Download pdf - LEARNINGS FROM THE 2015 PONGOLA SILO FAILURE · Keywords: silo failure, silo stabilisation, silo strengthening, silo design, buckling strength, compression buckling . Literature search

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


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.


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.


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


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


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.


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.


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.


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.


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)


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.


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.


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.


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.


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.


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


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.


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).


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;


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;


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.


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