Faults and Engineering Geology

TH E GEOLOGICAL SOCIETY OF AMERICA
ENGINEERING GEOLOGY (BERKEY) VOLUM E NOVEMBER 1860
FAULTS AND ENGINEERING GEOLOGY

B y G eo r g e D . L o ud erback
University of California, Berkeley, Calif.
CONTENTS
Page
In tro d u ctio n ........................................................................................................................................... 125
F a u lt movements................................................................................................................................... 127
Dim ensions............................................................................................................................................. 128
Various fa u lt phenomena..................................................................................................................... 128
Resultant physical conditions.............................................................................................................. 129
P ost-faulting changes............................................................................................................................ 131
Displacement of rock masses............................................................................................................... 132
Recognition of fa u lts ............................................................................................................................. 133
Use of topographic features................................................................................................................. 134
Some specific examples........................................................................................................................ 135
In tro d u ctio n ................................................................................................................................... 135
Seminole D am ................................................................................................................................ 136
Rodrigues D am .............................................................................................................................. 137
Pardee D am ................................................................................................................................... 137
Hales B ar D am .................................................................................... 138
Delaware Aqueduct....................................................................................................................... 138
San Jacinto T unnel....................................................................................................................... 139
Landslides connected w ith fa u lts................................................................................................. 140
Panama C anal............................................................................................................................... 140
A ctive fa u lts .......................................................................................................................................... 140
M ovements along an active fa u lt.................................................................... 142
General statem ent......................................................................................................................... 142
H orizontal movements.................................................................................................................. 143
V ertical movements....................................................................................................................... 144
M agnitude of movements............................................................................................................. 144
Movements norm al to the fa u lt trace......................................................................................... 145
Earthquakes................................................................................................................................... 146
Examples of structures on active fa u lts ............................................................................................. 146
General statem ent......................................................................................................................... 146
Tem escalDam ............................................................................................................................... 147
San Andreas D am ......................................................................................................................... 147
Upper C rystal Springs D am ......................................................................................................... 147
Coyote D am ................................................................................. 148
M orris D am ................................................................................................................................... 148
Railroad Tunnel............................................................................................................................. 149
Tunnels along Aqueduct Lines..................................................................................................... 149
References cite d ..................................................................................................................................... 150
IN T R O D U C TIO N
F a u lts a re ru p tu re s in ro c k m asses th a t a re a cco m p a n ie d b y d iffe re n tia l d is p la c e

m e n t o f o p p o site sides o f th e re s u lta n t fra c tu re s o r fra c tu re zones. T h e y a re th e re fo re
lik e ly to be o f seriou s co n ce rn to th e e n g in e e rin g g e o lo g is t.
I n th e s tu d y o f th e g ro u n d in re la tio n to a p ro po sed e n g in e e rin g s tru c tu re , th e
125
Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

by guest
126 G. D. LOUDERBACK
first concern is with the types of rock masses, their distribution, and mutual rela
tions. Each rock type has its own group of physical characteristics which may vary
within certain limits from place to place, and which must be determined for the
area under consideration by direct field examination. These characteristics may be
practically uniform within the area studied, or they may show certain gradations
or modifications which should be recognized and allowed for. Every part of such
rock mass cannot be observed or tested even with the most elaborate exploratory
work, and the only way an estimate of the conditions may be arrived at is by pro
jecting, through the unseen, the uniform or gradational properties of the seen on
the basis of generally known and locally observed characteristics of the rock type
under consideration. This is the customary, the necessary, and usually valid practice,
although sometimes, even after careful preliminary studies, unexpected or excep
tional conditions may be encountered in the later progress of actual construction
work.
The idea of the controlling importance of the rock type, and the differences in
the problems presented by different rock types are well illustrated by various writings
on engineering geology that have such sections, for example, as dam sites on lime
stone, sites on sandstones and shales, sites on crystalline rocks, etc.
In a given area certain structural features may characterize specific rock types,
such as particular types of jointage, stratification, spheroidal structure, etc. This
relationship of structural features to rock type may also hold for certain minor
features that must logically be included under the usual definition of a fault, such
as joints which, in the course of the folding of the strata, have suffered differential
displacement of their walls. Such minor phenomena are not likely to introduce any
problems, from the practical standpoint, beyond that presented by the occurrence
of the joint system itself. The faults of importance to the engineering geologist as
faults are those whose movements have produced special conditions that may give
rise to serious and difficult engineering problems. Such a fault may cut through any
mass and associated rock masses irrespective of their lithologic characters or pre
existent structures. It may have any location within the rock mass that it intersects,
and may have its own direction or attitude, irrespective of the strike or dip of the
affected rock. In certain areas, a number of faults may form a rather homogeneous
system within which the individual faults may be similar in strike, dip, and direction
of displacement. But even in such occurrences the spacing is rarely regular enough
to permit the prediction of the location of one fault from a knowledge of the location
of others.
While some faults produce very striking phenomena, such as scarps or offsets of
strata, which may be readily recognizable sometimes even at a distance, others
may be difficult or impossible to recognize by surface studies. They may yield no
characteristic topographic features, and the rock-surface intersection may not out
crop but be covered by soil, rock waste, or other masking material, and their existence
be unsuspected.
The zone affected by a fault of engineering significance may be but a few feet
wide, and the rock on either side of it may give no adequate hint of its proximity.
Some sites carrying a fault or faults have been examined with the aid of exploratory

by guest
FAULTS AND ENGINEERING GEOLOGY 127
borings, pits, or shafts placed at selected intervals, but the fault or faults lay within
the intervals and were not discovered. In each exploratory opening the rock dis
closed was considered suitable for the purpose in view, and as it was of the same
type in adjacent holes was projected through the intervening unseen ground and
the site accepted as satisfactory. Only after contracts were let and the proposed
excavation largely completed was the fault (or faults) discovered. Such discovery
generally results in increased, sometimes greatly increased total cost for the project.
In some cases where the site is still usable the total expense may be not much greater
than it would have been had the condition been recognized in the exploratory stage.
But even so, it may cause financial difficulty by raising the costs beyond the esti
mates, the readily obtainable finances, or the capital provided by a difficultly won
bond issue.
The late discovery of unsuspected conditions produced by faulting may neces
sitate the redesign of the structure, the shifting of its location, or even the aban
donment of the site. It is therefore, in general, an important task of the engineering
geologist to determine at an early stage the existence of any fault, or faults, that
affect a construction site, and to examine or explore the associated physical condi
tions sufficiently to arrive at a judgment as to their bearing on the project.
FAULT MOVEMENTS
The differential movements (faulting) which produce and characterize faults may
possibly occur immediately in connection with the initial rupture and then cease,
but apparently in the great majority of faults, especially the larger ones, total ob
served displacement represents movement over a long period of time. Most faults
are definitely related to some particular period and type of crustal disturbance and
are not reactivated during later periods characterized by different types of crustal
deformation (didstrophism). But there is evidence that some faults, produced during
one period of diastrophic activity, have been rejuvenated in a later period, during
which the movements may be different, even the reverse of those in the earlier period.
Fault movement may take place by gradual creep, and this may have been a common
mode in weaker rocks of shallow thrusts related to folding. But faulting frequently
progresses by a discontinuous succession of comparatively short displacements
separated by longer periods of inactivity.
Faults are often classified as active or dead faults. Those are considered active
which are undergoing movement now or have undergone movements in recent
geologic or in historic time and are considered liable to recurrent movements in
the future. Dead faults are those which were active in some earlier period of di-
astrophism but show no sign of having been active in more recent geologic time.
Whether a fault be active or dead, it has produced certain physical changes in
the rock masses through which it passes. There is no difference in the physical dam
age done to rocks as between active and dead faults, because in both types such dam
age was produced when the fault was active. The various rock conditions dependent
on faulting may therefore be described for both types of faults in one set of statements,
and such conditions must be reckoned with in engineering projects whatever the
activity status of the fault under consideration. They may be called the static aspects

by guest
of faulting; if the fault is judged to be active, the dynamic aspect is superimposed—

that is, the possibility or probability that renewed fault movement may take place
after the structure is completed (or even in course of construction). The possible
effects of such movement require special additional consideration.
DIMENSIONS
If a fault is defined as a fracture that shows relative displacement of the two
sides, certain phenomena corresponding to that definition may be seen only under
the microscope, and the total length of fracture may be but a small fraction of an
inch. From these minute structures at one limit there occur all sizes of faults up
to great ones hundreds of miles long. The small ones, up to a number of feet in
length, are generally of no concern from the engineering standpoint unless they
are sufficiently numerous to affect materially the strength or permeability of the
rock.
As to total displacement along faults, the small ones are naturally limited; in
large faults it may vary up to several thousand feet and even, in extreme cases,
up to a number of miles.
When a fault breaks through a rock mass it sometimes produces a clean, simple
fracture. If the surface of fracture (fault surface) approximates a plane surface it
is called a fault plane. The fault surface may be a curved surface. It is common
practice to call the curved fault surface, because the radius of curvature is usually
large, the fault plane, and to give its attitude by dip and strike. In situations where
a very limited section of a fault is under consideration, as at a dam site or a site
for some other surface structure, this practice is usually entirely satisfactory, but
in some problems it is very desirable to discover and take into consideration the
curvature of the fault surface. Such a problem is that of projecting a fault observed
at the surface to the level of a proposed tunnel which may be many (several hundred
or a thousand or more) feet below.
Very commonly the effects of faulting are not limited to a simple fracture, but
the rocks are disturbed or modified for some distance measured transversely to the
tault surface or general direction of rupture. The mass within which such modifica
tions prevail is called the fault zone or zone of faulting. In the fault zone the rock
may show divergent or subparallel fractures, brecciation, crushing, and other
phenomena. The limits within which the original rock structure and texture have
thus suffered damage or even destruction vary greatly. Narrow zones up to 10 or
20 feet are frequent, but broad zones are encountered, sometimes measured by
hundreds of feet, more rarely, thousands.
VARIOUS FAULT PHENOMENA
The rock masses adjacent to the fault surface are called the walls, and, if the
fault surface is other than vertical, the overlying mass is called the hanging wall,
the underlying, the foot wall. If the fracture is the direct result of shear, or if, during
fault movement, the opposing faces of the fracture are in contact, especially if held
together by a component of pressure transverse to the fracture surface, the mutual
reactions of the opposing faces in the course of displacement act to smooth both

by guest
faces and sometimes to polish them. The more or less polished surfaces are called
slickensides, and it is remarkable what small displacements will sometimes result:
in highly polished surfaces. Minor irregularities in hardness or variations in resistance
to deformation in the contact surfaces often produce striations or grooves on the slick-
ensided surface. These striations or grooves are of value in determining the actual
line of movement in the fault surface, and sometimes even give evidence of the
direction of movement along the line. This can sometimes be recognized by the
consistent tailing out of rubbed or smeared crystals or other recognizable discrete
particles along the slickensided surfaces, or the minor steplike or other relation of
the main fracture to minor cross fractures or gash structures in the walls which are
consistent with movement in only one direction.
Caution must be used in attempting to determine the direction of displacement
in the fault itself by such phenomena, even though the inferences are correct for
the part of the surface actually observed. For it sometimes happens that the fault
movement is not throughout a simple translation, but different parts of the mass
may partake of local movements involving rotation, or some of the blocks into
which the mass may be broken may show displacement diverging in various ways
from the main direction of fault movement.
RESULTANT PHYSICAL CONDITIONS
The effects of faulting on rock masses are quite varied, dependent on the character
of the rock, the way the disruptive forces are applied, and the general physical
conditions under which the action takes place. Furthermore, under certain condi
tions, other geological agents may be enabled, as a result of the faulting, to bring
about additional modifications in and near the parts affected by faulting.
The simplest result of faulting is a simple fracture. This is, in the main, limited
to small faults or small offshoots of larger faults. The crack may be tightly enough
closed so that water seepage is negligible, and the strength of the rock mass may
not be appreciably affected. But fractures are often open to a greater or less degree
(fissures) and may allow an active seepage or even a rather free flow of water through
them. In many instances where a fault zone is developed, with many subparallel
and sometimes branching and anastomosing fissures, especially in harder and more
brittle rocks, large quantities of water may be transmitted, and very serious prob
lems have thereby arisen in tunnel construction, especially in deep ones where high
hydrostatic pressure prevails. Such conditions may result in long delays and expen
sive and sometimes dangerous operation. The danger generally arises from sudden
and often unexpected flooding, and possibly more or less explosive blowouts. In
dam construction such fracture zones produce problems of preventing undesirable
underflow and development of uplift pressure.
The surfaces of fault movement may, especially in softer, nonbrittle rocks, become
more and more numerous until the mass within the fault zone is dominated by
movement surfaces. In my own experience, the best examples of such conditions
occurred in serpentine and in shale. Separable pieces of the material were usually
lenticular and commonly only a fraction of an inch in length, but sometimes up to
an inch or more. On removal from the mass the fragments were found to be covered

by guest
entirely by smooth, slickensided surfaces. In my notes I have come to refer to such

masses as panslicked to indicate briefly that the whole surface of every fragment
was completely slickensided. Such material, especially under the influence of moisture
but without the necessity of free seepage, can at depth exert effective ground pres
sure, giving rise to serious difficulty in tunnel construction, and it has caused failure
of a concrete lining of an already constructed tunnel. Naturally it is poor foundation
for surface structures, and has given rise to landslides. It has not proved possible
to cure the condition by grouting.
The deformative movements that give rise to faulting are not limited in their
expression to slipping along one or many fault surfaces. Those movements are
frequently associated with distortion, shattering, or crushing of the rock material
between the fault surfaces of a complex zone and out into the country on one or
both sides of any fault surface.
A noteworthy type of distortion, best observed in the weaker stratified rocks
and almost nonexistent in the stronger massive rocks, is called drag. While this
may occasionally be initiated by deformation along the zone before actual fracture
occurs, in the main it is the result of the interaction of closely appressed moving
walls of the fault on each other by which the rock in each wall is bent, kneeded,
or smeared back from the direction in which it is moving or, otherwise expressed,
is bent in the direction in which the opposing wall is moving. Stratification or other
structural features cut by the fault are bent or curved in opposite directions in the
two walls, and, if the rock is sufficiently flexible, the deformed structures tend to be
come tangent to the fault surface, and plastic facies are smeared out along that
surface. This phenomenon is a useful indicator of the direction of movement along
the fault. If the original bedding approximates the attitude of the fault surface,
local folds may be produced in the vicinity of that surface and are known as drag
folds.
Shattering is the cracking of the rock along various fractures, sometimes numerous,
which may show little or no differential displacement of the walls along the fracture
surface, and may therefore develop no slickensides. The cracks may be more or
less regular in form and orientation or irregular, and they may lie at any angle to
the fault surface.
Regularly arranged fractures may be considered as sets of joints resulting from
the strains leading up to or coincident with the faulting. They may be of the com-
pressional or shear type and may be arranged in complementary sets as visualized
in the strain ellipsoid; or they may be of the tensional or tear type, sometimes the
result of drag along the fault, and extend some distance from the fault surface into
the wall and be more open at the fault surface and narrow away from it {gash frac
tures).
The more irregular fractures break the rock into many irregular fragments. They
are best developed in the more brittle rocks. The mechanical action producing them
probably involves a combination of compressional and tortional effects. The result
ant material may often be designated breccia and is one type of fault breccia, and
it may well supply a complex of channels for active water flow.

by guest
Conditions of strain materially aided by differential movement which introduces

rotational effects may give rise to crushing. This involves any or all of the following
types of action: flattening, squeezing out or smearing out of fragments, grinding,
abrasion, granulation, mashing, pulverization, milonitization, and the alteration
chemically of the mineral constituents.
A zone characterized by such phenomena is a crush zone. If it shows larger frag
ments in a matrix of ground-up material it is called a fault breccia or, if the larger
fragments have been rounded by the abrasion, a., fault conglomerate. Where the
material has been finely ground it may be called gouge (or fault gouge), a common
type of which with a clay base (often produced by mineral alteration) is called a
clay gouge. It is evident that such zones will frequently be zones of weakness and
sometimes pose very serious problems in engineering construction. Clay gouges,
sometimes moist and plastic, may form good barriers to water flow. In some mines
the removal of gouge, and in some tunnels the necessary cutting through a gouge,
has been followed by serious flooding of water coming from aquifers or fracture
zones beyond.
The various effects above described, especially multiple and complex fracturing
and crushing, often accompanied by deleterious chemical changes, may so weaken
the rock material within the fault zone that it becomes distinctly inferior founda
tion with reduced bearing capacity, sometimes subject to squeezing out under
imposed pressure, or of transmitting ground pressure dangerous to certain kinds
of structures. If appropriately located, as in the abutment area of a dam, or in the
foundation, especially if dipping outward, of some surface structure built above a
slope, it may furnish conditions favorable to landsliding, which may cause trouble
during construction or may later menace the stability of the structure and even
lead to its failure.
POST-FAULTING CHANGES ,
In describing the physical effects of faulting, various types have been noted
where conditions favorable to the access and free flow or seepage of water are pro
duced. This makes possible further changes in the rocks, brought about by water
action. The water responsible for the changes may be of surface origin (meteoric)
or of deep-seated origin (juvenile).
Faulting may give access to mineraliferous waters under such physical conditions
that deposition takes place. The fault fractures and associated fissures in the sur
rounding country may be partly or entirely filled by secondary mineral deposition,
and, under favorable circumstances, the ground may become as free from water
flow and active seepage as it was before faulting, and may be satisfactorily strong
and firm from the engineering standpoint. In some instances the mineral-bearing
solutions may not only form deposits in the fissures, but may invade and modify
the composition of the country rock (metasomatic replacement). This action may
improve or may deteriorate the rock from the standpoint of its physical character
istics. In any case, the altered rock would have to be considered on its own merits
as distinct from the unaltered rock, and their mutual boundary would probably be
very irregular, although possibly arranged according to a definite pattern.

by guest
Post-mineralization faulting may, of course, alter the whole picture and produce
anew in the altered mass any of the features described as the result of faulting.
The most common types of alteration produced by meteoric water are those
typical of weathering, especially solution and decomposition. Solution in fracture
zones of such soluble rocks as limestone may give rise to underground caverns,
which may prove highly troublesome under dam sites or in connection with various
other types of structures. Rocks with a soluble natural cement may become greatly
weakened and made permeable by its solution, and may even be reduced to a
mass of loose sand. Decomposition weathering may be very slight in some fault
zones, sometimes merely discoloring the fracture surfaces, but it may be very ad
vanced, materially weakening the rock even to the point of transforming it into a
mass of impure residual clay.
In the arid regions fractures are often found filled with readily soluble minerals,
the most commonly troublesome of which is gypsum. Impounded water, having
access to such fissures, may progressively dissolve the filling, giving rise to water
flow and leakage after the structure is completed, and under certain circumstances
the material between the fissures may wash out under the effects of the flow, or
may soften up and squeeze out or slide, under the effects of water under pressure.
This is believed to have been an important contributory cause to the failure of the
St. Francis dam in California, and has caused trouble in connection with other
types of structures.
Sulphides, especially pyrite and marcasite, deposited in or about fault zones,
when reached by oxidizing surface waters, may give rise to the production of sul
phuric acid, which may attack and alter or weaken the surrounding rock or the
concrete or mortar of an artificial structure.
DISPLACEMENT OP ROCK MASSES
Irrespective of the results of faulting on the physical properties of any of the
rocks affected by the action, the mere displacement of rock masses may bring about
unsatisfactory conditions or raise problems in engineering construction. This may
happen in terranes where layers or masses of rocks of significantly different physical
properties are involved, and a weak or pervious rock is moved so as to adjoin or
lie in prolongation of a strong or impervious body. As one type of example, it may
be recalled that a consequent stream normally crosses a series of strata transversely,
and a rock layer approaching its valley from one side will usually be found on the
immediately opposite side with very similar if not identical characteristics. If it
is planned to place a dam or bridge across the valley of such a stream, a satisfactory
abutment on one side is likely to be matched by a corresponding satisfactory abut
ment on the opposite side. But if a fault lies along the streamway, it may happen
that it has so offset the strata that, where a satisfactory abutment is found on one
side, the opposite side may present only inferior or unsatisfactory material for such
purpose. Furthermore, in the case of a dam, foundation conditions may change
from satisfactory to unsatisfactory along the valley bottom.

by guest
RECOGNITION OF FAULTS
The existence and local effects of faults should be determined, insofar as feasible,
early in the course of deciding on the specific location and plans of an engineering
structure because their presence may materially affect the suitability or treatment
of the site and determine the design and cost of the structure. The first search should
be made by surface geological studies. If definite faults are discovered, supplementary
exploratory openings are usually required to determine, more definitely than can
generally be done by surface examination, their characteristics and especially the
physical effects they have produced in the rock masses. It often happens that evi
dence for even a significant fault zone in an area critical for an engineering structure
may not be visible at the surface. This may be the result of masking by soil, by
rock waste cover, or by alluvial or other post-faulting deposits. When the absence
of faults cannot be reasonably inferred from surface observations, subsurface ex
ploration becomes essential. Depending on local conditions, this may consist of
trenching, boring, shaft sinking, or tunneling, or a combination of two or more of
these methods. With comparatively thick depositional cover, geophysical explora
tion may be helpful in locating fault zones and depth of cover, and thereby of direct
ing other types of investigation.
As a general rule, faults are most easily recognized in stratified rocks, whether
sedimentary, igneous, or metamorphic, and particularly when the course of the
strata is readily visible. When the fault is transverse {dip fault) or oblique {oblique
fault) to the strike of the bedding, the offset of the strata is quite distinctive, and
the location of the fault, and frequently the extent of the displacement, are easily
determinable. A fault whose strike parallels the strike of the bedding {strike fault)
is more difficult to recognize unless its dip differs from the dip of the beds and ex
posures are available that disclose the dip component. Bedding faults, which parallel
the bedding, produce no offset of the strata, and their recognition must depend upon
the identification of such phenomena as slickensides, a gouge belt, a zone of brec-
ciation or crushing, or other characteristic products of faulting.
Because of the breaking, crushing, and other action disruptive of a rock and its
component minerals, fault zones are often subject to more rapid and complete dis
integration and chemical weathering than the country rock in which they he. They
may therefore be covered by a greater thickness of weathering products, and their
characteristic fault features may be hidden from surface observation even in an
area where the general soil covering or weathering effects do not seriously hinder
the recognition of the stratification of the country rock or the possibility of follow
ing the trend of certain beds across country. With such conditions of rock exposure,
the sudden termination of a characteristic bed, or the termination of a recognizable
sequence of beds, and the recognition of its continuance in offset position will usually
indicate fault displacement. If lenticular members are present at different horizons
in the formation, care should be taken not to mistake their sporadic occurrence
for fault offset. If offset is actually present, the approximate position of the fault
may be recognized, and its actual position, width of zone, and physical character
istics may be determined by trenching if the covering is thin. If the covering is

by guest
thick between the separated ends of the strata, as often occurs in a valley crossing,
some other method of exploration will be necessary.
Without good exposures, faults are often difficult to recognize in massive igneous
rocks. Sometimes the offset of a distinctive dike or vein may give the same kind of
clue given by a distinctive layer in a stratified formation.
USE OP TOPOGRAPHIC FEATURES
Topographic features often indicate or suggest the presence of a fault. In many
localities a fault line or zone is accompanied by a comparatively steep rise or scarp.
If the faulting is recent such a scarp may have originated from the actual differential
displacement of the ground on the two sides of the fault by virtue of the fault move
ment {fault scarp). Fault scarps are naturally subject to erosion. The original slope
may thereby be lowered {degraded fault scarp), and such original characteristics as
slickensides removed from the surface. Furthermore, the base of the scarp may
retreat from the fault line so that no fault can be found at the bottom of the slope,
but only at some distance away.
A scarp related to a fault may also be produced by erosion by virtue of differen
tial resistance, on the two sides of the fault, of different types of rock which have
been brought into juxtaposition by the fault displacement. Sometimes even a reverse
scarp is produced—that is, one that shows topographic elevation on the downthrow
side of the fault. Erosional scarps related to fault lines, whether direct or reverse,
are called fault line scarps.
Scarps accompanying faults may be very helpful in locating such faults. It is
well, however, to search for evidence of offset or other phenomena characteristic
of faulting before deciding on the existence and location of a fault suggested by
topographic forms. If surface observation does not furnish the deciding evidence,
subsurface exploration should be undertaken if the suggested fault may be of import
to the project under consideration.
It may be recalled that comparatively steep elements of slope (scarps) may be
caused by geological agencies other than faulting. Erosion of superimposed layers
of markedly different resistance in a normal stratigraphic sequence, or of a dike in
a formation of distinctly greater erodability may produce excellent and sometimes
long scarps. Marine or lake abrasion and lateral stream action of former times may
give rise to steep rises (cliffs), sometimes between flat slopes (terraces). Certain
volcanic flows may have original (or degraded) terminal scarps. Although the dis
crimination of such scarps from fault scarps or fault-line scarps is often easily made,
it is sometimes difficult. All these types have at times been incorrectly appraised
and have led to claims of the existence of a fault or faults where no corresponding
fault existed.
The shearing and crushing of rock along a fault zone may produce a belt
of markedly lowered resistance to weathering and erosion compared with the rock
on both sides of the fault zone, and this is likely to have topographic expression in
the development of a natural trench or roll of the surface. In areas of massive rock
such phenomena may be the only clues to the location of a fault. As the actual
structure may be hidden by soil or other weathering products, and as other con

by guest
ditions may give rise to like topographic appearance, positive confirmatory evidence
may be obtainable only by trenching or other subsurface exploration. A weakly
resistant stratum flanked by members of greater resistance to weathering, or a
less resistant dike in the midst of a more resistant formation, may produce the same
topographic form.
Such zones of weakness are favorable to the development of subsequent streams,
and they may even determine the course of a consequent stream, at least in part.
When such has occurred, direct field observation of the fault trace is usually pre
vented by a covering of alluvial deposits. A fault along a stream course, which
would cut through the foundation of a proposed dam from reservoir site to the
downstream country, may give rise to serious problems of dam design and con
struction, and a knowledge of its existence and characteristics may be of the great
est importance.
Under favorable conditions the existence of such a fault may be recognized by
the evident offset of strata which cross the valley, as observed and mapped on the
valley sides and interstream areas; or a fault may enter the valley obliquely and
its course be recognized beyond the alluvial cover. However, frequently no such
surface evidence is available. Unless the general characteristics of the area or the
exposures lead to the conclusion that no fault is present, subsurface exploration is
desirable. Where a valley is crossed by beds or other tabular rock masses, which
are sufficiently characteristic to be recognized in samples, a linear group of vertical
borings may indicate the position of a fault by the offset relation, vertically or/and
horizontally of a contact intersected in the borings. If a fault is so located, more
definite examination of its physical effects may be made by sinking a shaft at the
indicated position, if water and other conditions permit that type of exploration.
Valuable information as to the presence or absence of faults and width of the
zones have been obtained, especially in massive rocks, by the core boring of inclined
holes from the two banks through the bedrock so as to explore the full width under
the stream deposits. In the use of core boring for this and other exploratory purposes,
care should be taken not to overlook the existence of a fault zone or other bad ground
by the neglect (as has occurred in actual practice) of low or no recovery of core from
certain lengths of the bore hole. Another method, which often gives excellent ob
servational opportunities, is the sinking of a shaft in the side of the valley and tunnel
ing through bedrock entirely across the alluvium-covered foundation. Before under
taking either of these methods, the depth of the alluvium should be determined by a
number of vertical bore holes or by geophysical tests to assure that the inclined
bore holes or the tunnel will be entirely in bedrock. Even if the tunnel is in bedrock,
if serious fault Conditions are present, difficulty of exploration and observation may
be encountered as a result of water flow or heavy ground in the fault zone. This,
in itself, would be important information as to foundation conditions.
SOME SPECIFIC EXAMPLES
INTROD UCTIO N
A number of proposed dam sites have been abandoned or condemned, even after
considerable work had been done on them, on the discovery that they have been

by guest
136 G. D. LOTJDERBACK
seriously affected by faulting, often because of considerations of safety, sometimes

because of the heavy expense involved in appropriate treatment. This does not
always mean that a broad zone or zones of physical damage to the rocks is present.
Bromehead (1938) reported that two dam sites selected by the water engineer were
condemned by the Geological Survey of Great Britain because faulting had brought
two pervious sandstones in contact. A near-by site in unbroken shale was recom
mended. The possibility of an alternate site should always be considered, if par
ticularly troublesome conditions are recognized. However, a fault-affected area may
sometimes be the only available site, or for other important reasons the most suit
able site. It should, of course, be recognized that many types of conditions, even
quite serious conditions, produced by faulting can be and have been satisfactorily
controlled by appropriate engineering treatment, and the main question may narrow
down to the ability and willingness to meet the necessary expense.
The above general statements apply also to types of engineering projects other
than dams. In the planning of the great Catskill Aqueduct, one of the principles
used in selecting the route was to avoid, insofar as possible, fault and crush zones
(Berkey and Sanborn, 1923, p. 53), and this principle was instrumental in at least
two instances in aiding determination of the choice of route. Such choice however,
is not always possible.
A few examples will be given of fault conditions that have been met in specific
projects.
SEM INOLE DAM
This dam, on the North Platte River, Wyoming, was planned as a concrete arch
dam, 261 feet high from foundation to roadway, with storage capacity of reservoir
1,020,000 acre feet. Two sites were carefully examined, but the upper one, while
considered feasible, was discarded in favor of another where apparent depth to rock,
solid enough for foundation, would be less. The chosen site is entirely in Precambrian
granite. Foundation investigations included 17 diamond-drill holes, 4 vertical shafts,
and 8 horizontal drifts into the abutments. Five drill holes were directed at angles.
The dark-gray granite bedrock is extremely hard but broken by numerous seams
and fractures. The natural solid rock slopes of the canyon were steep. The river fill
of sand, gravel, and boulders was 25 feet deep (Keener, 1937).
Notwithstanding the extensive exploration, unexpected conditions were met in
the course of excavation. Considerable delay was experienced in the construction
of the diversion tunnel, which was planned in continuation of the horizontal part
of the spillway tunnel in the right abutment. After going through 30 feet of sound
granite from the upstream portal, a belt of disintegrated granite with numerous
mud seams was encountered. A series of shear planes intersected the spillway inlet,
requiring the removal of some 30,000 cubic yards of material. After river diversion,
scaling of the cliffs had to be co-ordinated with removal of overburden from the
canyon floor. In this process a slide of 2500 cubic yards developed on the left and
went down into the upstream side of the foundation excavation. In the course of
renewed excavation, two faults, intersecting at approximately a right angle, were
discovered. The soft material developed at the intersection of the two fault zones

by guest
was believed responsible for precipitating the slide. One fault cut through the portal
of the spillway tunnel and crossed the river gorge at the upstream toe of the dam
and dipped 50° downstream; the other followed the left bank and dipped 70° under
the left abutment. Both fault zones were badly decayed, ranged from 2 to 10 feet
in width, but the granite of their walls was in satisfactory condition. As recom
mended by a consulting board, the excavation of the zones to 30 feet below general
bedrock surface and back-filling with concrete, and the grouting of the walls was
considered entirely satisfactory treatment (Warner, 1937; Engineering News-Record,
1939).
ROD RIG UES DAM
This is on the main branch of the Tijuana River, Baja California. The rocks ex
posed in the walls of the gorge were rhyolite on the west slope and part of the east
slope, and granite on the eastern part of the east slope. In the narrowest parts of
the canyon the walls exposed fresh but considerably broken rock. On the upper,
gentler slopes the rock was badly weathered. Exploration work consisted of 9 bor
ings in the stream bed and 52 test pits on the side slopes. Five of the borings indicated
satisfactory rock at not over 50 feet. From four borings no core was obtained; from
two of these, sharp stone chips were recovered; from the other two, only weathered
fragments. It was planned to erect an Amburson dam, 187 feet above stream bed,
240 feet above lowest foundation rock. Excavation of the stream bed exposed a
fault zone 20 feet wide, which had not been indicated by surface or exploratory
studies. It lay nearly parallel to the stream bed along its eastern margin. Badly
disintegrated rock was also found in mid-stream portion of the excavation for the
cutoff wall. Doubt as to the feasibility of the site led to the calling in of a consulting
board. The site was finally approved with special provisions for meeting the condi
tions, among which was excavation of the cutoff trench to 112 feet below original
rock surface (150 feet below stream bed), and a trench along the fault zone 300 feet
below stream bed. The ingenious engineering provisions to meet the conditions by
modification of the dam design are described in the reference (Williams, 1932; 1933).
PARDEE DAM
On the Mokelumne River in California this concrete dam rises 351 feet above
river bed (358 above bedrock). The total capacity of the reservoir is 240,000 acre
feet. The rocks of the canyon, in which the dam site is located, are a stratified series
originally of volcanic breccias, ash beds, lava flows, and terrigenous and mixed
(ashy and earthy) sediments now all more or less metamorphosed. They strike
obliquely across the river and dip northeast (upstream) at 54° to 84°. The volcanics
have become greenstones, and the fine-grained sediments, more or less well-developed
slates. The volcanic types predominate, and the more massive facies form the strong
est rock masses of the area and determine the steep-sided canyon walls along this
part of the stream course. The clay-slate belt is in general only about 15 to 18 feet
thick.
The original plan called for an arch dam. Foundation and abutment exploration
disclosed an incipient slide in the right (north) canyon wall and showed that the
lower part of the slate belt was broken and sheared and that a fault, roughly paral-

by guest
lei to the stratification, had produced a soft gouge zone that suggested recent move
ment. This weak belt would have passed under the dam from the south side along
the planned trench of the cutoff wall, leaving diagonally along the river crossing,
and passing into the floor of the reservoir. Such conditions were considered un
desirable for an arch dam. The problem was rather simply solved by shifting the
dam location a moderate distance downstream, and redesigning the structure as
a gravity dam, though a curved alignment was retained. The dam was completed
in July 1929 and has been in uneventful operation since.
H ALES BAR DAM
This dam lies on the Tennessee River on cavernous Mississippian limestone. It
is evident from the history of the project that the exploratory work gave an in
complete understanding of foundation conditions. Construction was started in 1905
with the expectation that it would be erected in two years, but difficulty with the
foundation delayed completion until late in 1913, and added greatly to the estimated
expense. Eleven days after completion, marked leakage was recognized, and the
leaks progressively increased in size and number. A variety of methods were used
to stop the leakage but gave only temporary relief. In 1939 the dam was purchased
by the Tennessee Valley Authority, which undertook a careful survey of the condi
tions and developed a plan to build a cutoff without emptying the reservoir. At
that time attempts at careful measurement of the amount of leakage were made
by the T. V. A. and the U. S. Geological Survey and resulted in figures of 1720 and
1650 cu. ft. /sec. respectively, or about yo the normal low water flow of the river.
A four-year repair program was undertaken which finally stopped all but an in
significant amount of the leakage. It had been found that beneath the dam are two
important fault zones, each characterized by a large east-dipping thrust opposed
by several west-dipping thrusts beneath it. The small thrusts terminate on contact
with the larger one at the upper end and against a poorly defined gently east-dipping
shear plane at the lower end. The major cavitation beneath the dam is confined to
the area between the large thrust above and the shear plane below. An excellent
description of the conditions and history of the project is given by Frink (1946).
DELAWARE AQUEDUCT
An outline account of the fault crossings in the course of the 85-mile long Delaware
Aqueduct, built to carry water from the west slope of the Catskills to New York
City through a concrete-lined pressure waterway 300 to 2000 feet below the surface,
is given in the Engineering News-Record (1941). Although in selecting the route
careful consideration was given to avoiding as many as possible of the known fault
zones, it was found to be neither economical nor practical to by-pass all of them.
To give an idea of the types of difficulties encountered, two of the crossings may
be described here briefly.
North of shaft 12 they passed into Manhattan schist which gradually became
slabby and jointed, and the tunnel became wet. While mucking out a shot on August
10, 1939, an inflow of disintegrated and decayed rock, about 1000 cu. yds., and a
large inflow of water came into the tunnel. A timber bulkhead was erected 26 feet

by guest
back of the face and a timbered top drift begun. On August IS, the drift advanced
only 10 feet, when another inflow of decayed rock and water occurred, breaking the
timber bulkhead and filling the tunnel back 56 feet from the face. A 16-foot concrete
bulkhead was then constructed 75 feet back from the face, and borings made through
it into the fault zone. Several holes met large flow at high pressure, but both flows
and pressure diminished as the flow continued. Ultimately at this Purdy Lake fault
zone, steel plate interlinings, surrounded by heavily reinforced concrete, were in
stalled for a distance of 440 feet.
Preliminary borings indicated that at the Eastview fault the tunnel would pass
from Manhattan schist to Inglewood limestone with faulted contact. On May 17,
1940, the quality of the rock became so poor that a small drift, begun on the west
side of the face, penetrated rock so badly decayed that it resembled a mixture of
sand, clay, and gravel. It was impervious and did not run, but during excavation of
the drift a joint opened in the hard schist on the opposite face of the main tunnel
admitting 750 gal/min. of water. A 20-foot concrete bulkhead was constructed 28
feet back of the face, and the space filled with concrete. Pipes were carried through
to rock face for grouting, and a total of 64,338 bags of cement was used at pressures
up to 900 lbs/sq. in. Borings through the bulkhead indicated the water was not
sealed off, so a second grouting was tried, using 31,287 bags of cement, but was also
unsuccessful. Borings were drilled 155 feet entirely across the fault zone. At that
time 550 gal/min. were coming in, gradually diminished to 150, and the pressure
dropped from 175 to 25 lbs/sq. in. A 6- by 6-foot drift was driven on each side of
the tunnel at invert grade and reached hard limestone about 100 feet ahead of
original advance and ended 30 feet farther on. In the west drift a flow of 1000 gal/
min. filled the end of the drift with rock detritus, temporarily stopping progress.
After a few days inflow ceased. Steel-plate interlining was used here also.
SAN JACINTO TUNNEL
This tunnel is part of the aqueduct of the Metropolitan Water District of South
ern California which was built to carry water about 240 miles from the Colorado
River to the Los Angeles district. The most difficult construction job along this
line was the 13j-mile tunnel through the San Jacinto Mountains. The prevailing
granitic rock, in part firm and undisturbed, was cut by a number of major and
minor faults. With advance of headings, difficulty arose through inflow of water
under high pressure, and heavy ground conditions in the faulted and fractured
zones. In-flows up to 15,800 gal/min. in or near a single working face continued
with discouraging persistence. Water pressures as high as 600 lbs/sq. in. caved in
headings, or brought down the arch; along part of the tunnel water had to be pumped
out against an 800-foot head through a shaft that was flooded repeatedly while work
was under contract, at one time to a depth of 647 feet. The peak discharge of the
entire tunnel reached 40,000 gal/min. in February 1938. Pioneer tunnels from the
two working shafts, 10 by 10 feet and parallel to the main bore, offset 75 feet, were
driven to give advance information as to location and extent of bad ground, to im
prove drainage, afford additional entry and possible means of detouring utilities
and traffic in delayed sections. They did not always give advanced indication of

by guest
the true nature of the rock, nor always accomplish the drainage in the main tunnel,
owing to broken formation and consequent sudden change in the rock. Heavy surges
of water brought in quantities of fault filling and debris. The Goetz fault not only
yielded heavy flows of water but on one occasion released about 3000 cu. yds. of
sand and detritus into the tunnel. Repeated relocation of portions of the tunnel
were necessary (Leadbetter, 1938).
LANDSLIDES CONNECTED W ITH FAULTS
Many landslides are produced without the assistance of faults. These usually
occur in weaker formations as a result of oversteepened slopes, reduction of fric
tional resistance by access of water, removal of basal or lateral support, or over
loading especially at the top of the slope. But faults are frequently instrumental in
giving rise to conditions favorable to landsliding even in formations that otherwise
would be strong enough to resist such action. This is the result of the production of
crush, gouge, or shattered zones and often the development of planes of weakness
that may be so related to erosionally or artificially developed slopes that even large
slabs or other masses may break loose and move down along or into such slopes.
PANAMA CANAL
In the course of its construction and maintenance, the Panama Canal has been
plagued by many and large slides, both fault related and not, which delayed con
struction and added millions of dollars to the cost. An excellent account of these
slides was given by MacDonald (1915). In the Culebra Cut, he says that the largest
and most important slides developed from structural breaks and deformations.
The rocks involved are weak, including massive, partly indurated, volcanic clays,
friable, bedded tuffs, and soft, brittle, slippery lignitic shales, all with a relatively
high ground-water content, and low cohesion and crushing strength. Rocks were
further weakened by faulting. Deforming movements caused much rupturing of
the weak rocks so that from a dense, fine-grained, relatively impervious condition
they locally became much sheared and subject to invasion by large quantities of
ground water. The movements once started sometimes lasted months, and in some
instances more than a year. These slides involved millions of cubic yards of material.
His “fault-zone” type of slide occurs in rocks strong enough to stand at a steep
slope except where large blocks were weakened by diagonal, canalward sloping
faults. Several large slides of this type dropped from 20,200 to 300,000 cubic yards
of rock into the cut.
ACTIVE FAULTS
As stated earlier, active faults are those which are undergoing movement now
or have moved in recent geologic or in historic time, and are liable to recurrent
movements in the future. The physical injury to rocks already described must be
discovered and dealt with whether the fault is active or dead. The active fault, how
ever, introduces an added hazard, in that it may move and cause rupture of the
ground with vertical or/and horizontal offset or other deformation, and may be
accompanied by earthquake shock. This is the dynamic aspect of faulting.

by guest
It is not always possible with a reasonable degree of certainty to determine whether

a given fault is active or not. However, where considerable capital is involved in a
structure underlain by a fault, or where damage to the structure would threaten
great damage to other property or loss of life, if the status of the fault in question
is in doubt, it is only reasonable precaution to act on the basis that it is active.
In the study of a fault as to its active-dead status, it is desirable to determine, if
possible, the diastrophic period and phase to which it owes its existence. Although
there is evidence that under certain conditions some faults which originated in an
earlier period have been reactivated, sometimes with a different type or direction
of movement in a later period of disturbance, most faults have an active life history
that is dependent on a local or regional phase of acute diastrophic activity which
is limited in time. For example, in the Tennessee River area, where many dam sites
have been geologically studied, numerous typical faults are thrusts genetically as
sociated with the late Paleozoic compression and folding of the region. Where they
are overlain by later sediments, as for example the terrace deposits, or where similar
faults of apparently the same age disappear under thousands of feet of Tertiary
and Cretaceous strata in central Alabama, they have produced no disturbance in
these later formations (Fox, 1940). Such faults may be confidently considered dead.
In some regions, certain groups of faults are considered dead, while others are
recognized as active. In the Great Basin both types exist. Among the former are a
number that cut Paleozoic or Mesozoic strata, but nowhere affect Tertiary or Qua
ternary formations that may overlie them. The active group cuts all formations,
including the Recent, with which they come in contact, have direct expression in
the topography, and a number have undergone movements accompanied by earth
quakes in recent historic time.
In some regions with many faults whose life spans have fallen within different
periods of diastrophic activity, it may be difficult to assign some particular fault,
which may be under study, to a definite status as to activity. Stratigraphic or other
appropriate evidence may be lacking.
Various methods have been used to determine whether or not a fault is active,
and geologists have not always agreed on the tests to be applied, and therefore have
sometimes come to different opinions concerning individual faults related to en
gineering works. In this connection, it may be recalled that on the fault map of
California, published by the Seismological Society of America, somewhat different
criteria were used for faults in the northern part of the State and those in the south
ern part.
Criteria that have been used for arriving at a judgment that a fault is active are
geological, historical, and seismological.
The best geological criterion is based on evidence of recent displacement along a
fault, especially if it is connected with a long series of displacements of which it is
the end member. If a fault shows evidence of repeated movements the last of which
cuts late Quaternary or Recent sediments, for example, it is not likely that it ceased
activity permanently at that point, and we must believe that future movements are
practically certain. The actual observable evidence includes such of the following

by guest
as may be available in the particular case: fresh or youthful scarps, offset streams
or alluvial fans, a line of horizontal depressions (not the effect of erosion) or sag
ponds, and deformed or displaced Recent deposits.
Historical evidence lies in the records of earthquakes whose descriptive accounts
permit reasonable reference to a particular fault, and especially records of actual
displacements of the ground along a fault line.
A seismographic method of learning what faults of a region are active is to de
termine the locations of centers of origin of recurring small earthquakes. These
earthquakes, although they cause no damage and may even be too small to be felt,
give definite records on seismographs and indicate that the faults along which they
are generated are unstable and subject to repeated adjustments, including possibly
large ones. Certain areas are now served by seismographic stations so located that
the epicenter (point on the earth’s surface immediately above the origin point) of
an individual earthquake may be determined by a form of seismographic triangu
lation from two or more stations at which the shock is recorded. Where faults are
close together this method may not be sufficiently accurate to distinguish between
them.
The most convincing occurrences are naturally those in which geological, histor
ical, and seismological types of evidence are all available and concordant for the
fault under consideration.
If a fault is known to be active or is considered probably active, especially if it
is known that in the past it has produced considerable displacements of the ground
accompanied by major earthquakes, it would seem to be only good judgment and
wise precaution to place no structure on or through it. The need of human society,
however, may leave no choice. Many roads, including important highways, streets,
and railroads, have to cross such faults. Essential water supplies for certain important
cities or districts can be made available only if their conduits or tunnels cross or
penetrate active faults. Dams may be badly needed for flood control or even water
storage in certain fault-line valleys of active faults. In many cases considerations
of convenience or economic pressure for land use have led to the erection of struc
tures where the physical necessity is not so apparent, and sometimes this has been
done in ignorance of the danger involved.
In any particular project, serious consideration should be given to the necessity
of the use of a site involving an active fault and to the possibility of an alternative
location. If there is no appropriate alternative and the needs demand the taking of
a “calculated” risk, plans should be based on the known characteristics of fault
activity in general, and in particular those of the fault at the site, insofar as can
be learned, and a consideration of their possible effects on the proposed structure.
Types of activities that have been observed along active fault lines will be outlined.
MOVEMENTS ALONG AN ACTIVE FAULT
GENERAL STATEMENT
Faults are usually represented on map or section by lines, and they may be simple
fractures, but more or less broad zones of shearing are often associated with faults
that have shown important activity. In the event of movement along a fault with

by guest
a broad shear zone, it might be expected that all parts of the zone would be affected
by the shear. In the ideal case of uniformly distributed displacement, a straight
line normal to the fault surface would, after the fault movement, be a straight line
inclined to the fault surface. Irregularities in the physical condition of the material
in the shear zone might be expected to produce a more or less curved or broken
line. At the time of the California earthquake of April 18, 1906, this, however, did
not occur. In the main, the greater part of the shear was limited to a remarkably
narrow zone. Along certain stretches the fault trace for miles gave the appearance
of a furrow turned up by a plow.
Apart from these fresh ruptures which can be seen for only a comparatively short
time after the actual fault movement, there are found in the midst or at the edge
of the broader fault zones lines or comparatively narrow zones of elongate depres
sions, sag ponds, scarps, offsets, etc., which indicate belts along which the more
recent fault movements have been concentrated.
Looking toward the future, the next important fault movement will probably
take place along an established zone that has suffered more recent movements.
This, however probable, cannot be guaranteed. The opening of new lines, as must
have occurred in the past, cannot be considered beyond the range of possibility,
although within a well-developed sheared belt with rather closely spaced zones of
rupture and shearing the opening of new lines in the immediate future, for which
plans are being made, would seem less likely than the use of established ones.
HORIZONTAL MOVEMENTS
The most important active faults in the coastal region of California give evidence
that the most persistent and characteristic type of movement has been an essentially
horizontal displacement along the strike. This was first definitely recognized at the
time of the 1906 earthquake when the fault movement along the San Andreas rift
produced offsets in roads, fences, and other linear features that lay across the fault
trace within the limits of recognizable fault activity. Later geological studies along
both the San Andreas and the Haywards fault lines have made this characteristic
evident by the frequent and repeated occurrence of offset streams, offset alluvial
cones, offset spur ends, furrows, sag ponds, and elongated depressions formed with
out differential vertical displacement of the country on opposite sides of the shear
zone.
Naturally, such types of evidence must be treated critically. The characteristic
furrows, sag ponds, and elongated depressions are best developed on the lower-
grade erosion or alluviated slopes, or in broad valley bottoms. On the steeper canyon
sides they may be destroyed or masked or prevented from forming by local step
faulting, landsliding, and so forth.
In stream valleys where the floor is silt, sand, or gravel-covered, and fully oc
cupied by the stream as a bed during flood time, such features are not to be expected.
In some, they can be followed down to the valley floor, where their direction indicates
that the fault passes under the floor, yet no fault-trace characteristics can be seen
on the valley floor itself. The absence of these features under the conditions men
tioned cannot be taken as indicative that recent movements have not taken place
or that the fault is inactive. Depressions or ponds caused by slumping, landsliding,

by guest
solution, or erosion should not, however, be confused with the phenomena produced
by the type of faulting under discussion.
In considering a possible offset alluvial cone, one should make certain that its
position, composition, structure, and size relate it to the stream from which it is
supposed to have been removed, and that it was not produced by some other former
stream since beheaded or otherwise put out of action.
Excellent examples of offset streams are frequently met, but their best develop
ment requires somewhat special conditions. The most favorable situation is where
a small consequent stream flows from the hills onto a broad valley or valley plain
with the active fault zone at or near the base of the hills. Under such circumstances
a large stream may maintain its general course unless a massive spur is offset across
its original debouchment. In a number of places where fault-line valleys occur in
the midst of a hill country, a canyon or a not very wide-bottomed valley has been
produced by a subsequent stream working on rock crushed and weakened along the
fault zone. Tributaries enter such valleys flowing in directions transverse to their
axes. But usually it is not possible to recognize definitely old offset courses that
may have crossed the fault zone and continued into the country beyond.
VERTICAL MOVEMENTS
Active faults with important vertical components of movement are usually more
easily recognized in the field than those whose movements are essentially horizontal,
because they produce scarps in the landscape. Such faults are common in the Great
Basin. Their recency is attested by the fact that they cut recent alluvial cones and
glacial deposits where available. In the weaker deposits, even where the movement
is of very recent origin, the scarp front may be crumbled down or degraded. When
they are associated with hard rock country the upper and older part of the scarp
may be eroded and therefore degraded, but the lower part will usually show a more
or less unmodified surface, with characteristic grooving and slickensiding. Along
some of the faults we have record of movement in recent historic time accompanied
by earthquake. A well-described example is that associated with the Pleasant Val
ley, Nevada, earthquake of October 2, 1915, when a fault along the base of the
Sonoma Range broke the surface along a 22-mile stretch and produced a fresh
scarp 5 to 15 feet high (Jones, 1915).
Some faults have moved with a combination of noteworthy horizontal and vertical
components, and even those that show movement dominantly horizontal or vertical
may have some movement along both components along parts of their course.
MAGNITUDE OF MOVEMENTS
It would be of great practical value if a probable magnitude, based on physical or
historical data, could be given for the maximum movement that might be reason
ably expected along a particular fault. At present it cannot be calculated on a theo
retical basis, and historical data are very meager. Records of definitely recognized
and measured fault displacements in connection with earthquakes do not go back
as much as a century in any part of the world. In the geological study of a fault
it is usually only possible to measure the summation of the effects of a series of

by guest
successive movements. It can hardly be expected that the movement during any
one event could be determined unless measurements were made within a compara
tively short time of its occurrence.
For various reasons measurements of individual fault movements are compara
tively few. Some of the important measurements of maximum movements are:
Maximum fault movements
Associated with Important Earthquakes
Year Place of occurrence Horizontal Comp, (ft.) Vertical (ft.
1872 California 18 23
1891 Japan 13 19
1897 Assam 35
1899 Alaska 47
1906 California 21 3
1906 Formosa 8 6
1915 Nevada 15
1931 New Zealand 6 9
In the California earthquake of 1906 the movement over most of the course of
the fault was horizontal; where the maximum horizontal movement occurred there
was no vertical component.
MOVEMENTS NORMAL TO THE FAULT TRACE
With the movements described above there is often combined a local compres-
sional or tensional effect transverse to the line of the fault.
Compressional effects were noted at a number of localities at the time of the 1906
disturbance along the San Andreas fault. One example may be given. The Upper
Crystal Springs Dam, subjected to an 8-foot horizontal offset at the time, carried
a roadway on its crest transversely across the fault zone. Board fences flanked the
roadway; so did a telephone line. The fences on both sides of the road were broken
and the boards buckled and shoved over one another, and the telphone wire cross
ing the lake sagged, showing that the movement brought the poles closer together.
Although no measurement was made, there was evidently a definite compression
normal to the fault line. A compressional effect which contracts or narrows the
country of the fault zone, within the limits of magnitude observed or suggested by
local phenomena, is not likely to cause any serious damage to an earth-fill dam;
but a spreading or tensional effect might lead to serious results.
There are historic records of fault fissures that remained open after the fault
movement and so continued until they were gradually filled by caving of the walls,
rain wash, and the like. The danger of such action to an earthem dam is that the
opening of the ground will cause a corresponding opening in the dam, permit the
entrance of water from the reservoir, and cause erosion along the fine of separation
with consequent widening of the opening, demolition of the dam, and destructive
flooding of the country downstream.
Another form that the spreading effect may take is the subsidence of a belt or
wedge of ground within the general shear zone. Many localities show evidence of
such action in the past. The depressions are usually elongated in the direction of
the trend of the fault zone and may vary in width from a few feet to many yards.
The practical effect of such action on a superimposed dam would be about the same

by guest
as a vertical component in a primary fault movement, except that in its simplest

form it develops a relatively sunken belt between two lines of vertical or steep
differential displacement.
EARTHQUAKES
Active faults which move discontinuously by a series of saltations or flings, with
any two of the series separated by a considerable time interval, are considered to
be operating under conditions where continuous fault movement is inhibited, nor
mally by frictional resistance, even though the diastrophic deformation of the region
may be progressing at a fairly steady rate. When the strain in the rock mass develops
sufficiently to overcome the resistance to fault movement, the elastically deformed
walls of the fault move in opposite directions to or toward a temporary position
of no strain (theory of elastic rebound). Then a similar strain proceeds to build up
again in preparation for the next fault movement. Each instance of fault movement
by such elastic rebound produces an earthquake. If the fault extends into the crust
of the earth to such a depth that the resistance to its movement, determined pri
marily by the strength of the rock and the high general pressure, permits the de
velopment of a great elastic strain before faulting takes place, the resulting
earthquake will be of high intensity (Louderback, 1942).
In the planning of a structure on or across an active fault, the possible effect of
such an earthquake should be taken into consideration, in addition to the possible
effects of permanent ground displacement caused by faulting, or of landslides pre
cipitated by the fault movement, or by the earthquake, or both.
Certain violent earthquakes, especially affecting broad alluviated areas, are in
ferred to have been accompanied by marked topographic changes resulting from
elevation and depression of considerable areas of land. One notable occurrence is
the Cutch earthquake of June 16, 1819, in India. An area of perhaps 2000 square
miles was depressed, and flooded by sea water, and not far away a 50-mile long
barrier was formed by the elevation or tilting of part of a plain and depression of
an adjoining part, producing a rise between them of about 21 feet (Baker, 1846;
Oldham, 1898). Another example is afforded by the New Madrid, Missouri, earth
quakes of December 10, 1911, January 23, and February 7, 1912. Among the various
permanent physical effects, the most striking was the sinking of large areas of the
bottom lands of the Mississippi, Little, and St. Francis rivers in amounts varying
from 2-3 feet up to 15-20 or more. Many of these areas became flooded, forming
lakes. One of the noteworthy is Reelfoot Lake, which is 20 to 25 miles long, 4-5
miles wide and reaches depths of 20-30 feet (Fuller, 1912).
It is not clear how such types of land changes could be predicted nor what protec
tive measures could be taken.
EXAMPLES OP STRUCTURES ON ACTIVE FAULTS
GENERAL STATEMENT
Although a number of structures have been placed on active fault zones, there
appear to be no good examples of serious effects on large structures of the permanent
displacements caused by faulting. As is well known, the transient disturbances

by guest
caused by fault movements—earthquakes and seismic sea waves (tsunami)—have

in the course of historic time destroyed countless thousands of buildings and other
structures, and taken hundreds of thousands of lives. But most of the faults re
sponsible for such destruction were located in mountainous regions or at the base
of scarps or under the sea, and did not furnish attractive or suitable sites for important
structures. The destroyed structures usually lay some miles, or in the case of tsu
nami possibly hundreds of miles, away. However, in the expansion of land occupancy
in some regions, and the extension of engineering projects, in recent years many
structures, including large and important ones, have been placed on active fault
zones. To date, their period of existence has not been long enough for them to have
been subjected to the full effects of a major fault displacement. In regard to such
structures plans at the present time must necessarily be based on inferred possibilities
of damage or destruction and not on the type of actual experience in connection
with numerous structures that have been subjected to earthquakes of high intensity.
In the Coast Range of California several dams have been erected in the fault-line
valleys of active faults. Some of these were built before the dynamic characteristics
of the local faults were recognized—that is, before 1906. Although these faults have
moved, no real demonstration has been given of the effects of a moving fault under
a dam with high reservoir level.
te m e sc a l da m
Over Temescal Creek in Oakland, Temescal Dam, the oldest of the group, was built
in 1868-1869. It is an earth fill, 80 feet high above stream bed. It overlies the Hay
wards fault, which in October 1868 actively moved and gave rise to a destructive
earthquake. As appears usual with long faults, only a part of the fault line was
displaced, and the actual rupture did not extend as far north as the dam site. No
surface displacement has been recognized on the fault since that time.
SAN ANDREAS DAM
South of San Francisco, the San Andreas Dam was an earth dam 97 feet high above
the stream bed at the time of the earthquake of April 18, 1906. The zone of fault
displacement passed through a knoll that formed the left abutment of the main dam,
and did not produce offset in the dam itself. A brick and cement tunnel from the
overflow wier, 7 to 8 feet in diameter, with walls 17 inches thick, passed through the
affected abutment and was cut in two by the fault movement, offset about 5 feet,
and badly crushed near its outlet end for about 28 feet.
UPPER CRYSTAL SPRINGS DAM
South of the San Andreas Dam in the same fault line valley is the Upper Crystal
Springs Dam. It is an earth-fill structure about 90 feet high. The fault movement of
1906 sheared the structure along a line 20 feet (downstream side) to 60 feet (upstream
side) from the east end of the embankment and caused a displacement of about 8
feet. Little was learned concerning the behavior of a dam affected by fault displace
ment because the structure was not acting as a dam at the time. In 1888 the Lower
Crystal Springs Dam, a gravity concrete dam, was completed on San Mateo Creek,
the original outlet of the fault-line valley. This dam is not in the fault-line valley

by guest
and is therefore not underlain by the fault. After its completion the Upper Crystal
Springs Dam was cut through, and the water surface, determined by the new dam,
was at the same level on both sides of the old dam, which since has been used only
to cany a roadway across the reservoir (Louderback, 1937).
COYOTE DAM
About 1936 a dam designed to rise 120 feet above stream bed was erected on Coyote
Creek about 25 miles southeast of San Jose, California. It was recognized that the
dam would cross a fault-line valley and be underlain by an active fault, and, as the
lower course of the stream passes through a fertile, cultivated valley and through
the eastern part of the city of San Jose, a serious question of safety as to life and
property was raised. A consulting board, appointed by the State Engineer, was of
the opinion that with proper precautions a safe dam could be built. The fault zone
is broad, extending across the valley bottom and out into the left abutment. There
was evidence in the approaches that the more recent fault movements were limited
to the valley bottom at the dam site, but a definite location for any future movement
was not determinable. It was the judgment of the board that the impervious member
that resists the water must be of such a type that total rupture would not take place
in the event of fault movement. This precluded the use of a rigid type of dam, a
concrete water face, or a core wall of concrete or narrow puddle. It advised rolled
earth with deep cutoff, an impervious member of generous width, so as to leave a
liberal width of impervious earth in contact with impervious earth even after a
large (20 foot) displacement; protection of the impervious member by rock and gravel
fills on both upstream and downstream sides, and a liberal freeboard in the light of
possible vertical displacement. To provide against possible opening of a fissure in
the dam at the time of faulting, protective fills of small broken rock were placed
against the impervious member, which could enter such opening and prevent erosion.
A spillway of large capacity was to be placed in the natural formation of the right
abutment which was free from gouge and crush zones. To prevent slumping or
sliding under earthquake influence, low slope angles for core and rock blankets were
recommended, and special treatment for the ground slope above the crest of the
spillway lining. Protection against wave wash from earthquake (or storms) was
provided by an upstream rock blanket. Special care for the location and treatment
of the outlet pipe, and the upstream location of operation valves, were recommended,
as well as deep excavation and backfilling of crush and gouge zones under the cutoff.
The foregoing qualitative outline of recommendations is given to show how the
various possible movements and induced disturbances of an active fault, inferred
from a careful study of the characteristics and history of the fault and of similar
faults of the region, were considered in the engineering treatment of a specific struc
ture. (For further detail, including quantitative values and references to construction,
see Louderback, 1937.)
MORRIS DAM
The ground conditions encountered at the Morris Dam site on the San Gabriel
River, Los Angeles County, California, were quite different from those dealt with
at the Coyote Dam. An exploratory tunnel across the valley below the stream deposits

by guest
discovered a fault cutting across the dam site along the west edge of the channel
floor. The fault, which dipped 55°-58° NW., was explored by a 262-foot drift. It
showed seams of gouge from 1 to several inches thick, and a crushed or brecciated
zone from 2 to 8 feet wide. The enclosing rock on both sides was a fresh, strong
dioritic gneiss. During the planning stage it was uncertain whether the fault was an
active fault or not. It cut ancient rocks, and the river floods might well have removed
any surface evidence of recent activity, as commonly happens along known active
faults. The fact that a number of active faults are recognized in the region, that an
epicenter of a minor earthquake had been located not far from the site, and that the
failure of a dam might cause great loss of life and property led to precautionary meas
ures being taken. Because of the general satisfactory foundation conditions, affected
by a narrow and definite fault zone, a concrete dam was erected, 245 feet high from
stream bed to roadway (328 feet above lowest point of foundation), and an ingenious
“open joint” constructed above the trace of the fault to allow free movement without
serious injury to the dam. When the river gravels were excavated to bedrock in the
course of dam construction, it was found that the deeper gravels had not been dis
turbed by the fault. The condition of the gravels led to the inference that they had
been there undisturbed for some thousands of years, and the possibility of future move
ment was slight. The open joint was, however, retained in the dam design (Cal
ifornia Division of Water Resources, 1931; Peugh, 1935).
RAILROAD TUNNEL
The railroad tunnel near Wright Station in the Santa Cruz Mountains, California,
was intersected by the movement along the San Andreas fault in 1906, on a part of
the fault line where displacement was about a quarter of the maximum observed.
In this 6200-foot tunnel a definite fault fracture was observed, making an angle of
80° with the trend of the tunnel, at 400 feet from the northeast portal. On this fracture
the tunnel was offset 5 feet. For the remaining 5800 feet the tunnel had not been
shifted by a simple movement of translation, but the departure from its original
alignment decreased somewhat irregularly from the 5-foot maximum to zero at a
distance of about 5100 feet. Subsidiary disturbances included caving of rock from
the roof, crushing in of upright timbers, and heaving of the rails. Some of the ties
were broken in the middle (Lawson, 1908, p. 111-113).
TUNNELS ALONG AQUEDUCT LINES
Tunnels along the main aqueduct lines, carrying water from the Sierra Nevada or
the Colorado River to the municipalities of the San Francisco Bay and the Los
Angeles regions of California, pass through known active faults. Such locations were
not chosen by preference but because of practical necessity. These tunnels are all
of fairly recent construction, and no significant fault movements have taken place
since their construction. In the event of fault movement the tunnel affected will
certainly be locally damaged and, if the offset is sufficient, will be entirely closed.
The provisions made to prevent serious distress in the populous areas served, if and
when such condition arises, consist in the development of terminal water storage
and in having part of the supply from another source or from the same source but
conveyed by a different line with the idea that both lines would not be affected si

by guest
ISO G. D. LOUDERBACK
multaneously. The purpose, of course, is to have available a sufficient supply to last

long enough to allow repair or reconstruction necessary to return the damaged aque
duct to use.
REFERENCES CITED
Baker, W. E. (1846) Remarks on the AUah-Bund and on the drainage of the eastern part of the Scinde
Basin, Bombay Geog. Soc., Tr., p. 186-188.
Berkey, Charles P., and Sanborn, James F. (1923) Engineering geology of the Catskill water supply,
Am. Soc. Civil Eng., Tr., vol. 86, p. 1-91.
Bromehead, C. E. N. (1938) Geology of reservoir-dam sites, 2d Congress on Large Dams, Tr., vol. 4,
p. 113-119.
California, Division of Water Resources (1931) Report of the consulting hoard on the safety of proposed
Pine Canyon dam, Los Angeles County, California, 22 pages.
Engineering News-Record (1939) Mining under Seminole Dam, vol. 122, p. 490-492.
------------(1941) Passing faults on the Delaware Aqueduct, vol. 127, p. 150-154.
Fox, Portland P. (1940) Geology of Chicamauga Dam, T.V.A., Geol. Div., Tech. Mon. 47, p. 147-193.
Fuller, Myron L. (1912) The New Madrid earthquake, U. S. Geol. Survey, Bull. 494.
Frink, John W. (1946) The foundation of Hales Bar Dam, Econ. Geol., vol. 41, p. 576-597.
Jones, J. Claude (1915) The Pleasant Valley, Nevada, earthquake of October 2,1915, Seismol. Soc. Am.,
Bull., vol. 5, p. 190-205.
Keener, Kenneth B. (1937) Seminole Dam I. Site exploration and design studies, Eng. News-Record,
vol. 118, p. 395-399.
Lawson, Andrew C. et al. (1908) The California earthquake of April 18, 1906, Carnegie Inst. Wash
ington, Pub. no. 187, vol. 1. .
Leadbetter, B. C. (1938) Driving an extremely difficult tunnel, Eng. News-Record, vol. 121, p. 669
672.
Louderback, George D. (1937) Characteristics of active faults in the central Coast Ranges of California,
with application to the safety of dams, Seismol. Soc. Am., Bull., vol. 27, p. 1-27.
------------(1942) Faults and earthquakes, Seismol. Soc. Am., Bull., vol. 32, p. 305-330.
MacDonald, Donald F. (1915) Some engineering problems of the Panama Canal in their relation to
geology and topography, U. S. Bur. Mines, Bull. 86, 82 pages.
Oldham, R. D. (1898) A note on the AUarBund in the north-west of the Rann of Kuchh, Geol. Survey
India, Mem., vol. 28, p. 27-30.
Peugh, Verne L. (1935) Technical report on the construction of Morris Dam, Pasadena Water Dept.,
218 pages, 72 plates.
Warner, J. H. (1937) Seminole Dam progress, Eng. News-Record, vol. 119, p. 817-822.
Williams, Charles P. (1932) Foundation treatment at Rodrigues Dam, Am. Soc. Civil Eng. Pr., vol.
58, p. 1375-1385.
------------(1933) Unusual features of the Rodrigues Dam construction, Am. Water Works Assoc., Jour.,
vol. 25, p. 355-366.

by guest

Faults and Engineering Geology

Uploaded by

Copyright:

Available Formats

Faults and Engineering Geology

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Faults and Engineering Geology

Uploaded by

Copyright:

Available Formats

TH E GEOLOGICAL SOCIETY OF AMERICA

ENGINEERING GEOLOGY (BERKEY) VOLUM E NOVEMBER 1860

FAULTS AND ENGINEERING GEOLOGY

F a u lts a re ru p tu re s in ro c k m asses th a t a re a cco m p a n ie d b y d iffe re n tia l d is p la c e ­

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

of faulting; if the fault is judged to be active, the dynamic aspect is superimposed—

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

entirely by smooth, slickensided surfaces. In my notes I have come to refer to such

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Conditions of strain materially aided by differential movement which introduces

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

seriously affected by faulting, often because of considerations of safety, sometimes

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

It is not always possible with a reasonable degree of certainty to determine whether

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

as a vertical component in a primary fault movement, except that in its simplest

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

caused by fault movements—earthquakes and seismic sea waves (tsunami)—have

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

multaneously. The purpose, of course, is to have available a sufficient supply to last

Downloaded from https://pubs.geoscienceworld.org/books/chapter-pdf/3744726/9780813759418_ch06.pdf

You might also like

F a u lts a re ru p tu re s in ro c k m asses th a t a re a cco m p a n ie d b y d iffe re n tia l d is p la c e