Analysis of Soil Deformation Due to Oil and Gas Exploration and Hydrocarbons Micro Seepage in Ngasem District, Bojonegoro Regency

Land subsidence is the result of the extraction of liquid in the pores of the rock from the compression of the sediment. The decrease occurred due to the exploitation of oil and gas wells in the area which took and caused reduced sediment compression. Aside from the result of excessive extraction of hydrocarbons, hydrocarbon micro seepage can also trigger an anomaly in the ground level. This anomaly is due to the abundance of hydrocarbons in the reservoir and the dynamic nature of oil and gas accumulation, which results in the absence of a perfect rock seal. This study aims to determine the average rate of subsidence in Ngasem District and to analyze the relationship between soil deformation and oil and gas exploration activities and hydrocarbon micro seepage. The method used in this research is InSAR Time Series with LiCSAR Sentinel 1 SLC and Generic data. Atmospheric Correction Online Service (GACOS) which was used from 2014 to. Meanwhile, maps of the distribution of hydrobarone micro seeps were obtained from Exxon Mobil and Exploration Wells for the East Java Basin were obtained from the Ministry of Energy and Mineral Resources. The results showed that the deformation of the land surface in the Ngasem sub-district was very diverse due to the presence of geological structures, exploration wells and potential seepage of micro-hydrocarbons. From the results of deformation processing in the Ngasem District area, it has the highest rising speed in the study area at 77mm/year while for the greatest decrease at -85 mm/year. Exxon Mobill’s exploration wells experienced a subsidence of -30mm/year, while the potential for micro-hydrocarbon seepage increased by 60mm/year. From the results of the analysis of the decline that occurred in exploration wells due to oil and gas exploitation.


Introduction
Land subsidence is the result of the extraction of liquid in the pores of the rock from the compression of the sediment.This phenomenon can be found in various parts of the world including in California, Arizona, Nevada, Italy, Mexico, Bangladesh, China and the coasts of Texas.Several studies [1];[2]; [3] argue that the extraction of hydrocarbons has a very large role for the Harris-Galveston area.The research has facts from the Goose Creek Field which showed a 1 meter drop from 1975 to 1925.
Apart from the result of excessive extraction of hydrocarbons, according to [4] that hydrocarbon micro seepage can also trigger an anomaly in the ground level.This anomaly is caused by the abundance of hydrocarbons in the reservoir and the dynamic nature of oil and gas accumulation, which results in the absence of a perfect rock seal.The phenomenon of hydrocarbon micro seepage is one of the phenomena of mass migration of hydrocarbons that occurs due to a large pressure differential from the reservoir [5]; [6].This phenomenon refers to the process of mass migration of hydrocarbons which occurs very slowly, invisible, but 1245 (2023) 012002 IOP Publishing doi:10.1088/1755-1315/1245/1/012002 2 alkane gas accumulates [7].Hydrocarbon micro seeps predominantly consist of methane (CH4), ethane (C2H6), propane (C2H8), butane (C2H10) and pentane (C2H5) gas content [8].These hydrocarbons can interact with the stratigraphic column and produce a variety of physical, chemical, mineralogical, geobotanical and microbiological variations on the Earth's surface [9].
The use of the Global navigation satellite system (GNSS) in determining deformation is considered inadequate due to the large area of a Basin which can be included in several districts causing difficulties in determining subsidence caused by oil and gas exploration and anomalies caused by hydrocarbons.Utilization of technological developments can provide a separate solution for observing soil deformation using Earth Observation (EO) from the European Commission, opening up new opportunities in monitoring land subsidence that occurs on the surface of the earth.Newer satellite constellations, such as Sentinel from the Copernicus program, provide free data services that allow observations of the entire Earth's surface at unprecedented frequencies and by using a variety of sensors, both optical and radar [10].Radar technology on the Sentinel 1 Synthetic Aperture Radar (SAR) satellite works in all weather and data acquisition during the day and night, where the sensor can penetrate clouds to allow detection of subsidence.Therefore, in this study, spatial and temporal analysis of land subsidence will be carried out by utilizing unwrapped interferogram data from SAR Sentinel 1 satellite imagery that has been processed.Where later the area of Ngasem Subdistrict can be distinguished where there is an anomaly due to seepage of hydrocarbons and a decrease that occurs due to oil and gas exploration.It is hoped that this research will better understand the phenomena that occur, and be able to analyze the differences in land subsidence that occur in the study area.

Study Area
The research area is included in the Cepu Block area which is an oil field on the border of Central Java and East Java shown on Figure 1.The Cepu Block is an oil and gas contract area covering the Bojonegoro Regency -East Java, Blora Regency -Central Java, and Tuban Regency -East Java.Prior to the recent discovery of sizable oil reserves in the Cepu region and its surroundings, namely in the Regencies of Bojonegoro and Tuban.One of them is the Banyu Urip project which is under the auspices of Exxon Mobil Cepu Limited in Bojonegoro Regency. 3

Materials and Methods
This study uses the SBAS Time Series InSAR method to be able to find out the deformation that occurs in the Ngasem District area.The data used is LiCSAR Sentinel 1 SLC data with an acquisition date of 1 January 2016 -9 August 2022.After obtaining LiCSAR image data from the COMETLiCS web portal (https://comet.nerc.ac.uk/comet-lics-portal/) then Pre-processing is carried out which will pass through several stages shown on Figure 2. The first stage is Convert Geotiff and Downsample.This stage converts to single precision floating point format without using additional information to carry out further processing.For the downsampling process it can be used for LiCSAR products (~100m) so that it will reduce the amount of data and is very useful for testing process parameters and making data adjustments that will be used in high resolution or higher.After changing the next step is Mask and Clip Interferogram to cut areas that are only used in research Furthermore, the data processing is divided into 2.The first processing uses the Small Baseline Subset (SBAS) method which will use the LiCSBAS software.Furthermore, the data from the SBAS results were processed using GIS Software (ArcMap) to overlay and clip in certain areas so that only cross-sections could be made in the study area.LiCSBAS will perform a quality check on each LiCSAR image to reduce errors that occur during processing.Next, iterations are carried out to detect bad unwrapped interferogram data.Unwrapped data will have errors that will significantly affect the time series processing and must be removed before processing.By using a comparison of 3 images and 3 interferograms, you will see a comparison of the smallest RMS and the closest pixel to be selected and not selected as error data.After that, calculation/estimation of speed is performed using pixel data from the surface to generate the required data transfer of the stack of unwrapped interferograms.This step is used to cover data shortages resulting from the acquisition of the Sentinel 1 image which is less than perfect or does not exist at a certain time so that later it is hoped that using this method can at least approach the data that was acquired directly.From the iterative data, there is still some residual noise such as tropospheric noise residue, ionosphere noise and orbital error so that it will be reduced by using a 2-dimensional Gaussian kernel with space and time variables.Using this method is expected to facilitate the application of filters for short time series data and will be able to distinguish the original nonlinear displacement.
The results of the deformation velocity generated from LiCSBAS were then evaluated with maps of exploration wells and maps of hydrocarbon micro seepage.exploration data per year from 2014 -2021 which will be compared with land subsidence data around exploration wells.From these results, statistical processing is carried out to see the link/correlation between the 2 variables whether they are related or not.This will be a benchmark for the surrounding area if large-scale extraction of hydrocarbons results in very significant reductions.Actions must be taken to prevent extreme declines for the area.

Result and Discussion
From the data processed by LiCSBAS software, there are outputs in the form of speed, average coherence, cumulative displacement, standard deviation, number of interferograms that do not experience repetition and maximum length of interferogram relationships.Processing speed data using GACOS atmospheric correction is very influential on the results obtained.Even though the interferogram error has been controlled using a loop closure quality check, GACOS is still needed because it can correct signal turbulence caused by tropospheric total delays.The outcome data is very useful for interpreting in that area the results obtained are very good or vice versa [11].
LiCSAR data using Sentinel 1 satellite data has a temporal resolution of 12 days and a spatial resolution of 5 meters.From this spatial resolution, Sentinel 1 data is used to compare micro seepage for detailed analysis which is very difficult due to the difference in resolution of the 2 results.In the dissertation belonging to [4] active remote sensing can only analyze large land surface rise due to the large reservoir area and not too small to be interpreted by the Sentinel 1 satellite.With a temporal resolution of 12 days and exploration well hydrocarbons are taken every day Land subsidence will occur every day and can be monitored continuously by satellite.
If we compare it with plotting on the standard deviation data shown on Figure 3 before and after GACOS, the output from LiCSBAS shows some changes, especially in the unwrapped interferogram.In the scatter plot comparison of Figure 3, there are significant changes in some of the data, which initially had a standard deviation of 20, dropping to 15 or 10.This shows that the higher the standard deviation, the farther the range of values is from the average value.This makes the data will at one point tend to experience large deviations.Making the standard value closer to 0 will make the value more accurate due to the smaller range of values.These results are in line with the conclusion [11] that GACOS is highly recommended to improve accuracy in time series.However, in several studies, the results of the GACOS correction actually resulted in a negative effect on the image, so that the relationship between corrected and uncorrected GACOS is very important for deformation speed [12].
The results of the coherence value show how far each pixel is similar between the master and slave images with a value between 0 to 1. Loss or blank data at some pixels (zero coherence) is caused by decorellation so that the coherence value decreases.Decorellation has various types such as spatial decorellation.This correlation is due to the better height of the target due to the increasing length of the baseline.In addition, there is Doppler centroid decorellation, this decorelation occurs due to the speed of the vehicle changing, causing variations in speed when it returns to sensing the place [13].According to [14], decorelation can be caused by the volume factor due to the uneven surface which causes the backscatter to spread large, this usually occurs in vegetated surface areas.In addition to volume decorellation, satellite systems can produce decorelations due to the existing system.However, this decorellation will only have an impact on coherent interferometrics when the return signal is very weak.Figure 4 shows some areas that do not have pixel information due to low coherence values due to decorrelation.One example of low coherence (zero) is shown in several areas of the Ngasem District.Ngasem Subdistrict, where the majority of land use is built-up land and rice fields.From rice fields and production forests, vegetation growth or leaf movement can cause different phases of the two data collections.However, differences in conditions during planting, growing and harvesting also show phase variations throughout the time series data and can reduce coherence values, this condition occurs around Ngasem District which is located in a paddy field planted with rice.
From the data processed by LiCSBAS software, there are outputs in the form of speed, average coherence, standard deviation, number of interferograms that do not experience repetition and maximum length of interferogram relationships.The outcome data is very useful for interpreting in that area the results obtained are very good or vice versa.The results of the coherence value show how far each pixel is similar between the master and slave images with a value between 0 to 1. Loss or blank data at some pixels (zero coherence) is caused by temporal decorrelation so that the coherence value decreases.Figure 4 shows some areas that do not have pixel information due to low coherence values due to temporal decorrelation.One example of low coherence (zero) is shown in several areas of the Ngasem District.Referring to Figure 4 with the subtitle maxTLen which shows the length of the interferogram connections in the research area based on data from 2016 to 2022 processed by LiCSBAS using a predetermined default threshold, the results are 2.2 years.By following the default threshold for coherence of 0.05 and unwrapped data of 0.3, there are several interferogram pairs that do not fit into the predetermined criteria so that from the image the interferogram pairs shown in Figure 5 have 2 different colors indicating acceptance and rejection of the threshold. .The red color in the image shows that the interferogram pair on that date is not used so that it will reduce the time series and make the start of processing from 2016 -2022.Figure 5 Meanwhile the blue color can be continued for processing because it meets predetermined threshold standards.However, the black straight line upwards is a gap in the interferogram network due to noise or no data on that date.The existence of a gap in the interferogram pair date 20151114_20151208 20151208_20160406 20160711_20160921 20160921_20161108 20161108_20161202 20161202_20161226 20161226_20170119 20170119_20170224 20170425_20170507.From the deformation results it was found that the highest deformation speed in the study area was at 77mm/year which is depicted in red.Meanwhile, the biggest decline is at a speed of -85 mm/year which is depicted in green.In the Ngasem sub-district itself, there is the Kahyangan Api and there is also an exploration well owned by Exxon Mobil Bojonegoro.From the results of processing using the SBAS method on LiCSAR data, the results show that the subsidence in the area around the exploration wells varies greatly.In the cross-section of the Southeast -Northwest direction that passes through the existence of exploration wells, the decrease in height varies and there tends to be an increase from -10 mm/year to -40 mm/year.The exploration well area itself has decreased by -30 mm/year.An incision that is in the direction of Southwest -Northeast which passes through several geological structures in the form of anticlines and synclines and passes through faults and ends in the Kahyangan Api.From the results of the incision, it shows an increase of 45 mm/year in the anticline area while the decrease in the heavenly area is -40 mm/year.It is known that the structure of the anticline and syncline greatly influences the deformation and can be seen from the results of the incisions of the two anticlines that they both have an increase of 45 mm/year while the syncline itself has an increase of only 20 mm/year.Meanwhile for faults, the difference between the foot wall and the hanging wall is very visible in the incision.For the hanging wall of the fault there is absolutely no change, while for the foot wall it has decreased by around -10 to -20 mm/year so that it can be concluded that the fault is a normal fault.In the Kahyangan Api area, it experienced a very significant decrease with a decrease of -40mm/year, which was the largest decrease in the incision.This is due to the continuous reduction of the gas contained in the formation thereby reducing the existing pressure and compressing the soil.
From the results of Exxon Mobil data regarding micro-hydrocarbon seepage, the bottom yield in the southern area of Ngasem District has a very high seepage potential.To test the potential results with the phenomenon of deformation, cross sections are made in the direction of Northwest -Southeast.From the results of the cross section that there is an increase of 60mm/year.In line with research [4] that in areas experiencing micro-hydrocarbon seepage will experience anomaly in ground level.Where the area in the middle will experience the highest increase and will decrease exponentially to areas that do not experience seepage.From the results it shows that the highest increase is at 4.7 kilometers from Kahyangan Api and shows the highest increase in that area of 60mm/year.From the processed data, there are several places that have decreased or vice versa experienced an increase in the soil surface.According to the results obtained, the resulting soil deformation in Ngasem District comes from gas and hydrocarbons that are in formations that are pressing upwards.From the speed time series results, the highest potential for hydrocarbon seepage shows uplift results.This is in line with research from [15] that the migration of hydrocarbons is dynamic and there is no perfect seal cap, causing oil and gas traps in formations to be expressed directly by symptoms of anomalies on the earth's surface.Due to the accumulation of hydrocarbons in the reservoir.Causes a large pressure difference from the reservoir source and causes an anomaly on the ground surface due to the large upward pressure force from the reservoir [16].
The highest potential of the seepage zone is in the south of Ngasem District where the area has 2 geological structures that can become hydrocarbon traps.The first structure is anticline and syncline geological structure.The anticline structure in this area becomes a hydrocarbon trap which stores gas and hydrocarbons in it causing great pressure and suppresses the seal rock so that the uplift phenomenon is faster than the syncline structure which has no pressure.This is IOP Publishing doi:10.1088/1755-1315/1245/1/0120029 very much in line with what was stated by [4] that oil and gas will tend to press against the formation so that the hydrocarbons seep onto the statigraphic surface.
In line with the existence of the Kahyangan Api from the Ngasem District which continues to emit gas to the surface.As a result of the gas pressure being released continuously so that the formation experiences a lack of pressure and tends to solidify.As a result of this, the area around Kahyangan Api continues to experience very fast land subsidence -24.6 mm/year.The decline was also accelerated by Exxon Mobill's exploratory wells which produced 2 million barrels per year.With very large production, even though it has been replaced with the same specific gravity, over time it will lose the same pressure as hydrocarbons and cause a decrease that tends to be significant [17].
The amount of oil and gas production from exploration wells greatly affects land subsidence in the area.According to [18] that the extraction of oil that is in the formation will cause a decrease even though it has been replaced with drilling mud so that the pressure in the formation will tend to be stable.However, the pore pressure that is created due to oil and natural gas will not be the same as the drilling mud so that it will experience a loss of density and experience compression so that it will experience a significant decrease.From the results of oil and gas production, to see how much the level of relationship between the 2 variables is, the Pearson correlation method is used.From the correlation results, the value is -0.774.This value can be interpreted as a relationship between 2 variables where an increase in one variable causes a decrease in the value of the other variable.Vice versa, the smaller the value of a variable, the greater the value of other variables so that the relationship between the two variables between vertical deformation and oil and gas production is reversed [19].This can be seen from Table 1.From 2016 to 2021, Exxon Mobill's production will increase every year, accompanied by faster land subsidence.In 2019, when well production rose by 300,000 thousand barrels, it caused a significant decrease of -81.09.The decrease is 30 mm different from the previous year.According to [20] the enhanced oil recovery in this formation did not take hydrocarbons from the formation causing a sudden extreme decline in 2019.

Conclusion and Recommendation
In the results of this study, the deformation of the land surface in Ngasem District is very diverse due to the geological structure, exploration wells and potential seepage of microhydrocarbons.From the results of deformation processing in the Ngasem District area, the rate of increase is the highest in the research area at 77mm/year while the largest settlement is at -85 mm/year.In Exxon Mobil's exploration wells, there is a subsidence of -30mm/year while for potential seepage of micro-hydrocarbons has increased by 60mm/year.

Figure 1 .
Figure 1.Research Location Map

Figure 3
Figure 3 Scatter Plot of Standard Deviation Data Before and After GACOS

Figure 6 .
Figure 6.Land Deformation Cross Section Map

Table 1 .
Oil and Gas Production and Land Subsidence in Exxon Mobill Exploration WellsWhen compared to the results of oil and gas production at Exxon Mobil's wells taken from 2016 -2021 sourced from Oil and Gas in the annual figures released by the Ministry of Energy and Mineral Resources as shown in Table4.1.Oil and gas production always increases from year to year.Oil and gas production tends to increase resulting in a decrease in the area around exploration wells as shown in Table1.