M.J. Wilmott, T.R. Jack, J. Geerligs
Novacor Research & Technology Corp.R.L. Sutherby, D. Diakow
NOVA Gas Transmission Ltd.B. Dupuis
Foothills Pipe Lines Ltd.
Calgary
A soil probe has been developed and field tested that measures several soil characteristics active in the development of pipeline corrosion.
The Novaprobe incorporates several sensors in the single probe tip for in situ measurement of soil resistivity, soil redox potential, soil temperature, and pipe-to-soil potential. Other probes have only been able to measure individual parameters.
Further development of the probe will incorporate a solid-state pH electrode.
External corrosion and environmental cracking problems pose a significant threat to the integrity of gas-transmission pipelines. Considerable effort is being spent to understand the mechanisms which cause corrosion and environmental cracking and to identifying the conditions under which they occur.
To date, extensive excavation programs have determined correlations between conditions and materials at specific sites and the probability of integrity loss through a specific corrosion mechanism.
Targets for excavation are chosen by in-line inspection (ILI) for parts of the system which are amenable to this technique. Where ILI is impractical, selection of potential sites remains largely less scientific.
Corrosion of underground metallic structures such as buried steel pipelines is an electrochemical process, the rate being a function of the soil pH, resistivity, temperature, and redox potential. The presence of bacteria also influences the corrosion rate.1
Previously developed similar tools have, as stated, measured only one or at best two such characteristics.
In association with the American Gas Association, Deuber and Deuber developed a probe for field measurement of redox potentials.2
Costanzo and McVey developed this probe further.3 Their probe consisted of two platinum electrodes coupled to a saturated calomel reference electrode.
A similar probe for field determination of redox potentials has also been developed.4 Measurements of soil redox potentials, soil resistivity, and soil pH have led to the development of soil aggressivity models for steel exposed to soil environments.1 3-6
Construction and first field applications of the Novaprobe are described here. A comparison is made between local resistivity measurements obtained with the probe and the resistivity obtained using the Wenner four-point method.7
Also compared are the aggressivity models developed in earlier work with the observed corrosion measured during this field program.
DEVELOPMENT
The Novaprobe was developed jointly by NOVA Gas Transmission Ltd. and Novacor Research & Technology Corp., both of Calgary, taking approximately 18 months from conception to first field application.
The probe consists of a hollow steel shaft with an instrumented tip and a handheld meter. It has a cylindrical design with all sensor connecting wires running through the center of the hollow shaft.
The data-acquisition system consists of a handheld unit weighing approximately 600 g (21 oz). The 18-in. probe tip (Fig. 1)(69436 bytes) consists of a series of metal rings used for various measurements. The complete probe can be almost any length.
The present design of the probe permits data collection up to 1.9 m (6.3 ft) deep. Longer probes are currently under construction.
Four stainless steel rings measure soil resistivity, two platinum rings measure soil redox potential, and a thermistor measures the soil temperature.
The probe tip also contains a copper/copper sulfate reference electrode, against which all potentials are measured. This reference electrode can also be used to measure pipe-to-soil potential if a suitable connection to the pipe is available.
Future probe designs will include a metal/metal oxide electrode for measurement of local pH.
Measurements of redox potential, pH, temperature, and pipe-to-soil potential all involve passive potential measurements with respect to the copper/copper sulfate reference electrode. Soil resistivity is measured with an ac-potential-drop technique with an ac signal of 1 khz, which provides a local resistivity value at probe depth.
In the field, the probe is pushed to the desired depth by hand. In most locations, a guide hole is made with a pipe-locating rod or auger so that only the tip of the instrumented probe need be inserted into undisturbed soil. In some locations, frozen or very rocky soil has prevented insertion of the probe.
Fig. 2 shows the probe being used in the field.
FIELD APPLICATIONS
Data obtained with the probe at three separate locations in Western Canada are presented in Tables 1, (21341 bytes)2, (16246 bytes)and 3 (22178 bytes).
Data in Table 1 (21341 bytes) were collected at sites in British Columbia, the data in Table 2 (16246 bytes) from sites in Alberta, and Table 3 (22178 bytes) data are from southwest Saskatchewan.
The redox potential data presented in these tables have been corrected for pH and the reference cell potential using the following equation: EH = Ep + 0.3v + 0.050 (pH-7), where: EH = the redox potential (NHE) and Ep = the potential measured with the probe with respect to the copper/copper sulfate electrode.
Where soil pH was not determined, redox potential values are reported with respect to the copper/copper sulfate reference electrode.
TEMPERATURE
Temperature readings from the probe can map temperature profiles around a gas-transmission pipeline. Gas compression leads to elevated temperatures with the pipe becoming warmer than the soil in the right-of-way (ROW).
In Table 1 (21341 bytes), data obtained at Location 6 were recorded at a distance of 6.2 m from the pipe.
At this location, the soil temperature was measured at 11.5 C. at pipe depth. Closer to the pipe, temperatures up to 23 C. were recorded.
Table 2 (16246 bytes) indicates that the observed soil temperature falls as the gas travels farther. Site 2 was 32 km downstream of Site 1. The temperature fell from 26.4 C. at Site 1 to 20.9 C. at Site 2.
SOIL RESISTIVITY
The soil probe is equipped with four metal rings which are used as electrodes for local four-point resistivity measurement. Because the spacing of these rings is only a few centimeters apart, the resistivity measured is for the soil immediately surrounding the probe.
Both soil resistivity and the oxidation-reduction potential (ORP) readings can change markedly between unsaturated and waterlogged soil. These readings can be used to find the top of the local water table around the pipe.
In Table 1, (21341 bytes) shallow readings (70 and 64 cm depth) at Locations 1 and 2 gave very high soil-resistivity measurements (28,600 and 49,800 ohm cm, respectively) as well as relatively high ORP readings.
These data are consistent with an unsaturated soil which has gas as the continuous phase in the pore spaces between soil grains. This condition results in poor electrical conductivity and allows oxygen penetration and exchange with the overlying air.
In contrast, deeper measurements (117 and 98 cm) gave lower resistivity readings (3,780 and 3,880 ohm cm) and more negative ORPs indicating more anaerobic conditions.
The resistivities in the saturated zone converted to conductivities (0.02 to 0.3 Ms/cm; mS = milleSiemens) agree well with values obtained for local groundwater samples taken at the same location (0.09 to 0.20 mS/cm). Resistivity measurements at a depth of 80 cm at Location 3 (Table 1)(21341 bytes) were 7,440 ohm cm.
Taken together, these resistivity measurements indicate that the top of the saturated zone at Sites 1 and 2 was between 70 and 80 cm depth at the time of the survey.
Data presented in Table 3 (22178 bytes) were obtained on a pipeline in Saskatchewan over a distance of approximately 250 m. Readings were recorded every 16 m.
Readings 11 and 14 were obtained at the base of a small hill, whereas Readings 12 and 13 were recorded at the top of the same hill.
The resistivity values obtained at the base of the hill are lower than those obtained closer to the top, a result of better drainage and lower water saturation for hill top soil.
Readings 3-10 of Table 3 (22178 bytes) were recorded in a low area where dried white salts were observed on the soil surface. A low soil resistivity measurement at a depth of 1.7 m indicates a subsurface salt seep near where Reading 7 was taken.
Soil-resistivity measurements made at locations in Alberta (Table 2)(16246 bytes) were consistent with values of conductivity for ground water samples taken at the same locations.
Soil resistivities are normally measured in undisturbed soil along the edge of right-of-ways by a standard Wenner four-pin measurement.7
In this method, four metal pins arranged in a line, are pushed a few centimeters into the soil. A pin spacing of 1.5 m obtains a resistivity reading which is based on the cumulative properties of the soil to a depth of about 1.5 m.
Wenner four-pin resistivity values obtained at the site described by the data in Table 1 (21341 bytes) (17,000-25,000 ohm cm) therefore represent a blend of saturated and unsaturated soil resistivities.
This is consistent with the local soil-probe readings for these zones being lower in the saturated zone and higher in the unsaturated zone than those given by the standard four-pin method.
Pipelines are protected by means of a dual system of coatings and cathodic protection.
Many pipelines are coated with insulating polyolefin tape. The impressed current cathodic-protection system is intended to protect the exposed steel should coating defects occur.
When polyolefin tapes disbond, however, they form a barrier which shields the cathodic-protection current from the pipe surface.8 How far cathodic protection currents can penetrate under a disbanded coating as a function of the conductivity of water trapped under the disbonded coating has also been determined.8
The local resistivity values obtained with the soil probe can therefore be used directly to determine how effective cathodic protection will be for a tape-coated pipeline in a given soil environment if disbondment of the tape occurs.
SOIL REDOX POTENTIAL
ORP readings indicate how reduced or oxidized is the environment surrounding the tip of the soil probe. In most soil environments, the ORP reflects the balance between the rate of oxygen entry and the rate of its consumption by biological or chemical processes.
In an unsaturated zone near the surface of the soil, the continuous phase between the soil particles is gas. In this case, diffusion of oxygen from the overlying air results in positive ORP readings (Table 1,(21341 bytes) Locations 1 and 2, Site 1).
In the underlying soil zone saturated with water, oxygen arrival is limited by its solubility in water and the rate of diffusion. Where the rate of uptake of oxygen exceeds the rate of its arrival, ORP readings drop to negative values corresponding to a reducing environment more or less free of oxygen.
The aerobic/anaerobic interface coincides with the height of the local water table in such a site as noted in the previous discussion of resistivity.
Based on ORP data, this appears to be around 70 cm below surface in the site documented in Table 1 (21341 bytes). This is consistent with the conclusion based on resistivity measurements.
Another example of reduced water saturation being evident in the ORP readings can be seen in Readings 12 and 13, Table 3 (22178 bytes). Readings at a depth of 1.7 m under a small hill on the right-of-way indicate that the soil is drained more effectively than the surrounding, less-elevated area.
This interpretation is supported by the increased soil resistivity for these two readings.
Because the rates of oxygen ingress and of oxygen consumption determine the steady-state oxygen concentration in a soil matrix, ORP readings are also influenced by the soil type and permeability.
Soil analyses for selected locations probed in Table 1 (21341 bytes) are given in Table 4 (13482 bytes).
An ORP reading of -600 mv (NHE), taken at a depth of 1.2 m in a highly impermeable clay sediment, under a standing pool of water at the edge of the right-of-way, indicates very reduced, highly anaerobic conditions (Table 1, (21341 bytes) Location 6).
Soil analysis shows that this soil was rich in clay and plagioclase but had less sand content than other locations showing less negative ORP values (Table 4)(13482 bytes). The sediment in Location 6 appeared to be impermeable, had the consistency of a modeling clay, and had the dark gray color characteristic of anaerobic environments.
IMPLICATIONS
FOR CORROSION
Previous studies have attempted, with some success, to correlate soil properties with the corrosion of exposed unprotected steel.1-6 Soil probes have been described specifically for this purpose.2 3 4 9
Booth, et al., evaluated 87 soils with weight-loss coupons over a period of several years and correlated their findings with soil measurements taken in part with a soil probe.4-6 Their findings allowed them to identify aggressive and nonaggressive soils based on ORP and resistivity readings.
Aggressive soils were those having an ORP of less than 400 mv (NHE) for non-clay soils or 430 mv (NHE) for clay soils or a resistivity of less than 2,000 ohm cm.
In borderline cases, water content was used to designate a site. Levels greater than 20% were deemed aggressive.
In Booth, et al.'s, definitions, clay soils are those having a fine particle size and low permeability, rather than those having a substantial fraction of clay minerals present. These factors were used correctly to identify the probability of severe corrosion in 81 of the 87 sites considered.
Starkey and Wight had previously developed a finer scale of gradations for soil aggressivity.1 Based on ORP readings, their scheme placed increasingly aggressive soil in the following sequence of ORP (NHE) and anticipated corrosion: <100 mv, severe corrosion; 100-200 mv, moderate corrosion; 200-400 mv, slight corrosion; >400 mv, no corrosion.Field observations reported here can be correlated to the corrosion seen on unprotected steel coupons located at the sites where probe readings were made. These observations are inconsistent with the parameters established by Booth, et al.
Based on the criteria of Booth, et al., all the sites documented in this study have aggressive soil conditions at depths greater than 1 m. ORP readings are Table 3 (22178 bytes), where high values of ORP were obtained in the well-drained soil at a small hill.
Given that all the sites reported in this study were locations having wet soil conditions, the prevalence of anaerobic conditions (that is, aggressive conditions) is not surprising.
Resistivities give mixed results in terms of predicted soil aggressivity, but only Reading 12, Table 3(22178 bytes), shows resistivity greater than 2,000 ohm cm with a correspondingly benign ORP reading. This is the only location which would be deemed nonaggressive according to the criteria of Booth, et al.
Corrosion data are available for some of the sites described here based on observation of the pitting and thinning of unprotected steel coupons installed during construction of the pipeline.
Corrosion was found to be negligible in the site described in Table 1 (21341 bytes) despite ORP readings generally
Corrosion was also insignificant in the organic soil of Site 1 (Table 2)(16246 bytes) despite ORP readings (Table 2)(16246 bytes) showed significant corrosion consistent with an aggressive designation.
Predictions based on earlier work apparently require some caution. All wet sites will be designated aggressive on the basis of ORP, yet two of the three sites for which corrosion information exists fail to match the simple aggressive/nonaggressive designations predicted by the method of Booth, et al.4-6
The criteria of Starkey and Wight provide a better fit in these cases, predicting slight to moderate corrosion.1
USES OF THE PROBE
Although the probe was designed specifically to characterize corrosive soil environments for gas-transmission pipeline right-of-ways, the use of the probe is not limited to this.
For example, the probe can be used to select sites for further investigation in stress-corrosion cracking (SCC) field studies.
To develop correlations between site characteristics and the chance of Environmental cracking on line pipe, extensive excavation programs must be conducted. This process is well established for SSC sites based on very extensive exploratory excavation programs conducted largely by TransCanada PipeLines Ltd.10
The soil probe has a potential role for targeting excavation sites within a selected location. Soil characteristics at pipe depth can now be added to information on site topography, soil type, and drainage in the selection process.
Subsurface resistivity measurements by the soil probe can define the saline zone and the degree of upward migration at past or future pipeline construction sites. This would avoid more time-consuming and expensive sampling programs.
A subsurface saline seep was identified at the site described in Table 3 (22178 bytes) as being quite localized at pipe depth compared to surface indications. Focusing special excavation practices to control salt migration at the source of the seep could save time and money in construction.
As well as applications for pipeline operations, the soil probe has applications in environmental areas. The aerobic/anaerobic nature of a site can determine the rate and extent of natural or accelerated remediation, especially for hydrocarbon-contaminated soils.
Monitoring the ORP and temperature by soil probe may be a way to observe the conditions and process of remediation in the field.
The probe has been shown to be useful at a number of field locations and that information regarding soil aggressivity can be easily obtained.
Both TransCanada PipeLines and ARCO Exploration & Production Technology Plano, Tex., have shown interest in the probe. TransCanada has used the probe to characterize corrosion sites on its system, and ARCO is considering construction of a probe.
ACKNOWLEDGMENTS
The support of NOVA Gas Transmission Ltd.'s research advisory committee is acknowledged in the development of this probe.
REFERENCES
Starkey, R.L., and Wight, K.M., "Anaerobic Corrosion of Iron in Soil," AGA, 1945.
Deuber, C.G., and Deuber, G.B., "Development of the Redox Probe. Final Report to the AGA, Research Project PM-20," New York, 1956.
Costanzo, F.E., and McVey, R.E., "Development of the Redox Probe Field Technique," Corrosion, NACE, 14 (268T-272T), 1958.
Booth, G.H., Cooper, A.W., Cooper, P.M., and Wakerley, D.S., "Criteria of Soil Aggressiveness towards Buried Metals, I, Experimental Methods," British Corrosion journal, Vol. 2 (1967), pp. 104-08.
Booth, G.H., Cooper, A.W., and Cooper, P.M., "Criteria of Soil Aggressiveness towards Buried Metals, II, Assessment of Various Soils," British Corrosion Journal, Vol. 2 (1967), pp. 109-13.
Booth, G.H., Cooper, A.W., and Tiller, A.K., "Criteria of Soil Aggressiveness towards Buried Metals, III, Verification of Predicted Behaviour of Selected Soils," British Corrosion Journal, Vol. 2 (1967), pp. 114-18.
Wenner, F., Bur. Stand. Sci. Paper No.12,469,1916.
Jack, T.R., Van Boven, G., Wilmott, M., Sutherby, R.L., and Worthingham, R.G., "Cathodic Protection Potential Penetration under Disbonded Pipeline Coating,"Materials Performance, Vol. 33, No. 8 (1994), pp. 17-21.
Kajiyama, F., "Apparatus for Measuring the Corrosivity of Soil," Japanese Patent Application 58-208654, 1982 (in Japanese).
Delanty, B., and O'Beirne., J., "Major field study compares pipeline SCC with coatings," OGJ, June 15, 1992, p. 39.
Copyright 1995 Oil & Gas Journal. All Rights Reserved.