Assessing and Mitigating
CORROSIO N in Closed Loop Cooling Water Systems PAM BOSCHEE, OIL AND GAS FACILITIES SENIOR SENIOR EDITOR
C
losed loop cooling water systems that are critical to the operation o machinery and process equipment depend on heat exchangers. Te quality o the water used in the system can adversely affect the integrity and reliability o the heat exchangers, resulting in accelerated corrosion rates, equipment ailure, and downtime. A case study presented at the NACE Corrosion 2014 conerence in March determined that the root cause o corrosion ailures in a heat exchanger operating in an onshore liquefied natural gas train system stemmed rom inadequate system cleaning during commissioning and startup, ineffective microbial control, and subpar monitoring o pH, iron, total suspended solids (SS), and total bacteria counts (BC).
Root Cause Analysis of Heat Exchanger Failure Santuraki and Al-Sayed (2014) described a cooling water system in which resh cooling water is supplied at a temperature o 38°C and a pressure o 7 bar, while the return water temperature is 48°C at a pressure o 5 bar. Te average flow rate o the supplied resh water is 95,000 ton/hr. A oncethrough seawater cooling system cools the return water using plate and rame heat exchangers ( Fig. 1). Sodium nitrite, blended with azole (or yellow metal protection), was used or corrosion inhibition, and hydrated sodium borate was used as a buffer solution. A nonoxidizing biocide, isothiazolinone, was used or microbial control. Historically, only nitrite residual and pH were measured. Afer a modified corrosion program was introduced, monitoring o the ollowing parameters was done at set requencies: nitrite residual, electrical conductivity, pH, SS, iron content, and BC. o determine the cause o the heat exchanger ailure, the authors studied the inspection reports, historical water parameters, sludge analysis, microbial analysis, and metallurgical analysis o the heat exchanger tubes. A 69-cm long straight section o a 13-mm outside diameter tube rom a ailed heat exchanger was used as the sample in the case study. Te tube was retrieved afer the tube bundle was pulled out or inspection. Te shell and tube design heat exchanger was a nitrogen compressor cooler, with nitrogen on the tube side and cooling water on the shell side. Te tubes were Unified
Numbering System (UNS) K01200, made o carbon steel ASM A179, and the shell was UNS K03006, made o carbon steel ASM 106 Grade B. Beore the discovery o the leak, the heat exchanger had been in service or 2 years.
Findings of Inspection and Testing Te tube bundle was ound to be covered by a large amount o orange deposit when it was pulled out ( Fig. 2). Te resh deposits were wet and sticky and had a distinct hydrocarbon smell. An underlying black deposit was ound when the orange layer was removed. Afer the deposits were removed or analysis, the tube samples were cleaned by sandblasting. Afer cleaning, the external surace o the tube displayed deep, irregular depressions filled with metallic and nonmetallic products (Fig. 3). Te deepest depression was 2.07 mm, about 90% o the tube wall thickness, corresponding to an average corrosion rate o 44 thousandths o an inch (mils) per year (considering the heat exchanger’s service lie o 2 years). A deposit analysis included loss on ignition (LOI), scanning electron microscopy (SEM) and energy-dispersive X-ray (EXDA) analysis, Fourier transorm inrared spectroscopy (FIR), and microbial examination. LOI at 550°C was 3% and 4% or external tube (shell side) samples A and B, respectively. Te SEM-EDXA results showed that the corrosion product comprised mainly iron and oxygen, with some copper and sulur in the pits on the tubes. Samples taken rom the pits or microbial analysis were negative or total aerobic bacteria, molds, yeast, and sulate-reducing bacteria (SRB). Te authors noted that because the samples were taken 2 days afer the tube bundle was removed, bringing into question the viability o microbes, the negative results do not eliminate the possibility o microbial-induced corrosion. A sulfide spot test, used to determine the presence o sulate-reducing bacteria, was negative. However, the authors said the absence o SRB was inconclusive, because the sulfide might have reacted to orm other compounds, or might have been washed away, because the sample was not resh.
Missed Opportunities for Remediation Fig. 4 shows periods o low nitrite (corrosion inhibitor) residuals, sometimes below 300 mg/L. Although nitrite levels were restored to the threshold limit (500 mg/L), the cumulative periods o low nitrite residuals were significant.
Seawater supply
Plate heat exchanger
Fresh cooling water tank
Fresh cooling water return
Seawater return
Process area
Fig. 1—A simplified process flow diagram of the fresh water cooling system.
April 2014
•
Oil and Gas Facilities
21
It was ound that when nitrite levels were low, the addition o the chemical was delayed rom 3 to 4 weeks, because o the lead time required or delivery. Unexplained depletion o nitrite occurred on some occasions. A confirmed case o leakage rom an extracted air cooler was documented as the cause in one nitrite depletion event. Following the low nitrite events, increasing trends o iron were noted. No coupons were available
or determination o corrosion rate. Te system also showed high levels o SS and had requent heat exchanger blockages. Santuraki and Al-Sayed said that the increase in iron content during and immediately ollowing the periods o low nitrite levels was an indication o increased corrosion within the system. Te increases in SS seen afer the low nitrite conditions were believed to be the results o the presence o the insoluble iron floating around in the system.
The Widening Competency Gap in Corrosion Management s shown in the accompanying article, effective asset integrity management can stumble as a result o various actors. Sometimes, the people responsible or the work drop the ball when it comes to putting together the “large” picture necessary or the assessment o a problem. In the case o the ailed heat exchanger, missteps occurred as early as precommissioning and continued through a 2-year lie cycle. Ali Morshed, corrosion engineering specialist at Saudi Aramco, highlighted the shortcomings o the traditional education and training o corrosion engineers at the NACE Corrosion 2014 conerence in March. High oil prices in recent years, in tandem with the increasing number o companies that offer integrity management services, have increased the demand or competent and experienced corrosion engineers. Morshed said that university and on-the-job training have been considered the mainstays in preparing engineers or asset integrity management. However, in some cases, the traditional training methods have produced engineers who could not carry out their routine and daily tasks competently and efficiently. Te initial competency o a new graduate engineer depends on the contents, qualities, and the organization o his or her education and training. In many cases, the situation is exacerbated by the retirement o the more experienced and competent colleagues without the transerring or sharing o knowledge and experience with others. “Universities do a very good job o teaching the students about electrochemistry, corrosion basics and mechanisms, metallurgy, cathodic protection and chemical treatment basics—the engineering related to corrosion. But, they do not do as well in teaching ailure risk assessment, the integrity review process, risk-based inspection, production o inspection scopes, or how to determine and use corrosion key perormance indicators—the implementation o corrosion management principles,” Morshed said. Operators and service companies have tried to bridge the resultant knowledge gap by providing mentoring or onthe-job training or their novice corrosion engineers by their
A
22
Oil and Gas Facilities
•
April 2014
more experienced and senior corrosion engineers. In theory, this approach should work. However, the working environment and the workload are ofen such that the senior colleagues do not have adequate time to properly and comprehensively train the new engineer, he said. In most cases, the on-the-job training is haphazard and random, which does not promote the transer o corrosion management knowledge and skills in an organized, structured, and effective manner. Adding to the difficulty is that the newly recruited engineers have been asked to not only act as the project corrosion engineers (sometimes, with limited, i any, supervision and guidance), but have also been required to perorm the project management duties, such as time and resource estimates, time and cost control, and dealing directly with the clients, Morshed said. He proposed solutions to enable better preparedness o new engineers, including • Universities should incorporate an introduction to “asset corrosion management” into their engineering courses to emphasize the distinction between corrosion engineering and corrosion management. • Only senior corrosion engineers who are conversant in corrosion management should be selected to serve as the mentor or the novice engineer. • Te senior corrosion engineer should establish clear and well-defined learning objectives or the on-the-job training, placing emphasis on corrosion management (since the engineering aspects have likely been handled by the university education). • More senior and experienced corrosion engineers should have opportunities or sharing their knowledge with their colleagues across the company and the industry through “lunch and learn” events, public presentations, or by producing brie technical guidance documents. Morshed said that the goal is to “highlight to the trainee corrosion engineers that any balanced asset integrity management system comprises both corrosion engineering and corrosion management components.”
800 700 L / 600 g m500 , e400 t i r 300 t i N200 100 0
Fig. 2—The heat exchanger tube bundle covered by orange deposits.
1 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 2 0 / 2 0 1 / 2 0 1 / 2 0 1 / 2 0 1 / 2 0 1 / 2 0 1 / 2 0 1 / 2 0 1 / 2 0 1 / 2 0 1 / 2 0 1 / 2 0 1 / 2 0 1 / 2 0 1 / 3 1 5 3 0 / 6 4 6 8 4 1 9 / 3 / 7 8 / 1 1 / 2 / 1 3 / 2 8 / 1 1 1 1 / 1 1 / 1 1 / 1 2 / 1 3 / 1 3 / 2 4 4 4 / 1 5 1 1 1 1
Fig. 4—The historical levels of nitrite residuals in the system. 1,000,000
Fig. 3—Deep depressions visible on the heat exchanger tube samples after cleaning.
I the amounts o SS in the system are not kept under control, SS will eventually be deposited in the low areas and dead legs, which can lead to interruption o chemical treatment by preventing the inhibitor rom reaching the pipe wall (Herro and Port 1993) and promote microbial underdeposits. Santuraki and Al-Sayed said that some o the iron was mill scale, which was attributed to inadequate cleaning during the commissioning stage o the system. Pieces o wood chips and remains o welding rods were among the debris recovered when the heat exchanger tube bundle was pulled out. Cleaning activities, including the use o bag filters and physical cleaning o the heat exchangers, were done to reduce the iron and SS to acceptable limits. Te authors’ microbial analyses showed consistently high BCs at 1 million colony-orming units (CFU)/mL (Fig. 5). Pseudomonas spp., a slime-orming bacteria, was also ound. Te authors ound that the BC in the system had not been measured historically. otal coliorm counts were mistakenly measured. When the coliorm counts were repeatedly negative, biocide dosing was stopped. Te dosing was restarted when a new corrosion monitoring program was introduced beore the heat exchanger ailure. When the biocide injection was restarted, the system volume was miscalculated, resulting in underdosing o biocide (the total system volume was 30,000 m 3, but was miscalculated as 4,000 m3). Te system volume was measured when repeated biocide dosing was ineffective in reducing the microbial levels to the threshold limit o 1,000 CFU/mL. At that time, a microbial analysis o the resh cooling water showed that the microbial population comprised mainly Pseudomonas spp.
Contamination of the System Te analysis o the cooling water indicated the presence o hydrocarbons, chlorides, and high dissolved oxygen levels. It was later discovered that two extracted heat exchangers were leaking and pumping a significant amount o oxygen
24
Oil and Gas Facilities
•
April 2014
L m100,000 / U F 10,000 C , t n 1,000 u o c 100 a i r e t 10 c a b l 1 a t o 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 T 2 0 2 0 2 0 0 1 0 1 2 0 2 0 2 0 0 1 2 0 2 0 2 0 0 1 2 0 0
/ 7 / 4 / / 2 / 2 5 / 2 / 9 / / 2 2 / 9 / 6 / / 2 0 / / 2 7 1 0 1 2 1 8 1 2 2 5 1 1 2 3 4 / 4 / 4 / 5 / 5 / 5 / 5 / 5 / 6 / 6 / 6 / 6 / 7 / 7 / 1 7 / 1
Fig. 5—The historical levels of bacterial counts in the system.
into the cooling water system. It was also believed that several hydrocarbon contaminations resulted rom previous leakages rom heat exchangers.
Conclusions and Recommended Practices Te findings in the case study highlighted contributing actors in plant downtime; reduced heat transer efficiency; expensive equipment cleaning, inspection, and replacement; and increased water and chemical usage. Corrective actions included isolation o leaking heat exchangers, cleaning and inspection o heat exchangers plugged by slime or debris, system blowdown rom low points and dead legs, sidestream filtration, increased nitrite residuals, biocide injection, and increased water chemistry monitoring. Te authors’ recommendations included • Before commissioning, a closed loop cooling system should be cleaned by flushing with service water, especially at low points and dead legs, combined with sidestream filtration to remove mill scale and suspended solids. A high dose o nitrite can help to orm a stable passive film in the system. A high dose o nonoxidizing biocide should be used, especially i the system will be filled with water and lef or extended periods o time beore the plant is ully commissioned. • For proper treatment aer commissioning, a nitrite program, containing sodium nitrite as a corrosion inhibitor or carbon steel, blended with azole or yellow metal protection and sodium borate as a buffer solution, is cost-effective.
• Maintain residual nitrite levels from 600 mg/L to 800 mg/L at all times. I sulate and chloride irons are present in the water, the minimum nitrite residual level should be at least five times the total o the chloride and sulate concentrations combined. • Regular biocide treatment is required, preferably with the alternating use o two biocides to prevent the microbes rom becoming immune to a single biocide over long-term use, to keep microbial levels below 1,000 CFU/mL. Selection o biocides should be based on laboratory screening using samples rom the resh cooling water system. Biocides that are known
to promote oaming, such as older generations o quaternary ammonium, should be avoided in closed loop systems. • Immediately upon opening the heat exchangers during inspection, deposits and swab samples should be collected and transported in a rerigerated container overnight or laboratory bacteria analysis. • Regular monitoring of water parameters, such as nitrite, pH, SS, BC, and iron, are required. When SS increase beyond a threshold value (depending on the system), sidestream filtration is recommended or cleaning the system. OGF
For Further Reading Santuraki, M. and Al-Sayed, A.A. 2014. Improving Reliability of Closed Loop Cooling Water Systems, A Lessons Learnt Approach. NACE Paper 4291. In NACE Corrosion 2014 Conference Proceedings, 9–13 March 2014, San Antonio, Texas.
SPE International Oilfield Corr Conference and Exhibition
Herro, H.M. and Port, R.D. 1993. Te Nalco Guide to Cooling-Water Systems Failure Analysis . Boston, Massachusetts: McGraw-Hill.
ter Now!
12–13 MAY 2014 ABERDEEN EXHIBITION AND CONFERENCE CENTRE ABERDEEN, UK
New Challenges for a New Era
www.spe.org/events/ofcs
i
April 2014
•
Oil and Gas Facilities
25
