|Below: a diagram of hammer placement on the pile, from the Whipple and Vines monograph.
Below: a pile driving log from offshore, 1976. As the hammer lowers, blow by blow, the pile into the soil, the number of blows required to advance the pile one foot is counted and recorded for each foot of penetration. With a uniform soil, the resistance would increase uniformly as more pile-soil contact is made along the pile shaft and the toe resistance increases, but soils are anything but uniform, and the increase is best described as a general trend with interruptions.
Offshore pile driving was an impetus to many advances in the prediction of pile drivability, capacity and driving stresses. Central to that was the wave equation; the TTI (Texas Transportation Institute) wave equation program was a tool that came to maturity in the oil field. Click here for more information on the TTI wave equation program and research effort.
Sometimes these interruptions could be dramatic. In 1980, during one platform installation in Brazil, as a Vulcan 560 hammer was driving pile for the Italian contractor Micoperi, the pile toe encountered an underground cavern. The pile ran, the shackle broke and the Vulcan hammer assembly fell and went to the sea floor. (Micoperi was, much to Vulcan's disappointment, able to salvage the hammer and bring it back into service.
A Cautionary Note about Foundation Design and Platforms
In a recent issue of GeoStrata, the following was stated about the aftermath of the three hurricanes that came through the Gulf in 2004-5 (Ivan (2004), Katrina and Rita (2005)):
Driven steel pipe piles are the foundations for approximately 4,000 steel jacket platforms that rest on the sea floor of the Gulf of Mexico. More than 100 platforms were destroyed in these three hurricanes combined, which essentially tripled the number of platforms destroyed by hurricanes in the history of oil and gas production in the Gulf of Mexico...
While the damage to offshore platforms from these recent hurricanes was unprecedented, there were few if any failures of pile foundations either axially or laterally...The photograph (not included) shows the leg of a steel jacket near the mudline. The outer jacket that the pile was inserted through failed, while the pile remained intact. The "flange" in the jacket leg was formed by the rocking motion of the platform after the leg ruptured.
The pile foundations performed better than expected. Structural modelling for individual platforms with loads based on hurricane hindcasts indicates that numerous platforms should have failed in the foundation. These results include platforms that actually failed in the jacket above the foundation as well as platforms that did not fail. While there is some conservatism built into the pile foundations for offshore structures, there is now a wealth of information to better understand the actual capacity of full-scale pile foundations under design loading. Offshore piles are about an order-of-magnitude larger than those in the pile load tests originally used to develop the methods for predicting capacity.
A more realistic assessment of foundation capacity will be an important contribution from these recent hurricanes, particularly since design loads are increasing and foundation capacity is likely to limit the design of new platforms and the re-qualification of existing platforms. Ongoing research aims to improve these predictions. (Gilbert, R.B., "Offshore Experiences in Recent Hurricanes." GeoStrata, January/February 2007, pp. 28-31)
This is one of those situations where simple "hindsight" doesn't tell the whole story.
Offshore pile foundations were a major step up in size from what had been installed in onshore and coastal construction. Both the size and the lengths of the piles were beyond both those installed in load tested piles and those in actual service. This created a hard "push" on every aspect of pile foundation design and installation. We've documented here what Vulcan experienced in the relentless upward size push that took place in the 1960's and 1970's, to say nothing of the expansion of manufacturing and transport of the piles themselves.
Analytical techniques were likewise pushed. The wave equation, along with dynamic analysis of piles, went from E.A.L. Smith's "simple" program to the standard method of analysing pile drivability and verifying pile capacity dynamically in no small measure because of offshore demands. Dynamic formulae, already inadequate for onshore applications, were simply useless for the long wave-transmission elements that are offshore piles. Static load testing--tensile or compressive--were and are impractical for offshore platforms. As Dr. Gilbert points out, static design methods (both axial and lateral) were stretched as well, and the effects of dynamic loading on piles were no where nearly as well understood as they are now.
The foundations of conventional offshore platforms were, as with many other aspects of these structures, a leap in the dark. That they survived the worst natural disaster in U.S. history is a tribute to the engineering judgement of the designers of the platforms, along with many others. From a hammer manufacturer's standpoint, that judgement was further pushed by the fact that some piles for these structures were not driven to their design penetration due to equipment failures or drivability problems.
A humorous way of illustrating this came from our offshore sales manager. He was out on a derrick barge. To return, his baggage was loaded on a helicopter. He followed, sitting down with his travelling companions. The pilot revved up the rotor, the chopper rose a few feet and return to the helipad. The pilot revved up the rotor again with the same result. Back on the helipad, he turned to his passengers and said, "If we get rid of 12 1/2 pounds of fuel, we can be flying."
"I'm gettin' off," our sales manager said.
"Why?" the pilot asked.
"If you had said ten or fifteen pounds of fuel, that's okay. But 12 1/2 pounds is too precise." He did get off and caught a crew boat for a long journey back to shore.
It's obvious that the engineers that designed these platforms didn't count on the geotechnical equivalent of 12 1/2 pounds of fuel to get the job done.
Engineering judgement has always been considered a key to success in geotechnical design. The advent of probabilistic method of design (such as LRFD) represents an attempt to quantify that "judgement" and by doing so reduced the need for the level of conservatism in the design that characterises geotechnical engineering.
But the events of 2004-5 should give pause to this march for one reason. Any probabilistic method depends upon the accuracy of the assumptions and the understanding of the potential conditions to which a structure is subjected. These hurricanes--especially Katrina--subjected these structures to loads which were not anticipated. As Dr. Gilbert points out elsewhere in the article, "Hurricane Katrina was relatively slow moving and intense. Waves greater than 100 feet high were inferred from measurements at an offshore buoy--the largest ever recorded at a buoy in the Gulf of Mexico...One consequence of these three storms is that the design criteria for offshore platforms will likely be modified."
Modified indeed. Both shallow and deep foundations are subject to changes in the conditions from prosaic ones such as a long term change in the water table level to dramatic ones such as Hurricane Katrina. If we blindly allow the application of present knowledge of potential loads to a structure without any kind of engineering judgement allowed for that which we don't know, we will have failures that are both catastrophic and expensive.
In his pioneering article on the wave equation, the Australian engineer David Victor Isaacs spoke about a "factor of ignorance." That factor is still with us, and this should be kept in mind both by those that design and those who write the codes and standards that are imposed on those that do.
Article written February 2007