| |
|
 |
|
 |
|
|
|

|


Numerous causes
will result in the failure of a threaded pipe connection - a moderate to high
corrosion rate being often responsible. One fundamental and very obvious
reason, of course, is the threading process itself - which removes 50% or more
of the pipe wall beginning day one. See Technical Bulletin
# P-1 about the effect of wall loss in threaded applications.
Other less common reasons include the
failure of the thread sealant, poor machining of the threads, poor quality of
the pipe or fittings, vibration, stress, improper assembly, or excessive
operating pressures beyond design. In many cases, a metallurgical analysis may
be required to identify the exact failure mechanism.

A major cause of
thread failure within a building or process plant environment is galvanic
corrosion - where the carbon steel pipe directly meets a brass valve, or is
transitioned to copper pipe. Here, the microvolt difference in electrical
potential of the metals will produce a small current between them - the result
of which is to greatly accelerate the deterioration of the more reactive and
often termed "less noble" carbon steel pipe.
In effect, an extremely small DC
electrical circuit is created, with the steel pipe serving as the anode, the
brass fitting or copper pipe acting as the cathode, and the water serving as a
weak wire connection completing the circuit. In simplest terms, a very weak
battery is created.
"Galvanic"
corrosion occurs between any two dissimilar metals in contact with each other
and water, and typically attacks the steel pipe to a degree somewhat dependent
upon existing corrosion conditions. Galvanic corrosion is, in fact, defined as
an electrochemical reaction of two dissimilar metals in the presence of an
electrolyte, typically water, and where a conductive path exists. It is
visually recognizable in its latter stages by some degree of deposit buildup
where the dissimilar metals meet at the threads - creating a microfine leak. At
that point, however, most of the damage has already occurred and replacement is
required.

The presence of
chemical water treatment, since it works by inhibiting the electrical pathway
of the corrosion process itself, can be very effective at slowing down a
galvanic condition. Substantial differences in the effectiveness to slow a
galvanic corrosion condition may exist between different chemical treatment
programs, and even the most well maintained program can still result in
threaded joint failures.
Galvanic
corrosion is, however, far less likely to cause problems where chemical
corrosion control is very well maintained, and where uniform corrosion rates of
1 MPY or less exist. A piping system having a low general corrosion rate will
often show no evidence of any galvanic condition at the hundreds of carbon
steel to brass valves normally in service. CVI has documented hundreds of
examples where no greater wall loss could be found at direct steel to brass
valve connections even after many decades. Closed piping systems, almost by
definition as having corrosion rates of 1 MPY or less, rarely show problems
attributable to galvanic activity - moving the greatest concern always to open
cooling tower systems and process
plants.
In contrast, a problem condition
at the pipe threads caused primarily by a high corrosion rate will be
substantially accelerated due to galvanic activity - thereby turning an already
bad situation even worse. In many cases, blame is incorrectly placed upon the
direct steel to brass connection as the source of problems, when a high
corrosion environment is fundamentally at
fault.
The cause of a thread leak can
often be identified by a close look at the threads themselves. The failure of
multiple sections of threaded pipe at steel to steel joints such as at elbows,
tees, and reducing bushings, without a significant increase of failures at the
steel to brass or copper joints, always suggests a moderate to high corrosion
condition as the fundamental cause. This is especially common for cooling tower
or open process water piping.
Conversely,
observing multiple examples of corrosion products at only the brass valved side
of a thread nipple would indicate that galvanic activity is the major force
involved. In many cases, however, a leak condition will involve some
combination of both galvanic activity and normal corrosion losses.

Pipe schedule is
very important wherever threaded piping is involved, and heavier schedule 80
pipe is always recommended where a higher corrosion rate might be expected -
such as at a steam condensate or condenser water system. Schedule 80 should be
specified exclusively for any threaded pipe serving a cooling tower or open
process water system due to the higher corrosion rates commonly found today -
this regardless of system operating pressure. Should a galvanic condition
exist, heavier pipe will offer longer service life to a degree again greatly
dependent upon corrosion activity.
Age is
also an important factor - since even a moderate corrosion rate will not
protect a decades old piping system from thread failure. A section of 2 in.
schedule 40 pipe having an initial wall thickness of 0.154 in., less its thread
cut of 0.085 in., leaves only 0.069 in. of available pipe wall for service over
its entire lifetime. At a moderate corrosion rate of 3 MPY, such pipe will last
roughly 20 years before completely wearing through. Yet, the first signs of
leakage can be expected years earlier - normally as the interior wall wears
roughly to within 15-25 mils of the
threads.
Internal pressure is still
another factor. While plant process piping may vary greatly in pressure,
cooling water systems usually operate at low to moderate pressures. Higher
operating pressures exceeding 300 PSI rarely exist except at many high rise
commercial office properties - but add greatly to the leak potential of any
threaded connection in such applications. Therefore, failures of any lower
level threaded piping should not be assumed an isolated event, but rather a
preview of a system wide weakness that will extend to the upper floor areas
given sufficient time.
The below gallery
of threaded pipe failures are due primarily to the galvanic action of the brass
or copper attacking the carbon steel pipe. Brass to steel problems far outweigh
that of copper to steel. For the overwhelming majority of examples shown, a
high corrosion condition exceeding 5 MPY was also found - greatly accelerating
the galvanic effect.
A general rule to
follow is that the worse a leak site looks, the worse it probably is. Once a
pinhole leak is produced at the threads as shown below, the leak often
temporarily seals itself. High corrosion is still proceeding its course,
however, and should not be assumed to have stopped or slowed. The consequences
of a 2 in. line failure operating at 200 PSI can be devastating to any building
property or plant operation in terms of water damage, and dictates the need to
immediately address any such problems found.


Galvanic activity
is well recognized as a serious potential threat to any piping system.
Generally, the use of electrically isolating fittings, called dielectrics, are
specified in the design and construction plans for all connections between
dissimilar metals at most piping systems. In reality, however, dielectric
fittings are rarely found - a fact only revealed years after construction when
a decade or more of advanced wall loss has taken place.
CorrView
International recommends establishing a strict specification for the use of
dielectric fittings, as well as a close review of all construction work to
ensure their proper installation. (The bottom right side photo represents an
inproper installation of the dielectric installed between the brass valve and
copper pipe, rather than between brass and black steel. Attack is always from
the brass or copper metal against the steel, and this is where electrical
isolation must exist.
Direct connections
of dissimilar metals should be upgraded wherever possible to include dielectric
fittings even though no external indication of a galvanic problem may
exist.
©
Copyright
|
|
|
|
|
|
|

|
|
|
 |
|
|