Correlation of excess pore water pressure ratio on flow liquefaction phenomenon in Sibalaya – Central Sulawesi Province

. On September 28, 2018, liquefaction in Sibalaya damaged irrigation canals and hit 51.2 hectares fields and roads. This phenomenon was triggered by an earthquake of magnitude 7.5 that shook Sigi Regency, Central Sulawesi Province. As an increase in soil pore water pressure is one of the causes of liquefaction, the pore water pressure ratio ( 𝑟 𝑢 ) at the study site was analyzed. Then, this figure is used as a point of reference to determine the liquefaction potential. Standard Penetration Test (SPT) data and laboratory tests from four boreholes resulting from soil investigations in 2021 with a maximum depth of 20 meters were utilized in this study, along with microtremor data down to the depth of bedrock from test results in 2023. Based on the data, the liquefaction potential was assessed using the one-dimensional nonlinear site response approach, the GQ/H + PWP model, and the DEEPSOIL V.7 program. At multiple layers of boreholes, 𝑟 𝑢 is more than or equal to 0.8, indicating that the Sibalaya liquefaction flow area is still susceptible to liquefaction.


Introduction
Sibalaya, located in the Sigi Regency of the Central Sulawesi Province, exhibits a pronounced susceptibility to liquefaction, as indicated by its classification as a high liquefaction vulnerability zone [1].Depicts the Pakuli Formation, in which Sibalaya is encompassed, as presented in the Systematic Geological Map of The Pasangkayu Quadrangle Sulawesi [2].This geological formation, characterized as a young Holocene sedimentary rock, possesses a higher potential for liquefaction compared to Pleistocene sedimentary rocks [3].
On September 28, 2018, an earthquake measuring 7.5 Mw caused flow liquefaction in Sibalaya, thereby highlighting the continued activity of the Palu Koro Fault [4], situated approximately six kilometers away.This factor contributes to the heightened likelihood of future recurring liquefaction events in the area.In contrast to Jono Oge, Petobo, and Balaroa, Sibalaya features a lower overall number of residential structures and lacks mass graves for victims [5].The impact of flow liquefaction spans a 51.2-hectare area in Sibalaya, characterized by landslide crowns observed in the Gumbasa irrigation canal.Fig. 1 illustrates the extent of the region affected by flow liquefaction.Subsequent to the liquefaction event, excess pore water pressure *Corresponding author: fikri.faris@ugm.ac.id dissipates within the liquefiable layer, leading to land subsidence [6].Previous studies have assessed the potential for liquefaction in the Gumbasa irrigation canal, utilizing data from the N SPT (Standard Penetration Test) values and laboratory test results conducted by Andiny et al. [7], Pratama et al. [8], and Widyatmoko et al. [9].The analysis of liquefaction potential employed a simplified procedure developed by Idriss and Boulanger [10].In this study, the N SPT values and laboratory test results are based on soil investigation data from 2021, supplemented with additional microtremor data.The liquefaction potential is evaluated by examining the increase in pore water pressure during an earthquake, following the methodology employed by Mei [11] and Jalil et al. [12].The accumulation of excess pore water pressure during seismic shaking, causing a decrease in effective stress, leads to liquefaction [11].Consequently, the Palu earthquake resulted in the accumulation of excess pore water pressure, triggering a loss of soil strength and stiffness and resulting in extensive damage [12].Although previous research has examined the relationship between increased pore water pressure and liquefaction, no studies have specifically investigated the occurrence of flow liquefaction in Sibalaya.
This study presents the PWP ratio (  ) using the open-source software DEEPSOIL V.7 to determine the liquefaction potential at flow liquefaction in Sibalaya.This method serves as a reference for assessing liquefaction potential attributed to the influence of excess pore water pressure induced by seismic activity.Due to the unavailability of ground motion data for the Palu earthquake on September 28, 2018, Jalil et al. [12] utilized the ground motion data from the Kobe earthquake available in the DEEPSOIL software.Similarly, in this study, synthetic motion from the Amberley earthquake in New Zealand is employed.Future research endeavors aim to address these limitations to ensure a more accurate representation of the actual conditions.

Material study
In this study, the liquefaction potential was analyzed at four specific boreholes situated in Sibalaya, namely borehole BM 02, borehole BM 04, borehole BM 05, and borehole BM 06.The spatial distribution of these boreholes is illustrated in Fig. 1.The land investigation data utilized in this analysis includes N SPT values and laboratory test results obtained from studies conducted by the Ministry of Public Works and the Housing Republic of Indonesia in 2021.The soil investigation data collected through core drilling is limited to a depth of 20 meters.To supplement this data, an additional soil investigation was carried out using a microtremor method conducted by the Palu Class I Geophysical Station of The Meteorology, Climate, and Geophysics Agency in 2023.
Microtremor investigations are carried out at the same point as the borehole, aiming to get soil condition data from the surface to the bedrock.Based on the results of microtremor measurements, the bedrock depth of each point is different.The value of bedrock shear wave velocity (Vs) is above 750 m/s [12].At BM 02, the bedrock depth is 125.5 meters with a Vs value is 806 m/s ; at BM 04, the bedrock depth is 119.4 meters with a Vs value is 759 m/s; at BM 05, the bedrock depth is 150.9 meters with a Vs value is 861 m/s ; At BM 06 the bedrock depth is 117.1 meters with a Vs value is 782 m/s.In this research area, the value of Vs 30 ranges from 249 m/s to 305 m/s.This value can be classified as the soil site class belonging to the medium class (SD) [13].
Two damping ratio values are used in the soil model in this study.For soil conditions from the surface to the soil boundary before bedrock, a damping ratio, D, equal to 5%, is used [12] .For bedrock characterized by a specific gravity of about 22 kN/m 3 and a Vs value of about 800 m/s, D equal to 0.5% is used [14].

Nonlinear site response analysis
In this study, the nonlinear site response analysis was conducted using the generalized/Hyperbolic constitutive soil model (GQ/H+PWP) method.The selection of the GQ/H soil model was based on several advantages it offers: it provides the maximum shear modulus at zero shear strain, offers the maximum shear stress as the shear strain approaches infinity, allows for adaptable control of nonlinear behavior within the boundary conditions, and maintains a formulation simplicity for site response analysis [15].The GQ/H+PWP model was previously developed by Mei [11].A series of shear tests were performed to evaluate and validate the Vucetic and Dobry (1986) [16] modifications.
For the One-Dimensional (1-D) site response analysis and potential liquefaction assessment, the DEEPSOIL V7.0 software was employed.This software facilitates the analysis of actual soil conditions, encompassing the soil surface down to the bedrock.The specific PWP model utilized depends on the soil type at the research location, with each PWP model requiring different input parameters.A comprehensive overview of the PWP models available in the DEEPSOIL software, along with the corresponding input parameters, is presented by Hashash [17] , as shown in Table 1.Based on the soil investigation at the study site, the Vucetic / Dobry PWP model was chosen because it is sand.PWP model park & ahn can also be used on the sand.However, it was not selected because this method is challenging to use.A cyclic shear test is required for each sand to calibrate the damage parameter at the initiation of liquefaction (D  ) and the calibration parameter (α) [18].The parameters used in the Vucetic/Dobry PWP model are considered easy to apply to sandy soils.As found in research by May et al. [18] explained that Vucetic / Dobry PWP model [16] developed a unique relationship between   , cyclic shear strain amplitude (ϒc), and multiple loading cycles (fNc).This relationship is the origin of the   calculation equation, as shown in Equation 1.
where  2, where Vs is the shear wave velocity of the coating and H is the layer thickness.To enhance   , The thickness of the layer must be decreased, and the minimum frequency for all layers is a minimum of 30 Hz [17].
Equations 5-6 of the Vucetic / Dobry PWP Model are influenced by the magnitude of Shear wave velocity (Vs) and Fine Content (FC) [11].Vs was derived using microtremor measurements taken at the exact research location at the position of the investigated borehole.In contrast, FC was collected by laboratory examinations of soil at the studied borehole. = 3810.s −1.55  (5)  = ( + 1) 0.1252 (6) The fitting parameter value for modulus degradation (ѵ) in Equation 7depends on the relative density value (  ) in per cent as obtained in Equation 8 [10] , with a minimum value of ѵ is 1 and a maximum value is 3.8 [11].
Jalil et al. [12], stated that the coefficient of consolidation (  ) of sand ranges from 0.02 to 0.1 m 2 /s; Therefore, the researchers correlated the range of   values for sand with sand properties based on the magnitude of the Vs values which were divided into several classes, namely loose, medium, dense, and very dense [13].Summarizing all of the model parameters for the analysis is presented in Tables 2-5, f = 2, p = 1,   = 0.05, and Max   = 0.95 applies to all layers.

Input ground motion
As it is known that strong earthquakes are one of the causes of liquefaction of flow and ground movement in Sibalaya, Sigi Regency, Central Sulawesi.So that the correct ground motion input when performing nonlinear analysis of PWP increase at the research site is very important.The ground motion used in this study is synthetic ground motion from the New Zealand Amberley Earthquake.This earthquake occurred on November 13, 2016, at 11:02:56 UTC, with a magnitude of 7.8 Mw and a depth of 23 Km.
Synthetic Ground Motion The Amberley earthquake was chosen because it is comparable to the two biggest earthquakes in Central Sulawesi in the last 30 years.The first was an earthquake measuring 7.9 Mw with a depth of 24 kilometres on January 1, 1996.The second was an earthquake on September 28, 2018, with a magnitude of 7.5 Mw and a depth of 20 kilometres.The history of the largest earthquake that occurred in the study area during the last 30 years is presented in Fig. 2.  In this study, the selection of synthetic motion also refers to the range of Peak Ground Acceleration (PGA) values.This value is based on the history of earthquakes over the last 30 years in the study area.The PGA value is calculated using the attenuation of Kanno [21], which considers the earthquake's depth, magnitude value, and distance to the epicentre.This attenuation equation can be seen in Equations 9-10.The synthetic ground motion has been adjusted to the location distance from the fault and the shear wave velocity value at the bedrock to obtain an appropriate target spectrum.Equation 9 for D ≤ 30 km and Equation 10 for D ≥ 30 km represents the Kanno attenuation [21].

PGAm calculation using kanno attenuation
This study calculates surface PGA (PGAm) values from six earthquake sources on Sulawesi Island.One earthquake incident was an earthquake that resulted in the worst impact in the last 30 years.Then the other five earthquake sources with a strength of more than 5 Mw occurred within a radius of 10 km from the research location.The PGAm is calculated using Kanno attenuation.The results of these calculations obtained the range of PGAm values.The lowest to the highest PGAm value based on the earthquake at the study site was 0.1 g to 0.67 g.Using the DEEPSOIL software reveals that the PGA input of the Amberley earthquake's synthetic ground motion is 0.5 g.BM 02 had a PGAm synthetic motion value of 0.33 g, while BM 04, BM 05, and BM 06 each had a value of 0.40 g, 0.39 g, and 0.55 g, respectively.The PGAm value of the Amberley earthquake falls within the range of PGAm values calculated by Kanno attenuation for the history of earthquakes at the study site.

Input nonelinear site response analysis and excess pore water pressure parameter
Based on soil investigation, it can be concluded that the four boreholes analyzed belong to the sandy soil type.This can also be seen in the soil type column from the soil investigation presented in Tables 2-5.In conducting a nonlinear site response analysis, the researchers wants to model the soil conditions according to the actual site conditions.Soil conditions from surface to bedrock are described with parameters appropriate to field conditions.For this reason, researchers used DEEPSOIL V.7 software to model the soil conditions at the site.The selection of the PWP model in this software follows the type of soil in the study area.The parameters must be entered according to the selected PWP model.These parameters are obtained from calculating equations and the results of soil investigations on the site.Some parameters must be entered into DEEPSOIL, which is presented In Tables 2-5.
The thickness of each layer of the soil profile must be considered when using the DEEPSOIL software.This affects the maximum allowable frequency value by Equation 2. The soil profile in the BM 2 consists of 45 layers with a total thickness of approximately 143.60 m.It includes 1.5 m of silt and sand on top, 4.5 meters of gravel: silt, gravel, conglomerate, 4.5 meters of clay and sand layer, 3 meters of sand and gravel, 6.3 meters of medium soil, soft rock hard soil, 105.7 meters of hard, and bedrock 18.1 meters.The depth of the groundwater table in the BM 2 is 9.1 m.The shear wave velocity profile, shear strength, and soil profile for each borehole are presented in Tables 2-5.
The soil profile in the BM 4 consists of 47 layers with a total thickness of approximately 157 m.It includes 1.5 m of Silt, gravel, and sand on top, 3 meters of gravel and sand, 4.5 meters of clay and sand layer, 3 meters of sand and gravel, 15.1 meters of Sand, Silt, and gravel, 4.5 meters of medium soil, 95.3 meters of soft rock hard soil, and bedrock 37.6 meters.The depth of the groundwater table in the BM 4 is 13.05 m.
The soil profile in the BM 5 consists of 49 layers with a total thickness of approximately 172.9 m.It includes 1.6 m of sand on top, 2.9 meters of sand, silt, and gravel, 15 meters of sand and clay layer, 2.4 meters of medium soil, 129 meters of Soft rock hard soil, and bedrock 22 meters.The depth of the groundwater table in the BM 5 is 16 m.
The soil profile in the BM 6 consists of 40 layers with a total thickness of approximately 145 m.It includes 1.6 m of gravel and silt on top, 10.4 meters of sand and gravel, 1.6 meters of sand and conglomerate layer, 1.4 meters of sand and gravel layer, 1.5 meters of sand and conglomerate layer, 3.1 meters of sand and gravel layer, 97.5 meters of Soft rock hard soil, and bedrock 27.9 meters.The depth of the groundwater table in the BM 6 is 18.65 m.
Carter and Bentley [22] correlated to determine the unit weight value in this investigation.The kind of soil determines the correlation between normal and wet conditions.As seen in the blue layer in Tables 2-5.The unit weight used for soil layers below the water table is believed to be wet.In addition, Equations 3-8 describe how to obtain the additional input parameters.
In numerous studies, soil liquefaction occurred before   = 1 [11].Hence a maximum   of 0.95 was determined.Jalil et al. [12] mentioned that the excess consolidation ratio (OCR) was set to 1, while the Plasticity Index (PI) was determined based on the findings of soil laboratory tests.Olson et al. [23] explains that Darendeli's [24] modulus reduction and damping curves were utilized for all sand, silty sand, and sandy silt layers.

PWP analysis results
After inputting the parameters of the soil investigation and inputting motion on the DEEPSOIL software, the pore water pressure ratio (  ) per soil layer is obtained.This study presents the   graphs value for significant layers is presented in Fig. 3-6.The graph represents the increase in   following an earthquake with a specific PGA value.
The maximum   in boreholes BM.02 is 0.95.According to some researchers, such as Jalil et al. [12] and Olson et al. [23], at   > 0.8, liquefaction was reached.The BM 02 has the potential for liquefaction at a 9.75 to 15,975 meters depth on layers 10-13.This can be seen in Fig. 3.At the same time, the groundwater level is at a depth of 9.1 meters.  began to increase to 0.8 when the earthquake shook at 40 seconds.The maximum   in boreholes BM.04 is 0.92.In borehole BM 04, liquefaction has potential at a depth of 15.75 to 17.25 meters.The   obtained was greater than 0.8, which fell in layer 15.This can be seen in Fig. 4. At the same time, the groundwater level is at a depth of 13.05 meters.The   began to increase to 0.8 when the earthquake shook at 55 seconds.
The maximum   in boreholes BM.05 and borehole BM 06 are 0.62 and 0.66.The   does not reach 0.8, meaning the borehole has no potential for liquefaction.When the earthquake shook at 120 seconds in the borehole BM 05,   went 0.6 at a depth of 15.75 meters, namely on layer 20.This can be seen in Fig. 5.While the groundwater depth of 16 meters.When the earthquake shook at 130 seconds, In BM 06,   reached 0.6.This condition is at a depth of 53.6 meters, namely at layer 27.This can be seen in Fig. 6.The depth of the groundwater table was 18.65 meters.
In the previous study, several researchers analysed the potential for liquefaction around the Gumbasa irrigation.One of which was Pratama et al. [8].This research was conducted in Jono Oge, Sigi, Central Sulawesi, 18 km from this research location.The research was performed using the Idriss-Boulanger simplified method.This study conducted Analysis at four borehole points; in previous studies, research was carried out at five.This study result had two boreholes with liquefaction potential and two without liquefaction potential.In comparison, in a previous analysis, three boreholes had low and moderate liquefaction potential, and two had no liquefaction potential.
Research on the liquefaction potential at Jono Oge, Sigi, Central Sulawesi, was also carried out by Jalil et al. [12] in one borehole.The study considers increasing PWP to identify liquefaction potential.Similar to this research, previous research was also assisted by DEEPSOIL software using the GQ/H+PWP method.The result of the last analysis is that there is potential for liquefaction at a depth of 16-17 meters from the ground surface, with a groundwater depth of 14 meters.In this study, the potential liquefaction at borehole BM 02 was 9.75 to 15.975 meters with a groundwater depth of 9.1 meters.Jalil et al. [12] stated that the liquefaction potential at a depth of more than six meters does not affect the surface.However, this will have an effect if cracks penetrate the inner layer.
The ground motion used in the previous research was the Kobe Earthquake which is the default from DEEPSOIL software with an input earthquake PGA value of 0.8 g.This study uses an Amberley earthquake synthetic ground motion.At the same time, it has been adjusted to the distance between the site and the Palukoro fault and the shear wave velocity value.In this study, the input earthquake PGA value was 0.5 g.This PGA value is based on the history of earthquakes in Central Sulawesi for the last 30 years.

Conclusion
This study presents an analysis of the liquefaction potential at four specific boreholes along the Sibalaya segment of the Gumbasa irrigation canal.The land investigation data utilized comprises N SPT values and laboratory test results conducted by the Ministry of Public Works and Public Housing of the Republic of Indonesia in 2021.Additionally, supplementary soil investigations using microtremor measurements were conducted by the Class I Geophysics Station of the Meteorology, Climatology, and Geophysics Agency in Palu in 2023.The assessment of liquefaction potential was carried out using a one-dimensional nonlinear site response approach, specifically employing the GQ/H + PWP model and the DEEPSOIL software.The findings of this study indicate a significant liquefaction potential in Sibalaya.
The analysis reveals liquefaction potential in borehole BM 02 and borehole BM 04, specifically at depths ranging from 9.75 to 15.975 meters and 15.75 to 17.25 meters, respectively, with corresponding groundwater depths of 9.1 and 13.05 meters.In contrast, no liquefaction potential is observed in borehole BM 05 and borehole BM 06, as the pore water pressure ratio (ru) did not reach the threshold of 0.8 during the seismic shaking.
Given the unavailability of ground motion data for the earthquake that occurred in the study area, synthetic ground motion from the Amberley earthquake in New Zealand was employed.The peak ground acceleration (PGA) value of the synthetic ground motion corresponds to the PGA value associated with the historical earthquakes in the study location.
It is crucial to select the appropriate analysis procedure in nonlinear analysis to ensure that the results accurately reflect the actual soil conditions.This selection depends on the accuracy of the input soil profile, soil parameters, and ground motion data, highlighting the importance of precision in these aspects.

Fig. 2 .
Fig. 2. History of the largest earthquake strength (Mw) in Sulawesi in the last 30 years around the study area.

Table 2 .
Input parameters in non-linear model analysis (GQ/H +u) using DeepSoil V7 software in bore hole BM 02.

Table 3 .
Input parameters in non-linear model analysis (GQ/H +u) using DeepSoil V7 software in bore hole BM 04.

Table 4 .
Input parameters in non-linear model analysis (GQ/H +u) using DeepSoil V7 Software in bore hole BM 05.

Table 5 .
Input parameters in non-linear model analysis (GQ/H +u) using DeepSoil V7 software in bore hole BM 06.