SOME INITIAL RESULTS RECEIVED FROM THE DIKE MONITORING SYSTEM AT THE RED RIVER DIKE SECTION IN THÁI BÌNH PROVINCE

TRẦN CÁNH¹, ANDREAS WELLER², RONALD LEWIS³

¹Institute of Geological Sciences, VAST, 84, Chùa Láng Str., Đống Đa, Hà Nội.
²Institute of Geophysics, Clausthal Technic University, Germany.
³ Planungsgesellschaft Scholz + Lewis mbH, Dresden, Germany.

Abstract: This article displays an experimental dike monitoring system which installed at a major point weak dike section of Red River left dike in Thái Bình Province. The system consisting of geophysical and geotechnical equipments includes: 1. Multi-electrode induced polarization (IP) instrument for observing structural varieties of underground and resistivity changes caused by the variation of water content; 2. The tensiometer TS for observing water pressure and temperature in the dike body; 3. The electromagnetic sensor system (FDR-Sensors) for determining the water content at different depths in the dike body. This system has been operating since April 2006 and the interesting information above the variations of geophysical parameters were received in the flooding season of 2006-2007 and 2008-2009 years. The authors will present some initial results received from this dike monitoring system.

I. INTRODUCTION

The dike system in the Red River Delta includes more then 4000 km of river dikes, that have an important role for protecting against flood and for socio-economic development in North Việt Nam. This system has been constructed over 200 years ago and consolidated through different generations. The satisfied guarantee of this dike system in annual flood is mission in the strategy of socio-economic development of the country.

The river dike lines lie over different topographical areas and different geological units containing many hidden threats for the dike body and its fundaments, that can cause different damage events in the flood season.

According to the annual reports of the National Dike Management Organ (2003) there are more than 200 km of main point dikes according to more than hundred weak dike places in the dike system in the Red River Delta, where one must prepare the intended projects with materials and power for protecting again flood [6].

Vietnamese and German geophysical researchers have been interesting the question on safety of dike system in both countries since 1995, the applied results of geophysical methods for river dike can be found out in the works [1-5, 10-12].

A new research direction has been drawn from the strong flood in 2002 year in the Germany and Việt Nam, when some river dike sections were broken. This demands the creation of a dike monitoring system. This article presents some initial results received from this work.

II. DIKE MONITORING SYSTEM

1. General principles of dike monitoring system

In the flood season, the water saturation in dike-component parts strongly increases, which makes mechanic-physical character of dike materials in weak section not to attain the criteria for dike safety. Data on variety of geophysical field and geologic-geotechnical parameters in “weak foundation soil zone” is valuable information for calculating a three-dimensional model to estimate dike states (up-to standard or in possibly broken unstability). The tension of river water on the dike body and foundation can be directly measured by tensiometer. Indicators of water saturation increasing in direct ratio with the increase of dike-forming material conductivity are measured by a multi-electrode resistivity tomography data acquisition system with a series of specially made electrode-sensors touching ground into dike body. Measurement results will be standardized in comparison with tension data by tensiometers.

The acquired data may be used to evaluate the state of the dike body during a high water event. The resulting models are useful to predict stability criteria and to warn disasters for dike.

Finally, a monitoring system which can be operated automatically can be integrated in a complex warning system. The remote system provides relevant data concerning the state of the dike. The studied dike section and locations installed with measuring instruments is showed in Figs. 1a, b and c.

Figure 1a. Experimental dike section and the locations the monitoring equipments installed at km 169+150 m, left Red River dike in Ngô Xá, Thái Bình Province.

Notes: 1. Electrical cables and multi-electrode array installed in the dike cross section (depth 0.5 m from dike surface) with 50 electrodes and 1 m spacing; 2. Locations where is installed steel pipe for measuring soil density; 3. Locations where are installed temperature sensors (T) and frequency domain reflectometry sensors (FDR); H. Place of Data logger box for T and FDR; T. Station for preserving equipments and periodically measuring IP.

Figure 1b. In the place where river water level is measured daily.

Figure 1c. In the place where groundwater level at the dike foot is measured daily.

2. Used technologies and instruments

The new characters of the monitoring system are that, the measuring instruments have high accuracy and little using power.

The dike monitoring system consists of following technologies:

1. Multi-electrode induced polarization instrument (IP) for observing structural varieties of underground and resistivity changes caused by the variation of water content. The technical parameters of equipment can be seen in [7].

2. Tensiometer (T) for observing the water pressure and temperature in the dike body [8].

3. Electromagnetic sensor system based on the principle of Domain Frequence Reflexometry sensors (FDR-Sensors) for determining the water content at different depth in the dike body [6].

The physical principle and technical parameters of T sensors and FDR sensors can be found in [6, 7]. The Figure 2 indicates two FDR-sensors and T-sensors, which have been installed at weak dike section in Thái Bình Province.

4. Two places with equipments for observing river water level and groundwater at the land-side at the dike foot (Fig. 1b, c).

The IP measurement is carried out by the software in notebook (Fig. 3a). The T and FDR-sensor systems work automatically and data stored in memory of data logger.

a

b

c

d

Figure 2. The electro-magnetic sensor systems and tensiometer T 8.

Notes: a. FDR 6 sensors system installed in dike body at land side; b. FDR sensor; c. FDR 6 sensors system installed in dike body at river side; d. Tensiometer for measuring temperature in dike body.

a b c

Figure 3. Reading data from monitoring instruments at dike in Thái Bình.

Notes: a. Multi-electrode IP measurement for dike cross section; b. Receiving data
from T and FDR sensor systems; c. Reading data of groundwater levels at the dike foot.

III. INITIAL RESULTS

1. Results of geoelectric tomography survey

The results of geoelectric tomography survey reflect temporal changes of the water saturation of different dike material layers in the dike cross-section. The resistivity images measured at different times also indicate the change in dike body structure. In flood times, if the measures will be carried out daily or continually, it will receive disaster changes in the dike structure.

The measurement is carried out monthly /weekly/daily by the equipment 4 P Hp (made in Germany) with configuration: Half-Wenner forward and backward.

The received data were processed by the software DC2DSIRT that reduce the effect of dike topography.

Figure 4 shows the resistivity images in dike cross section received in the months July, September and November, 2008. In these months the water level is high in river, reaching from 2.60 to 3.97 m that is the water levels of alarm grad II and over grad II in this dike section (alarm grad I is +2.48 m, II = + 3.12 m, III = + 4.13 m).

The resistivity images show that the dike structure consists of clay with the resistivity value of 5 ÷ ~ 80 Wm. The sand layer lies in the depth of +1.0 to -3.5 m and the resistivity value of 15 ÷ 20 Wm having high permeability in flood time.

The resistivity distribution in Figure 4, right side, reflects little change in the dike body structure. In the felt side of Figure 4, we can see difference between normal resistivity distribution (measured in January, 2008) and measured resistivity distribution in percent. The change of resistivity distribution in dike structure varies over ± 20%.

The geoelectrical data contain image resistivity part, which is related to soil porosity. This will be analyzed later.

2. Results of measurement by tensiometer T8

Figure 5a shows the change of water pressure in dike soil and of river water levels and groundwater levels from 01/7/2007 to 11/9/2007. There are two flood times showed in Figure 5a and in Table 1. The Figure 5a shows that the change of river and groundwater is influenced by tide effects and water pressure indicated by T8a and T8b has positive value in flood time from 29/7/2007 to 11/8/2007 with the flood alarm of grad II. In other times the water pressure has negative value.

Figure 5b reflects the change of temperature in river side (T8a) and in land side (T8b) of dike body from 24/3/2006 to 25/10/2007. The temperature graphics increases to maximum (27.5^o C) in October 2006, after that it decreases to minimum (24.8^o C) in April 2007.

The data measured by T8a and T8b from 07/12/2008 to 23/7/2008 have similar results.

3. Results of measurement by FDR sensor system

Figure 6a presents monitoring results measured by FDR 8 sensors system in water side and graphics of river water and groundwater levels. The extremes in water level graphics correspond to flooding days of 2008 showed in Table 1.

The curves of FDR 8 sensors reflect the change of dike soil permeability order water saturation in different depths. Strong changes belong to the soil layer in the depth of 0.95 m and 1.55 m. During high water time the water saturation varies from 0.045 to 0.063 S/m.

Figure 6b shows progresses of soil permeability in land side measured by FDR 6 sensors. In land side the soil layers have permeability smaller than in river side, these values change from 0.025 to 2.42 S/m.

The comparison of permeability variations at two dike sides is showed in Figure 6c. Data received in the same depth (1.05 and 1.07 m), soil permeability at river side has value greater than that at land side (at river side: 0.063 S/m; at land side: 0.037 S/m).

In brief, the FDR sensor systems allow to measure automatically the change of dike material permeability. These data are useful for dike management.

Figure 4. Resistivity images of dike cross section measured from 08/7/2009
to 05/11/2009 and these changes compared with normal resistivity section in %.

Figure 5a. Results of measurement by tensiometer T 8 system
from 01/7/2007 to 11/9/2007.

Figure 5b. Temperature changes of dike body from 24/3/2006 to 25/10/2007.

Figure 6a. Permeability variations of dike soil measured by FDR 8 sensors at river side and the changes of river and ground water levels from 22/02/2008 to 15/4/2009.

Figure 6b. Permeability variations of dike soil measured by FDR 6 sensors at land side and the changes of river water and groundwater levels from 22/02/2008 to 15/4/2009.

Figure 6c. Comparison of permeability variations at two dike sides.

Table 1. Flood date of 2008 year

No	Date	Month, year	River water level (m)
1	17 - 26	7/2007	2.43 - 2.48
2	30/7 - 9/8	7 & 8/2007	2.83 - 3.04
1	7, 8, 9, 10, 11	7/2008	2.60 - 3.00
2	26, 27, 28, 29	7/2008	2.80 - 3.34
3	10,11,12,13,14,15	8/2008	2.50 - 3.70
4	25, 26, 27, 28, 29, 30	9/2008	1.80 - 2.50
5	1, 2, ...., 10, 11	11/2008	2.50 - 3.97

IV. CONCLUSIONS

For the first time a dike monitoring system is installed at main points of weak dike section of the Red River left dike in Thái Bình Province. In nearly 3 experimental years, this system has worked continuously under tropical conditions (strong rain, high humidity, high temperature…). The system has been repaired some time and the instruments have been improved corresponding with Vietnamese conditions.

The analysis of monitoring data in flood seasons 2006-2009 years has been carried out in first step. The above initial results indicate that the monitoring system with new technologies and instruments can provide with changes of dike body in flood season.

The combined analysis of data and modelling in order to evaluate dike safety and to warn disaster events are tasks in the coming time.

Acknowledgements: The joint project “Development of a dike monitoring system” is sponsored by BMBF, Germany, VNM 05/001 and by MOST, Việt Nam, 42/823/2007/HD-NDT. The authors would like to thank the leaders and managers of national and local organs for useful assistance in the realization of this Project.

REFERENCES

1. Canh Tr., A. Weller, 1995. Co-project report “Application of geophysical methods to study on dike in Việt Nam”. Sponsored by Volkswagen-Stiftung, Germany, Archives of DDG GmbH Leipzig, Gemany and Inst. of Geol. Sci., Hà Nội.

2. Canh Tr., A. Weller, 2003. Final report of Co-project “Development of a geo- electrical technology to find termite nests for protecting dikes”. Sponsored by Volkswagen-Stiftung, Germany. Archives of Inst. of Geophysics, TU Clausthal, Germany and Inst. of Geol. Sci., Hà Nội.

3. Cánh Tr., Tuyên D.V., Weller A., Preusse S., Vũ N.T., 2003. Application study of multi-electrodes geoelectrical technology for finding hidden defects at dike in Thái Bình. Proc. of Vietnamese-German Worksh. on Dike Monitoring, p. 47-53. Hà Nội.

4. Cánh Tr., Ph.K. Vân et al, 2007. The dike monitoring system as a tool to protect actively against floods. Proc. of 5th Symp. on geophysical sciences and technics of Việt Nam, Hồ Chí Minh City.

5. Hennig Th., Weller A., Tr. Cánh, 2005. The effect of dike geometry on different resistivity configuration. J. of Applied Geophysics, 57 : 278-292.

6. Report on assessment of quality of dike and sluice buildings before storm: flood season year 2003. Archives of National Dike Management Office. Hà Nội.

7. Sentek technologies, 2006. EnvirScanPlus, continuous moisture monitoring solutions. www.sentek.com.au.

8. Tensiometer for direct measurement of soil water tension, 2006. UMS GmbH, Grmunder Str. 37, Muenchen, Germany.

9. Tuyen D.V., Canh T., Weller A., 2000. Geophysical investigations on river dike in Việt Nam. Euro. J. of Env. and Eng. Geophysics. 4/3 : 195-206.

10. Weller A., Hennig Th., Canh T., 2003. Topographic correction for dike geometryusing averaged three-electrode configurations. Proc. of Vietnamese-German Worksh. on Dike Monitoring, p. 31-36. Hà Nội.

11. Weller A., Tr. Canh, Ng.Tr. Vu, 2006. Multi-electrode measurements at Thái Bình dike (Việt Nam). Near surface Geophysics, 4/2.

12. 4 Point light 10 w. Technical data, operating instruction. Version 4.19, 2008, LGM Lippman. Germany, www.l-gm.de