PERMFILE57051

• Permit No: M1999002
• IBM Index Class Name: Permit File
• Doc Date: 6/18/1999
• Doc Name: OBJECTORS EXHIBITS
• Media Type: D
• Archive: No

306
edge of the azea in ++'hich settlement observations
were made. It till be noted that the pt'ofiles For
the years 1950 through 1953 exhibit a form similar
to that of a loaded, clamped plate of which the
clamped edge is subsiding at a rate of about one-
halE inch per year. The postulated plate consisted
of rock strata with an initial thickness of at least
900 Eeet. The initial do Formation of such a plate
would be very small but a progressive decrease of
its thickness associated ++ith the dropping of strata
or packets of strata into the underlying caciq~
++•ould lead to increasing deformation oC the plate
and hence increasing settlement of points on the
surface of [hc ~*tound. On the hypothesis that sub-
sidence was the surface expression of the deforma-
tion of a roughly circular clamped plate, the
episode of rapid and catastrophic subsidence of
February, 1954. can br interpreted as the result of
the failure of the remaining strata tchich could no
longer carry their o+vn weight plus that of the over-
lying unconsolidated deposit.
It is not possible ro determne xccwately the
diameter of the area undrrgoin~ accelerated subsid-
ence, corresponding to the diameter of the postu-
lated clamped plate, because settlement
observations were not made throughout the entire
bowl of subsidence. Hotvevcr, available data pro-
vide abasis Eor an estintatc of maximum and mini-
mum values. As indicated by Figure 5, the clamped
edge of the postulated plate was descending at an
approximately constant rate, whereas the remain-
der of the plate was undergoing deformation at a
generally increasing rate. Hence on the map shown
in Figure 6 (reproduced from Peck's 1954 report),
the location of the clamped edge can be assumed
to coincide approximately with the contour repre-
senting zero difference bctw'cen the 1949 rate of
subsidence and the 1953 rate. The distance from
this contour to the point of maximum subsidence
Gcfore sinkhole formation is approximately the
radius oC the clamped plate subject to deformation.
IF this point +. as located near the center of the
sinkhole area, the radius .vas about 750 feet. How-
ever, the center of subsidence preceding Cailure was
not necessarily identical ++ith the center of the
sinkhole, because the location of the sinkhole was
probably' determined by the location of that part
of the cavity with maximum original vertical di-
mension, which +cas not necessarily at the center of
the subsiding plate. As a matter of lac[, settlement
obse•n•ations made in 1950 suggest that maximum
srulcmcnt in 1970 UCCUffc(I In a^ arCa SCPI'faI ItLtll-
dred feet to the southwest of the center uC the
sinkhole (Fig. 2). If this area was the site of maxi-

~rinelield Subsidence at Windsor, Ontario
mum subsidence throughout the years 1951
through 1953, the effective diameter of the
clamped plate may not have exceeded 650 feet.
Although it is impossible to determine with pre-
cision the time at which the thickness of the cavity
roof began to decrease, the records of well aban-
donment provide some general information. OF the
nine wells drilled within the area of the bo+vl of
subsidence or at its periphery, all but two had been
abandoned by 1932 o+ving to severe damage pro-
duced by rockfalls. Two suniced into the 1940'x.
One of these, located about 500 feet to the south-
east++'ard of the center of the sinkhole, w'as aban-
doned in 19-}0 owing to a rockfall at a depth of
965 fee[. The second, located about 100 feet east
oC the first, +cas abandoned in 19-}3 because of
incparable damage at a depth of 1200 feet. Hence
i[ seems that the su~ata at depths less than 1000 to
1200 feet +vere still fairly intact in the earh• 1940's
beneath the site of the future bo+vl of subsidence;
othencise rockfalls ++'ould ha+'e taken place in the
nco remaining wells as higher elccations. By 1950,
+chcn the bo+cl of subsidence first became per-
ceptible, progressive failure must ha+e reached a
level +, ell above the top of the Opper Salt.
Subsidence like that at Sandwich is commonly
attributed to sloping above a large cavity' located at
depth, followed by the collapse of the remainder
oC the roof of the cavity. Ho+vever, if it +cere
assumed that the Sandt+'ich subsidence +vas due to
the presence of a single high cavity, into t+'hich
much of the roof rock, with a total thickness in
excess oC 1000 feet, had dropped, it would be vir-
tually impossible to account Eor the Eact that bulk-
ing of [he debris had not caused the cavity to be
filled before major subsidence could take place.
An alternative hypothesis is that early ca+'ities
were extensive and locally deep, but contained
abundant salt pillars which provided temporary
and partial support for a sagging roof. According to
this hypothesis, if strata ovcrly'ing a cavil}' became
separated along bedding planes from the rocks
above and sagged into the atcin', they did not de-
scend far before coming to rest at least temporarily
on salt remnants. ',Then the next set oC strata in
turn became detached Erom the ocerly'ing rock,
they also came to rest in a nearly' intact state and
onh' slightly below their original positiun. These
strua must hau• bic•n thoroughly' fracuued and en-
tirely' lacking in lcnsile strength, but the space
+chich they occupied need not h:nc been greatly in
excess of their original culume. It cm, fur instance,
be pustulated that the 150- to 200-tout thick
stratum oC insoluble rock bcl+cecn the Lower and