MRCSP Phase I Geologic Characterization Report - Midwest ...

APPENDIX A: PENNSYLVANIAN COAL BEDS IN THE MICHIGAN BASIN
131
zontally into the sides of hills. Although portions of these mines can
reach depths of more than 2,000 feet beneath the surface, most are
still above drainage, or above the level of the lowest streams in the
area. For sequestration purposes, it may be important to delineate
those coal beds that are below the lowest level of streams (called
“below drainage”) to ensure the sequestered CO 2 gas will not leak to
the surface through fractures or updip along bedding partings.
Coal thickness data in each state were obtained mainly from deep
diamond-drill core holes and density logs from oil and gas exploration
wells on file at each state survey. Oil and gas data are particularly
useful for the deeper parts of the basin, although only a fraction
of the wells drilled have density logs through the Pennsylvanian
strata because, in general, coal-bearing strata are cased-off prior to
logging. The manner in which thickness data was compiled in each
state varied also because the topography, geology, mining history,
and available data were wide-ranging across the region. Data from
Maryland, Ohio, Pennsylvania, and West Virginia represent coal
beds more than 500 feet beneath the present-day land surface. In
eastern Ohio, this may also be close to the below-drainage depth
because of minimal- to moderate-topographic relief in this area. In
northern West Virginia, many of the oil and gas wells used were
cased off to depths of more than 500 feet; so much of the thickness
represented may also be below drainage. In Kentucky, data were
only compiled for coal beds more than one-foot thick and more than
500 feet below drainage. Coal beds in Maryland occur only in two
synclinal basins preserved (one isolated; see Figure A15-3) near the
eastern edge of the Appalachian basin proper. Thus, coal thickness
data are shown as individual data point values not connected to the
main contoured net-coal map for the Appalachian basin.
No data were compiled for southern West Virginia because that is
an area of active mining, and it was felt the area should be restricted
from consideration for sequestration in that state at this time. If
these criteria were used also in Kentucky, no sequestration potential
would exist in that state because areas of thickest coal at depth are
also areas of mining at or near the surface. Likewise, parts of the
thick, cumulative coal thickness in Ohio and Pennsylvania are areas
of current near-surface mining. Hence, considering these multifaceted
elements related to CO 2 sequestration into coal beds, further
analysis, public discussion, and additional data collection and mapping
will be required in Phase II of the MRCSP project.
16. PENNSYLVANIAN COAL BEDS IN THE MICHIGAN BASIN
The Pennsylvanian Saginaw Formation is a coal-bearing interval
that is restricted to the central part of the Michigan basin due
to erosional truncation. Along with other Pennsylvanian strata in
Michigan, the Saginaw is isolated from correlative coal-bearing
strata in the Appalachian basin portion of the MRCSP study area. In
general, the Saginaw Formation is equivalent to Lower and Middle
Pennsylvanian (e.g., Pottsville and Allegheny formations) strata in
adjacent basins.
Although commercial coal production occurred in the Pennsylvanian
“Coal Measures” of the central Michigan basin in the late
1800s and early 1900s (maximum production of over two million
short tons in 1907), all coal beds in the basin are now considered
noneconomic (Ells, 1979). Total current reserves are estimated at
126.5 short tons (Kalliokowski and Welch, 1976) in individual coal
beds typically less than three feet thick and laterally discontinuous
on a scale of hundreds to thousands of lateral feet (Kelly, 1936).
Complete sections of Pennsylvanian rocks in Michigan do not exist
as outcrops or in known diamond drill-hole cores. Stratigraphic
relationships are mostly inferred from subsurface geophysical logs,
well cuttings, drillers’ reports, and rare cores collected during drilling
of oil, gas, and water wells (Vugrinovich, 1984). Small-scale,
spatial,lateral and vertical lithologic variability characterizes Pennsylvanian
strata in Michigan. Various stratigraphic nomenclature
has been proposed for these strata; however, no scheme is universally
accepted due to the paucity of data, lack of lateral continuity of
most units, and complex facies relationships.
Biostratigraphic data are limited for the Saginaw Formation. Age
ranges for the unit indicate a late Morrowan (Early Pennsylvanian)
through late Desmoinesian (Middle Pennsylvanian) age, with the
majority of the Saginaw being Morrowan age (Arnold, 1949; Wanless
and Shideler, 1975; Vugrinovich, 1984). However, recent palynological
analysis indicates an Early Pennsylvanian, Atokan age for
upper portions of the Saginaw (R.M. Ravn, pers. comm., 2005).
The predominantly arenaceous facies of the Parma Sandstone is
considered the basal formation of the Pennsylvanian Subsystem in
the Michigan basin, although some workers included the Parma as
the basal portion of the Saginaw Formation (Ells, 1979). Strata of
the Saginaw Formation proper consist of heterolithic sandstones,
shales, coals, and limestones with a regionally prominent limestone
unit, the Verne Limestone Member, separating pre- and post-Verne
cycles (Kelly, 1936). The overlying Grand River Formation is a
laterally discontinuous, predominantly fine- to coarse-grained sandstone
unit (Kelly, 1936).
Wanless and Shideler (1975) subdivided the Pennsylvanian strata
of the Michigan basin, in ascending order, into a sand-dominated
basal “A” unit, a predominantly fine-grained shale, siltstone, sandstone,
limestone, and coal-bearing “B” unit, and a predominantly
sandy “C” unit, even though substantial lithologic variability occurs
laterally in each of these units. Vugrinovich (1984) conducted an
extensive subsurface investigation in a ten-county area of the central
Michigan basin, and proposed an informal revision of Wanless and
Shideler’s nomenclature. Recently, a regional hydrogeologic study
by Westjohn and Weaver (1998) concluded that the “assignment of
sandstones or other Pennsylvanian rocks to either the Saginaw Formation
or the Grand River Formation is difficult, if not impossible,
because there are no lithologic differences or stratigraphic horizons
that mark a change from one formation to the other.”
ORIGIN OF NAMES, TYPE SECTION, SIGNIFICANT
EARLIER STUDIES ON THIS INTERVAL
Lane (1902) named the Saginaw Formation for a series of Pennsylvanian-age
sandstones, shales, coals, and limestones observed
in drill holes for coal exploration in the central Michigan basin.
Kelly (1936) provided more detailed descriptions of the Saginaw
from very limited outcrop exposures along the Grand River in
Eaton County, from coal mine shafts and pits in the Saginaw Valley,
and from other areas of the central Michigan basin as well as.
Kelly interpreted the complex lateral and vertical facies changes
as indicators of small-scale spatial variations in local depositional
environments that ranged from fresh water to brackish and marine,
and concluded also that portions of the Saginaw were deposited as
cyclothems. Ells and others (1964) and Ells (1979) presented the
most accepted stratigraphic terminology and stratigraphic relationships
for Pennsylvanian units. Shideler (1969) and Wanless and
Shideler (1975) studied the stratigraphy, sedimentology, and paleo-