TEXTILES.ORG
By Bob Koerner
Prior to 1959, the chemical industry used sugar cane (mainly from Cuba) for its source of the carbohydrate sucrose used to produce ethyl alcohol for the manufacture of munitions and alcohol.
When Fidel Castro came to power in 1959, this nearby source was abruptly eliminated and the alternative was molasses. Of course, imported molasses required onshore storage tanks for the unloading of molasses tanker ships. Two 37m (120-ft)-diameter, side-by-side tanks were designed for a site at the Wilmington Port Authority (“property owner”) at the confluence of the Delaware and Christiana rivers in Wilmington, Del. (see Figure 1).
The two tanks were each to be 12.1m (40ft) high, and with the high-viscosity molasses at a specific gravity of 1.5, this is equivalent to a ground surface loading of 180kPa (3750 psf). The site, however, consisted of approximately 20m (65ft) of saturated organic clay. Three soil borings (center and at opposite ends of the site) were taken along with numerous undisturbed soil samples. Blow counts were extremely low (0-5 blows/300mm) for 20m and then a firm granular soil layer was encountered. There was some visual evidence of horizontal stratification within the organic clay but it was inconclusive.
The water table was at the slag-covered ground surface and site flooding was not unusual. The undisturbed samples were sent to the consultant’s in-house soil testing laboratory and revealed the following average properties for use in the sand drain design:
- void ratio, e = 1.72
- coefficient of (vertical) hydraulic conductivity, kv = 0.508 × 10-8 cm/sec
- coefficient of (vertical) consolidation, cv = 0.0405 cm2/min
- “assumed” coefficient of (horizontal) consolidation, ch= 0.081 cm2/min; i.e., ch = 2 cv
The bearing capacity of these geologically recent deposited silty clay soils was extremely low with consolidation times far in excess of the owner’s requirements. Thus, sand drains, followed by surcharging, were designed by the consulting engineering firm (“consultant”) hired by the chemical company (“tank owner”) involved.
Note in Figure 2 that the proposed site for the molasses storage tanks was directly across from three large two-story warehouses. The warehouses were supported on steel H-piles driven into the dense granular soil beneath the saturated silty clay. There was a service road, with a underlying water main providing service to the warehouses. It was located between the proposed storage tanks and the warehouses. The warehouse structures themselves were steel framed with brick facing on all four sides. The first floor was reinforced concrete placed directly on the concrete pile caps. There was an asphalt overlay on the concrete floor.
Using standard vertical consolidation theory and the above laboratory-obtained “cv-value,” the curve of Figure 3a shows a 90% consolidation time of 41.2 years. This was clearly unacceptable to the tank owner. However, using Barron’s theory of sand drains at n = 10 for horizontal flow resulted in slightly more than one year, and for combined flow slightly under one year. There was discussion of further reducing the time using an v = 5 sand drain array (which would have resulted in a four-month time—see Figure 3b). But one year was considered acceptable. In the end, sand drains of 300mm (12in.) diameter at a triangular spacing of 3.05m (10ft) were installed.
The installation started by placing a 1.0m-thick porous slag blanket over the entire site (Figure 2). This served as both a working platform for the large pile-driving crane and subsequently to allow for expulsion of drainage water coming from the sand drains as surcharge was placed. The installation of the sand drains was completed rapidly, with each drain taking approximately five minutes to drive, fill, withdraw, and move to the next location.
Upon completion of the sand drain installation, three open-standpipe piezometers were installed. The initial rate of surcharge placement was approximately 0.5m (1.6ft) per week with no fill placed during weekends. With a target of 10m (33ft) of surcharge, this would take 20 weeks and, depending on the piezometer readings, surcharge would then be stripped off the site and construction of the two tanks would begin. This timetable met the tank owner’s schedule since much of the molasses was coming by tanker ships from Africa.
As surcharge fill commenced, the three piezometers responded immediately. Unfortunately, there was almost a direct correspondence of surcharge stress and excess pore water pressure increase. Over weekends there was only a nominal decrease in piezometric levels. Pore water levels in the standpipe tubes rapidly rose to elevations above the level of the surcharge.
During this time, there were several water main breaks in the service lines connected to the main waterline under the service road—see Figure 4 (1). After about four weeks, the surcharge rate of filling was decreased to about 0.25m (0.82ft) per week. There was still essentially no decrease in piezometric water levels and they continued to rise assurcharge fill was placed. Pressure gages were eventually fitted to the standpipe tubes. Even further, bricks in the warehouse building’s veneer started to come loose and many could even be removed by hand—see Figure 4 (2).
A few weeks later, the warehouse superintendent noted that the floor under his desk was rising upward—see Figure 4 (3). Cracks in the asphalt-covered warehouse floor confirmed this upward heave. A futile attempt was made to position large lead ingots on the floor, but they also rose.
Finally, on a late Friday afternoon a large tension crack opened up on the top of the surcharge fill—see Figure 4 (4). The edge of the crack closest to the warehouse was approximately 150mm (6in.) lower than the far side and the crack visually lengthened (parallel to the warehouses) for 40–60m (131–197ft). It appeared as though it was “unzippering.” As shown by the arrows in Figure 4 a massive soil foundation shear failure was in progress. It occurred when the surcharge was 5.5m (18ft) high.
[Using minimum shear strength values, along with the piezometer measured pore water pressures, it was actually predicted to fail with a factor-of-safety of less than one when it was at 4.2m (13.8ft) high, Koerner (1963)].
The removal of surcharge began immediately and continued until the top of the slag drainage blanket was exposed. By the following Monday morning, the building floor stabilized (surveying was used throughout the weekend) and apparent equilibrium was reached.
After discussion among all parties involved, the sand drain/surcharge plan was aborted. An alternative design using two 40m-diameter interconnected steel sheet pile ringwalls were driven to a depth of 12.2m (40ft). The two steel molasses storage tanks were then constructed on top of the slag working blanket and within each of the two sheet pile containment ringwalls.
Large, flexible connections from ships moored at the Christiana Pier to the storage tanks were installed and small amounts of molasses pumped into the newly constructed tanks. At this point the molasses load in the tanks was serving as a gradually increasing surcharge load for the soil mass contained within the ringwalls. It took more than four years for the tanks to be utilized at their full storage capacity. In the meantime, numerous tanker ships moored in the Delaware River served as temporary molasses storage vessels at a great cost to the storage tank owner.
Conclusion
For about the next eight years, lawsuits were filed and the following two judgments were eventually reached in this case:
- The property owner sued the tank owner in Delaware court and won a $1.2 million lawsuit for building, street, and utility damages and repairs.
- The tank owner sued the consultant/designer in out-of-state court and won a $6 million lawsuit, which included the prior judgment, the alternative sheetpile ringwalls, the rental of numerous tanker ships (called “demerage”), the loss of capacity of the storage tanks, and maintenance/repair of the storage tanks.
The reason the court found theentire judgment to go against theconsultant was that the consultant did not convey to the tank owner that any risk was involved. Indeed, sand drains were still new in the early 1960s, but the consultant felt confident that this was the correct approach. (It is not known how the field inspector’s predicted failure height was addressed in the courtroom deliberations, but it probably was a factor.)
Of course, the essential technical question is: Why did the organic silty clay soil not consolidate as designed?
The author (see sidebar starting on pg. 25), who was the inspector on the project, does not know but suspects that “smear” creating a much lower “ch value” than expected was the major item. The issue of smear zone properties adjacent to the sand drain itself is unanswered today, nearly 50 years after this case history occurred.
A final comment:
This comment has to do with communication between the consultant and the client, usually the owner of a facility or structure. The lack of such communication before this sand drain project began was the pivotal point in the large financial judgment rendered. Not only is such communication required, it must be fully understood and accepted by the client. Only then (in the event of a failure) can the consultant avoid a judgment of the type described here.
In short, we must be good teachers as well as good designers!