[modula id=”4206″]

A Hidden Threat

There is a general understanding by those working within the building operations, plant engineering, and water treatment fields that drained carbon steel pipe will corrode to a higher degree than pipe which is always filled with water.  This produces a tremendous vulnerability to the piping system, and one which is typically hidden from view.

While the actual degree of wall loss for drained pipe is in most cases undefined and only speculated upon, a presumption that it does exist often prompts preventative action.  For those responsible for the operation of building or plant properties in Northern sub-freezing climates, freeze protection for less than 24/7 operations is required – thereby introducing new corrosion concerns as well.

• Corrosion Activity Varies

Such wall loss varies greatly, and is, unfortunately, far from being predictable.  Pipe which is drained and left in a wet condition has been documented to corrode at as much as 10 or more times the rate of pipe left filled with chemically treated water.  Yet in some cases, roof level pipe which has been drained and left to dry has exhibited only slightly greater corrosion and pitting activity – reducing service life, but not significantly threatening building or plant operations.  In some instances, pipe left drained over decades of winter lay-up will show only a moderate increase in corrosion and pitting activity.  In others, 15 years of winterization will produce sufficient damage to require full pipe replacement.

In fact, the actual amount of wall loss at drained pipe will vary greatly depending upon many different factors and physical conditions, and is still not well understood or explained by most piping and corrosion authorities.  Two similar properties having similar condenser water systems and operating under similar drain down conditions may show dramatically different pipe corrosion characteristics.  Therefore, in order to minimize this potential corrosion threat to the best degree possible, some preventative action is always advised.

Various influences on the degree of corrosion occurring in drained down piping are:

• • • Inherent corrosion resistance of the pipe
    • Length of downtime
    • Fluctuation in water level inside the drained pipe
    • Pipe temperature or ambient temperature
    • Number of years drained
    • Chemical inhibitor and inhibitor dosage level
    • Preventative actions taken
    • Infiltration of outside air
    • Horizontal or vertical orientation of the pipe
    • Moisture content within the piping
    • Location top or bottom of the piping system
    • Length or travel of the piping system

• Attempts To Mitigate The Loss

Recognizing the probability of such high wall loss, most building and plant managers will take some action to safeguard or at least retard the potential corrosion damage.  Most commonly, high dosages of the normal chemical inhibitor are applied to the system immediately before drain down – the theory being that higher chemical levels will coat the pipe to provide greater corrosion protection.  A maintained molybdate level of 10 PPM therefore may be boosted to 50 PPM or higher in the hope of providing greater corrosion protection.

In reality, such effort is virtually worthless.  Even a molybdate inhibitor level of 50 PPM is far less than the 100 PPM or greater maintained in any closed system – capable of producing only 1 mil per year (MPY) corrosion rates under much more ideal conditions.  Our ultrasonic testing of condenser water systems which have carefully followed such procedures for decades have shown no benefit whatsoever, with virtually the same high wall loss shown as exists at unprepared piping.

A more effective method, sometimes employed, is to insulate and heat trace the pipe, and then maintain the pipe filled with chemically treated water – an expensive option for a seasonal cooling system.  Given that substantial pipe wall deterioration will exist wherever the water level ends, it becomes necessary to maintain the entire piping system full back to the cooling tower, heat tracing it accordingly.  Otherwise, such action only accomplishes moving the point of pipe wall failure to somewhere else.

• More Effective Options

Displacing the air and oxygen by a blanket of nitrogen gas is one very effective method for protecting carbon steel pipe, but one rarely used in building and plant operations.  Quickly drying the pipe to extremely low relative humidity, is another option.  A new class of highly effective chemicals called Vapor Corrosion Inhibitors (VCI) offers probably the greatest protection at reasonable expense and effort, and can provide almost total corrosion protection.

Few other options exist.  Without utilizing nitrogen gas, VCI products, drying the system completely, or some other proven method of lay-up protection, any drained piping system can be expected to show greatly accelerated deterioration over time.

• Areas Of Deterioration

The greatest level of metal deterioration often exists at the interface between water and air – where saturated moisture exists in an abundance of oxygen.  For any vessel where the level of water is constantly raised and lowered, such effect is especially pronounced.  In a typical piping system, variations in temperature may expand and contract the water and pipe to create a small washing effect over a localized area to produce severe wall loss. Areas of horizontal pipe are especially vulnerable due to incomplete draining.

This exact corrosion scenario is commonly found by our ultrasonic testing of domestic cold water storage tanks and especially at steam condensate tanks, and provides an ideal illustration to the similar problem at piping.  Performing thousands of wall thickness measurements along a defined X – Y grid allows us to produce a virtual three dimensional image of the interior tank wall, and often provides clear illustration of any weakness at the tank surface.

In the below example showing the left and right sides of a 22 year old, 3,000 gallon cold water make-up storage tank, ultrasonic testing showed a very defined deterioration of wall thickness along the entire length of the shell and heads.  This area of deterioration directly corresponded to the high and low limits of its fill valve, producing a very localized 30% deep penetration.

This same corrosion effect was also shown at the front and rear heads, further confirming this common problem.

• Ultrasound Identifies The Problem

Pipe corrosion problems are often complex conditions developing over many decades of service.  Identifying their severity and extent is rarely evident from observation, and typically requires a moderate degree of investigation for even the most simple corrosion forms.

Ultrasonic testing (UT) provides a high level of information with no interference to operations, and at very moderate cost.  With sufficient test points taken in the appropriate areas, and given proper analysis of the resulting data, UT can provide an extremely clear and accurate evaluation of pipe condition which may have escaped detection for decades.

The below seven sets of data are taken directly from an investigation into a 55 year old piping system at a New York City property.  With a total replacement of the central chiller plant and cooling tower planned, consulting engineers justifiably raised concern for the original 28 story condenser water riser piping.  Due to its age, this pipe was expected to have been extra heavy originally, and assuming a moderate corrosion rate based upon its prior history of being well maintained, was anticipated to fall somewhere at or below the ASME specification for standard pipe, which has been often seen in other similar properties.  No operating problems or leaks had been reported.

This 18 in. condenser water piping system was constructed of extra heavy ASME A72 wrought iron pipe having an original wall thickness of 0.510 in.  Installed in 1949, the pipe had been annually drained 1-3 floors inside the riser shaft every winter for freeze protection between the months of November and March.

Initial testing, beginning at the lower floors, showed excellent results; with the wall thickness approaching that of new 0.500 in. thick extra heavy pipe in areas, and still far exceeding specifications for new standard pipe – which was planned should the pipe require replacement.  Further testing, however, showed significant difference.  As the various sets of graphs and statistics show below, this piping system suffered greatly varying wall loss depending upon the location of the pipe.

• Basement Level Pipe

The results at left were taken from the very bottom of the return riser in the basement chiller room where the pipe has always remained full of chemically treated water. Here we show a consistently high remaining wall thickness generally exceeding 0.450 in. – which is excellent.  Thinner pipe having a wall thickness of 0.375 in. would be installed today for such service. We show a very low corrosion rate and some moderate pitting at this return line to the cooling tower, as evidenced by the top bar graph of the 12 wall thickness readings taken.  The lower graph quickly illustrates the overall condition of the piping in this area. Virtually unlimited service life was found to remain.

• Roof Level Pipe

Testing at the roof level area showed the beginning effects of drained pipe.  Overall corrosion rate is approximately tripled, although pitting activity is similar. For this and all other areas of roof level piping, we can show that the average wall thickness is at virtually the thickness specifications for new standard pipe. While losing much pipe wall, the original use of extra heavy material now benefits the building property 55 years later. Extremely long service life still remains.

• Drained Pipe – 1 Floor Below Roof

Entering the pipe shaft, a generally inaccessible area, proved critically important to this investigation.  Our testing identified a very high corrosion rate and severe degree of pitting activity as shown by the wall thickness profile to the left. While some wall thickness readings were higher, extremely low thickness values exist to define this pipe as existing under minimum acceptable standards.  Here, the corrosion rate is over 10 times higher than at the basement level. Testing showed that even the use of extra heavy pipe could not withstand decades of such high corrosion and pitting activity. This pipe was judged as unacceptable for further reliable service and recommended for replacement.

• Drained Pipe – 2 Floors Below Roof

Further testing downward into the shaft revealed similar results, as shown in the accompanying graphs. We documented conditions of extremely high wall loss and pitting at both supply and return risers in this and other upper areas of piping near the planned drain down point. Extensive testing at this and other upper floor riser areas showed identical results.  We identified no further service life existing and recommended pipe replacement.

• Drained Pipe – 4 Floors Below Roof

Attempting to identify the cut-off point at which the severe pitting stopped and higher more uniform wall thickness existed, UT testing still showed low wall thickness conditions 4 floors below the roof level. The risers were drained down to their final level based upon a rough estimate of static head pressure from basement gauges. Therefore we can speculate that final water level was not always the same, and that a variation in water level existed over the years. This pipe was shown in slightly better condition, but still existing below minimum standards and requiring replacement.

• Filled Pipe – Midpoint Of Building

Testing further down the risers again identified consistently good pipe at the return riser to the cooling tower, but not at the supply side piping.  We were able to establish that good quality return riser pipe existed at 5 floors below the roof level and below to the basement area. UT testing showed a very low corrosion rate and some mild pitting expected for a 55 year old piping system.  Average wall thickness far exceeded the 0.375 in. specification for new standard pipe, and some thickness measurements were found as high as 0.500 in. We were able to estimate virtually unlimited service life for this return side pipe.

• End Of Lower Horizontal Run

Surprising difference was consistently found at the supply downfeed to the refrigeration machines.  Unlike the high 0.450 in. wall thickness found at the return riser, our testing at the supply pipe found less than half that wall thickness and far less remaining service life. Pitting activity was slightly higher at the supply side piping as well. While some service clearly remains for the supply side riser piping, it provides far less reliable service and may not fulfill the long term needs of the new chiller plant, thereby requiring replacement.

• Complex Corrosion Problem Is Explained

As a result of this ultrasonic investigation, which due to its convenience and mobility allowed testing in areas possibly overlooked otherwise, a very severe deficiency at the top 5-6 floors of riser piping was detected.  Initially unexplained, the far lower wall thickness at the lower floors of the supply riser only was identified as due to a constant leak all winter at the pump packing glands.  This constant leak would drain down the supply side pipe through the pump, which due to its check valve at the discharge back to the roof, prevented the loss of water from the return side riser piping only.

Our ultrasonic testing well documented a very common problem for northern climates which drain their cooling tower piping for freeze protection.  It also showed that overlooked events as simple as a minor leak at a pump seal can produce enormous consequences given sufficient time.

In this investigation, we initially determined the top 5-6 floors of piping in need of replacement due to the effects of draining down.  Such losses were expected to cease at the point at which water level would be held full, as evidenced by the results from the return riser piping.  However, the totally unexpected consequence of slowly draining down the supply side piping, and exposing its entire length to the effects of drain down produced a lesser but still significant wall loss – raising the question of whether any of the original condenser water pipe would be suitable for future service.

Such events underscore the need to treat any ultrasonic evaluation as an investigation where results shape and change the direction of testing as progress is made.  This is substantially different from a procedural evaluation where testing is pre-defined at different areas based upon random thought rather than experienced prediction.

A contrasting look at this procedural difference is provided in this 2006 CorrView International, LLC article in World Pipelines.

© Copyright 2023 – William P. Duncan, CorrView International, LLC