INSTITUTE OF ENGINEERING (IOE)

TRIBHUVAN UNIVERSITY (TU)

DEPARTMENT OF CIVIL ENGINEERING
PULCHOWK CAMPUS, LALITPUR

Project work on

Probabilistic (simplified SPT performance-based) liquefaction Hazard

evaluation of the Kathmandu valley.

Submitted To

Assoc. Prof. Dr. Indra Prasad Acharya

Campus chief
Pulchowk campus, lalitpur

Submitted By
Chakra Bhandari (078/MSGtE/004)
Email: rcbhandari42@gmail.com
15 June 2023
Abstract:
This thesis/research work aims to conduct a probabilistic seismic hazard analysis (PSHA) of Nepal and
simplified SPT performance-based probabilistic liquefaction hazard evaluation of the Kathmandu Valley.
Large-scale liquefaction has occurred in the area during previous earthquakes, notably those in 1833 and
1934, which left behind severe damage. Although there was very minor liquefaction damage from the
2015 earthquake (Mw7.8), surface manifestations were seen across the valley. These historical
earthquakes highlight the need to understand, analyze and mitigate liquefaction hazards in Kathmandu
valley. Reliable seismic hazard & liquefaction assessment is a critical element in development of policy for
seismic hazard mitigation and risk reduction. By employing Standard Penetration Test (SPT) data from soil
borings and PGA from PSHA of Nepal, this study will aim to demonstrate the advantages of performance-
based probabilistic seismic hazard and liquefaction analysis. The study is seeking to use the probabilistic
liquefaction models from Boulanger and Idriss (2012) and Cetin et al. (2004) model to perform liquefaction
analysis of Kathmandu valley. For PSHA, logical tree approach will be used by considering 23 areal zones
and 14 linear seismic sources (active faults) in R-crisis software. Entire area of Nepal will be divided into
1.2 * 0.6 degrees grid size and the peak ground acceleration (PGA) will be calculated for the return period
of 10, 108, 225, 475, 1033, 2475, 4975 and 20000 years. These PGA will be used as Characterization of
Earthquake Loading and SPT-N value as resistance characterization of soil for liquefaction Initiation.
Compared to traditional methodologies, the performance-based procedures provide more precise and
reliable predictions of the likelihood of liquefaction in regions with varying seismicity. In order to provide
real-time maps of the findings, the study will attempt to python scripting model and Q-gis.

Keywords: probabilistic seismic hazard, Liquefaction hazard, surface manifestations, Seismicity,

Earthquake, soil boring, Probabilistic liquefaction models, PSHA – Peak ground acceleration.

1. Introduction
Nepal lies in seismically very active zone with records of occurrence of about 17 percent of largest earthquakes
in the world. It is located on the subduction region formed by the convergence of Indian and Eurasian plates at
the depth of 4-18 km with low dip angle of about 10 degree, where the Indian plate is moving towards the
Eurasian plate at the rate of 30-40 mm/year resulting in the accumulation of large amount of strain energy
which can generate Mw 8+ earthquakes. With 800 out of 2400 km of the Himalayan range falling within Nepal's
boundaries, the country's narrow width of about 193 km can be divided into various zones: Indo-Gangetic
Plain, Sub-Himalayan, Lesser Himalayan, Higher Himalayan, and Tibetan Tethys. These zones are separated
by four fault systems: Main Central Thrust (MCT), Main Boundary Thrust (MBT), Himalayan Frontal Thrust
(HFT), and South Tibetan Tethys system (STDS) (Figure 3).These fault systems exhibit significant seismic
activity, with historical magnitudes ranging from 6.5 to 8.0. MFT, separating Sub-Himalayan and the Indo-
Gangetic Plain, is a very active thrust which subsume several active faults like SanguKhola fault, Jumla Fault,
140 km Long Barigadh faults, 17 km Long Dorpatan Fault, 15 km long JimrukKhola Fault, 10 km Long
Kulekhani Fault, 7 km Long Sunkoshi-Roshi Fault and historically, the maximum recorded magnitude in this
zone is 6.5. MBT, which is the boundary between Sub-Himalayan and the Lesser Himalayan,includes 80 km
long, Rangunkhola Fault,120km long Surkhet Ghorahi Fault, 60 km long Arun Khola Fault, 40 km long
Hetauda faults, 85 km long Udayapur faults and SaptakoshiMechi fault running from Dharan to Mechi.
Historically, the maximum recorded magnitude in this zone is 8.0. MCT is another active thrust, lies between
1
the Lesser Himalayan and Higher Himalayan, and major faults identified are Dharma Fault, 10 km long Talphi
fault, 20 km long tibrikot fault, 20 km long Dhaulagiri faults, Thakkhola fault, Bari-gad faults. Historically, the
maximum recorded magnitude in this zone is 7.5. (Bhusal et al. 2019)
Earthquake is one of the most devastating natural disasters which can cause lots of damages within few
minutes in terms of human life and properties. Since 1255, Nepal has experienced many earthquakes of
magnitude greater than or equal to 5 magnitude. Among them major earthquakes greater than Magnitude of 6.5
occurred in 1255 Jun. 7, 1505 Jun. 6, 1720 Jul. 25, 1833 Aug. 26, 1911 Oct. 14, 1916 Aug. 28, 1934 Jan. 15,
1936 May 27, 1945 Jun. 4, 1950 Aug. 15, 1954 Sep. 4, 1988 Sep. 20, 2011 Sep. 18, 2015 April 25, and 2015
May 12. Two strong shocks– M7.8 and M7.3 of Recent Gorkha Earthquake 2015 (April 25 and May 12)
followed by series of aftershocks prompted devastating impact claiming lives and damages of properties in
Nepal. These earthquakes have wide range of effects including soil liquefaction, especially in Kathmandu
valley. The Kathmandu Valley in Nepal is characterized by a geological composition of Precambrian to
Devonian rock formations in the basement and surrounding hills, overlain by Quaternary sediments and recent
alluvium in the valley. The sediment deposits can reach up to 500-600 meters in thickness. The valley's sandy
and alluvial soils, combined with shallow groundwater depths ranging from 0.5 to 5 meters below the surface,
make it highly susceptible to liquefaction during seismic events. Past earthquakes, such as those in 1833 and
1934, caused significant liquefaction damage in the valley. However, during the 2015 Gorkha Earthquake
(Mw7.8), liquefaction was limited and localized to certain areas. Interestingly, some sites that had liquefied in
the past did not exhibit liquefaction during the 2015 earthquake, indicating variations in liquefaction
susceptibility. Liquefaction effects pose significant hazards to infrastructure, including transportation
networks, bridges, and lifelines. Effects such as sand boils, lateral spreading, settlement, and landslides can
cause damage and disruption. Notable events like the 2011 Christchurch earthquakes, the Tohoku earthquake,
1964 Nigata earthquake and 1934 Bihar-Nepal earthquake highlight the need to understand, analyze and
mitigate liquefaction hazards. The ability to forecast liquefaction is crucial in the field of civil engineering.
Predicting the possibility of seismic risks like liquefaction and the harm they cause will help to save lives, cut
down on losses, and raise living standards for people all across the world. Understanding their sources and the
circumstances in which they will exist is crucial for understanding and predicting dangers. Because we don't
fully understand the phenomena, it has been impossible to predict when liquefaction would occur. Recent
improvements in the amount of data available to researchers as well as our understanding of what causes
liquefaction have resulted in the creation of new and improved empirical models for forecasting liquefaction
and its likelihood of occurring (Cetin, Seed et al. 2004; Boulanger and Idriss, 2012; Juang, Ching et al. 2012).
The purpose of the study described is to use a method that uses Standard Penetration Test (SPT) data
from site soil borings and PGA value from PSHA of Nepal to perform performance-based probabilistic
liquefaction hazard analysis of the Kathmandu. The use of performance-based techniques to analysis and design
is another new breakthrough in engineering. Performance-based methods to engineering have an edge over
more conventional deterministic ones because they let the engineer compartmentalize and systematically
account for uncertainty. Using the framework created by Kramer and Mayfield (2007), this study will compare
two performance-based liquefaction initiation models: the probabilistic Cetin et al. (2004) model and the
probabilistic Boulanger and Idriss (2012) model. Lack of essential research and evaluation in several key areas
regarding liquefaction hazards is one of major issue in the Kathmandu valley. Some of the research gaps include
the absence of a real-time map, the absence of performance-based liquefaction evaluation, the lack of a study
on the timing of liquefaction and its utility in hazard evaluation, the absence of liquefaction return period
analysis specific to Kathmandu Valley, and the failure to consider the effect of structures within the study area.

2
 2. Data and Methods

2.1 Study Area

Nepal lies within 26.5-30.5N latitude and 80-89E longitude, towards the southern collisional boundary of the
Indian and the Eurasian Plate, converging at 19 mm per year (Jouanne et al. 2004). The Proposed area for
liquefaction susceptibility evaluation is Kathmandu valley, which is located in Nepal's central region, Bagmati
province with coordinates ranging from 85°11′34"E to 41 85°31′19"E longitude and 27°32′16"N to
27°48′47.75"N Latitude (Khatakho et al. 2021). The Kathmandu Valley comprises three main historical cities
Kathmandu (the capital of Nepal), Lalitpur, and Bhaktapur. The Kathmandu valley includes the metamorphic
nappe and the sediments above it that are fossiliferous (Stöcklin & Bhattarai, 1981). Late Pliocene to
Pleistocene thick basin-filled sediments occupy the Kathmandu valley (YOSHIDA, 1984). Fluvio-lacustrine
deposits that are quite thick cover its basin itself. Lacustrine deposits, formed by the mechanical and chemical
weathering of rocks encircling the adjacent hills of the valley, predominate in the southern half of the valley.
The worn materials were carried by several carrying organizations before being eventually laid out on the
lakebed. As a result, the finished product is typically a clay layer covered in sand and silt overburden. As the
valley composed of Sandy and alluvial soils in most parts and Groundwater in the valley is found at shallow
depths ranging from 0.5 to 5 m below the ground surface, which attributes the Kathmandu valley to be highly
susceptible to liquefaction during seismic events.

Figure 1(a) - Map showing study area

3
 Figure 1(b) - Map showing location of borehole data (From Subedi et al. 2022)

2.2 Theory of planned behavior: Conceptual Framework

The proposed research/study consists of 2- phase action plan.

Phase 1- Earthquake loading determination (PGA & Mw):
Probabilistic seismic hazard analysis (PSHA) will be performed using 23 areal zones and 14 linear
seismic sources in R-crisis software to obtain following outputs: De-aggregation analyses of PGA and
magnitude (mean/modal value) and magnitude - dependent seismic hazard curves based on PSHA
Phase 2- Liquefaction initiation evaluation:

Two Probabilistic model (Simplified Procedure): 1.) Cetin et al. model, 2.) Boulanger and Idriss
Probabilistic Model will be used for prediction of liquefaction initiation of study area. PGA obtained
from PSHA of Nepal will be used as Characterization of Earthquake Loading and SPT-N value as liquefaction
resistance characterization of soil

4
 Figure 2- Flowchart of conceptual framework

2.3 Probabilistic Seismic hazard analysis of Nepal

For PSHA of Nepal, logical tree approach will be used by considering 23 areal zones and 14 linear
seismic sources (active faults) in R-crisis software. Earthquake catalogue are collected from different sources
start from 1255 to 2023 May having magnitude greater than 4 Mw. All the data are converted into moment
magnitude by appropriate relationship, repeated events are removed manually. Minimum magnitude considers
as 4.5 Mw and maximum magnitude calculated from empirical relation developed by researcher. Declustering
using Gardner and Knoff (1974) windowing algorithm and assume that the earthquake catalogue followed
possion’s distribution. The seismicity of the Nepal is very complex and no also no any specific ground
attenuation relationship is derived so, in the current research mix type of GMPEs derived for active shallow
crustal and subduction interference wiil be used by giving the weightage by considering the uniform velocity
760 m/sec. Entire area of Nepal will be divided into 1.2 * 0.6 degrees grid size and the peak ground acceleration
(PGA) will be calculated for the return period of 10, 108, 225, 475, 1033, 2475, 4975 and 20000 years.

2.3.1 Earthquake Catalogue, Declustering Earthquake Catalogue and Completeness

Nepal lies within 26.5-30.5N latitude and 80-89E longitude and thus, all the historical earthquake data
within the area till 2023 May 23 are collected. Earthquake catalog provides the basic data for delineating
seismogenic sources and computing the seismicity parameter, especially the mean seismic activity rate λ, the
Gutenberg-Richter b value, and the maximum expected earthquake magnitude Mmax. The earthquake catalogue
5
are formed merging the data from U.S. Geological Survey, Department of geology and mines, International
seismological center (ISC), and various Literatures including Parajuli 2015, Thapa and Guoxin. The repeated
catalogue are removed manually. The study will consider the incompleteness of the earthquake catalog which
includes both historical earthquakes and the instrumental catalog, with different levels of completeness (Kijko
and Smit, 2012; Kijko et al., 2016). Ambraseys and Douglas (2004) and Szeliga et al. (2010) provide
information regarding historical major earthquakes for Himalaya and adjacent regions. The different types of
magnitude (e.g., body wave magnitude, surface wave magnitude, duration magnitude, and local magnitude)
contained in the catalog will be converted into moment magnitude 'MW ', using the different empirical
relationships shown in the Table 1. In this study, minimum magnitude will be taken as 4.0 since quakes lesser
in magnitude are less significant from engineering point of view. Maximum magnitude depends on fault
dimensions and stress drop, and most of the empirical scaling relations are developed based on this assumption,
developing the relations as the function of rupture area and slip length.
Table 1- Magnitude scaling relations used for conversion of different types of magnitude into MW

Catalogue of main shocks can be used in estimating seismic risk by virtue of statistical model when
aftershocks are removed from total event listing and assumed that the earthquake occurrence the
Poisson’s distribution. There are several declustering algorithms that have been proposed over the
years. Up to now, most users have applied either the algorithm of Gardner and Knoff (1974) or
Reasenberg (1985), mainly because of the availability of the source codes and the simplicity of the
algorithms. In this study, we are planning to use technique proposed by Gardener and Knop off 1974
was applied to identify aftershocks.

2.3.2 Earthquake Source Models

Proper seismic source zoning is essential in PSHA. Seismogenic sources with different characteristics are
generally delineated according to tectonic settings, seismological and geological attributes, and fault
characteristics. In the contest of Nepal, by considering different sources in hazard analysis, different Literatures
have provided different value of PGA for same locations. The major reason behind this discrepancy is the lack
of enough information about the sources of earthquakes. So, this study will use a mixed types of source model
that combines three different seismic sources: areal, linear, and smoothed gridded sources.
The area of a seismic source zone is delineated using the basic principle that seismicity within a single
source zone is uniform and homogenous. It is therefore assumed that every point within the source has an equal
potential to generate a future earthquake (Baker, 2015). Recent studies (Chaulagain et al., 2015; Ram and Wang
GX, 2013) have delineated 23 areal seismic source zones (Figure 3) for the study region based on quantitative
analysis of the earthquake distribution, fault information, and seismotectonic settings. In this study, the same
(23) seismic areal source zones will be utilized. Usually, major earthquakes are associated with active faults,
and in seismic hazard assessment these active faults are considered as linear seismic sources. Although, there
are numerous methods for earthquake source simulation (e.g., Zhang WB et al., 2006), for the Himalayan-
Tibetan region, Styron et al. (2010) identified active faults using remote sensing imagery, aerial photographic
interpretation, paleoseismic studies, and field investigations. For our study region, 14 active faults will be

6
assigned as linear seismic sources. The areal and linear sources considered in the research are depicted in figure
3 below.

Figure 3. Map of the study area showing seismic sources and the spatial distribution of declustered earthquakes
(red open circles) and some historical earthquakes (MW>6.0) (black filled stars). The yellow star marks the
location of the recent devastating 2015 MW7.8 Gorkha Earthquake. Brown rectangles are seismic source zones
numbered from Z1 to Z23. Purple solid lines are active faults modified from Styron et al. (2010). The
numbering of faults from f1–f8 is arbitrary. MFT: Main Frontal Thrust, MBT: Main Boundary Thrust, MCT:
Main Central Thrust, STD: South Tibetan Detachment.
Table 2 - Active faults parameters that will be used in this study

7
2.3.3 Maximum Earthquake Magnitude and Focal Depth
The maximum expected earthquake magnitude (Mmax) for every seismic source zone will be computed on the
basis of historical (e.g., 1505 MW 8.2 and 1934 MW 8.2 earthquakes) and recent earthquake data, but also
considering great earthquakes that occurred in nearby seismic zones with similar tectonic settings. The estimate
of Mmax is computed following the algorithm of Kijko and Singh (2011). This estimation takes into account the
uncertainty in the earthquake magnitude determination. For smoothed gridded seismicity, the same range of
Mmax will be used for the respective grids. Accurate determination of earthquake location along the mountain
belt is usually difficult because of insufficient station coverage. The ISC catalog reveals that earthquakes are
distributed throughout the thickness of the Tibetan and Indian crust ranging between 0 and 90 km. Moreover,
some recent studies have found that the hypocenters of most of the earthquakes in Nepal are located at depths
ranging between 0 and 30 km (Pandey et al., 1999; Priestley et al., 2008). In addition, hypocenter relocation
results from a nearby broadband seismic array found that the focal depths for the Gorkha earthquake sequence
ranged primarily between 10 km and 20 km (Adhikari et al., 2015; Bai et al., 2016; Letort et al., 2016). Thus,
in this analysis, the average focal depth of 15 km will be assigned as a constant depth for all the source models.

2.3.4 Seismicity Model and Parameters evaluation

To estimate the earthquake magnitude exceedance rates, a Modified Gutenberg-Richter (MGR)-Poisson model
will be chosen as a seismicity model for all seismogenetic sources except characteristic earthquakes. Herein,
the seismicity is expressed as (Cornell and Van Marke, 1969):

where, λ0 is the rate of earthquake activity with threshold magnitude M0 (here M0 = 4.0), β is a parameter
equivalent to the b value (β = 2.303×b), and Mmax is the maximum expected magnitude for the sources. In
computing the λ(M) value, the uncertainty of both Mmax and β are taken into account.
In this study, the essential seismicity parameters will be computed by taking into account the incompleteness
of the instrumental catalog and the historical earthquake records. The Gutenberg-Richter b value and its
standard error for every source zone are determined following the joint maximum likelihood function of Kijko
and Smit (2012) based on the historical and instrumental earthquake catalog. This new approach is the
extension of the Aki-Utsu (Aki, 1965; Utsu, 1965) b value estimator which takes into account both
incompleteness of the earthquake catalogs and temporal variation of seismicity. It provides straightforward
approximations for the standard errors and confidence intervals (Kijko and Smit, 2012).

2.3.5 Attenuation of Ground Motion

The ground motion prediction equations (GMPEs), also known as attenuation relationships, are essential in
estimating the ground motion parameters (e.g., PGA, SA). These equations usually give ground motion
intensity as a function of many variables such as earthquake magnitude, source-to-site distance, and soil
condition. In general, the attenuation relationships are obtained empirically using a statistical regression method
on a particular set of strong motion data.
Since the no definite GMPEs is developed in the contest of Nepal, analysts are obligated to use different
GMPEs developed for subduction zone, active shallow crustal regions in other parts of the world. The complex
seismo tectonic setting of the Himalayan orogenic belt further makes the selection process of GMPEs a very
difficult task for analysist. Jonathan et al. developed the selection of ground motion prediction equations for
the global earthquake model and they suggested to review the GMPEs derived data (local or global), saturation
with magnitude, magnitude dependent distance scaling and mimic effects anelastic attenuation etc. before
selection of GMPEs. For proper estimation of hazard, the latest ground motion attenuation relationship must
8
be used which characterized the seismic sources by the details of the fault – rupture model. Various researchers
have used different GMPEs for hazard analysis of Nepal. (BECA, 1993) used attenuation law Kawashima et
al. (1984), Pandey et al. (2002) and Maskey (2005)have used Youngs et al. (1997) GMPE. Pandey et al. (2002),
in their studies, made the GMPE to fit with approximately horizontal acceleration recorded in Uttarkashi
earthquake of Mw =6.8 and Chamauli earthquake of magnitude Mw = 6.6. Similarly, Thapa and Guoxin (2014)
used attenuation relationship developed by CEA (2005). Chaulagai et al. (2015) have used Boore and Atkinson
(2008), Chiou and Youngs (2008), Cambell and Bozorgnia (2008), Boore and Atkinson (2003) and Youngs et
al. (1997) are used in logical tree for seismic hazard. Parajuli (2015) used Youngs et al, Gregor et al (2002),
Boore and Atkinson (2008), Kanno et al (2008) and zhao et al (2006) attenuation model for his research. V.L
stevens et al. (2018) used 4 GMPEs with equal probability; Abrahamson et al (2016), zhao et al (2006), Boore
and Atkinson (2003), Boore et al (2014) for subduction interface events and Asimki (2017), Chiou and Youngs
(2014), Boore et al (2014), Chiou and Youngs (2008), Boore and Atkinson (2008) considered for active shallow
crustal region of Nepal. Rahaman and Bel (2018) have used GMPEs considering for both active shallow crustal
and subduction interface. In their study they have used Abrahamson et al (2014), Ambraseys et al (2005), Chiou
and Youngs (2014) and Zhao et al. (2006) for active shallow crustal and Abrahamson etal. (2016), Boore and
Atkinson (2003), Youngs et al. (1997), Zhao et al (2006) for subduction interface by giving weightage. Eleven
different GMPEs, which are used by previous researchers for the hazard analysis of Nepal, will be incorporated
in the present study by the use of logical tree structure which is shown in figure 4.

2.3.6 Logic Tree Structure

In this study, three types of seismogenic source models and two sets of GMPEs (each consisting of four dif-
ferent GMPEs) will be combined using the logic tree. Equal weights (1/3) will be assigned to each of the source
models, as there is no sub-stantial observed data for the study region in evaluating the ranking of the weight.
For the linear source model, equal weight (1/2) will be assigned for the Characteristic Earthquake and
Gutenberg-Richter seismicity models as there is no reason to prefer one model over the other. Figure 4 displays
the standard logic tree framework used for the linear source model in this analysis.

Figure 4: Systematic Logical Tree for Hazard calculation

9
2.4 Liquefaction evaluation of Kathmandu valley
The history shows that valley has experienced extensive liquefaction during past earthquakes,
including those in 1833 and 1934, which caused significant damage. Although the 2015 earthquake
(Mw7.8) resulted in limited liquefaction damage, surface manifestations were observed in various
parts of the valley. The survey team (Subedi et.al) observed (after mainshock 7.8 Mw april 25) that
the liquefaction triggered by 2015 Gorkha Earthquake appears to be very limited and localized (The
localized areas where liquefaction observed are manamaiju, ramkot, bungamati, hattibanm imadol,
mulpani and duwakot ). Total of 25 sites were observed liquefied during Gorkha Earthquake 2015 ,
out of which 15 sites were either within or on the valley edges .The intersting fact is that when these
sites revisited (after Mw 7.3 aftershock may 12), none of the site reflected the recurrent liquefaction
phenomenon. The Tundikhel area liquefied previous 1934 earthquake found not liquefied this time
2015 earthquake.
2.4.1 Liquefaction susceptibility and Initiation
A soil is considered susceptible if it is able to liquefy. Some soils will not liquefy under almost any
circumstances and others cannot liquefy due to the conditions and environment in which they are found. To
understand which soils are susceptible to liquefaction and which soils are not, it is constructive to examine
some of the criteria that affect liquefaction. These criteria include historical, geological, compositional, and
state criteria (Kramer 1996).The evaluation of liquefaction susceptibility is particularly challenging when
dealing with fine-grained soils of marginal plasticity and coarse-grained soils with high fines content. Recent
observations of liquefaction in fine-grained soils, despite being classified as non-susceptible according to the
Chinese criteria (Seed and Idriss, 1982), have prompted extensive research on the liquefaction susceptibility of
such soils. Two significant studies have been conducted to establish criteria for assessing the liquefaction
susceptibility of fine-grained soils. These criteria demonstrate consistency in many conditions but exhibit
variations in others, highlighting the inadequacy of the Chinese criteria. Both studies utilized field observations
and laboratory tests, conducted by renowned experts in the field of geotechnical engineering. According to the
Chinese criteria as reported by Kramer is defined as (Kramer 1996), soil may be liquefiable if: Fraction finer
than 0.005 mm ≤ 15%, Liquid Limit (LL) < 35%, Natural water content (wc) ≥ 0.9LL and Liquidity Index (LI)
≤ 0.75. However, the Chinese Criteria has come to be viewed as obsolete (Bray and Sancio 2006), and newer
criteria have been developed by Bray and Sancio, stating that a soil may be liquefiable if: Plastic Index (PI) <
12 and Water content (wc) / LL > 0.85. The Kathmandu Valley is characterized by two distinct geological
successions. The basement and surrounding hills consist of Precambrian to Devonian rock formations, while
the valley is overlain by Quaternary sediments and recent alluvium. The sediment deposit thickness in the
central part of the basin reaches up to 500-600 meters. Different authors have classified the stratigraphy of the
sediment deposits, with Sakai (2001) proposing a new classification comprising Tarebhir, Lukundol, Itaiti,
Bagmati, Kalimati, Patan, Thimi, and Gokarna Formations in the southern, central, and northern parts of the
groundwater basin, respectively. As the valley composed of Sandy and alluvial soils in most parts and
Groundwater in the valley is found at shallow depths ranging from 0.5 to 5 m below the ground surface, which
attributes the Kathmandu valley to be highly susceptible to liquefaction during seismic events.
Liquefaction initiation, also known as liquefaction triggering, occurs when specific conditions align with the
susceptibility of a soil. Liquefaction triggering is associated with the buildup of pore water pressure equaling
the effective confining stress, but various definitions exist for this point. Cetin and Bilge (2012) found a range
of definitions for liquefaction triggering based on pore pressure ratio (ru) and cyclic strain (γ). The variation in
definitions is partly due to the fact that the onset of liquefaction varies among different soil types, with
contractive soils liquefying at smaller shear strains. As liquefaction initiates, the strength of the soil decreases.
10
Contractive soils can develop increased pore pressure, leading to a decrease in overall strength, while dilative
soils can develop negative pore pressures, enhancing their strength and making them less prone to liquefaction.
Understanding the conditions that cause liquefaction to initiate is crucial alongside assessing its susceptibility.

2.4.2 Characterization of Seismic / Earthquake Loading

The cyclic stress ratio is most commonly represent the earthquake loading and evaluated using the “simplified
method” first described by Seed and Idriss (1971), which can be expressed as

Where, amax = peak ground surface acceleration, g = acceleration of gravity (in same units as amax), σ vo = initial
vertical total stress, σ'vo = initial vertical effective stress, rd = depth reduction factor, and MSF = magnitude
scaling factor, which is a function of earthquake magnitude. It should be noted that two pieces of ground motion
information, amax and magnitude, are required for estimation of the cyclic stress ratio. The PGA, obtained from
PSHA of Nepal will be used to compute CSR for different return period.

2.4.3 Characterization of Liquefaction Resistance

The cyclic resistance ratio is generally obtained by correlation to insitu test results, principally from
standard penetration (SPT), cone penetration (CPT), or shear wave velocity (Vs) tests. Of these, the
SPT has been most commonly used and will be used in the remainder of this Study. A number of SPT-
based procedures for deterministic (Seed and Idriss, 1971; Seed et al., 1985; Youd et al., 2001, Idriss
and Boulanger, 2004) and probabilistic (Liao et al., 1988; Cetin et al., 2004) estimation of liquefaction
resistance have been proposed. The borehole data with SPT-N counts will be collected from NGS
office and KUKL.

2.4.4 Probabilistic Assessment of Liquefaction Initiation Potential

Following two commonly used empirical approaches will be used for this study.
a.) Cetin and Seed et al. Probabilistic model

Figure 5: SPT-based probabilistic correlations for the CRR of sands for M7.5 and σv =1 atm: Cetin et al.
(2004)
11
Where,

\
Table 3: Posterior statistics of model parameters, accounting for measurement errors (Cetin et.al 2004)

b.) Boulanger and Idriss Probabilistic model

Figure 6: Curves of CRRM7.5, σv =1 atm versus (N1)60cs for probabilities of liquefaction of 15, 50, and
85% (Idriss and Boulanger, 2008)

12
 Where,

ɸ-1 is the inverse of the standard cumulative normal distribution, and PL is the probability of
liquefaction.

3. Expected Results and discussion

a.) Probabilistic liquefaction hazard analysis of Kathmandu valley and Liquefaction susceptibility
dynamic or real-time map for different earthquake scenarios (varying MW and PGA).

Figure 7- Typical liquefaction mapping

13
 b.) Probabilistic Seismic Hazard Analysis of Nepal and hazard map of spectral acceleration

Figure 8 (a) - Tentative expected output of PSHA model (seismic hazard curves and disaggregation of
magnitude and PGA)

Figure 8 (b) – Typical Hazard maps of spectral acceleration

14
 References

Mori, T., Shigefuji, M., Bijukchhen, S., Kanno, T., & Takai, N. (2015). Ground motion prediction
equation for the Kathmandu Valley, Nepal based on strong motion records during the 2015 Gorkha
Nepal earthquake sequence.
Rahman, M. M., & Bai, L. (2015). Probabilistic seismic hazard assessment of Nepal using multiple
seismic source models.
Shrestha, S. (2015). Probabilistic seismic hazard analysis of Kathmandu City, Nepal.
Marasini, N., & Okamura, M. (2014). Numerical simulation of dynamic centrifuge tests on seismic
behavior of residential building on liquefiable foundation soil in Kathmandu Valley.
Bhandary, N. P., Marasini, N., Okamura, M., & Adhikari, S. R. (2015). Investigation of soil
liquefaction in Kathmandu Valley (through in-situ (SPT, BPT tests) or lab testing data (cyclic simple
shear test, cyclic triaxial test)).
Subedi, M., et al. (2004). Probabilistic liquefaction potential of Kathmandu soil based on standard
penetration tests.
Kramer, S. L., Sideras, S. S., & Greenfield, M. W. (2012). The timing of liquefaction and its utility in
liquefaction hazard evaluation.
Kramer, S. L., & Mayfield, R. T. (2010). Performance-based liquefaction hazard evaluation.
Franke, K. W., et al. (2014). Simplified standard penetration test performance-based assessment of
liquefaction and effects.
Kramer, S. L., & Mayfield, R. T. (2010). Return period of soil liquefaction.
Mayfield, R. T., Kramer, S. L., & Huang, Y. (2014). Simplified approximation procedure for
performance-based evaluation of liquefaction potential.
Seed, et al. (2001). Recent advances in soil liquefaction engineering and seismic site response
evaluation.
Kramer, S. L., et al. (2016). Modeling soil liquefaction hazards for performance-based earthquake
engineering.
Alavi, A. H., et al. (2008). Energy-based numerical models for assessment of soil liquefaction.
Qiu, Z., & Jinchi, J. (2005). OpenSees three-dimensional computational modeling of ground-structure
systems and liquefaction scenarios.
Arulanandan, K., & Li, X. S. (2002). Numerical simulation of liquefaction-induced deformations.
Shahir, H., & Pak, A. (1997). Estimating liquefaction-induced settlement of shallow foundations by
numerical approach.

15
Glaser, S. D., & Chung, R. M. (1995). Estimation of Liquefaction Potential by In Situ Methods.
Earthquake Spectra, 11(3), 431.
Holtz, R. D., & Kovacs, W. D. (1981). An Introduction to Geotechnical Engineering. Prentice Hall.
Ishihara, K. (1985). Stability of natural deposits during earthquakes. In Proceedings of the 11th
International Conference on Soil Mechanics and Foundation Engineering (Vol. 1, pp. 321-376).
Kramer, S. L. (1996). Geotechnical Earthquake Engineering. Prentice Hall.
Kramer, S. L., & Seed, B. H. (1988). Initiation of Soil Liquefaction Under Static Loading
Conditions. Journal of Geotechnical Engineering, 114(4), 461-478.
Mohamad, R., & Dobry, R. (1986). Undrained monotonic and cyclic triaxial strength of sand.
Journal of Geotechnical Engineering, 112(10), 941-958.
Vaid, Y. P., & Chern, J. C. (1983). Effect of static shear on resistance of liquefaction. Soils and
Foundations, 23(1), 47-60.
Zehgal, M., & Elgamal, A. W. (1994). Analysis of Site Liquefaction Using Earthquake Records.
Journal of Geotechnical Engineering, 120(6), 996.
Zorapapel, G. T., & Vucetic, M. (1994). The Effects of Seismic Pore Water Pressure on Ground
Surface Motion. Earthquake Spectra, 10(2), 403.
http://ceor.princeton.edu/~radu, Soil Dynamics and Geotechnical Earthquake Engineering at Princeton
University

http://ceor.princeton.edu/~radu/soil/velacs, Velacs Research Project.

http://geosystems.gatech.edu/Research/gpr.html, Georgia Institute of Technology, Liquefaction research.
http://nisee.ce.berkeley.edu, National Information Service for Earthquake Engineering
http://peer.berkeley.edu, Pacific Earthquake Engineering Research Center.
http://quake.wr.usgs.gov, United States Geological Survey Earthquake Information
http://rccg01.usc.edu/GEES/velacs/velacs.html, Velacs Research Project.
http://wrgis.wr.usgs.gov, USGS, Western Region Geologic Information Server.
http://www.abag.ca.gov/bayarea/eqmaps/liquefac/bayaliqs.gif, Hazard map for the bay area in San
Francisco, ABA 1983

16
17