Groundwater Geophysics in Burundi to support the Technical Cooperation Project
Report of the project:
Within the framework of the project geophysical surveys were conducted using Transient ElectroMagnetics (TEM), Magnetic Resonance Soundings (MRS), Electrical Resistivity Tomography (ERT), and geophysical borehole logging. Aims were to
- investigate the spatial extent and connection of aquifers,
- identify water strikes and aquiferous strata within the boreholes, and
- recommend sites for water abstraction and monitoring boreholes.
Source: Department of Field Support, Cartographic Section, Map No. 4048 Rev. 8 UNITED NATIONS May 2015
The measurements were carried out in all three project areas Gitega, Kirundo and Rumonge.
Gitega
In Gitega the overabstraction and possible contamination of the aquifers is feared. The survey was restricted to the area Nyanzare (abstraction boreholes F7.xxx, > 50% of the groundwater production for Gitega). There groundwater was expected to be drawn from weathered quartzite and schist connected to fractures. The investigations detected abundant ferricrete layers at surface and the typical high resistivities of these layers revealed thicknesses of several metres.
Underneath the resistivities dropped to some tens Ωm and the gamma log detected thin (some metres) layers with radiation differences (150 – > 400 gAPI) at two places (GI-PZ01 and GI-PZ03, the measurement at GI-PZ03 was hampered by the borehole’s construction). The low resistivities point to weathered schist as aquiferous strata and not to quartzite. The lineaments detected by satellite images are related to fractures within the schist (profiles 1 and 2) or separate possibly schist from quartzite as indicated by the sharp resistivity contrast at profile 3.
Only very slight hints for high resistive bedrock were detected and, therefore, the depth and nature of the bedrock remains unclear. The hydrogeological analysis showed short residence times and recent recharge of the waters drawn by the abstraction boreholes F7.xxx. Local recharge is likely through the detected ferricrete layers at surface (Taylor et al., 2000) and, therefore, measures against local contamination are essential.
Rumonge
In Rumonge the elevated iron contents of some borehole waters hamper the development of a groundwater based water supply scheme for the town. The currently installed abstraction boreholes and the newly drilled piezometers draw their waters from different aquifers. One extended aquifer system comprises the alluvial layers of the Murembwe river resembling a delta and the upstream catchment. The first electromagnetic profile revealed no separating structures within this aquifer system but a discontinuous deepening of the alluvium toward the lake with abrupt deepening at a lineament mapped by satellite images.
The productive aquifer zones are found in bedrock fractures (RU-PZ01), in weathered saprolite/regolith above the bedrock (RU-PZ05) or the alluvial layers (RU-PZ04, Hotel David, Savonor). The gamma logs of the piezometers showed a sharp increase of radiation at different depths interpreted as weathering front and water is found either side, in the weathered zone and below in fractures. Water level monitoring and hydrochemistry revealed recharge from precipitation between RU-PZ04 and RU-PZ05 diluting the mineralised groundwater. The waters drawn from boreholes close to Lake Tanganyika contain very high amounts of iron (>10mg/l) and those toward the mountains do not.
The source of the iron still remains unclear but it is contemplated whether it is related to lake sediments. The second electromagnetic profile runs NW/SE parallel to the graben shoulder and connects piezometers RU-PZ03, RU-PZ01 and RU-PZ02. The profile starts close to a well hitting granite at about 22m, then it crosses an alluvial filled deltaic structure, then again granite is detected and the profile ends close to piezometer RU-PZ02 with schist at depth and the low transmissivity of the piezometer reveals schist not to be a promising aquifer except in fractures.
Kirundo
In Kirundo investigations aimed at locating suitable aquifers to provide safe water to the communities. The ERT profiles crossed lineaments mapped by satellite images in order to detect aquifers possibly related to these lineaments. The profiles revealed high resistive ferricrete crusts at some places but only one profile shows this crust all along the surface. At depth where full saturation is expected the resistivities are quite low, at two profiles below 150 Ωm. The ERT profiles detected no resistivity increases at depth possibly related to granitic bedrock with certainty. Two TEM profiles were measured along and crossing one valley (crossed by ERT profile 2) where five new piezometers were drilled and tested for hydrochemistry and transmissivity.
The results discovered an extended aquifer within the valley and the depth to the aquifer increases with increasing elevation as expected. The aquifer comprises weathered bedrock but high resistive granite at the aquifer’s bottom is nowhere detected. The investigation depth of the electromagnetic survey was about 100-150m and from the drillers’ reports at these depths unweathered granite is seen. It appears unlikely that unweathered granite shows low resistivities. The high uranium content of some groundwaters and the low resistivity at depth might indicate a deeply weathered granitic zone possibly related to the mapped lineament.
Report