A Case Study To Detect the Leakage of Underground Pressureless

A comprehensive method combining Ground Penetrating. Rader (GPR), electric potential survey and geochemical survey is introduced in the leakage detect...
0 downloads 0 Views 179KB Size
Environ. Sci. Technol. 2002, 36, 1077-1085

A Case Study To Detect the Leakage of Underground Pressureless Cement Sewage Water Pipe Using GPR, Electrical, and Chemical Data G U A N Q U N L I U , * ,† Y O N G G A N G J I A , † HONGJUN LIU,† HANXUE QIU,† DONGLING QIU,‡ AND HONGXIAN SHAN† Department of Environmental Engineering, College of Geoscience, and College of Chemistry and Chemical Engineering, Ocean University of Qingdao, Qingdao 266003, Shandong, P. R. China

The exploration and determination of leakage of underground pressureless nonmetallic pipes is difficult to deal with. A comprehensive method combining Ground Penetrating Rader (GPR), electric potential survey and geochemical survey is introduced in the leakage detection of an underground pressureless nonmetallic sewage pipe in this paper. Theoretically, in the influencing scope of a leakage spot, the obvious changes of the electromagnetic properties and the physical-chemical properties of the underground media will be reflected as anomalies in GPR and electrical survey plots. The advantages of GPR and electrical survey are fast and accurate in detection of anomaly scope. In-situ analysis of the geophysical surveys can guide the geochemical survey. Then water and soil sampling and analyzing can be the evidence for judging the anomaly is caused by pipe leakage or not. On the basis of previous tests and practical surveys, the GPR waveforms, electric potential curves, contour maps, and chemical survey results are all classified into three types according to the extent or indexes of anomalies in order to find out the leakage spots. When three survey methods all show their anomalies as type I in an anomalous spot, this spot is suspected as the most possible leakage location. Otherwise, it will be down grade suspected point. The suspect leakage spots should be confirmed by referring the site conditions because some anomalies are caused other factors. The excavation afterward proved that the method for determining the suspected location by anomaly type is effective and economic. Comprehensive method of GRP, electric potential survey, and geochemical survey is one of the effective methods in the leakage detection of underground nonmetallic pressureless pipe with its advantages of being fast and accurate.

Introduction Pressureless cement pipes mostly belong to drainage pipes, which transport industrial and life sewage water with large * Corresponding author phone: ++86-532-203-2571; fax: ++86-532-203-2102; e-mail: [email protected]. † Department of Environmental Engineering, College of Geoscience. ‡ College of Chemistry and Chemical Engineering. 10.1021/es001954s CCC: $22.00 Published on Web 01/26/2002

 2002 American Chemical Society

amounts of poisonous materials. The leakage of such pipes will contaminate soil and groundwater and affect the health of local residents. Presently, there have been many kinds of instruments developed to detect underground pipes. These instruments include the 801 Metallic Detector made in the U.S.A., the PL-801GXII pipe detector produced by Fuji, a Japanese company (1), the FM9800 series pipe detector of Metrotech company (2), and the RD400 series pipe detector of Radiodetection, Ltd. in the U.K., and a variety of magnetometers. These instruments detect metallic pipes effectively but not nonmetallic pipes. There is also a meter designed to listen to leakage and a noise correlation instrument to investigate the leakage of high-pressure tap-water pipes. GPR surveys have been used to do this previously (3). But it is very difficult to find the leakage of pressureless nonmetallic pipes, and there is no well-developed method for this problem. Therefore, observation by excavation is always used to map the leakage effectively, but it consumes great manpower and material resources and costs a large amount of money. In America, Al-Saigh et al. (1995) applied the natural electric potential method to find the leakage of a reservoir dam (4). Since the amount of leakage was very large, the anomaly was very evident. The method to detect the small amounts of seepage has not yet been reported. Geochemical prospecting is an economical technique to find mineral mines, because the anomalies caused by large mines are eminent. It is possible to use geochemical surveys to investigate the leakage of pipes, but pure geochemical surveys may need a large amount of work and money and is not too reliable. In recent years, ground penetrating radar (GPR) has developed rapidly and is being applied to the investigation of geology, engineering, then environment, agriculture, and archaeology, etc. (5-8). Liu et al. (1999) employed GPR to probe the range of a mined-out area (9). Tadensz and Ulrych (1994) combined GPR and the geoelectric method to map the underground pollution range caused by industry at Bahia Petroleum and a Chemical Center in Brazil (10). Benson and Mustoe (1998) combined GPR, geoelectric, VLF, and geochemical data to map groundwater pollution in northeastern Arizona (3). Benson and Mustoe (1996) integrated GPR, geoelectrical, and geochemical data to outline leakage from underground storage tanks in north central Arizona (11). In this paper, GPR, electric, and geochemical surveys are combined to find leakage from a pressureless cement pipe. A practical example proves that the combined GPR, electric, and geochemical surveys is a useful scheme for detection of leakage from a pressureless cement sewage water pipe.

General Conditions of Experimental Site and Water Quality Inside the Pipe The experimental site is located at the margin of a mountain front alluvial and flood plain and the Yellow River alluvial plain. The elevation of the south area is higher than that of the north and the area dips to the northeast. The pipe is 27.5 m above sea level in the south part of the study site and 3.7 m in the northeast part. The terrain is gentle, and the vegetation was wheat when the work was taken. The location of the study site is shown in Figure 1. The pipe is mostly buried in silty clay and silty sand and locally in fine sand or soil and sand. The depth of the water table varies from 23.5 to 5.0 m from south to north, respectively, in the dry season. It is slightly shallower during times of high flow. VOL. 36, NO. 5, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1077

FIGURE 1. The location of the study site. The length of the pipe to be investigated is about 30 km, the diameter is 1.5-2.0 m. The thickness of the overlying soil is generally 2-5 m, with the thickest layer between 5 and 7 m. Along the pipe line, there is a check well every 80-120 m. The characteristic components in the sewage water include petroleum, ammonia, nitrogen, volatile phenol, cyanide, benzene, COD, chloride, and total hardness. The contents of petroleum is 2-10 mg/L, ammonia 10-100 mg/ L, volatile phenol 0.1-2.0 mg/L, cyanide 0.021-0.20 mg/L, benzene 0-0.2 mg/L, COD 30-150 mg/L. Chloride was 1-8 g/L during 1989-1993, and chloride is 1-1.4 g/L after 1993, and total hardness 8-16 mol/L.

Prospecting Principle The leakage of wastewater from a pipe causes compositional differences between the immediate media surrounding the pipe and the background media. The leakage flow takes cations from the solution and increases their concentration in the direction of flow, while anions are detained in the opposite direction. Thus, an electrofiltration field is formed by the water flow. As a result of the differences of the concentration of groundwater and leakage water, the diffusion speeds of cations and anions are different, creating a diffusion electric field. These changes produce a contrast in the leakage zone’s electromagnetic, electrical, and chemical properties compared to those of the background media. These differences can be detected by GPR, electric, and geochemical methods. The effects of the natural electric fields produced by the electrofiltration field and the diffusion electric field can be measured at the surface. This information can be used to determine the range of leakage. Although the natural electric potential and GPR methods are each capable of detecting leakage from pipes, they each have advantages and disadvantages. The natural potential method is usually affected by irrigation because the potentials caused by the leakage of pipes and irrigation are superimposed. GPR mainly shows contrasts in the electromagnetic properties of the media, and it is sometimes difficult to distinguish the anomalies on radar profiles produced by geological differences and leakage. Therefore, it is better to use these two detection methods cooperatively to improve the accuracy of prospecting. 1078

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 5, 2002

FIGURE 2. The location of GPR, electrical survey lines, and sampling. GPR and electrical surveys are applied to help determine the range of leakage. Samples are then taken from the anomalous area and chemically analyzed to determine the type of leakage. This scheme can be used to investigate the subsurface quickly, save money, and yield relatively reliable results.

Arrangement of Survey Lines Figure 2 shows the location of GPR, electrical survey lines, and sampling. Only the survey lines discussed in this paper are shown. The measurement sections were selected by local residents, local Environmental Bureau, and authors of this paper. GPR Survey. The radar system employed in this investigation was a Pulse EKKO-100 produced by Canadian Sensor

and Software Co. Radar profiles were selectively obtained along the pipe. For each prospecting section, three profiles were designed. One was over the pipe, and the other two were located at the right and left sides 3 m away from the pipe. After many tests, the conventional common offset method was employed (5). The central frequency of the antennae was 100 MHz for the area where the top of the pipe was shallower than 3 m and 50 MHz otherwise. The separations between 50 and 100 MHz antennae were 1.0 and 2.0 m, respectively, when measurements were taken along each survey line, and the measurement interval was 1.0 m. Electrical Survey. Electrical surveys were conducted using a high accuracy, high sensitivity excitation potential JS-3A instrument with nonpolarization electrodes. Before the survey was started, tests were carried out to determine the measurement interval. For 1 and 2 m intervals, the amount of field work was too large. For 8 m interval, the anomalies maybe cannot be detected. Therefore, the interval of 4 m was selected. For each survey section along the pipe one measurement line each on the right and left sides were arranged. Area electrical surveys were performed for the possible leaking positions to define the properties and ranges of anomalies. The line and point intervals of the area measurements were 10 and 4 m, respectively. The number of lines of area surveys at each site was 3-5, according to primary results of surveys. If electric potential values were positive along one line, an additional measurement line outside the line was not necessary. Otherwise, one additional line was arranged. The natural electric potential method measured electric potential. One electrode was fixed at a position far from the pipe where the electric field was stable and ground connection condition was ideal. Another electrode was moved along the survey line. The electric potential between the two electrodes was measured point by point. The interval was 4 m. Sampling Water and Soil Samples. Water and soil samples were taken at each possible leaking section detected by GPR or electrical methods. Chemical examination and analyses were carried out for the samples, and the characteristics of the anomalies were classified. Chemical examination and analyses were done by College of Chemistry and Chemical Engineering, Ocean University of Qingdao. Generally, a well 3 m away from the pipe line was bored at the center of an anomaly. Occasionally, an additional sample was taken in a well 5 or 10 m away from the pipe line. One reference sample was also collected 1 km far from the pipe line (see Figure 2). Soil and water samples were generally obtained at positions above, at, and below the pipe line since the water and soil background might vary greatly in such a large pipe line region. Soil samples were always collected, but water samples were obtained only where water was encountered at the sampling positions. The soil samples collected above the pipe line were selected as reference samples.

Interpretation and Results It was a complicated problem to judge whether the pipe leaked or not. All the geophysical and chemical exploration data as well as the geological conditions and possible sources of pollution needed to be considered and integrated. Interpretation of Geophysical Data. 1. Interpretation of GPR profiles. Based on the waveform, continuity of events, reflection properties, the width of waveform, and the size of the anomalous range, three types of GPR reflections were classified. Type I: The waveforms are disorderly, reflections are very strong, the width of the waveform is large, the continuity of events is poor, and the range of an anomaly is large (Figure 3).

FIGURE 3. The radar profile of section 40 of type I.

FIGURE 4. The radar profile of section 32 of type II. Type II: The waveforms are less disorderly, reflections are strong, the continuity of events is not good, and the range of an anomaly is small (Figure 4). Type III: Waveforms are simple, the continuity of events is good, and the characteristics of reflections are consistent (Figure 5). 2. Interpretation of electric potential curves. Potential curves are divided into three types according to the shape of the curve, negative magnitudes of anomalies, and the width of anomalies. Type I: The curve is “U”-shaped, the negative magnitude of the anomaly is less than -4.0 mv, the difference between the background value and negative magnitude is greater than 8 mv, and the width of the anomaly is wider than 30 m (Figure 6). The contours of the natural electric potential form a closed ellipse distributed along the direction of the pipe. The anomaly is negative, and the smallest value of the anomaly is less than -25 mv (Figure 7). Type II: The curve is V-shape shown in Figure 8, the magnitude of the anomaly is less than -2.0 mv, the difference between the background value and negative magnitude is greater than 5 mv, the width of the anomaly is larger than 20 m. Type III: The curve is gentle, varies slightly at local sections, and the potential is positive (the range of 12 50013 000 m in Figure 6). VOL. 36, NO. 5, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1079

FIGURE 8. Natural electric potential curve of section 32 of type II.

FIGURE 5. The radar profile of section 20 of type III.

FIGURE 6. Natural electric potential curve of section 54 of type I.

FIGURE 7. Natural electric potential contour of section 40 of type I. Using the above classifications, all 61 prospecting sections were classified, and 11 abnormal sections were identified (Table 1). Relative to the abnormal data sets, leaking positions were determined by using the chemical exploration results. Using Geochemical Results To Judge Leakage. Whether the geochemical data are abnormal or not is determined by 1080

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 5, 2002

comparing the analyzed results of a sample with those of a reference sample collected far from the pipe line. Many samples taken from different horizontal distances and different depths from the pipe were analyzed so that the spatial variation of the chemical components could be determined. If all or most of the monitoring components of a sample are higher than those of the background, it is classified as abnormal. If only a few components are slightly higher, then there is no leakage. The investigation of surface pollution sources is very important in determining the leakage. Samples were analyzed from locations along the traverses of the 11 abnormal sections (Table 1). Only sections G54, G9, and G32 have both water samples and soil samples. The other sections only have representative soil samples. According to characteristics of the wastewater in the pipe, the components selected as analytical indexes for soil and water samples were the following: petroleum, benzene, toluene, phenol, cyanide, chloride, organic matters, total hardness, and pH. Analysis methods used to analyze soil and water samples are listed in Table 2. Among the components, benzene, toluene, petroleum, chloride, phenol, and cyanide were particularly important indexes. The analytical chemical results of the soil samples and the water sample are listed in Tables 3 and 4, respectively. From the analytical results of the water sample and the soil samples, anomalies of chemical exploration were classified by three types, depending on the number and concentration of anomalous components. Type I: The characteristic components, petroleum, benzene, toluene, phenol, cyanide, chloride, organic matters, and pH, were all checked out whenever possible, and the concentrations of 4 or more than 4 of the characteristic components are higher than 25% of the concentrations of the reference sample. Reference samples were collected at the places not affected by the wastewater. Type II: Most characteristic components are checked out and their concentrations are higher than the reference components. The concentrations of 3 characteristic components of 8 are higher than 25% of the concentrations of the reference sample. The other components are only slightly greater than the reference values. Type III: The concentrations of most characteristic components are close to that of the reference sample, having the concentrations of 0-2. Characteristic components of 8 are higher than 25% of the concentrations of the reference sample. Other components’ concentrations are slightly greater than the reference values. Anomaly classifications of GPR, electrical surveys, and chemical surveys are generalized in Table 5.

Exploration Results and Excavation Verification Comprehensive Analysis of Exploration Results. The typical detection results associated with cross sections G20, G32,

TABLE 1. Exploration Results of Geophysical Surveysa section

range, m

anomaly center

GPR

G9 G16 G17 G20 G31 G32 G40 G50 G54 G57 G59

29753-29850 25226-25626 25062-25162 24006-24102 19400-19745 19220-19400 15160-15320 11389-11509 10100-10374 8626-8935 8592-8691

29798 25326 25140 24034 19538 19300 15235 11469 10153 8841 8676

II II II III III II I I I III III

anomaly type of ES CON II II II I II II I I I II II

result from GP

anomaly type of CS

leakage possibility

II II II II III II I I I III III

III I III I III III I I I III II

III II III II III III I I I III III

/ / / / / I I I / / /

a ES, natural electric potential curve; CS, chemical survey; GP, geophysical prospecting; CON, natural electric potential contour; “/” represents the natural electric potential area survey was not carried out.

TABLE 2. Analysis Methods Used To Analyze Soil and Water Samples component pH petroleum chloride volatile phenol cyanide organic matter COD toluene benzene total hardness ammonia-nitrogen nitrate sulfate

analysis method

soil sample

water sample

x x x

x x

pH glass electrode ultraviolet spectrophotometry potential titration silver nitrate titration colorimetric method of 4-aminoantipyrine isoniacin-pyrazolone colorimetric method dichromate titration permanganometric method gas chromatography gas chromatography ethylenediamine tetraacetic acid (EDTA) titration Nessler colorimetric method ultraviolet spectrophotometry EDTA titration

x x x

x x x

x

x x

x x x x x

TABLE 3. Chemical Analytical Results of Soil Samplesa sample no.

soil sample position from pipe

pH

petroleum, mg/kg

chloride, mg/kg

phenol, mg/kg

cyanide, mg/kg

organic matter, %

toluene, mg/kg

benzene, mg/kg

N

reference reference reference 9-1 9-2 16-1 16-2 17-1 17-2 20-1 20-2 31-1 31-2 32-1 32-2 40 50-1 50-2 50-3 50-4 50-5 54-1 54-2 54-3 57 59-1 59-2 59-3

1 km west above 1 km west middle 1 km west below 3 m east above 3 m east middle 3 m east below 3 m east above 3m west below 5 m west above 3 m west below 3 m west middle 3 m west middle 3 m west above 3 m east below 3 m east above 3 m west below 3 m west above 5 m west below 10 m west middle 25 m west middle 3 m east middle 5 east above 5 east middle 5 east below 3 m east middle 5 m west above 5 m west middle 5 m west below

8.44 8.33 8.46 7.59 7.82 7.80 8.57 7.64 7.62 8.10 7.97 8.22 8.14 8.96 8.26 7.82 7.56 7.63 7.62 7.44 7.61 7.99 7.99 7.92 8.50 8.45 8.62 8.72

24.0 64.5 31.0 126.5 85.0 48.2 18.3 35.0 32.0 84.6 69.5 79.6 99.0 441.0 97.2 65.4 84.1 33.1 46.0 16.0 24.0 64.0 87.4 112.0 40.0

442.0 325.0 325. 0 608.7 136.0 141.6 220.7 24.4 103.2 697.7 225.7 79.8 163.4 215.9 646.3 297.4 307.6 276.2 373.6 438.2 164.2 585.0 1885.0 2255.0 307.8 221.0 325.0 273.0

0.015 0.020 0.021 0.0061 0.0012 0.0070 0.0110 0.0132 0.0504 0.002 0.0018 0.0023 0.004 0.007 0.0064 0.005 0.022 0.020 0.016 0.002 0.020 0.027 0.027

0.009 0.006 0.007 0.006 0.007 0.006 0.004 0.005 0.006 0.008 0.006 0.006 0.004 0.007 0.009 0.006 0.004 0.010 0.054

0.72 0.69 1.10 0.61 0.47 1.05 0.28 0.49 0.51 0.41 0.68 0.51 0.43 0.21 0.28 1.20 0.35 0.46 0.35 0.25 0.31 0.47 0.29 0.11 0.27 0.49 0.58 0.51

/ / / 0.71 3.05 0.96 0.80 4.51 0.08 / / / / / /

/ / / 1.38 1.80 0.22 0.22 / / / / / /

2 1 5 3 2 1 5 5 2 1 2 3 4 2 4 3 2 1 2 4 3 2 2 3 3

a Note: “-” represents 30 m

>3 of 8 components usually exceed background by 25%

V-shape magnitude < -2.0 mv difference > 5.0 mv width > 20 m

3 of 8 components usually exceed background by 25%

gentle shape potential positive

0-2 of 8 components usually exceed background by 25%

II

III

FIGURE 11. Natural electric potential curve of section 40 of type I.

FIGURE 10. Natural electric potential contour of section 32 of type I. those of the reference sample (Table 4). The analysis results of the water and the soil samples from the two simple wells indicate that there is no chemical anomaly. The anomaly belongs to our type III classification. From 19 298 to 19 328 m, the electric potential anomaly is consistent with the direction of the pipe (Figure 10). The potential anomaly is the result of a known iron pipe, which transports petroleum. The diameter is 0.4 m, the depth is 1 m, and the angle between the iron pipe and the pipe line is 49 degree. The anomalies of radar and electric potential are both of type II. However, the anomalies are not caused by the leakage of the pipe. The chemical anomaly is of type III. There is no leakage associated with this section of data. 3. Section G40. The GPR events near 15 274 m are continuous (Figure 3), but the events between 15 220 and 15 244 m are disorderly and strong. From 15 220 to 15 244 m at a depth of 2-4 m, which is the depth of the pipe, the section shows anomalous type I behavior. The natural electric potential curve is plotted in Figure 11. The background potential field is 2.0-3.0 mv, the abnormal shape of the anomaly is “U”-shaped, whose width

is about 70 m. The anomaly is located at 15 215-15 287 m, and the magnitude of the anomaly is -12.0 mv at 15 235 m. The electric potential anomaly is of type I. The natural electric potential contour is shown in Figure 7. In the range 15 225-15 275 m there is a closed low anomalous area distributed along the pipe line. The minimum anomaly of -28.70 mv is located at the position of the pipe at 15 235-15 239 m. A well is located at 15 235 and 3 m west. A soil sample was obtained below the pipe. It was yellowish-brown clay. The contents of organic matter, toluene, petroleum and cyanide are 1.2%, 4.51 mg/kg soil, 441 mg/kg soil, and 0.008 mg/kg soil, respectively. These contents are higher with respect to the reference sample, but the content of phenol is 0.004 mg/ kg soil, which is low. The chemical anomaly fits our type I classification. The radar, electric potential, and chemical anomalies are all of type I. Therefore, the pipe line most likely leaks along this section. 4. Section G54. In the range from 10 133 to 10 171 m, the GPR reflections are strong, the apparent period is long, and waveforms are disorderly (Figure 12). The radar anomaly is of type I. The natural electric potential curve is plotted in Figure 6. The background potential field is 1.0 mv, the abnormal shape of the anomaly is “U”-shaped, whose width is about 60 m. The anomaly is located at 10 156-10 221 m, and the magnitude of the anomaly is -8.5 mv at 10 189 m. There is an anomaly on the east side of the pipe too, but it is not as obvious as that on the west side. The electric potential anomaly is of type I. A well is at 10 153 and 5 m east. The sample taken from the well was mild clay in the shallow part and mediumcoarse sand in the lower part. The pH value of the soil is slightly lower than 8 and decreases with depth. The variation VOL. 36, NO. 5, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1083

FIGURE 12. The radar profile of section 54 of type I. of phenol is the same trend as that of pH. The content of organic matter is high in the shallow part but low in the deep part. The soil from the deep part is coarser than the soil from the shallow part. As a result, the adsorptive force is different at different depths for organic matter. The contents of chloride, petroleum, and cyanide increase rapidly from the shallow to the deep part of the soil sample. But the contents of chloride, petroleum, and cyanide decrease rapidly from the shallow to the deep part of the reference sample. The contents of COD, petroleum, chloride, and total hardness of the groundwater are much higher than those of the reference water sample (Table 4). Some chemical components, such as 24.8 mg/L of COD, 3659.3 mg/L of chloride, NH4-N, and NO3-N, are higher than those of the sewage water drained to the sea in a recent period of time. From the analysis of the water and soil samples, the chemical anomaly is of type I. In conclusion, the anomalies of GPR, electrical, and chemical surveys are all of type I. There is most likely some leakage within this section of pipe. Verification by Excavation. The possible leakage sections G40, G50, G54 of type I, G16, G17, G20 of type II, and G57, G59 of type III were excavated to verify the interpreted results. The excavations revealed that along G40, G50, and G54 some leakage locations were found. The diameter of the leakage points varied from 2 to 5 cm. At the anomalous position from the electric potential contour of G16, there was a black abnormal body, which contained organic matter of high content. Investigations and excavations show that there was no leakage on section G20. The beer and the chemical fertilizer factories must have caused the anomalies on section G20. The anomalies on G17, G57, and G59 were caused by outside pollutants, not the leakage of the pipe, because they are close to wastewater ditches.

Natural electric potential can be measured quickly, and it is interpreted easily. Thus, at the primary stage of a survey, it is applied to map possible leakage locations. Only the possible leakage locations need to undergo more detailed area surveys. This can accelerate the investigation rate. If both the GPR and electrical results are of type I, chemical surveys are necessary to conduct. Otherwise, they do not need to be considered. Since the components of soil and water vary greatly along the long pipe line, it is better to collect reference soil and water samples at each anomalous location. Analyzing pollution situations around the anomalous locations can help us understand the causes of indicated anomalies. The chemical components in water and soil react under the action of bacteria. The factors that affect the chemical constituents in soil are more than those that affect the chemical components of water. The components in water reflect the leakage better than those in soil. For example, section 54 is better assessed from the water chemistry than the soil chemistry. As a result, water samples should be collected if possible. It is better to conduct electrical and GPR surveys in the dry season and during periods without irrigations in order to minimize the influence of precipitation and irrigation. We end this paper with some conclusions. The exploration and determination of the leakage in the nonmetallic pressureless pipe is one of difficult issues. Ground survey is a procedure of inferring and identifying. The differences of media properties between leakage influencing area and background are the theoretical foundation of using GPR, electric potential, survey. It is obvious that these two survey methods can be used to determine the anomalous area fast and effectively. GPR and electric potential survey results are classified into three types, and chemical survey schemes are decided according to the combination character of two geophysical methods. This can greatly reduce the blindness of the chemical survey and may save the exploration cost. Classifying chemical exploring result types aids us to focus on the most suspected spots of leakage because the chemical survey gets more direct results. However, the influence from other pollution must be eliminated. The site condition must be carefully observed and recorded. By verification of excavation, it is proved that the leakage only happens among the spots at which the anomalies are all of type I in all three survey methods. It shows that the classifying of the anomaly type is proper and appropriate. This analyzing result greatly reduces the scope of excavation and maintenance for the user.

Acknowledgments Special thanks to Dr. Zhou Hui at Ocean University of Qingdao for reviewing the English manuscript. Funding was provided by Key National Nature Science Foundation under Contract 40036010 and by National Nature Science Foundation under Contract 499760028.

Literature Cited Discussion The results from combining GPR, electrical, and chemical surveys yields a relatively quick, accurate, and economic technique for detecting leakage from pressureless cement sewage water pipes. The field and interpretation work of 13 km of this detection took 1 month, respectively. The leakage of wastewater from a pipe causes compositional differences between the immediate media surrounding the pipe and the background media. The leakage can be detected by GPR, electrical, and chemical surveys because of the electromagnetic and chemical differences. 1084

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 5, 2002

(1) Wang, D. Y. The application of a novel pipe detection technique. Beijing Water Conservancy 1996, 1, 56-58. (2) Cheng, H. Metrotech FM9800 series pipe detector. Bulletin Mapping 1996, 2, 33-40. (3) Benson, A. K.; Mustoe, N. B. Integration of electrical resistivity, GPR, and VLF surveys to help map groundwater contamination produced by hydrocarbons leaking from underground storage tanks. Environ. Geosci. 1998, 5(2), 61-68. (4) Al-Saigh, et al. The detection of leakage of dam with the natural electric potential method. Dynamics Geological Sci. Technol. 1995, 4, 31-34. (5) Wu, Q. The advantages of ground penetrating radar. Dynamics Geological Sci. Technol. 1995, 10, 1-3.

(6) Liner, C. L.; Liner, J. L. Ground penetrating radar: a near-surface experience from Wathington County, Arkansas. The Leading Edge 1995, 14(1), 17-21. (7) Zhou, H.; Sato, M. Application of vertical radar profiling technique to Sendai Castle. Geophysics 2000, 65(2), 533-539. (8) Zhou, H.; Sato, M. Fracture imaging and saline tracer detection by crosshole borehole radar data migration. In Eighth International Conference on Ground Penetrating Radar; Noon, D., Stickley, G. F., Longstaff, D., Eds.; SPIE: 2000; Vol. 4084, pp 303-307. (9) Liu, H. J.; Jia, Y. G. The application of GPR to map mined-out area. J. Geological Hazards Environ. Preservation 1999, 10 (9), 73-76.

(10) Tadeusz, U. J. In search of plumes: A GPR Odyssey in Brazil; The 64th SEG Annual Meeting, NS1.7, 1994; pp 569-571. (11) Benson, A. K.; Mustoe, N. B. DC resistivity, GPR, and soil and water quality data combined to assess hydrocarbon contamination. Environ. Geosci. 1996, 3(4), 165-175.

Received for review December 11, 2000. Revised manuscript received November 7, 2001. Accepted November 16, 2001. ES001954S

VOL. 36, NO. 5, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1085