-
REFEREED PAPER
LEARNINGS FROM THE 2015 PONGOLA SILO FAILURE
LAWLOR WK
RCL Foods, Westville, South Africa
[email protected]
Abstract
During June 2015 the refined sugar silo at the Pongola Sugar
Mill suffered a severe buckling failure. The failure occurred with
the silo in operation and full of sugar. During the months which
followed the silo was stabilised, strengthened, the sugar was
removed, the damaged sections were safely dismantled and a thorough
investigation into the cause of the failure was undertaken. This
paper reports on the steps taken to safely dismantle the silo and
on the various mechanisms by which silos can fail which were
considered during the investigation. Most importantly, the paper
provides a list of recommendations to be followed to reduce the
likelihood of future silo failures. Keywords: silo failure, silo
stabilisation, silo strengthening, silo design, buckling strength,
compression buckling
Literature search
The only mention of a silo damage in the SASTA proceedings was
in 1992 when Saunders RR described how the incorrect operation of
the dust extraction fans during the filling up stage caused one of
the Noodsberg silos to implode. The extent of the damage was not
described, nor was any mention made of the repair of this
implosion.
Introduction The Pongola refined sugar silo and service tower
were constructed in 2005/2006 and have been in operation since
March 2006. The silo was constructed from rolled 3CR12 plate welded
into strakes and installed on a concrete base. The penthouse
contained a concrete floor and roof. The service tower, constructed
from carbon steel, contained a bucket elevator which fed sugar to
the top of the silo where a distribution system evenly distributed
the sugar into the silo via a distributor and 12 inlet pipes. The
silo was used to condition sugar by passing dry, warmed air upwards
through the sugar which was constantly discharged from an inverted
cone base through 12 discharge pipes. The shell of the silo
consisted of five sections of different thicknesses; 16 mm at the
base, decreasing to 12 mm, 10 mm, and 8 mm in the middle and 6 mm
at the top. The silo was 43.5 m tall and had a diameter of 8.8 m
and was lagged with 100 mm of phenolic foam and corrugated sheeting
in order to keep the internal temperature as constant as possible.
When full the silo could contain 2 000 tonnes of sugar.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
524
-
Figure 1. The Pongola Refined Sugar Conditioning Tower and
Service Tower before the failure
In July 2015, buckles in the lower half of the service tower
were observed which prompted an inspection of both the service
tower and conditioning silo. From the deformation on the brackets
which connected the service tower to the silo it became clear that
it was the silo which had failed and was pulling the service tower
over. Closer examination of the silo revealed a definite “kink”,
visible approximately half way up the steel shell even though the
insulation was still in place.
Figure 2. Tell-tale buckles in the service tower as a result of
the silo pulling it over
Once this had been determined, a surveyor was commissioned to
take readings of the top of the silo every 12 hours in order to
detect if any further movements were occurring. These readings
indicated that no further movement was occurring. In order to
reduce the load on the silo and to enable an internal inspection,
an attempt was made to remove the sugar from the silo through the
discharge system. From the surveyor’s measurements, it immediately
became apparent that as the sugar level dropped the silo leaned
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
525
-
over more in a direction away from the service tower. At this
point the silo was taken out of service. The surveyor was able to
determine that the top of the silo had deflected horizontally by
approximately half a metre. Given that a cylinder is such a rigid
structure, this revelation which was not apparent to the naked eye,
suggested that the silo had suffered a massive deformation at the
kink point. Following a brief review of the strength of buckled
cylinders and failed silos, it became clear that the silo had
suffered a serious deformation and would certainly have collapsed
if the sugar it contained was not “holding it up”. The contents of
a silo offer a path for the transfer of force across a weak point,
however, there was no way of knowing how close the silo was to a
total and catastrophic collapse. Within the factory the silo is
located in a very central position. A 50 m radius fall zone was
identified and this included the packing station, the evaporator
station, molasses tanks, boiler water filter station, and the main
cane truck entrance roads.
Figure 3. Fall zone of the silo
At this point the silo and Pongola Sugar Mill were in a very
precarious position. It was not obvious what the best way forward
should be. Without the sugar inside the silo, it would very likely
not be self-supporting, but without removing the sugar the silo
could not be dismantled. To make matters worse, no hot work could
be carried out on the silo because of the danger of sugar dust
explosion. The use of a large crane to hold the silo up was
investigated, however, no crane large enough could be sourced, and
no crane operator was prepared to work with such an unstable
structure. No work was allowed on the silo, the fall zone was
demarcated with danger tape with access for essential personnel
only, and the cane truck road was re-routed. At this point the most
attractive suggestion was to evacuate the factory and use a bull
dozer in an adjacent sugar cane field to pull the silo over.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
526
-
Steps Taken to Safely Remove the Sugar and Dismantle the Silo
Collaboration Once the severity of the situation became apparent,
it was decided to call upon the best and most experienced brains in
the business to provide input into the decision making process.
This included input from Professor Michael Rotter, an expert in the
field of silo failures, structural design experts, rigging
specialists, health and safety practitioners, and mechanical and
process engineers. This collaboration took the form of face to face
meetings, brainstorming, sharing of past experiences and technical
calculations. Following this process a plan of action was
established with a view to: Stabilising the silo to allow work to
commence on and around it; Removing the lagging for inspection and
strengthening of the “kink”; Removing the sugar; Recovering the
sugar into one tonne bags; and Safely dismantling the silo in
preparation for a rebuild. Stabilisation As an initial step towards
stabilising the silo, it was decided to install guy cables
connecting anchor points to both the silo and service tower. The
intention of the guy cables was to provide horizontal support to
the silo and service tower, and to provide a mechanism for
continuous monitoring of the movement of the silo as indicated by
the readings on the load cells installed on certain critical guy
cables. The readings from the load cells were incorporated into the
factory DCS system to allow continuous monitoring of the loads on
the cables. Any alarm condition would activate a siren and initiate
the evacuation of the fall zone. Five anchor points were
constructed: four of them consisting of brackets cast in concrete
blocks, and one of them fabricated from structural steel within the
factory. These anchor points were connected via cables to brackets
at the top of the silo bolted through the concrete penthouse roof
and to a collar bolted around the silo just below the kink. The
anchor points were also connected to brackets bolted to the top of
the service tower. The guy cables were selected to have a breaking
strain of 26 tonnes each so that if the silo started to fall over,
the cables would snap before the concrete anchors were pulled out
the ground causing secondary damage to the factory. Removal of
Lagging Once the guy cables had been installed and the load cells
outputs were providing continuous monitoring of the cable tensions,
it was decided that the risk of sudden unexpected collapse had been
mitigated and it was now safe to work on and around the silo. An
external scaffold was erected around the silo up to the kink, and
the lagging was removed to reveal the failure.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
527
-
Figure 4. Sketch of anchors and cables for stabilisation of the
silo and service tower
Figure 5. Anchor for stabilisation cables and load cells
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
528
-
Figure 6. Silo and service tower with cables installed
Figure 7. Output from one of the load cells showing variation in
tension between day and night
A massive compressive buckle was found to have formed at the 24
m height over 270 degrees of the circumference of the silo. Only
the side adjacent to the service tower was not buckled, which
suggested that the service tower had offered support to the
silo.
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
2016/01/26 00:00 2016/01/31 00:00 2016/02/05 00:00 2016/02/10
00:00 2016/02/15 00:00 2016/02/20 00:00
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
529
-
The severity of the buckle immediately confirmed that without
the sugar in the silo it would not be self-supporting. It was
therefore necessary to design and install an external brace, or
exoskeleton, to provide sufficient strength to support the mass of
the silo above the buckle so that the sugar could be removed.
Figure 8. Lagging removed showing part of the buckle in the silo
shell
Installation of Exoskeleton The exoskeleton, which consisted of
brackets, vertical columns and circumferential collars, was riveted
into place with no hot work undertaken on the silo shell. The
exoskeleton was designed to allow the silo to be self-supporting
without the help of the sugar it contained. This would allow the
silo to be emptied of sugar so internal inspections could be
undertaken, and it would allow hot work to be undertaken on the
silo shell. The limited working range of the rivets used required
the brackets to be completely flush with the shell of the silo.
Given that the silo was round and deformed, and the brackets flat,
each bracket had to be scored, bent, shaped and welded individually
to suit its position. A further challenge was the rigging into
place of the 6 m long vertical columns. This was achieved by
removing the scaffold boards and lowering the column through the
lattice structure of the scaffold. When it was at the desired
height, the scaffold ledgers were removed to allow the column to be
tacked onto the brackets. Once the columns were in place the
collars and cross members were installed by welding them onto the
columns.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
530
-
Figure 9. Silo with “exoskeleton” installed
Removal of sugar One of the forces a silo needs to withstand is
the vertical downward drag force of the product as it flows
downwards. This force can be approximated by considering the
pressure force of the product and the coefficient of friction
between product and silo shell. However, this is very much an
approximation and there is much uncertainty in the result. Because
of this uncertainty, and the limited strength of the exoskeleton,
it was determined that removing the sugar from the existing
discharge system would be too risky, and an alternative method of
sugar removal was sought. It was determined that the sugar, which
was free flowing and well-conditioned, could be removed through two
inch holes mechanically cut into the shell of the silo. By cutting
these holes in progressive rings 1 m below the sugar level, all
vertical drag forces could be minimised.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
531
-
Temporary funnels fashioned from tin sheeting, pop rivets and
duct tape were used to direct the sugar into plastic irrigation
hoses which transported the sugar by gravity to the ground level.
The sugar was therefore drained out of holes, 1 m at a time, until
the sugar level was below the level of the buckle. This process
took several days to complete. At this point, the top half of the
silo was supported by the exoskeleton and because the lower half
was not damaged, the remaining sugar could be safely discharged
from the sugar discharge system.
Figure 10. One of the two inch holes and funnels used to extract
sugar from the silo
Recovery of sugar At the time of decommissioning the silo it
contained approximately 1 800 tonnes of sugar. This sugar, although
well-conditioned, had been contaminated by the drilling of the
holes for the exoskeleton attachments and the cutting of the holes
for the removal of the sugar. It was decided to recover this sugar
in one tonne bags so that it could be easily transported and
reprocessed. A temporary one tonne bagging station was set up
outside of the fall zone (50 m from the silo), and temporary
conveyors were used to transport the sugar from the outlet of the
hoses to the bagging station. The silo was finally safely emptied
eight months after it had been taken out of service. Dismantling of
the silo Once the silo had been made self-supporting and emptied of
sugar, hot work could be undertaken, and cranes with sufficient
capacity were readily available so that the dismantling of the silo
could be undertaken without further complications.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
532
-
Figure 11. Temporary conveyors transporting sugar to the one
tonne bagging plant
Figure 12. Using cranes and hot work to dismantle the emptied
and strengthened silo
Comments As far as the author and all the local and
international experts involved are aware, this is the first time
the above procedure has been undertaken in order to safely
stabilise, support, empty, and dismantle a damaged and unstable
silo, whilst still being able to recover the sugar. This silo
failure did not end in a catastrophic collapse, however, given the
severity of the buckle, it could easily have done so if the correct
experts had not been consulted, and if the correct decisions had
not been taken.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
533
-
Investigation into the Cause of the Silo Failure Running in
parallel with the stabilising, emptying and dismantling of the silo
was an extensive investigation into the cause of the failure. There
are several possible mechanisms which can lead to the failure of a
silo. Most of these are highly complex and require both
mathematical modelling and empirical results to fully describe. In
the following sections these mechanisms will be introduced, simply
described and their basic concepts discussed. Defective Design In
order to design a sugar silo the loads on the silo must be properly
understood. The silo must support its own weight and that of the
penthouse and feed equipment, and it must accommodate the forces
imposed on it by the sugar it contains. The sugar exerts both
normal pressure and frictional drag forces on the walls of the
silo. The normal pressure increases with depth but, unlike a fluid,
the increase of pressure with depth is not linear but tapers to an
asymptotic value. The normal pressure from the sugar is resisted by
circumferential or hoop stress in the silo walls.
Figure 13. Sugar pressure on the walls of a silo (Rein 2007)
The frictional drag forces on the walls of the silo induce
vertical compressive stress in the silo walls which are cumulative
below the level of sugar.
Figure 14. Calculated compressive and hoop stress in the Pongola
silo (Juvinall 1991)
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
534
-
Silos are susceptible to buckling, not only because they are
slender structures, but because of inevitable or unavoidable
imperfections in their wall geometry. An out of round of one wall
thickness can reduce the silo’s resistance to buckling by
approximately 70 %. Based on the expected depth of imperfections,
the vertical compressive stress at which a silo buckles can be
calculated using the appropriate design code, and it can be found
to be as low as 10 % of the material yield strength.
Figure 15. Graph showing the effect on buckling strength of
imperfections in the shell (Sodowski 2011)
For this reason it is the vertical compressive stress in the
silo walls which determines the wall thickness rather than the hoop
stress, yielding or bursting. The detailed design should also take
into account ground conditions, wind and seismic action and should
be in accordance with the appropriate standards which in South
Africa are: a) SANS 10160-1: Basis of structural design; b) SANS
10160-2: Self-weight and imposed loads; c) SANS 10160-3: Wind
actions; d) SANS 10160-4: Seismic actions; e) SANS 10160-5: Basis
for geotechnical design and actions; f) SANS 10160-6: Actions
induced by cranes and machinery; g) SANS 10160-7: Thermal actions;
and h) SANS 10160-8: Actions during execution. These codes are
based on the Eurocodes but are notably reduced in content. By SANS’
own recommendation where these codes are lacking or deficient, the
Eurocodes should be referred to. It is noted that the South African
Codes do not specifically consider silos and, therefore in the case
of a silo design at least the following Eurocodes should be
referred to: a) EN 1991 – 4, 2006. (Provides material properties of
sugar); b) EN 1993 – 1 – 6, 2007; and c) EN 1993 – 4 – 1, 2007.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
535
-
It is interesting that in 2007 the Eurocode was updated (by a
team led by Professor Rotter). These updates limited the aspect
ratio of silos to less than three to one and also outlawed the use
of inverted cone discharges. Defective Fabrication or Construction
Apart from obvious errors such as the use of the wrong thickness
plates, the buckling resistance of a silo depends strongly on the
quality of the fabrication. As shown in Figure 14, the most
important feature of fabrication quality affecting buckling is
small deviations from perfect shape. These imperfections can be
introduced during the construction of a silo and can be as a result
of bolted lap joints, weld shrinkage depressions, local flats near
welds and ovalling of circular strakes. A certain amount of
construction damage/imperfection is an unavoidable result of a
typically complex construction methodology, and it is accommodated
specifically in the design codes but needs to be managed closely.
For this reason a construction methodology which minimises this
risk and careful quality inspection are important during the
construction of a silo. Wind Action When wind passes over a silo it
produces a force in the downwind direction that is proportional to
the wind speed and the projected area of the silo. This force
results in a bending moment which the silo and its anchors must
contain, and it can result in wind buckles which typically develop
in the thinnest strakes and propagate downwards into the thicker
strakes. This bending moment will also contribute to the
compressive stress in the downwind silo wall.
Figure 16. Indicative sheer force and bending moment caused by
wind
SANS 10160 - 3 proposes that a one in 50 year wind speed (with
gusts) should be used as the design load to be applied. It is
interesting to note that when the silo strake thicknesses are
selected on buckling resistance it is likely the silo will be
strong enough to easily survive such a wind.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
536
-
Figure 17. Map of fundamental value of the Basic Wind Speed from
which the velocity of a wind
with a mean return period of 50 years is determined (SANS 10160
– 3, 2011)
A less obvious effect of wind action is vortex shedding. When
wind passes a silo, because it has viscosity it compresses and
slows down and effectively “sticks” to the silo. This is referred
to as the boundary layer. As the wind passes the silo, because of
the curvature of the silo wall this boundary layer separates from
the leeward side of the silo. As this separation occurs vortices
are formed on either side of the silo which produce periodic forces
on the silo perpendicular to the wind direction. This phenomenon is
known as vortex shedding and it is the reason flags flutter as the
wind passes their flag pole.
Figure 18. Diagram showing vortex shedding
(www.wikipedia.org)
The frequency at which a silo will shed vortices is a function
of the wind speed, the silo diameter and the Strouhal Number.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
537
-
The alternating forces from vortex shedding are not massive,
however, if the frequency of their oscillations match the natural
frequency of the silo it is possible for the silo to be damaged by
its own resonance and the progressive input of energy over time. A
silo has two modes of resonance, namely its natural bending
frequency and its natural ovalling frequency.
Figure 19. Diagram showing the mode of vibration for resonance
at the natural ovalling frequency
Excitation at the natural ovalling frequency over an extended
period of time will result in the resonance of the silo which may
lead to fatigue cracks in the silo walls where the ovalling
movements are the greatest. Excitation at the natural bending
frequency over a period of time will result in the resonance of the
silo which will cause it to swing from side to side, which will
introduce an alternating compressive stress in the silo walls which
could lead to a buckling failure. Again, it is interesting that
with typical silo dimensions, winds in excess of a one in 50 year
wind are required to achieve vortex shedding at a frequency that
matches the typical natural frequency of a silo.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
538
-
Figure 20. Diagram showing the mode of vibration for resonance
at the natural bending frequency
Force Applied by Service Tower Typically, sugar silos are built
with an adjacent service tower of smaller diameter which contains a
bucket elevator to transport the sugar to the top of the silo and a
spiral staircase for access to the top of the silo. The service
tower is typically connected to the silo via a short walkway. This
walkway can be connected to both silo and service tower via rigid
connections or via flexible/pinned joints. Because the silo is
insulated and temperature controlled and the service tower is not,
and is exposed to the sun, the service tower and silo will attain
different temperatures and therefore differential thermal expansion
will occur. Where the service tower and silo are constructed of
different materials this effect can be increased. Where the
interconnection is rigid differential thermal expansion will cause
stresses in both the service tower and silo. Since the wall
thickness of the silo and service tower is at its least value at
the top, this stress is likely to be relieved by very local
deformation in the vicinity of the walkway. Although these loads
are unlikely to be large enough to cause serious damage to the
structures, the effect of this differential expansion combined with
other loads (wind, seismic, normal operation) could lead to
unwanted damage, and therefore rigid walkway connections to the
silo and service tower should be avoided.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
539
-
Figure 21. FEA model showing the stresses in the silo as a
result of differential thermal expansion
Eccentric Discharge Eccentric discharge is the term used to
describe the discharge from a silo from one side only, when the
sugar flows preferentially down one side and remains stationary
throughout the rest of the silo. This could be caused by deliberate
discharge from one side only, blocked discharge pipes, or by small
variations in sugar moisture, temperature, or grain size. The
effect of this is that the pressure from the sugar exerted on the
walls of the silo is high where the sugar is stationary, and
notably lower where the sugar is flowing.
Figure 22. Eccentric flow channel and the resulting pressure
pattern
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
540
-
In response to this uneven pressure distribution, the silo shell
is deformed into an out of round shape with partial flattening of
the shell adjacent to the flow channel. This deviation from the
perfect shape drastically reduces the buckling strength of the silo
and if severe enough, can reduce the silo’s buckling strength to a
value below the compressive load it is experiencing in normal
operation. This would lead to an elastic buckle which would grow
fairly rapidly and thereby transfer stress into the adjacent shell
sections which would also buckle. This would result in a series of
buckles growing from the first buckle circumferentially around the
silo and can lead to total collapse of the silo. Because the middle
of the silo is furthest from the stiffening effect of the roof and
floor, and therefore more susceptible to being pulled out of round,
this type of failure often occurs at the mid-way point regardless
of the plate thicknesses.
Figure 23. Picture of compressive buckling failure caused by
eccentric discharge
Seismic Action Seismic events result in horizontal and vertical
displacements and accelerations of the base of a silo. The silo and
its contents have mass and inertia which resists these
accelerations, according to Newton’s Second Law of Motion (Force =
Mass x Acceleration). These forces produce bending of the silo as a
vertical cantilever, causing vertical compressive forces on one
side, and vertical tensile stresses on the opposite side of the
silo. These forces increase from the top of the silo to its base.
Clearly, the acceleration from a seismic event acting on a silo in
operation can impart additional vertical stress which, when
combined with the existing vertical stresses, may cause the silo’s
buckling strength to be exceeded and cause the silo to buckle and
collapse. SANS 10160-4 contains a seismic map of Southern Africa
which provides the maximum acceleration to be used in the design of
structures depending on their location.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
541
-
Figure 24. Seismic Hazard Map showing peak accelerations with a
10 % probability of being exceeded in 50 year period (SANS 10160 –
4, 2011)
A second mechanism by which a seismic event can damage a silo is
by seismic excitation or resonance. If a seismic event contains
vibrations whose frequency matches the natural bending frequency of
the silo, the silo will resonate. The overlap of seismic
frequencies with the natural frequencies of the silo, and the fact
that the progressive input of energy over time can cause the
amplification of vibrations above ground level, may lead to large
oscillations. If the vertical compressive stresses associated with
these oscillations exceed the buckling strength of the silo, it
will buckle and fail. Therefore a seismic event which does not
contain sufficient acceleration to cause damage to a silo may, if
its duration is long enough, damage it and cause it to collapse by
causing it to resonate at its natural frequency. The most effective
defence against seismic action is the inclusion of damping into the
structure. While the levels of seismic activity in Southern Africa
do not require this consideration, in parts of the world where
seismic activity is more prominent, damping should be strongly
considered. It should be noted that the area of earthquake analysis
is a highly specialised field and the outcomes can be uncertain.
Vacuum Implosion Silos typically operate with several large fans.
These include forced draft, induced draft and dust removal fans. If
for any reason the fans are out of balance and a vacuum is
established inside the silo, the walls of the silo can be sucked in
and buckle. There is a documented case of this happening at the
Noodsberg Silo during its commissioning (Sanders, 1992).
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
542
-
Figure 25. Frequency analysis of June 2016 seismic event
compared to the natural bending frequency of the Pongola silo
(South African Council of Geoscience)
Sugar Dust Explosion The literature search revealed only one
documented sugar dust explosion in Southern Africa. This occurred
at Mhlume on 25 June 1997 (Dale and Knoetze, 1999). The explosion
occurred in the silo outfeed bucket elevator located in the silo’s
service tower. The exact cause of the explosion was not
conclusively determined but the damage to the bucket elevator and
service tower was extensive. Given the volume contained in a sugar
silo it is clear that any explosion or fire inside a silo would be
catastrophic and potentially cause severe damage to the silo and
its surrounds. So What Caused the Pongola Silo Failure? The damaged
Pongola Silo was analysed extensively in order to determine the
most likely cause of the failure. It was inspected by experts in
the following fields: Structural Engineering; Metallurgy; Silos;
and Welding. The following tests were undertaken on material and
samples cut from the silo: Material thickness tests; Material
property tests; Weld quality tests; Finite Element Analysis; and
Laser scans of the inside to determine out of round. The following
data was gathered and reviewed: Design records; Operational
records; Maintenance records; Inspection records; Construction
records;
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
543
-
Wind data from the South African Weather Services; and Seismic
data from Council for Geoscience Applied Science Solutions.
Following all of the inspections, tests, data review and analysis
of the various experts’ opinions, the following sequence of events
was agreed to be the most probable cause of the failure. Since the
silo survived the strongest seismic event of its life on 05 August
2014 it must have been structurally sound at this time. This was
confirmed during the 2014/15 offcrop inspection. It is believed
that following the start of the processing season in 2015 an
eccentric discharge occurred. This event was severe enough to cause
a stable buckle to form in the shell of the silo at the mid height
position. It is expected that this buckle would have been only two
or three plate thicknesses deep. This was deep enough to
significantly reduce the buckling resistance but not deep enough to
cause the silo to fail in normal operation. This buckle was hidden
by sugar on the inside and lagging on the outside so went
unnoticed. A seismic event on 16 June 2015 occurred during normal
operation, and although it can be shown by calculation that it was
not strong enough to cause a healthy silo to fail, it can be shown
by calculation that this seismic event was strong enough to cause a
silo weakened by a stable buckle to fail in buckling. The cause of
failure was therefore a seismic event acting on a weakened silo,
weakened by a stable buckle formed previously by an eccentric
discharge event.
Conclusions The following is a list of recommendations based on
the learnings from the Pongola silo failure and resultant
investigation designed to reduce the likelihood of future silo
failures:
Silo design should be in accordance with all applicable SANS
codes and the Eurocodes which specifically address silo design;
All possible loading conditions should be understood and
considered in silo design;
Construction methodology should minimise the likelihood of out
of round imperfections;
Quality control during construction should focus on out of round
inspection;
Quality control during construction should confirm the correct
plate thicknesses are used at the various levels;
Any connections to the silo, for example, from the service
tower, should be flexible and allow for thermal expansion;
Instrumentation on the discharge of silos should measure the
flow of sugar through each discharge pipe and alarm and trip when
the flows differ or pipes block completely. This is to prevent the
occurrence of eccentric discharge;
Annual internal inspections should focus on identifying any
minor buckles, dents or flattening. Laser scanning can be
undertaken to achieve this;
Annual inspections should also measure silo shell plate
thicknesses to confirm they have not been reduced by erosion or
corrosion;
Any minor buckles, dents or flattening should be repaired using
a suitable repair procedure which reinstates as far as possible the
original round shape of the shell;
Operators should be aware that the incorrect operation of the
silo fans can lead to implosions and damage, and therefore strict
fan operating procedures should be established and adhered to. All
vacuum breakers (and explosion doors) should be well maintained and
regularly confirmed to be in good condition;
If a kink in a silo is identified the silo should be taken out
of service immediately but the sugar should not be removed until it
is certain the silo is able to support its own weight;
In the event of a kink in a silo, extreme caution should be
undertaken as this represents a very unstable structure which could
collapse at any time;
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
544
-
All precautions to prevent sugar dust explosions should be
followed (See Dale and Knoetze, 1999); and
The fields of silo design and failure are fairly specialised and
if some of the fundamental issues are not fully appreciated the
wrong decisions can be taken. Therefore specialists in these fields
should be accessed without hesitation.
Acknowledgements
The author would like to thank the following for their
invaluable contribution to averting a catastrophic disaster and
achieving a safe and successful outcome. Professor Michael Rotter;
Lovemore Riggers and Site Works; DRA Structural Engineers; Pongola
Metal Works; Rodcol Civils; Pongola Factory Management; South
African Weather Services; and Council for Geoscience Applied
Science Solutions
REFERENCES
Dale TB and Knoetze TP (1999). Sugar dust explosion at Mhlume: A
case study. Proceedings of the
South African Sugar Technologists Association, June 1999 (Paper
73:289 to 295).
Rein P (2007). Cane Sugar Engineering, Verlag Dr. Albert Bartens
KG, Berlin, Germany.
Juvinall RC and Marshek KM (1991). Fundamentals of machine
component design, John Wiley and Sons Publishers, Singapore.
Rotter JM (1985). Buckling of ground supported cylindrical steel
bins under vertical compressive wall loads. Metal Structures
Conference 1985 Melbourne.
Sadowski AJ and Rotter JM (2011). Steel silos with different
aspect ratios: II behaviour under eccentric discharge, Journal of
Construction Steel Research, 67(10), 1545 – 1553.
Sanders RR (1992). A decade of refining at Noodsberg.
Proceedings of the South African Sugar Technologists Association,
June 1992 (177 to 181).
SANS 10160-3 (2011). Basis of structural design and actions for
buildings and industrial structures: Part 3: Wind actions, South
African National Standards, SABS, Pretoria.
SANS 10160-4 (2011). Basis of structural design and actions for
buildings and industrial structures: Part 4: Seismic actions and
general requirements for buildings, South African National
Standards, SABS, Pretoria.
Lawlor WK Proc S Afr Sug Technol Ass (2017) 90: 524-545
545
Cover pageContentsSearchDisclaimer 2017Officers-SASTA 2017SASTA
Instructions to Authors 2017Prize Award Winners 2017Editorial Panel
2017Sponsors and ExhibitorsPlenary Session (Chair: Carolyn
Baker)Review of South African sugarcane production in the 2016/2017
season: light at the end of the tunnel?Ninety-second annual review
of the milling season in Southern Africa (2016/2017)A financial
estimation of the mill area-scale benefits of variety adoption in
South Africa: A simplistic approach
Agriculture Session 1: Entomology (Chair: Des Conlong)Cacosceles
(Zelogenes) newmani (Thomson) (Cerambycidae: Prioninae), a new pest
in the South African sugar industryThe effect of an improved
artificial diet formulation on Eldana saccharina Walker rearing,
growth and developmentEstimating the potential economic benefit of
extending the harvesting cycle of dryland coastal cane by
chemically suppressing eldana levelsA cellular automaton model for
simulating Eldana saccharina infestation in sugarcaneTimeframe for
the development of borer resistant genetically modified
sugarcaneTowards optimising crop refuge areas in transgenic
sugarcane fields
Agriculture Session 2: Soils and Nutrition (Chair: David
Sutherland)The fertility status of soils of the South African sugar
industry – 2012 to 2016: an overviewMass and composition of ash
remaining in the field following burning of sugarcane at
harvestEffects of surface-applied lime and gypsum on soil
properties and yields of sugarcane ratoon cropsPrediction of soil
nitrogen mineralization to crop fertiliser nitrogen
requirementsFactors controlling the solubility of phosphorus in
soils of the South African sugarcane industry
Agriculture Session 3: Agronomy (Chair: Sanesh Ramburan)Analysis
of long term rainfall in the Felixton Mill supply area and
investigation of Derivatives as a hedging mechanism against
droughtAn experimental and crop modelling assessment of elevated
atmospheric CO2 effects on sugarcane productivityThe investigation
of a suitable summer breakcrop after Imazapyr application for
integrated management of Cynodon dactylonNitrogen use efficiency of
selected South Sfrican sugarcane varietiesA web-based decision
support tool for analysing monthly sugarcane growth rates in South
AfricaMycanesim® Lite: A simple web-based sugarcane simulation
toolOptimum harvest age of sugarcane at Kilombero Sugar Company
under high minimum temperature
Agriculture Session 4: Plant Breeding I (Chair: Kerry
Redshaw)The effect of Eldana saccharina damage on sugarcane
breeding populations and the implications on sugarcane
breedingIdentifying elite families for the Midlands sugarcane
breeding programmes in South AfricaMolecular phylogeny of
sugarcane: Discovering a new speciesEffect of self-trashing on
Eldana saccharina Walker damage in sugarcane and implications for
resistance breeding
Agriculture Session 5: Plant Breeding II (Chair: Derek
Watt)Performance of imported genotypes and implications for
utilisation in SASRI breeding programmesThe agronomic performance
of tissue culture (NovaCane®) versus conventional seedcane under
rainfed conditionsAn investigation into stored seed viabilityA new
origin of sugarcane: The undiscovered species
Agriculture Session 6: Engineering (Chair: Peter Lyne)Modified
"Twin-stacker" cane loading systemPBS vehicles in the South African
sugar industry: opportunities and limitationsA simple
spreadsheet-based irrigation electricity cost calculatorYield
variability mapping for a cut and stack system
Agriculture Session 7: Crop Management (Chair: Rowan
Stranack)Irrigation scheduling demonstration trials are an
effective means of promoting adoption: Pongola case studyPositive
influence of Demonstration Plot Extension Methodology in a rural
sugarcane communityHere, there or everywhere? An investigation into
nematode trial sampling
Agriculture Session 8: Economics (Chair: Kathy Hurly)Determining
the cost of post-harvest deterioration in a South African sugarcane
supply chainCaneTEC®: An economic conversion tool for sugarcane
experimental and commercial production scenariosA new
decision-making framework for developing variety-specific chemical
ripening recommendationsCost benefit analysis of a herbicide
tolerant and insect resistant genetically modified sugarcane
variety under coastal conditionsBiogas from sugarcane - a system
for sustainabilityA time-series analysis of large-scale grower
input costs in the South African sugarcane industry: 2000/2001 -
2014/2015
Factory Session 1: Energy (Chair: Nico Stolz)A strategy for
monitoring and reporting continuous energy consumption in a typical
raw sugar millExperiences of reducing the steam consumption in
sugar plantSolar live steam generation and solar bagasse drying for
South African sugar mills
Factory Session 2: Milling and Diffusion (Chair: Warren
Lawlor)"Sleeve-Kamal" an innovative three piece sugar mill roller
for high performance and lower operating costMonitoring juice
holdup in a cane diffuser bed using electrical conductivity -
evaluation on a laboratory scaleMonitoring juice holdup in a cane
diffuser bed using electrical conductivity - evaluation on a plant
scaleExperiences with the millability of drought-affected cane
varieties for the 2016 season
Factory Session 3 papers do not appear in the Proceedings as
they were non-refereed commercial talksFactory Session 4: Rawhouse
(Chair: Paul Schorn)An investigation into the viscosity of
c-massecuite using a pipeline viscometerDynamic simulation on a
spreadsheet as a tool for evaluating options for mixed juice flow
controlAre gums produced in the factory? Quantification of gums
isolated from mixed juice and final molasses
Factory Session 5: Posters (Chair: Dave Love)Can NIRS detect
quaternary ammonium compounds in refined sugar?A benchmark energy
indicatorAnalysis of sulphites in sugar by ion chromatographyAn
effective viscosity modifier for improved production outputAnalysis
of Vitamin A in fortified sugarFactory control using NIRS: Are we
there yet?The effect of rotoclone bacterial slime on the refined
sugar turbidity increase experienced at the Noodsberg refinery
Parallel Session: Sugarcane Biorefinery and Downstream Products
(Chair: Anne Stark)Lignocellulose biorefineries as extensions to
sugar mills: Sustainability and social upliftment in the green
economyThe development of a partial equilibrium economic model of
the South African sugar industry in a biorefinery scenarioAn
economic analysis of the potential bio-polymer industry: the case
of sugarcaneEconomic recovery of biobutanol - A platform chemical
for the sugarcane biorefineryReactive extraction and reactive
distillation: A new recovery process development for levulinic acid
from fermentation brothsNitrogen-doped carbon nano-tubes synthesis
from biorefined sugarcane bagasseOrganic acid treatment of
sugarcane residues for the production of biogenic silicaThe
development of a screening tool to identify new products for the
South African sugarcane industryInclined perforated drum dryer and
separator for cleaning and drying of sugarcane bagasseConversion of
sugarcane bagasse into carboxymethylcellulose (CMC)Preparation and
characterisation of cellulose nano crystals (CNCs) from sugarcane
bagasse using ionic liquid (1-butyl-3-methyllimidazolium hydrogen
sulphate)-DMSO mixturesSugar cane juice concentration and
separation with hydrate technology
Factory Session 6: Refinery I (Chair: Steve Davis)Energy
footprint and operating costs, a comparison of ion exchange resin
and activated carbon in the application of sugar
decolourisationAutomation of white pans at the Tongaat Hulett
refineryPowdered activated carbon (PAC) with membrane filter press
for secondary decolourisation system to produce refined sugar in
backend refineryWhere do you go to (my saccharides)? A preliminary
saccharide analysis of refinery streams
Factory Session 7: Refinery II (Chair: Stephen Walford)The
transfer of non-sucrose species into sucrose crystals: can it be
useful?Optimisation of white sugar colour management through the
utilisation of on-line colour camerasLearnings from the 2015
Pongola silo failure To bee or not to bee (stung): Hulref's
intervention in reducing bee stings