Assessment of the Cathodic Protection of Crude Oil Storage Tanks in the Arabian Gulf Antonio Martinez Niembro Abu Dhabi Operating Company P.O. Box 303, Abu Dhabi, United Arab Emirates Fax: 971 2 6064442 E-mail:
[email protected]
Shaikha Mohamed K. Al-Zaabi Abu Dhabi Operating Company P.O. Box 303, Abu Dhabi, United Arab Emirates Fax: 971 2 6064442 E-mail:
[email protected]
ABSTRACT
This paper describes the diagnostic and troubleshooting methodology used to evaluate the effectiveness of impressed current cathodic protection on the underside of above ground storage tank floor plates. The case history includes the field data for several types of measurements performed to demonstrate if the current flow is in the proper direction and if the amount of current arriving at the floor plates is sufficient to offer protection. It is realized that in order to confirm effective cathodic protection, the voltage drop in the soil between the floor plates and the reference electrode must be considered. An innovative tank-to-soil potential monitoring method using permanent reference electrodes and potential monitoring conduits directly under the tank floor plates was implemented for one of the tanks to obtain accurate data and implement an effective monitoring programme.
The subject tanks (Crude Oil Storage tanks 7 and 11) of this paper are single shell, vertical, cylindrical, steel welded with double deck pontoon-floating roofs. The tanks were designed and fabricated to BP Standard 163 and BS 2654. Tank dimensions and capacity are 79.25m (260 ft) diameter, 19.51m (64 ft) height and nominal capacity 600,000 BBL. Both tanks were commissioned between 1967 and 1969. In 1990, both tanks were retrofitted wit h new floors, common, sketch and annular plates after jacking up the shells. The original Coal Tar Epoxy lining was replaced with Glass Fiber Reinforced Epoxy Epoxy (GFRP). Further actions to mitigate underside corrosion were implemented by installing a new berm seal around the periphery of the tank rim. The principal cause of the external corrosion on the bottom plates was from water leakage resulting from severe internal corrosion corrosion through to perforation. Internal failures were deemed to have happened due to ineffective internal CP. CP. However, it should be noted that that underside corrosion can happen even in the absence of leaks due to internal corrosion. Economic, environmental and safety considerations dictate the use of cathodic protection for corrosion control of the soil side of above ground storage tank bottoms. Certain types of cathodic protection (CP) designs may not be effective in providing adequate CP current to the tank bottoms in certain soil environments. On DAS Island, the electrical current for the tank bottoms, bottoms, provided by deep wells, may find lower resistance paths through buried piping or earthing systems. The use of bitumen sand precludes deep anode cat hodic protection systems to mitigate corrosion under the tank bottom plates. The oiled sand pads have an advantage only when no cathodic protection is installed, and some disadvantages are: − − − −
It traps moisture between the steel and the bitsand. It shields the CP current from getting to the steel. Reference cells do not measure the true potentials of the tank bottom. Moisture must be excluded permanently to be effective.
P. Barham et al, concluded that tank to soil potentials at the periphery of large diameter tanks have little correlation to actual potentials near the center of the tanks, however, perforated non-metallic pipe beneath the tank is an easy and useful method of measuring valid tank to soil potentials. They also installed angle-drilled deep groundbeds beneath the tanks to provide better current distribution and more uniform tank-to-soil potentials 4. Other technologies were tested to improve the effectiveness of corrosion protection methods for AGST with secondary containment of double-bottom storage tanks. The work of S.R. Rials and J.H. 5 Kiefer concluded that Vapor Phase Corrosion inhibitors are effective in controlling corrosion in moisture saturated conditions, further they recommended that installation of ICCP designs should not be used in the annular space of double-bottom storage tanks. M. Surkein and J. Collins proposed two different cathodic protection designs utilizing zinc ribbon anodes for small diameter tanks (35m) with double bottoms. The results showed that sacrificial anodes with high quality sand can be effective to achieve cathodic protection according to industry 6 accepted standards . M. A. Al-Arfaj describes a study of five tanks to measure the effectiveness of zinc ribbon cathodic protection to single and double bottom tanks using the 100mV depolarization 7 criterion . Crude oil operators in the Middle East are increasingly installing dedicated cathodic protection systems to mitigate AGST underside corrosion problems. That is the case of Abu Dhabi National Oil company (ADNOC) projects in Ruwais and Habshan (UAE). These included close anodes technology for new tanks with secondary containment HDPE liners under the tanks; BP-AMOCO in UAE installed in 2002, a dedicated MMO anode system for an 80 meter diameter tank. TotalFinaElf in Total South Pars Phase 2&3 Project installed the anode loops system under 23 crude oil and water treatment tanks in 1999. Also in 2002, a water transmission pipeline project in UAE used dedicated close anodes for the associated water storage tanks. R. L. Garret stated that monitoring cathodic protection levels of AGST bottoms is a challenge and concluded that the current actually picked up by the tank bottom is an important interpretation tool and also stated that oiled sand pads have an advantage only where no cathodic protection is installed
K. Kendell 11 in a review of failure mechanisms prevalent in the Middle East summarizes the various technical issues and offers a coherent selection methodology for protection of both new and existing tank bases. Kendell further describes complicating complicating factors which affect obtaining accurate measurements of cathodic protection, such as the liquid level in the tank must be sufficiently high to create intimate contact between the tank bottom and the pad because empty tanks often indicate erroneous test results. Sealing against moisture entrance has been a matter of numerous attempts to keep moisture free between the tank foundation and its bottom plates. F. Habiby et al, describe a way of sealing the gaps by applying a suitable sealant of two-component polysulfide in a circular groove around the periphery 12 of the tank rim . Many other attempts were developed to mitigate corrosion by water ingress under the tank; however none of them have proved to be a total successful. 1.3. Relevant Codes and Standards
API RP 651 in section 5.3.5 clearly says that: Clean sand is the most common material used as cushion beneath storage tank bottoms. API RP 651 further says that experience in the industry is varied as to the effectiveness of oiled sand for corrosion control. The presence of oiled sand has not been proven as an effective corrosion control measure and thus does not eliminate the need for cathodic protection and may cause CP to be less effective because of the higher resistivity of the oiled 13 sand . NACE RP0193 Section 5.6 says that: Clean, fine sand is the preferred tank pad material; and Section 5.7 says that on-grade tanks that are set on solid concrete or asphalt pad foundations generally require 14 specialized measures for corrosion protection because cathodic protection may be ineffective . BS 7361 in Section 4.5.1 says that: The top of the foundation mound may be provided with a bitumen-sand carpet or some form of coating may be applied to the underside of the tank bottom. During operation, the bottom of the t ank is subjected to flexure and settlement. The bitumen-sand is therefore liable to be damaged, thus allowing parts of the tank bottom to come into contact with the soil of the foundation mound. If the environment is corrosive, cathodic protection may be applied to
2. CASE STUDIES AND FIELD TESTS
The tanks were undergoing a major overhaul; hence all piping was disconnected from the t anks but remained electrically connected through CP bonding. In normal operation, no electrical isolation (isolation joints) exists between tanks and the associated pipelines. The earthing systems remained electrically connected to the tanks. These lines, tanks, and the associated structures are integrated to the two closest CP systems. The impressed current Pt/Ti anodes for these tanks are installed in deep wells located approximately 150 meters from the periphery of the tank and at a depth of 40 m. 2.1 Case Study Crude Oil Storage (COS) Tank 7
An “as found” potential survey was completed for COS Tank 7, including the associated underground pipelines and earthing systems. systems. The potential survey was conducted using using a portable copper/copper sulphate reference electrode, permanent silver/silver chloride (SSC) reference electrodes, and potential monitoring conduits installed under the tank. Current measurements at negative drain points and at earthing systems were also taken. 2.1.1. Potential Survey
As found potential measurements, detailed in Table 1 (“As Found” Column 1), were completed for both tanks. The potentials were measured with respect respect to a portable reference electrode and the reference electrodes installed under the bottom of COS Tank 7. As found potentials were measured during welding works to evaluate the influence of welding currents on the CP readings (data contained in Table 1). After interrupting the welding machinery, the potentials measured with the permanent reference cells did not have a significant change; which means that welding works most probably do not have a detrimental effect or corrosion on the soil side of the steel plates.
electrolyte can produce an important loss to the copper earthing rod, if that current is drained continuously to the surrounding soil. Such current may find a less resistance path through this earthing well, flowing to the piping network and finally coming back to the transformer rectifier. The calculated CP current directly collected by the floor plates through the soil is 1.4 amperes (2% of the total current delivered by the two closest rectifiers); such a low current cannot polarize the tank floor plates to a satisfactory level of cathodic protection (it is estimated that the tank will require around 50 – 60 A or more current to be effectively protected against corrosion). The low level of cathodic protection current picked up by the floor plates can be explained by different reasons, such as the shielding effect of the bitumen sand, low voltage drop per unit distance surrounding the tank floor plates, and/or high resistivity soils under t he foundation of the tank. 2.2 Case Study COS Tank 11
COS Tank 11 was also undergoing a major overhaul. This tank is surrounded by a High Density Polyethylene (HDPE) membrane from the berm to the bund for retaining any oil spillage inside the bund (no membrane is installed directly under the tank) and no permanent reference electrodes exist, so the analysis was done with the portable reference cell located in the relevant cathodic protection pits at N, S, E and W locations. These pits were installed in such a way that the reference cell contacted the soil directly below the HDPE membrane. 2.2.1. Potential Survey
As indicated previously for COS Tank 7, the same analysis is done for COS Tank 11. The “ON” potential readings at the periphery of the tank shown in Table 3 are not the true potentials, as they have to be corrected for the IR drop; hence, it cannot be concluded that the tank is being cathodically protected as it is shown with the DC current measurements. The fact that the tank, at the time of the evaluation, was empty can affect the distribution of current and potentials under the tank. Potential readings taken at earthing wells demonstrate that some wells (high negative potentials) are exposed to the voltage gradient of the t he deep well anode groundbeds, while others not (low negative
3. ANALYSIS OF RESULTS 3.1 Bitsand Shields the Protective Current and Traps Water
The bitsand, a mix of 86% sand, 8.5% Portland cement and 5.5% asphalt is a barrier for the cathodic protection current; hence any ingress of moisture between the steel plates and the bitsand would initiate a corrosion process that cannot be stopped. Water leaks from the process side, rain water during construction, condensed water from the humid environment, moisture entering during the filling/emptying cycles of the tanks and/or water ingress by the effect of capillary from the soil can be a real source of water ingress. Chemical analysis of the oily sand (50mm thickness) and the underlying foundation (500-900mm) in two tanks revealed very high content of chlorides 0.1-0.2% in the oily sand and 0.25-0.85% in the foundation. Sulphate content also showed very high concentration in both tanks, in the order of 1% in the sand and a wide range in the foundation between 1-8%. Ranges of pH contrary to expected were 8-9 in the foundation and 10 -11 in the bitumen sand. Seven locations per tank were analyzed. The material used in forming the pad under the floor can also cause underside corrosion of tank floor. It may contain chemical compounds that are corrosive to steel (i.e., sulphur compounds become very corrosive when moistened), the presence of clay will cause electrochemical corrosion resulting in pitting at each point of clay concentration. In view of the above results, the feasibility of an effective CP system with clean uncontaminated sand shall be evaluated in order to extend the life of new or refurbished tank floors. One important advantage of the clean sand pad foundation is that it dries out faster than bitsand and does not allow the water ingress to remain in contact with the steel, hence, less corrosion on the tank tank bottom. The sand also allows CP current to reach the bottom when it is wet, and does not shield, as happens with the bitsand.
3.3 Location of CP Groundbeds (boreholes)
The deep well ground beds or “boreholes” are located remote from from some tanks. This means that the current delivered by the groundbeds and distributed through the soil to the underground steel structures, including the tank floor plates, will not be able to reach the underside of the tanks because of shielding and relative high resistance compared to other structures. In both cases, the underground structures between the CP systems and the floor plates collect most of the protective current, minimizing distribution to the bottom plates. 3.4 The OFF Criteria For Protection Shall Be Applied
Cathodic protection criteria for underground steel structures shall consider the true polarized structure to environment measured at the structure surface for valid interpretation of the potential measurements. Voltage drops other than those across the structure-to-electrolyte boundary must be considered. Furthermore, in the case of COS tank floors with large diameter and due to the sealing system around the perimeter of the tanks, the space between the plates and the bitsand, most probably would have differing oxygen concentration generating differences in potentials, hence corrosion. The anaerobic conditions will require that the accepted protection criteria be –950mV (CSE) instead of –850mV free of IR drop. 3.5 Monitoring CP
The standard method for evaluating the effectiveness of CP is the tank-to-soil potential measurements with a reference electrode. For adequate representation of potential, the reference cell is to be placed as close as possible to the tank bottom. A problem associated with monitoring CP is the inability to place the reference cell under the tank for old tanks. For old tanks, most testing has relied upon readings at the perimeter of the tank, which may yield erroneous results because of potential gradients created in the soil by the groundbeds.
4. ECONOMIC CONSIDERATIONS 18
The method developed by R. A. King is to evaluate the relative cost of two or more corrosion control options taking into account the risk of failure and repair/refurbishment of the plant or equipment. In this case the two corrosion control options concern a crude oil storage tank and are: a) replace bitsand beneath new tank bottom b) install a distributed (mesh) CP system beneath the COS tank. The basis for the assessment is calculation of the differential costs between the two options. All common costs are ignored. Consequently the cost differential is the difference between: Option A = Cost of the corrosion control option as additional capital expenditure (CAPEX) + operation expenditure (OPEX) + Repair cost x % risks of failure; and Option B = Cost of the corrosion control option as additional CAPEX + OPEX + Repair cost x % risks of failure. For Option A, the cost of the corrosion control option is the cost of replacement of the bitsand and the marginal cost of the current derived derived from the existing CP systems. systems. The CP current from the existing existing groundbed systems is inadequate to prevent all corrosion because of the high resistivity of the bitsand layer. However this current drain to the tank represents a power cost. The repair costs are based on the percentage area of tank floor that will need to be replaced should a failure mechanism operate. For Option B, the cost of the corrosion control option is the cost of replacement of the contaminated bitsand with soft sand and the cost of installation of the dedicated CP system and its running costs. The repair costs are based on the percentage area of tank floor that will need to be replaced should a failure mechanism occur. The failure probability is the same as for Option A but the extent of underfloor corrosion is different because the plates are protected by the CP system. Estimates of the costs and time requirements have been input to test out the spreadsheet, see Table 5. It is interesting to note that Option B becomes cost attractive if the risk of failure of the internal lining leading to extensive underfloor corrosion corrosion exceeds 1 in 4 (~25%). (~25%). It is also obvious that after a second second
5. CONCLUSIONS
Cathodic protection current delivered by the deep wells do not provide the necessary protection for the floor plates of both case studies; this is probably due to preferential current paths to piping network, the existence of the shielding bitumen sand, long distance between the deep wells and the floor plates and/or high resistivity soil below the foundation of the tanks. Permanent reference cells were installed under one of the tanks; however the measured “ON” potentials do not represent the true potential of the floor plates in contact with the bit umen sand. Instant Off measurements will eliminate the IR drop, and future readings shall be performed using free voltage drop criteria. The risk based cost analysis approach concludes that the cost of the corrosion control option of replacement of the contaminated bitsand with soft sand and the installation of the dedicated CP system becomes cost attractive if the risk of failure of the internal lining leading to extensive underfloor corrosion exceeds 1 in 4 (~25%).
6. RECOMMENDATIONS
Evaluate the possibility of increasing the cathodic protection levels on the floor plates of the tanks using the existing CP systems, without affecting the associated piping and the earthing system. The measured potentials must be corrected to eliminate the IR drop error. Because cathodic protection may be ineffective due to bitumen sand shielding effect, evaluate other specialized measures for corrosion protection of the underside floor plates exposed to a corrosive environment, i.e. external coating of the floor plates and/or use of vapor phase corrosion inhibitors. Regular inspection and maintenance of the berm/annular sealant shall be ensured to limit the ingress
REFERENCES
1. J.G. Davis, Cathodic protection in Refineries, Chemical Pl ants and Storage Terminals, Materials Performance, Sep 1990. 2. D.H.Kroon, Tank Bottom cathodic protection with secondary containment, Paper 579, NA CE Corrosion 1991. 3. V. Brown, R.W. Stephens and R.D. Kygar, Use of MMO Anodes for Impressed Current Cathodic protection of AGST with PE leak detection membranes, Paper 384, NACE Corrosion 1992. 4. P. Barham, Ch. Bickford, G E. Mikish and R. W. Stephens. Improved Cathodic Protection Current Distribution Under Large Diameter Above Ground St orage Tanks using AngleDrilled Deep Anode Groundbeds, Paper 383, NACE Corrosion 1992. 5. S. R. Rials and J. H. Kiefer, Evaluation of Corrosion Prevention methods for Aboveground Storage Tank Bottoms, MP, January 1993. 6. M.B. Surkein and J. Collins, Evaluation of Cathodic Protection Systems Design for Tanks with double bottom utilizing galvanic anodes, Paper 368, NACE Corrosion 1995. 7. M. A. Al-Arfaj, The 100mV Depolarization Criteria for Zinc Ribbon anodes on externally coated tank bottoms, NACE, MP January 2002. 8. R.L. Garrett, The effect of different types of sand Pads on cathodic protection of AGST th bottoms, The 7 Middle East Corrosion Conference, Bahrain, 1996. 9. Saleh A. Al-Zubair, G.T. Al Nassir, A.A. Bukhamseen, Saudi Aramco Experiences with a non-conventional ICCP anode system to protect the exterior of a Storage Tank Bottom, The th 8 Middle East Corrosion Conference, Bahrain, 1998. 10. C. K. Meier and J. H. Fitzgerald, CP Monitoring, Installation, and leak Detection under
TABLE 1 - AS FOUND, AND “ON” POTENTIALS AT COS TANK 7
Tank
COS 7
Point 1 2 3 4 R1 R2 R3 R4 R5 R6 R7 R8 R9
Location North East South West
PERMANENT REFERENCE ELECTRODES
As Found (-mV SSC) 997 1112 1092 1104 999 1002 986 991 995 993 994 994 994
Welding OFF (-mV SSC)
997 1000 959 990 995 992 995 995 995
TABLE 2 - DC CURRENT AT CP BONDINGS AND EARTHING SYSTEM COS 7
DC Current at DC Current at CP bonding / Earthing CP bonding (A) Earthing Cable (A) Cathodic Protection Bondings (CPB) CPB-1 0.8 CPB-2 0.0 CPB-3 2.2 CPB-4 0.0
Direction of DC current
Rectifier Rectifier
TABLE 3 - AS FOUND POTENTIALS AT COS TANK 11
Tank
COS 11
Point 1 2 3 4 5 6 7 8
As Found (-mV CSE) 1092 1115 1053 991 1014 1087 1032 998
Location North East South West North-East South-East South-West North-West
TABLE 4 - DC CURRENT AT CP BONDINGS AND EARTHING SYSTEM COS 11
CP bonding / Earthing
CPB-1 CPB-2
EB-1 EB-2 EB-3
DC Current at CP bonding (A)
DC Current at Earthing Cable (A)
Cathodic Protection Bondings 0.0 1.0 Earthing Bars Cable 1 Cable 1 Cable 1
0.0 0.0 0.0
Direction of DC current
Rectifier
-
TABLE 5 - Risk Based Differential Cost Analysis
COST RELATED INPUTS
Storage tank diameter Cost of steel for f loor plates Thickness of floor plates installed Cost of corroded floor plate removal Cost of steel floor plate installation Thickness of bitsand/soft sand Cost of bitsand Cost of soft sand Cost of installation of bitsand/softsand Cost of internal lining (material + install.) Thickness of the annular plates Annular plate replacement by tank jacking CP current drain to tank
Units 80 m 500 $/tonne 7 mm 2 20 $/m 2
45 $/m
3
20 $/m
3
15 $/m
2
10 $/m
110 $/m2 12 mm 500 $/m
25 $/m2
Inspection interval A - Risk of failure of internal lining Area of floor replaced - no CP system Area of floor replaced - with mesh CP B - Risk of failure of wall seals % length of annular plate replaced - no CP % length of annular plate replaced - with CP Tank floor area Annular plate area/lined wall area
30 days 10 days 3
0.25 hours/m
2
0.25 hours/m
30 days 2 0.5 hours/m 30 days 30 days 2 0.1 hours/m
2
5 mA/m 5V
0.05 $/kWhr 10 years 33 % 80 % 5% 15 % 10 % 0% 2
5027 m 2 248.2 m
Assumes extensive underfloor corrosion subsequent to leakage Assumes only internally corroded plates require replacement Assumes annular plates are 1m width
Tank circumfere circumference nce % Area of annular plates to tank floor area
` 2
3 hours/m
2
12 V
Power costs for CP systems
Units 60 days 2 1 hours/m
1 mA/m
Installation cost of mesh CP system T/R voltage for mesh CP system
Time to mobilise floor plate contractor Time to remove corroded floor plates Time required to source bitsand Time required to source soft sand Time required to install bitsand/soft sand Time required to install floor plates Time required to mobilise lining contractor Time to install internal lining Time required to mobilise jacking contractor Time to remove and install annular plates Time required to source CP contractor Time required to install CP system
300 mm
Voltage of T/R supplying CP drain current CP current density for mesh system
TIME RELATED INPUTS
251.3 m2 4.938 %
TABLE 5 - Risk Based Differential Cost Analysis (Cont.)
OPTION A: replace bitsand and floor plates
Initial bitsand layer Initial storage tank floor Initial internal lining CP system installation cost Initial CAPEX for floor plates Initial CAPEX for annular plates OPEX CP drain from deep groundbeds FAILURE A: Bitsand and floor replacement FAILURE A: Internal lining replacement FAILURE B: Bitsand and annular replacement FAILURE B: Internal lining replacement Risk Based Differential Cost Analysis Option A
CAPEX for Corrosion Control OPEX for CP drain from groundbeds No failure Internal lining failure Wall seal failure Combined internal lining and seal failures Risk Based Cost of Option A Do ub ubl e Cycl e Cost of Op ti ti on on A Tank Downtime for Refurbishment Option A
No failure Internal lining failure Wall seal failure Combined internal lining and seal failures Risk Based Duration for Option A
$'000 80.4 464.0 580.2 0 1124.6 164.6 0.26 435.5 464.2 13.7
B: replace bitsand, floor plates, mesh CP Initial soft sand layer Initial storage tank floor Initial internal lining Mesh CP system installation cost Initial CAPEX for floor plates Initial CAPEX for annular plates OPEX of mesh CP system FAILURE A: Soft sand and floor replacement FAILURE A: Internal lining replacement FAILURE B: Bitsand and annular replacement
2.9
FAILURE B: Internal lining replacement
$'000 72.88 464.0 580.2 125.7 1242.7
0.55 26.8 29.0 0. 0 0.0
% ----47.05 33 15 4.95
$,000 80.4 0.26 0.0 296.9 2.5 45.4 425.43 850.86
Option B CAPEX for Corrosion Control OPEX of mesh CP system No failure Internal lining failure Wall seal failure Combined internal lining and seal failures Risk Based Cost of Option B Seco nd nd Cycl e Cost of Optio n B
% ----47.05 33 15 4.95
$,000 198.5 0.55 0.0 18.4 0.0 2.8 220.30 242. 04 04
% 47.05 33 15 4.95
Days 0 171 61 46 277
Option B No failure Internal lining failure Wall seal failure Combined internal lining and seal failures Risk Based Duration for Option B
% 47.05 33 15 4.95
Days 0 67 60 30 157
Note common costs are not considered. These include: tank drainage, desludging, degassing, tank inspection, routine maintenance of the CP systems, wall seal inspection. It is assumed that contaminated bitsand is removed and replaced with clean bitsand and that all materials requirements are sourced coincidentally.
