Technical Bulletin # C-03

Roof Level Corrosion

Proof That Annually Draining Down Condenser Water Piping Will Dramatically Increase Corrosion Rate Activity

It is a generally recognized fact that fully drained or partially drained piping systems are far more susceptible to corrosion than those containing treated water, or even untreated water.  Given a moist environment in combination with the presence of abundant air and oxygen, partially drained pipe has been documented to corrode at a rate two to ten times that of other water filled pipe of the same type, and located within the same circulating system.  Condenser or open water systems clearly suffer the greatest in Northern climates where draining is often unavoidable.

In cases where condenser water piping is drained down within the interior of a building to protect it from freezing, it is common to measure significantly higher corrosion rates at the rooftop or outside level.  In those cases, roof level piping will require replacement many decades before the remainder of the system.  The buckets of iron oxide and scale typically removed from strainers, condenser heads, and heat exchanger tubes at every spring start-up are partially the result of such higher off-season corrosion activity.


Coupon Testing Impossible

Corrosion coupon racks, typically the only form of corrosion monitoring used, cannot be employed at the most vulnerable area of drained down pipe due to a lack of water flow.  They are rarely installed in outdoor or at roof level locations.

Of course, corrosion at the outer surface of the exposed rooftop pipe will also occur if not properly coated, insulated, and protected - a common maintenance problem often identified as a contributing factor to an overall higher measured corrosion rate.  However, it is most often the pipe's interior, having been totally or partially drained over many years, which places the system at greatest risk of advanced failure.


Case History

The below comparison of wall thickness measurements and estimated corrosion rates, taken from an actual 1996 ultrasonic piping investigation of a New York City commercial property, clearly illustrates the differences which, to some degree, always exists between drained and filled piping within the same condenser water system.

In this example, the left side of the page represents test results taken from an ultrasonic evaluation of a section of 18 in. extra heavy condenser water pipe located in the sub-basement mechanical equipment room area of a 42 story New York City office building, and never drained of chemically treated water.

The right side of the page represents the same exact pipe at the 30th floor outdoor shaftway, where it has been drained every year during its five month winter season.  The pipe is the same in all respects - having been extra ASTM A 53 seamless heavy stock with an initially specified wall thickness of 0.500 in., a solid history of excellent water treatment maintenance, and having been in service for 45 years.   Descriptions of the various charts and bar graphs precede each set of report data.


Comparison # 1 - Wall Thickness Measurements

In a direct comparison of current wall thickness measurements, test results show significantly lower remaining pipe metal at the drained test site, shown below to the right.  Also, the greater deviation between thickness measurements at the drained piping illustrates the much higher level of pitting activity at that location.  Please note the differing scales for wall thickness at the left side of each graph - required due to the large difference in wall thickness.

c03_01 c03_02
Filled Pipe Drained Pipe




Comparison # 2 - Average Corrosion Rates

The below calculations are based upon the average of all recorded ultrasonic wall thickness measurements illustrated above.  Differences in average measured pipe thickness, corrosion rate, and remaining pipe life are dramatic.  As shown, the annual draining of this system has actually increased the average corrosion rate of the roof level piping four times that of the rest of the system. As a result, the lifetime of the piping has been reduced by nearly ten fold!

c03_04 c03_03
Filled Pipe Drained Pipe




Comparison # 3 - Maximum Corrosion Rates

The below tables show the same basic set of calculations as in Comparison # 2, except that they are based upon the lowest measured wall thickness value of each set.  Such data represents a weak link or worst case scenario, and offers an estimate of when the most aggressive corrosion activity will deteriorate the piping past its safe recommended limit.

Shown below, differences in corrosion rate, percentage of loss, and retirement date are even more pronounced.  In fact, the minimum wall thickness of the constantly filled pipe significantly exceeds the standard specifications of any new pipe installed today, while the upper pipe exists at nearly half that value.

c03_05 c03 06
Filled Pipe Drained Pipe




Comparison # 4 - Original vs. Remaining Values

While the original pipe wall thickness and minimum allowable thickness values remain constant for both test locations, a major difference is obvious in the amount of pipe wall remaining.  For the subject property, ultrasonic testing indicated that the minimum measured pipe wall thickness of the drained piping is nearing its minimum allowable safe operating limit.  Yet the water filled pipe offers almost unlimited remaining service.

c03_07 c03_08
Filled Pipe Drained Pipe




Comparison # 5 - Corrosion Rate vs. Location

The below graph is based upon data taken from a separate ultrasonic investigation of a 37 floor condenser water piping system at a downtown New York City office building.  The blue line represents the average corrosion rate based upon all wall thickness measurements; the red line represents the highest corrosion rate based upon the lowest individual wall thickness.  Since the condenser piping was installed exposed to the elements within an outside stairway, it had been drained completely to the bottom for all 39 prior winter seasons.

Ultrasonic thickness testing was performed at the roof and at each floor down to the basement. As expected, testing showed a high corrosion and pitting condition throughout most areas.  When the test results were sorted based upon the physical floor location, highest to lowest, we found a clear trend - with noticeably higher corrosion and pitting activity at the upper areas of the pipe, and the only low and uniform examples of corrosion found at the lowest 4-5 floors.  Similar to many past investigations of drained vs. filled pipe, an almost 10 to 1 relationship was shown to exist.




Although the entire piping system was drained, the higher loss of pipe wall thickness at the upper floors was attributed to the abundance of fresh air and oxygen migrating through the cooling tower openings and down the risers.  The lowest sections of pipe, as expected, showed the lowest corrosion activity for the exact opposite reason.  Similar results have been found in many other CVI investigations, as well as by other investigators.


Useful Recommendations

There are currently only four feasible methods toward reduce the corrosion rate within a piping system which is partially or fully drained over any period.  One or more may apply depending upon conditions.


Increase Inhibitor Level


The easiest, though least effective measure, is to greatly increase the level of the standard chemical inhibitor just prior to draining.  This theoretically leaves a heavier coating of rust protection on the piping to provide partial protection against oxidation while drained.

In reality, CVI has found little benefit through this action, and most properties which do follow such procedures generally still find an excessive corrosion loss at drained pipe locations.

Most water treatment companies offer special lay-up inhibitors, although it is difficult to judge their effectiveness in the field.  We have never failed to identify far greater corrosion activity at a roof location regardless of the form of standard lay-up chemical treatment provided.




Vapor Corrosion Inhibitors


A second and extremely effective method is to introduce a supplemental chemical rust inhibitor specifically formulated for the purpose of preventing corrosion during lay-up periods.  Numerous formulations exist, including the newest development of powders called Vapor Corrosion Inhibitors (VCI). 

These gas producing products place a protective and penetrating layer of corrosion inhibitor on the surface of the metal to provide virtually total corrosion control for up to two years.

They require a general, though not airtight, sealing of the piping system in order to allow the protective element to sublime, that is - move from solid directly to gaseous phase and then migrate throughout the piping.

We highly recommend Cortec products, with further information available at




Nitrogen Gas


A third rarely used, but effective method, is to fill the empty system with a blanket of nitrogen gas, displacing the air and oxygen, and stopping the corrosion process almost entirely.  Its effectiveness relies upon being able to drive all air from the system to replace it with only nitrogen.

This procedure requires an airtight piping system - which may be difficult to achieve at the cooling tower end - the area most in need.  It is rarely used in HVAC applications.




System Drying


Since corrosion cannot continue without a presence of moisture, reducing that moisture content below a certain dew point will inevitably stop or substantially lower all corrosion losses.  Drying, however, is often difficult due to the physical configuration of a piping system, and especially for open systems at the cooling tower itself, which would require temporary closure. 

Areas of the system containing larger reservoirs of water can take months to dry.  For clamped piping system, substantial water will remain at all horizontal lines due to the groove which raises an internal bump trapping water along the bottom.  Vertical risers are therefore the easiest to address most effectively.

Drying can be accomplished by one or a combination of heating, desiccant use, or the introduction of heated super dry air to -100 °F into the system.  Custom configured equipment is available.


Clean System Essential

An extremely important step prior to any lay-up procedure is to chemically clean and sterilize the entire system.   By removing unwanted rust, dirt, and microbiological matter, the above inhibitor methods will work much more effectively due to the increased amount of contact between the chemical inhibitor and base metal.

High corrosion and pitting rates present a significant threat to every building or plant property which drains its piping for even short intervals.  Most often, that wall loss is not recognized until a failure occurs, or interest to learn the source of start-up deposits is raised.  It is a problem often not addressed due to a combination of non-awareness, physical difficulty, questionable effectiveness of result, and an expenditure benefitting only a potentially limited amount to the overall piping system.


©  Copyright CorrView International, LLC

Company Info


CorrView International, LLC

P.O. Box 8513
Landing, NJ  07850
Ph:    973-770-7764
Fax:  973-770-6576

For More Information

Our Visitors

Join For Site Updates

New content is regularly added to the site, when time allows. 

For the latest updates and other advisories you are invited to join our mailing list.




For More Information