Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 195
Working Paper F Liquefaction and Landslides LIQUEFACTION As applied to seismic problems, liquefaction has become a catch- all word referring to various types of earthquake-caused failures of saturated cohesionless soils. Four different Manifestions of liquefac- tion have been identified (National Research Council, 1985~: 1. Flow slides from slopes. 2. Loss of foundation bearing capacity, leading to large settle- ment author tilting of structures. 3. Lateral spreading, that is, a movement of gradually sloping ground toward low points. 4. Ground oscillation, where ground overlying saturated sand breaks up into jostling "plates. All of these phenomena may be accompanied by sand boils small volcanoe-like mounds or craterlets from which sand and water spurt to the surface. The first two manifestations of liquefaction are dramatic but less common. When they do occur, there is considerable potential for damage and, in the case of flow slides, for loss of life. Flow slides may occur in natural ground, but are also likely in man-made deposits, such as earth dams, mine tailings darns, and fill placed behind waterfront retaining structures. 195

OCR for page 195
196 The remaining manifestations are less spectacular but much more common. Lateral spreading frequently disrupts pipelines, roads, railways, and canals, and if occurring beneath a structure, can cause extensive damage and even loss of life. Ground oscillation and associated sand boils can present an enormous clean-up problem if they occur in a built-up area. If accompanied by ground settlement, damage and disruption can also occur. All aspects of seismic liquefaction have been reviewed and dis- cussed in a major report (National Research Council, 1986~. It is important, for the subsequent discussion, to distinguish two situa- tions: . Level ground where no shear stresses are required for equilib- rium following an earthquake. . Slopes (which include building foundations) where shear stresses are required for static equilibrium. Ground with a very gentle slope (< 5°) may, depending on the circumstances, fall into either situation. Liquefaction Susceptibility A range of criteria and methods exists for evaluating the sus- ceptibility of a soil to liquefaction as a result of earthquake ground shaking. The simplest method considers just two factors: the geo- logic age of the deposit and the depth to the water table. Table F-1 presents such a set of criteria from Youd et al. (1978~. Other exam- ples appear in ATC-13 (Applied Technology Council, 1985~. These ratings are based on observations and experience during actual earth- quakes, and rate the susceptibility of a soil deposit as a whole. Only portions of a deposit would actually experience liquefaction. In Table F-1, latest Holocene refers to the most recent 1,000 years, with the earlier Holocene extending back to 10,000 years. Experience suggests that deposits older than about 130,000 years will not liquefy. As indicated, the depth of the water table is also a very important factor. The information in Table F-1 is directly useful for preparing liquefaction hazard maps. A procedure for combining this information with the expected ground-shaking hazard is described by Youd and Perkins (1978~. A more quantitative method for assessing liquefaction suscepti- bility makes use of penetration resistance as measured by the Stan- dard Penetration Test (SPT). In Figure F-1, the horizontal axis is the blow count in the SPT, corrected for the depth at which the blow

OCR for page 195
197 TABLE F-1 Considerations Used in Producing a Map of Liquefaction Susceptibility in the San Fernando Valley Deoth to Groundwater (ft) Age of Deposit 0-10 10-30 < 30 Latest Holocene Earlier Holocene Late Pleistocene High Lowa Nil Moderate Low Nil Low Nil Nil aLatest Holocene deposits in this basin generally are not more than 10-ft thick. Saturated deposits in the 10- to 30-ft intermural are earlier Holocene sediments. SOURCE: Youd et al. (1978~. count is recorded and the energy delivered to the drib rods when per- form~ng the test. The vertical axis is the ratio of the dynamic stress occurring during an earthquake to the vertical elective overburden stress in the soil. The dynamic stress is commonly computed from a simple expression involving the peak acceleration at ground surface and the unit weight of the soil. The data points on the plot represent actual observations during earthquakes, and a curve has been drawn separating cases of liquefaction from those where no liquefaction was observed. If a new situation is represented by a point plotting above this curve, liquefaction is to be expected. The data in Figure F-1 apply for an earthquake with a mag- nitude of 6.5. Corresponding curves have been developed for other magnitudes (see Figure F-2~: the larger the magnitude, the greater the duration of shaking and hence the greater the susceptibility to liquefaction for a given (Nl)60 and Tab/ a'. These figures apply for clean sands; relations for taking into account the influence of fines have also been developed. It is unlikely that a program of penetration tests would be under- taken in connection with a large-scale loss estimation study. However, data from previously drilled borings can be used to evaluate the liq- uefaction susceptibility of deposits in a study area and thus serve as a basis for preparing liquefaction hazard maps. Other and more sophisticated methods for evaluating liquefac- tion susceptibility have also been developed. There are more precise techniques for measuring penetration resistance, such as the Cone Penetration Test (CPT). If very good undisturbed samples can be obtained, various types of laboratory tests can be done. Theoretical

OCR for page 195
198 06 0.5 0.4 T~V a' 0.3 0.2 0.1 o . _ ~ / · ~ a I'1 1 · ~ I · ~ + a / I o / 0 · ~ a so ~ 0 .~ ~ / He o o , · ~ Pon Time riccn data Jcpanese data Chinese date too Pi N ES CCl~J7E>J ~ 5 5 % Ctl~rese 3ulIding Code (clay ccotent-O) ver-ir.al No L`=efccti~ L-ue~c.'cn Ocuef~:t~on 1 ~ ~ ~ 0 0 0 ! 0 10 20 30 (~1)60 40 50 FIGURE F-1 Relationship between stress ratios causing liquefaction and (N. )60 values for clean sands for magnitude 7.5 earthquakes. Source: Seed et al. (1984~. methods are also available. While these techniques are of value for evaluating specific sites or particular earth structures (e.g., earth dams), they are not appropriate for large-scale loss estimation stud- ies.

OCR for page 195
05 8 O 04 .o = cr v' Cal — c ° . _ g 5 . _ 0 0 _ _ 0 (,0 c - 't 0 3 b' c `~ 0 2 ~ . _ In ~ O ~ J · _ — _ ~ — lo. — A) ~ o 199 I I I / I / I .' ; ' / I J I l l l l _ _ . _ 1.1 _ ~ I ~', / / / / / ,, 1 1 1 10 , . ~ / 1 ~11 ~ ~ / go/' `\~/ V' ~~ hi/ ' ,' / - 20 30 Modified Penetration Resistance, Nl -blows/ft 40 FIGURE F-2 Chart for evaluation of liquefaction potential of sands for earth- quakes of different magnitudes. Source: Seed and Idriss (1982~. Consequences of Liquefaction Methods for evaluating liquefaction susceptibility are essentially determunistic in nature, and do not indicate directly how likely liquefaction might be during an event of given intensity nor how widespread liquefaction might be over a given deposit. Furthermore, the methods are based heavily on observations as to the occurrence

OCR for page 195
200 or nonoccurrence of some manifestation of liquefaction, without ref- erence to the severity of the occurrences. Indeed, it is possible, even likely, that liquefaction actually occurred beneath the surface in some of the cases identified as "no liquefaction," but these liquefactions did not appear at the surface of the ground. Ishihara (1985) has shown that the thicknesses of a liquefying layer and of an overlying nonliq- uefiable layer both affect the likelihood that liquefaction is observer! at the surface; Figure F-3 provides initial guidance in this matter. The ATC-13 report gives a ground probability failure matrix, reproduced here as Table F-2, based on expert opinion. The matrix obviously is oriented to situations in California, but for comparable soils should also apply elsewhere. Liso et al. (1988) performed a detailed statistical analysis of the case studies upon which Figure F-1 is based. It was concluded that the boundary curve in Figure F-1 might correspond to about 50 percent probability of liquefac- tion. This study also provided curves for estimating the probability of liquefaction for a point falling at any point of a raV/a' versus (N~60 diagram. However, these several results still do not get at the questions of how widespread and damaging liquefaction may be for a given deposit. In the ATC-13 report, some very scant data are cited to the effect that damage to buildings on poor ground (such as liquefiable sand) is 5 to 10 times greater than damage to buildings on firm ground, for the same intensity of ground motion. Thus, for facilities on the surface, the ATC report proposes to evaluate a mean damage ratio (MDR) as: MDRgrour~d = MDRfirm ground X PtL] X 5, where PtL] is the probability of liquefaction for the deposit of interest. For buried structures (e.g., pipelines), the ATC report proposes using a factor of 10. Youd and Perkins (1987) introduce the concept of a liquefaction severity index (LSI). They relate LSI to the extent and magnitude of movements and other manifestations of liquefaction that can be expected; their descriptions are reproduced in Table F-3. They also propose an equation relating LST to the magnitude and epicentral distance for an earthquake. However, this equation is applicable only for late Holocene floodplains and deltas associated with rivers having channel widths greater than 10 meters and for seismic conditions in California and Alaska. Thus the method is not directly applicable to other parts of the country. In addition, the method still leaves

OCR for page 195
201 12 1 1 10 Cal I , 8 CO a) Ct ~ _ :D Or ._ o in CO a) 4 By ._ 6 _ s 3 2 1 Max. ace. _ 200 gal ~- c ._ ~ 1 AS c ,o ~ ' 0 a, ~ J _ _ 0 1 - Max. ace. . 300 gal ., / \ I Max. ace. 1 400-500 gal / - - ~/ .. .. . I I I l i I 2 3 4 5 6 7 8 9 Thickness of surface layer, H 1 (m) 10 FIGURE F-3 Proposed boundary curares for site identification of liquefaction- induced damage. Source: Ishihara (1985~. the problem of relating LSI to quantitative measures of damage to facilities e LANDSLIDES Earthquake-~nctuced landslides have caused tens of thousands of

OCR for page 195
202 TABLE F-2 Ground Failure Probability Matrix for Poor Ground (in percents Zone Type of Deposit Probability of Ground Failure by MMI VI VII VIII IX X XI : XII la Stream channel, tidal channel 5 20 40 60 80 100 100 lb San Francisco Bay mud and fill over bay mud 3 15 30 40 60 80 90 2a Holocene Alluvium, water table shallower than 3 m (10 ft) 2 10 20 30 40 60 80 2b Holocene Alluvium, water table deeper than 3 m ([Oft) 0.5 2 5 7 12 25 40 3 Late Pleistocene Alluvium 0.1 0.5 1 2 4 7 10 aEstimates are based on consensus of the ATC-13 Project Engineering Panel. SOURCE: Applied Technology Council (1985). deaths and billions of doDars of losses worldwide in this century. In many earthquakes the resulting landslides have caused as much or more damage than the other effects of ground shaking. Over half of the damage caused by the 1964 Alaska earthquake was the result of landslicles. In Japan, of the deaths caused by large earthquakes since 1964, more than half have been attributed to landslides. In an earthquake in the Peruvian Andes in 1970, an avalanche was triggered that buried two cities and killed at least 20,000 people. The 1987 earthquake in Ecuador caused landslides that clogged rivers and destroyed sections of the trans-Andean of} pipeline. In 1959, the Hebgen Lake, Montana earthquake set off a mam- moth landslide that dammed the Madison River. Major efforts were made to reduce the possibility of rapid erosion when this natural dam was overtopped, to prevent catastrophic downstream flooding. The 1971 San Fernando earthquake caused very damaging slides in earth dams and structural earthfi~] in the western part of the San Fernando Valley most of which were associated with liquefaction. In addition there were several hundred rockfalIs, soil falls, and de- bris flows that caused considerable damage to highways and roads. Blockage of roads is a common occurrence whenever earthquakes shape steep terrain.

OCR for page 195
203 TABLE F-3 Qualitative Assessment of Abundance and General Character of Liquefaction Effects as a Function of LSI for Areas with Widespread Liquefiable Deposits LSI Abundance and General Character of Liquefaction Effects 10 30 70 90 5 Very sparsely distributed minor ground effects include sand boils with sand aprons up to 0.5 m (1.5 It) in diameter, minor ground fissures with openings up to 0.1 m wide, ground settlements of up to 25 mm (1 in.~. Effects lie primarily in areas of recent deposition and shallow groundwater table such as exposed stream beds, active flood plains, mud flats, shore lines, and 80 on. Sparsely distributed ground effects include And boils with aprons up to 1 m (3 ft) in diameter, ground fissure with openings up to 0.3 m (1 ft) wide, ground settlements of a few inches over loose deposits such as trenches or channels filled with loose sand. Slumps with up to a few tenths of a meter displacement along steep banks. Effects lie primarily in areas of recent deposition with a groundwater table less than 3 m (10 ft) deep. Generally sparse but locally abundant ground effects include sand boils with aprons up to 2 m (6 ft) diameter, ground fissures up to several tenths of a meter wide, some fences and roadways noticeably offset, sporadic ground settlements of as much 0.3 m (1 ft), slumps with 0.3 m (1 ft) of displacements common along steep stream banks. Larger effects lie primarily in areas of recent deposition with a groundwater table less than 3 m (10 ft) deep. 50 Abundant effects include sand boils with aprons up to 3 m (10 ft) in diameter that commonly coalesce into bands along fissures, fissures with widths up to 1.5 (4.5 ft), fissures generally parallel or curare toward streams or depressions and commonly break in multiple strands, fences and roadways are offset or pulled apart as much as 1.5 m (4.5 ft) in some places, ground settlements of more than 1 ft (0.3 m) occur locally, slumps with a meter of displacement are common in steep stream banks. Abundant effects include many large sand boils (some with aprons exceeding 6 m [20 It] in diameter that commonly coalesce along fissures), long fissures parallel to rivers or shorelines usually in multiple strands with many openings as wide as 2 m (6 ft), many large slumps along streams and other steep banks, some intact masses of ground between fissures displaced 1-2 m down gentle elopes, frequent ground settlements of more than 0.3 m (1 ft). Very abundant ground effects include numerous sand boils with large aprons, 30 percent or more of some areas covered with freshly deposited sand, many long fissures with multiple strands parallel streams and shore lines with openings as wide as 2 or more meters, some intact masses of ground between fissures are horizontally displaced a couple of meters down gentle slopes, large slumps are common in stream and other steep banks, ground settlements of more than 0.3 m (1 ft) are common. SOURCE: Youd and Perkins (1987~.

OCR for page 195
204 TABLE F-4 Relative Abundances of Earthquake-Induced Landslides in 40 Historical Earthquakes Worldwidea Very abundant (> 100,000~: Rock falls Disrupted soil slides Rock elides Abundant (10,000 to 100,000~: Soil lateral spreads Soil slumps Soil block slides Soil avalanches Moderately common (1,000 to 10,0003: Soil falls Rapid soil flows Rock slumps Uncommon (100 to 1,000~: Subaqueous landslides Slow earth flows Rock block slides Rock avalanches aLandslide type listed in order of decreasing total numbers. SOURCE: Wilson and Keefer (1985~. An excellent, recent summary about earthquake-~duced land- slides and their consequences has been prepared by Wilson and Keefer (1985~. Table F-4 assembles data concerning the relative abundance of different types of landslides, while Table F-5 categorizes different types of earthquake-induced landslides together with their charac- teristics. Likelihood of [an~lides Information relating the occurrence of landslides to characteris- tics of earthquakes has been summarized in ATC-13 (Applied Tech- nology Council, 1985~. Building on a concept proposed by Legg et al. (1982), and utilizing expert opinion, ATC developed the probability matrices reproduced in Table F-6. Each box in this table represents a different degree of inherent stability for a slope, characterized nu- merically by the yield acceleration, ac, at which movement starts. The slope failure states (SFS) relate, in a probabilistic manner, land- slide displacement to shaking intensity as a function of initial slope stability. The matrices are for dry summer conditions in California;

OCR for page 195
Q C 3 m ._ - m C) L. o C) ._ CO ._ s" V U) :~ ._ C) o Q) C~ s" - C. V] _ $~ ~ U. a, ._ o a C~ o _ ._ L. ~ ~ a' - e ._ - 4 ~ o P4 C. 3 3 o o _ _ a, ~ ,= CQ U] ~ a, a) a' _ :^ o . U~ s~ o~ o~ ~o s~ ~2, ~ p~ 3 ~ o X X X X ~ X X X X X C) o ._ r~ a~ ._ a' ~s X X X X X X X ~ 0 3 0 bO ~ _ _ ~ ~,o ~ ' 5 ~ ~ ~ ~ ~ E CO a~ ~ V ~ _ a: ~ ~ ~ ~ ~ ~ aS °° ~ 0 es ~ — C43 ~ — ~ .= ~ ~ ~ ~ ~ ~ ~ 205

OCR for page 195
a' a Q :^ ._ o o V US 3 a, V] A. C. U: o _ ._ ~ ._ ma o _` . o v U3 1 ¢ I hi; 3 3 3 o o o ~ ~ ~ ~ C a a :^ ~ S ~ ,, ~ ~ ~ E ~ 3 ~ 3 > c ~ ~ ~ X ~ X X X X X - ~ X X X ._ - a~ as X X X ~- ~C bO ._ ~C ._ bO ._ X C~- o ~v ~s o ._ ~ - O L. ~ o _ U] ~V — 2 . ~ c ~ 3 u = D 3 ~ ~ C ~ O n ~ a: a, = 3— ~ == —o 3 o 0 a cn v ~ ~ 206

OCR for page 195
3 0 ~ ~ ° 3 sat— At, Act Q oo o ~ ~ '" Ct ~ ~ S =~ ~ .= X ~ ~ ~ X ~ ,= X ~ X ~ X ~ a. X C-- X no, ~ ~ s ,, s C ~ "E D ~ ~ c: if-, 1 ~ ~ e ~ - I; ~ 3 IS _ ~ 3 ~ ~— C ~ 8 ~ CQ 207 CO red - .. . o ~ =m ~ .E to :> ~ ° 3 ~ ~ cat ~ ~ 3 o ~ ~Q ~ be ~ O ~ a:} ·— O-Y be "0 or .c _ ~ 3 O u.~,,~...E os _ o U] o ~ CO _ ~ o ~ ~ ~— .. ,' oo .~5 ~ .~ ~ E ~ £ o ~ ~ {~5 U ~ CO o5 _ · _ ~ C. ~ o° ~ == V V _ O ° al ~ ~ ~ ~ ~o 3 "~ h=.= s~ ~ ~ £ aS ., _, .. ~ ~ o ~T:) ~ c,, 3 ~ ~ ~ ~ {S ·_ {e _ ·- >. . ·- ~c y t=t, ~ ~ E _ ~ —o ,~ ~ X ~ o - oo -

OCR for page 195
208 for applying to a wet season it is recommended that the ~fM! be in- creased by one unit. Yield acceleration has been used in conjunction with a Newmark (1965) analysis to prepare a regional map of seismi- cally induced landslide susceptibility as a function of bedrock type, slope steepness, and seasonal groundwater-leve! conditions (Wiec- zorek et al., 1985~. Going a step further, ATC also used expert opinion to relate landslide severity (i.e., SES) to the mean damage ratio (MDR) at affected facilities (see Table F-7~. Thus, the mean damage ratio from landslides is: MDRL,s = As, PtSFS] x CDF~s, SFS where PtSFS] comes from Table F-6, the central damage factor CDF~s is from Table F-7, and the products are summed over all slope stability states. This ATC method is logically sound, but at this stage it involves considerable judgment and has not yet been tested for an actual large-scale study. Mapping Landslide Hazards During the Bay Area Project of the 1970s, a landslide hazard map was developed for the San Francisco Bay Area (Nilsen and Wright, 19793. The indicated hazardous areas were identified on the basis of evidence of past sliding (not necessarily during earthquakes) and topography. Wieczorek et al. (1985) produced a map of earthqual~e-caused landslide susceptibility for one of the San Francisco Bay Area coun- ties. In this approach, the nonseismic data needed are: maps showing the distribution of geologic materials; estunates of the wet and dry strength characteristics of each of the age or stratigraphic cIassifi- cations obtained from the geologic maps; estimates of wet and dry season depths to saturated soil; and maps showing topography, with the contour intervals assigned to one of six percentage slope ranges. These geologic-based susceptibility data are then combined with a consideration of ground motion. Faults capable of producing suffi- cient motion to cause slides are identified. (Because the map was in- tended to serve several purposes rather than being tied to a scenario- based disaster response planning study, the effects of the different earthquakes were not plotted discretely on the final map.) Several

OCR for page 195
- ~ or o · · o v L' en - ~ * on · - 4. · - · o - · o - u) m _ __ __ _ O Wd: o v v o en Q . __ _ _ iC _ NC ~e S :) ~e~e =~t ° O ~ . O O o O O O O O ~ 0 0 O O 0 O _ O O O O O O O _ O _ _ 1 1 1 1 1 - 1 1 1 v 1 1 ~1 1 v 1 1 ~ 1 1° 1 1 ~ 1 I E , ~ 1 1 o:1 1 1 ~ 1 1 ~ 1 1 o 1 I ~ 1 1 ~ 1 1 1 Ct 209

OCR for page 195
114d2 - ~ ~4 . - 4= o v - ~ m E~ ~ . -- e O e ~ = O ~ o: O 0~ O O O O O ~ O ~0 0 `-, ~t g! O o2 ~ ~ ~ ~ ~ · ~ ~ ~e ]-. ~ i ~ , !! ~ 0 ', E E ~ · _ ~ ~ E b e E: O ~ ~ ~ O ~ E , v 0 ~ ~ E~ ~ ~ ~t E ~ _ ~ U: V ~ e!1 ° ! E _ 4} vl od o - E - - ~ cn S 0 ~ ~ ~ 0 ~ eG ~ E ·' '~ ~!i ~ §~3 E ~ ~ ~ ~ o 1' ewl E ~ c ct " £_ l' ~ o ~ ° 3 v E ~ ;~ t' ~e 3] ilv 1 ~ _7 E'=~ ~ j~~°~ ~e ~~; , Ii2111 ,i'ni, I t I t I I IC — E G Cl ~ ~ ~ ° 210

OCR for page 195
211 TABLE F-7 Relation Between Landslide Severity and Facility Damage Factora Central Slope Failure State Damage Factor (percent) Light Moderate Heavy Severe Catastrophic o 15 50 80 100 aEstimates are based on consensus of the ATC-13 Project Engineering Panel. SOURCE: Applied Technology Council (1985~. . historic earthquake records are adapted to represent the size of earth- quake assigned to each fault. Simple slope stability analysis is used to determine the yield or critical acceleration, ac, necessary to overcome slope equilibrium. The severity of the slide, in terms of the amount of displacement, is then computed using the method of Newman (1965) as adapted by Wilson and Keefer (1983), which accounts for the way in which suc- cessive accelerations of critical or greater size act over time, against the restraining influence of friction, to move the slide downhill. The results are displayed on a map and divide the study area into high, moderate, low, and very low earthquake-caused landslide hazard zones. Liquefaction was beyond the scope of this method, but liquefaction susceptibility was also plotted on the same map from the work of Youd and Perkins (1985~. The four descriptive landslide sus- ceptibility categories are defined quantitatively in terms of predicted movement, relative to a benchmark amount of displacement of 5 cm (2 in.~. This was considered a conservative estimate of the threshold of movement causing major damage to average building foundation conditions, based on Youd (1980~. The other factor determining the assignment of a site into one of the four zones was the critical acceleration causing the movement. For each of these four levels of susceptibility, an estimate is provided of the percentage of the area of that zone that would fait when the presumed earthquake occurs. This estimate of the extent of failure within each landslide zone is derived from Youd (19803.

OCR for page 195
212 Figure F-4 shows the maximum distance of several types of landslides ~ a function of magnitude and was assembled by Wil- son and Keefer (1985) U8=g data from California. These authors also used Newmark's sliding block theory to relate the likelihood of slides to the intensity of ground motions, and produced a map (see Figure F-5) giving the probability of coherent slides (in either hilly terram or saturated soils) for a magnitude 6.5 earthquake on the Newport-Inglewood fault. This type of mapping is still in the developmental stage, and does depend heavily on historical data con- cerning earthquake-induced landslides. However, the work points the way to the type of analysts that can be used for mapping landslide hazards.

OCR for page 195
- - l i I I I I I I I T~ I \ ~ · I \ - e Y \~ ·  \ C \ ~ CO—4% cr. i i ~ ~ 1 1 ~ 1 O O O O O O O LO ~ ~ ~ C~ o o o o .O ~ ~ o L`~ ~ ~ o o SH319W011> Nl 3~0Z 3Unid(18-llOV] WOU] S3Ol1S lN383H00 30 3ON~lSIC WtlWIXVW i ! i 1 _ ~ ~ _ \ \ · ~ _ - \ - \ - V~e \ LL \ - \ J\ ~ ~ ~ . a:N U) o o o o o L£) _ o ln _ . oo _ o oo _ LO — r~ ~ O UJ _ . _ CD Z _ o, LO _ ln _ o .D LO o ~ o a' o, ID co o oo LO r~ o r~ o, LD o o O O O O O ~ ~ — LO ~ o o o SH313W011> Nl'3NOZ 3UnidnU-llf)~] WOU] SllV] H0 S3Ol1S 031dnUSIC J0 33N~lSIC WnWIXVW 213 - - UJ z 6 .= ~ o ~ .. ~ O _ ~ ~ V =_ _ c<3 O ~ U2 ~ 4= ~ U '_ a a) ~ ~ ~ CS, _ s C~ U) ".= " - V ° a2 -D vrY >~. · - ° ~ V 0 _ ~ 0 , . U' _ ~ ~ C~ _= _ ~ ~ - ~ V>^ ~ C, V^¢ CQ bO~ ~ ~ ~ 0 ~ ~ o~ ._ ~ ~ '6 ~ V ~- ~ ~ _ ~ ° tS, Cc~ U) ~ _= ~ ,= ~ ~

OCR for page 195
 - \ -  . - %e ~: \: ~--~- \ `;~ \ \ 1 . ~ . 1 1 1 1 1 1 1 ~ o o o o o o ~ c~ ~ ~ O O O Lf) ~ ~ ~ ~ c SH313W011> Nl 3NOZ 3UnidnH-llnV] wo~ 3allsaNv1 ~o 3~vlsla wnwlxv o o o 1 1 - 1 1 1 1 1 1 ~—- .= _ _ ,,, 13 = ~ ~ o ro c, _ ~ c Q) _% c .m ~ o `- ~ ~, I . : I , I I ---T-- I I ~ T I ~ , . a~ I o _ . 1 Ln 1= ~ o 1 1 ~ o _ ·o o CD .o o, U~ Ln o, ~ . o 1 1 1 1 1 1 1 1 1 1 ~ o o o o o o o ~ ~ _ o o o o ~ ~ ° ~ ~ ~ o 0 SH313W011~ Nl 3NOZ 3UnidnU-llnV] WOU] SMO13 HO SOV3UdS 1~31~1 jO 33NUlSlO WnWIXVW 214 ·r, _ o _ CO o _ oo _ ~ . — UJ o Q —r~ Ln Z _ , . CD 10 I ~ — LO ,c,— ~ ~ . o ~, ~ o _ ~ ~ ¢ ~ ~ m ~ V) =: . ~4 0 ~ ~ oo oB ~— _4 O a) 0O V ~ ~ ~ - t-. ~ ~ · - ~ {~ ~ O C5) -= ~ I-. · _- oo— C5> ~ ~ v ~ ~ ~ ~ o i ~ ~v ~ d p~ ~ ;> ._ ~ .. :^ C~ U] ,~ O ~ Co—U2 == O - _` _ ~? ~ ~ _= ·— ~ oo C7) O Ce _ ~ ~ ·" _`o ~ ~ ~ oo ._ .= _m ~ ~ U:~ ·— O ~ ~ V ~= — m O _ ¢! ~ ~ _' ~4 ~ ~ 2 ~ .~ · - _ ¢ - o, ~ fi ~ v

OCR for page 195
215 A I. 1 1 I. 1 rat ,, M O J A ~ .E: 'ID E Syt R , , ¢. ~ A::;:: A E ~ ~ A ~ 1, ' ~ off I ~ =~9~ ;j~—~5~\ '-iffy SALES _~' ~ ! \ ~ G . . -a - rut ~ , \- ~ ~~ ~ ~~ 3!\ 72- RIVE ,, I"""_", ~-~^ i,^~ 'ii' ~o5~85~ . ~ ~ V~__ _ ~ i; \ °~ , ~1 i , ~^ ~ - R