surface since it can be constrained by a great deal of projected mining data as well as a good quantity of borehole data. An unfaulted bLSC surface was created with all the borehole, mining and outcrop data.
  2. The bLSC surface was fitted to the borehole and well picks pointset ‘surface-tools-fit-to pointset’ using 2 iterations and setting control nodes.
  3. In selected regions with extreme disparity between the projected mining and seismic data (areas where there is no well control but good mining control) the bLSC was fitted to the borehole+mining+outcrop derived pointset (using 2 iterations and without setting control nodes).
  4. In the selected regions identified, the seismic data points were then edited to fit the depth surface (transfer the Z property subtract the difference using the calculator). A new combined data file for bLSC was then created with the revised seismic-depth data points included.
  5. To maintain the consistency of thicknesses within the seismic interpretation the following method was used for the remaining four surfaces:
  • A thickness isopach from original seismic pick to the revised bLSC pick was created.
  • The thickness isopach was added onto the bLSC depth-corrected seismic pick to maintain consistency.
  • The revised seismic depth picks were merged with the other data sources.
  1. When the faulted modelled surfaces were created, a fit to borehole/well point iteration was performed for each surface (inserting control nodes) to ensure that the modelled surfaces honoured the well/borehole points.

The depth conversion process introduced significant uncertainty into the dataset. Due to the observed variability in seismic velocities, future work would benefit from building a more complex velocity model representative of individual wells.

Previous models used by this model

Small parts of existing models were utilised by this model namely:

  • an area of the Lower Limestone Formation in central-west Fife with no constraining seismic datasets from Monaghan (2013)[4].
  • the base of the Lawmuir Formation (= base West Lothian Oil-Shale unit) at the margins of the Clyde Plateau Volcanic Formation blocks around Glasgow after Millward and Stephenson (2011)[5].
  • The base Upper Limestone Formation and base Lower Limestone Formation at the margins of the western Central Coalfield after McCormac (2012[6], 2013[7]) and/or Monaghan (2013)[4].

Pointsets were extracted from regions of these surfaces and included within the combined data file for each stratigraphic surface.

In the Firth of Forth, the base Upper Limestone Formation was not previously interpreted on seismic data and is not present in the Firth of Forth 1 well, nor in any borehole that ties to the seismic data. For the purposes of this study a standard thickness of 220 m was added to the base Limestone Coal Formation seismic interpretation (based on the formation thickness in boreholes to the north and south of the Firth of Forth). Subsequent reductions to that thickness were made to ensure consistency with overlying surfaces and rockhead. The result is that the base Upper Limestone Formation surface is particularly poorly constrained between the offshore seismic and onshore datasets at the north-eastern end of the Midlothian-Leven syncline. In this area the unit as modelled is very thin and steeply dipping and could be improved upon by further work.

References

  1. 1.0 1.1 1.2 MONAGHAN, A A, and POULIQUEN, G. 2012. Model metadata report for Midland Valley Lithoframe 250k model. British Geological Survey Internal Report, IR/12/008.
  2. 2.0 2.1 MONAGHAN, A A. 2014. The Carboniferous shales of the Midland Valley of Scotland: geology and resource estimation. British Geological Survey for Department of Energy and Climate Change, London, UK.
  3. FORSYTH, I H, HALL, I H S, and MCMILLAN, A A. 1996. Geology of the Airdrie district. Memoir of the British Geological Survey, Sheet 31W (Scotland).
  4. 4.0 4.1 MONAGHAN, A A. 2013. Model metadata report for the Midland Valley of Scotland Regional Model 2012/13. British Geological Survey Internal Report, IR/13/013. 34pp.
  5. MILLWARD, D, and STEPHENSON, D. 2011. Bedrock GSI3D models from interpreted data in geologically complex Carboniferous terrains: A work in progress from the Clyde catchment area, Midland Valley of Scotland. British Geological Survey Internal Report, IR/11/052. 58pp.
  6. MCCORMAC, M. 2012. Clyde Catchment GOCAD® bedrock models, 2009 to 2012; background information, model descriptions and methodology. British Geological Survey Internal Report, IR/12/046. 32pp.
  7. MCCORMAC, M. 2013. Clyde Catchment GOCAD® bedrock regional model, 2013. British Geological Survey Internal Report, IR/13/014. 11pp.