Cost of Corrosion

COST

OF CORROSION IN OIL

PRODUCTION AND REFINING

Robin Tems

Ahmed M. Al Zahrani

Robin Tems is the Downstream Corrosion Team Leader and an Engineering Consultant in Saudi Aramco’s Corrosion Technology Technology Unit, Materials Engineering and  Corrosion Control Division, Consulting Services Department, in Dhahran, Saudi Arabia. He holds a BS (Hons) degree in Chemistry and Metallurgy (1974) and a PhD in Corrosion (1978), both from the University of  Newcastle upon Tyne, England. He is licensed as a  profes  pro fessio sional nal memb member er of the Inst Institu itute te of Mat Materi erials als,, Minera Minerals, ls, and Mining (MIMMM), as a Chartered Engineer (CEng), and as a Chartered Scientist (CSci) in the United Kingdom. Before joining Saudi Aramco in 1992, he worked for Mobil  Research and Development Corporation, in Dallas, Texas, and Stavanger, Norway, focusing on corrosion and  metallurgy in diverse applications, ranging from sweet  North Sea fields to deep sour gas in the U.S. offshore. He has a long association with Saudi Aramco, serving as the Mobil team member to Shareholder Review Team No 1 beginning in 1982. Robin has over 30 years experience and  over 30 publications on corrosion in the oil, gas and   pollut  pol lution ion con contro troll indus industri tries. es. Ahmed M. Al Zahrani is a Downstream Corrosion Engineer in Saudi Aramco’s Corrosion Technology Unit, Materials Engineering and Corrosion Control Division, Consulting Services Department, Dhahran, Saudi Arabia. He holds a BS in Chemical Engineering from KFUPM, Dhahran, Saudi Arabia, and an MS in Materials Science & Engineering from Pennsylvania State University, University, U.S.A. Ahmed is a member of the Specialist Development 

2 SAUDI ARAMCO JOURNAL OF TECHNOLOGY  SUMMER 2006

Program/Downstream Corrosion Specialty. Ahmed joined  Saudi Aramco in 1995 and has worked for several  departments, including the Ras Tanura Refinery, Abqaiq Plants, Shedgum Gas Plant, Process & Control System Dept. and Consulting Services Dept. He is on a one year assignment with the Chevron Technology and Research Company. Ahmed has published more than eight papers at  international conferences and in technical magazines. He is a member of the National Association of Corrosion Engineers (NACE) International and the Electrochemical Society.

6 USA BCL-NBS

5

Kuwait 87

Turkey

USA NBS

Average 4.2% 4

   t   n   e   c 3   r   e    P

NACE-DOT + indirect

Egypt

UK Hoar Germany

USA NACE-DOT Kuwait 92

2 USSR

USA Uhlig

Japan

1 0 1940

1950

1960

1970

1980

1990

2000

2010

Year Fig. 1. Cost of corrosion, % GNP.

ABSTRACT

In 2003, Saudi Aramco initiated a study to define the cost of corrosion throughout core operations with the objective of focusing plant, engineering, and research investment in corrosion control to the areas that had the largest economic impact on corporate performance. This paper reviews available literature on cost of corrosion, reports the methodology adopted for the study, and highlights key conclusions from the study. Saudi Aramco corrosion costs were in line with industry experience. For Saudi Aramco’s five domestic refineries, 36% of maintenance budget was due to corrosion. For gas sweetening plants, it was found that 25% of the maintenance budget was committed to corrosion control. For gas fractionation plants, 17% of the maintenance budget was due to corrosion. For production operations onshore, corrosion is responsible for 28% of  maintenance costs. Offshore, corrosion accounts for 60% to 70% of maintenance costs. For plants or systems that were capacity limited, the biggest single corrosion cost impact on the total cost of operations is deferred production. The total cost of corrosion including deferred production costs for gas fractionation plants was estimated to be five times direct corrosion maintenance costs. The total cost of  corrosion including deferred production costs for refineries ranged up to three times direct costs. In a gas sweetening plant, indirect costs including deferred production costs were found to be 50 times the direct cost of corrosion for the specific case reported. INTRODUCTION

While it has long been recognized that corrosion can have a serious impact on plant operation and safety, there has been little understanding of the exact cost impact on a company’s operations. Without such knowledge of the true cost impact of corrosion, it is hard to evaluate new technologies and corrosion control initiatives that must be considered on a sound economic basis. Similarly, increases in corrosion control manpower and training can only be justified when an economic analysis confirms the need. In 2003, Saudi

Aramco initiated a study to define the cost of corrosion throughout core operations with the objective of focusing plant, engineering, and research investment in corrosion control to the areas that had the largest economic impact on corporate performance. The methodology adapted in this study included a detailed evaluation of published literature, coupled with a survey of all key operations. REVIEW OF PUBLISHED STUDIES

• The average value of the cost of corrosion is 4.2% of the gross national product (GNP) of a country, based on 12 national surveys (Fig. 1). • The most recent USA study estimated the cost at 6% of  the GNP. • The range of data is 1.8 to 6%, or approximately: Average Value ± 50%. Considerable uncertainty exists in the collection and analysis of cost of corrosion data. Two studies of the same industry performed concurrently gave results that differed by 300%, which again is: Average Value ± 50%. • Lower corrosion costs do not invariably mean that corrosion control has been optimized. Rather, it can indicate an inability to track costs or, in the worst case, insufficient corrosion control spending that leads to later disasters with large economic consequences and loss of life. • The cost of corrosion in onshore oil production ranges 0.7   e    d0.6   u   r   c    f 0.5   o    l   e   r 0.4   r   a    b   r 0.3   e   p    ) 0.2    3    0    0    2 0.1    (    $

0 USA, 1978

USA, 1998

Kuwait, 1987

Arabian Gulf, 1998

Europe, offshore

Fig. 2. Comparison of upstream oil production costs.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY  SUMMER 2006 3

1.4    f 1.2   o    l   d   e  e 1   r  s   r  s   a  e    b  c 0.8   r  o   r   e  p   p 0.6    )   e    d    3    0  u 0.4    0  r    2  c    (    $

0.2 0 USA, 1978

USA, 1998

Kuwait, 1987**

Turkey

Major Operator

Fig. 3. Comparison of downstream refining costs.

from $0.03 to $0.62/bbl (Fig. 2). Locations reporting the lowest costs experienced later corrosion-caused disasters that are not reflected in the estimate. • Offshore production costs including capital expenditures are estimated to be $0.45/bbl by one operator. Another study attributed 60% of offshore maintenance costs to corrosion. • The annual cost of corrosion in the U.S. pipeline industry is $9,860 per kilometer. Offshore, 41% of major leaks in the USA were due to corrosion. Onshore, 20% of major leaks were due to corrosion. A United Kingdom study of pipeline leaks found that 19% were due to external corrosion, 24% were due to internal corrosion or stress corrosion cracking, and 7% were due to materials deficiency. • Based on national and company surveys, the cost of  corrosion in refining ranges from $0.15/bbl to $1.34/bbl with an average value of $0.89/bbl (Fig. 3). Locations reporting the lowest costs experienced later corrosioncaused disasters that are not reflected in the estimate. One company found that 56% of leaks were due to corrosion. A NACE Task Group effort estimated the cost of  corrosion in refining to range between $0.16/bbl and $1.96/bbl, with a mid range cost of $0.85/bbl. • The net avoidable costs of corrosion are generally estimated to be between 15% to 25% of the cost of corrosion. • Based on experience from the automobile industry, investment in corrosion technology extended usable equipment life by 50% and reduced corrosion control costs by 27%. • For new construction, corrosion resistant alloys (CRAs) can represent the largest investment. In a study of two offshore fields, CRAs accounted for 54% to 82% of the capital expenditures attributed to corrosion control. Capital investment in corrosion control significantly reduces operating and maintenance costs. In recent years, there have been several national “Cost of Corrosion” studies covering a range of countries. Additional papers have dealt with specific industries or projects such as the development of offshore fields 1-14. Figure 1 presents a summary of data from national 4 SAUD I ARAMCO JOURN AL OF TECHNOLOGY 

SUMME R 2006

studies. The average value of the cost of corrosion is 4.2% of the GNP of a country, based on twelve national surveys. This average is a simple arithmetic average and does not consider the size of each national economy. The most recent USA study estimated the total cost of corrosion to be 6% of  the GNP. The range of data is 1.8% to 6%, or approximately the average value ± 50%. Considerable uncertainty exists in the collection and analysis of cost of corrosion data. Two studies8,13 of the same industry performed concurrently gave results that differed by 300% or by the average value ± 50%. The uncertainty in the data arises for several reasons. First, not all corrosion-caused failures are correctly identified. Often, corrosion is only identified and recorded as a cause when repetitive failures occur and an in-depth analysis is performed. Second, there are two components to the cost of  corrosion, the direct cost and the indirect cost. The direct cost of corrosion includes the annual expenditures spent on maintenance and repairs, plus corrosion control techniques such as coatings, linings, chemical treatment, cathodic protection, high alloys, and non-metallic materials. Inspection and testing costs are also part of the cost of corrosion. Indirect costs result from planned or unplanned plant shutdowns, excess capacity, redundant equipment, loss of  product, loss of equipment efficiency due to corrosion, contamination of product, and over-design. Indirect costs due to “lost or deferred production” can reach many times the cost of replacing or repairing corroded equipment but are often difficult to estimate. Methodologies used in cost of corrosion surveys have varied significantly. Early studies focused on direct costs and failed to account for indirect costs. Further, early studies tended to survey only corrosion control providers (e.g., coatings suppliers, chemical suppliers, etc.) and therefore did not include the significant costs of corrosion management in end-user companies. Hence, more recent cost of corrosion surveys estimate the total cost to be much higher than previously estimated. Many recent studies use the Battelle Columbus Laboratories/National Bureau of Standards (BCL/NBS) input-output economic model. The model divides a national economy into 130 sectors and is able to evaluate the interaction within and between sectors by using fractional multiples or coefficients for each part of a business sector. Many published papers include the coefficients for each sector in their results. From these coefficients, it is a relatively straightforward matter to back calculate corrosion costs per unit volume using published production and refining capacity data for each country for the period in question. In the oil and gas production industry, corrosion costs

are generally expected to increase in the future due to development of deep wells producing high temperature, high-pressure, aggressive environments. At the same time, conventional fields are aging and producing much higher water-cut percentages, with the result that corrosion increases. In the worldwide refining community, the cost of  corrosion is increasing due to the purchase of “opportunity crudes” on the spot market, resulting in unpredictable changes in the corrosivity of refinery production streams. United States

Uhlig, 1949. The first cost of corrosion study was carried out in the United States by Uhlig in 1949. The study estimated the annual cost of corrosion in the USA to be $5.5 billion, which represented 2.1% of the GNP. The study surveyed the cost of corrosion control products sold which Uhlig defined as direct costs and also included a component to account for the effect of corrosion on non-owners. BCL/NBS, 1978. The benchmark study was conducted in 1978 by Battelle Columbus Laboratories (BCL) for the National Bureau of Standards (NBS). The estimated annual cost of corrosion was $70 billion, which represented 4.2% of the USA GNP in 1975. The analysis indicated that 15% of the cost of corrosion was avoidable if effective corrosion prevention technologies were applied. The error in the estimation of direct cost of corrosion was about ± 30%. The error in indirect costs was speculated to be much higher but was not reported. Significant error was expected in life cycle costs due to the uncertainties of life prediction. A more sophisticated methodology was employed for the 1978 BCL/NBS study than that used by Uhlig. An inputoutput economic model was adapted for the study. The model divides the U.S. economy into 130 sectors and is able to evaluate the interaction within and between sectors by using fractional multiples or coefficients for each part of a business sector. The model also allows easy generation of  “alternative worlds” where there is no corrosion or where best corrosion control practices are employed. While this methodology is preferred for the determination of costs in a national economy over multiple industry segments, the NBS report states that for limited segment studies, a simple summation of costs taken over all components of  equipment, production and other similar factors can be sufficient. For such studies, the data are collected by interview and literature survey. Evaluation of the coefficients reported for the refining and upstream sectors combined with published data on production rates allows calculation of per barrel corrosion costs. Upstream corrosion costs were $0.21/bbl (1975 $) which is equivalent to $0.62 (2003 $). Refining costs were $0.44/bbl (1975 $) which is equivalent to $1.30 (2003 $).

NBS 1995. The 1995 study was based on the data obtained for the original BCL/NBS 1978 study, adjusted for changes in the economy. Department of Transportation (DOT) 1996. In an analysis of major pipeline incidents in the USA between 1985 and 1994, corrosion was the leading cause of offshore pipeline failures, responsible for 41% of leaks. Onshore, corrosion was the second leading cause being responsible for 20% of leaks. DOT tracks pipeline leaks causing fatalities, injuries, or damage over $50,000. NACE-DOT, 1998. The most recent cost of corrosion study was carried out in the period 1999 to 2001 for the base year of 1998. The study (NACE-DOT) was conducted by CC Technologies on behalf of NACE International and the U.S. Department of Transportion. The direct cost of  corrosion was estimated to be $276 billion, which represented 3.1% of the GNP in 1998. The overall indirect cost was estimated to be at least equal to the direct cost of  corrosion. For the specific industry sub-sector of highway bridges, the indirect cost of corrosion was estimated to be 10 times the direct cost of corrosion. NACE-DOT reported the total cost of corrosion including indirect costs to be 6%. The NACE-DOT study evaluated industrial sectors comprising 27.55% of the U.S. GNP. For these sectors, direct corrosion costs of $137.9 billion were identified. Costs were estimated for the remaining 72.45% of the U.S. economy using a non-linear extrapolation to give the total direct cost estimate of $276 billion. The non-linear extrapolation considers the extremely limited potential for serious corrosion impact in businesses such as miscellaneous services, wholesale and retail trade, and financial sectors. The cost of corrosion estimation method used to obtain the total direct cost of corrosion was the BCL/NBS inputoutput model. Sub-studies used a variety of analytical models. The annual direct cost of corrosion control methods and services was estimated in a sub-study that took basic industry and government sales data, added assumed values for items such as services, and extrapolated. These data are presented in Fig. 2. The cost was estimated to be $121 billion. Based on this estimate, organic coatings accounted for the bulk of corrosion control costs, 88%. The cost of corrosion in the U.S. pipeline industry transporting gas, oil, and chemicals was estimated to be $9,000 per kilometer (1998 $) or $9,860 (2003 $). Of this, 52% was spent on maintenance and 38% on replacement of  corroded aging pipelines. Ten percent of the annual cost was attributed to the cost of property damage or injury, etc., resulting from corrosion failures. A survey of three transmission companies estimated that 15% of operating and maintenance costs were due to corrosion prevention measures such as cathodic protection, coatings, and inspection.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY SUMMER 2006 5

Education & training Research & development

Monitoring 5%

Services

Corrosion Inhibitor 68%

Anodic & cathodic protection

Non-metallics

Inspection 17%

Corrosion inhibitors Metallic coatings Metals & alloys Organic coatings

0

20

40

60

80

100

Fig. 4. Distribution of spending on corrosion/control technologies, U.S.A., 1998, percent.

For upstream, onshore facilities, the cost of corrosion is $0.22 (2003 $) per barrel of oil produced. This is based on data from two field studies by the same operator. Distribution of costs for each field is shown in Fig. 3 and Fig. 4. Downhole repair costs due to corrosion were estimated to be an additional $0.16/bbl resulting in a total onshore cost of $0.38/bbl. Capital costs attributable to corrosion resulted in the total cost of $0.50/bbl (2003 $). The study did not include a systematic evaluation of the cost of offshore production. Refinery corrosion costs in the U.S. were estimated from extremely limited data. Operation and maintenance costs due to corrosion were reported to be $0.34/bbl (2003 $). Turnaround costs were estimated at $0.26/bbl. Total corrosion cost was $0.59/bbl. Costs for crude unit corrosion control ranged from $0.01 to $0.12/bbl and was dominated by inhibitor costs. Inhibitor costs were estimated to average $0.05/bbl across the refining industry. EPRI, 1998: Reliability of Data. The Electric Power Research Institute (EPRI) initiated a study of the cost of  corrosion in the electric power industry concurrent with the NACE-DOT national survey. EPRI estimated the cost of  corrosion in the electric power industry to be $17.3 billion or 0.20% of the U.S. GNP which compares with 0.24% found by the BCL/NBS study. The NACE-DOT study estimated $6.9 billion for the same sector. EPRI report 1004662 claims these differences are due to ignoring indirect costs in the NACE-DOT study. The text of the NACE-DOT report claims to include indirect costs for the electric power industry. The EPRI reported cost of corrosion represents 300% of the NACE-DOT reported value, indicating considerable uncertainty in such data. NACE Task Group STG 34. This NACE Task Group performed a worldwide follow-up to the NACE-DOT study referenced above. The evaluation estimates the cost of  corrosion in U.S. refineries to total approximately $3.7 billion. The cost per barrel was estimated to range between $0.16/bbl and $1.96/bbl with a mid point value of 

6 SAUDI ARAMCO JOURNAL OF TECHNOLOGY  SUMMER 2006

Engineering Staff 7%

Corrosion Repairs 3%

Fig. 5. Cost breakdown for a producing field in U.S. — $0.22/bbl.

Corrosion Inhibitor 54%

Inspection, Monitoring & Staff 41%

Corrosion Repairs 5%

Fig. 6. Breakdown of corrosion costs in a U.S. producing field — $0.22/bbl.

$0.85/bbl. Maintenance costs due to corrosion were estimated to range between $0.15/bbl and $1.75/bbl. Losses in operating capacity were estimated to range between $0.01/bbl and $0.21/bbl. As part of the NACE Task Group’s work, presentations were made by various companies. One company established a procedure to document the cause and cost of leaks in refinery and chemical plant operations through a detailed investigation of every leak. Fifty-six percent of leaks were due to corrosion. External corrosion leaks accounted for 19% of leaks and were due to either corrosion under insulation or stress corrosion cracking. The average leak cost for all facilities studied was reported as $2.40/bbl, which gives a cost of leaks due to corrosion of $1.34/bbl. This does not include any allowance for ongoing operations, maintenance, inspection, or capital costs due to corrosion that are not directly associated with a leak.

Research & development Anodic & cathodic protection

Corrosion inhibitors & oils Metals & alloys Surface treatments Organic coatings 0

10

20

30

40

50

60

70

Fig. 7. Distribution of spending on corrosion control technologies, Japan, 1977, percent.

United Kingdom

A cost of corrosion study was performed in the United Kingdom (UK) in 1970 by T. P. Hoar. Hoar employed a methodology that surveyed principal users, suppliers, and specialists. The annual estimated cost was 3.5% of the UK GNP. The study reported that hydrocarbon and petrochemical corrosion costs could be reduced by 8% through improved corrosion engineering, though specific oil and gas companies surveyed reported that corrosion costs had been reduced by up to 50% by the establishment of corrosion control groups. The purchase of coating materials was approximately equal to alloy purchases. UK pipeline leaks have been analyzed by Advantica for the period 1961-200015. Internal and external corrosion was responsible for 43% of the leaks. Of failures originating from the process-side, the great majority (95%) were due to stress corrosion cracking caused by wet town gas which is no longer transported in the system. Another 7% of leaks were due to inadequate materials. Japan

A Japanese study using the Uhlig methodology was completed in 1977 using 1974 as the base year. This methodology ignores direct corrosion management costs and all indirect costs. It therefore results in low estimates. The survey estimated that the cost of corrosion was 1.8% of the Japanese GNP. The distribution of corrosion control costs is shown in Fig. 7. The major portion of the expenditures was spent on paints and coatings and surface treatment, which was approximately 88% of the total. Kuwait

The cost of corrosion estimation for the State of Kuwait was carried out for the base year of 1987 using the BCL/NBS input/output (I/O) analysis model. The total cost of corrosion was 5.2% of the Kuwait GNP, 18% of which was deemed avoidable.

Evaluation of the coefficients reported for the refining and upstream sectors combined with data on production rates allows calculation of per barrel corrosion costs. Upstream corrosion direct costs were $0.11/bbl (1987 $) which is equivalent to $0.16 (2003 $). Refining direct and indirect costs were $0.10/bbl (1987 $) which is equivalent to $0.15 (2003 $). The 1992 study extrapolated data from the 1987 study to account for effects of the Gulf War. It is noteworthy that the corrosion control costs in the refining sector pre-date major disasters at the Mina AlAhmadi Refinery in June 2000. The Mina Al-Ahmadi fire resulted in six dead and over 50% of the 480,000 bbl refinery destroyed. Both the option to entirely demolish the refinery and to attempt repair were evaluated. Estimates of  the cost to repair the Mina Al-Ahmadi Refinery ranged from $412 million16 to $4 billion. The fire occurred due to a vapor cloud explosion caused by a corrosion leak in a 10 inch condensate line. Other studies of the cost of corrosion in upstream operations have estimated costs to be $0.03/bbl in 1998 rising to $0.44/bbl by 2010 (2003 $). These estimates precede some major corrosion-caused incidents in the upstream sector. Turkey

A 1985 study estimated the cost of corrosion to be 5.1%. The study used the BCL-NBS input output model which allows extraction of costs for individual sectors, when supporting data are available. On this basis, the cost of  corrosion in refining was found to be approximately $0.20/bbl (1985 $) or $0.31 (2003 $). Offshore Costs

A study by AGIP reports the capital costs of new projects and the distribution of capital expense on corrosion control technologies. In the study of two offshore fields, one in the North Sea, one in the Mediterranean, found that between 54% to 82% of the capital cost was expended on corrosionresistant alloys. Coatings represented only a small portion, between 5% to 11%, of the corrosion capital expense. These data are presented in Fig. 8. The overall cost of offshore production, including capital and ongoing operation and maintenance expense, was estimated to be $0.45/bbl (2003 $). A Norwegian study estimated that 60% of offshore maintenance work was due to corrosion. AGIP estimates the total cost of coating a new platform of medium size for the Adriatic Sea (114,000 square feet of  deck painting; 22,000 square feet of structural painting) to be $393,000 (1993 $) or $473,000 (2003 $). Maintenance painting costs were $140,000 (1993 $) or $169,000 (2003 $). Splash zone coating cost $3/ft2, areas above the splash zone cost $2.70/ft2, and helidecks cost $4.8/ft 2. SAUDI ARAMCO JOURNAL OF TECHNOLOGY  SUMMER 2006 7

9

Inhibition equipment

  e 8    l   c    i    h   e    V 7   r   e   g   n 6   e   s   s   a    P 5    f   o   e   g    A 4

Corrosion allowance Coatings Cathodic protection Corrosion resistant alloys 0

20

40

60

80

100

Mean Median

3 1970

1975

1980

Fig. 8. Distribution of capital spending on corrosion control technologies, offshore oil fields.

1985

1990

1995

Year

Fig. 9. Average life of automobiles, USA8.

Savings Possible

Avoidable costs of corrosion. A study by Cherry of the cost of corrosion in Australia, supported by comparison with data from other countries, found that the overall avoidable cost of corrosion ranged between 0.8% to 1.5% of the GNP or in the region of 25% of the cost of corrosion. The range of savings estimated by the BCL/NBS study was 15%. NACE-DOT estimated avoidable corrosion costs to be 50%. The cost of corrosion prevention was estimated to be half of the potential savings. Thus, the net avoidable corrosion cost was 25%. The range of possible savings reported by various workers in oil and gas production and refining ranges from 8% to 75%. Benefits of increased investment in corrosion engineering. The benefits of increased investment in corrosion engineering are demonstrated by an evaluation of data from the U.S. automobile industry 8. The automobile industry has become one of the major success stories in corrosion engineering management over the past 25 years. Before 1950, corrosion of automobiles in the USA was only an issue in coastal areas. The increasing use of deicing salts combined with increased industrialization and acid rain resulted in a major reduction in automobile lives because of  corrosion. In the late 1970s, automobile manufacturers started to increase the corrosion resistance of vehicles by using corrosion-resistant materials, employing better manufacturing processes, and designing more corrosionresistant vehicles through corrosion engineering knowledge. Because of the steps taken by manufacturers, today’s automobile has very little visible corrosion and most vehicles survive structurally until a vehicle wears out mechanically. The decrease in cost, as a percentage of the GDP in the USA, indicates the success of the automobile industry in controlling both corrosion and the cost of preventing corrosion. The percentage of GDP due to motor vehicle corrosion has decreased from 0.37% in 1975 to 0.27% in 1998, which is a reduction in the cost of corrosion of 27%. 8 SAUDI ARAMCO JOURNAL OF TECHNOLOGY  SUMMER 2006

Figure 9 shows the increasing life of automobiles due to improved corrosion performance. This increasing life is also recognized by comparing the warranties offered against perforation. American Motors offered the first 3-year warranty against perforation in the U.S. market in 1977. Today, almost every manufacturer offers at least a 5-year warranty. One manufacturer even offers a 12-year warranty. These improvements in lifetime required an initial investment in technology by automobile manufacturers and steel manufacturers. Improvements were made in design and in materials selection. One of the large costs was a change to using two-sided galvanized steel. Steel manufacturers were not tooled to provide this product, so the cost of  providing the required material was relatively high. This is reflected in Fig. 10 where the cost of corrosion control increased between 1975 and 1991 while the lifetime of cars also steadily increased. Since 1991, the percentage cost of  corrosion control has fallen dramatically as steel manufacturers converted plants to provide the required product at a competitive price. In summary, between 1975 and 1998, the automobile industry increased the life of its product by 50% and reduced the cost of corrosion control by 27%. These 4

  n   o 3.5    i    t   c   e 3    t   o   r   %    P  , 2.5   e   n   l   c   o   i    i   s   h 2   o   e   r   r   v 1.5   o   r    C   e    f   p 1   o    t   s   o 0.5    C 0 1970

1975

1980

1985

1990

Year

Fig. 10. Cost of corrosion in automobile manufacture, USA8.

1995

2000

results were only achieved through investment in corrosion control and working with suppliers to provide the required materials.

% of cost attributable to corrosion

METHODOLOGY

In addition to the extensive literature analysis, an extensive plant and field survey was performed covering all company operations. Cost of corrosion data were gathered from plants using an MS Excel™ based survey form. Plants were encouraged to form their own local teams to complete the survey form. Such teams usually included representatives from engineering, maintenance and operations, as well as administration personnel familiar with budgetary operations. Multiple site visits were made to each location to introduce the idea of the survey, to follow-up on questions concerning completion of the survey, and to investigate areas of uncertainty in the submitted responses to the survey. Some plants included a detailed evaluation of every work order in their response to the survey form. Others included reviews of specific equipment files and turnaround and inspection (T and I) reports. Personal interviews were performed with key long-service personnel to provide historical data. Loss Prevention databases, central engineering databases, corporate data sources, and vendor data were reviewed to correlate significant cost items with field data. To focus on collecting the most important data within the required timeframe, the scope of the survey was limited to corrosion in existing plants. The survey did not include investment in corrosion control in new capital developments. Costs in new construction projects include factors such as corrosion-resistant alloys (CRAs) and nonmetallic materials, corrosion allowances, increase in structural dimensions to accommodate weight of cathodic protection systems, purchase of HIC-resistant steel, stress relief operations, installation of platforms for inspection and corrosion monitoring, corrosion monitoring systems, cathodic protection, and coatings systems. The economic analysis method used was the additive methodology used by T. P. Hoar in the UK survey9. Where available, deferred and lost production costs were included in the analysis. Cost data were collected and reported for each year over a five year period, or for a shorter period if  only limited data were available. All costs were adjusted to a base year of 2003, using the U.S. gross domestic product deflator values.

2001

2000

1999

1998

Maintenance NDE

70

60

60

65

Operational NDE

20

19

21

21

Capital

30

5

5

5

Engineering NDE

11

11

6

6

NDE = net direct expense Table 1. Offshore upstream operations — percentages of costs attributable to corrosion.

% of cost attributable to corrosion

2001

2000

Maintenance NDE

26

31

Operational NDE

23

21

Capital

1

1

Engineering NDE

16

13

NDE = net direct expense Table 2. Onshore upstream operations — percentages of costs attributable to corrosion.

generally very low water cuts result in some of the lowest corrosion control costs in the industry. Due to the dynamic nature of upstream production, these costs will increase as fields age. Table 1 and Table 2 present percentages of costs attributed to corrosion for offshore and onshore oil operations. As would be expected, offshore operations have 60% to 70% of their maintenance effort dedicated to corrosion issues. This figure agrees with the Norwegian data12. Offshore corrosion concerns focus on corrosion measurement and management in subsea pipelines, offshore produced water injection, utility seawater handling, and platform maintenance including coatings and the expanded use of non-metallics.

Coatings 32.7% Cooling Water 0.1%

Cathodic Protection 26.1%

Monitoring 8.5%

Chemicals 32.7%

UPSTREAM PRODUCTION

Production in Saudi Arabia is different from most other operating areas of the world in that economies of scale and

Fig. 11. Percentage of direct cost for specific corrosion control technologies, onshore production, 2000-2001.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY  SUMMER 2006 9

Cathodic Protection Coating 70%

Coating Chemicals

Monitoring Cooling Water

Chemicals 22%

Monitoring 5%

Cathodic Protection 3%

Fig. 12. Percentage of direct cost for specific corrosion control technologies, offshore production, 1998-2001.

Onshore oil operations report approximately half as much maintenance effort being expended because of  corrosion—about 29%. Specific sectors of onshore oil operations, notably those handling seawater injection, report approximately 60% of their maintenance budget is due to corrosion. Onshore and offshore operational costs attributed to corrosion are about the same, in the region of 21%. Offshore engineering costs jumped 5% for years 2000 and 2001 due among other factors to the inclusion of in-line inspection (ILI) costs in the budget for those years, though the total Engineering NDE budget is reported as essentially constant during this time. Figures 11 and 12 show the distribution of direct cost of  corrosion control technologies used in the upstream oil operations sector. Coatings dominate the offshore sector. GAS SWEETENING PLANTS 1

Gas sweetening plants receive associated gas from oil production, non-associated gas from gas wells, and gas condensate from gas oil separation plants (GOSPs). These plants remove acid gases (CO2 and H2S) from the gas then compress and chill the sweetened gas. The gaseous fraction (mostly methane) is injected into the Kingdom’s sales gas grid. The liquid fraction (C2+) is sent to the gas fractionation plants for further processing. For gas sweetening plants, it was found that 25% of the maintenance budget was committed to corrosion control. For plants that were capacity-limited, deferred production represented the most significant cost. For a gas sweetening plant, a day of deferred production in one unit at one plant would be on the order of 47% of the annual direct cost of  corrosion for that unit. A specific case was reported for one sweetening plant where the total indirect costs of corrosion 10 SAUDI ARAMCO JOURNAL OF TECHNOLOGY  SUMMER 2006

Fig. 13. Percentage of direct cost for specific corrosion control technologies,  gas sweetening plants, 1999-2001.

including deferred production were 50 times the direct cost of corrosion19. Common failure mechanisms in amine units included velocity enhanced damage due to operation at throughputs in excess of the original design, selective weld grooving, hydrogen induced cracking (HIC), and on rare occasions, stress corrosion cracking (SCC). Corrosion of rich/lean exchangers was common. Corrosion in reboiler circuits in stripping columns occurred frequently. All plants experienced corrosion under insulation, concrete degradation, and utility system corrosion, including firewater and steam generating systems. The average percentages of direct costs attributable to corrosion are shown in Table 3. One plant performed a thorough analysis of work orders for two time periods. For the periods studied, 49% of maintenance work orders were found to be due to corrosion. Plants that did not perform such a detailed analysis estimated that only 14% of work orders were due to corrosion. Year-to-year variations in the capital cost of corrosion occurred due to varying degrees of upgrade that were required at each plant to repair corrosion damage. The distribution of costs for various corrosion control techniques are presented in Fig. 13. Chemical costs represent 43% of these costs over the three-year period. Coatings costs represent 23%. Chemicals are used in control of boiler water treatment programs and for inhibition of gas wells that are % of cost attributable to corrosion

2001

2000

1999

Maintenance NDE

24

27

23

Operational NDE

9

7

7

Capital

27

7

29

Engineering NDE

22

22

20

NDE = net direct expense Table 3. Gas sweetening plants — average percentages of costs attributable to corrosion.

Cathodic Protection

Coating

Chemicals

Cooling Water Monitoring

Fig. 14. Percentage of direct cost for specific corrosion control technologies,  gas fractionation plants, 1997-2001.

one source of gas to the plants. Firewater system degradation has been problematic at all plants. Several systems have been replaced using nonmetallic materials for buried sections of pipework. Boiler failures due to chelate corrosion were a major problem in the past. Corrosion under insulation is reported by all plants. Hydrogen induced cracking of acid gas flare headers is a common issue. Corrosion deposit accumulation in flare headers was identified as a major problem at one plant resulting in an unplanned shutdown to effect repairs. GAS FRACTIONATION PLANTS 1

Gas fractionation plants receive liquid hydrocarbons from the gas sweetening plants and oil processing facilities. The liquids proceed through a series of fractionation modules to refine the products. Products are ethane, liquid propane, butane, pentane and hexane. Most of these products are consumed as petrochemical feed stock in-Kingdom. Excess ethane not required for petrochemical feedstock is injected into the Kingdom’s sales gas grid. This sector is one of the most thoroughly documented of  all sectors surveyed. Local plant investigators evaluated the causes of work orders for all or parts of the periods studied.

% of cost attributable 2001 2000 1999 1998 1997 to corrosion

Maintenance NDE

22

17

16

19

13

Operational NDE

4

4

4

4

3

Capital

0

0

0

5

0

Engineering NDE

5

5

5

5

5

NDE = net direct expense

These gas fractionation plants may be considered to be mildly corrosive as the processed fluids have been pretreated in other plants. Even so, over five years, 17% of  maintenance was due to corrosion. This includes maintenance hours spent on T and I’s. Table 4 presents a breakdown of the percentage of budget attributable to corrosion. Direct cost components are shown in Fig. 14. As would be expected for these types of plants, the significant corrosion expenditures are on corrosion monitoring, with coatings and chemical treatment being next most significant at 19% each, averaged over the five-year study period. For the last two years, coatings costs have become more important and represent approximately 30% of the expenditures. Indirect costs from lost production in this sector include deferred production costs from plant downtime for T and I’s, and deferred production costs due to corrosion and materials failures. The total cost of corrosion including indirect costs was approximately five times the value attributed to direct costs. Environmental concerns have also been raised when a sulfur plant was unavailable and the gas was routed to flare. Environmental cracking has resulted in major expenditures at all the plants. Caustic stress corrosion cracking has occurred in Merox units or immediately downstream of these units when unintentional carryover of  caustic occurred reaching non-stress relieved pipe. Solutions include replacement of pipe systems with stress relieved pipe and changes in equipment opening procedures from steamout to water washing. Liquefied gas streams may additionally contain water and hydrogen sulfide meaning that some systems must be designed for sour service. One plant required replacement of part of a de-ethanizer tower on two occasions due to hydrogen induced cracking (HIC). A reflux drum in the same system also had to be replaced due to HIC. The system was not designed to experience wet sour conditions because it was protected by a desiccant bed. Over time the desiccant became saturated and therefore ineffective.

% of cost attributable to corrosion 2001

2000

1999

Maintenance NDE

36

33

40

Operational NDE

32

12

6

Capital

0

3

4

Engineering & Inspection NDE

14

11

14

NDE = net direct expense Table 4. Gas fractionation plants — average percentages of costs attributable to corrosion.

Table 5. Refineries — Percentages of costs attributable to corrosion.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY  SUMMER 2006 11

1.50 A

B

C

D

Cathodic Protection

E

1.25    l   e   r   r 1.00   a    b   r   e0.75   p    t   s   o0.50    C

Chemicals Coating

0.25

Cooling Water

0.00 1997

1998

1999

2000

2001

Monitoring

Fig. 15. Direct cost of corrosion, each refinery. Fig. 17. Percentage of direct cost for specific corrosion control technologies, refineries.

$20.00

REFINERIES $15.00

$10.00

$5.00

$0.00 1995

1996

1997

1998

1999

2000

2001

Fig. 16. Variation in leak costs by year, Refinery XYZ.

Another plant also replaced a de-ethanizer column. Further modification was performed at a later date, cladding the column with stainless steel. Sulfide stress cracking affected an LPG storage sphere that had been constructed from material with hardness exceeding the limits of MR0175. Sulfide stress cracking also affected a valve with a Type 410 stainless steel trim. Day-lighting projects to prevent external corrosion failures of buried in-plant piping have been completed at two facilities. Corrosion under insulation (CUI), insulation quality, and inspection for CUI are concerns at several facilities. CUI has affected liquid product tanks and piping delivery systems that operate intermittently. Propane and butane piping that operates continuously has been found to be in good condition. Utility water systems experience corrosion challenges. These include the seawater cooling system, cement-lined firewater system, and offshore insulated drain lines. Nonmetallic piping is being used in some renovation projects for buried sections of pipework. Nonmetallic materials are also being used for offshore applications such as deck gratings. Although of a higher initial cost, the use of  nonmetallic gratings has been supported by a total cost of  operations analysis, because the nonmetallic gratings have a significantly increased lifetime.

12 SAUDI ARAMCO JOURNAL OF TECHNOLOGY  SUMMER 2006

Overall, Saudi Aramco refineries had corrosion control costs in line with expectations, based on the analysis of  worldwide refinery data. There were broad differences in data reported from refinery to refinery, as shown in Fig. 15. Similar variations are reported by other operators. Figure 16 shows variation in cost for one non-company refinery over a period of years. The variations between refineries are expected due to differing complexities of each plant, differing ages of each refinery, and variations in record keeping and recovery of records. The experience level of  corrosion control personnel was directly related to their ability to identify and track corrosion costs. Based on published surveys, the cost of corrosion in refining worldwide ranges from $0.15/bbl to $1.34/bbl with an average value of $0.89/bbl. A NACE Task Group estimated the cost of corrosion in refining to range between $0.16/bbl and $1.96/bbl, with a mid range cost of  $0.85/bbl, confirming the range reported in published data. Saudi Aramco costs were in line with industry expectations. Over the three years from 1999 to 2001, over the five refineries, 36% of maintenance costs and 13% of  engineering costs were due to corrosion. Normally about 6% of operational costs would be due to corrosion. In 2000 and 2001, this was impacted by a major maintenance program at one refinery. Direct costs for corrosion control equipment are shown graphically in Fig. 173. Indirect costs due to deferred production during downtime were the largest cost factor in the total cost of  corrosion. The total cost of corrosion for refineries including deferred production costs was approximately three times the direct cost of corrosion. Corrosion in crude units, amine units and utility units including boilers was significant at most or all facilities.

Desalter operation, corrosion under insulation and positive materials identification issues were areas of concern at most facilities. These corrosion control challenges have prompted the development of research into the performance of  corrosion control chemicals17 and the development of a chemical alliance program18. Significant factors in the cost of corrosion were construction issues that led to a sulfide stress cracking failure, fatigue failure, and weld problems. Economies made during design and construction at one plant resulted in higher costs over the long term, emphasizing the need for a life-cycle cost approach to major project design. At a second plant, positive materials identification (PMI) issues resulted in 50% of bolting in a hydrocracker reactor being off  specification. Since 1990, four fires have resulted from corrosion at caustic injection points immediately upstream of the heater in crude units. Failure of storage tank bottoms due to soil side corrosion is a major concern at several facilities. Corrosion challenges in handling sulfuric acid provide an incentive for greater use of non-metallics. Non-metallics are also finding increasing use in utility water systems such as buried firewater piping. Corrosion under insulation remains a challenge especially for equipment that cycles through a wide range of temperatures and especially for systems subject to occasional dousing during tests of seawater firewater systems. CONCLUSIONS

Undertaking an extensive cost of corrosion study in Saudi Aramco’s facilities has allowed greater focus to be placed on corrosion control and the development of focused technologies to address critical issues. The need for a lifecycle costing approach to corrosion control in project design has been clearly established. The need for a career ladder for the development of corrosion engineers has been emphasized. For plants or systems with a high level of corrosion challenges, 60% to 70% of the plant maintenance budget is dedicated to corrosion control and in the region of 25% of  the plant operating budget. For plants with limited corrosion problems, even so, 15% to 25% of the plant maintenance budget is dedicated to corrosion and about 5% to 10% of the plant operating budget. By far the largest effect on the total cost of corrosion is the value of deffered production. It is often hard to obtain accurate data in this regard. A specific instance has been recorded where the indirect costs were 50 times the direct costs. Deferred production costs of three to five times the direct costs are readily documented.

REFERENCES

1. Tems, R.D. and Al Zahrani, A.: Paper 444, Corrosion 2006, National Association of Corrosion Engineers. 2. Benett, L.H. ed.: PB-282 631, Economic Effects of  Metallic Corrosion in the United States, U.S. Department of Commerce, U.S. Government Printing Office, Washington, D.C., USA, 1978. 3. Cakir, A.F.: Progress in the Understanding and Prevention of Corrosion, 1990, 671-677. 4. Cavassi, P. and Cornago, M.: The Cost of Corrosion in the Oil and Gas Industry, JPCL-PMC, May, 1999, 30-42. 5. Cherry, B.W.: The Cost of Corrosion, Materials Forum, Vol. 19, 1995, 1-7. 6. Gysbers, A.C.: STG 34 Cost of Corrosion Analysis, NACE Committee Survey, January 2003. 7. Iwawaki, H., et al.: Paper 02483, Corrosion 2002, NACE International, Houston, Texas. 8. Koch, G.H., et al.: FHWA-RD-01-156, Corrosion Cost and Preventive Strategies in the United States, U.S. Department of Transportation, Virginia, USA, March, 2002. 9. Hoar, T.P.: Report of the Committee on Corrosion and Protection, Department of Trade and Industry, London, UK, 1971. 10. Al-Kharafi, F., Al-Hashem, A. and Al-Matrouk, F.: Economic Effects of Metallic Corrosion in the State of  Kuwait, Kuwait Institute for Scientific Research, Kuwait, 1995. 11. Al-Matrouk, F., Al-Hashem, A., Al-Kharafi, F.M., and El-Khafif, M.: Industrial Corrosion and Corrosion Control Technology, Kuwait Institute for Scientific Research, 1996, 567-579. 12. Steinsmo, U. and Heggelung, H.: Norwegian Oil Review, 1993. 13. Syrett, B.S.: EPRI Report 1004662, 2002. 14. Vieh, P.H.: Analysis of DOT Reportable Incidents, 1985 through 1994, 9th Symposium on Pipeline Research, Houston, Texas, September/October 1996. 15. UK Onshore Pipeline Operators Association (UKOPA) Pipeline Fault Database, Pipeline Product Loss Incidents 1961 – 2000, Advantica Report Reference R 4798, June 2002.

SAUDI ARAMCO JOURNAL OF TECHNOLOGY SUMMER 2006 13

16. Coco, J.C. (editor), March Report: 100 Largest Losses in the Hydrocarbon – Chemical Industries, 1972 – 2001, 20th edition, March Risk Consulting Practice, February 2003. 17. Yanabi, Y.T., Tems, R.D., Khuraidah, A.S., Al-Jabran, A.A. and Origuerra, R.O.: Laboratory Evaluation of  Neutralizing and Filming Amines to Mitigate Corrosion in Crude Distillation Unit Overhead Systems, 11th Middle East Corrosion Conference, Bahrain, February 2006. 18. Almajnouni, A.D., DiGiacomo, T., Tems, R.D., Owen, T. and Moore, M.A.: The Implementation of Alliance Initiative in Refining Industry, Petrotech 2006, Bahrain,  January 2006. 19. Carswell, R.J.: High Temperature Ultrasonic Scanning, 11th Middle East Corrosion Conference, Bahrain, February 2006.

14 SAUDI ARAMCO JOURNAL OF TECHNOLOGY SUMMER 2006

ACKNOWLEDGEMENTS

The authors wish to acknowledge the Saudi Arabian Ministry of Petroleum and Mineral Resources, and the Saudi Arabian Oil Company (Saudi Aramco) for granting permission to present and publish this paper. The work involved in the cost of corrosion study was an extensive effort including the contributions of many individuals including Mohammed A. Al-Anezi, Gasem M. Fallatah, Faisal M. Al-Faqeer, Robert E. Palmer, Ahmad S. Al-Rammah, and Rakan Al-Shammary, as well as the individual teams at each plant surveyed.