JNIOSH

Abstract of Special Research Report (SRR-No.35)

National Institute of Occupational Safety and Health, Japan

Safety for Construction of Small-to-medium Sized Excavation Using Observational Procedure

Introduction

SRR-No.35-1
Yasuo TOYOSAWA

: Accidents due to slope failure frequently occur at excavation sites. About 20 workers die annually in accidents attributed to slope failure or rock fall during slope-cutting work in Japan.
    The accident associated with earth failures (e.g. trench failures and slope failures) is one of the major accidents in construction work. These accidents occasionally involve serious accident in which more than 3 workers are injured. The numbers of accident in construction work have been gradually decreasing, yet they still account about 500 accidents in every year. Among them, failures and slope failures are major causes of accidents that account for about 3-6% of accidents in construction work.
    According to the investigation of fatal accidents caused by earth failures, almost 40% of these accidents occurred during slope excavation work. So, this report focused on prevention of the slope failure accidents.
    Main objectives of this research project are to understand the situation of accident in slope excavation works, to understand of failure mechanism of the slope and to establish the effective countermeasures for preventing slope failure accidents.
    For this purpose, a series of researches has been carried out as follows:
  1) Analysis of labor accidents caused by slope failure during slope-cutting work
  2) Stability of trench excavation under construction machinery load
  3) Experimental analysis on shallow failure of slope in consideration of relationship between ground permeability and precipitation intensity
  4) Stability of shotcreted slope during the maintenance
  5) Physical modeling of slope failure during slope cutting work
  6) Application of Tilt-sensor in the slope excavation field for the measurement of slope movement just before the collapse
  7) Development of the Short Pipe Strain transducer (SPS) and experimental analysis on its applicability for safety monitoring in slope work
  8) Development of monitoring and alarm system for slope failure using laser beam and optical sensor

Analysis of Labor Accidents Caused by Slope Failure During Slope-cutting Work

SRR-No.35-2
Kazuya ITOH, Yasuo TOYOSAWA and Noriyuki HORII

: Accidents due to slope failure frequently occur at excavation sites. About 15 to 20 workers die annually in accidents attributed to slope failure or rock fall during slope-cutting work in Japan. Study of such accidents from 1989 to 2002 was carried out and found that about 131 fatal accidents occurred due to slope failure and rock fall. Analyses of such accidents were done to recommend effective countermeasures for reduction of such accidents. The profiles of the slope failures were studied and classified by various factors, such as geology, type of construction works, slope height, gradient, scale of failure, and rain fall conditions, etc. These factors were investigated and analyzed in this research and following summary is made.
  1) About 70% of labor accidents caused by slope failure during construction works occurred in retaining wall construction. Regarding types of work, many accidents occurred during the works related to retaining walls which were carried out near slopes, as well as excavation works of the slope and ground.
  2) About half of the deaths were caused by suffocation and crushing. This means that victims died when they were buried alive under earth and sand from failed slopes or their chests were crushed by such earth and sand pressure.
  3) According to the results of assessments of failed ground, slope failure that leads to labor accidents often occurred at the slope angle equal to 60 degrees or more but less than 75 degrees. This indicates that labor accidents tend to occur in construction works where the slopes are too steep. Natural slope failures often occurred at gradients, which varies from 40 to 49 degrees.
  4) In 60% of cases, the collapsed soil amount was less than 50 m3 and it was understood that many accidents occurred where the small-scale slope failures took place.
  5) Regarding the correlation between rainfall and accidents, in about 60% of the accident cases, there was rainfall within three days before the occurrence of the accident. It turned out that there is a close relationship between rainfall and slope failure.

Stability of Trench Excavation Under Construction Machinery Load

SRR-No.35-3
Yasuo TOYOSAWA, Kazuya ITOH and Junjie YANG

: There are many labor accidents in which the workers are killed and buried under collapsed ground when the construction machinery such as a drag shovel falls or topples over the edge of the trench during excavation. Based on the labor accident reports, during the period of 1994 to 2002, there were approximately 30 workers killed every year in Japan due to this type of failure. To prevent the loss of life and injury caused by the ground collapse, it is necessary to gain more understanding about the effect of construction machinery load, geometry of ground excavation (e.g. slope angle and excavation depth) and characteristic of ground strength on the stability of trench excavation.
    This study examined the stability of the edge of excavation under the load of construction machinery by the following procedure: 1) The load of construction machinery was modeled; 2) Failure behavior and depth of excavation were studied by conducting centrifuge model tests using an in-flight excavator, with the magnitude of the machinery load, position of the load, etc. as parameters; and 3) The test results were evaluated by upper bound solution.
  1) Modeling of Construction Machinery Load:
    Based on the investigation of 287 different types of the drag shovels used in the trench excavation, the construction machinery load was modeled by a rigid block (U-shape) with b = 0.6 m (in prototype scale) and B/b = 5.
  2) Geotechnical centrifuge modeling:
    The stability and failure mechanism of ground excavations have been investigated using the centrifuge modeling by many researchers (Kusakabe 1982, Taylor 1984 and Toyosawa et al. 1994). However, the stability and failure mechanism of trench excavation under the construction machinery load have not been fully investigated. Therefore, in this paper, a series of centrifuge modeling tests was conducted on various types of ground models under different magnitudes and locations of the machinery load in order to examine the stability and the failure mechanism of trench excavation under the construction machinery load.
  3) Evaluation by upper bound solution:
    For the sandy ground (low cohesion, high friction angle) and the Kanto loam ground (high cohesion, low friction angle), it is possible to express the failure mechanism with the composite slip surface using straight and logarithm lines. The safety factor and failure pattern calculated from the upper bound solution agree well with the centrifuge test results.

Experimental Analysis on Shallow Failure of Slope in Consideration of Relationship between Ground Permeability and Precipitation Intensity

SRR-No.35-4
Satoshi TAMATE, Naoyuki ITOH and Akira ENDO

: Rainstorms frequently induce slope failures that take a lot of lives and destroy many social assets. Recent increases in both the amount of rainfall and the intensity of precipitation cause severe damages in Japan as well as other countries throughout the world. In particular, heavy rain events induce shallow failures in slopes within short durations. Therefore, the risk of failure must be analyzed in particular sites to reduce losses and save peoples lives. However, preliminary evaluations are not conducted well because simple methods of analysis are not well considered.
    This study conducts a series of experimental simulations of slope failure caused by precipitation in a centrifuge. 30mm/hr of the intensity of precipitations (rp) and 300mm of the amount of rainfall (Rp) were decided as the standard conditions for heavy rain in prototype, and then these conditions were simulated in 50g centrifuge model tests. Pneumatic spray nozzles producing fine mist atomization with mean droplet diameter of 10 - 50μm were used to simulate the equivalent impact pressure on the ground surface in the prototype. A newly developed rain simulator composes 17 pneumatic spray nozzles that enable to give 1500mm/hr of the model precipitation intensity (rm). Measurement of slope deformation by installation of conventional displacement transducers is not performed in these tests because these instruments disturb the uniform distribution of precipitation. In addition, serious problem is that spray mist sticking around the instrument accumulates so that enlarged raindrops fall on the model slope with a high impact force. Therefore, shallow strain transducers were also developed to monitor the progress of slope failure caused by these insertions in the surface.
    Two kinds of soil materials, which are Toyoura sand and Silica100, are used in the tests to simulate different permeability (ksp) in the model grounds which are 4.0×10-2(cm/sec) and 1.48×10-3(cm/sec), respectively. Index of relative permeation (Ir) is defined as the ratio of ksp to precipitation intensity (rp), i.e. Ir=ksp/rp. In addition, two kinds of liquid are also used to give different viscosity in precipitation to investigate the relationship between rm and Ir . Slope inclination was given to be 30 degrees and 45 degrees.
    Uniform precipitation was achieved by arranging the jetting angle of the nozzles. Shallow failures were observed in the lower permeable ground conditions where Ir=1.78. Precipitation mobilized the saturated layer in shallow section of the slope so that pore pressures increased by occlusion in the ground. Since predominant vertical permeation varies the thickness of the layer of saturation in the slope, the failure thickness decreases as the angle of the slope inclination increases. Meanwhile, water dissipated in the high permeable ground condition where Ir=50, and then slope failure did not occur. Consequently, the potential of slope failure in the shallow section basically decreases as the value of Ir increases. The potential of slope failure also increases in certain ground that has a highly permeable surface overlaying on soil with low permeability. Accordingly, it was clarified that the potential risk of slope failure by heavy precipitations must be evaluated by considering both rp by weather forecasts and ksp by preliminary site investigations.

Stability of Shotcreted Slope During the Maintenance

SRR-No.35-5
Kazuya ITOH, Yasuo TOYOSAWA, Masafumi SUZUKI and Naoaki SUEMASA

: The maintenance of slope stabilization constructed during the rapid economic growth period of Japan (1955-1973) is increasing in recent years. In that period, most of the slope stabilizations were done by shotcrete, which can protect the slopes from water intrusion and erosion. Every year, some accidents take place during the maintenance of such shotcreted slopes. Hence, it could be said that such type of labor accident might increase from now onward.
    In this research, at first, statistical analyses of labor accidents occurred due to the slope failure during construction work within 14 years period from 1989 to 2002 were carried out to explore the basic statistical characteristics of damage consequences during maintenance repairing. And some of labor accidents caused by slope failure, which is covered with shotcrete and which is occurred at the construction sites during the removal of old shotcrete, are examined to find the cause of failure.
    Secondly, the results of geotechnical centrifuge model tests which simulated the failure are mentioned. Especially, this paper focuses on the influences of weathered soils which exist behind the shotcrete and the process of the removal of old shotcrete on the stability of slope by conducting a series of centrifugal model tests. In the case of the process for smashing shotcrete from the top of the slope to the downward direction, Slip failure of the weathered soils under separated mortar-shooting is initiated from the top area, leading to unstable behavior of the separated mortar-shooting and the whole slope failure occurred. On the other hands, in the case of the process for smashing shotcrete from the toe of the slope to the upward direction, sliding of the smashed shotcrete occurred all of sudden and smashed shotcrete and the weathered soils behind that also slid down. It is important to consider the thickness of weathered soils behind the shotcrete for the stability of shotcrete. Finally, the results of numerical models using the distinct element method (DEM) which simulated the centrifuge model tests are mentioned to arching effects of particle sizes with weathered soil thickness on the slope failure.

Physical Modelling of Slope Failure During Slope Cutting Work

SRR-No.35-6
Kazuya ITOH, Yasuo TOYOSAWA, Surendra B. TAMRAKAR,Sahaphol TIMPONG and Noriyuki HORII

: Excavated slopes become safe only when they are either excavated to a safe slope angle orwhen protective measures are taken. During slope cutting work, slopes are at greater risk of failure. In many cases, slope failure suddenly occurs without any warning signs and the workers do not have time to escape and hence accidents take place. In Japan, there are about 30 to 40 slope failure accidents every year, causing casualties among workers and damage to property. More than 50% of such accidents occur during excavating or leveling the lower parts of the slope toe. To prevent such accidents, the mechanism of slope failure caused by slope cutting work should be clarified.
    In this study, field tests of slope excavation were carried out; a model slope was prepared by compacting soil with a small bulldozer and trimming the slope surface by backhoe. The height, width and slope angle of each model slope were 5 m, 3.5 m and 45 degrees, respectively. The slope was excavated vertically downward from the slope toe using the backhoe. The excavation was continued until complete slope failure occurred. As a result, it was found that in all cases spalling or local failure of the excavated slope was observed before the slope failed completely. This mechanism is consistent with some accidents due to slope failure during slope cutting work. Because the horizontal and vertical movements could be measured during the field test, it is possible to predict the movement of the slope just before failure.
    Centrifuge model tests were also conducted to verify the field test and clarify the slope failure mechanism caused by the slope cutting work. In addition, the centrifuge model test may be able to reproduce the field test.

A Study on Prediction of Slope Failure Using Highly Accurate Tilt-Sensor

SRR-No.35-7
Yasuo TOYOSAWA, Kazuya ITOH, Surendra B. TAMRAKAR, Takaaki ARIKI, Takashi KUNIMI, Atsushi NISHIJYOH, and Satomi OHKUBO

: Sudden failures of slope during or just after slope excavation works cause many serious accidents which sometimes take the lives of workers. Loss of lives and property damages could be prevented if early predictions of failure could be made. Although many instruments have been developed to measure the movement of landslide, however, most of them are either difficult to set up or are too expensive to use at small to medium-sized slope excavation sites. In this study, a simple and costeffective slope monitoring system using the tilt-sensor was developed, and the mechanism of slope failure was also investigated.
    The tilt-sensor consists of highly sensitive accelerometers, which are made up by anodic bonding of three layers of silicon (two fixed and one movable) and glass. By measuring the output voltage, it can measure the change in the inclination of the sensor within the range of 20 degrees with the sensitivity of 100mV/deg.
    Two types of prototype model of tilt sensors; small compact type and stand-alone type were used. The small compact prototype model comprises of small size accelerometers (10×8×5mm), which are mounted on a small plate (22×18mm) supported by a flat longitudinal bar. Whereas the stand-alone prototype model consists of fabricated tilt-sensors, which are rested on a thick plate (120×120mm) supported by a large diameter (48.6mm) pipe.
    The small-scale full size model slopes of River sand were prepared in the laboratory and the fullscale model test of Kanto loam and Narita sand slopes were prepared in the field. The tilt-sensors were set up both on the slope surface and on the top of slope. Stepwise toe excavation was then followed at regular time interval and continuous measurement of tilt-angle was made until slope failure occurred. By analyzing the trend of tilt-angle movement (slope movement), it was found that in all cases, at each step of excavation, the changes in the tilt-angle were observed. The amount of change in the tilt-angle gradually increased during the progress of the excavation and it changed rapidly just before the slope failure. Large increment in the tilt-angle just before the slope failure showed the applicability of the tiltsensor for the possible early prediction of slope failure during the excavation works.

Development of the Short Pipe Strain Transducer (SPS) Comprising a Screw End and Experimental Analysis on its Applicability for Safety Monitoring in Slope Works

SRR-No.35-8
Satoshi TAMATE and Akira ENDO

: Slope failures frequently cause labour accidents in construction sites. It is also known that even a small collapse can cause serious injury to workers. In particular, vertical cuts at the toe of slope are carried out to build retaining walls for expansion of load width, and unstable slopes appear even short term. Slope failure must be avoided for safety. Consequently, temporary constructions are needed to support slopes. However, this process is sometimes omitted because of difficulties resulting from a lack of sufficient lateral reaction force. Since many slope works are going on, finding methods are needed to ensure safety for workers in slope works. One of the solutions considered is the provision of a warning system that provides the time required to escape through accurate monitoring. A slip surface appears at slope failures, and then major deformations develop along the slip. However, shallow deformation arises at the same time even though higher sensitivity is required to measure its small values. The Short Pipe Strain transducer (SPS) was developed, and then its applicability was investigated by Centrifuge model tests and Prototype model tests. It is compact in size with a length of 1m and a diameter of 15mm. SPS outputs the strain based on the response of bending deformations. Since the SPS penetrates by itself without pre-borings, installation into ground is easily conducted by a hand operated drill. Moreover, SPS strongly contacts with surrounding soil and this contributes to high sensitivity.
    Prototype model tests were carried out in several sets of the model slopes, which were composed of loose sand, less compacted loam and compacted loam. 45 degrees of inclination and 5m of height were given in the model slopes. SPS was set into the shallow section of slopes near the shoulder. Vertical cuttings were carried out from the toe of slopes to simulate failure. Centrifuge model tests were also conducted by giving the equivalent conditions of prototype model tests to provide an additional test of SPS.
    Clear increases in the responses of strains (rs) were measured with progress of slope cuttings in both prototype tests and centrifuge tests prior to failure. Sensitivity of rs depends on distance between SPS and failure parts. In loam ground, creep strain curves were observed. A couple of minutes could be provided by monitoring the 2nd or 3rd creep. In sand ground, a liner relationship between rs and settlement by conventional displacement transducers was observed.
    Accordingly, it was confirmed that SPS does not only measure the slope movement but also senses an increase of the potential risk of slope failure. Slope inspections must be conducted to ensure safety during works, and SPS is applicable for subsidiary warning systems to saves workers' lives.

Development of a Monitoring and Alarm System for Slope Failure Using a Laser Beam and Optical Sensor

SRR-No.35-9
Kazuya ITOH, Yasuo TOYOSAWA, Minenori TAKEYAMA and Tetsuya SANO

: Accidents due to slope failures frequently occur at excavation sites. In Japan, approximately 20 workers are killed each year by slope failure or rockfall during slope cutting work. Lives could be saved and damage reduced if early warning of slope failure could be made. Although many monitoring instruments for landslides have been developed to measure the movement of the slope just before failure, most of them are either difficult to set up in the field or are too expensive to use in small- to medium-sized slope cutting sites. To prevent accidents due to slope failures, it is necessary to develop a system for monitoring small- to medium-sized slope cutting sites.
    In this study, a slope failure monitoring and warning system using a laser beam and optical sensor was developed and its application was examined. Firstly, the laser beam and optical sensor were selected to satisfy the requirements of a precise measurement system with simple installation in the field at low cost. Based on trial tests, an amorphous silicon optical sensor, 635 nm laser beam, and sharp cut filter (transition wavelength: 620 nm) were selected in this study. Next, a model test of slope excavation was carried out to investigate the slope movement just before failure. The height, width and slope angle of the model slope were 2.2 m, 1.35 m, and 60 degrees, respectively. The results showed that the relationship between the output voltage of the optical sensor and deformation of the slope was found to be linear within a limited range. Finally, a monitoring and warning system for slope failure was developed based on these results.

Labor Accidents Caused by Slope Failures and Preventive Strategies

SRR-No.35-10
Yasuo TOYOSAWA and Kazuya ITOH

: In Japan, slope failures have caused a number of fatalities and economic loss throughout time. Most of the human-induced slope failures occurred due to the destabilization of slope when the lower part of a natural slope was removed or the slope was steepened during construction works.
    Risk assessment was stipulated in the Article 28-2 of ISHL (Industrial Safety and Health Law) in 2006. To implement the risk assessment for preventing labor accidents caused by the slope failures, it is necessary to understand the hazards both in terms of quantitative and qualitative aspects. In particular, when the ground condition has a high degree of uncertainty, it is important to take a comprehensive approach. The construction safety should be considered as part of planning from commencement of a project. In general, the first and foremost priority would be design and construction procedures. The appropriate design should be based on good site investigation and justified evaluation for slope stability. In addition, even if the safe construction procedures are conducted, a slope monitoring system may be adopted to compensate the remained chance of slope failure. This report presents the effective risk assessments in combination with 1) review of the necessity of operations on the excavated slopes in construction sites, 2) site investigation and the evaluation of slope stability during the construction, 3) safe construction procedures and 4) slope monitoring as follows.
  1) Review of the necessity of operations: Although most of the slope excavations are excavated by the excavators, the workers may have to enter dangerous areas near the excavated slope. The contractor must review this necessity and alternate construction methods requiring no operation below the excavated slope should be employed.
  2) Site investigation and evaluation slope stability under the construction: It is necessary to evaluate the stability of slope and risk of slope failure not only after the completion of construction but also under the construction.
  3) Safe construction procedures: It is a fundamental requirement to employ a construction method that can ensure safety throughout excavation of slopes from the start to the completion of excavation. The authors recommend a simpler and cheaper method with the use of retaining walls installed after sheet piles or pressing the H-section steel beams into the slope to prevent the slope failure.
  4) Slope monitoring: The construction should be conducted by assuring safety of the slope throughout the construction process based on the slope monitoring data such as ground deformations obtained from displacement sensors, high precision tilt-sensors and etc.

Other Publications

Kiyose District Map

Umezono 1-4-6,Kiyose,
Tokyo 204-0024 Japan
TEL: +81-42-491-4512
FAX: +81-42-491-7846

Noborito District Map

Nagao 6-21-1, Tama-Ku,
Kawasaki 214-8585 Japan
TEL: +81-44-865-6111
FAX: +81-44-865-6124