 |
| Titel |
Visualisation and Quantification of Transport in Barrier Rocks with Positron Emission Tomography |
| VerfasserIn |
J. Kulenkampff, C. Gajewski, M. Gründig, J. Lippmann-Pipke, H. Mittmann, Michael Richter, M. Wolf |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250022866
|
|
|
|
|
|
| Zusammenfassung |
| In tight barrier rocks laboratory observation of radionuclide transport and determination of
transport parameters is a demanding and interminable task, because of slow rates, small
concentrations, and intricate chemical interactions. The validity of results from
common laboratory methods, like flow- and diffusion experiments on small samples, is
limited by the heterogeneity of the pathways and adherent upscaling issues, because
homogeneous conditions have to be presumed for these input-output investigations. But
nano-pores or micro-fractures could be present, which would provide pathways for
heterogeneous transport processes. Transport properties of these pathways are most
influential boundary conditions for reactions between fluid components and crystal
surfaces.
We propose Positron Emission Tomography (GEO-PET) as an appropriate method for
direct observation of heterogeneous transport of radiotracers in tight material on the
laboratory scale. With high-resolution PET scanners, which are common instruments of
biomedical research (“small animal PET”), it is possible to determine the spatio-temporal
distribution of the tracer activity with a resolution of almost 1 mm during about three periods
of the tracer half-life (half-lives of some applicable PET tracers: 18F: 1.8 h, 124I: 4.2 days,
58Co: 70.8 days). The PET tracer is applied as ion in solution or as marker for compounds,
like colloids.
The most considerable difference between PET applications on geomaterial
compared to biological tissue is the stronger attenuation and scattering of radiation
because of the higher density of rock material. After travelling the positron attenuation
length in dense material (about 1 mm), the positron annihilates in contact with
an electron, transmitting two photons with 511 keV, propagating in antiparallel
direction. The sample size of geomaterial is limited by the attenuation length of
these photons. By applying an appropriate attenuation correction it is possible to
investigate transport processes in rock cores with diameters up to 10 cm. Then at
least 20% of the initial annihilation events are recorded as coincidences. However,
one single photon of the annihilation radiation may be recorded while the other
is absorbed; therefore, the signal to noise ratio is degraded by attenuation. Other
sources of noise are scattered events, and the loss of one coinciding photon due
to gaps between the detectors and other detection probability reasons. Also, the
ratio of random coincidences increases with the noise level and impairs the image
quality of the tomographic reconstruction. The reduction of these reconstruction
artefacts by enhanced data correction methods is an important requirement for the
development of the GEO-PET method. An other problem is the development of
special methods for the quantitative evaluation of the extensive spatio-temporal data
sets.
We present results from high-resolution PET for tomographic process observation during
transport of colloids and conservative tracers in macroscopic samples of clays, saline rocks,
and granites (diameter 5 to 10 cm, length 5 to 20 cm). In most cases we observed localized
zones of transport, even in a homogenized compressed clay sample. This reflects
the non-representative sample volume, which probably is not achievable for any
laboratory method. However, at least the PET tomograms reveal these deviations from
representativeness. Up to now, break-through-curve parameters can be determined from
spatially resolved tracer concentration measurements at distinct regions of the sample,
without mandatory penetration of the complete sample extension. A multiscale model-based
inversion scheme for continuous scale-dependent parameter determination is currently
developed. |
|
|
|
|
|
|