An Evaluation of the Long-Term Stagnancy of Porewater in the Neogene Sedimentary Rocks in Northern Japan

A groundwater dating for very old porewater using Cl and He was applied to the Koetoi andWakkanai formations distributed in the northernmost part in Japan.Measured Cl/Cl in theKoetoi Formationwas 2.6± 2.0× 10 and that in theWakkanai Formation was 8.1 ± 2.5 × 10. These values are similar to Cl/Cl in situ secular equilibrium calculated from chemical compositions of core suggesting that Cl ions and porewater have remained in the formations for much longer than half-life of Cl . He concentration in porewater ranged from 1.1 × 10 to 2.6 × 10 (ccSTP gw−1) and it is much higher than water saturated with air indicating that both formations contain very old porewater. However, the possibility of mixing of young water was indicated because He concentration was lower than that calculated bymultiplication of in situHe production and time after the uplift.This possibility was also supported by Cl, δD, and δ18O data. After combining information on Cl/Cl, He, and δD and δ18O, it was inferred that the porewater in the deep part of the Wakkanai Formation might have been stagnant since the uplift. The porewater in the Koetoi Formation and the shallow part of the Wakkanai Formation were found to be affected by young surface water.


Introduction
Geological disposal is one of the most promising methods for high-level radioactive wastes (HLW).In geological disposal, the transportation of radionuclides by groundwater is one of the most important phenomena for safety assessment of the HLW disposal (groundwater scenario).Some geological formations with low hydraulic conductivity are considered potential host rock for radioactive waste disposal [1][2][3] because these formation will be able to isolate the radionuclides for a long time.The migration of water and radionuclides needs to be understood in the safety assessment of radioactive waste disposal.The groundwater ages (the residence time of it in the subsurface) and their special distribution in the potential host rock can provide information about the migration of porewater in the formation over long time frames. 36Cl is one of the most effective tracers to evaluate the age of very old groundwater because of the long halflife and chemical stability of chloride ion in groundwater. 4He is also considered a useful tracer to investigate very old groundwater.Both of them have been used for the investigation of old groundwater [4][5][6][7][8] and sometimes a combination of 4 He and 36 Cl can provide the accumulation rate of 4 He including "external flux" and make 4 He age more quantitative [9,10].
Many studies have been conducted to determine the age of flowing groundwater by using 36 Cl and 4 He [9,11,12].However, there have been only a few studies that investigated the porewater age in formation with low hydraulic conductivity [1,[13][14][15].The evaluation of porewater age in formations with low hydraulic conductivity is much more difficult compared to that for flowing groundwater because of the following two reasons.
One reason is the technical difficulty.In formations with low permeability, it is difficult to obtain water samples by pumping.Thus, porewater and/or elements such as He gas and Cl − ions have to be extracted from core samples to measure their concentrations and isotopic ratios.Some techniques have already been developed to measure the concentration and/or isotopic ratios in the porewater for He [16,17], Cl − ions [2,18], and D and  18 O [19,20].It would be useful to accumulate the knowledge and experience to address the shortcomings in these techniques.The second reason is the fundamental difficulty of evaluation of porewater of sedimentary rocks.The characteristics of the rock, especially the porosity, change as a result of consolidation during subsidence.In addition, porewater is affected by water from underlying formation during subsidence [21] and this makes the interpretation of the data complicated.In such cases, the evaluation of absolute groundwater age may be difficult.However 36 Cl and 4 He concentrations still contain useful information about the stagnancy and mobility of porewater.Thus, the case studies about evaluation of very old porewater in the formations with low permeability especially in sedimentary rocks are needed to identify what we can learn from using 36 Cl and 4 He data.Addressing the technical issues in the application of the methodologies used for the evaluation of flowing groundwater to very old porewater may also provide very important information for improving dating method for very old porewater.
In this study, the stagnancy of porewater in the Koetoi and Wakkanai formations that were main target rocks for the Horonobe Underground Research Laboratory (URL), in the northern area of Hokkaido Prefecture in Japan, was investigated by 36 Cl and 4 He.The formations are Neogene sedimentary rocks and are characterized by low permeability.The possibility of surface water mixing (intrusion and/or diffusion) was discussed using 4 He concentrations and water isotopes (D and  18 O).Some technical problems were identified regarding the application of 4 He and 36 Cl in this study of these formations.These problems will be further discussed below.
In accordance with Japanese government policy, the Horonobe URL sites will not be considered as candidate sites for a HLW repository.

Geological Setting
2.1.General.The study area is in the Tempoku basin located in northwestern Hokkaido, on the eastern margin of a Neogene to Quaternary sedimentary basin ( [22,23], Figure 1).This basin is part of an active foreland fold-and-thrust belt, developed near the boundary of the Okhotsk and Amurian plates (Figure 1).Seismic reflection surveys indicate that the current compressive E-W neotectonic stress to the west of the Horonobe area was established at around 2 to 3 Ma [24,25].The sediments in the basin consist of (from youngest to oldest) the Sarabetsu formation (alternating beds of conglomerate, sandstone, and mudstone, intercalated with coal seams), the Yuchi formation (fine-to medium-grained sandstones), the Koetoi Formation (Neogene to Quaternary diatomaceous mudstones containing opal-A), the Wakkanai Formation (Neogene siliceous mudstones containing opal-CT), the Masuporo Formation (the upper part: Neogene siliceous mudstone; the lower part: Neogene conglomerate, sandstone, and mudstone), the Magaribuchi formation (Paleogene sandstone and mudstone), the Haboro formation (Paleogene sandstone, mudstone, and coal-bearing beds), the Hakobuchi formation (Cretaceous sandstone, mudstone, tuffaceous sandstone, and coal-bearing beds), and the Yezo formation (Cretaceous sandstone and mudstone).All of the above sediments are overlain by late Pleistocene to Holocene deposits.
The Koetoi, Wakkanai, and Masuporo (upper part) formations, which are composed mainly of homogeneous siliceous rocks, were deposited successively in a marine environment with a thickness of ∼2 km as indicated in Figures 1 and 2 [22,26].The burial and subsidence of these formations occurred throughout the Neogene and Quaternary.Subsequent uplift and denudation started at about 1.3-1.0Ma with the deposition of the Sarabetsu formation [27].The lithological boundary between the Koetoi and Wakkanai formations is a gradual and conformable transition that represents an iso-burial depth of about 1 km at the time of maximum burial [27] (Figure 2).The porosity of these formations decreased from 40-60% to less than 10% with increasing burial depth as a result of compaction during diagenesis of the burial (Figure 2).A large amount of porewater was expelled from these formations as a result of the porosity reduction during the burial.The porewater from deep burial depth may have migrated upwards and replaced the porewater at shallow depths during subsidence.
Based on results from outcrop and core observations, fault zones well develop in the Wakkanai Formation, which comprise fault core(s) of incohesive fault rock (gouge/breccia) with a thickness of decimeters or less, together with surrounding damage zones meters or less in width where fractures are densely developed but fracture mineralization is limited [29].On the other hand, in the Koetoi Formation, fault zones are less common but shear fractures without fault rocks and fracture mineralization are prevalent [30].Many flow anomalies are detected at fault zones in the Wakkanai Formation and at a few well-connected shear fractures in the Koetoi Formation by flowing fluid electric conductivity (FFEC) logging in boreholes [29,30].The transmissivities of the flow anomalies range 1 × 10 −7 to 2 × 10 −9 m 2 s −1 in the Ketoi formation and 2 × 10 −10 to 3×10 −6 m 2 s −1 in the Wakkanai formation [30].The hydraulic conductivity of the formations from packer tests (interval: meters to decameters) range 9 × 10 −10 to 5 × 10 −6 m s −1 for the Koetoi formation and 1 × 10 −12 to 5 × 10 −5 m s −1 for the Wakkanai formations [22].The hydraulic conductivity of rocks (matrices) from laboratory hydraulic tests is 2 × 10 −11 to 5 × 10 −10 m/s and 5 × 10 −13 to 3 × 10 −11 m s −1 for the Koetoi and Wakkanai formations, respectively [22].
The formation process of the porewater chemistry in the studied section of the Koetoi and Wakkanai formations is summarized as follows: the origin of the porewater is sea water [31].The sea water had reached isotopic equilibration with the surrounding rock (clay/silica minerals) through    [27].Porosity and bulk chemistry data are from Ishii et al. [22] and Tsuji and Yokoi [26].
long-term water-rock interactions [32][33][34][35], resulting in D = −30‰ to −20‰ [36].In addition, the sea water had been significantly diluted by dehydration of clay/silica minerals during burial [33,35,37].During the uplift and denudation after ca.1.3-1.0Ma, the altered sea water has been diluted with surface meteoric water in the Koetoi Formation and the shallow levels of the Wakkanai formations [32,36], where 14 C is detected from the diluted sea water [38,39].The intrusion of surface water probably accelerated due to a temporary increase in the hydraulic gradient during glacial periods (likely 20 ka) by advection through fractures [23,32].

Sampling Points.
The core samples were collected from deep drillings at HDB-9, 10, 11, and HCD-2 (Figure 1).The lengths of these boreholes are 520 m, 550 m, 1020 m, and 700 m, respectively.The boreholes intercept the Koetoi Formation and the shallow part of the Wakkanai Formation.Sampling depths are shown in Table 1.HDB-9, HDB-10, and HDB-11 were drilled to vertical direction, while HCD-2 was drilled to slanting direction. 4He Concentration in Porewater.The measurement of 4 He and Ne concentrations and 3 He/ 4 He was conducted according to the procedures proposed by Osenbrück et al. [16].

Measurement of
The freshly drilled cores were used for measurements.Just after the cores were lifted up from the drill holes, core was taken from the core-tube and cut into 0.1 m long cylindrical sections using a dry cutter.The surface of the core was wiped with waste to remove the drilling fluid.The cut core was put into a stainless steel vessel with copper tube (Figure 3) and the vessel was sealed airtight.The inside of the vessel was vacuumed through the copper tube with a pump to remove the air in the head space of the vessel.The appropriate vacuum time, which can remove air in the head space and preserve He and Ne concentration in the porewater, was determined from preliminary experiment [17].In the preliminary experiment, some kinds of sand stones were suspended into He saturated water and they were used as simulated samples.The relationships between vacuum time, pressure in the vessel, and concentration of He in porewater were investigated for simulated samples.It was found pressure inside the vessel decreased according to the vacuum time in four steps: (1) sharp initial decrease, (2) temporal stability, (3) relative sharp decrease again, and (4) very slow decrease.In the steps (1) and (2) a fraction of air in the head space was found remaining and concentration of He in porewater began to decrease significantly after pressure reached to step (4).Thus, in the sampling, we monitored the pressure inside the vessel and stopped the vacuum just after confirming step (4) in the pressure.The copper tube was pinched off with an iron clamp after pumping and then the vessel was sent to the laboratory.The sampling procedure described above was carried out within 60 min after the core was lifted up to the surface to prevent the loss of dissolved gas by diffusion to the air.In the laboratory, the vessel was kept at room temperature (20 to 25 ∘ C) for 3 to 5 months to allow the gases dissolved in the porewater to diffuse into the head space of the vessel.
The amount of He and Ne in the head space and 3 He/ 4 He was measured using a VG 5400 (GVI now Thermo Fisher Scientific, Massachusetts, USA) mass spectrometer with a purification system.The head space gas was purified using a cold trap, Ti getter, and cryostat and purified gas was measured using the VG5400.The peak intensity corresponding to 4 He was measured and it was compared with that of purified air with a known amount.The amount of 4 He was determined from the results of comparison described above and was converted to He concentration using water content as estimated by the weight of rock core before and after drying.The analytical precision was ±5% (1 sigma), while the data contained the other errors such as sampling.
The procedure described above was repeated for some samples to evaluate residual He.However, less than 2% of He was extracted by the second procedure compared with the amount of He extracted by the first procedure.Thus, dissolved He in the porewater was confirmed to have been extracted almost completely to the head space by the procedure described above. 36Cl/Cl in Porewater.The freshly drilled cores were obtained in the study area and the core was vacuum packed with an aluminum bag just after the core was lifted up to prevent drying and oxidizing of the sample.The packed cores were then sent to the laboratory.The surface of the core was trimmed by a dry cutter to remove parts contaminated by drilling fluid and the trimmed core was subjected to the following procedures.

Measurement of
The trimmed core was put into the cylindrical vessel and compressed to pressure of 300 MPa.Porewater was extracted by maintaining the pressure until shrinkage stopped.The extracted porewater was acidified with 2.0 mol/L HNO 3 solution and then 0.3 mol/L AgNO 3 solution was added to obtain AgCl precipitation.The precipitation was rinsed with pure water and dissolved by 3.0 mol/L NH 3 solution.BaNO 3 solution was added to reduce the SO 4 2− ion concentration with BaSO 4 precipitation [43] and then AgCl precipitation was obtained by adding HNO 3 and AgNO 3 solutions.In some cases, column chromatography was applied to separate Cl and SO 4 2− ions [44].Purified AgCl was used for measurement of 36 Cl/Cl using accelerator mass spectrometry (AMS) in the Australia National University [45] or the Purdue University [46].The analytical precision of 36 Cl/Cl measurement was about ±1 to 4 × 10 −15 .

Measurement of Ion Concentrations and 𝛿D and 𝛿 18 O.
The porewater was extracted by compression using the same procedure described in Section 3.2.Ion concentrations in the extracted porewater were measured using an ion chromatograph (IC-7000, Yokokawa Analytical Systems, Tokyo, Japan).The D and  18 O values for the water were measured by isotope ratio mass spectrometer (Iso-Prime: GV Instruments Ltd., Manchester, UK).The isotope ratios of hydrogen and oxygen are expressed in the following -notation:  = () sample /() standard − 1 where  is the isotopic ratio (D/H or 18 O/ 16 O) relative to the international standards VSMOW [47].The analytical precision of D was ±0.5‰ and that of  18 O was ±0.1‰. 36Cl/Cl in the Koetoi and Wakkanai Formations.Depth profiles of Cl − concentration and 36 Cl/Cl in the Koetoi and Wakkanai formations are shown in Figure 4.Both formations were folded during tectonics events and the surface was removed by denudation.Therefore, the depths of the boundary between the Koetoi and Wakkanai formations vary depending on the location of each borehole.In Figure 4, the boundary of the Koetoi and Wakkanai formations is indicated as 0 m, and the relative depth from the boundary is indicated by the vertical axis.In the case of HDB-9, the surface was denudated and the Koetoi Formation was not present and only the Wakkanai Formation was found, even in the shallow part of the drilling.As the boundary of both formations was estimated to be near the surface of HDB-9 from geophysical logging data [40], the length of the bore hole was expressed as the depth from the boundary of both formations. 36Cl/Cl in the Koetoi Formation ranged from 0 to 4.6 × 10 −15 and the average of 5 points was 2.6 × 10 −15 . 36Cl/Cl in the Wakkanai Formation ranged from 3.6 to 15.9 × 10 −15 with an average of 8.1 × 10 −15 for 21 points.The error bars indicated in Figure 4 represent the uncertainties in the 36 Cl/Cl measurement using AMS.All 36 Cl/Cl data were found to have relatively large errors because the 36 Cl/Cl in both formations was very low (below 10 × 10 −15 ).However, it was clearly shown that 36 Cl/Cl in the Wakkanai Formation had higher values than those in the Koetoi Formation. 4He Concentration and 3 He/ 4 He in the Koetoi and Wakkanai Formations.A depth profile of 4 He concentrations is shown in Figure 5: the vertical axis (depth) was expressed in the same manner as in Section 4.1 (Figure 4).The 4 He, 3 He/ 4 He and Ne concentrations are shown in Table 1.The Ne concentrations ranged from 2.5×10 −8 to 3.1× 10 −6 cc STP g w −1 and were different from the Ne concentration in water saturated with air (WSA: 2.0 × 10 −7 cc STP g w −1 air saturated fresh water at 25 ∘ C [48]).The difference indicated that contamination by air or degassing of dissolved gases might have taken place during the drilling and/or sampling procedures.The procedure to correct the 4 He concentrations shown in Figure 5 using Ne concentrations is now described.When the Ne concentration exceeded 2.0 × 10 −7 cc STP g w −1 , the sample was considered to be contaminated by air.In this case, the following equation was applied to correct the 4 He concentration:

Depth Profile of
where [He] co was the He concentration after correction, [He] sam and [Ne] sam were the He and Ne concentrations, respectively, in the sample, [Ne] air was the Ne concentration in WSA, and [He/Ne] air was the ratio of He and Ne in the air.Some data with high Ne concentrations (over 1 × 10 −6 cc STP g w −1 ) were eliminated because such high Ne concentrations indicated the samples were highly contaminated.
When the Ne concentration was below 2×10 −7 cc STP g w −1 , the gas dissolved in porewater was considered to dissipate by degassing.In this case, the following equation was applied to correct the 4 He concentration: where [He/Ne] sam was the ratio of He to Ne in the sample.In (2), the degassing rate of He and Ne was assumed to be the same, although this was not verified.Degassing was very significant in some samples and (2) was not deemed to satisfactorily correct for the degassing of He in samples where significant degassing occurred, as described in Section 5.2.
As shown in Figure 5, He concentrations for all samples obtained in this investigation were much higher compared with that in WSA (4 to 5 × 10 −8 cc STP g w −1 ) indicating that porewater has stayed in subsurface for a long time.
The relationships between 3 He/ 4 He and 4 He concentrations are shown in Figure 6.The 3 He/ 4 He observed in this investigation ranged from 2 to 10 × 10 −7 .These values of 3 He/ 4 He were lower than He in atmosphere and higher than that produced from rock in Koetoi and Wakkanai formations calculated from Li, U, and Th (Section 5.2).This indicates that He in the porewater might be affected by He from the other sources, such as the mantle.

Depth Profiles of 𝛿D and 𝛿 18 O and Relationships between
Cl − Concentrations and D.Depth profiles of D and  18 O are shown in Figure 7. Expected end members, which are sea water and shallow spring water, were also shown in Figure 7 as a dashed line [49].In the deep part (below −200 m) of the Wakkanai Formation, D and  18 O reached −20 and +4‰, respectively.The  18 O values in the deep Wakkanai Formation were greater than that of sea water suggesting interaction between water and rock might have taken place in the past, as mentioned in Section 2.1 (Geological Setting).In the Koetoi Formation and in the shallow part of the Wakkanai Formation D and  18 O were lower than in the deep part of the Wakkanai formation and were very similar to the values of spring water in the boundary of the Koetoi and Wakkanai formations and in the upper part of the Koetoi Formation.The part of porewater in both formations was considered to be affected by meteoric water because D and  18 O had similar values to spring water.The relationships between D and  18 O were indicated in Figure 8.The values for sea and spring water are also indicated in Figure 8.

Discussion
As mentioned in Section 2.1 (Geological Setting), subsidence may have caused the porosity of the formations to decrease and water to be expelled and migrate upwards.The porewater has reached its present hydrochemical state as a result of complex mixture of processes such as sea water confined in the formation during sedimentation, mixing with water from underlying formations, water-rock interaction, and expelling of porewater by consolidation.Thus, the evaluation of the absolute age of porewater is difficult because quantitative information describing all the phenomena described above is not available.However, the 36 Cl and 4 He can provide the information in the migration of porewater in long-terms.In addition, stagnancy of porewater in both formations could be discussed in terms of 36 Cl and 4 He since the uplift (1.3-1.0Ma, [31]) because the characteristics of the formation including porosity have not changed since the uplift.
The stagnancy and mobility of porewater since the uplift are now mainly discussed using 36 Cl, 4 He, and D and  18 O. 36Cl/Cl. 36Cl/Cl in porewater decreases through the decay of 36 Cl and increases through the reaction of 35 Cl(n,) 36 Cl [9]. 36Cl/Cl approaches an in situ secular equilibrium state (( 36 Cl/Cl) se ) with increase of residence time of Cl − ions in porewater.Figure 9 showed the example of relationships between time versus 36 Cl/Cl values and as indicated in Figure 9, in both cases, initial 36 Cl/Cl higher than ( 36 Cl/Cl) se (Figure 9(a)) and lower than ( 36 Cl/Cl) se (Figure 9(b)) ( 36 Cl/Cl) se becomes nearly equal to ( 36 Cl/Cl) se if Cl − ions have remained in a formation for over 1 million years.Conversely, 36 Cl/Cl may show higher values compared with ( 36 Cl/Cl) se if significant amount of Cl ions flows into the formation with porewater from underlying formations because ( 36 Cl/Cl) se of underlying formation (the Masuporo formation) was much higher than that of the Wakkanai formation as described in Section 5.1.1.

Evaluation of Porewater Stagnancy with
In this study, it was considered that 36 Cl/Cl could be applied to understand the stagnancy of porewater since the uplift because characteristics of pore probably have not changed significantly since the uplift.Even in this case, absolute value of 36 Cl/Cl age since uplift could not be estimated because it is impossible to know the 36 Cl/Cl in porewater just after the uplift.However, as described above, if measured 36 Cl/Cl were compared to ( 36 Cl/Cl) se and they showed agreement, we could know Cl − ion has stayed in porewater for a long time compared to half-life of 36 Cl (3.01× 10 5 years: [50]).

Calculation of In Situ Secular Equilibrium (( 36 Cl/Cl) 𝑠𝑒 ).
( 36 Cl/Cl) se in the Koetoi and Wakkanai formations was calculated using the chemical composition of the rocks. 36Cl is produced in -situ when thermal neutrons are absorbed by 35 Cl; the reaction is expressed as 35 Cl(n,) 36 Cl.The reaction of 39 K(n,) 36 Cl also produces 36 Cl.However it is generally insignificant compared with 35 Cl(n,) 36 Cl [9,50].Thermal neutrons are involved in the 35 Cl(n,) 36 Cl reaction produced by (, n) reactions, and -rays are produced by the alpha decay of U and Th.The thermal neutron flux in rock depends on the concentrations of U and Th and the concentration of the light elements related to (, n) reactions.
In addition, porosity has significant effect on thermal neutron flux because H 2 O in pore has very large absorption cross section of neutron.The ( 36 Cl/Cl) se can be calculated as follows [51]: The thermal neutron flux () was determined by following equation [52,53]: where  has units n cm −2 y −1 ,  is the neutron production rate (n cm −3 y −1 ), and Σ  is the macroscopic absorption cross section for thermal neutron absorption (cm −1 ).Σ  can be calculated by multiplying the number of atom in unit volume of rock and absorption cross section.The thermal neutron production rate () was calculated by following equation [52]: where   () is the neutron production rate from the reaction with cosmic muons (likely to be negligible below 100 m depth from the ground surface);  is the neutron production rate from the spontaneous fission of 238 U (=0.429); and  U and  Th are the neutron production yields for (, n) reactions in the light elements for 1 ppm of U and Th.Values of  U and  Th from Heaton et al. [54] were used in (5).
The parameters used for the calculation and calculated ( 36 Cl/Cl) se were shown in Table 2.The calculated ( 36 Cl/Cl) se was 4.6 ± 1.9 × 10 −15 and 7.6 ± 1.3 × 10 −15 for the Koetoi and In both cases, 36 Cl/Cl approaches to ( 36 Cl/Cl) se over 1 million years.15 × 10 −15 was chosen because it was a calculated ( 36 Cl/Cl) se for the Masuporo formation that is underlying to the Wakkanai formation.0.76 × 10 −15 was chosen because it is a 36 Cl/Cl value for sea water [42] and ( 36 Cl/Cl) se = 7.6 × 10 −15 was adopted because it was calculated value for the Wakkanai formation.
Wakkanai formations, respectively.As shown in  36 Cl/Cl uncertainties in the measurement of 36 Cl/Cl are relatively large due to small value of 36 Cl/Cl and it is impossible to know 36 Cl/Cl value just after the uplift.The distribution of 36 Cl/Cl (Figure 4) clearly shows that 36 Cl/Cl values in the Wakkanai Formation were significantly larger than those of the Koetoi Formation. 36Cl/Cl at the bottom of the Koetoi Formation may have been affected by higher 36 Cl/Cl in the Wakkanai Formation if significant

He/ 4
He, and the calculation results.The data from Shibata et al. [28] for the cross section of the reaction with neutrons were used in this study.amounts of groundwater and Cl − ions migrated into the Koetoi Formation from the Wakkanai Formation.However higher 36 Cl/Cl values in the Koetoi Formation were not observed.The clear difference in 36 Cl/Cl for the two formations suggested that migration of groundwater and Cl − ions from the Wakkanai Formation to the Koetoi Formation has not been significant since the uplift.( 36 Cl/Cl) se of underlying formation of the Wakkanai formation (the Masuporo formation) was also calculated and it was 15 × 10 −15 .That value was almost as twice as that of the Wakkanai Formation.Thus the migration of groundwater and Cl − ions from the Masuporo Formation to the Wakkanai Formation is significant; the distribution of 36 Cl/Cl may be affected, especially the deep part of the Wakkanai formation.However, the distribution of 36 Cl/Cl in the Wakkanai Formation did not show a dependence on the depth of the sampling points.This indicated that porewater in the Wakkanai Formation investigated in this study has also not been affected by the water from underlying formations since the uplift.The results described above suggested that Cl − ions and thus porewater may have stayed in the pores since the geological setting reached its present state (1.3-1.0Ma). 36Cl for the Evaluation of Porewater in Sedimentary Rock.Sedimentary rocks, when compared with granite, have lower U and Th concentrations and a higher content of water (porosity).Therefore, the values of ( 36 Cl/Cl) se are generally very small (below 10×10 −15 ) and thus the error bars in the measurement 36 Cl/Cl became relatively large.In the case where sea water is a source of porewater, both initial 36 Cl/Cl [42] and ( 36 Cl/Cl) se are small and differences within 10 × 10 −15 have to be measured and evaluated.

The Issues Identified regarding the Use of
In such cases, the measurement of very low 36 Cl/Cl values with high accuracy is required.Measurements using high accelerating voltage were reported to be effective in obtaining a high accuracy [55].The improvement in the precision of the separation between Cl and S as a pretreatment to AMS measurement may also contribute to high accuracy measurement [44]. 4He.The 36 Cl/Cl results indicated that porewater in the Wakkanai and Koetoi formations has been old in both formations.In this section, the stagnancy of the porewater in both formations since the uplift was checked using 4 He concentrations.

Data Selection for
Evaluating.The effect of degassing or contamination on 4 He concentration was corrected using the Ne concentration, as described in Section 4.2.However, 4 He concentration might not be completely corrected by this procedure.Sometimes a lot of bubbles were observed on the surface of the core (Figure 11) when it was pulled up from subsurface.This indicates that degassing was vigorously taking place.This was especially true for HDB-11, where a significant amount of bubbles were found on the surface of the core.This could be explained by the borehole length of HDB-11, which was much longer (up to 1000 m) than other e point bubble was observed boreholes, and thus the big difference in pressure between subsurface and surface caused the vigorous degassing.
In this study, the data from sample cores with significant amounts of bubbles in the sampling procedure were eliminated from further evaluation; the degassing behavior of He and Ne might be different with such vigorous degassing and the He concentration might not be corrected by the Ne method.In addition, the He concentration in the deeper part of HDB-11 was actually much lower than that at similar depths in the other boreholes.The data obtained from core samples with significant amounts of bubbles are indicated in Figure 5 and Table 1 and these data were not used for evaluation of groundwater stagnancy. 4He.Increases in the rate of He concentration in porewater by in situ production depend on the concentration of U and Th, porosity, and density of the rocks.These parameters were obtained from core measurements and thus could not be applied to the porewater before the uplift because these parameters had been significantly changed during the consolidation that occurred in the subsidence of both formations.However, parameters obtained from the cores can be applied to porewater since the uplift because they may not have significantly changed after the uplift.Thus, in situ production rate of 4 He calculated below was used for evaluating the stability of porewater since the uplift. 4He production rate can be calculated by ( 6) [56].

Calculation of In Situ Production Rate of
where A is the production rate of He by U and Th (atoms g −1 y −1 ) and [U] and [Th] are the concentrations of U and Th (g g −1 ), respectively.In the saturated zone, 4 He produced by U and Th in the rock is accumulated by groundwater in contact with the rock.Thus the increase in the rate of He concentration by in situ He production can be calculated using (7) [13].
[He] ac-in where [He] ac-in is the accumulation rate from in situ He production,  He is the release coefficient of He from rock and set as 1.0,   and   is the density of rock and water, respectively, and  is the porosity of the rock.
[He] ac-in for the Koetoi and Wakkanai formations is 1.4 ± 0.6 × 10 −12 and 2.6 ± 0.5 × 10 −12 (cc STP g w −1 y −1 ), respectively.The parameters used to calculate [He] ac-in and the values of [He] ac-in at each sampling points were listed in Table 2.The difference of production rate for Koetoi and Wakkanai formations was mainly due to the difference in porosity. 3He/ 4 He and the Effect of Mantle He. 3 He/ 4 He is a useful indicator of the origin of He [35,[57][58][59].In the following, 3 He/ 4 He of He produced in both the Koetoi and Wakkanai formations was calculated according to Tolstikhin et al., [60].

5.2.3.
In subsurface, 3 He is mainly produced by n- reaction of 6 Li that produces 3 H followed by -decay of 3 H that produces 3 He and the production rate of 3 He from the rock is expressed by [60] 3 where 3 He ac the production rate of 3 He from 1 g of rock (atoms g rock −1 s −1 ),  is the neutron flux in the formation (n (cm 2 ) −1 s −1 ), (Li) is the reaction cross section in the reaction of 6 Li to 3 He (barn: converted to cm 2 in the calculation), and [ 6 Li] is the concentration of 6 Li (atoms g rock −1 ).Thus 3 He/ 4 He produced from the rock can be calculated by combining (6) and (8).The parameters used for the calculation are also shown in Table 2.The neutron flux calculated in 36 Cl/Cl eq was used for this calculation (Section 5.1.1).The calculated results for 3 He/ 4 He produced in the Koetoi Formation ranged from 3.1×10 −9 to 5.7×10 −9 and in the Wakkanai formation ranged from 4.0 × 10 −9 to 1.38 × 10 −8 .
As shown in Figure 6, the calculated 3 He/ 4 He was much lower than the measured values.This indicates that not only He produced in situ (radiogenic) but also He from other origins accumulated in the porewater.He with a high 3 He/ 4 He (up to 1 × 10 −5 ) was observed in Japan and was ascribed to He affected by the mantle [61][62][63].In the evaluation of porewater stagnancy with He, we will compare the measured He concentration and the He concentration calculated with in situ accumulation rate (Section 5.2.4).Therefore, the effect of mantle He should be removed for the evaluation.The He concentration without mantle effects He ([He] w ) was calculated using the following procedure.The ratio of [He] w in porewater (  ) was calculated from 3 He/ 4 He using  3 and ranged from 0.92 to 0.99.He concentration without the effect of mantle He was calculated using where [He] m is measured He concentration.The calculated [He] w is also listed in Table 3.
As shown in Figure 6, 3 He/ 4 He in borehole HCD-2 was much higher than in other boreholes.We could not explain why 3 He/ 4 He was different in borehole HCD-2 even though the distance from the other boreholes is within 2 to 3 km.However it is possible that the effect of water coming from underlying formations may be different for each borehole.Some previous studies indicated the possibility that faults play an important role in transporting mantle fluid [62] and this may also be applicable to the present study. 4He.The 36 Cl/Cl suggested that Cl − ion has stayed in the porewater for a long time compared to half-life of 36 Cl since the uplift of both formations (<1.3 to 1.0 Ma) (Section 5.1.2),and thus there is a possibility that porewater has remained in pore since the uplift.In this study, we evaluated the stagnancy of porewater since the uplift as follows.The He concentration obtained by the multiplication of 4 He accumulation rate by radiogenic reactions (calculated in Section 5.2.2) and time (1.0 million years, since the uplift) was defined as [He] aup in this study.If the porewater has stayed in the pore over 1 million years since the uplift, the He concentration should significantly exceed [He] aup because He had accumulated in porewater before uplift and, in addition, He might be accumulated by external flux.Thus, if the He concentration is lower or nearly equal to [He] aup , it could indicate the porewater was affected by younger water.

Evaluation of Porewater Stagnancy Using
In Figure 12, measured [He] w and [He] aup are compared for each of the sampling points and the dashed line is the line of  = .For some points, the rock characteristics listed in Table 2 were not measured, and only He concentration and 36 Cl/Cl were measured.In these cases the [He] aup average for the Koetoi or Wakkanai formations was used for the plot.In the Koetoi Formation, all the data plots were near the dashed line.This result suggests the possibility of young water mixing into the Koetoi Formation and nearly all porewater in the Koetoi Formation may be affected by young water.In the Wakkanai Formation, some data plots were near or under the dashed line.This result indicated a part of porewater in the Wakkanai Formation is also affected by young water.
Thus, in contrast to 36 C/Cl results, the 4 He results indicate the possibility of young water mixing into the Koetoi Formation and a part of the Wakkanai Formation.This possibility will be discussed further in Section 5.3 by incorporating D and Cl − concentration.

The Problems Identified When Dating Groundwater
Using 4 He.As described in Section 5.2.1, the degassing of gases from porewater in the core samples was identified as one of the most significant issues for porewater evaluation with 4 He in this study.In some samples, correction of the He concentration using the Ne concentration seemed to be effective.However, in some samples significant amount of bubbles were observed, indicating vigorous degassing.For these samples, the correction was not effective because the degassing rate of He and Ne may be different for such formations contains high methane concentration [23], and the gas stripping of methane might cause the vigorous degassing observed.Information about the effect of coexisting gases is also important in understanding the degassing behavior of He and Ne.As mentioned above, the targeted porewater contained a lot of methane in this study.Methane could interfere with the accurate measurement of noble gases or it could be a cause of contamination in the preparation system for noble gases measurement because it may take long time to be removed completely by the Ti getter that is frequently used in the preparation system.Additional procedures may be required to remove methane from groundwater or porewater with high methane concentrations, such as the porewater found in the study area.In addition, Cl − concentration decreases significantly in the boundary part.Thus it could be inferred that surface water may intrude along boundary part of both formations, as indicated in the past studies [23,32].Thus the intrusion of surface water along the fracture in boundary area of both formations may contribute the formation of He, Cl − , D, and  18 O profiles.In addition diffusion may also affect their profile.The relationships between Cl − and D are indicated in Figure 13.The relationships are roughly linear, indicating that D and Cl − concentrations changed with the same mechanism, mixing of porewater and surface water.

Intrusion and/or Diffusion of
Relationships between D and He concentration are shown in Figure 14.In Figure 14, the data plotted under or close to the dashed line in Figure 12 were indicated with open circles.Figure 14 clearly shows that D has a correlation with He concentration, if the data from HDB-11 are excluded.This correlation indicates that the possibility of young water intrusion and the inferences from the He results are supported by the D results.Some samples from HDB-11 had lower values of He concentration than from the other boreholes even though they had relatively large D (-30 to -20‰).This could not be explained by mixing because of the large D.The degassing may explain why some data from HDB-11 showed lower He concentrations.The degassing of He has taken place in the sample from HDB-11 even though data from the cores with bubbles on the surface were eliminated.

Evaluation of Porewater Stagnancy Since the Uplift Using
All the Information. 36Cl/Cl measurements showed that Cl − ions may have stayed in porewater for a long time.However, He concentrations suggested the possibility of young water mixing to porewater in the Koetoi Formation and part of the Wakkanai Formation.The D,  18 O, and Cl − results support the possibility and evidenced the mixing of porewater and surface water.
By considering all the information described above, originally the porewater in the Koetoi Formation was considered to be very old.However, it was affected by the mixing of young surface water.Part of the porewater in the Wakkanai Formation was also affected by the mixing of young surface water.At in the boundary of two formations, He, D, and Cl − concentration values were smaller than in the deep part of the Wakkanai Formation.This suggests that young surface water flowed into the top part of the Wakkanai Formation.This suggestion is in good agreement with the results of previous work [36].It is inferred that physical differences between the two formations created a high permeability zone at their boundaries.The intrusion of surface water described above and the diffusion to the surface and/or boundary area of both formations are considered to contribute for formation of He, Cl − , D, and  18 O profiles in the studied area.
In the deep part of the Wakkanai Formation, D and 36 Cl/Cl were found to be stable with depth.This suggests that porewater in the deep part of the Wakkanai Formation has not affected water from underlying formations. 36Cl/Cl measured agreed with calculated ( 36 Cl/Cl) se , indicating Cl − ions had remained in porewater for much longer time than the half-life of 36 Cl.In addition, 4 He concentration greatly exceeded [He] aup and this confirmed that groundwater in porewater in the deeper part of the Wakkanai formation was old.According to the results described above, it could be inferred that porewater in the deep part of the Wakkanai Formation may have been stagnant in the formation since the uplift.
The study showed that a combination of 36 Cl/Cl and 4 He measurement could provide useful information for the evaluation of the stagnancy of porewater and the mixing of surface water in sedimentary rocks.The geological history should be understood when determining the evaluation period because the change in the characteristics of the rocks and/or the water emanating from underlying formations during subsidence make the interpretation of 36 Cl/Cl and 4 He data very complicated.
The comparison of measured 36 Cl/Cl and calculated ( 36 Cl/Cl) eq could provide important information about the stagnancy of porewater.However, this information should be combined with other indicators such as 4 He and D because 36 Cl/Cl was not sensitive to surface water with low salinity intruding into old saline water.In this study, He analysis did not reveal the absolute age of the porewater.However He analysis was found to be very useful to find the intrusion and flow path of surface water.This may be a very useful property for use in site selection for the disposal of radioactive wastes.

Conclusions
A groundwater dating for very old groundwater using 36 Cl and 4 He was applied to porewater in the Koetoi and Wakkanai formations distributed in the north area of Hokkaido Prefecture in Japan.In addition, the stagnancy of porewater since the uplift was discussed with measurement results of 36 Cl, 4 He, and D and  18 O in the Koetoi and Wakkanai formations with low permeability.
Porewater from the core sample was obtained by compression and 36 Cl/Cl in porewater was measured.The 36 Cl/Cl measurement found that 36 Cl/Cl in porewater was equal to 36 Cl/Cl in secular equilibrium.This result indicated the Cl − ions in both formations have remained in the formations for a longer time with half-life of 36 Cl.This also indicated the possibility that porewater has been stagnant since the uplift.
He concentration in porewater ranged from 1.1 × 10 −6 to 2.6 × 10 −5 and much higher than water saturated with air indicating the porewater in both formations contains very old groundwater.He concentration was applied to confirm the stagnancy of porewater after the uplift.In all samples in the Koetoi Formation and part of the shallow part of the Wakkanai Formation, He concentration clearly showed lower or same values than calculated He concentration by multiplication of in situ He production and time after the uplift (1.0 million years).This suggested young water intrusion into the Koetoi and the shallow part of the Wakkanai formations.The young surface water intrusion was also supported by D results.This intrusion and diffusion were inferred to form the depth profile of each material in the both formations.It was inferred that the porewater in deep part of the Wakkanai Formation has been stable since the uplift.Furthermore, the porewater in the Koetoi Formation and the shallower part of the Wakkanai Formation was found to be affected by young surface water.
In this study, the absolute age of porewater could not be evaluated using 36 Cl/Cl and 4 He because of the complicated geological history.However, 36 Cl/Cl and 4 He can provide useful information on porewater stagnancy since the uplift.More information and studies are needed to understand the applicability of age tracers to porewater age determination in sedimentary rocks.
This study revealed some technical issues regarding 36 Cl/ Cl and 4 He measurements.Measurements and pretreatment methods that allow very low 36 Cl/Cl (below 10 × 10 −15 ) measurements are required.This would enable the investigation of the relative contributions of groundwater originating from sea water contacting with rocks with low U and Th concentrations.The degassing of dissolved gases during sampling was identified as one of the most significant problems for the correct measurement of 4 He, and therefore the development of sampling methods that prevent degassing is required.In addition, understanding the degassing behavior of He and Ne is considered important in applying numerical corrections to the quantification of the degassing of He using Ne values.The effect of coexisting gases on He and Ne degassing behavior is especially important because the degassing behavior may be affected by the degassing of coexisting gases.

Figure 1 :
Figure 1: Maps and cross-sections showing the location of the Horonobe underground research facility and the boreholes (after [27, 40]): (a) location and geological maps and (b) geological cross section of the Horonobe area.The plate boundaries and the direction of plate movement are from Wei and Seno [41].

Figure 2 :
Figure2: Geological column and variations in porosity and bulk chemical composition with maximum burial depth.Sb, Yc, Kt, Wk, and Mp indicate the Sarabetsu, Yuchi, Koetoi, Wakkanai, and Masuporo formations, respectively.The age of the studied section is based on FT ages of intercalated tuff beds reported by Ishii et al.[27].Porosity and bulk chemistry data are from Ishii et al.[22] and Tsuji and Yokoi[26].

Figure 3 :
Figure 3: Appearance and schematics diagram of core sampling vessel used for dissolved gases measurement.

Figure 4 :
Figure 4: Depth profile of Cl − concentration 36 Cl/Cl in the Koetoi and Wakkanai formations.* Depth: the boundary of the Koetoi and Wakkanai formations is indicated as 0 m, and the relative depth from the boundary is indicated by the vertical axis.

Figure 5 :
Figure 5: Depth profile of 4 He concentration in the Koetoi and Wakkanai formations plots with open circle: significant amounts of bubbles were observed on the surface of the core during the sampling procedure.

Figure 7 :Figure 8 :
Figure 7: Depth profile of D and  18 O concentration in the Koetoi and Wakkanai formations.

Figure 10 :
Figure 10: Comparison of calculated ( 36 Cl/Cl) eq and measured 36 Cl/Cl; the dashed line in the figure is the line of  = .

Figure 11 :
Figure 11: The core sample on which bubbles were observed.

Figure 12 :
Figure 12: Comparison of [He] w and [He] aup : [He] w is defined as the He concentration without the effect of mantle He, as calculated by (10); [He] aup is defined as the He concentration obtained by the multiplication of the accumulation rate of 4 He by radiogenic effects (calculated in Section 5.2.2) and time (1.0 million years).

Figure 13 :
Figure 13: The relationships between D and Cl − concentration in the porewater in the Koetoi and Wakkanai formations.

Figure 14 :
Figure 14: The relationships between D and He concentration in the porewater in the Koetoi and Wakkanai formations.Plots with open circle: the data plotted under or close to the dashed line in Figure 12.

Table 1 :
Sampling points (depth) investigated in this study and He and Ne concentrations, 3

Table 2 ,
we have calculated ( 36 Cl/Cl) se of 4 and 17 samples for the Koetoi and Wakkanai formation, respectively.The uncertainties described above are one sigma of 4 and 17 samples for the each formation.5.1.2.Evaluation of Stagnancy of Groundwater with36Cl/Cl.The average of calculated ( 36 Cl/Cl) se value for each formation is indicated in Figure4.In addition, calculated values of ( 36 Cl/Cl) se (Table2) are compared to measured36Cl/Cl values in Figure10.The ( 36 Cl/Cl) se calculated in Table2did not use the average values for each formation but used the measured values at each sampling point because the porosity and concentration of each element are slightly different to each other.Thus, in Figure10, ( 36 Cl/Cl) se and measured36Cl/Cl were compared at each sampling point.Nearly all data in Figure10are plotted on the  =  line, indicating that measured36Cl/Cl agreed with ( 36 Cl/Cl) se at each sampling point.This agreement suggested that36Cl/Cl had almost reached secular equilibrium state at all sampling points indicating much longer time compared to half-life of36Cl has passed since Cl − ion stayed in the porewater.In this case, we can not precisely indicate the period that Cl − ion stayed in the porewater from

Table 2 :
Parameters used for the calculation of (36

Table 3 :
(10)arison of [He] w and [He] aup ; [He] w was calculated from(10)and [He] aup was calculated by the multiplication of the accumulation rate of4He by radiogenic effects (calculated in Section 5.2.2) and time (1.0 million years).
Surface Water into the Koetoi and Wakkanai Formations.As shown in Figure7, D and 18O reached -20 and +4‰, respectively, in the deep part of the Wakkanai formation.Much data were plotted near (−20, +4) in D versus 18O plot (Figure8), and this could be inferred as representing one of the end members of porewater in the Koetoi and Wakkanai formations.The other data were plotted on a line from (−20, +4) to (−72, −10.8), the value of spring water from shallow wells.Ths indicates that porewater and surface water were mixed in the Koetoi Formation and part of the Wakkanai Formation.As described in Section 4.3, in the boundary part of the Koetoi and Wakkanai formations, D and 18O approach to those of surface spring water.