Inverted cone silos are not immune to silo blockages, but strict compliance with Eurocode 1991-4 loading requirements confine any problems to the past at a modest cost. Aurecon explains how adherence can yield benefits for silo operators.
The concept of an ‘inverted cone’ silo was first developed in the mid-1970s and since that time, thousands of this type of silo have been constructed. Most inverted cone silos were designed using loads derived from the German code DIN1055, Australian Standard AS3774 or the American code ACI313.
While there are a number of versions, the basic ‘inverted cone’ concept involves the construction of an upward pointing 60º cone structure within the silo which forces product to be discharged around the perimeter of the silo, adjacent to the silo wall. This geometry guarantees eccentric discharge which imposes highly asymmetric loads on the silo wall during silo operation.
In the past 10 years, it has become increasingly apparent amongst cement companies that many inverted cone silos are exhibiting severe structural distress. A number of well-publicised spectacular failures occurred such as those at Hover in Germany and Davao in the Philippines. More importantly, many companies reported operational issues such as extensive wall cracking, blockages resulting from water ingress, lumps of concrete breaking off inside the silo and other such problems.
There was a general tendency to attribute these problems to poor-quality construction, particularly in slipform construction where lack of cover and lowstrength concrete were often confirmed by inspection.
When released in 2005, the new Eurocode 1991-4 radically changed the assessment of loads on silo walls for silos with eccentric discharge. Concerned about the possible shortcomings of AS3774, Aurecon Group commenced a three-year investigation into the performance of inverted cone silos to gain insight into the validity and applicability of the new Eurocode.
Aurecon’s initial investigation was reported by McKay and Durack1 at the Cemtech Conference 2006 in Rome. It documented analytical investigations undertaken to compare the traditional design standards, AS3774, DIN1055 and ACI313 to Eurocode, using as a case study, an Ф18m, 10,000t-capacity inverted cone cement silo.
The investigation showed that wall hoop direction moments and shear forces derived by AS3774 were only around one-third of that predicted by Eurocode. McKay and Durack concluded that wall reinforcement normally provided for such silos was seriously deficient to resist both wall moments and shear.
As can be seen in Figure 1, non-uniform pressure results from the formation of flow channels above the discharge point. These flow channels cause a reduction in wall pressure adjacent to the opening and an increased wall pressure on either side of the flow channel, producing severe bending moments and shears in the silo wall.
Concerned by its findings, Aurecon approached Holcim subsidiary, Cement Australia, and sought approval to measure wall deflections on a 10,000t, Ф 8m silo at Devonport. The Devonport silo was specifically selected because it was somewhat unusual insofar that after filling by rail over a period of days, the silo is fully discharged at the rate of 600tph into a ship. This permitted survey of an entire discharge cycle.
The silo was instrumented with prism targets and continuously surveyed over one full discharge cycle of 16 hours, using high-precision surveying techniques. The results of this work were reported by Durack, McKay and Davies2 at Cemtech 2007 in Prague.
Very significant wall displacements were recorded, ranging from -15mm to +28mm, far in excess of what should have been occurring, based on loads derived from AS3774. The authors concluded that despite the uncertainties and limitations discussed throughout the paper, the movements were of a similar magnitude to those predicted by Eurocode.
Aurecon was subsequently engaged by Cement Australia to inspect, analyse and report on its entire suite of inverted cone silos at plants throughout Australia, some of which were experiencing operational problems, including cracking, water ingress and blockages. Fifteen inverted cone silos were inspected, ranging in size from 5000-35,000t capacity, with diameters ranging from 14 to 30m. Silo loadings were calculated according to Eurocode and the design of each silo reviewed. All but one of the silos had been designed according to AS3774.
Aurecon concluded that:
Aurecon’s report made a number of recommendations including:
Concurrent with the work of Aurecon, others equally concerned by the new Eurocode were undertaking extensive research and development.
Claudius Peters Technologies GmbH (CPT) engineers Hilgraf and Krause3, 4 reported some excellent research work, and argued that a correctly operated CPT EC silo did not experience the maximum loads predicted by Eurocode, due to the size (diameter) of flow channel that formed, and its location which barely touched the silo wall. Despite this, Hilgraf and Krause also recommended that loads derived for structural design should adhere to Eurocode.
Buschmann5 from IBAU Hamburg similarly focused on emphasising the need for correct operation of its central cone silo and claimed that “aeration is configured so the flow channels formed during emptying do not touch the walls, or only touch them minimally.”
Ultimately, both systems involve an inverted cone and sequential discharge from close to the perimeter of the silo.
In our view, therefore, it is an inherent feature of that geometry that eccentric discharge must therefore occur, and for Eurocode compliance, the pressure distributions used for structural design must be the same for either system.
Buschmann correctly observes that Cement Australia’s Townsville cement silo featuring IBAU Hamburg discharge equipment was designed by Aurecon according to the AS3774 code, and that it continues to operate without problem after 16 years of service. The reason for this is likely to be due to a combination of factors: the silo was post tensioned and was constructed to a very high standard, its discharge system is well maintained, and the silo has been operated strictly according to the supplier’s recommendations. Thus, the silo has probably never been subjected to the maximum possible loads predicted by Eurocode. Nevertheless, knowing what we now know, Aurecon considers the AS3774 code deficient in determining loads on inverted cone silos and recommends that it should not be used in isolation going forward.
With all the research and debate ongoing, it was evident to Aurecon that the cost ramifications of Eurocode design compliance needed to be put in perspective. Accordingly, three silo examples were studied.
The first example involved two dualcell ring silos constructed in Singapore for Pan United Cement6. Both silos were designed by Aurecon, the first in 2001 in compliance with DIN1055 and AS3774 and the second in 2007, compliant with Eurocode. Both were constructed by Leighton Contractors (Singapore) Pte Ltd and operate without problems. The silo geometries are similar but the second silo is marginally smaller in diameter due to site space constraints.
Using tendered rates applicable to the 2007 silo, the costs were directly compared. Strict application of Eurocode resulted in:
The increased structural cost was less than SGD200,000, representing approximately 1.5 per cent of the civil/structural cost and just 1.2 per cent of the total silo cost.
In the case of a dual-cell ‘ring’ silo, the presence of the inner wall limits the pressure that can develop on the outer wall and the size of the flow channels that can develop. The cost impact of Eurocode compliance is therefore understandably minimal.
Obviously, on a single-cell silo, the load impact of Eurocode is much greater. Two single-cell silo scenarios were considered:
Both silos are located at Cement Australia’s Bulwer Island plant in Brisbane and the height of the storage zone in both silos is virtually identical at 42m.
The 25,000t cement silo was designed by Aurecon in 2007 and construction was completed in January 2009.
The silo is Eurocode compliant. To study the cost impact of Eurocode compliance, a comparative design was developed based on loads derived from Australian Standard AS3774. The exercise revealed that Eurocode compliance results in:
The second comparison related to the 10,000t silo, designed by Aurecon in 1996 in accordance with loads derived from AS3774. Its wall design was repeated using loads derived from Eurocode 1991-4, with the following findings:
Using the rates tendered in 2008 for the adjacent 25,000t silo, the cost premium for Eurocode compliance for a 10,000t silo was estimated at AUD625,000 or 9.5 per cent of the structural cost including piling.
The major portion of the additional cost related to the requirement for prestressing and the resulting thicker high rise wall and buttresses.
Many operators of inverted cone silos experience operating difficulties resulting from silo blockages. Market demands often mean silos cannot be shut down to remove blockages, so operation continues regardless, often resulting in loads not envisaged by the equipment supplier.
Many of the silos inspected by Aurecon in Australia, New Zealand, Indonesia and the Philippines exhibited significant cracking, internal spalling and water ingress. Certainly, some of these problems may have been attributable to poor quality construction, but in the case of the Australian silos inspected, that was generally not the root cause of problems.
It was very evident to Aurecon that previously well-accepted loading codes, DIN1055, AS3774 and ACI1313 were deficient in the way they dealt with eccentric discharge and this has potentially left the cement industry with a significant problem. However, uncompromising compliance with Eurocode loading requirements in future designs will eliminate many of the currently experienced problems. Depending on the size and type of inverted cone silo (single, dual or multicell), the structural cost impact for this compliance ranges from as little as 1.5 per cent for dual-cell ring silos to a maximum of around 9.5 per cent for the smaller single-cell silos (Ф <20m and not previously post-tensioned).
1 McKAY, H and DURACK, J (2006) Implications of the new Eurocode EN1991-4 on the Design of Cement and Raw Meal Storage Silos. Cemtech Conference, Rome.
2 DURACK, J, McKAY, H and DAVIES, N (2007) Measured Wall Movement during Discharge Compared to Predictions for an Inverted Cone Silo. Cemtech Conference, Prague.
3 HILGRAF, P and KRAUSE, I (2007) Improvement of Process Technology to Reduce Static Loads on Cone Silos. IEEE/ IAS/PCA Conference, Charleston, SC.
4 HILGRAF, P and KRAUSE, I (2007) Process Technology and Static Loads for Cone Silos. In: Cement International, 5.
5 B USCHMANN, H (2009) Townsville Cement Silo. In: International Cement Review, October.
6 McKAY, H (2010) New silo for United Cement. In: International Cement Review, January.