See discussions, stats, and author profiles for this publication at: [Link]

net/publication/237172333

Howe Lake: A hydrodynamic blowout structure

Article in Canadian Journal of Earth Sciences · February 2011

DOI: 10.1139/e82-097

CITATIONS READS

42 942

3 authors, including:

D. J. Gendzwill
University of Saskatchewan
39 PUBLICATIONS 677 CITATIONS

SEE PROFILE

All content following this page was uploaded by D. J. Gendzwill on 17 June 2014.

The user has requested enhancement of the downloaded file.

 Howe lake: a hydrodynamic blowout structure

E. A. CHRISTIANSEN'
Saskatchewan Research Council, 30 Campus Dr., Saskatoon, Sask., Canada S7N OX1
D. J . GENDZWILL
Department of Geological Sciences, University of Saskatchewan, Saskatoon, Sask., Canada S7N OW0
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14

AND
W. A. MENELEY~
Saskatchewan Research Council, 30 Campus Dr., Saskatoon, Sask., Canada S7N OX1
Received August 20, 1981
Revision accepted December 23, 1981

Howe Lake (formerly called "Goudie Lake") is an unusual feature of southeast Saskatchewan. Almost perfectly circular
(diameter 295 n ~ )it, is much deeper (26 m) than almost any other prairie lake. Initially noted in 1952, numerous investigations
culminated in deep drilling and geophysical programs in the 1969-1976 period. The lake is the surface expression of a deep,
funnel-shaped depression filled with coarse colluvium at the bottom and finer material near the top. A central pipe extends to a
depth of at least 138 m. Radiocarbon dates and the history of deglaciation give evidence that the structure was formed between
12 000 and 12 500 years BP, shortly after the last ice sheet melted in the area. The structure was probably formed by a
hydrodynamic blowout of water from the Mannville Group 400 m below surface. Extreme overpressure of water in the Mannville
Group was probably induced by recharge from meltwater of the continental glacier then standing a few kilometres to the north of
Howe Lake. The blowout was initiated at Howe Lake by fractures in the subsurface, related to a salt solution-collapse structure.
For personal use only.

Howe Lake probably served as a valve controlling the artesian pressure in the Mannville Group over a large area, until the active
ice margin had retreated northeastward exposing
. -
lower elevation discharge areas. Hydrodynamic blowouts provide another
process for the origin of breccia-filled pipes.

Le lac Howe (appelk autrefois "lac Goudie") prksente une forme inhabituelle pour le sud-est de la Saskatchewan. Etant presque
parfaitement circulaire (diamktre 295 m), il est beaucoup plus profond (26 m) que la grande majoritk des lacs des prairies. C'est
en 1952 qu'il fut remarqut et depuis de nombreuses Ctudes se sont accumulkes suite aux programmes de forages profonds et de
gkphysique qui furent rCalisCs entre 1969 et 1976. Le lac traduit en surface une dkpression profonde en forme d'entonnoir
remplie de dCp8ts colluviaux a grain grossier au fond et de matkriaux plus fins prks de la surface. Une pipe centrale plonge jusqu'a
une profondeur d'au moins 138 m. Des datations au radiocarbone et l'histoire de la dkglaciation indiquent que la structure s'est
formke entre 12 000 et 12 500 annkes avant le prksent, peu de temps aprks la fonte de la calotte glaciaire. La structure rtsulte
probablement d'un claquage hydraulique dB ?Ijaillissemeit violent ei soudain de l'eau du groupe ~ a n n v i l l gisant
e h 400 m sous la
surface. Une surcharge excessive de la ~ressionde l'eau du erouve Mannville fut induite probablement par une recharge h meme
les eaux de fonte du ilacier continentaiqui stagnait 5 quelGes cilomktres au nord du lac Howe. Le claquage fut amorck au lac
Howe par les fractures h la subsurface relites ?I une structure d'effondrement provoqute par la dissolution d'une assise de sel. Le
lac Howe a probablement servi de soupape contrdlant la pression artksienne dans le groupe Mannville sur une grande surface de la
rkaion, ius~u'autemps de recul du front actif du glacier vers le nord-est mettant hjour les zones de dkcharge de faible altitude. Les
claquages hydrauliqies constituent un autre prckessus pour expliquer lloriginkdes pipes rernplies de brkches.
[Traduit par le journal]
Can.J. Earth Sci., 19, 1122-1 139 (1982)

Introduction controller of surveys, Saskatchewan Department of

Howe Lake (Figs. 1, 2, 3) is anomalously circular Natural Regina.
(diameter 295 m) and deep (26 m), unlike most prairie AS a result of this report, A. J. Budding, geologist,
lakes. Saskatchewan Department of Mineral Resources con-
The anomalous nature of H~~~ ~ a was k ~first ducted a field investigation of the Howe Lake area on
recognized on aerial photographs in 1952 by M ~ . October 5 and 6, 1953. A mine detector survey and the
Gandre, an oil company exploration geologist who field investigation failed to uncover any meteoritic
brought the lake to the attention of A. 1. ~ ~ ~then~ material.
~ k The i meteoritic
~ , impact explanation for Howe
Lake was considered unlikely by Budding, who in-
'present address: E. A. christiansen consulting ~ ~ dB~~
. , formed Bereskin that the feature seemed to be of glacial
3087, Saskatoon, Sask., Canada S7K 3S9. origin.
'Present address: W. A. Meneley Consultants Ltd., Box Howe Lake was brought to the attention of the senior
7167, Saskatoon, Sask., Canada S7K 451. author in 1960 by F. M. Atton, Saskatchewan Depart-
OOO8-4077/82/061122-18$01 .OO/O
01982 National Research Council of Canada/Conseil national de recherches du Canada
 CHRISTIANSEN ET AL. 11:

LAC L A RONGE

€7
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14

SASKATCHEWAN

h 1,
For personal use only.

' ...-,..-., .-...-... - - . . "102'

0 100 200
km
FIG. 1 . hcation of the Howe Lake structure.
ment of Natural Resources, who sounded the lake
bottom that year. His soundings showed the lake was
26m d e e ~with a central mound 6 m high that he
attributed'to the Coriolis effect (F. M. ~tto;, personal b
communication, 1960).
This paper is an account of investigations of the -- - - - = - -- - .+7 - -.

nature, age, and origin of this depression.

I History of data collection
Geologic information
In February, 1962 an attempt was made to drill into
the lake bottom from the ice surface using cable-tool
equipment. After the "drilling string" was lowered
through 18 m of water to the lake bottom, the drilling
tools sank an additional 20 m into soft gyttja. Extremely
cold and stormy weather and a limited budget- prevented -

further work that year.

Two test holes (Fig. 3, test holes A, D) were drilled
into the surrounding upland in 1969 to establish the
stratigraphy under the upland prior to making the second
attempt at drilling into the lake bottom. In February,
1970 Shelby and excavator samples were taken to a FIG. 2. Vertical ( a ) , oblique (b), and ground ( c ) photo-
depth of 138 m below the ice surface (Fig. 3, test hole C) graphs of Howe Lake taken in 1958, December 2, 1970, and
using cable-tool equipment, but the base of the collu- 1970.
vium in the Howe Lake structure still had not been
penetrated.
 1124 CAN. J. EARTH SCI. VOL. 19. 1982
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14
For personal use only.

mE Rotary test hole ~'"O-Contour on lake bottom,

interval 10 feet
a 3 Rotary probe hole
aC Cable tool test hole -1e3o- Contour on present day surface,
interval 5 feet
A-D Cross section

FIG.3. Morphometric map of bottom of Howe Lake. Lake bottom soundings provided by F. M. Atton.
In February, 1972, a third attempt was made to drill of the drilling fluid to the ice surface. The suction hose
the lake fill and to penetrate at least one of the marker was placed in the lake, and the cutting samples were
beds established under the upland, using rotary drilling permitted to spill out on the lake bottom. When the
equipment. This attempt at reaching the base of the lake drilling bit encountered the harder sediments below the
fill also failed because the casing column buckled in the gyttja and colluvium, a bit sample was taken by pressing
60 m of water and soft sediment. After abandoning site the bit into these harder deposits. After 12 probe holes
C (Fig. 3), the drilling equipment was moved to site B, had been drilled, two additional test holes were drilled
and a bedrock marker bed was penetrated. During the into the adjacent upland (Fig. 3, test holes E, F).
drilling of test hole B, a distinct drilling contrast was
noted between the lake fill and the underlying till and Geophysical information
bedrock. This contrast in drilling rate was the basis for After the 1969 and 1970 test drilling programs, a
subsequently conducting the probe hole drilling pro- seismic survey similar to the one conducted at Crater
gram. Lake (Gendzwill and Hajnal 1971) was carried out. The
Probe holes were drilled without the return circulation shooting plan is shown in Fig. 4a. Broadside profiles,
 CHRISTIANSEN ET AL. 11:

'
with shots on one of the lines and geophones across the toon Group, Fig. 8). These stratigraphic units we1
lake on the other line, resulted in two profiles with defined by Christiansen (1968a,b).

1I reflection points halfway between the lines under the

middle of the lake. Normal vertical profiles were also
shot north and south of the lake. A digital seismic
The Sutherland Group is composed of 8-21 m of
interbedded till and silt. The uppermost part of the unit is
olive gray, with yellowish brown staining along joint
recording system and fourfold common depth point surfaces.
redundancy were used. The Floral Formation is composed of 15-25 m of till,
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14

These data were processed in 1976 using a ray-tracing sand, and gravel. The uppermost part of the till of the
migration program and a band-pass filter of 50-96 Hz to Floral Formation is hard, oxidized, pale yellow, and
bring out more detail (Figs. 4 b-d). jointed with yellowish brown staining on joint surfaces,
In February, 1973 a gravity survey was carried out on whereas the till in the lower part of the unit is gray. The
the ice of Howe Lake and the surrounding area (Fig. 5). contact between the Floral Formation and the underlying
A magnetometer survey was also conducted but a Sutherland Group is distinct.
magnetic map is not included with this report because no The Battleford Formation is the youngest till in the
1 variations greater than 10 y (10 nT) were observed. area and is composed of 3-4m of soft, bouldery,
unstained, and unjointed till, which is readily distin-
Stratigraphy guished from the underlying Floral Formation.
Introduction
The stratigraphy of the Howe Lake area is shown in Holocene sediments
Because the Howe Lake structure postdates the
Figs. 6-8. The upper part of the Howe Lake structure is
developed in the Riding Mountain Formation, in the Pleistocene deposits (Fig. 8), the colluvium and gyttja
Tertiary-Quaternary sediments, and in the Pleistocene have been assigned a Holocene age.
deposits (Fig. 8). The depression of the structure was The colluvial deposits extend in depth from 45 to
>138 m (Fig. 8), hence are more than 93 m thick. Near
For personal use only.

'
filled with Holocene colluvium and gyttja.
the bottom of test hole C (Figs. 3, 8), the colluvium
I Riding Mountain Formation consists of blocks (3cm diameter) of material that
The Riding Mountain Formation is about 160 m thick originated from strata much closer to the surface. These
in the Howe Lake area (Fig. 6) and is composed of blocks commonly have a poorly sorted sand and gravel
noncalcareous, gray silt and clay of Late Cretaceous matrix.
age. Several examples of the displaced blocks are shown in
Tertiary-Quaternary sediments Fig. 9. Figure 9a shows oxidized till from the upper part
Outliers of Tertiary-Quaternary sediments, inform- of the Floral Formation, originally only 3-7 m below the
ally named "Wynyard Group" by Cheny and Whitaker surface, now at a depth of 86 m in test hole C. Another
(1969), occur throughout east-central Saskatchewan, piece was recovered from a depth of 128 m. Figure 9b
and a large outlier of these sediments occurs in the Howe shows a distinctive Tertiary-Quaternary grayish brown
( Lake area (Fig.7). The Tertiary-Quaternary sediments clay, normally occumng 90 m below the surface, from a
in the Howe Lake area (Fig. 8), in ascending order are: depth of 108 m in test hole C. Figure 9c shows a mklange
gravel, sand, clay, and silt. of material in a single chunk recovered from a depth of
Well rounded, basal chert and quartzite gravel, 2-5 m 128 m. It includes till derived from a depth of less than
thick, lies disconformably on the Riding Mountain 30 m, and Tertiary-Quaternary grayish brown clay
Formation. The sand overlying the basal gravel is fine to derived from 90 m depth, in a sand and gravel matrix.
medium grained, composed mainly of quartz and These samples show that material has fallen as much as
dark-colored minerals giving it a "salt and pepper" 120m and that materials from different strata have been
appearance. The sand unit ranges in thickness from 6 to thoroughly mixed.
9 m. The coarse, blocky material near the bottom of test
Overlying this "salt and pepper" like sand is 5-8 m of hole C is replaced by smaller blocks with a finer sand and
calcareous, grayish brown clay, which becomes silty gravel matrix at shallower depths. In the upper part of
toward the top of the unit. the colluvium, the blocks do not exist, and the material
The uppermost silt layer of the Tertiary-Quaternary is principally sand and silt.
sediments is about 50m thick, and is composed of silt Gyttja occurs between 16 and 45 m below the ice
that becomes clayey toward the base and sandy toward surface (Fig. 8). The lower 13 m of "Firm Gyttja" is
the top of the unit. sufficiently cohesive to be sampled with a Shelby tube
and is olive-gray. The upper 16 m of "Soft Gyttja" is
Pleistocene deposits black organic fluid, which was penetrated by the weight
The Pleistocene deposits include the Sutherland of the drilling tools. The top of the soft gyttja was taken
Group and the Floral and Battleford Formation (Saska- from Fig. 3.
 1126 CAN. J. EARTH SCI. VOL. 19, 1982

Section Sectkn
22 23

52

0 P
4,"

NORTH LlNE
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14

:4 = 7 2'
SOUTH LlNE

(a
( b ) SOUTH LlNE
SHOT POINTS
For personal use only.

FIG.4. Seismic sections. C = Cretaceous; M = Manville; D = Devonian; P = Prairie Evaporite. (a) Shot point location map.
(b) Vertical siesmic reflection profile for south line. (c) Broadside seismic reflection profile for middle line with shot points on
south line and receivers on north line. (d) Vertical seismic reflection profile for north line.

Howe Lake structure origin, one might expect a large bowl-shaped volume of
Description brecciated rock to exist under the lake, as described by
The Howe Lake structure is funnel shaped, tapering in Innes (1961). The gravity results show that no such
diameter from about 295 m at the top to less than 50 m at low-density breccia exists. The method of Talwani and
the base (Figs. 8, 10). The volume of the funnel (Fig. Ewing (1960) was used for computations, with densities
10) is about 2 X lo6 m3. The funnel surface dips radially estimated from experience. No in situ density measure-
toward the center at an angle of about 25" near the shore, ments were possible.
steepening to nearly vertical in the center. The water and The structure below the depth of drilling is revealed
unconsolidated material filling the funnel help to support by seismic data described above (Fig. 4) and by others
the steep slopes from collapsing. At the center of the (Fig. 11) released by the Saskatchewan Department of
funnel, the structure develops into a narrow breccia- Mineral Resources (Sawatzky 1967).
filled pipe, whose diameter is less than 50 m (Figs. 8, Reflections from the Devonian strata reveal a com-
10). The pipe was drilled to a depth of 138 m, but plex structure on the north, middle, and south lines. Line
undoubtedly extends to a greater depth. A (Fig. 11) about 2km south of Howe Lake clearly
Interpretation of the gravity data is in agreement with shows it to be a synform structure with characteristic
the near-surface geological model (Fig. 5 ) , showing that features of Prairie Evaporite salt removal. About 120 m
all the mass is accounted for. If the lake were of meteoric of salt has been removed. The synform trends north-
 CHRISTIANSEN ET AL.

I
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14

( C ) MIDDLE LINE (S-N)

( d ) NORTH LINE
SHOT POINTS
For personal use only.

I south, with a change in strike attitude near the lake, but

FIG.4. (Concluded)
salt-karst sinkhole. A small fault or faults may occur
the lake is near the edge of the synform, not the center. under the lake.
The three lines of Fig. 4 show the Cretaceous strata to
be dipping west or northwest. The west end of the north Age
line is about 60m lower than the east end of the south The erosional surface of the Howe Lake structure
line, owing to the removal of Devonian salt. truncates all till formations in the area, including the
Removal of salt occurred in two episodes: near the Battleford Formation (Fig. 8), and pieces of the tills
end of the Devonian Period, and in Late Cretaceous were recognized in the colluvium filling the funnel (Fig.
time. Minor irregularities, probably small faults (Fig. 9). The Howe Lake structure, therefore, is younger than
4c), occur under Howe Lake, but there is no evidence the youngest till, the Battleford Formation. According
on any line of block displacements of Cretaceous rocks to Christiansen (1979), the site of the Howe Lake
commensurate with the dimensions of the lake as were structure was deglaciated about 12 500 years ago (Fig.
found at Crater Lake (Gendzwill and Hajnal 1971). 12), exposing the Battleford Formation. Consequently,
1 From the seismic data we conclude that Howe Lake is
located on the flank of a salt-removal synform, but that
the Howe Lake structure must be younger than 12 500
years.
,
I
the lack of a significant shallow structure under the lake
indicates that the lake itself is not the expression of a
Seven radiocarbon dates were obtained from the fill
in the Howe Lake structure (Table 1). Radiocarbon
 1128 CAN. J. EARTH SCI. VOL. 19. 1982

samples 529-53 1were derived from gyttja, 532 and 533

from wood, and 534-536 from carbonaceous silt within
the colluvial part of the fill. Radiocarbon dates 534-536
were rejected because they are older than the Battleford
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14
For personal use only.

FIG. 5. (a) Gravity contour map over Howe Lake and

vicinity. Contour interval is 0.2 mGal(0.2 X lo-' m/s2). (b)
Calculated gravity profiles for various units. D = gravity effect
of colluvium; C = gravity effect of firm gyttja + D; B =
gravity effect of soft gyttja + C; and A = gravity effect of
water + B. Profile A matches the observed gravity very
closely. (c) Geological model. D = colluvium; C = firm
gyttja; B = soft gyttja; and A = water.
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14
For personal use only.

0 Electric logs only

8 Electric logs,cutting samples
@ Electric logs,cuttinq samples,and cores

km
Contour interval on bedrock surfocr. 100 t e e t

FIG. 7 . Bedrock geology and topography of the Howe Lake area. Modified from Cheny and Whitaker (1969).

Formation (20000 years BP; Christiansen 1968b), Origin

which is older than the fill in the Howe Lake structure. A number of plausible origins of the Howe Lake
Presumably the carbonaceous silts were contaminated structure will be discussed below. They include meteor-
with "old" carbon. itic impact, collapse, and hydrodynamic blowout.
The time-depth curve in the gyttja is shown in Fig.
13. If the curve is projected to the base of the gyttja, then Meteoritic impact
it can be inferred that the colluvial sedimentation ceased Meteoritic impact for the origin of Howe Lake was
and the deposition of gyttja commenced about 12 000 dismissed for the following reasons: (1) there is no
years ago. Because the Howe Lake structure is younger raised rim around Howe Lake; (2) the diameterldepth
than the Battleford Formation, which was deposited in ratio does not lie on the Baldwin curve for meteoritic
the Howe Lake area from about 20 000-12 500 years craters (Baldwin 1949; Beds 1958); (3) meteorites
BP, and older than the gyttja that is presently filling it, create a bowl-shaped volume of breccia to a depth of
we conclude that the Howe Lake structure formed about one third the diameter of the crater (Innes 1961),
between 12000 and 12 500 years ago, shortly after the whereas the colluvium in Howe Lake is funnel shaped
ice retreated from the area. and greater in depth; (4) the gravity results show no
 1130 CAN. J. EARTH SCI. VOL. 19, 1982

TABLE1. Radiocarbon dates from Howe Lake fill

13A31 13s 3A08V 1333

Depth below ice s

0 0 0
8
0
8
0
2'
0
8
Years BP Sample No. Material surface (m)

4180k 130 S-529 Gyttja 33.5

5465k 135 S-530 Gyttja 38.4
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14

8405k 185 S-531 Gyttja 43.0

S-532 Wood 47.3
lo2lok 225 S-533 Wood 47.9
31 710k1420 S-534 Carb. silt 76.2- 77.7
32 830k 1630 S-535 Carb. silt 89.0- 90.5
36 940k 3680 S-536 Carb. silt 111.3-112.8

evidence of fractured ground adjacent to the Howe Lake

structure as in other craters (Innes 1961); (5) no
meteoritic material was found nor any magnetic anoma-
ly; and (6) no secondary impact features such as shatter
cones or high pressure minerals were found.
Collapse
A collapse origin for the Howe Lake structure, similar
'U
to that of Crater lake as described by Christiansen (1971)
For personal use only.

and Gendzwill and Hajna1(1971), is also dismissed for

the following reasons: (1) a seismic profile under Howe
Lake (Fig. 4c) does not reveal a collapse structure; (2)
no large intact blocks of till or Upper Cretaceous rocks
were found in displaced positions under Howe Lake as
were found in Crater Lake; and (3) the fill in Howe Lake
is colluvium containing pebbles and boulders mixed '
together in a fashion unlikely to have occurred as a result
of collapse.
Hydrodynamic blowout
Howe Lake may have formed as an elutriation cone
resulting from upward discharge of groundwater from
the Mannville Group. This is the hypothesis we propose
for the origin of the Howe Lake structure; it is developed
in the next section of this report. Although the Howe
Lake structure can no longer be regarded as a collapse
feature, as suggested by Christiansen et al. (1973), the
hydrodynamic blowout hypothesis described herein
offers another mechanism for the non-magmatic origin
of breccia-filled pipes.
Hydrodynamic blowout
Hydrodynamic blowout remains the only plausible
explanation for the origin of the Howe Lake structure.
This hypothesis evolved through more than a decade of
observations of flowing holes and artesian springs in the
region, and from the observations made at Howe Lake
during the attempt to drill into the neck of the funnel in
1972. Experience with a violent blowout at Brightwater
-3 0
'0
O
0
g 8
'D
8
5
0
o
0
?
Marsh (Meneley 1977) provided a working model for 13A31 v3s 3mav 1333

the hypothesis.
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14
For personal use only.

LITHOLOGY PECTRIC LOB $CAE,

IW SPONTANEOUS REStST*IICE
Till. oxidized Silt POTENTlAL
FWt
a Till, unoxidmd Cloy
- O
30
4O
q$% t
Grnval ~olluvium Yln. 0 25
6mm m l m to, dh
Q smd ~ ~ ~ -~*L . =l n Q R I m T A L O*CI p ~ a r wmo w n 18 mt
SALE
-
RG.8. ( a , b) Cross sections across Howe Lake structure.
 1132 CAN. J. EARTH SCI. VOL. 19, 1982

Hydrodynamic blowout structures result from the

upward discharge of groundwater from an underlying
confined artesian aquifer, through a naturally or arti-
ficially produced discontinuity in the overlying confin-
ing beds. The process of groundwater discharge erodes a
funnel-shaped elutriation cone, which eventually be-
comes partly filled with breccia composed of particles
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14

too large to be expelled by the upward flowing water. A

more or less cylindrical conduit extends from the aquifer
up to the base of the cone. The size and shape of the cone
are functions of the flow rate and the strength of the
materials through which the structure forms. The fol-
lowing conditions are required: (1) the area is underlain
by an extensive artesian aquifer capable of delivering at
least 0.04m3/s; (2) the hydrostatic pressure in the
aquifer approaches, or even exceeds, the critical uplift
or lithostatic pressure; and (3) a localized discontinuity
must be present to initiate venting through the overlying
beds.
Brightwater Marsh
Two such blowout structures were produced inadver-
tently when drill holes encountered a confined aquifer
For personal use only.

beneath Brightwater Marsh lowland about 30 km south

of Saskatoon (Meneley 1977). The drill holes, one made
in 1963, the other in 1976, each encountered high
artesian pressure in a gravel aquifer at a depth of 52 m.
Despite efforts to contain them, the holes blew out of
control. The 1963 hole flowed for 6 months before being
plugged. It had developed into a large funnel about 18 m
in diameter at the top. The sides of the funnel sloped
gradually towards the center, where a vertical hole went
down to a depth of 44 m. The bottom of the hole was
filled with coarse rubble, and water was flowing at
0.02-0.04 m3/s before it was plugged. The 1976 hole
blew out despite many precautions during its drilling,
and had an initial flow rate estimated between 0.3 and
b Q crn f 0.6 m3/s, driven by a hydraulic head of at least 25 m
r h of uncontrolled flow a crater
above surface. ~ f t e 22
with a diameter of 9 m formed suddenly by collapse of
the frozen surface layer a few minutes after the drilling
rig, which was attempting to control the flow, was
removed from the hole. The flow rate diminished
steadily until the hole was plugged by grout 2 months
later, at which time it was flowing at 0.02 m3/s. By that
time the hole had reached dimensions similar to those of
the 1963 crater. More than 1500 m3 of sand and silt had
been washed out of the craters and deposited in a thin
blanket east of the area. Nevertheless, a year later there
was little evidence of the sediment visible on the land.

FIG.9. Photographs of blocks of till ( a ) ;Tertiary-Quater-

nary clay (b);and till, Tertiary-Quaternary clay, and collu-
vium in Howe Lake structure (c).
 CHRISTIANSEN ET AL. 11
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14
For personal use only.

aE Rotary test hole ,17001 Contour on top of drift and bedrock,

interval 50 feet
e3 Rotary probe hole
aC Cable tool test hole
A-D Cross section

FIG. 10. Topography of drift and bedrock underlying the fill in the Howe Lake structure.

Howe Lake water wells and observation wells to the northeast

The Howe Lake structure fits the hydrodynamic suggest that the hydraulic conductivity of the sand beds
blowout model in terms of its shape, structure, and should be about 1 x 1W m/s. Because the sand beds in
sedimentary fill as outlined above. In addition, the three the Mannville Group are at least 40 m thick near Howe
hydrodynamic requirements could have easily been Lake, the potential discharge from a blowout structure
satisfied at the time of glacial retreat. would be of the order of 0.8 m3Is, provided that the
The first requirement is satisfied at Howe Lake by the hydraulic head was at least 400 m above ground surface.
presence of the Mannville Group at a depth of about The second requirement may be met by considering
400 m below the present ground surface (Fig. 6). This the ice frontal position that existed at the time that the
unit includes extensive sand and silt beds derived largely Howe Lake structure formed (Fig. 12). An indeter-
from the F'recambrian Shield area to the northeast. minate thickness of stagnant or mobile ice covered the
Although no measurements of hydraulic conductivity area northeast of Howe Lake, creating a recharge upland
have been obtained near Howe Lake, measurements in (Fig. 14). At that time, groundwater flow in the
 CAN. 1. EARTH SCI. VOL. 19, 1982
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14

0 I00 200 300 400 SpO

METRES
For personal use only.

Howe Lake

FIG.11. Seismic contour map in seconds of reflection time for marker bed near top of Devonian rocks, and seismic section near
Howe Lake showing evidence for a regional salt-removal synform.
 CHRISTIANSEN ET AL.
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14

SASKATCHEWAN
For personal use only.

_ _ - A -

LAND
ICE

FIG. 12. Deglaciation of Howe Lake structure. From Chistiansen (1979, p. 927).

Mannville Group would be in a southerly and south- the Mannville Group at the time the Howe Lake
westerly direction, and there would be a strong tendency structure originated. Whether the hydrostatic pressure
for upward groundwater discharge to occur near the exceeded the lithostatic pressure at that time cannot be
stagnant ice margin. Thus, throughout the large inter- determined unequivocally.
lobate area shown in Fig. 12, it appears probable that The third requirement must provide the key to the
high-pressure, flowing, artesian conditions existed in origin of Howe Lake. The Howe Lake structure is on the
 1136 CAN. 1. EARTH SCI. VOL. 19, 1982

- HOWE LAKE SW~-ZZ-~O-IO-~~ flow is able to expel large particles, and the elutriation
cone forms rapidly (Fig. 15a). However, the upward
S-529 (Gyttja) velocity decreases inversely with the area of the cone,
3
wl and the discharge also decreases as a function of the
W
'2
logarithm of time, because the discharge is comparable
3 to that of an artesian well flowing under constant head-
3 15c- GYTTJA variable discharge conditions (Walton 1970). As the
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14

W
m
c-g Om structure becomes larger, the overflow from the elutria-
W
W
i~ COLLUVIUM tion cone decreases greatly in velocity, with a corre-
T sponding reduction in its ability to carry coarser parti-
I
!-
cles. The upper conical section forms by the slumping of
4
0
-. I
the wall materials toward the central conduit. The fine
o ~ X I O O Oyeors B P ~ IS particles produced are carried over the lip of the
S-Soak. Research Council Radiocarbon Laboratory
structure and deposited beyond the perimeter. As the
FIG. 13. Relationship of age of gyttja and depth below ice coarser fraction, comprising boulders from the till and
surface, 1970. blocks of till, moves down the slope of the cone into the
central conduit, the weaker particles are broken apart by
flank of a local synform in the subsurface (Fig. l l ) , the stronger ones as in an autogenous mill. The coarser,
related to dissolution of salt from the Prairie Evaporite strong particles tend to accumulate in the central conduit
Formation. Fractures along the flank of the synform, or and move downward under the influence of gravity,
differential movement at local irregularities along it, opposed by the upward velocity of the water. As the
could have created higher permeability flow paths to the coarse particles obstruct the central conduit, the upward
surface. velocity increases and the walls of the conduit tend to be
For personal use only.

Even without fractures, the synform would disrupt the eroded, enlarging it and allowing the coarse particles to
hydraulic continuity of the Mannville Group, causing a move downward. This enlargement tends to be self-
tendency for upward groundwater flow on the up- limiting because the upward velocity decreases by the
gradient perimeter of the lower permeability zone. This square of the conduit diameter, and the large particles
has been demonstrated by Freeze and Witherspoon can only partly obstruct the flow.
(1967). At some time after its formation, the lower part of the
' ~ n b t h e rfactor that may have contributed to the conduit is filled with the coarser clasts, which tend to be
formation of this structure might have been the explo- finer grained upward (Fig. 15b). It would tend to be an
sive venting of natural gas from the Mannville Group or "open work" gravel and cobble deposit, with perhaps a
some deeper zone, creating a diapiric structure such as few blocks of till accidentally preserved. As the upward
those described by Hedberg (1980). Hedberg also flow diminishes, less material is expelled and more
described the formation of gas hydrates, which may material moves down the slope and into the neck. At this
created a barrier to upward discharge of methane.
According to the theory developed by T6th (1980), there
would be a tendency for hydrocarbon migration toward
stage the neck appears as a "sand boil" in the bottom of
the lake. By this time the upper conical section is in
a more or less stable equilibrium and the amount
I
the margin of the continental ice sheet because of the of material slumping down the slope is diminished.
hydrodynamic gradient in that direction. This could Smaller fragments of till and pebbles move downward
result in the accumulation of gas hydrates beneath the through the sandy "comminution residue" in the con- '
frozen margin of the ice. el tin^ of these gas hydrates duit. Samples of this sort of material were recovered
as the ice margin retreated would allow the gas to from the Howe Lake structure (Fig. 9). The increased
dissipate, which may have been a factor in creating a head loss produced by sedimentation of finer grained
gas-discharge induced structure. Although gas venting material into the central conduit further decreases the
may have initiated the structure, upward groundwater upward discharge, allowing more rapid sedimentation
discharge is essential to produce the observed character- of finer grained material. This leads to a progressive
istics of the sediments now reserved within the dis- plugging of the central cone, which terminates the
charge funnel. development of the hydrodynamic structure. However,
if the Mannville aquifer were still overpressured, the
Evolution of hydrodynamic blowout structures hydraulic head would increase rapidly as the flow
Hydrodynamic blowouts are formed when high- decreases, and this would rejuvenate the structure,
pressure water from an aquifer exits forceably through a expelling the fine material in the neck, and increasing
small opening to the surface. Initially the high upward the rate of erosion. Thus, an eventual dynamic equili-
 CHRISTIANSEN ET AL. I 12

brium could result, with a quasi-stable elutriation

structure as a major focus for groundwater discharge.
The major artesian springs in the Qu'Appelle Lakes
(Christiansen et al. 1977) and in Brightwater Marsh
(Meneley 1977) are believed to be this type of structure.
The scenario may also end in two other ways (Fig.
15c). First, the hydraulic head may dissipate because of
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14

continued melting of the ice sheet. This is certainly

possible to explain the Howe Lake structure, given the
position of the lake and its time of formation. The
second possibility would also satisfy the observed
evidence at Howe Lake. If another hydrodynamic
blowout structure broke out at some distance from Howe
Lake, the resulting hydraulic interference would de-
crease the amount of discharge and lead to the termina-
tion of the structure. Both alternatives would result in a
structure as shown in Fig. 15c.
The cone of influence of a hydrodynamic blowout
structure such as that at Howe Lake may be approxi-
mated by considering the structure as a well completed
in a leaky artesian aquifer. Using the hydraulic conduc-
tivity estimated previously and a steady state flow rate of
0.08 m3/s, the cone of pressure influence of the Howe
For personal use only.

Lake structure would be of the order of 20-30 km. Thus

hydrodynamic blowout structures at spacings less than
40-60km would tend to interfere with one another.
Once formed, a hydrodynamic blowout structure would
persist because it would be a weak point in the confining
layer. It would react instantaneously to an increase in
hydraulic head by increasing discharge. One such
structure would provide a hydraulic head control valve
that would be effective over a large area as long as the
high-pressure glacial discharge existed. These factors
probably account, in part, for the comparative rarity of
these structures.
Results of a chemical analysis of the water in Howe
Lake are shown in Table 2. The low concentrations
indicate that there is no upward flow of sodium chloride
water into Howe Lake from the Mannville Group at this
time. Instead, the low concentration of the lake water
indicates there must be sufficient groundwater discharge
out of Howe Lake to mitigate any concentration of salts
by evaporation of lake water.
TABLE2. Water chemistry in Howe Lake, January
26, 1970

(mg/L) (mg/L)
Bicarbonate 182 Calcium 56
Sulphate 214 Magnesium 46
Chloride 3 Sodium 18
Potassium 14
NOTE:
pH = 7.0; sum o f constituents = 532 mg/L.
 CAN. J. EARTH S(3. VOL. 19, 1982

mic section in Fig. 11 was released by Mobil Oil Canada

Limited. Messrs. Matt Fedak and Walter Tarnowetsky,
who own the land adjacent to Howe Lake, not only
provided access to their land but encouraged the study
by arranging for trailer parking, snow removal, and
many other contributions. Mr. F. M. Atton provided
sounding information of the Howe Lake bottom and was
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14

always available for discussions on this unusual feature.

Cost of geophysical data collection and processing was
paid by Natural Sciences and Engineering Research
Council of Canada grants to D. Gendzwill. Cartography
was done by Carlos Sardinha.

BALDWN,
R. B. 1949. The face of the moon. University of
Chicago Press, Chicago, IL.
BEALS,C. S. 1958, Fossil meteorite craters. Scientific
American, 7, pp. 32-39.
CHERRY,J. A., and WHITAKER,S. H. 1969. Geology and
groundwater resources of the Yorkton area (62 M, N),
Saskatchewan. Saskatchewan Research Council, Geology
Division, Map 9 .
CHRISTIANSEN, E. A. 1968a. Pleistocene stratigraphy of the
Saskatoon area, Saskatchewan, Canada. Canadian Journal
of Earth Sciences, 5, pp. 1167- 1173.
For personal use only.

, - -.. 1968b. A thin till in west-central Saskatchewan,

(b)
llrnlhd local runott Canada. Canadian Journal of Earth Sciences, 5, pp.
329-336.
197 1 . Geology of the Crater Lake collapse structure in
southeastern Saskatchewan. Canadian Journal of Earth
Sciences, 8, pp. 1505-1513.
1979. The Wisconsinan deglaciation of southern
Saskatchewan and adjacent areas. Canadian Journal of
. . Earth Sciences, 16, pp. 913-938.
CHRISTIANSEN, E. A., GENDZWILL, D. J., and DERRY, D. R.
Mannvlllb Group CoHmrhm bcomhq I h r 1973. Pipe-shaped structures. Economic Geology, 68, pp.
1334-1336.
CHRISTIANSEN, E. A., ACTON,D. F., LONG, R. J., MENELEY,
W. A., and SAUER, E. K. 1977. Fort Qu'Appelle geolog.
Saskatchewan Museum of Natural History, Interpretative
0 so W) 1% MI,"
t Report 2 , 8 3 p.
no vartlcal exaggemth FREEZE, R. A., and WITHERSPOON, P. A. 1967. Theoretical
analysis of regional groundwater flow. 2. Effect of water-
f i ~ 15.
. Schematic evolution of the Howe Lake hydro- table configuration and subsurface permeability variation.
dynamic blowout structure. Water Resources Research, 3, pp. 623-634.
GENDZWLL, D. J., and HAJNAL, Z. 1971. Seismic investi-
Conclusion gation of the Crater Lake structure in southeastern Sas-
Howe Lake is interpreted as a hydrodynamic blowout katchewan. Canadian Journal of Earth Sciences, 8, pp.
structure, formed by groundwater discharge from the 1514-1524.
Mannville Group during the final deglaciation of Sas- HEDBERG, H. D. 1980. Methane generation and petroleum
katchewan 12 000- 12 500 years ago. migration. In Problems of petroleum migration. Edited by
W. H. Roberts, 111 and R. J. Cordell. American Association
Acknowledgments of Petroleum Geologists, Studies in Geology No. 10,273 p.
INNES, M . L. S. 1961. The use of gravity methods to study the
Test drilling was done by Torchinsky Consulting underground structure and impact energy of meteorite
Limited, Saskatoon; Elk Point Drilling Limited, North craters. Journal of Geophysical Research, 66, pp. 2225-
Battleford; and Hayter Drilling, Limited, Watrous, 2239.
Saskatchewan. Seismic data for Fig. 4 were collected by MENELEY, W. A. 1977. Control of Brightwater Marsh flowing
Century Geophysics and computer-processed by Seis- testhole. Saskatchewan Research Council, Saskatoon,
can Delta Limited, both of Calgary, Alberta. The seis- Sask., Internal Report.
 CHRISTIANSEN ET AL. 11:

SAWATZKY, H. B. 1967. Composite seismic map, Yorkton water: a mechanism of the transport and accumulation (
area, Saskatchewan. Saskatchewan Department of Mineral petroleum (the generalized hydraulic theory of petrolew
Resources, Drawing No. N-3. migration). In Problems of petroleum migration. Edited t
TALWANI, M., and EWING,M. 1960. Rapid computation of W. H. Roberts, III andR. J. Cordell. American Associatic
gravitational attraction of three-dimensional bodies of arbi- of Petroleum Geologists, Studies in Geology No. 10,273 I
trary shape. Geophysics, 25, pp. 203-225. WALTON,W. C. 1970. Groundwater resource evaluatio~
T~TH, J. 1980. Cross-formational gravity-flow of ground- McGraw-Hill Inc., New York, NY, 664 p.
Can. J. Earth Sci. Downloaded from [Link] by University of Saskatchewan on 06/17/14
For personal use only.

View publication stats