GRIMSEL现场的⽓体渗透试验
第25卷第4期岩⽯⼒学与⼯程学报V ol.25 No.4 2006年4⽉Chine Journal of Rock Mechanics and Engineering April，2006 GAS MIGRATION TEST AT THE GRIMSEL
TEST SITE
Fujiwara A1，Okamoto S1，Tsuboya T1，V omvoris S2，Marschall P2，Ando K3，Shimura T3
(1. Radioactive Waste Management Funding and Rearch Center(RWMC)，Tokyo，Japan；
2. Cooperative for the Disposal of Radioactive Waste(Nagra)，Wettingen，Switzerland；
3. Obayashi Corporation，Tokyo，Japan)
Abstract：The Radioactive Waste Management Funding and Rearch Center(RWMC) in Japan has been conducting (rearch and development) for the disposal of radioactive waste since its establishment in 1976. More recently，RWMC has focud on important sociological and technical issues for geological disposal，such as monitoring，remote-handling and site investigation technology. Among the issues，the evaluation of gas migration from a repository is important for the developme
nt of more reliable safety asssments. This issue has been studied in a large scale in-situ experiment，the “gas migration test(GMT) in engineered barriers”at Nagra′s Griml test site(GTS) in Switzerland. GMT has focud on issues related to waste-generated gas migration through the engineered barriers in a silo-type disposal. A reduced scale
EBS(engineered barrier system) consisting of a concrete silo with a permeable gas vent and bentonite/sand buffer has been emplaced and instrumented in a specially constructed silo cavern at the GTS. The concrete silo was 2.5 m in height and
2.5 m in diameter and was located in an approximately 4 m diameter shaft excavated from the upper cavern. After
saturation of the EBS and water permeability tests，gas was injected into the silo during two main injection periods. After the first gas injection，repeat water permeability tests were performed to identify changes in EBS performance. The cond gas injection ud a cocktail of gas tracers which were sampled from multiple points in the EBS and geosphere. Following the end of gas injection，the EBS was carefully excavated and characterized using a range of techniques.
Key words：disposal of radioactive waste；transuranic(TRU) waste；in-situ test；engineered barrie
r system；gas migration；
two-pha flow modelling
颈部淋巴结肿大疼痛CLC number：TL 942+.2 Document code：A Article ID：1000–6915(2006)04–0781–07花的图片头像
GRIMSEL现场的⽓体渗透试验
Fujiwara A1，Okamoto S1，Tsuboya T1，V omvoris S2，Marschall P2，Ando K3，Shimura T3
(1. 放射性废弃物管理基⾦会与研究中⼼(RWMC)，⽇本东京；
2. 放射性废弃物处置公司(Nagra)，瑞⼠维蒂恩格昂；
3. Obayashi Corporation，⽇本东京)
摘要：⾃1976年以来，⽇本核废料管理基⾦会与研究中⼼(RWMC)⼀直在进⾏核废料处置的研究。最近，RWMC ⾮常关注核废料处置库的监测、遥控以及现场勘查技术等重要的社会公众与技术问题。其中，处置库溢出⽓体是
Received date：2005–08–30；Revid date：2005–12–20
Foundation item：Funding from the Ministry of Economy，Trade and Industry(METI) in Japan.
Corresponding author：Fujiwara A(1950–)，male，M. S. E.，obtained master's degree in 1975 at Department of Engineering，Kyoto University，Kyoto. He is prently a project manager of RWMC，and his main rearch activities cover the engineered barriers for radioactive waste disposal. E-mail：
782 岩⽯⼒学与⼯程学报 2006年⼀项重要的安全可靠性评估项⽬。这项名为“⼯程屏障的⽓体渗透试验”(GMT)的研究是在瑞⼠Nagra的Griml 现场实验室进⾏的。GMT集中研究在粮仓形的处置库中由核废料产⽣的⽓体在⼯程屏障中的渗透问题。在Griml 现场，⼀个按⽐例缩⼩的⼯程屏障库由⼀个具有透⽓孔的混凝⼟仓和砂–膨润⼟缓冲材料组成，缓冲材料被放置在⼀个专门建造的洞室中。混凝⼟仓直径为2.5 m，⾼为2.5 m，放置在直径为4 m的从上到下开挖⽽成的竖井中。
在⼯程屏障饱和及⽔渗透试验之后，⽓体分两个阶段被注⼊地窖中。⽓体第⼀次被注⼊之后，重复进⾏⽔渗透试验，以便观察⼯程屏障库的变化。第⼆次⽓体注⼊的时候，在⼯程屏障及岩⽯圈中采⽤⼀种可以多点取样的鸡尾酒⽓体跟踪器。在⽓体注⼊之后，⼩⼼挖掘出⼯程屏障库，然后⽤各种⽅法来进⾏鉴定，并采⽤实验室以及建模程序来进⾏现场的模拟研究。
关键词：放射性废物处置；超铀废物；原位试验；⼯程屏障体系；⽓体渗透；两相流模拟
1 INTRODUCTION
1.1 RWMC activities
The Radioactive Waste Management Funding and Rearch Center(RWMC) has been conducting broad rearch and development(R&D) for management of various radioactive wastes ranging from low level radioactive waste(LLW) and uranium bearing waste，to transuranic(TRU) waste and high level radioactive waste(HLW) since its establishment in 1976. More recently，RWMC is given a role to gather and provide information related to radioactive waste disposal，and pursue R&D on sociological and esntial technical subjects. Sociological subjects for HLW geological disposal are prervation of repository records and monitoring of repository. In addition，RWMC has been conducting R&D on esntial technical subjects，such as remote technology(remote emplacement of bentonite buffer，remote welding of waste containers) and site investigation technology(advanced geophysical exploration，site investigation flow diagrams). For TRU wastes disposal，barrier technologies are focud on. They are immobilization of I–129，waste packaging，long-term alteration of engineered barrier，behavior of C–14 and gas migration.
1.2 The gas migration test
Among the barrier technologies，the evaluation of gas migration in a repository is a quite important issue to develop more reliable safety asssment for a TRU waste repository. This issue has been studied in a large scale in-situ experiment called“Gas Migration Test(GMT) in Engineered Barriers”which is performed at Nagra's Griml Test Site(GTS) in Switzerland. GMT was initiated in 1997 and field testing was completed in December 2004. The project focus on issues related to waste-generated gas migration through the engineered barriers in a silo-type disposal concept. The objectives of GMT[1] are：
(1) Asss the function of the system EBS and adjacent geosphere as a whole with respect to migration of waste-generated gas.
(2) Evaluate models(conceptual and numerical) applicable to gas migration through barriers under realistic in-situ conditions.
(3) Provide data for further improvement of the EBS design with respect to gas migration(including the prence of a vent).
(4) Demonstrate the construction and the emplacement of an LLW-silo system under realistic in-situ conditions.
The major activities performed during the test were as follows：
(1) Site lection and excavation of the GMT access drift and silo cavern.
(2) Characterization of the geosphere at the GMT site.
(3) Emplacement and instrumentation of the EBS(concrete silo and bentonite/sand buffer—20% bentonite and 80% sand).
如何做好本职工作(4) Backfilling of the upper cavern and aling with a concrete plug.
(5) Saturation and pressurisation of the EBS and backfill.
(6) Gas injection into the silo(with and without tracers).
(7) Water permeability testing of the EBS before
第25卷第4期Fujiwara A，et al. Gas Migration Test at the Griml Test Site ? 783 ?
and after gas injection.
香蕉饼的做法大全(8) Depressurisation of the system and careful excavation and characterization of the EBS.
The overall experiment concept and the different components of the composite system(EBS/upper cavern and access
drift/geosphere) are shown in Fig.1.
Fig.1 Overall concept showing simplified layout and
消愁解闷components of composite system
2 CONSTRUCTION PROCEDURES
2.1 Site lection
During the planning of the experiment a t of requirements for a suitable site were developed. The included：
(1) Hydrogeology
The effective hydraulic conductivity of the bulk rock needed to be greater than 1×10－10 m/s and the
rock should be at least moderately fractured.
(2) Construction
Good access for transport of excavated rock and construction materials.
(3) Operation
No interference with current active tests.
Pre-screening of the different experimental areas at GTS identified the ventilation experiment(VE) area as being well-characterized and potentially meeting all the requirements. Existing data were collated；and scoping calculations were then performed. Finally，the suitability was confirmed by drilling two investigation boreholes at the lected site.
2.2 Excavation of the cavern and site characteri-
In June 1998，a new access drift was excavated from the VE tunnel. The drift axis was approximately perpendicular to the strike of a major shear zone that had been identified in the VE tunnel and veral investigation boreholes. It was planned that the shear zone would cut across the
upper cavern and silo cavern. The access drift was excavated by drill and blast with a nominal 2.8 m×2.8 m ction. After the drift intercted the shear zone，the upper cavern (approximate 7.5 m in diameter) was excavated.
Several boreholes were drilled and tested from the upper cavern prior to excavating the silo cavern to a depth of 4.5 m with a diameter of approximate 4 m，using a smooth wall blasting method. Some difficulties were encountered where the shear zone intercted the top of the silo cavern，but the were overcome and the excavation was completed in September 1998. Following excavation，further boreholes were drilled and tested and some tests were performed in the existing boreholes to identify changes due to excavation. A long-term pressure monitoring system was then installed in the boreholes around the GMT cavern.
2.3 Construction of the concrete silo and emplace-
ment of bentonite/sand buffer
The detailed layout of the EBS is shown in Fig.2 together with a photograph of the in-situ compaction of the buffer below the concrete silo. The 20/80 (bentonite/sand) buffer was emplaced as a ries of “lifts”which were compacted in-situ by using “stamper”machines(Fig.2). Lifts were typically 5–9 cm in
thickness after compaction. The EBS was divided into layers consisting of 3 to 9 lifts. The bentonite/sand layers were numbered 2–12 and monitoring instruments were emplaced at the top of each layer. The bentonite/sand in layers 8–10 was traced with lead-nitrate for a gas flowpath visualization experiment performed during the cond ries of gas injections.
Monitoring instruments were installed at the top of each bentonite/sand layer and included：piezometers，time domain reflectometers(TDRs) for the measurement
784 岩⽯⼒学与⼯程学报 2006年
(a)
(b)
十二星座双鱼座Fig.2 EBS layers and photograph of compaction of layers below concrete silos
sand and on the silo and rock walls and temperature nsors.
The reinforced concrete silo was cast in-situ after the compaction of Layer 4. The silo was constructed as an upper and lower half with a construction joint as shown in Fig.2. Instruments were installed on the top，sides and within the top and walls of the silo. Four TDR probes were located inside the concrete silo straddling the construction joint to monitor performance of the joint during the test. A permeable M1 mortar gas vent was t into the silo top.孤篇压全唐
After silo construction and instrumentation，Layers 5–12 of the bentonite/sand and their associated 2.4 Upper cavern backfill and plug construction
A sand layer(Layer 13) was emplaced above the top of the bentonite/sand. This layer contained a fiber-optic monitoring system and gas collectors. The upper cavern was then backfilled with gravel. A 2.3 m long concrete plug was constructed to al the GMT cavern from the access drift.
2.5 Data acquisition system(DAS)
The EBS instruments and nsors in the geosphere and backfill were monitored by the GMT data acquisition system(DAS) which provided near-real time access to the data via the World Wide Web. In all，there were some 96 piezometers，72 TDRs，10 temperature-nsors and 11 total pressure cells installed in the EBS and backfill. In addition，there were some 31 pressure monitoring intervals
and 4 TDRs in boreholes in the geosphere.
3 EXPERIMENTAL PROCEDURES
3.1 Saturation and pressurisation of the EBS
海伦凯勒的故事During the characterization of the site，it was established that the natural inflow to the GMT cavern was approximately 5 L/d. This was insufficient for saturation of the EBS and backfill within the experimental timescale. Artificial saturation was initially performed by water injection into gravel layer，Layer 1，at the ba of the EBS and then into the backfill in the upper cavern. Saturation started in August 2001，and from January 2002，the water injection into the backfill was maintained so that the upper cavern was pressurized to approximately 550 kPa(absolute).
During flooding and pressurisation of the backfill，it was found that piezometers and TDRs between levels 7 and 10 showed much faster respons than at other levels，indicating preferential saturation of Layers 8–10 around the vent and silo top. This was probably due to reduced swelling of the bentonite caud by the prence of lead nitrate. Saturation in other layers was slow due to the low permeability of the untraced 20/80 mixture. At the end of the saturation stage in October
2002，Layers 8–10 were
第25卷第4期 Fujiwara A ，et al. Gas Migration Test at the Griml Test Site ? 785 ?
was only partially saturated [2]. As it was expected that the gas flow would be focud in the highly saturated layers(becau of the location of the gas vent on the silo top)，it was decided to end the saturation stage and perform the pre-gas water tests. In fact ，saturation of the low permeability untraced layers continued from August 2001 to depressurisation of the EBS in April 2004.
3.2 Pre-gas water test(WT1)
Prior to gas injection ，a ries of pumping tests were performed to determine hydraulic properties of the EBS. The tests included injection/withdrawal tests from the concrete silo and backfill. 3.3 Gas injection pha 1
Nitrogen was injected into the silo at increasing mass flowrates from 0.025 to 5.000 mg/s starting in January 2003(e Fig.3). The injections were designed to target different process ：gas dissolution and transport in the water pha ，two-pha flow ，interface opening and pathway dilation. The first TDR respons to gas injection were en in TDRs on the silo sides and later within the bentonite/sand above the silo(e Fig.3). During the higher rate gas injections(1 mg/s and above)，TDR respons became focud in Layer 8(clo to the silo top) at the north side of the silo.
Pressure breakdowns were en at the start of the 0.05，1.00 and 5.00 mg/s injections ，respectively. The breakdowns corresponded to periods of low effective stress on the silo top. During the high rate gas injections ，although the backfill pressure ro strongly indicating an influx of gas to the upper cavern ，effective stress stayed roughly constant throughout the injection as shown in Fig.4. In total ，approximately 1.2 m 3 of gas(at backfill pressure) was injected.
After shut-in of the gas injection ，free gas was removed from the silo to determine the volume of water displaced during gas injection. The best estimation of the produced gas volume was about 13 litres(at 650 kPa). This reprented approximately 1% of the total gas injected. After the gas/water exchange ，the pressure in the backfill was then reduced from about 650 kPa to 580 kPa so that conditions would be similar to tho prior to WT1. 3.4 Post-gas water test(WT2)
Repeat water testing after the reduction in EBS pressure showed the influence of residual gas in the bentonite/sand and an incread resistance to flow in/out of the silo. Analysis with two-pha flow models suggested that there had been a reduction in
Fig.3 Measured volumetric water content ，effective tress ，and pressure during the two-pha gas injection
P r e s s u r e (a b s o l u t e )/k P a
E f f e c t i v e s t r e s s /k P a
V o l u m e t r i c w a t e r c o n t e n t /%
(a) volumetric water content
(b) effective stress
(c) pressure
786 岩⽯⼒学与⼯程学报 2006年Fig.4 Effective stress on silo top and change in pressure in silo and backfill during 5.00 mg/s gas injection
intrinsic permeability between the two water tests.
3.5 Gas tracer injections
A cond ries of gas injections using nitrogen and five different gas tracers(2% volume of krypton，neon，sulphur hexafluoride and isobutane，respectively) were performed between Januar
y and April 2004. Tracer gas were detected in the EBS，backfill and geosphere as shown in Fig.5. The concentrations of the different gas are affected by the different solubilities and diffusivities. Fig.5 shows the measured concentrations from four of the samplers for sulphur hexafluoride (high molecular weight and low solubility) and krypton(high molecular weight and high solubility). In total，approximate 0.8 m3 of gas(at backfill pressure) was injected.
At the end of gas injection，the volume of gas within the silo was again measured. The measured volume was between 10 and 11 litres，again about 1%
Fig.5 Gas tracer sampling locations in the EBS and geosphere(tracer recovery plots for two of the five gas tracers：sulphur hexafluoride and krypton)
E
f
f
e
c
t
i
v
e
s
t
r
e
s
s
/
