The development of new oil production areas and deep water activities requires more efficient and reliable cathodic protection (CP) designs. In this regard, laboratory tests that simulate service conditions can be very valuable. Preliminary results have shown that it is possible to obtain, on a short term basis, meaningful data that can be applied to the design of offshore CP systems. Current practice and challenging aspects of the methodological aspects of using laboratory electrochemical techniques to develop CP guidelines for offshore structures are addressed. The discussion focuses on those aspects that are more relevant to obtain on a short-term basis useful information that could provide for the successful long-term operation of new CP systems. A protocol recently introduced by Hartt et al. was utilized to evaluate the potential of relatively short-term laboratory data for predicting mean current densities. New ways of predicting long-term current densities based on accelerated laboratory tests are discussed.

This paper is intended to discuss some of the methodological aspects of applying laboratory electrochemical testing to develop offshore cathodic protection guidelines. A new protocol recently introduced by Hartt et al¹ will be utilized to evaluate the potential of relatively short-term laboratory data for predicting mean current densities. The advantages of using electrochemical laboratory techniques should not be underestimated if we take into account the substantial economical impact that may result by anticipating long-term cathodic protection performance from relatively short-term data.

The problem of selecting the most appropriate methods to provide data of good quality for offshore CP designs in new development areas is not an easy task. The current state of the art in the instrumentation technology allows the monitoring of the performance of offshore CP systems. The experience and data accumulated from such monitoring systems can be of great value in order to improve the performance of upcoming CP designs. However, the expansion of oil production operations to deeper waters and new development areas needs quality data that could provide for the successful long-term operation of the new CP systems.

The exposure of instrumented panels in the targeted areas is a good alternative, the data can be obtained in the real conditions where the structure will operate. It is an approach that several major oil companies and research institutes have chosen to obtain their current density requirements. Another alternative is to conduct laboratory testing. The advantage of obtaining these requirements with short-term laboratory techniques are obvious. The use of such techniques would imply cost savings and also would provide quality data with the degree of confidence that the CP designers need.

Experimentation with natural environments in the laboratory is a technical challenge. The experimental design should be capable of reproducing similar kinetics of nucleation and growth of calcareous deposits in the specific geographical areas. It is well recognized that it is the quality of the calcareous deposits that determines the current requirements for CP designs. Additionally, when sacrificial anode materials are involved in such experiments, care should be taken to provide the anodic material the minimum conditions for a proper activation and steady dissolution.

Deepwater operations started in the early nineties accompanied by a dramatic move to deeper waters between the years 1994 and 1997. Such a short period of time is clearly not enough to accumulate all the necessary data to accomplish new designs. The most experience in offshore has been accumulated in shallow and/or warm waters which have environmental characteristics substantially different from those one can encounter in deepwaters. Temperature, hydrostatic pressure, oxygen concentration, velocity, resistivity, salinity, biological activity are some of the environmental factors that can be significantly different.

The principal environmental factors that determine the nature of calcareous deposition are temperature, velocity and the dissolved oxygen concentration. The interaction of the two last factors produces the so called term oxygen availability, which is in other words the effective amount of oxygen being transported and subsequently reduced on the cathode. The dissolved oxygen concentration in sea water is interconnected with the structure of the ocean currents and the biological activity in the respective site. New developments in the area of modeling kinetics of the calcareous deposition should bring the current status of interpretation of the electrochemical data into a new level that would allow one to predict long-term performance from data obtained in the minimum possible time frame.

Coupling through a resistor: the cathode and the anode are coupled through an external resistor of a known value. The resistor is sized to get either a target slope when following the slope parameter method² or a desired initial protection current density³. The value of the external resistor is such that it controls the total resistance of the circuit. The use of a resistor allows simulation in the laboratory the resistance of space frame type structures with multiple galvanic anodes¹. In this kind of setup there is no need to reproduce the real cathode/anode area ratios that can be found in the field. However, the anode exposed area is recommended to be higher than a minimum value so as to avoid anode passivation⁴.

Direct Coupling: The anode and the cathode are coupled without an external resistor. The current density and the potential of the cathode are a function of the electrolytic and circuit resistance. In this kind of setup the anode/cathode area ratio is often close to values found in the field^5,6. Additionally, current density is usually measured using a Zero Resistance Ammeter (ZRA) or the voltage drop across a precision resistor fixed externally between the anode and the cathode. The cathode or anode potentials are determined directly or by indirect calculation using the shunt resistor or the determined resistance of the circuit.

Potentiostatic: Potentiostatic techniques are useful when investigating the chemical reactions that take place on the cathodes at some specified potentials representative of values found in real structures⁵. Of particular interest is the composition and corrosion protective characteristics of the calcareous deposits formed at different potentials and the influence on these properties of such environmental variables as velocity, sea water chemistry , pressure and temperature.

Polarization of an electrode in aerated water to a specified cathodic potential is invariably accompanied by oxygen concentration polarization such that current density decreases with exposure time. Presumably, some properties of the calcareous deposits formed at potentials closer to -1.03V are unique in comparison to less negative potentials and it is probably a consequence of some aspect of its chemistry or structure⁷.

This technique is also used to characterize passive tendencies on sacrificial anode alloys. In this case, the anode is potentiostatically or potentiodynamically polarized to potentials less negative than the open circuit potential. The presence of a shoulder or plateau in the anodic polarization curve indicates a tendency for passivation and therefore a tendency of the anode to polarize to potential values less negative than the protection potential. When conducting a single polarization curve the information lacks of a long-term insight.

This kind of evaluation can also be conducted presetting several samples of the same anode at different potentials⁴. Resulting anode current is determined over several time intervals at each set potential thus obtaining the variation of the resulting polarization curve versus time. In this case the data reflects long-term information. However, depending on the environmental variables being evaluated the required time to obtain accurate long-term information may vary considerably.

Galvanostatic: This technique is widely used to evaluate anode current capacity and efficiency. The NACE standard TM0190-98 describes the procedure to determine the current capacity of aluminum alloys in the laboratory. Using a galvanostat the anode current is preset at a specific level and the current capacity is determined by integrating the current versus time data which is equivalent to the total charge transferred during the test and also the anode weight loss. The mentioned standard provides results which are only applicable to environmental conditions close to the conditions indicated in the standard. Conversely, free corroding techniques are more versatile since they can provide anode and cathode performance data at the same time. Additionally, free-corroding techniques better simulate the way an anode functions in practice, whereas standard galvanostatic techniques preset specific current value that is not likely to occur in the field.

The selection of a test duration in laboratory testing is based on the compromise between two contrary trends: (1) A need for obtaining credible data in the shortest test duration possible and (2) The attainment of stable calcareous deposits on offshore structures protected with sacrificial anode systems is a long term process and therefore the protection current densities decrease even after 10⁵ hours¹. Two principal groups of sea water environments are defined by temperature and they are widely recognized as warm and cold and/or deep waters.

There are differences between the composition of the calcareous deposits in warm and cold waters. Since the solubility of calcium carbonates in warm temperatures is less, the calcareous deposits at these temperatures are richer in calcium carbonates and have a lower content of magnesium hydroxide. The opposite is found at cold temperatures where the calcareous deposits are less protective because the Ca:Mg ratio is lower than at warm temperatures. The current densities reported for structures in cold waters are higher than in warm waters and the differences found can be of one order of magnitude¹. From reference 3 it can be noted that the precipitation kinetics of calcareous deposits in cold water is slower than in warm waters over a period of ~ 10⁵ hours (~ 10 years).

The results provided in reference 3 suggest a way to interpret results obtained with laboratory testing on a short term basis. As was mentioned before, field and laboratory data show that calcareous deposits can still be growing after a period of approximately 10 years. Additionally, the data cited in reference 3 indicate a linear relationship between current density and time in log-log coordinates. Taking these results in consideration, short term data measured over a period of approximately one year could be mathematically processed following the cited linear relationship and extrapolated to a period of 10 years or even further. The methodology used in reference 3 as a protocol for defining mean current densities will be utilized in this paper to qualitatively evaluate data obtained in laboratory tests carried out over a period of one to one and half months. The results of this analysis are discussed in the following sections.

Because, of it variability, seawater is not easily simulated in the laboratory for corrosion testing purposes⁸. Stored seawater is notorious for exhibiting a corrosion behavior that is different from that of the water mass from which it was taken. This is due in part to the fact that the minor constituents, including the living organisms and their dissolved organic nutrients, are in delicate balance in the natural environment. This balance begins to change as soon as a seawater sample is isolated from the original water mass. When tests are conducted in the laboratory with transported sea water the researcher needs to realize the importance of the continuous refreshing of the natural seawater. Measures should be taken to keep the environment as live as possible. These measures can include agitation and aeration of the stored seawater while not in test. It is also important to store the environment samples during a short period of time. A period of two weeks may be reasonable to keep the stored environment alive while replenishing the respective experimental facility at an appropriate rate.

The principal cathodic process in sea water is oxygen reduction. The oxygen is transported from the bulk to the surface across the boundary layer by diffusion. Accordingly, the oxygen concentration changes from a minimum at the surface to the level of the bulk concentration at the outer edge of the boundary layer. In other words, the transport of oxygen is determined by a concentration gradient. However, during the cathodic protection process, calcareous deposits will form on the metal surface. Therefore, in order to reach the metal surface, oxygen will diffuse through the calcareous film.

In order to simulate appropriate hydrodynamic conditions in the laboratory, the type of flow that applies to the specific application has to be considered. Sea water flow is, in general, turbulent⁹, and this turbulence controls the transport of oxygen to the metal surface. A model for addressing the combined influences of electrolyte turbulent flow and presence of calcareous deposits was developed for the mass transfer of oxygen to the metal surface⁹. The model has been applied successfully to different flow geometries like steel plates and cylindrical shaped bodies⁹ and pipes¹⁰. The model analytically includes the Sherwood (Sh), Schmidt (Sc) and Reynolds (Re) numbers and it is described by the following equation:

---- (1)

where:
i_L is the oxygen limiting diffusion current,
D is the diffusion coefficient of oxygen,
n number of exchanged electrons,
F is the Faraday constant
C_b is the bulk oxygen concentration,
x is a dimensional parameter related to the flow geometry,
Sh is the Sherwood number,
t is the calcareous deposit thickness,
p is the porosity constant of the calcareous deposits
The Reynolds number is expressed as:
 

---- (2)

where:
D_e is the equivalent diameter
v is velocity
n is the kinematic viscosity

Laminar flow occurs when the Reynolds number is less than approximately 2100. The flow is said to be in a transition zone with a Reynolds number between 2100 and 4000 and fully turbulent for values over 4000. Schmidt number characterizes the diffusional properties of the electrolyte.

Rather than keeping in the laboratory a surface shear stress equal to a value representative of a service situation, flow modeling for offshore cathodic protection has been focused on providing a model that can describe and predict the oxygen transport and its limiting diffusion current densities upon turbulent flow conditions. Selection of an appropriate Reynolds number which will produce turbulent conditions seems to be the starting point in simulating service conditions in the laboratory. Equation (1) has been utilized to asses the calcareous deposits porosity¹⁰ and also to predict the current density decay trends for cathodically polarized steel specimens in defined flow fields¹.

An experimental facility has been developed to simulate low to warm temperatures, various levels of dissolved oxygen and velocities typical for deep waters. The experimental apparatus is shown in Figure 1. This apparatus was developed within a program which has expanded to include joint industrial sponsorship. This facility is provided with fresh sea water obtained from an offshore site in the GOM.

A computerized acquisition system collects current and potential data while the carbon steel cathodes are in electrical contact through an external resistor with Al-Zn-In anodes. The measurements are conducted using a constant high initial polarization current density which is set at the start of the experiment. The potential data is measured versus a Ag/AgCl electrode. A special board carries out the analog-digital conversion of the current signal measured using a zero impedance ammeter.

As seen in Figure 1, velocity can be set using a stirring unit located at the edge of a plexiglass tube contained in the electrochemical cell. The cathodes and anodes are located inside the aforementioned tube. The electrochemical cell was continuously replenished with fresh seawater. Given the geometry of the cathodes (square shape) and accordingly their Reynolds number, the tests were conducted in the transitional zone from laminar to turbulent conditions.

The flexibility of the experimental apparatus allows investigation of the effects of temperature, oxygen and velocity typical for deepwaters, on the current and polarization characteristics of various Al based anodes using statistical experimental design. The principal experimental responses that could be subjected to statistical analysis are the cathode current density, cathode potential and anode current capacity.

To qualitatively analyze data, the protocol recently introduced by Hartt¹ et al was utilized. According to the mentioned protocol, current density and time are plotted on log - log coordinates. This type of plot allows identification of different stages of the calcareous deposits precipitation process¹. The first stage consists of a decrease in current density produced by the precipitation of a thin, Mg-rich deposit within the first several minutes of exposure. Figure 2 shows this kind of plot for an Al-based sacrificial anode and four different environmental conditions.

As shown in Figure 2a the first stage is absent. The second stage corresponds to the time required for calcite or aragonite to nucleate. Subsequently, a decrease in current density occurred showing the start of nucleation of Ca-rich deposits. The stage of nucleation and growth of calcareous deposits exhibits a second and more steeply decrease in current density.

As shown in Figure 2a, all the mentioned stages except the first one were confirmed in the log-log plot. From the environmental conditions presented in this figure, Condition 1 exhibited more clearly the precipitation stages mentioned above. This result could be attributed to a beneficial effect of a higher high initial polarization current density Additionally, in that test there was no velocity which can lead in turn to a thicker high pH zone near to the cathode surface. Conditions 2, 3 and 4 exhibited the nucleation period and presumably the second decrease in current density is about to begin but more clearly for condition 2. For conditions 2, 3 and 4 a higher initial polarization current density may help to overcome the detrimental effect of velocity. The effect of velocity consists in wiping away the hydroxide ions produced upon the cathodic reduction of oxygen. Figure 2b, on the other hand, shows a totally different scenario; these data can be identified more likely with the stage where no nucleation of calcareous deposits has occurred.

Figure 3 shows the same type of diagram but for an Al-based anode of a different composition. In this case, a pronounced decrease in current density was observed. However, this time the effect was not related to the calcareous deposits nucleation and growth but to anode passivation. Figure 4 shows that for this anode the decrease in current density was accompanied by loss in cathode polarization.

An attempt to predict a mean current density based on the procedure presented by Hartt et al¹ was conducted using the data obtained for the condition 1 shown in Figure 1 and for four different anode compositions. As shown in Figures 2a and 3, the second and more pronounced current density decrease was regressed to a linear relationship. Three lines are shown in the diagram: (1) c = 0, (2) c = 3 and (3) c = -3. In the protocol introduced by Hartt, sigma means the standard deviation of data collected from several sources (laboratory and field testing). In the present analysis sigma represents the standard deviation of a single set of data with no contribution of different environmental conditions.

The equation utilized in the analysis is as follows¹:

where: c is the level of conservatism,
s is the standard deviation of the laboratory data,
a and b, intercept and slope obtained from the log-log linear fitting results,
i current density mA/cm²,
T time hours

The data points utilized in the linear fitting were taken starting at a test duration of 100 to 200 hours. Table 1 shows the mean current densities obtained for four different anode compositions. The present calculations were carried out for a conservatism degree of c=3s. It is interesting to note that the mean current densities were calculated to predict 20 years of service and are pretty much in the range with the value reported in 3 for a time of 20 years and zero degree of design conservatism.

Log-log plots indicate that for most of the environmental conditions studied, the growth stage of the calcareous deposits was not attained and consequently, longer test durations are required. Based on these findings, tests durations from 6 months to 1 year may be required. On the other hand, the obtained results for condition (1) in Table 1 show that the Hartt protocol may work when processing early data corresponding to the growth stage (~200 to 2000 hours).

Positive results in that direction could be very beneficial for short-term oriented laboratory programs. It is clear that the aforementioned time range will not be enough for the majority of the environmental combinations that can be found in the field and even more for deepwaters. A range of 1 to 1.5 years can be still considered within the growth stage and also a short-term duration as compared to the prospective of 20-30 years CP designs. The new mean current protocol could provide useful results in the 1 to 1.5 years time range.

The results presented in this paper were obtained within a feasibility program, which has been expanded for multiclient industrial sponsorship. The extension of this program will introduce several experimental modifications from what was learn in the feasibility program. First, based on the discussion above the test durations will be extended to longer durations. Secondly, more attention will be paid to simulation of the high initial polarization current densities by selecting a range of slope parameters. This will prevent being in the overprotection range. In a situation like this, when the system is demanding a current density over the limiting diffusion current, the system can adjust to that current but by increasing the reduction of water contribution to the overall cathodic process. This situation is likely to occur in environments with substantially low dissolved oxygen concentration and low velocities. In cases like this, the polarization process shows a horizontal line in the cathode polarization vs current density plots. The effect of passivation shown in Figures 3 and 4, could be presumably related to the anode area exposed to the environment (~2 cm²). For the extension of the program the anode exposed area will be increased to 40 to 100 cm².

In a methodology proposed for applying laboratory results in designing cathodic protection systems², the maintenance or mean current density was arbitrarily defined as the final current after 30 days. This decision was made, considering that the current densities reached stabilization during the aforementioned test duration period. Conversely, to what was thought at the time the referred paper was written it has been proved that current densities can continue to decrease even after 10⁵ hours. As a matter of fact, the current densities obtained within this program at cold temperatures did not reach even the growth stage of the calcareous deposits.

In the extension of the program the new mean current density protocol¹ will be utilized. Even though, there are still uncertainties as to when current densities will level out, current field data shows that stabilization of current densities in warm and cold water is not likely to be attained in a period of 10⁵ hours. Consequently, extrapolations of current density data from early growth stages can give quite valuable orientative values.

The current practices to determine mean current densities for CP designs imply the measurement the current densities on the respective protected steel cathodes over a long period of time. This is because the nature of growth of the calcareous deposits is an slow process. Field and laboratory data show that even for a period of 10 years current densities do not exhibit a threshold value. It is obvious that laboratory measurements can not be based on a such long term prospective. Nevertheless, the long-term current densities need to be measured or at least estimated to provide for a more accurate calculation of the mean current density. An estimation of the long term current densities, theoretically or better, experimentally, will allow definition of the a long term shape for a plot similar to the log-log plots shown in Figures 2 and 3

For example, a way to estimate long-term current densities in the laboratory could be the simulation of the long-term structure and composition of calcareous deposits. Accordingly, a reasonable approach could be the development of an experimental procedure that could accelerate the nucleation and growth kinetics of calcareous deposits for a given set of environmental conditions. A wise use of modern electrochemical techniques, including DC and AC methods, combined with advanced structural characterization techniques as SEM and XRD seems to be the necessary tool to solve this challenging problem. This research strategy is planned to be included in a joint industrial sponsorship program as a continuation of this investigation.

Free corroding techniques seems to be the most appropriate way to obtain the current requirements to develop up-to-date offshore CP guidelines in the laboratory. The use of electrochemical techniques in the laboratory to develop offshore is a technical challenge due to the complexity of the environments that are simulated. The definition of the variables that can not be overlooked should still be addressed with quantitative results. New mathematical approaches need to be developed in order for a better understanding and interpretation of laboratory results.

Based on the experimental study conducted herein the following conclusions were made:

1. Plot of current density vs time data in log-log coordinates can be a valuable tool to qualitatively analyze the various stages of calcareous deposition on cathodically protected surfaces. A sharp decrease in current density could be an indication of anode passivation. The data analysis suggests also that the use of initial polarization current densities higher than 200 mA/m² can be of benefit to shorten the time to reach a defined trend within the growth stage of calcareous deposits for the experimental conditions that involve velocity.
2. Data obtained by coupling sacrificial anode materials of several compositions with carbon steel cathodes in the laboratory, showed that in general, test durations of 30 to 45 days are not sufficient to provide mean current densities meaningful for a long-term CP design prospective (20-30 years) for low temperature waters.
3. Application of the new Hartt protocol for mean current densities to data corresponding to the early portion (200 to 1000 hours) of the growth stage of the calcareous deposits exhibited consistency with values reported by the author of the protocol. This result suggests that duration of the exposures in laboratory tests can be optimized so that excessive test durations (more than 1 year) could be prevented. The values reported in this work were obtained assuming a high level of conservatism (c = 3s) and were comparable with values obtained in reference 3 with a low level of conservatism (c = 0s). This strategy will be incorporated within the scope of the present program extension.

The authors are grateful to the Texaco for providing the funding base for the research program from which this paper was developed and also to Mr. G. Farquhar from Texaco GED for his support and critical guidance during the execution of the program. The authors also wish to express their thanks to their colleague Dr. Russell D. Kane for his extensive reviews and contributions during the performance of this work and development of this paper.

References  Top

1. W. H. Hartt and E. Lemieux, "A Principle Determinant in Cathodic Protection Design of Offshore Structures: The Mean Current Density", Corrosion 99, paper 627, NACE International, Houston, Texas.
2. "Design of Galvanic Anode Cathodic Protection Systems for Offshore Structures", NACE International Publication 7L198, July 1998.
3. Carlos M. Menendez, H. R. Hanson and R. D. Kane, "Methodology for Deep Water Cathodic Protection Designs Based on Laboratory Results", Corrosion 99, paper 358, NACE International, Houston, Texas.
4. C. F. Schrieber, "Pitfalls Associated with the Laboratory Testing Versus Actual Performance of Aluminum Galvanic Anodes", Corrosion 93, paper 529, NACE International, Houston, Texas.
5. S. Rossi, P. L. Bonora, R. Pasineti, L. Benedetti, M. Draghetti and E. Sacco, " Laboratory and Field Characterization of a New Sacrificial Anode for Cathodic Protection of Offshore Structures", Corrosion, Vol. 54, No.12, 1998.
6. C. F. Schrieber, "The Aluminum Anode in Deep Ocean Environments", Corrosion 89, paper 580, NACE International, Houston, Texas.
7. W. H. Hartt, C. H. Culberson and S. W. Smith, "Calcareous Deposits on Metal Surfaces in Seawater-A Critical Review", Corrosion, Vol. 40, No. 11, 1984.
8. S. C. Dexter, "Seawater", Metals Handbook Ninth Edition, Volume 13: Corrosion.
9. S. W. Smith, K. M. McCabe and D. W. Black, "Effects of Flow Parameters on the Cathodic Protection of a Steel Plate in Natural Seawater", Corrosion, Vol.45, No.10, 1989.
10. D. Hugus and W. H. Hartt, "Effect of Velocity on Current Density for Cathodically Polarized Steel in Seawater", Corrosion, Vol.55, No.2, 1999.

Click on the images for bigger images

FIGURE 1 - Experimental Facility

(a) warm temperature

(b) cold temperature

FIGURE 2 - Current density vs Time plot for one Al-based sacrificial anode alloy (B) in log-log type of coordinates. Data represents four sets of conditions and two high initial polarization current densities: (1) 300 mA/m², high O₂, no velocity, (2) 200 mA/m², high O₂, velocity, (3) 200 mA/m², low O₂, velocity, (4) Intermediate O₂, velocity.

FIGURE 3 - Current density vs Time plot for one Al-based sacrificial anode alloy (C) in log-log type of coordinates. Data represents four sets of conditions and two high initial polarization current densities at warm temperature: (1) 300 mA/m², high O₂, no velocity, (2) 200 mA/m², high O₂, velocity, (3) 200 mA/m², low O₂, velocity, (4) Intermediate O₂, velocity.

FIGURE 4 - Plot of Potential vs Current Density for Anode C in the following conditions: 200 mA/m², low oxygen, velocity and 25C.