Sunteți pe pagina 1din 9

C A S E S T U D Y OF DEBRIS FLOW

By Robert W. Day, ~ Fellow, ASCE

ABSTRACT: A debris flow is defined as soil with entrained water and air that
Downloaded from ascelibrary.org by GADJAH MADA UNIVERSITY on 12/04/14. Copyright ASCE. For personal use only; all rights reserved.

moves readily as a fluid on low slopes. A case history of a debris flow adjacent to
the Pauma Indian Reservationis described. Two contributingfactors in the debris
flow were the cutting down of trees and the constructionof haul roads. The debris
flow travelled a long way (370 m) in the depositionalarea before strikinga house.
Fortunately,by the time the debris flow reached the house, the energy was nearly
spent and there was no damage to the structural frame. The material involvedin
the debris flow was classifiedas silty sand. Results of laboratory tests,indicate that
this type of soil is susceptible to debris flow. There is a range of coefficient of
permeabilityfrom about 10-2 to 10-6 cm/s that corresponds to soils that have an
optimum soil permeabilityfor long-distanceflow in the depositional area. Based
on the historicalmethod, it is probable that there will be future debris flow in this
area.

INTRODUCTION

Debris flows cause a t r e m e n d o u s a m o u n t of damage and loss of life


throughout the world. A n example is the loss of 6,000 lives from devastating
flows that occurred in Leyte, Philippines, on N o v e m b e r 5, 1991, due to
deforestation and torrential rains from tropical storm Thelma. D u e to con-
tinued population growth, deforestation, and poor land-form practices, it
is expected that debris flows will increase in frequency and devastation.
Debris flow is defined in this paper as soil with entrained water and air
that moves readily as a fluid on low slopes. Debris flows can include a wide
variety of soil-particle sizes (including boulders) as well as logs, branches,
tires, and automobiles. Other terms, such as mudflow, debris slide, m u d
slide, and earth flow, have been used to identify similar processes. While
categorizing flows based on rate of m o v e m e n t or the percentage of clay
particles may be important, the mechanisms of all these flows are essentially
the same (Johnson and Rodine 1984).
The historical method is one means of predicting debris-flow activity in
a particular area. For example, as Johnson and Rodine (1984) indicated,
many alluvial fans in southern California contain previous debris-flow de-
posits, which in the future will again experience debris flows. However,
using the historical method for predicting debris flow is not always reliable.
For example, the residences of Los Altos Hills experienced an unexpected
debris flow mobilized from a road fill after several days of intense rainfall
(Johnson and H a m p t o n 1969). Using the historical method to predict debris
flow is not always reliable, because the area can be changed, especially by
society's activities.
Johnson and Rodine (1984) stated that a single parameter should not be
used to predict either the potential or actual initiation of a debris flow.
Several parameters appear to be of prime importance. Two such parameters,
which numerous investigators have studied, are rainfall a m o u n t and rainfall

~Chf. Engr., Am. Geotech., 5764 Pacific Center Blvd., Ste. 112, San Diego, CA
92121.
Note. Discussion open until January 1, 1995. To extend the closing date one
month, a written request must be filed with the ASCE Manager of Journals. The
manuscript for this paper was submitted for review and possible publication on
October 13, 1992. This paper is part of the Journal of Performance of Constructed
Facilities, Vol. 8, No. 3, August, 1994. 9 ISSN 0887-3828/94/0003-0192/$2.00
+ $.25 per page. Paper No. 4436.

192

J. Perform. Constr. Facil. 1994.8:192-200.


intensity. For example, Neary and Swift (1987) stated that hourly rainfall
intensity of 90 to 100 mm/h was the key to triggering debris flows in the
southern Appalachians. Other important factors include the type and thick-
ness of soil in the source area, the steepness and length of the slope in the
source area, the destruction of vegetation due to fire or logging, and other
Downloaded from ascelibrary.org by GADJAH MADA UNIVERSITY on 12/04/14. Copyright ASCE. For personal use only; all rights reserved.

society-induced factors such as the cutting of roads.


There are generally three segments of a debris flow: the source area,
main track, and depositional area (Baldwin et al. 1987). The source area is
the region where a soil mass becomes detached and transforms itself into a
debris flow. The main track is the path over which the debris flow descends
the slope and increases in velocity depending on slope steepness, obstruc-
tions, channel configuration, and the viscosity of the flowing mass. When
the debris flow encounters a marked decrease in slope gradient and depo-
sition begins, this is called the depositional area.
The purpose of this paper is to describe a case history of a debris flow
and investigate the variables involved in the flow.

CASE HISTORY
The best method to understand a debris flow is to study actual failures
such as the case history described here. The location of the debris flow was
adjacent to the Pauma Indian Reservation, in San Diego County, California.
The debris flow occurred in January of 1980, during heavy and intense winter
rains. A review of aerial photographs indicates that alluvial fans are being
built at the mouths of the canyons in this area due to past debris flows.
The debris flow in January of 1980 hit a house, which resulted in a lawsuit
being filed. The writer was retained in June of 1990 as an expert for one
of the cross defendants, a lumber company. The case settled out of court
in July, 1990.
The lumber company had cut down trees on the Pauma Indian Reser-
vation and it was alleged that the construction of haul roads and the lack
of tree cover contributed to the debris flow. These factors were observed
to be contributing factors in the debris flow. For example, after the debris
flow, deep erosional channels were discovered in the haul roads, indicating
that at least a portion of the soil in the flow came from the road subgrade.
Fig. 1 presents a topographic map of the area, showing the location of
the house and the path of the debris flow. The debris flow traveled down
a narrow canyon (top of Fig. 1) that had an estimated drainage basin of 0.8
km 2 (200 acres). The complete extent of the source area could not be
determined because of restricted access. The source area and main track
probably extended from an elevation of about 980 m (3,200 ft) to 400 m
(1,300 ft) and the slope inclination of the canyon varied from about 34 to
20~ At an elevation of about 400 m (1,300 ft), the slope inclination changes
to about 7~ corresponding to the beginning of the depositional area.
Fig. 2 presents a photograph of the house, which is a one-story single-
family structure, having a typical wood-frame construction and a stucco
exterior. It was observed in the area of the site that the debris flow was
uniform with a thickness of about 0.6 m (2 ft). There was no damage to the
structural frame of the house due to two factors. The first factor was that
the debris flow impacted the house and moved around the sides of the
house, rather than through the house, because of the lack of any opening
on the impact side of the house (Fig. 2). The second factor was that the
debris flow had travelled about 370 m (1,200 ft) in the depositional area
before striking the house and then proceeded only about 15 m (50 ft) beyond
193

J. Perform. Constr. Facil. 1994.8:192-200.


Downloaded from ascelibrary.org by GADJAH MADA UNIVERSITY on 12/04/14. Copyright ASCE. For personal use only; all rights reserved.

FIG. 1. Debris Flow

the house. By the time the debris flow reached the house, its energy probably
had been nearly spent.
A sample of the debris-flow material was tested for its grain-size distri-
bution. The soil comprising the debris flow was classified as a nonplastic
silty sand (SM), with 16% gravel, 69% sand, and 15% silt and clay. The
coefficient of permeability of the soil is estimated to be 10 .3 cm/s. The
material in the debris flow was derived from weathering of rock from the
source area. Geologic maps of the area indicate that the water shed is
composed of metamorphic rocks, probably the Julian Schist (Zellmer, per-
sonal communication, 1990).
Based on the historical method, for the site shown in Fig. 1, it seems
probable that there will be future debris flows. Perhaps the best solution is
to restrict the construction of houses in the depositional area. Notice in the
center of Fig. 1 that two houses were built very close to the path of the
194

J. Perform. Constr. Facil. 1994.8:192-200.


Downloaded from ascelibrary.org by GADJAH MADA UNIVERSITY on 12/04/14. Copyright ASCE. For personal use only; all rights reserved.

FIG. 2. Photograph of Site

debris flow, but they are situated at an elevation about 15 m (50 ft) above
the depositional area and, hence, were unaffected by this debris flow.

EXPERIMENTAL MODELING
It has been observed at many debris-flow sites that, in general, the finest
soil particles travel the farthest from the source area. As shown in Fig. 1,
when the debris flow reached the depositional area, it had travelled about
370 m (1,200 ft), on a 7 ~ slope. The purpose of the laboratory testing was
to investigate the type of soil that would be most likely to travel the farthest
in the depositional area and to determine a correlation between the perme-
ability of the soil and the distance travelled in the depositional area. Distance
travelled in the depositional area is important because the further the debris
flow travels, the greater is the potential for damage.
The laboratory-test set-up consisted of a cylinder, having a diameter of
6.4 cm (2.5 in.) and a length of 48 cm (19 in.), that was suspended 7.6 cm
(3 in.) above a 0.3-m (12 in.) square plate, which was tilted 23 ~ At the
start of the experiment, the bottom of the cylinder was sealed (using a
rubber stopper) and 500 grams of distilled water was poured into the top
of the cylinder. Then dry soil was slowly poured into the top of the cylinder.
The amount of soil poured into the cylinder was just enough so that all the
soil was submerged and all the water occupied the soil voids. In essence,
this process created a saturated to nearly saturated, very loose or normally
consolidated soil within the cylinder. The rubber stopper was then pulled
out of the bottom of the cylinder so that the soil could flow out the cylinder,
down the tilted plate, and out onto a level surface. In terms of modeling,
the soil within the cylinder represents the source area material, the vertical
drop of the soil out of the cylinder and the flow down the tilted plate (23 ~)
are the main track, and the depositional area is the level surface.
Four different types of soils were used in the experiments. The grain-size
distribution curves are shown in Fig. 3. The first soil is an angular gravel
(GP) with all particles between 12 m m and 25 m m in size. The second soil
is a uniform, nonplastic, sand (SP). The third soil is a silt (ML) having a
liquid limit of 30 and a plasticity index of 5. The fourth soil is clay (CH),
consisting of commercially available montmorillonite, having a liquid limit
of 460 and a plasticity index of 430.
195

J. Perform. Constr. Facil. 1994.8:192-200.


Downloaded from ascelibrary.org by GADJAH MADA UNIVERSITY on 12/04/14. Copyright ASCE. For personal use only; all rights reserved.

8AND , j
81LT CLAY~
J O.AVEL JCOANsEI MEDIUM J FINE
81EVE $1ZES-U.8, STANDARD
J
3 / 4 " 112" II4 '~ 4 8 10 18 20 30 40 5 0 8 0 100 200
100 100

DO 90

80 80

lO 70

80

Co z
O~ 50 50 "~

40 4O z

30 30

J. Perform. Constr. Facil. 1994.8:192-200.


20 20
GRAVEL
1 SAND
10 10

0
IO.0 1,O 0.1 .01 .OOI
PARTICLE SIZE-MILLIMETER8

FIG. 3. Grain-Size Distribution Curves


Downloaded from ascelibrary.org by GADJAH MADA UNIVERSITY on 12/04/14. Copyright ASCE. For personal use only; all rights reserved.

TABLE 1. Data Summary


Distance
Water Did water Soil that traveled on Coefficient of
Test Dry density content Degree of drain out of flowed level surface permeability
number Soil type (pcf) (%) saturation soil? (%) (cm) (cm/s)
(1) (2) (3) (4) (5) (6) (7) (8) (9)
la 72.3 44.2 91 Yes 0 lO (E)
lb Gravel 74.2 44.5 96 Yes 0
2a 84.3 35.7 98 Yes 10-20 0 2.8 x 10 z ( C )
2b Sand 84.7 35.3 98 Yes 10-20 0.8
~o 3 Silt 50.6 77.6 91 No 100 26.4 6.6 10 -5 (F)
4 Clay 13.9 407 99 No 0 1 x 10 " (E)
b 2.9 10 -4 (F)
5 6% clay, 94% sand 63.9 59.9 100 No 70-80
6 6% clay, 94% silt 46.6 93.8 98 No 100 15.0 7.2 x 10 6 ( F )
a
7 75% sand, 25% silt 88.6 32.2 98 100 9.7 9.8 x 10 -4 (C)
8 50% sand, 50% silt 78.7 40.7 98 No 100 19.8 1.2 x 10 4 ( F )
9 25% sand, 75% silt 67.6 51.6 95 No 100 30.5 9.5 x 10 5 ( F )

J. Perform. Constr. Facil. 1994.8:192-200.


"A small a m o u n t of water did drain out of soil.
bFiow did not reach level surface.
Note: Test n u m b e r s 5 - 9 are soil mixtures. In c o l u m n (9), (C) = constant head p e r m e a m e t e r ; (F) = falling h e a d p e r m e a m e t e r ; a n d (E) = e s t i m a t e d
coefficient of permeability.
Table 1 summarizes the results of the experiments using the laboratory
set-up and the tests are individually discussed as follow:

9 Test number 1 - - T h i s test was performed on the angular gravel.


Downloaded from ascelibrary.org by GADJAH MADA UNIVERSITY on 12/04/14. Copyright ASCE. For personal use only; all rights reserved.

Although some gravel particles rolled down the tilted plate, the
behavior cannot be described as a flow. Rather, the water simply
flowed freely through the gravel pores due to the very high perme-
ability. Two tests were performed on the gravel (tests la and lb)
to check the reproducibility of results.
9 Test number 2 - - T h i s test was performed on the uniform sand. A
portion of the sand did move, although, once again, the water flowed
freely through the pores between the sand particles. The water
exiting the sand pores helped to mobilize the part of the sand that
flowed. Two tests were run on the sand (tests 2a and 2b) to check
the reproducibility of results.
9 Test number 3 - - T h i s test was performed on the silt. For this test,
the silt readily flowed from the cylinder, down the tilted plate, and
out onto the level surface (depositional area).
9 Test number 4 - - T h i s test was performed on the soil having mont-
morillonite clay particles. For this situation, at a moisture content
of 407%, the clay would not even come out of the cylinder. The
clay was much too plastic and it would not flow. Field observations
also indicate that soils having high clay contents are not susceptible
to debris flows. For example, Ellen and Fleming (1987) indicated
that soils with a clay content over 35% were not susceptible to debris
flows.
9 Test numbers 5 and 6 - - T h e next step in the laboratory testing was
to use mixtures of different proportions of the four main soil types.
Test number 5 was performed on a mixed sample of 6% montmo-
rillonite and 94% sand. Even with only 6% montmorillonite, the
soil would not flow off the tilted plate. Test 6 was performed on a
mixed sample of 6% montmorillonite and 94% silt. The addition of
6% montmorillonite to the silt reduced the flow distance by about
one half. Both test numbers 5 and 6 show that the presence of
montmorillonite greatly reduces the ability of soil to flow.
9 Test numbers 7 - 9 - - T h e s e tests were performed with different mix-
tures of sand and silt. All of these soils experienced debris-flow
characteristics.

Fig. 4 presents a plot of the distance the soil flows in the depositional
area [Table 1 column (8)] versus the coefficient of permeability [Table 1
column (9)] for the silt or silt mixtures. The coefficient of permeability
plotted in Fig. 4 was determined from either a constant-head or falling-head
permeameter. Fig. 4 shows that there is an optimum soil permeability that
favors a debris flow. If the permeability is too high (greater than 10 -2 cm/
s), then the water flows freely between the soil particles. A low permeability
(less than 10 -6 cm/s) indicates the presence of clay particles, which inhibits
flow due to its plasticity and reduces the ability of the soil to absorb water.
There is a range of permeabilities, from about 10 -2 to 10 -6 cm/s, that
corresponds to soils such as very fine sands, inorganic silts, and mixtures
of sands, silts, and clays. The permeability of these soils is more conducive
198

J. Perform. Constr. Facil. 1994.8:192-200.


35 | .
I Ir 1F - i
I i | I * I
Downloaded from ascelibrary.org by GADJAH MADA UNIVERSITY on 12/04/14. Copyright ASCE. For personal use only; all rights reserved.

30

25 SILT OR
SILT MIXTURES

i,o
z

E
?
m
UJ
C3

f =-I I z _ I I ~_ I - I I
I ~,
1.0 10- 2 10- 4 10- 6

COEFFICIENT OF PERMEABILITY ( c m / s e c )
FIG. 4. Debris-Flow Distance versus Coefficient of Permeability for Laboratory
Tests

to a great distance of flow in the depositional area. These results are sum-
marized as follow:

Coefficient o f P e r m e a b i l i t y f r o m 10 2 to 10 - 2 c m / s
Types of soil are clean gravel, clean sands, and clean sand and gravel
mixtures. The soil is too permeable for a debris flow, because the water
flows rapidly through the soil pores. If a debris flows occurs, deposition
most likely will not be far from the source area.

Coefficient o f P e r m e a b i l i t y f r o m 10 - 2 to 10 - ~ c m / s
Types of soil are very fine sands and inorganic silts, mixtures of sand,
silt, and clay. The soil condition is favorable for a debris flow. The debris
flow can travel great distances.
199

J. Perform. Constr. Facil. 1994.8:192-200.


Coefficient of Permeability from 10 -6 to 1 0 - 9 cm/s
The types of soils are sandy clays, silty clays, and clay. The soil is too
impermeable, and thus it is difficult for clays to quickly absorb water. Clays
also have plasticity. It is unlikely that a debris flow will develop.
Downloaded from ascelibrary.org by GADJAH MADA UNIVERSITY on 12/04/14. Copyright ASCE. For personal use only; all rights reserved.

SUMMARY
A case history of a debris flow adjacent to the Pauma Indian Reservation
has been described. Two contributing factors in the debris flow were the
cutting down of trees and the construction of dirt-haul roads. The debris
flow travelled a long way (370 m) in the depositional area before striking
the house. The material involved in the debris flow was classified as a silty
sand. Laboratory tests show that this type of soil is susceptible to long-
distance flow in the depositional area. There is a range of soil permeability,
from about 10 -2 to 10 -6 cm/s, that is more conducive to a long-distance
flow in the depositional area.
At the Pauma Indian Reservation, alluvial fans are being built at the
mouths of the canyons due to past debris flow. Even if logging is restricted
and haul roads eliminated, it is probable that there will be future debris
flow. Perhaps the best solution is to restrict the construction of houses in
the alluvial fan depositional area.

APPENDIX. REFERENCES
Baldwin, J. E., Donley, H. F., and Howard, T. R. (1987). "On debris flow/avalanche
mitigation and control, San Francisco Bay area, California." Debris Flows~ava-
lanches: process, recognition, and mitigation, The Geological Society of America,
Boulder, Colo., 223-226.
Ellen, S. D., and Fleming, R. W. (1987). "Mobilization of debris flows from soil
slips, San Francisco Bay region, California." Debris flows~avalanches: process,
recognition, and mitigation, The Geological Society of America, Boulder, Colo.,
31-40.
Johnson, A. M., and Hampton, M. A. (1969). "Subaerial and subaqueous flow of
slurries." Final Rep., U.S. Geological Survey (USGS) Contract No. 14-08-0001-
10884, USGS, Boulder, Colo.
Johnson, A. M., and Rodine, J. R. (1984). "Debris flow." Slope instability, John
Wiley & Sons, New York, N.Y., 257-361.
Neary, D. G., and Swift, L. W. (1987). "Rainfall thresholds for triggering a debris
avalanching event in the southern Appalachian Mountains." Debris flows~ava-
lanches: process, recognition, and mitigation, The Geological Society of America,
Boulder, Colo., 81-92.

200

J. Perform. Constr. Facil. 1994.8:192-200.

S-ar putea să vă placă și