Guest Post Written by J. P. Schell, P. Eng (Retired)
On June 18, 2016, I received an
e-mail from Jim Gordon, P. Eng. (Retired) attached to which was a CBC News item
entitled “Nalcor investigates inconsistent patterns on Muskrat Falls transmission lines". Gordon, who has frequently commented on aspects of the
Muskrat Falls project on the Uncle Gnarley Blog, asked what I thought of the issue.
My comments are as follows.
From the pictures included in the CBC
News item it is apparent the outer strands of the line were squeezed too close
together, causing one strand to “pop out” by an amount which is about half the
strand diameter as shown in the photo.
(Source: CBC)
Transmission line conductors are
formed by wrapping aluminum wire strands around a central wire as illustrated in the image below. (Source: internet). Layers of strands are added, with the wrap direction
alternating to avoid line coiling, until the desired thickness is attained.
There are three situations where a
“popped out” strand could have occurred: a) during stringing, b) during
transportation, and c) in the factory. The CBC item indicated that
approximately 170 km of the DC transmission line had been strung before the
work was halted, which represents about 340 km of conductor.
During the stringing operation it is
difficult to understand how one strand of conductor could “pop out”. It might
be possible because of an accidental mishandling on one reel, but not on about
300 to 400 reels, one after the other. There is nothing in any stringing
procedure that can conceivably cause this to occur, reel after reel after reel.
That the problem occurred during transportation
is also difficult, if not impossible, to comprehend. Certainly, it will not
happen during normal shipping or even very rough shipping. If the reel breaks
in transportation, or pieces are damaged and fall off, the conductor is often damaged.
What does not happen to the conductor is a condition where over the full length
of the reel, one and only one strand “pops out”.
The only logical place for this to
happen is in the factory during the stranding procedure.
When each strand comes off its reel
or spool, it is fed into a stranding machine where it is preformed and coiled
around the layer below it, layer after layer, until the final layer is
completed. As the first conductor is formed, all the procedures are (presumably)
checked, measured, and corrected as necessary, until the output from the
stranding machine meets the specification. Then the production run can commence
after which, to a great extent, automation takes over.
Obviously, sometime
during the stranding procedure, something must have occurred that caused this
one strand to be incorrectly coiled and it continued, reel after reel after
reel, without stopping until at least 340 km of conductor had been fabricated.
Possibly it even continued for the full production run.
Now what?
If the conductor is replaced there
will be a significant delay of not less than 6 months, and possibly a year or
more. There will be a large cost associated with the defective wire which was
likely paid up front. In addition, protracted and expensive lawsuits may ensue.
The first question that needs to be asked is this: is this flaw acceptable? The
question that follows is: if it is not acceptable, can it be repaired in situ?
On the question of acceptability and in situ repair, I will confirm that I am
not a conductor expert, so my comments are opinions, not statements of fact. Nevertheless,
I would comment as follows:
At the location of the clamps and dampers this strand will tend to be squeezed,
distorted and flattened. As a result, the strand will be weaker than its
companion strands; it will be more prone to breaking. If it breaks, the
conductor’s overall strength will be correspondingly weakened and the broken
strand could slide or work its way out of the clamps. If that happens, it will
unravel.
Because the strand has “popped out”
of the conductor, it is probably very slightly longer than its companion
strands and therefore will tend to act somewhat differently. Rain or dew or
similar could get under the strand and freeze, thus pulling the strand even
further from the conductor, thus allowing weather better access to the inner
strands and isolating the strand even further from the conductor.
Because this strand has “popped out”,
it then becomes a target for being picked upon during the stringing process. If
anything adverse is going to happen, it will happen to this strand. In other
words, it is a target for distortion, for getting “in the way”, and for
breaking. I would not be surprised if reports were made of strands having
broken during stringing.
All conductors, if the wind speed and
wind direction are just right and are blowing along the strands, have a
tendency to gallop. Galloping occurs if the lifted strand funnels the wind just
enough such that the elongated shape becomes an airfoil — just slightly more
prone to fly.
Is there a solution to the problem?
First, I would state that I do not agree with a remedial action such as tightly
winding an aluminum tape or equivalent around the conductor in a direction
counter to the stranding. There are just too many things that could go wrong
with it.
This line will be in service for more
than 50 years — probably closer to 100 years — and under very severe weather
conditions. This one odd-ball strand
will be the one around which problems will tend to occur. For this reason, it would
be my opinion that the conductor should be replaced. For major transmission
lines, like those delivering power from Muskrat Falls, there is just too much
at risk if they are not.
Under the very reasonable presumption
that the flaw did originate in the factory, at the stranding machine, there are
many questions that need to be asked. All are related to two, which are
fundamental. First, why wasn’t it picked up sooner? And second: how was it that
a decision to stop the work wasn’t made until most, if not all, of the
conductor had been shipped and 170 km of the HVDC line was strung?
Under normal circumstances, it is
difficult to comprehend how 340 km (or even possibly the full production run)
of flawed conductor could be allowed to leave the factory. While it is
recognized that the process, once started, is essentially fully automated, it
is difficult to understand how reel after reel after reel could be stranded and
packaged and shipped without being picked up by normal factory quality
assurance procedures. One might accept the possibility of one or two defective reels,
but not more.
The factory process is basic. The conductor
for each reel is viewed by the factory work crews who attach the start
conductor to the reel, cut and fix in place the end conductor, and apply the
lagging. They would know it was a flawed conductor, particularly as the problem
occurred reel after reel after reel. Such a situation would not speak highly of
the supplier’s internal quality assurance procedures.
If, on the other hand, the factory QA
did pick up the flaw, then someone in management had to have let it go, knowing
that the flawed product would be shipped. It seems obvious that the shipment
would not have taken place if the purchaser’s inspector was knowledgeable and
doing a good job.
Presumably the purchaser (Nalcor, the
Owner? SNC-Lavalin, the Engineer?) hired an independent and qualified inspector
to undertake the factory inspection. If the supplier’s factory is located overseas,
or if the conductor’s place of origin is suspect, then the necessary for a
diligent inspection process is absolutely mandatory.
That said, 340 km (or more) of flawed
conductor having left the factory, it is apparent that the inspector, if there
was one, was lax in his undertaking. We might ask: What were his terms of
reference? Who wrote them? Nalcor? SNC-Lavalin? Was he supposed to inspect and
pass each reel before shipment? Did the inspector make daily visits? Weekly?
Monthly? Did he have the authority to stop production? Was he not required to make
the purchaser aware of the flaw? On what basis did he “approve” the conductor?
What did his inspection reports say? For that matter, do any inspection reports
exist?
It is indeed curious that even though
the supplier had to know that his conductor was flawed, it was shipped anyway,
and presumably paid for in full.
What was happening during stringing
is another story. It is impossible to believe that it wasn’t almost immediately
noticed, particularly when it was happening reel after reel.
Valard should have seen it first. Their
crews removed the lagging, connected the conductor to the pullers, manned the
reel brakes and, ultimately, took the conductors off the pulleys and attached
the clamps, spacer dampers and dead-ends.
Presumably, the Owner’s (Nalcor’s)
and/or the Engineer’s (SNC-Lavalin’s) inspectors were on site, along with Valard
QA personnel. Almost immediately, qualified inspectors would have recognized
the problem and informed the Owner. If an inspector wasn’t on site, the Owner’s
Engineer would have had to have been informed of that circumstance, too. Valard,
or any contractor, most assuredly would have reported this flaw if for no other
reason than that they would not want to be responsible for installing material
that they knew was faulty or potentially contained a serious flaw. It is simply
not in any contractor’s interest to install a product which they know or
suspect will have to removed and reinstalled at their own expense.
In short, from almost the very
beginning of the stringing process, the Owner’s Engineer at least had to have
been aware that something was not right with the conductor. Yet, the stringing
continued.
The disturbing conclusion is that Valard
had to have been told to continue stringing. Equally disturbing is that Valard
was likely informed that, though the conductor was flawed, the flaw was
acceptable and that Valard could continue stringing without consequence to the
Company. Logic dictates that the installation was approved in writing.
About 4–5 months after 170 km of line
was strung, the stringing was finally halted. The CBC News item does not say
what inspired this decision. Whatever it was had to have been serious. Likely,
the decision was made because the popped-out strand continually unravelled as the
stringing processed. It might have become obvious that the conductor — potentially
the entire shipment — was flawed, and the whole batch unusable.
Again, the role of the
Inspector/Engineer in the QA process is disturbing. So many questions demand
answers. Was the stringing being closely monitored? When did the unravelling,
or other problems, first occur? How often did it happen? Was it reported? If so,
how was it reported? Did it show up as just another item in the Inspector’s
routine inspection reports, or was it serious and passed up the line?
The stringing continued for several
months. The implication is that the Engineer did not consider the flaw to be
serious. If he did, he may not have insisted that the work be stopped, which, admittedly,
is a decision he has the right to make.
The role of the Owner was also
disturbing. The decision to continue with the stringing had to have originated
with the Owner. Was the decision the result of pressure to maintain the
schedule and to avoid cost overruns? Did no one pay attention to reparation
costs that would only increase as the stringing continued? Did no one consider
the inevitable problems that would occur having installed this flawed conductor?
This massive failure of Quality
Assurance can only enlarge an already serious problem of cost overruns on the
Muskrat Falls project. Mistakes of such magnitude, and the reasons why they
occurred, deserve proper airing. Those responsible must be held to account.
At the risk of repetition, perhaps
the CBC will ask Nalcor the following: First, what was the role of the Engineer/Inspector
in allowing, potentially, a full shipment of flawed conductor to leave the factory?
Second, what was the role of the Engineer in allowing a conductor, possessing a
flaw, to be continually installed for several months? And thirdly, what was the role of the Owner
in allowing a flawed product to be installed on his transmission line?
_______________________________________________________________________
Editor's Note:
J.P. (Joe)
Schell is a Professional Engineer with over 55 years of transmission line
experience with Shawinigan Engineering, Montreal Engineering [Monenco] including,
for 20 years, as a private Consultant. He has experience in all aspects of TL
design from initial studies and evaluations to detailed design and construction
supervision. Schell has worked on projects in all the Canadian provinces, the
Yukon and the NW Territories.
Internationally,
Mr. Schell has worked on TL projects in over 25 countries. He was the lead TL
engineer with ShawMont Newfoundland in St John’s for the design of the Bay
D’Espoir island wide transmission system, the lead designer for the 69 kV
Marble Mountain project for Deer Lake Power, and other lines in Newfoundland.
Other career
credits include lead transmission line engineer in the study stages of the Sete
Quedas 13,200 mw hydro development, Brazil. He was the lead TL design engineer
for a 400 kV super grid project in Iraq. Prepared contract documents for the
400 kV line from Shiraz to Sirjan in Iran and provided assistance to the World
Bank, in Washington, related to the rebuilding of the high voltage transmission
system in Lebanon.
Schell also provided design assistance to SNC-Lavalin in St
John’s for the early stages of the Muskrat Falls hvdc project.