IEG 환경지질연구정보센터

전체 글

(1)Geosciences Journal Vol. 16, No. 3, p. 313 − 326, September 2012 DOI 10.1007/s12303-012-0026-3 ⓒ The Association of Korean Geoscience Societies and Springer 2012. Estimation of the CO2 storage capacity of the structural traps in the southern Jeju Basin, offshore southern Korea, northern East China Sea. }. Gwang H. Lee* Bumsuk Lee Department of Energy Resources Engineering, Pukyong National University, Busan 608-737, Republic of Korea Bo-Yeon Yi Keumsuk Lee Korea National Oil Corporation, Anyang 431-711, Republic of Korea Myong-ho Park Han-Joon Kim Korea Ocean Research and Development Institute, Ansan 426-744, Republic of Korea Hai-Soo Yoo. } }. ABSTRACT: We analyzed 2-D seismic and well-log data from the southern Jeju Basin, offshore southern Korea, northern East China Sea to estimate the CO2 storage capacity of the structural traps in the area. Sand intervals with >10-m gross thickness were identified from the gamma-ray logs and their porosity was estimated from the neutron logs corrected for the shale effect. A total of 14 structural closures was delineated from the depth-converted maps of the sand intervals. Seismic inversion and multi-attribute transform were performed to predict the reservoir quality (i.e., porosity) of the closures away from the well control. The total storage capacity of the closures was estimated from the deterministic, volumetric method, based on the published storage efficiency parameters. The estimated CO2 storage capacity for the 14 closures is about 302 Mt, comparable to the CO2 emission (ca. 530 Mt) of Korea in 2009. Key words: CO2 storage capacity, Jeju Basin, structural traps. 1. INTRODUCTION Geological sequestration of CO2 involves pumping the CO2 into the subsurface formations to store it in the pore space of the rock in the same way that hydrocarbons are trapped in oil and gas reservoirs. At depths below about 800−1000 m, CO2 becomes supercritical and has a liquidlike density which increases the efficiency of CO2 storage in deep formations. Depleted oil and gas reservoirs, coal formations, and saline formations can be used for storage of CO2 (IPCC, 2005). CO2 can also be injected into active oil fields to improve production rates and increase productive lifetime of the field. This CO2-driven enhanced oil recovery (EOR) has become one of the earliest methods of CO2 storage because of the economic benefits to offset the production cost. In the long term, the large capacity and advantageous locations of deep saline formations will increasingly make them the preferred storage site as the amount of CO2 to be sequestered increases (White et al., 2003). The Donghae-1 field in the southern Ulleung Basin mar*Corresponding author: [email protected]. gin, East Sea (Japan Sea) is the only producing field in Korea. The Donghae-1 field is producing mostly gas and thus CO2-driven EOR cannot be used. It is also planned to be used for the national strategic petroleum storage after depletion. Therefore, the most likely option for the geological sequestration of CO2 in Korea is storage in deep saline formations. The onshore geological sequestration of CO2 may be ruled out in Korea because the onshore basins are shallow and the distribution of good reservoir-quality rocks is very limited. On the other hand, offshore basins of Korea (the Ulleung, Gunsan, and Jeju basins) contain thick and young clastic sediments (Lee et al., 2001; Lee et al., 2006; Shinn et al., 2010). The Gunsan and Jeju basins contain over 5000-m thick, Late Cretaceous and Tertiary, dominantly nonmarine sediments. The Ulleung Basin is the youngest and contains over 11 kilometers of Late Oligocene and younger, dominantly marine sediments. In this study, we analyzed 2-D seismic and well-log data from the southern part of the Jeju Basin off southern Korea, northern East China Sea (Fig. 1) to map the structural traps and estimate their CO2 storage capacity. Inversion and multi-attribute transform were applied to the data to predict the reservoir quality of the traps away from the well control. Kingdom Suite® (version 8.6) and Hampson-Russell® (version 4.4.1) were used for seismic and well-log data analysis and inversion and multi-attribute transform, respectively. 2. GEOLOGGICAL SETTING The Jeju Basin is located in the northern East China Sea shelf basin (Fig. 1). It is separated from other subbasins (the Socotra and Domi basins) to the south and north by the Hupijiao Rise and faulted basement blocks and bounded to east by the Taiwan-Sinzi belt (Lee et al., 2006; Gungor et al., 2012). Initial rifting in the Late Cretaceous created grabens and half grabens in the northern part of the proto-East China Sea shelf (Lee et al., 2006). The rifting was termi-.

(2) 314. Gwang H. Lee, Bumsuk Lee, Bo-Yeon Yi, Keumsuk Lee, Myong-ho Park, Han-Joon Kim, and Hai-Soo Yoo. Fig. 1. Structural and tectonic elements of the East China Sea (Zhou et al., 1989; Yang, 1992; Lee et al., 2006). Inset is bathymetry in meters. Study area is shown in Figure 2.. nated by regional uplift and folding in the Late EoceneEarly Oligocene. This regional deformation was extensive near the Hupijiao Rise and in the Jeju Basin. Rifting resumed in the Early Oligocene but was interrupted in the Early Miocene by uplift in the Domi Basin and areas adjacent to. the Hupijiao Rise. This uplift, although relatively minor, marks the transition to the postrift phase in the northern East China Sea shelf basin. The early part of the postrift phase (Early Miocene-Late Miocene) is characterized by eastward and southeastward tilting and regional subsidence..

(3) CO2 storage capacity of the southern Jeju Basin. 315. In the Late Miocene, compressional tectonism created an extensive thrust-fold belt along the eastern and southeastern parts of the Jeju Basin and the eastern part of the Domi Basin. Subsequent erosion completely leveled the thrust-fold belt, resulting in the prominent Late-Miocene unconformity. Most basement faults terminate at the Late-Miocene unconformity which marks the beginning of a new phase of regional subsidence that resulted in a broad continental shelf. The Jeju Basin is filled with nonmarine sediments with relatively thin shelf deposits on the top (Kwon and Boggs, 2002; Lee et al., 2006). The basin fill is thick (>4500 m) in the south and thins (<1500 m) toward the southwestern and central parts of the basin (Kwon and Boggs, 2002). The basin fill can be divided into five depositional sequences (Kwon and Boggs, 2002). The basal Oligocene synrift sequence consists of fluvio-lacustrine sandstone, mudstone, and conglomerate with thin coal beds. The Early to Middle-Miocene postrift sequence is composed of sandstone and mudstone with interbeds of conglomerate, coal, and fresh-water limestone. The Late-Miocene sequence consists of fluvial sandstone, siltstone, and mudstone. Pliocene and PleistoceneHolocene sequences consist of shallow marine or shelf sandstone and mudstone. 3. DATA Data used in this study consist of nine multi-channel seismic profiles with total length of about 430 km and sonic, density, gamma-ray, and neutron-porosity logs from four exploratory wells (J1-1, J1-2, J2-1 and J2-2) (Fig. 2). The seismic data for the J1 and J2 areas were acquired in the early 1980s and late 1970s, respectively, by various service companies for the Korea Petroleum Development Corporation (PEDCO) which was renamed to the Korea National Oil Corporation (KNOC) in 1999. Detailed information about the data acquisition and processing is unavailable. The sonic and density logs were used to generate synthetic seismograms for seismic-to-well tie and inversion. The gamma-ray log was used to delineate reservoir sand intervals. The age control is based on the published reports (KIGAM, 1997; Kwon and Boggs, 2002; Lee et al., 2006). 4. DATA ANALYSIS 4.1. Well-Log Data Analysis The density and sonic logs were edited for anomalous values and smoothed using a seven-point moving average to remove heterogeneities. The density-log data with density correction (DRHO) of over 0.1 g/cm3 were excluded in data analysis. The neutron-porosity logs were also edited for erratic values. When clays are part of the formation matrix, the neutron porosity is greater than the actual formation. Fig. 2. Seismic reflection data coverage, well locations, and structural closures with sand. Heavy gray lines and respective figure numbers indicate seismic profiles shown in other figures. Contours are bathymetry in meters.. porosity because the hydrogen within the clay structure and in the water bound to clay is detected in addition to the hydrogen in the pore space (Asquith and Krygowski, 2004). The neutron porosity should be corrected not only for this shale effect but also for the lithology if the formation is sandstone or dolomite since the neutron porosity is recorded as the limestone porosity index which assumes no hydrogen in the matrix. The correction for the lithology can be done using the neutron departure curves or appropriate algorithms (Helander, 1983); the neutron porosity decreases about 4−5% after the correction for the lithology (Schlumberger, 1989). The neutron porosity was corrected for the shale effect but the correction for the lithology was not made because no detailed information about how the neutron log was recorded and processed is unavailable. The first step in the correction for the shale effect is to determine the shale volume (Vsh) which can be computed from various logs. The most widely used is the gamma-ray empirical relationships between the response of the mea-.

(4) 316. Gwang H. Lee, Bumsuk Lee, Bo-Yeon Yi, Keumsuk Lee, Myong-ho Park, Han-Joon Kim, and Hai-Soo Yoo. Fig. 3. Histogram for gamma-ray values of four wells. Gamma-ray values range from 20 to 135 API with peak at around 50–60 API. Cutoffs for clean sand and 100% shale were assumed to be 40 and 100 API, respectively.. surement and the shale or clay volume (Asquith and Krygowski, 2004). We estimated Vsh from the gamma-ray log using the following equation (Asquith and Krygowski, 2004): GRlog – GRsand , IGR = -----------------------------------GRshale – GRsand. (1). where: IGR = gamma-ray index, GRlog = gamma-ray reading of formation, GRsand = gamma-ray for shale-free (clean) sand, and GRshale = gamma-ray for 100% shale. GRsand and GRshale can be determined from the clean sand and shale base lines from the gamma-ray log, respectively, or from the gamma-ray log corresponding to the cored rock samples. Because core data are unavailable and it is difficult to determine the clean sand and shale base lines directly from gamma-ray log curves, we plotted the distribution of the gamma-ray values for the four wells (Fig. 3) to determine the sand and shale cutoffs. The gamma-ray values range from 20 to 135 API with a peak at around 50−60 API. The cutoffs for clean sand and 100% shale were assumed to be 40 and 100 API, respectively. This assumption may not be very conservative but is reasonable because the gammaray distribution exhibits an overall normal distribution and the numbers of samples for <40 API and >100 API are about 7.5% and 5% of the total number of samples, respectively. We further assumed that Vsh is equal to IGR, which is valid for a first-order estimation of Vsh (Asquith and Krygowski, 2004). After Vsh was determined, the neutron porosity was corrected for the shale effect using the equation from Dewan (1983): φ Ne = φN – Vsh ⋅ φ Nsh , where: φNe = shale-corrected neutron porosity, φN = neutron porosity, and φNsh = neutron porosity of shale.. (2). Fig. 4. Neutron porosity (φN, black closed circles) and shale-corrected neutron porosity (φNe, gray open circles) vs. depth for J2-1 well. Difference between φN and φNe at shallow depths is over 10% and decreases with depth.. To be conservative, φNsh was assumed to be the average (approximately 20%) of the neutron porosity values of the formations with >120 API. Figure 4 shows φN (black closed circles) and φNe (gray open circles) vs. depth for the J2-1 well. The gap between φN and φNe decreases with depth..

(5) CO2 storage capacity of the southern Jeju Basin. 317. Fig. 5. (a) Seismic profile crossing J1-1 and J1-2 wells. (b) Acoustic impedance section of (a). Stratigraphic horizons were correlated with those from published reports (KIGAM, 1997; Kwon and Boggs, 2002; Lee et al., 2006). See Figure 2 for location..

(6) 318. Gwang H. Lee, Bumsuk Lee, Bo-Yeon Yi, Keumsuk Lee, Myong-ho Park, Han-Joon Kim, and Hai-Soo Yoo. 4.2. Seismic Inversion and Multi-Attribute Transform The acoustic impedance computed from seismic inversion can be correlated directly with quantitative rock and pore-fluid properties. Multi-attribute transform, taking into account the lateral and vertical variations of rock properties, further provides the link between the seismic attributes, including the acoustic impedance, and well-log data and help predict any measured or calculated logs away from well control (Schultz et al., 1994; Russell et al., 1997). Thus, seismic inversion and multi-attribute transform are the very valuable tools to characterize and quantify rock properties beyond the well control. We performed model-based inversion to compute the acoustic impedance (Fig. 5). Inputs to the inversion included. stacked seismic data, sonic and density logs, and interpreted horizons. In model-based inversion, an initial impedance model, constructed from the sonic and density logs at the wells and guided by the interpreted horizons, is perturbed iteratively until a good fit is obtained between real seismic traces and synthetic traces generated from the convolution of the source wavelet and the reflectivity computed from the impedance model. A wavelet was extracted first statistically from the seismic traces at or near the well location, assuming a constant phase. The synthetic trace constructed using this wavelet was correlated with the real seismic trace at the well to obtain the time-depth relationship. Then, the full wavelet, which produces the real seismic trace when convolved with the reflectivity at the well, was extracted. This final wavelet was assumed to be the source wavelet to. Fig. 6. (a) Time-depth relationship for J1 area. (b) Time-depth relationship for J2 area. Gray curves are from seismic-to-well tie. Dashed curves are time-depth relationships derived from seismic-to-well tie..

(7) CO2 storage capacity of the southern Jeju Basin. run the model-based inversion. The synthetic-to-seismic tie was further fine-tuned to get the more accurate time-depth relationships at the wells which were used for time-depth conversion. Two different time-depth relationships were obtained for the J1 and J2 areas, respectively (Fig. 6). The multi-attribute analysis finds a linear or non-linear transform between a subset of the seismic attributes, including the acoustic impedance, and the target log values (Hampson et al., 2001). The transform in the linear mode consists of an explicit combination of attributes derived by a multi-linear regression. In the non-linear mode, a neural network is trained, using selected attributes as inputs. With enough well-log data, neural network can predict better and provides higher resolution. We predicted φNe away from the wells using the multi-linear regression because the result from neural network is very noisy and does not reflect the stratigraphy observed in the seismic data. Inputs to the multilinear regression included φNe at the wells, the stacked seismic data, the acoustic impedance, and various seismic attributes, computed internally in the Hampson-Russell program. Figure 7 shows the predicted φNe for a segment of the seismic line crossing the J1-2 well and the φNe-log curve at the well. The multi-attribute transform did not predict the absolute φNe at the well, but the predicted φNe generally exhibits the similar pattern of variation to that of the real φNe: the high and low predicted φNe values correspond to the high and low real φNe values, respectively. The predicted φNe was used to determine the porosity of the structural traps or closures away from the wells.. 319. 5. DELINEATION OF THE CLOSURES A total of 13 sand intervals (S1−S13) with gamma-ray values of <40 API and gross thickness of >10 m were identified (Fig. 8). The gross thicknesses of the sand intervals were delineated at the inflection points on the gamma-ray log. The reflectors or horizons correlated to these sand intervals were interpreted and mapped (Fig. 9). The timestructure maps of the sand intervals were converted into depth (Fig. 10). A total of 30 potential traps or closures were identified from the depth maps: five in the J1 area and 25 in the J2 area. These closures occur below the Late Miocene regional unconformity and are associated with the thrusts or anticlines formed during the Late Miocene tectonic event. The closures that cannot be determined with confidence due to the wide seismic coverage were not included. The wells penetrated 20 of the closures or their immediate margins and confirmed the sand intervals in 12 of them. The average φNe values of the sand intervals at the wells for the 12 closures were assumed to represent the porosity for the entire closures. The porosity for the ten closures that were not penetrated by the wells was determined qualitatively from the results of the multi-attribute transform because the predicted φNe appears to represent only the relative porosity. Among these ten closures, two closures show relatively high predicted porosity (Fig. 11). These two closures and the 12 closures with sands at the well locations (Fig. 2) were analyzed for the CO2 storage capacity. The porosity of the closures where the porosity log is unavail-. Fig. 7. Predicted shale-corrected neutron-porosity (φNe) section from multi-attribute transform with shale-corrected neutron-porosity log (φNe) of J1-2 well overlaid. Predicted porosity generally exhibits similar pattern of variations to real porosity at well: high and low predicted porosity values correspond to high and low real porosities, respectively. See Figure 2 for location..

(8) 320. Gwang H. Lee, Bumsuk Lee, Bo-Yeon Yi, Keumsuk Lee, Myong-ho Park, Han-Joon Kim, and Hai-Soo Yoo. Fig. 8. Gamma-ray logs and 13 sand intervals with gamma-ray values of <40 API and gross thickness of >10 m delineated at inflection points. S5 is identified in both J2-1 and J2-2 wells.. able was estimated from the sand φNe-depth trend (Fig. 12) in the area. 6. ESTIMATION OF THE CO2 STORAGE CAPACITY The CO2 storage capacity of a rock formation depends not only on the physical rock properties and those affecting them such as temperature and pressure, among others (Kopp et al., 2009) but also on the allowable pressure change and the size of the affected area (Van der Meer and Egberts, 2008; Birkholzer et al., 2009). Because it is very difficult to quan-. tify all these factors, the concept of storage efficiency was introduced. The storage efficiency is defined as the percentage of the rock formation that is available to CO2 storage (EERC, 2009). The two most often applied systems of the storage efficiency for the deep saline formations in open systems are those of the Carbon Sequestration Leadership Forum (CSLF) (CSLF, 2007) and the National Energy Technology Laboratory (NETL) of the U.S. Department of Energy (DOE) (NETL, 2010). In open systems, injection into the formation does not cause a noticeable increase in the formation pressure, and pressure buildup is limited to the near.

(9) CO2 storage capacity of the southern Jeju Basin. 321. Fig. 9. Seismic profile showing sand intervals identified in J2-2 well and corresponding seismic horizons. See Figure 2 for location.. wellbore environment (EERC, 2009). The CSFL method calculates the volume of stored CO2 based on a capacity coefficient. However, there are no values in the literature for the capacity coefficient (CSLF, 2007). The NETL method calculates the volume of stored CO2 based on the storage efficiency (Esaline), consisting of the geologic and displacement parameters. The storage efficiency of the NETL (2010) is given by: Esaline = EAn/At ⋅ Ehn/hg ⋅ Eφe/φtot ⋅ EA ⋅ EL ⋅ Eg ⋅ Ed. (3). where: EAn/At = net-to-total area, fraction of the total basin or region suitable for CO2 storage, Ehn/hg = net-to-gross thickness, Eφe/φtot = effective-to-total porosity, EA = areal displacement efficiency, EL = vertical displacement efficiency, Eg = gravity displacement efficiency, and Ed = microscopic displacement efficiency. The multiplications of the first three terms (EAn/At⋅Ehn/hg ⋅Eφe/φtot) and the next four terms (EA⋅EL⋅EgEd) are the geologic and displacement parameters, respectively. The geologic parameter represents the efficiency for the effective pore volume. The first three terms of the displacement parameter can be multiplicatively combined into a single volumetric. displacement efficiency (Ev) term (EERC, 2009). Then, the storage mass of CO2 (MCO ) can be computed from the following equation: 2. MCO2 = ρCO2 ⋅ RV ⋅ Esaline = ρCO2 ⋅ RV ⋅ Gp ⋅ Ev ⋅ Ed ,. (4). where ρCO , RV, and Gp are the density of CO2 at in-situ conditions, rock volume, and geologic parameter, respectively. RV⋅Gp represents the available, effective pore volume (Vφe) of the total geologic unit and can be given by Eφe/φtot⋅Vφtot where Vφtot is the total pore volume. EERC (2009) proposed the values for Ev and Ed for various geologic settings, determined from numerical simulations run on models developed from the Average Global Database (AGD) that contains geologic properties of over 20000 hydrocarbon reservoirs. The Ev and Ed values for nonmarine (i.e., fluvial) environment proposed by EERC (2009) range from 0.19 to 0.53 and from 0.34 to 0.73, respectively. For a deterministic estimation, the medians for the two efficiency values can be used: 0.36 for Ev and 0.54 for Ed. Then, Equation 4 is simplified into: 2. MCO2 = 0.19 ⋅ ρCO2 ⋅ ( Eφ e/φ tot ⋅ Vφ tot ) .. (5). Our estimation of the CO2 storage capacity for the structural traps is based on Equation 5, assuming an open sys-.

(10) 322. Gwang H. Lee, Bumsuk Lee, Bo-Yeon Yi, Keumsuk Lee, Myong-ho Park, Han-Joon Kim, and Hai-Soo Yoo. Fig. 10. (a) Depth structure map of S4. (b) Depth structure map of S11. Closures with reservoir-quality sand are shown..

(11) CO2 storage capacity of the southern Jeju Basin. 323. Fig. 11. (a) Depth (black) cross section of NE-SW seismic profile showing closure for S11 and predicted porosity (gray). Predicted porosity is high in closure. (b) Depth (black) cross section of NW-SE seismic profile showing closure for S11 and predicted porosity (gray); seismic data are missing over closure but overall trend of predicted porosity suggests relatively high porosity. See Figure 2 for location.. tem. First, the total pore volume (Vφtot) of each sand interval in the closure was computed using the following equation: Vφtot = SRV ⋅ N/G ⋅ φ Ne ,. (6). where SRV is the sand rock volume and N/G is the net to gross ratio. The net thickness was determined by the gamma-ray cutoff of <40 API (Fig. 8). The N/G of the undrilled closures was assumed to be the same as those of the drilled closures for the same sand interval. The N/G ranges from 0.23 to 0.98. The SRVs for the 14 closures range from about 90×106 m3 to over 4000×106 m3, respectively. The total pore volume is about 2640×106 m3. The values of the effective-to-total porosity efficiency (Eφe/φtot) for nonmarine environment, proposed by EERC (2009), range from 0.64 to 0.77. These values are for the clastic rocks in fluvial environment that can encompass various types of lithologies. In this study, we took the maximum value (0.77) for Eφe/φtot because only the reservoirquality sand intervals were included for the total pore vol-. ume. The SRV was computed directly from the Kingdom program. Then, Equation 5 can be reformulated into: MCO2 = 0.19 ⋅ ρCO2 ⋅ ( 0.77 ⋅ Vφ tot ). (7). The in-situ temperature and pressure to estimate the density of CO2 at reservoir depths were estimated from various published data. The temperature in each closure was estimated, assuming the seafloor temperature of 12 °C (Yanagi et al., 1996) and the geothermal gradient of 2.7 °C/100 m (Xu et al., 1997). The pressure in each closure was estimated, assuming hydrostatic condition from the sea level and seawater density of 1025 kg/m3 (Weast, 1967−1968). The density of CO2 in each closure at in-situ conditions was taken from the CO2 density, temperature, pressure diagram of Bachu (2003), constructed from the equation of state for CO2 under high temperature and pressure proposed by Span and Wagner (1996). Table 1 summarizes the 14 closures and their associated sands and volumetric parameters and density of CO2 at in-situ conditions. Closures 11 and 13 are.

(12) 324. Gwang H. Lee, Bumsuk Lee, Bo-Yeon Yi, Keumsuk Lee, Myong-ho Park, Han-Joon Kim, and Hai-Soo Yoo. those that were delineated from the predicted porosity. The estimated total storage capacity for the 14 closures is about 302×109 kg (= 302 Mt), which is comparable to the CO2 emission (ca. 530 Mt) of Korea in 2009 (Energy Information Administration, 2011). 7. DISCUSSION. Fig. 12. Porosity vs. depth for sand (gamma-ray value of <40 API) in study area. Porosity of closures where porosity logs are unavailable was estimated from this trend.. The offshore sedimentary basins of Korea are the prospective sites for geological CO2 sequestration. However, despite exploration drilling and extensive seismic surveys, these basins have rarely been evaluated for CO2 storage except for the most recent study of the southwestern margin of the Ulleung Basin reported by Kim et al. (2012). Our study, although not regional, is the first attempt to evaluate the CO2 storage capacity of the Jeju Basin. We took a simple deterministic approach based on the published storage efficiency parameters for nonmarine sediments. The deterministic method, commonly employed by petroleum industry to estimate oil and gas reserves or volumetrics during the exploration stage, is known to give a reliable, first-order estimate (Jahn et al., 2008). The storage efficiency was included because the deterministic method alone tends to overestimate the CO2 storage capacity (EERC, 2009). We assumed an open system although the shelf sediments above the Late Miocene unconformity are likely to provide a regional seal and the faults that terminate at the Late Miocene unconformity may act as seals. The structural traps in the study area may form semi-closed systems but it is very difficult to develop storage efficiencies for closed or semi-closed systems because of their complex nature (EERC, 2009).. Table 1. Structural closures, associated sand intervals, their rock properties for CO2 storage estimation, pore volumes, in-situ conditions, and storage capacity Closure Sand 2 3 4 5 6 7 8 9 10 11 12 13 14 a. S2 S3 S4 S5 S5 S6 S7 S8 S9 S10 S11 S11 S12 S12. Area (×106 m2) 16.64 18.69 50.25 168.23 5.11 102.63 11.44 6.05 6.55 8.66 71.75 7.17 87.72 16.94. SRV: sand rock volume.. SRVa (×106 m3) 267.41 429.31 775.89 1822.83 91.03 1726.25 193.95 108.02 143.53 123.77 4324.19 129.57 3394.47 694.98. Gross (m) 27 30 18 11 103 17 46 28 44 26 79 79 44 44. Net (m) 11 6 9 9 68 14 35 14 37 6 50 50 35 35. N/G. Porosity. 0.41 0.20 0.50 0.82 0.66 0.82 0.98 0.50 0.84 0.23 0.63 0.63 0.80 0.80. 0.36 0.21 0.17 0.18 0.25 0.06 0.36 0.35 0.32 0.33 0.35 0.29 0.33 0.26 Total:. Pore Vol. (×106 m3) 39.22 18.03 65.95 268.45 15.02 85.30 68.43 18.90 38.62 9.43 957.89 23.78 891.05 143.73 2643.81. Pressure (MPa) 11.05 20.39 22.00 26.37 21.35 36.66 11.55 11.80 12.05 13.56 12.05 15.17 13.56 17.58. Temp. CO2 Den(°C) sity (kg/m3) 29.70 800 54.81 765 59.13 747 70.88 753 57.38 784 98.55 747 31.05 773 31.73 773 32.40 800 36.45 780 32.40 800 40.77 787 36.45 780 47.25 760 Total:. Capacity (×106 kg) 4590.37 2018.01 7207.44 29573.87 1723.34 9321.78 7738.43 2137.79 4520.37 1075.58 112111.34 2738.22 101681.15 15981.59 302419.29.

(13) CO2 storage capacity of the southern Jeju Basin. We also assumed that the sand identified at the well location is present in the entire closures. This assumption can be the main source of errors in our estimation of the CO2 storage capacity. The uncertainty in the seal integrity may not be a big source of errors because the storage efficiencies take into account the various uncertainties, including the seal integrity, based on a large number of data from many nonmarine sedimentary basins that are not very different from the Jeju Basin. The porosities predicted away from well control using 2-D seismic data especially between seismic profiles can also cause errors because seismic inversion and multi-attribute transform work much better for 3-D than 2-D seismic data. Nevertheless, our estimate represents a minimum storage capacity in the area because only the well-defined structural traps with thick sands were included. Other geological trapping mechanisms, such as hydrodynamic and stratigraphic trapping, do not require structures as long as the top and/or lateral seals are present (IPCC, 2005; White et al., 2003). The flow of CO2, following the natural hydrodynamic gradient, is typically very slow (110 cm/yr) and therefore CO2 cannot escape for hundreds of thousands to million years (Hitchon, 1996). Chemical trapping occurs when CO2 dissolves in subsurface fluids (i.e., solution trapping) and may then be involved in chemical reactions with the rock matrix or becomes adsorbed onto grain surfaces (i.e., mineral trapping) (Bachu et al., 2007). Chemical trapping can contribute up to 70% of the total storage (IPCC, 2005), but it operates slowly over a very long time scale and its contribution is negligible during the operational phase of CO2 injection (Bachu et al., 2007). 8. SUMMARY This study is the first attempt to evaluate the CO2 storage capacity of the Jeju Basin although it is limited to the structural traps found in the southern part of the basin. A total of 14 structural closures was identified for the potential CO2 storage sites. The total storage capacity estimated using the simple deterministic method, based on the storage efficiency for nonmarine sediments, is about 302 Mt. This is a minimum estimate because only the rock volumes of the sand with >10-m gross thickness were included and other geological trapping mechanisms, such as hydrodynamic and stratigraphic trappings, and chemical trapping were not considered. ACKNOWLEDGMENTS: This work was funded by the Construction of Carbon Storage Map and Selection of Demonstration Sites in Korean Offshore Areas (Ministry of Land, Transport and Maritime Affairs of Korea). Partial financial support to H.J. Kim and H.S. Yoo was provided by Korea Ocean Research and Development (PM56481). We thank the Korea National Oil Corporation (KNOC) for permission to publish the data. We also thank the anonymous reviewer whose suggestions greatly improved the quality of the work. Kingdom Suite®, provided by Seismic Micro-Technology, Inc. and Hampson-Russell®, provided by CGGVeritas, were used for data analysis.. 325. REFERENCES Asquith, G. and Krygowski, D., 2004, Basic Well Log Analysis (2nd Edition). American Association of Petroleum Geologists Methods in Exploration Series, No. 16, 244 p. Bachu, S., 2003, Screening and ranking sedimentary basins for sequestration of CO2 in geological media in response to climate change. Environmental Geology, 44, 277−289. Bachu, S., Bonijoly, D., Bradshaw, J., Burruss, R., Holloway, S., Christensen, N.P., and Mathiassen, O.M., 2007, CO2 storage capacity estimation: Methodology and gaps. International Journal of Greenhouse Gas Control, 1, 430−443. Birkholzer, J.T., Zhou, Q., and Tsang, C.-F., 2009, Large-scale impact of CO2 storage in deep saline aquifer: A sensitivity study on pressure response in stratified systems. International Journal of Greenhouse Gas Control, 3, 181−194. CSLF (Carbon Sequestration Leadership Forum), 2007, Estimation of CO2 Storage Capacity in Geological Media – Phase II Report, 43 p. Dewan, J., 1983, Essentials of Modern Open-Hole Log Interpretation. PennWell Publishing Company, Tulsa, Oklahoma, 361 p. EERC (Energy & Environmental Research Center), 2009, Development of Storage Coefficients for Carbon Dioxide Storage in Deep Saline Formations. 94 p. Energy Information Administration, International Energy Statistics, Total Carbon Dioxide Emissions from the Consumption of Energy, <http://www.eia.gov/cfapps/ ipdbproject/IEDIndex3. cfm?tid=90&pid=44&aid=8> Accessed November, 2011. Gungor, A., Lee, G.H., Kim, H.J., Han, H.C., Kang, M.H., Kim, J., and Sunwoo, D., 2012, Structural characteristics of the northern Okinawa Trough and adjacent areas from regional seismic reflection data: geologic and tectonic implications. Tectonophysics, 522−523, 198−207. Jahn, F., Cook, M., and Graham, M., 2008, Hydrocarbon Exploration and Production (2nd Edition). Elsevier, Developments in Petroleum Science 55, 444 p. Hampson, D.P., Schulke, J.S., and Quirein, J.A., 2001, Use of multiattribute transforms to predict log properties from seismic data. Geophysics, 66, 220−236. Helander, D.P., 1983, Fundamentals of Formation Evaluation. Tulsa, Oklahoma, OGCI Publications, 332 p. Hitchon, B., 1996, Aquifer Disposal of Carbon Dioxide: Hydrodynamic and Mineral Trapping – Proof of Concept. Geoscience Publishing Ltd., Sherwood Park, Alberta, Canada, 165 p. IPCC (Intergovernmental Panel on Climate Change), 2005, IPCC Special Report on Carbon Dioxide Capture and Storage. Cambridge University Press, 431 p. KIGAM (Korea Institute of Geoscience and Mineral Resources), 1997, Hydrocarbon potential of the northern East China Shelf Basin I: Korea Institute of Geoscience and Mineral Resources (KIGAM) Report, not paginated. Kim, Y., Lee, K., Jo, S., Kim, M., Kim, J.-S., and Park, M.h., 2012, A preliminary evaluation on CO2 storage capacity of the southwestern part of Ulleung Basin, offshore, East Sea. Economic and Environmental Geology, 45, 41−48. Kopp, A., Class, H., and Helmig, R., 2009, Investigation on CO2 storage capacity in saline aquifers Part 1. Dimensional analysis of flow processes and reservoir characteristics. International Journal of Greenhouse Gas Control, 3, 263−276. Kwon, Y.-I. and Boggs Jr., S., 2002, Provenance interpretation of Tertiary sandstones from the Cheju Basin (NE East China Sea):.

(14) 326. Gwang H. Lee, Bumsuk Lee, Bo-Yeon Yi, Keumsuk Lee, Myong-ho Park, Han-Joon Kim, and Hai-Soo Yoo. a comparison of conventional petrolographic and scanning cathodoluminescence techniques. Sedimentary Geology, 152, 29−43. Lee, G.H., Kim, H.J., Han, S.J., and Kim, D.C., 2001, Seismic stratigraphy of the deep Ulleung Basin in the East Sea (Japan Sea) back-arc basin. Marine and Petroleum Geology, 18, 615−634. Lee, G.H., Kim, B., Shin, K.S., and Sunwoo, D., 2006, Geologic evolution and aspects of the petroleum geology of the northern East China Sea shelf basin. American Association of Petroleum Geologists Bulletin, 90, 237−260. NETL (National Energy Technology Laboratory), 2010, Carbon Sequestration Atlas of the United States and Canada (Third Edition), 162 p. Russell, B., Hampson, D., Schuelke, J., and Quirein, J., 1997, Multiattribute seismic analysis. The Leading Edge, 16, 1439−1443. Schlumberger, 1989, Log Interpretation Principles/Applications. Schlumberger Education Services, Houston, 223 p. Schultz, P.S., Ronen, S., Hattori, M., and Corbett, C., 1994, Seismicguided estimation of log properties, Part 1. The Leading Edge, 13, 305−310, 315. Shinn, Y.J., Chough, S.K., and Hwang, I.G., 2009, Structural development and tectonic evolution of Gunsan Basin (Cretaceous-Tertiary) in the central Yellow Sea. Marine and Petroleum Geology, 27, 500−514. Span, R. and Wagner, W., 1996, A new equation of state for carbon dioxide covering the fluid region from the triple-point temperature to 1100 K at pressures up to 800 MPa. Journal of Physical Chemistry Data, 25, 1509−1596. Van der Meer, B. and Egberts, P., 2008, Calculating subsurface CO2. storage capacities. The Leading Edge, 27, 520−505. Weast, R.C., 1967−1968, Handbook of Chemistry and Physics, 48th Edition, CRC Press, Cleveland, Ohio, F-3. White, C.M., Strazisa, B.R., Granite, E.J., Hoffman, J.S., and Pennline, H.W., 2003, Separation and capture of CO2 from large stationary sources and sequestration in geological formations coalbeds and deep saline aquifers. Journal of Air and Waste Management Association, 53, 645−715. Yanagi, T., Shimizu, T., and Matsuno, T., 1996, Baroclinic eddies south of Cheju Island in the East China Sea. Journal of Oceanography, 52, 763−769. Yang, Q.L., 1992. Geotectonic framework of the East China Sea. In: Watkins, J.S. Zhiqinag, F., and McMillen, K.J. (eds.), Geology and geophysics of continental margins. AAPG Memoir, 53, p. 17−25. Xu, W., Jiao, R., Deyu, Z., Le, J., and Huang, Z., 1997, Geophysical research of deep tectonics and basin evolution in the East China Sea. In: Liu, B., and Li., S. (eds.), Proceedings of the 30th International Geological Congress Vol. 8: Basin Analysis, Global Sedimentary Geology and Sedimentology. VSP BV, Netherlands, p. 320−331. Zhou, Z., Zhao, J., and Yin, P., 1989, Characteristics and tectonic evolution of the East China Sea. In: Zhu, X. (ed.), Chinese Sedimentary Basins: Sedimentary Basins of the World 1, Elsevier, p. 165−179. Manuscript received November 28, 2011 Manuscript accepted June 26, 2012.

(15)