Verification of geotechnical reliability using load tests and integrity tests - Risk and Reliability in Geotechnical Engineering - page 528

of construction, 1% of working piles should be proof tested using static loading tests. With
these construction control measures in mind, a low FOS is commonly used. Zhang et al.
(2006b) described the piling practice in the language of Bayesian updating. A preliminary
pile length is first set based primarily on a dynamic formula. The pile performance is then
verified by PDA tests during the final setting tests. The combination of the dynamic formula
(prior) and the PDA tests will result in a posterior distribution. Taking the posterior distribu-
tion after the PDA tests as a prior, the pile performance is further updated based on the out-
come of the CAPWAP analysis, and finally updated based on the outcome of the proof load
tests. This exercise can be repeated if more indirect or direct verification tests are employed.

14.4 uPDatIng PIle CaPaCItY WIth IntegrItY teStS

14.4.1 reliability updating based on integrity tests
Integrity tests do not offer pile capacity values directly. When using reliability theory, how-
ever, the outcome of qualitative integrity tests is routinely used as the basis for verifying the
reliability of piles. Testing of toe debris in bored piles is used in this chapter to illustrate the
use of qualitative testing information for the verification of pile reliability. Toe debris is the
accumulation of loose or undesirable foreign materials at the position of pile rock socket. It is
detrimental to piles that are designed to found on rock, as the quality of the interface between
pile concrete and bedrock greatly affects the mobilization of the pile base resistance. There
are three categories of defective toes, namely soil inclusions, unbounded concrete aggregate,
and their combinations (HKCA 2003). Causes of toe debris include poor workmanship dur-
ing concreting, soil dislodgement from unlined portion of excavation, substandard bentonite
slurry, percolating groundwater, and objects falling from the ground surface. In order to
ensure a satisfactory pile toe condition, interface coring, which is the formation of access to
the pile shaft by drilling through a pile toe interface, is often conducted after pile construction
following guidelines such as those specified by Hong Kong's Buildings Department (2004).
Consider the case of single-bored piles with toe debris along with other types of imper-
fections such as cracks, necking, and voids. If thick debris exists, then the pile capacity or
settlement will be adversely affected. The toe debris, if present, is assumed to be uniformly
distributed over the pile cross section for simplicity. The pile performance is still uncertain
even when the pile is free from toe debris due to the presence of other types of imperfections
as well as many other sources of uncertainty such as spatial variability of soils, construction
effects, and load effects. Based on the total probability theorem (e.g., Ang and Tang 2007),
the probability of unsatisfactory performance of the pile, p f , can be calculated as follows:

pPFE = ( | )( 1 − pPFEp ) + (|) (14.15)

f d d

where p d is the occurrence probability of toe debris, F is the event of pile failure, E and
Ē are the events of toe debris presence and toe debris absence, respectively, and P ( F | Ē ) and
P ( F | E ) are the conditional probabilities of unsatisfactory performance of the pile given the
absence and presence of toe debris, respectively.
For a particular pile, the toe debris may be described by its thickness x . If x is taken to be
a discrete variable, the conditional probability of unsatisfactory performance of the pile is

n

∑

PF E (|) = P Fx Px E

i

( | )( | ) (14.16)

i

i = 1