Based on presentation to NACE Corrosion 2010 conference, San Antonio, Mar. 14-18, 2010.
A Ti-MMO anode grid system placed between an aboveground storage tank's bottom surfaces and the secondary liner with properly distributed power-feed connections can protect the tank's bottom plates against external corrosion. Such systems can achieve uniform potentials along the tank bottom surfaces.
Inappropriate distribution of power-feed cable, however, can cause a nonuniform potential profile. Proper distribution of power-feed connections along the grid is vital in designing AST-bottom CP systems.
High contact resistance of reference cell to soil and electrochemical instability of permanently installed reference cells also demand alternative methods to ensure proper protection of metal surfaces, such as installing an electrical resistance probe or more reliable permanent type reference cells such as high purity zinc. Installation of soil access holes through the periphery of the ring wall can also help address these conditions.
Limiting anode surface current densities to around 10 ma/sq m (for coated bottom plates with anode spacing of 1.2-1.5 m) helps reduce the depolarizing effects of anodic reaction products (oxygen). Power-feed cable connections distributed along the grid should limit the current of each power-feed cable to 3-3.5 amp. Conductor bars should be 6 m apart.
The most appropriate criterion for tank bottoms in the discussed project is the 150-mv CSE polarization shift from natural potentials, measured by polarization decay. Implementing the criteria discussed in this article can reduce the current to the optimum value, reduce system operating cost, and lead to higher anode life and less oxygen depolarizing.
Background
Leaks due to corrosion of aboveground storage tanks at oil production, transportation, storage, and export facilities affect the petroleum industry adversely. External corrosion of AST bottom plates is one of the industry's primary concerns.
This article discusses various aspects of preventing corrosion of AST bottom plates constructed with a secondary containment liner by using cathodic protection. Kuwait Oil Co. routinely uses impressed-current cathodic protection with Ti-MMO anode grid system between the tank bottom and nonconductive external liner for AST.
Corrosion-resistant paint (coal tar epoxy) applied to the external (soil) side of bottom plates reduces CP current requirement, but the coating is expected to be partially damaged during construction of the tank bottom and heat during welding at the lap joints.
This article emphasizes a model defining adequate protection criteria for the external bottom of newly constructed crude oil storage tanks at the tank farms of Kuwait Oil Co. The crude export facilities are one of the major expansion projects at KOC, which include 29 new aboveground storage tanks, five 48-in. OD gravity lines, a 56-in. gravity line, three 56-in. OD offshore export pipelines, plant piping associated with tank farms and metering stations, and offshore buoys.
KOC divided the project into two contracts awarded to two contractors to achieve targeted timelines. This article discusses the performance of the CP system provided for the external bottom plates of 29 AST (27 crude oil storage tanks and 2 water tanks).
Two contractors with their own designs carried out CP system installation. Table 1 shows the slight differences that occurred between the installed systems despite use of the same project specifications. The article compares CP performance for the tanks with respect to factors such as protective potentials, applied current densities, and current distribution. It also studies and discusses the cause of low potentials at certain tank surfaces.
Design factors
The environment under the tank bottom is subject to several variations when compared with other buried structures, some of which are not completely definable during CP design. Designers normally verify all extremes to match with the actual model but must consider many parameters in designing, operating, and analyzing the potentials of such systems.
Some of the major factors affecting the performance of a normal tank bottom CP system include:
• Soil properties. Soil resistivity of tank bottom cushion constructed with washed sand normally vary from the time of construction towards the end of the design life due to seasonal changes, water ingress to the bottom surfaces, compaction of soil during construction, or product level variations, product temperature, etc. Anodic reactions also vary the electrolytic characteristic of soil.
Variations in soil conditions can influence CP performance, requiring frequent monitoring to verify proper operation. Changes in soil properties can lead to difference in current requirements, current distribution pattern, and contact resistance of the reference cell to soil in potential monitoring. The potentials measured during the initial phase of operation, for example, are more electronegative than those recorded after a period of tank operation.
Proper reference cell-to-soil contact due to available moisture retained in the soil during construction could have led to the initial potentials, but reference cell resistance may eventually increase during tank operation due to variations in the soil cushion.
Dry, clean sand used for the cushion has higher resistance than normal soil. Soil resistivity range for this kind of dry sweet sand varies from a minimum of 100 ohm-m to about 1,000 ohm-m, leading to a design range of 100-1,000 ohm-m.
• Depolarization of structure due to anodic reactions.
Installing anodes directly under the structures might cause anodic reaction products to influence the structure differently than systems with anodes installed remotely or away from the protected structure.
The primary anodic reaction at the anode surface of a Ti-MMO type anode ribbon in sand is expressed as 2H2O → 4H+ + O2 + 4e–.
A cathodic depolarizer such as oxygen will influence the polarization of the bottom plate's external surfaces.4 Anode current output creates more depolarization due to oxygen generation. Protective potentials and anode operating output must balance to achieve optimum polarization. Equilibrium between the cathodic reactions and depolarization caused by generated oxygen will polarize the cathode surface.
The initial potential structure shift (electronegative) will be rapid with a small increment of current, but further improvement may require high current due to oxygen. The depolarization reaction may prevail over cathodic reactions on a high current output in which polarization cannot be maintained.
• Native state (natural) potential. The measured natural potentials of AST bottom surfaces are normally more electropositive than other buried structures. To an extent soil properties as discussed earlier help create the low potential scenario.
Tank construction usually takes time after the bottom plates are laid and installed on the soil cushion. During this time the bottom plates remain unprotected and will have more metal ion concentration (due to corrosion) at the surfaces, resulting in more electropositive potentials than expected. Damage or holidays in the containment liner may lead to mixed potential measurements of the tank bottom's surface and surrounding copper earthing networks.
• Temperature. Maximum operating temperature of crude oil inside the tank normally measures 35-40° C. Considering the huge volume of product within the tank, neither ambient temperature nor metal temperature, particularly of the bottom plates, will have much effect. The maximum anticipated temperature on the bottom plates will therefore not exceed 40° C.
Temperature plays a direct role in validating the criteria used to protect tanks by cathodic protection. The internationally accepted –850 mv CSE and 100 mv polarization shift criteria are only valid under normal ambient temperature conditions of 20-25° C.
Several studies have reported the inadequacy of –850mv CSE and 100 mv polarization shift criteria at higher temperatures. Barlo and Berry assert the polarization shift criterion should be adjusted to 150-250 mv from 100 mv at temperatures of 60° C., while Morgan3 suggested an adjustment of 2 mv/°C. for high temperatures. Most studies recommend increasing the –850 mv CSE criteria to –950 mv CSE for temperatures >60° C.5
Criteria for the subject tanks should therefore use the extreme operating temperature of 40° C.
• Accurate measurement of structure potentials.
Facilities installed during construction can measure tank bottom surface potentials (Figs. 1-2). Such facilities include permanently installed reference cells and through-the-soil access PVC tubes intended to insert portable retractable reference cells.
Soil properties, contact resistance of reference cells to soil, and drying of permanently installed reference cells are some of the factors affecting tank bottom potential measurement. As the permanent reference cell can become electrochemically unstable with time, readings observed from them need to be closely verified with those measured by other means (retractable reference cells, readings obtained from other permanent reference cells on same tank, previous potential readings at that particular point, previous protective potential trend, etc.).
Calibration of retractable reference cells (Fig. 3) can occur before measurements, but high-contact resistance of the electrode to soil remains a concern. Pouring water inside the perforated tube to achieve better contact is not desirable, as it will distort the CP current's path.4 Wrapping the porous cap with wet sponge can help reduce contact resistance.
The reference cells' installation above the anodes could lead to high IR drop while carrying out "On" potential measurements. Measuring "Instant-Off" potentials by interrupting the T/R output and measuring the potentials instantly avoids error due to IR-drop.
Since the measurement of one reference electrode represents the protection level of the tank surface close to the measurement, readings need to be taken on other permanent cells and through the perforated tubes along the surface to analyze the protection level of the complete bottom plate.
Analysis of structure-to-soil potential readings requires considering tank product level. The entire bottom plate would have proper contact with the soil while the tank is filled with product and hence require more current compared to an empty tank (where the bottom plates remain in poor contact with the soil). An empty tank, therefore, would show more electronegative potential than a tank filled with product.
• Current distribution. Achieving uniform current distribution to the entire surface of the tank bottom requires careful anode grid design. Many design-stage parameters fix the spacing between anode ribbons and conductor bars, the orientation and number of power-feed cables, etc. These parameters include potential profiles based on earth potential rise formulas, structure current requirement, anode current density, voltage drop, and attenuation calculations based on experience from similar construction.
Most designers distribute positive power-feed connections along the grid (Fig. 1). Another approach makes the connections to the central conductor bar (Fig. 4). Calculation of current attenuation along the anode ribbon in a grid requires complex calculations (finite-element analysis) only possible with computer software applications. Table 1 and Figs. 1 and 4 show the number and orientation of power-feed cables on the subject tanks as designed by different designers.
Adequate criteria
Values of –850 mv CSE polarized potential and 100 mv polarization shift stand as the widely accepted potential criteria for adequate protection.1 5 Recorded natural potentials (Tables 2-4, Figs. 5-8) were generally 200-250 mv more electropositive than the normal expected values of steel in soil, requiring an average polarization shift of 500-600 mv to achieve the –850 mv CSE criteria. These values exceed the minimum polarization criteria of 100 mv specified in various international standards.
The true criterion for cathodic protection of a structure is polarization of the cathodes on the structure to the open circuit potentials of the most active anode on the structure. ISO 15589-1 (Petroleum and Natural Gas Industries—Cathodic Protection of Pipeline Transportation Systems) recommends adjusting –850 mv CSE for high soil resistivity values, moving it to –750 mv CSE for resistivity values of 100-1,000 ohm-m and 650 mv CSE for resistivity values >1,000 ohm-m. German Standard DIN 30676 also recommends adjusting the criteria to –750 mv CSE for soil resistivity values >500 ohm-m.
With consideration to the soil resistivity and depolarization effect on a structure due to anodic reaction products, therefore, 100 mv polarization shift stands as the optimum criterion for satisfactory protection. The criterion, however, might still need to be validated for temperatures of 40° C. As discussed, Morgan3 recommends 2 mv/°C. for temperatures >25° C., while Barlo and Berry suggest adjusting criteria to 150-250 mv for temperatures of 60° C. A 150-mv polarization shift criterion should thus be satisfactory for the subject tanks. Considering variation in soil under the tank, a 150-mv polarization decay method can be adopted.
KOC project
The expansion of crude export facilities involved construction of 29 AST (Table 1). Tank bottom construction and CP system installations conformed to Figs. 1, 2, and 4 and Table 1.
Impressed-current cathodic protection using a Ti-MMO anode grid system between the tank bottom and non-conductive external liner cathodically protected the tanks. System design focused on ensuring a relatively uniform current distribution along the bottom surfaces.
The grid-type design consisted of anode ribbons of metal mixed oxide (MMO) laid in parallel and connected by resistance-welded conductor bars laid perpendicular to the anode ribbons (Fig. 1). The conductor bars provided redundant connections to ensure integrity of the grid, lower circuit resistance, and enhanced current distribution throughout the grid. Interconnection of all anode ribbons to the conductor bar used a spot welding machine specifically designed for the job.
Power-feed cable connection encapsulation was vital in cathodic protection of the positive cables' connection since the copper in the cable, if exposed, would discharge current leading to rapid corrosion. Factory splicing 16 sq mm Kynar-HMWPE cables to a piece of conductor-bar material preceded field spot welding of the conductor bars to the grid. The plain titanium conductor bar acts as a conductor and does not discharge current from its surface. Sand backfill between the anode ribbon bed and the tank underside should consist of conductive clean-washed sand free from contamination (chlorides, sulfates, etc) and with a pH value of 6-11.5.
Installing a perforated PVC pipe wrapped with geotextile fabric allows access to the soil (electrolyte) close to the surfaces for measuring the structure's protective potential by placing the half cell on the soil close to the structure and tank bottoms despite their being completely sealed. Nine permanently positioned half cells also lie under each tank. Test facilities with cable leads connected to the tank body measure the potential of the tank bottom.
Reference cells and test facilities follow Fig. 1:
• One permanent reference cell at the center.
• Four permanent reference cells equidistantly placed 1 m from the tank edge.
• Four additional permanent reference cells placed equidistantly midway between the center of the tank and the edge.
• Three perforated 3-in. diameter UPVC pipes for retractable reference electrodes.
Tank classification can also fall into four cases based on diameter and power-feed orientation (Table 1).
Conductor bar spacing of 6 m is common on all tanks. Tanks under Case 1 have an anode-to-anode spacing of 1.5 m and anode-to-tank bottom spacing of 0.5 m. Other tanks (Cases 2-3 and 4) have a spacing of 1.2 m and 0.35 m, respectively. Structure potential profile during design used earth potential rise formulas from Von-Beckman.2 Potentials at points on the structure (structure surface close to anode and midpoint of two anodes) for all cases and design results show equivalent potential profile in terms of anode spacing.
Case 1 tanks differ from others in number and orientation of power-feed points on the grid. All tanks, except Case 1 tanks, have seven power-feed cables near the central conductor bar (installed parallel to the anode ribbons) similar those seen in Fig. 4. Case 1 tanks used 18 power-feed cables distributed along the grid by maintaining a minimum of one power-feed cable/conductor bar and two each on three conductor bars at the center of the grid.
Results
In all the cases the natural potentials of the bottom plates became more electropositive with time. Table 2 shows a comparison of natural potentials over a period. Fresh metal installed on moist soil would result in more electronegative potentials compared to a tank bottom surface on dry soil after a period of tank operation.
As tank construction normally extends 1-2 years, the bottom surfaces will remain in contact with soil during this period, leading to an increase of metal ion concentration on the soil surfaces due to corrosion reactions and more electropositive potentials. High contact resistance of the reference cell to soil surfaces is also a factor.
Table 4 and Figs. 5-8 show results for operating a tank CP system for roughly 1 year at a current density of about 5-8 ma/sq m for tank bottom surfaces.
IR drop in the measuring circuit and "On" potentials dropped after a few months of tank operation, possibly due to the decrease of resistance between the bottom surface and anodes, where the current remained constant. When the tank is filled with product, the bottom plates will remain completely in contact with the sand cushion and after a period will become more pressed-sagged into the soil cushion, reducing resistance between the bottom surface and anodes. During product loading and unloading movement on the bottom plates leads to abrasion of coating with soil, coating erosion, and a lower IR.
Tanks with power-feed cables on the center line (Cases 2-4) experience improvement (increased electronegative) in the polarized potentials at reference cells near the power feeds over time, while those away from the power feeds remained the same on a few tanks and diminished (became more electropositive) on most. Diminishment of tank-bottom resistance to anode after a few months' operation increases current attenuation along the grid, resulting in diminished current throw from those anodes away from the power feeds.
Table 4 and Figs. 5-8 show a large reduction in potential of tank surfaces away from the power feed for most large-diameter tanks (i.e., 92 m diameter) with time. A comparative study of tank-to-soil potentials of surfaces near the power feed to surfaces farthest from the power feed showed that after 1 year operation potentials at the edges of the tank were 35-45% less than those close to the power feeds. In Case 3, however, it was only 15% less.
Since measured potential at the surface is directly proportional to current received at the surface, the readings in Table 4 and Figs. 5-8 show a large attenuation of current along the grid for Case 2 tanks with 92-m diameter, which only increased over time.
Case 1 tanks constructed with power feeds distributed along the grid result in a uniform potential profile, except a slight dip at the center due to lack of a nearby power-feed connection.
Case 1 tanks use at least one power-feed connection on each conductor bar and additional power feeds on central bars for a total of 18 power-feed connections. Analysis of potential profile, geometry of connections, and attenuation along the conductor bar show 19 power feeds as preferable. This arrangement can distribute 66 amp of design current to 3.5 amp/power-feed connection.
Given no considerable reduction in potentials on the edges of Case 3 tanks (Fig. 8), typical design of power-feed connections on the center conductor bar was satisfactory for tanks of 72-m diameter.
KOC analyzed potentials along tank surfaces for depolarizing due to oxygen generated by anodic reactions. Oxygen generated at anode surfaces is proportional to the current density at the anode surface. The current output of anodes near the power feed is greater than other areas of the grid, leading to a greater anodic reaction and subsequent oxygen generation near power feeds.
On all the cases shown in Table 4 and Figs. 5-8, limited depolarization occurred due to oxygen generation at the anode surface. Surfaces close to power feed result in better potentials than the rest of the area and potentials close to power feeders even improve after a period of tank operation, implying limited effect of oxygen depolarizer at the CP system's operated output.
References
1. NACE Recommended practice RP 0193-2001, "External Cathodic Protection of On-Grade Carbon steel Storage Tank Bottoms," NACE International, Houston, 2001.
2. Von Baeckmann, W., Shwenk, W., and Prinz, W., "Handbook of Cathodic Corrosion Protection," Third edition, Gulf Publishing Co., Houston, 1997.
3. Morgan, J., "Cathodic protection," 2nd edition, NACE, Houston, 1993.
4. Gummow, R.A., "Cathodic Protection—Technological Changes and Application Challenges," 11th Middle East corrosion conference, Manama, Bahrain, Feb. 26-Mar. 1, 2006.
5. International Standard ISO 15589-1, "Petroleum and natural gas industries—Cathodic protection of pipeline transportation systems, Part 1: On-land pipelines," Geneva, 2003.
The authors
Abdul Wahab Al-Mithin ([email protected]) is team leader inspection & corrosion, south and east, at Kuwait Oil Co. (subsidiary of KPC), Ahmadi. He holds a degree in metallurgical engineering from the University of Utah. He is a member of NACE International and Society of Petroleum Engineers.
Saleh Al-Sulaiman ([email protected]) is team leader inspection & corrosion, north and west at Kuwait Oil Co., Ahmadi. He holds a mechanical engineering degree from Kuwait University. He is a member of NACE International.
Amer Jarragh ([email protected]) is senior corrosion engineer at Kuwait Oil Co. He holds a chemical engineering degree from University of Missouri, Columbia. He is a NACE cathodic protection specialist, senior corrosion technologist, and certified coating inspector Level-III, Peer-review. He is a member of NACE International.
Hasan Sabri ([email protected]) is corrosion specialist Kuwait Oil Co. He holds a chemical engineering degree from University of Toledo, Ohio. He is a NACE cathodic protection specialist, senior corrosion technologist, and certified coating inspector Level-III, Peer-review. He is a member of NACE International.
Renish Rahim ([email protected]) is cathodic protection engineer at Kuwait Oil Co. He holds an electrical engineering degree from Kerala University, India. He is a NACE cathodic protection specialist and member of NACE International.
More Oil & Gas Journal Current Issue Articles
More Oil & Gas Journal Archives Issue Articles
View Oil and Gas Articles on PennEnergy.com