[modula id=”4206″]

Corrosion Coupons

Our extensive volume of work in the field of ultrasonic pipe testing has overwhelmingly documented the accuracy and reliability of this testing process, as well as its ability to provide more valuable information than any other pipe testing method typically employed.  Combined with proper data analysis, and given sufficient test points, ultrasonic testing (UT) provides an excellent overall evaluation of pipe wall loss, corrosion rate, and remaining service life.

A frequent conflict exists, however, when comparing corrosion rates from new metal coupons against corrosion rates estimated from actual wall loss measurements.  In virtually all cases, reported corrosion coupon rates will exist far below those calculated based upon actual wall thickness loss – often with the corrosion rate for coupons showing 10% or less of the true wall loss occurring.  Where questions arise, independent testing in the form of a metallurgical analysis will confirm the accuracy of ultrasonic testing over corrosion coupons in every case.

Unfortunately, corrosion coupons have been in service for so long that their function and accuracy are now accepted without question; their results relied upon almost blindly.  Very obvious faults in their ability to report a corrosion rate relative to the actual pipe, and clear examples where they have failed to return accurate data, are typically ignored by even those relying upon such results as part of their professional service.  While the use of corrosion coupons present no threat in monitoring a piping system that truly exists having a low corrosion rate, they can have severe consequences to a building property with a high corrosion problem, and is thereby misinformed – as CorrView International, LLC has proven time and again.

• Coupons Fail To Represent True Pipe Corrosion

By definition, corrosion coupons will estimate the potential for a chemical inhibitor to protect a new steel or metal surface, or conversely, the potential for a liquid to corrode a new metal surface.  They do not indicate the wall loss occurring to the pipe wall, as most authorities well recognize.  This is a significant difference from the commonly presumed definition of what is being measured and reported by corrosion coupons, and often vastly different from what a client understands corrosion coupon based information means – which is, a measurement of pipe wall loss.

That a corrosion coupon test result does not truly represent the corrosion rate at the pipe wall is rarely expressed to a client who has been recommended to use corrosion coupons as a guide of chemical treatment effectiveness.

In many cases we have been involved, reported corrosion c oupon rates are not even within the realm of possibility, with some reflecting numbers that would be even difficult to reproduce in the laboratory under ideal conditions.  Corrosion coupon rates reported at 0.1 or 0.3 mils per year (MPY) for open condenser water systems are commonly offered to clients as proof that the chemical water treatment is performing, when such numbers are not even feasible, and should be immediately suspected.  Aside from the inherent inaccuracy of corrosion coupons to report corrosion activity at the pipe surface, manipulation of the results and outright fraud are also occasional elements seen.

CorrView International has witnessed corrosion coupons that have been coated with clear lacquer, sealants, and even nail polish.  We have seen corrosion coupon reports that have been tampered with to reflect low corrosion rates where a higher rate was actually measured.  Aside from offering a check on the accuracy of a specific laboratory, the use of multiple corrosion coupon testing labs is often for the purpose of eliminating any fraud in the handling and reporting of the corrosion coupon rate.

Corrosion coupons unfortunately do not provide an accurate or realistic estimate of the actual corrosion activity taking place at the surface of the pipe.  As the corrosion rate of the piping system itself increases, it is common to see an even greater difference between coupon reported losses and actual pipe wall loss.  The presence of even a moderate amount of interior deposits will almost completely eliminate any relevance between a coupon based corrosion rate and the wall loss which is actually taking place at the pipe surface.  Plus, basing an entire chemical treatment program upon the results of corrosion coupons taken at only one point in the system is risky in itself, given the low probability that any one test area is representative of the entire piping system.

• Common Case History Scenario

A client manages an 11 year old, 42 floor office building in a major U. S. city.  Refrigeration is provided by a 24/7 condenser water system running the entire height of the building, with two individual package units on each floor serving A/C and computer cooling needs.  One 4 in. take-off from each 24 in. carbon steel riser supplies water to and from the A/C units, while a second 2-1/2 in. set of take-offs are available for future use, and valved off at most floors.  Main risers are standard grade ASTM A53 carbon steel of domestic manufacture and having a specified wall thickness of 0.375 in.

The property management firm is particularly recognized for very well maintaining its properties, and has a chief engineer that places chemical water treatment as one of his highest priorities.  Chemical treatment is maintained by a nationally well known and established chemical supplier, with no expense seemingly spared in installing a fully automated chemical injection and bleed system.  Annual costs for water treatment chemicals and service alone exceed $65,000.

On-site chemical testing is performed twice per day and all necessary adjustments made to maintain the chemical levels within design parameters.  A representative of the chemical company is on site once per month or more in order to discuss results and to adjust the treatment schedule if necessary.  Two independent and highly respected water treatment consultants are employed to review all data, analyze water samples, and make their own recommendations on a bimonthly basis.  Added expenses for the water treatment consultants approach the cost of the chemical water treatment contract.

The building was designed and built with an automatic full flow water filtration system capable of reducing particulates down to 100 microns – a tremendous benefit, and one not often seen specified today.  In addition, they have installed a large sand filter at the roof level MER to further remove the condenser water particulates to below 1 micron.  The water is clear and free of turbidity, and both filters are said to work flawlessly.

Corrosion activity within the condenser water system is monitored by one elaborate corrosion coupon rack installed at the roof level MER.  The coupon rack is constructed of PVC, and offers multiple coupon ports.  In addition to the coupons installed by the chemical treatment contractor and changed on both 45 and 90 day intervals, two other sets of coupons are provided and analyzed by both independent water treatment consultants.

Three independent sets of corrosion coupons thereby serve to define the effectiveness of the water treatment program, and as the basis of the decision making at this large building property.

• Tell Tale Problem Signs Exist

While this building property has always taken a proactive approach to the chemical water treatment, certain signs are present to suggest a hidden corrosion problem.  Some connections at the isolation valves to various cooling units show corrosion product at the threads.  Pinholes in some of the smallest diameter piping to supplemental tenant A/C cooling has required replacement.  Random copper cooling coils have failed due to pinholes.  Blowing down any dead leg on even a regular basis produces a high quantity of iron oxide mud before the water finally runs clean.  Rust and scale accumulates in the cooling tower pans sufficiently to require periodic removal.

The capture volume of the two condenser water filters is unknown since it discharges directly to floor drains rather than to a holding tank.  The sand filter inlet line is oriented perpendicular to the flow of water at the high velocity discharge of the condenser water pumps – thereby greatly reducing its effectiveness.

• Corrosion Coupon Results Suggest Otherwise

While there is some suspicion by the engineering department that a corrosion problem may exist, corrosion coupon results show no cause for concern.  Estimated corrosion rates of all three coupon laboratories agree to within 15% of each other, and report consistently low corrosion rates of about 0.4 mils per year (MPY), excellent by any standard.

Each set of coupon results are reviewed and compared with prior results, minor changes in treatment levels are recommended by the chemical treatment representative or water treatment consultants, and minor changes made to the treatment program if necessary.  The independent water treatment consultants provide quarterly reports generally supportive of the actions taken by the engineering staff.  As a result of the favorable corrosion coupon results, all parties concerned are pleased at the results of their own efforts, and of each other.

Lingering concerns due to the rust deposits found within the system, however, prompt a further investigation into the condition of the system using ultrasonic thickness (UT) testing – a testing method never before employed at this building property.

• A Different Piping Assessment Emerges

CorrView International, LLC initiates an investigation with the objective of providing the most thorough piping evaluation possible within the one day of field testing contracted.  Ultrasonic testing is performed according to our standard procedure, and addresses all major areas of the building including, tower area, pump room, risers at various intermediate floors, threaded pipe, top and bottom of the system, distribution lines to the A/C units, future taps at the risers, and miscellaneous pipe fittings.  A total of 45 individual piping locations are evaluated as part of this investigation.

Our ultrasonic testing immediately identifies a high wall loss, and a wide 60-100 thousandths variance in the high to low wall thickness measurements, which is a strong indication of a pitting condition.  Significantly, it also identifies random high thickness values suggesting that the pipe was received and installed above specification.  In fact, new pipe can be produced +/- 12.5 percent of its ASME factory specification and still be suitable for service.  For this older property, the use of standard 12 in. through 24 in. pipe having a specified wall thickness of 0.375 in. is found with random thickness values of 0.410 in. throughout most areas tested.  The pipe wall is therefore assumed to have been oversized by 10%, and all calculations based upon this new higher starting wall thickness of 0.412 in.

While in one respect a clear benefit to the building property by its heavier wall thickness, it has to be remembered that any calculation of corrosion rate must then take into account the higher wall loss from this higher assumed starting point – a common flaw in many corrosion reports.  References made in some prior consultant reports of pipe service life mistakenly use the ASME factory specification thickness of 0.375 in., when actually, the pipe originated much heavier and with a correspondingly much greater wall loss.  This error alone results in a much lower than true corrosion rate estimate by prior investigators.

• Roof Level Pipe

Testing at the 20 in. roof level machine room main header showed a high degree of pitting and high corrosion rate at this and all other areas investigated. An average corrosion rate of 4.5 MPY at the main riser piping documented through ultrasound was more than 10 times the 0.4 MPY corrosion rate predicted by corrosion coupons. Ultrasonic investigation showed dramatically different corrosion rates at this condenser water piping located only 10 ft. from the corrosion coupon rack installation. Various other visual examples suggested higher corrosion activity including thread leaks and cooling tower sediment.

• Bottom Of Condenser Riser

This bottom area of the condenser system at a 2 in. drain valve nipple showed very high corrosion and higher pitting activity. With nearly 275 PSI of pressure at this area and no shut off between here and the roof level cooling tower, the failure of this location represented a very serious threat which was previously unknown. Testing showed that even the use of extra heavy pipe could not withstand 11 years of such high corrosion and pitting activity. No concern had been raised in this area given the low 0.4 MPY corrosion rate shown by coupon reports.

• End Of Lower Horizontal Run

Lower parts of any piping system often demonstrate higher corrosion rates due to the settlement of particulates. Long horizontal runs further amplify the effect, as clearly shown here. This area of piping had been previously replaced five years earlier.  However, no significant concern over its premature failure was raised due to favorable corrosion coupon results. Ultrasonic testing showed an extremely high corrosion rate of over 22 MPY and current wall thickness values only existing 0.013 in. from the threads, and under 275 PSI of operating pressure.

• 15 Times Corrosion Under Reporting

Over its 11 years in service, our UT testing showed a consistent corrosion rate of between 6 and 9 mils per year – far above the 0.4 MPY estimate of corrosion coupons by a factor of more than 15 times.  At most locations, a high to low wall thickness spread of up to 100 thousandths or 0.100 in. was measured – confirming a serious pitting condition.  Yet no indication of pitting was shown in the corrosion coupon results.

Given that a 5 MPY corrosion rate at 24 in. pipe will produce an annual loss of approximately 121 lbs. of steel pipe for every 100 linear feet, we can roughly estimate that far greater than 12,000 lbs. of original pipe wall has already been lost due to corrosion over the life of this property thus far.  Much of this rust volume is removed in blowdown, cleanings, and through filtration, while some percentage will settle and accumulate within the system itself.  It is that remaining amount of corrosion debris which will cause future problems.

The same high corrosion rate is found throughout most areas tested, which helps to confirm the system wide presence of whatever problem exists.  Testing shows far less corrosion and pitting at the vertical risers, and testing of any horizontal lines clearly identifies much higher wall loss at the bottom and lower sides of the pipe – further suggesting a deposit related corrosion problem.  A slightly lower corrosion rate is found at the smallest diameter piping.  However, galvanic activity is observed at some brass valves, which would further reduce wall thickness at the threads, immeasurable through our UT testing methods, and therefore remaining an unknown.

Schedule 40 pipe is identified as the material used in all small diameter threaded applications, and all examples tested are documented to exist below minimum acceptable wall thickness standards.  With a beginning wall thickness of 0.154 in. for 2 in. schedule 40 black pipe, our testing showed remaining thickness values of between 0.118 in. and 0.130 in. at most examples.  Having an initial thread loss of 0.085 in. which must be deducted from its available wall, we can easily document that barely 0.033 in. remains at certain areas of the threads, and that such small diameter threaded pipe represents a high priority area for failure to occur.

The larger 2-1/2 in. future take-off lines at each floor show the same degree of corrosion as elsewhere, with slightly higher wall loss at the threaded end – suggesting galvanic loss at the steel to brass connection.  Current wall thickness values of between 0.140 in. to 0165 in. remain, but a higher thread loss of 0.115 in. exists for 2-1/2 in. and larger pipe – thereby leaving little steel at the threads themselves. No galvanic insulators are installed.

With corrosion and thread loss considered, most of the future take-off lines tested are shown in highly deteriorated condition and in need of replacement.  Some current wall thickness values of under 0.140 in. are found – meaning that only 0.025 in. remains at some areas of the threads.  Compounding the difficulty of replacing these two sections of pipe at 42 different floor locations is the fact that the pipe is directly welded to the riser, rather than being more appropriately threaded into a threadolet fitting.  Repair now requires burning off the existing pipe and welding in a new fitting – an expensive and time consuming effort.

• Questions About The Monitoring Program Arise

Our ultrasonic testing report is clear in result and overwhelming in the evidence presented of a system wide high corrosion and pitting problem.  This information dramatically conflicts with years of previously offered evidence of a low corrosion rate and no problem concerns, yet coincides with physical observances about the system.

Due to favorable corrosion coupon results, clearly evident problem signs such as leaking threads and pinhole failures were explained away as isolated failures due to causes other than a high corrosion rate.  The accumulation of rust deposits in the tower pans and drain down lines are also wrongly interpreted as the result of airborne particulates being captured by the cooling tower, or other causes.

With only one method of corrosion monitoring provided to indicate corrosion control effectiveness relied upon for its entire 11 years in service, this building property obviously existed under a false sense of security for many years.  Valuable time was lost where an earlier identified corrosion problem might have been easily brought under control.  Whether the corrosion problem was greater initially and is now a lesser problem, or whether the corrosion losses shown took place in more recent years cannot be identified at this late stage of investigation.  Undebatable, however, is the finding that a significant amount of wall loss has taken place; undetected and unreported through all previous monitoring and investigation efforts.

• Some Bad Assumptions, The Underlying Problem

Fundamentally, the reliance on corrosion coupons as the exclusive test method is at fault.  In establishing a corrosion monitoring program, the only questions raised about the corrosion coupons were of the accuracy of the various labs used, rather than if the coupon test method was returning accurate and representative data.  Enormous effort was essentially wasted in cross checking test results from a methodology not likely to provide a result representative of the true system corrosion rate.  All three labs provided accurate corrosion rate data from metal coupon samples not representative of the actual pipe loss.

It may have been reasonably assumed that corrosion coupons in a new piping system would have provided accurate corrosion loss estimates initially, but over time, such a relationship diminishes.  The observance of deposits in the tower pan and dead legs should have suggested an interior deposit problem very likely to further limit the accuracy of the coupons installed.  It should have also immediately suggested the need to establish alternative corrosion testing methods.

The placement of the corrosion coupon rack at only one location, at the roof area MER near the chemical feed station, clearly plays a part in not being able to identify true corrosion activity at the lowest points within the system, and elsewhere.  Furthermore, its construction of PVC plastic completely eliminates any electrical contact, and therefore significantly minimizes what corrosion is essentially defined as being – an “electro-chemical” reaction.

Significant other faults of corrosion coupons exist which are well defined in previously referenced articles on this Internet site.  Suffice it to say that in their isolated state, unaffected by particulates, rust deposits, flow rate, electrical currents, galvanic activity, under deposit corrosion, microbiological growths, and the many other factors which do exist within most typical condenser water systems, there is no other reasonable expectation than for corrosion coupons to produce artificially low corrosion rate results.

Ironically, the failure of corrosion coupons to produce an accurate and representative corrosion rate assessment in this case was quickly recognized and agreed to exist by both the chemical treatment contractor and independent corrosion consultants once evidence in our UT report was presented.  In this case, as is typical, blame was quickly and finally focused by all upon the corrosion coupon testing method.

• Common Scenario

The above case history is an extremely common scenario we have seen to various extent in many dozens of testing investigations.  Typically, a building property relies exclusively upon the results of corrosion coupons, and then realizes they have a severe corrosion problem years later.  Finding someone to pin the blame on is often the first reaction, with the plant engineer and/or water treatment contractor often receiving the greatest criticism.  Once recognized as a threat to building operations, the appropriation of major funds is required to correct the problem.

It is exactly this failure of corrosion coupons to accurately represent true corrosion activity within piping systems that the CorrView corrosion monitor was developed.  Various articles are available on this Internet site providing documentation to this finding.

© Copyright 2024 – William P. Duncan, CorrView International, LLC