Unexpected Hydrologic Perturbation in an Abandoned Underground Coal Mine: Response to Surface Reclamation? DENVER HARPER Indiana Geological Survey Bloomington, Indiana 47405, U.S.A GREG A, OLYPHANT Department of Geology Indiana University Bloomington, Indiana 47405, U.S.A EDWIN J. HARTKE Indiana Geological Survey Bloomington, Indiana 47405, U.S.A. ABSTRACT/A reclamation project at the abandoned Blackhawk Mine site near Terre Haute, Indiana, lasted about four months and involved the burial of coarse mine refuse in
Hydrologic Effects of Mine Subsidence In the Illinois Basin the principal bedrock aquifers of Pennsylvanian age include coalbeds and sandstones. Sandstone units are commonly discontinuous and well cemented, so that their transmissivities (ranging from about 4 x 10 -5 to 3 x 10 -4 m2/sec; Smith 1983) and their well yields (0.05-1.9 1/sec but typically less than 0.7 1/sec; Wangsness and others 1981; Cocroft 1984) are relatively low. Furthermore, coalbeds and sandstones are typically closely interbedded with aquitards, such as shales, underclays, and limestones, which significantly reduce vertical leakage into aquifers. Consequendy, even some large active underground coal mines have relatively small inflows of water. A survey of pumping records for 15 mines in Illinois (ranging in depth from 40 to 250 m) determined that 11 mines pumped less than 100 m3/d, and the maximum pumping rate was 5,000 ma/d (Cartwright and Hunt 1978). Furthermore, parts of some mines remained dry for many years after abandonment (Cartwright and Hunt 1978). In some localities the fractures opened by the subsidence of bedrock above underground coal mines provide avenues for the vertical movement of groundwater. There are two general mechanisms by which mine subsidence affects the overlying strata: (1) a block of strata overlying an individual room or group of rooms within the mine may be displaced downward into the voids, perhaps crushing any remaining unEnviron Geol Water Sci Vol. 15, No. 3, 179-187
shallow (less than 9 m) pits excavated into loess and till in an area of about 16 ha. An abandoned flooded underground coal mine underlies the reclamation site at a depth of about 38 m; the total area underlain by the mine is about 10 km2. The potentiometric levels associated with the mine indicate a significant (2.7 m) and prolonged perturbation of the deeper confined groundwater system; 14 months after completing reclamation, the levels began to rise linearly (at an average rate of 0.85 cm/d) for 11 months, then fell exponentially for 25 months, and are now nearly stable. Prominent subsidence features exist near the reclamation site. Subsidence-related fractures were observed in cores from the site, and such fractures may have provided a connection between the shallower and deeper groundwater systems.
derlying pillars of coal, or (2) unconsolidated material may be washed downward into the mine through fractures opened by the collapse of the roof strata into the voids. The first mechanism can produce broad sags on the land surface, even when the collapse occurs at a depth of more than 100 m, whereas the second can produce either sags or deep sinkholes, although sinkholes seldom form where a mine is more than 45 m deep. Where subsidence is caused by the displacement of bedrock, fractures within the displaced block may be either abundant or rare, depending on local conditions, but a system of continuous fractures connected with mine voids is necessary for the formation of sinkholes. Studies have shown that the subsidence associated with active and abandoned underground mines may increase the permeability of the rock strata for more than 100 m above the mines (Cifelli and Rauch 1986; Hobba 1981). Where the mines are topographically above major surface drainageways, as is common in the Appalachian region, the strata may be completely or partly dewatered. The fractures opened by mining may increase the importance of flow through interconnected fractures relative to intergranular flow, and they may increase the vertical permeability by cutting through all stratigraphic units. The magnitude of these effects depends on the relative abundance in the lithologic column of hard rocks (such as sandstone and limestone) and soft rocks (such as shale and mudstone). © 1990 Springer-Verlag New York Inc.
180
D. Harper et al,
T h e recent investigations in Illinois of bedrock disturbances within eight subsided areas over abandoned coal mines reveal little disturbance of the bedrock (Bauer 1984). T h e fracture frequency increases as the amount and compressive strength of the limestone within the sequence increases, but little fracturing may be attributable to the subsidence within shale sequences. The fractures in the bedrock within the subsided areas were found in the roof as much as 15 m above the mines. T h e water loss occurring within this zone during coring or rotary drilling indicated that the fractures were connected to the underlying mine voids. Because there was concern at the Blackhawk Mine reclamation site, where subsidence sags are large and abundant, that subsidence-related fractures might permit the downward escape of contaminated groundwater, monitoring wells were installed in the unconsolidated materials, bedrock, and abandoned underground workings to acquire water samples for chemical analysis.
Reclamation and Well Emplacement At the Blackhawk reclamation site in southeastern Vigo County, Indiana (Fig. 1), a pile of coarse refuse from a coal-preparation facility was reclaimed by eraplacing the top of the pile in shallow pits that were dug in unconsolidated surficial materials consisting of about 6 - 1 2 m of soil, loess, and glacial till (Fig. 2). The pits were less than 9 m deep (typically about 4.5 m) and covered a total area less than 16 ha in extent. Water was pumped from the pits during construction, but ponding occurred at times, especially after storms. T h e refuse in the pits and the rest of the pile were then capped with soil. This construction phase of reclamation lasted four months, from July 16 to November 17, 1984. Soon after the reclamation was completed, the lower part of the buried deposit of refuse became saturated. The pump tests performed on the shallow observation wells finished in the undisturbed surficial materials (primarily till) around the periphery of the Blackhawk reclamation site indicated transmissivity values ranging from 4.3 x 10 -s to 5.2 x 10 -6 m2/sec. The water levels in those wells showed normal seasonal variations, so that they typically rose from November through March and fell from April through October. T h e total annual variations ranged from 1.3 to 3.2 m; for example, see well 3D in Figure 3. Within the reclamation site (near well 15M in Fig. 1), shallow wells were finished in buried refuse (well 10G) and in till that immediately underlies buried refuse (well 10D). Whenever the water levels in wells
10G and 10D were measured at the same time, the levels were equal, but as the record for well 10D was more complete, it has been plotted in Figure 3. T h e water level in well 10D seems to record seasonal variations superimposed on an overall rise in level (Fig. 3). However, in contrast with the water levels of other shallow wells, the water levels in well 10D showed their greatest rises from February to October and then fell slightly from October to February. In .January 1988 the monitoring of wells 10G and 10D was discontinued at the request of the landowner; at that time the total observed change in the water level of well 10D was 3.2 m. T h e Blackhawk Mine, which is an underground mine in the Springfield Coal Member of the Petersburg Formation (Pennsylvanian), and an adjacent mine, the Dixie Bee Mine, with which it is probably hydrologically connected, underlie an elongate area that has a northeast-southwest trend and that is about 1.6 km wide and 6.4 km long (Fig. 1). The reclamation site is in the northeastern part of this area, where the mine is at a depth of about 38 m, or about 145 m above mean sea level. T h e coalbed dips to the southwest, so that the extreme southwestern part of the mine is about 82 m deep (about 100 m above mean sea level). As part of the monitoring project, seven observation wells (designated wells 1M, 3M, 7M, 9M, llM, 13M, and 15M) were emplaced at depths ranging from 34 to 43 m into abandoned workings of the Blackhawk Mine (Fig. 1). All wells were within an area of about 0.5 kin2; the maximum distance between any two wells was about 0.8 km. Well 13M was completed on October 24, 1984, which was 24 days before the completion of the construction phase of reclamation. The other wells were completed during or after April 1985, more than four months after the reclamation was finished. The monitoring wells were cased from the surface to the top of a competent limestone that lies 7.5-15 m above the coalbed; the rest of each hole was left open. Packers (truncated rubber cones) were used to seal the wells at the bottoms of the casings (Fig. 2). Three observation wells (wells 1M, 3M, and 13M) were finished in collapsed mine voids, and four (wells 7M, 9M, llM, and 15M) were finished in pillars of coal within the mine. Pump tests were performed on wells 1M and 3M (in collapsed voids) and wells 7M and l l M (in coalbed pillars) (Table 1). That no decline in the water level was measurable in wells 1M and 3M after continuous pumping at 0.6 1/sec for 60 min indicates connections to the voids with little resistance to the flow. T h e Bucktown Coal Member of the Dugger For-
181
Hydrologic Perturbation in Abandoned Mine
EXPLANATION ]
SURFACEMINE
m
SURFACEMINE LAKE
172
MAN-MADE LAKE (NOT SURFACE-MINED) (~
!FRENCH LAKE
KNOWNMETHANEEMISSION BLACKHAWK MINE
180 ELEVATIONOF LAKE SURFACE
®1 1M
~
Vigo County
¢
Z
9M
0
RECLAMATION SITE
175
DIXIE SEE MINE
,
I-
SURFACEMINE IN SPRINGFIELD COAL
SURFACE MINE IN OVERLYINGDANVILLE COAL
1 KM
Section 19
I
R 9
W
R 8
~/:LACKHAW"K N\ FIELD \\
(
%
W
Figure 1. Map showing the locations of the reclamation site, observation wells, underground mines, surface mines, and methane emissions. The location of cross section A]-A4 is also shown. The elevations of the lake surfaces are in meters (above mean sea level). marion (Pennsylvanian), a thin (about 1 m) noncommercial coalbed, lies about 9 m above the Springfield coal (Fig. 2). The interval between the Bucktown and Springfield coals contains thin beds of limestone, gray shale, black shale, and sandstone. The Bucktown coal is overlain by 9 - 1 8 m of gray shale with a uniform appearance, which is the uppermost bedrock unit in the burial area (Fig. 2). Pump tests were performed on wells 1B, 3B, and 7B, which were a few feet from wells 1M, 3M, and 7M, and which were finished in the gray shale (Table 1). The flow through the Blackhawk Mine is probably trimodal. The flow to a pumping well would probably be most rapid in tortuous networks of major roadways that remain open (that is, where the roof strata have not collapsed), and it would probably be slowest through intact (uncrushed) pillars. Intermediate flow velocities might occur through networks of rooms and roadways where the roof strata have collapsed and where the flow must pass through rubble and through open channels across the tops of the rockfalls. The complexity of the flow within abandoned underground mines is discussed by Aldous and others (1986) and Aldous and Smart (1988), who used tracer studies to investigate the flow.
At the Blackhawk reclamation site, fluorescent dye (rhodamine WT) was injected into well 3M, which was finished in a collapsed mine void. Because the measured concentration of dye in the well remained close to the initial concentration for six months, very slow flow was indicated.
Changes of Water Levels T h e elevations of the water levels in the monitoring wells emplaced in the mine voids (wells 1M, 3M, and 13M) have consistently been almost equal; since July 1985 the maximum difference at any given time in the water levels in any two of those wells has been 15 cm. The water level through time in well 1M is plotted in Figure 3. The changes of the water levels within the wells emplaced in pillars (wells 7M, 9M, llM, and 15M) lagged behind the changes in the wells emplaced in voids by one to two months. The level of well 9M, which had the greatest lag, is also plotted on Figure 3. Because of this lag, the water levels in wells 1M and 9M gradually diverge during the rising phase (September 1985 to August 1986) and gradually converge during tile falling phase (after August 1986). From October 1984 to May 1985, the water levels in
0. Harper et al,
182
A~
A2A3
RECLAMATION 81TE
WELL
mine
3M
co~
\-
.
.
.
.
.
.
.
.
A i200
WELL 7M
FRENCH ! LAKE
-180
-:
"3 -160
~
E _o E o
Y ~0,"~ /
= o
s~r!~oe mine
Z
I-< •140 >~ o_
\pillar
,,
voids and fractures
potentiometric surface associated with Springfield coal
0
II KM
.J uJ
120
I
100
Figure 2. Cross section across the Blackhawk and Dixie Bee Mines showing stratigraphic relationships above the Springfield coal. The height of the mine voids and the extent of fracturing related to subsidence are speculative. See Figure 1 for the location.
135t
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
184 1
.
.
.
.
.
.
.
.
.
.
.
..............
183-
~182-
o>m 181~180
llllJll=l~Jl~al~lll~llll,ll~l~ll~llll~lll
v
ii
7/85 10/851/86 4•86 718610/86 11874/87718710/871/88 4/887188 10/881/89
,oo
I
wells 1M, 3M, and 13M rose in elevation from about 166.8 to about 167.2 m (above mean sea level). Then, from May to September 1985, the water levels fell to a minimum level of 166 m (Fig. 3). This modest rise and fall of the water levels may have been a natural response to seasonal variations in the water balance. However, in September 1985, the potentiometric level associated with the underground mine began to rise again, and the water levels continued rising for almost a year with no indication of recessions during dry seasons (Fig. 3). In August 1986 the water levels began to fall and have continued to fall for more than two years, so that the levels appear to be returning to elevations close to those measured at the beginning o f the monitoring project. The rise in the water levels between September 1985 and August 1986 was linear, and the average rate of the rise was 0.85 cm/d. In contrast, the recession o f the past two years has been exponential and can be described by a standard recession equation: h =
,
,
,
,
.................
_ ....
I 7/85 10/851/86 4/86 7/8810/861/874/87 7/87 10/871/88 4/887/88101881/89 Date (months)
Figure 3. Graphs showing the elevations of the water levels in wells 1M, 3D, 9M, and 10D from April 15, 1985, to June 23, 1988.
hoe
where ho is the initial potentiometric level (that is, the level on September 13, 1985), and h is the potentiometric level t days later. The estimated value of the recession constant (k = 0.998, Fig. 3) is consistent with a discharge medium of very low permeability, such as probably exists in the materials that bound the under-
183
Hydrologic Perturbation in Abandoned Mine
Table 1. Permeabilities of coalbed, mine voids, and gray roof shale
Well No. 1Ma 3Ma 7Mb 11Mb 16Mc 1Bb 3Bb 7Bb
Material
Thickness (rn)
void void combed combed combed gray shale gray shale gray shale
< 1.8 < 1.8 1.8 1.5 1.5 15.4 18.0 8.8
Hydraulic conductivity (rn/sec)
7.8 1.3 3.5 3.0 4.7 8.5
--x 10 -6 x 10 -5 X 10 -7 × 10-6 × 10 -7 x 10 -6
Specific capacity (m3/sec)
Storage coefficient (assumed)
1.6 x 10 -5 2.7 x 10 -5
1 x 10 -4
4.2 x 10 -5 1.0 x 10 -5 6.8 x 10 -5
1x
1 x 10 - a 10 -5 1 x 10 -4 1 x 10 -5
aNo drawdown was observedunder maximumpumping capacityof 6 x 10-4 mS/sec. bEstirnates are based on a procedure presented in Bradbnry and Rothschild (1985). CEstimateis based on a procedure presented in Thompson (1987). ground-mined area. Note the relatively low conductivity values associated with pillars in Table 1. T h e total range of the variation observed to date in well 1M is about 2.8 m (166.0-168.8 m). At present, the static water level associated with the mine voids is about 1 7 - 2 0 m above the actual voids, which are about 149 m above sea level.
Possible Causes of Water-Level Perturbation T h e perturbation of the potentiometric levels associated with the underground mine probably resulted from some change of the inflows or outflows to the mine workings from the recharge and discharge areas, although other factors could also be involved. In westcentral Indiana, which is on the east edge of the Illinois Basin, local groundwater flow in Pennsylvanian bedrock is generally toward major streams, and there is also a regional southwestward component of flow toward the center of the Illinois Basin (Wangsness and others 1981; Funkhouser 1983). However, because the Pennsylvanian section contains numerous thin (and commonly discontinuous) aquifers interbedded with aquitards and because most wells contain long uncased intervals that penetrate more than one aquifer, producing detailed and reliable potentiometric maps for well-defined aquifers is often difficult or impossible (Cocroft 1984; Funkhouser 1983). In areas of Indiana that are undisturbed by mining, the recharge of Pennsylvanian bedrock aquifers may occur primarily along their subcrops, where they are in contact with unconsolidated sediments. As noted by Funkhouser (1983), the areas disturbed by surface mining commonly possess excellent recharge characteristics because the disturbed material in such areas may be porous and permeable rubble. There are several possible recharge sources and
discharge areas for the water in the underground voids of the Blackhawk Mine. T h e recharge may be derived from (1) the lateral flow from abandoned surface mines in the Springfield coal, (2) the vertical flow from abandoned surface mines in the Danville Coal Member of the Dugger Formation (Pennsylvanian), (3) the vertical flow through unfractured bedrock, and (4) the vertical flow through abandoned shafts and subsidence features and along annuli of the monitoring wells. T h e discharge from the abandoned workings is probably predominantly downdip into the unmined coalbed. Furthermore, fluctuations in the water levels could be complicated by gas pockets in the fractured bedrock above the abandoned workings. French Lake and other lakes developed in the abandoned pits and spoil deposits of the surface mines in the Springfield coal lie less than 300 m east of the Blackhawk Mine at places (Fig. 1). T h e level of French Lake is 8 - 1 1 m higher than the potentiometric level associated with the abandoned underground mine, so the recharge may occur laterally from French Lake through the intervening unmined coalbed. The hydrologic connection between French Lake and the unmined Springfield coal was established when an observation well (well 16M) was finished in the coalbed near the west edge of French Lake (Fig. 1); the water levels indicate that the potentiometric level of the unmined coalbed is slightly below the surface of French Lake. However, the level of French Lake, which is controlled by precipitation and a surface outlet into a stream, normally varies by less than 0.6 m during a year. Therefore, although the equilibrium level of the water associated with the underground workings may be determined by the elevation of French Lake, fluctuations in the level of the lake cannot account for the perturbation of 2.7 m in the head associated with the underground mine.
184
D. Harper et al.
Figure 4. Photograph showing the fractures in a core from well 13M of a black shale bed that immediately overlies the Blackhawk Mine. Southwest of the reclamation site, the Blackhawk Mine is overlain by a surface mine (Fig. 1), which was developed in the Danville Coal Member, more than 30 m above the Springfield coal (Fig. 2). The large lakes that fill the abandoned pits in this surface mine overlie the Blackhawk Mine and may have the potential for recharging the underground mine. But a county landfill exists in the surface-mine deposits, which apparently have very low permeability, so that little or no downward percolation of the groundwater occurs (Howard B. Dillon, oral communication, 1987). The land surface near the Blackhawk reclamation site is characterized by large areas of sag-type subsidence features, where some subsidence troughs are as large as 110 m long and as much as 1 m deep. The subsidence is probably caused primarily by the displacement of large blocks of bedrock. T o the southwest, where the mine is 4 5 - 8 2 m deep, subsidence features are small and scattered. Fractures definitely related to the failure of the roof strata were observed in the cores of a black shale bed that immediately overlies the coalbed (Fig. 4). Other fractures, possibly related to the failure of the roof strata, were observed in the black shale as much as 5.8 m above the coalbed. Mining reports indicate that the grade work conducted within the Blackhawk Mine, while it was active, involved the removal of the roof strata to a height of 4.6 m in places; if additional caving occurred within these areas, the voids might reach even higher into the roof strata. During the drilling of one hole, circulation was lost at 4.2 m above the mine void. T h e abandoned vertical shaft of the underground mine is also within the reclamation site, although its exact location is unknown. T h e method used to seal the shaft is also unknown, but the mine was abandoned in 1952, when shaft-sealing methods were unregulated and typically consisted of dumping debris into the hole. Thus, the abandoned shaft might constitute a vertical hydrologic connection that could allow recharge from higher levels.
Well 11M is within the redamation area and penetrates refuse, but it is where the original refuse pile (deposited on loess and till) was simply graded and capped with soil rather than moved and emplaced in pits. T h e chemical analyses of the water from this well indicate chemical contamination by contact with refuse, because the analyses are similar to those of the water from wells finished in refuse (well 10G) and unconsolidated material (well 10D) within the redamation area (Table 2). These analyses contrast with those of the water from the wells finished in unconsolidated material (well 3D) and gray shale in the upper bedrock (well 3B) that are outside the reclamation area (Table 2). They also contrast with the analyses of the water from the wells finished in the voids and pillars of the Blackhawk Mine that are immediately updip and downdip of the reclamation area (wells 7M and 3M) (Table 2). T h e contamination of the water from well 11M may have entered the underground mine by the downward percolation from the buried refuse, perhaps through subsidence-related fractures, or by the downward leakage through the well annuli around the faulty packers. After February 1986, the acidity of the water obtained from well 11M diminished to about 50 percent of its initially high concentration, and an alkalinity component developed, although no other chemical components changed. On September 16, 1986, a borehole television camera was lowered into well 1 IM, but the camera struck rock a short distance below the casing. There appeared to be a void of indeterminate lateral extent below the bottom of the casing that indicated a collapse of the underlying bedrock, but the water-level data indicate that the hydraulic connection with the lower bedrock has continued to the present. Thus, subsidence fractures affording vertical hydrologic connections may have formed within the reclamation area in the months preceding February 1986. The perturbation of the water levels could also be related to changes that occurred in the discharge areas, but at its shallowest point, the Blackhawk Mine is about 30 m deep, so there is no possibility of a direct discharge of water from the underground voids into surface drainage (Fig. 2). The potentiometric level associated with the underground workings has ranged from about 165.8 to 168.9 m (above mean sea level), which is comparable to the elevation of the valley in which Paint Mill Lake lies (Fig. 1). The surface of this lake, which is a manmade impoundment unrelated to surface mining, is maintained at an elevation of about 169.2 m by an overflow structure at the north end of the lake. The extreme south end of Paint Mill Lake almost directly overlies abandoned workings in the
Hydrologic Perturbation in Abandoned Mine
185
Table 2. Chemistry of water from selected wells Well
Number of samples
Average pH
pH range
Average aciditya
Average alkalinitya
Average sulfateb
3B 3D 3M 7M 10D 10G 11M
21 28 21 29 27 13 18
7.6 6.9 6.9 8.2 4.2 3.2 4.9
7.0-8.2 6.3-7.8 6.2-7.5 7.5-9.2 3.6-4.7 2.6-4.0 3.5-6.2
0 0 0 0 10,406 7,085 6,515
523 275 1,489 550 8 2 145
63 158 3,709 122 14,227 12,080 15,594
aReported in milligramsof CaCO3 equivalentper liter. bReported in parts per million.
Blackhawk Mine, but because the lake level is relatively constant, Paint Mill Lake is an unlikely source of the recharge event that perturbed the potentiometric level of the mine. The water levels associated with the abandoned underground workings could also be affected by pockets of gas. Where parts of abandoned mines remain dry, they may contain mixtures of nitrogen, carbon dioxide, oxygen, and methane. Methane from the coalbed or from a nearby oilfield (Blackhawk Field, Fig. 1) may have some effect on the confined groundwater system contained in the voids of the Blackhawk Mine. Anecdotal reports of gas include emissions from water wells, a gas-and-water outburst in an open field, and a gas-and-water outburst while a water well was being drilled (Fig. 1). It is not possible to determine the depth or geologic strata from which gas may have issued. Nevertheless, these reports indicate the possibility that methane gas, originating from either the Springfield coal or deeper strata, may become trapped within the roof strata in the voids created by the roof collapse above the abandoned Blackhawk Mine. If these gas pockets exist, they could affect the water levels within the mine workings because the accumulation of gas under pressure could displace the water. The release of the gas, either by naturally occurring outbursts or by drillers who accidentally penetrate the gas-filled voids, might result in a sudden fall of the water levels. However, if the growth of a gas bubble had caused the rise in the water levels observed at the Blackhawk reclamation site, then the rates of rise should have gradually decreased through time because of the compressibility of gas, whereas the observed rates of rise were constant between September 1985 and August 1986 (Fig. 3). The association of the perturbation in water levels with reclamation activity indicates another possibility: that water in highly permeable refuse emplaced in pits steadily percolated downward by gravity drainage.
The fact that the observed rate of the rise of water levels (0.85 cm/d) is consistent with the saturated conductivity of moderately fractured bedrock (about 10 -5 cm/sec) indicates that the pure gravity drainage of ponded water is a reasonable hypothesis.
Discussion The interconnected voids of the Blackhawk Mine constitute a confined aquifer with limited areal extent (10 km 2) but with very high permeability and approximately 7.6 × 109 1 of water. Because a conduit-type flow exists within the mine, the boundary of the mine approximates an isopotential line. Except where the roof strata are fractured, the permeabilities of the confining units are low. This large body of freeflowing water and of subsidence-related fracturing may, in places, predominate over the effects of heterogeneous stratigraphy. Knowing that the regional hydrologic gradient is to the southwest and assuming that the recharge to the underground mine is normally dominated by the flow from surface-mined areas to the east, we could draw a hypothetical flow net (Fig. 5). We believe that a temporary removal on the surface of relatively impermeable soil, loess, and till in a small area (16 hectares) resulted in a large (2.7 m) and prolonged (almost two years) increase in potentiometric levels above t h e equilibrium level associated with the underground mine. The isopotential lines near the underground mine were probably shifted southwest-. ward (Fig. 5), but data for evaluating details of the shift are inadequate. The removal during reclamation of relatively impermeable surficial cover may have permitted the introduction of a downward-moving slug of water into the bedrock. Such a slug may have been derived from direct precipitation and dewatering of areas surrounding newly excavated pits (during active reclama-
186
D. Harper et al.
EQUILIBRIUMCONDITION~ / / ~
\.~
'
AUGUST,1986
\
%
f
,.
I
X
\
I~ '
'
Figure 5, Maps showing hypothetical flow nets in the Springfield Coal Member near the Blackhawk Mine under equilibrium conditions (left) and under perturbed conditions caused by reclamation (right). The contour values designate potentiometric levels (meters above mean sea level). The contour interval of the isopotential lines is 5 m. tion from July 16 to November 17, 1984), or it may have subsequently been derived from the buried deposits of saturated refuse. Ten months after the completion of redamation (in September 1985), the arrival of the leading edge of the slug at the fractured bedrock overlying the mine workings initiated a linear rise in the potentiometric levels. The termination of the slug in August 1986 must have resulted from the loss of vertical connection (as by the physical sealing of subsidence fractures), because the water levels in the buried refuse (as recorded by well 10D) continued to rise until September 1987 (Fig. 3). Because of the confined nature of the water in the abandoned mine, the movement of a relatively small volume of water under sufficient head could probably have produced the observed changes in the water levels. After the entire slug had arrived in August 1986, the lateral discharge of water into the downdip part of the unmined coalbed caused the observed exponential decline of the water levels. The effects of the perturbation in pressure head were probably transmitted almost instantaneously by the mine voids across an area of 10 km 2. In a slightly different geologic or topographic setting (higher refief, laterally continuous permeable strata, or deeply incised bedrock surface), this perturbation might have resulted in the discharge of contaminated water to the surface through seeps or to overlying groundwater systems at a considerable distance from the site. In this particular setting, however, the constant rate of rise of the water levels between September 1985 and August 1986 (Fig. 3) indicates that no significant off-site discharge occurred. The monitoring program at the Blackhawk reclamation site was primarily focused on the postreclama-
tion geochemical sampling of the water in the unconsolidated surficial materials. As such, the monitoring program was inadequately designed to fully explain the large and prolonged perturbation of the water levels associated with the abandoned underground Blackhawk Mine. This large perturbation (whose mathematical form is strikingly simple) was entirely unexpected and indicates that subsidence may have significantly affected the vertical conductivity of the confining layers, even though shales dominate the stratigraphy. In the eastern part of the Illinois Basin, virtually nothing is known about the formation, extension, and persistence of the vertical fractures above numerous abandoned underground coal mines. Our observations at the Blackhawk reclamation site indicate that these fractures may have important transient environmental implications for reclamation and other construction activity and underscore the need for carefully designed field experiments to investigate the hydrologic role of subsidence fractures.
Acknowledgments We wish to acknowledge the contributions of Christopher R. Smith and Joseph G. Hailer, formerly of the Indiana Geological Survey, who helped to design and conduct the monitoring project. We wish to thank William J. Steen and John Clark, Indiana Division of Water, for their comments and suggestions regarding this report.
References Cited Aldous, P.J., and P. L. Smart, 1988, Tracing ground-water movement in abandoned coal-mined aquifers using fluorescent dyes: Ground Water, v. 26, no. 2, p. 172-178.
Hydrologic Perturbation in Abandoned Mine
Aldous, P.J., P.L. Smart, and J.A. Black, 1986, Groundwater management problems in abandoned coal-mined aquifers: a case study of the Forest of Dean, England: Quarterly Journal of Engineering Geology, London, v. 19, p. 375-388. Bauer, R. A., 1984, Subsidence of bedrock above abandoned coal mines in Illinois produces few fractures: Society of Mining Engineers of AIME Preprint 84-400, 8 p. Bradbury, K. R., and E. R. Rothschild, 1985, A computerized technique for estimating the hydraulic conductivity of aquifers from specific capacity data: Ground Water, v. 23, no. 2, p. 240-246. Cartwright, K., and C. S. Hunt, 1978, Hydrogeology of underground coal mines in Illinois: Proceedings of the International Symposium on Water in Mining and Underground Works, September 17-22, Granada, Spain. Cifelli, R. C., and H. W. Rauch, 1986, Preliminary results of research on aquifer dewatering effects by underground coal mining in north-central West Virginia: Abstracts for 2nd Workshop on Surface Subsidence due to Underground Mining, June 9-11, Morgantown, West Virginia, sponsored by U.S. Office of Surface Mining Reclamation and Enforcement and West Virginia University, p. 92-93.
187
Cocroft, J. E., 1984, Hydrogeochemical study of a mined region in the Carbondale Group (Pennsylvanian), southwestern Indiana: MA thesis, Bloomington, Indiana University, 198 p. Funkhouser, R.V., 1983, Hydrogeological study of the French Lake area, Vigo County, Indiana: MA thesis, Bloomington, Indiana University, 171 p. Hobba, W. A., Jr., 1981, Effects of underground mining and mine collapse on the hydrology of selected basins in West Virginia: West Virginia Geological and Economic Survey Report of Investigation RI-33, 77 p. Smith, C. R., 1983, Hydrology of the Carbondale and Raccoon Creek Groups, Pennsylvanian System, Vigo, Clay, and Sullivan Counties, Indiana" MA thesis, Bloomington, Indiana University, 93 p. Thompson, D.B., 1987, A microcomputer program for interpreting time-lag permeability tests: Ground Water, v. 25, no. 2, p. 212-218. Wangsness, D.J., R. L. Miller, Z. C. Bailey, and C. G. Crawford, 1981, Hydrology of area 32, eastern region, interior coal province, Indiana: U.S. Geological Survey Water Resources Investigation Open-File Report 81-498, 76 p.