Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

GUIDELINES FOR USE OF CPTU APPLICATION CLASSES

ACCORDING TO ISO 19901-8: (2014)

T Lunne
Norwegian Geotechnical Institute, Norway
R Soage-Santos
Gardline, UK
J Brink Clausen
Geo, Denmark
JJM Powell
Geolabs Limited, UK

Abstract
The ISO 19901-8, 2014 standard defines 3 Application Classes according to what the measured results of pie-
zocone penetration test (CPTU) are to be used for. Application Class 1 is the strictest and is intended for use
when parameters are to be defined for very soft to soft deposits. Application Class 3 is the least strict and can
be used mainly for defining layers in mixed bedded material and soil parameters in stiff to very stiff soils.
Application Class 2 is between Application Classes 1 and 3. The minimum accuracies of the measured values
of cone resistance, sleeve friction and pore pressure need to be within certain limits as given in the standard.
The ISO standard does not give a well-defined procedure for how to document if a test fulfils a certain class or
not. This lack of clarity causes uncertainties in interpretation of the Application Classes and in worst-case con-
flict between end user and contractor. The paper recommends a pragmatic procedure for Application Classes
evaluation that can be used as a guide for parties involved in testing and data evaluation of CPTU results.

1. Introduction discussed, including the requirements to validate that


In the latest ISO standard on Marine Soil the CPTU results are within a certain minimum
Investigation (ISO 19901-8, 2014), recommendations accuracy that depends on the required Application
and requirements with regards to the performance and Class. The standard gives no guidelines on how to
reporting of offshore CPTU tests are presented in validate data quality, only which possible factors that
terms of three Application Classes, see next section could influence the data quality and related accuracy.
for more detailed descriptions. It should be noted that The practical challenge related to this topic is setting
the defined Application Classes have been simplified up general guidelines on how to assess the data
relative to the ones given in ISO 22476-1:2012, which quality - hence, establish the achieved CPTU Appli-
was developed for CPT and CPTU with land based cation Class of a performed test more or less
use in mind. Further, it should be pointed out that the immediately upon completion of the CPTU. A
term Application Class refers to the actual CPTU number of methods are used throughout the industry,
results being obtained and not to the cone penetro- and the approach varies from company to company.
meter being used. Some cone manufacturers refer to One parameter required in the ISO standard, is
the ISO standard (ISO 2012) and state that their cone evaluating the “zero drift”, based on the readings
penetrometers are Class 1 cones based on laboratory before and after (a test). This is a simple and valid
calibration only. This is misleading since there are approach. This is a valuable input for evaluation of
many factors in addition to the cone penetrometer the uncertainty related to CPTU results, provided care
itself and the laboratory calibration that influences the is taken in making sure the conditions when taking
final accuracy of a CPTU result. Although to be able the readings before and after are similar.
to achieve an Application Class the cone itself must
fulfil the requirements. It will not be possible to prove and document with
100% confidence how accurate the CPTU measure-
This paper quickly reviews the developments in the ments are since the true values are not known. A
requirements for precision and accuracy in guidelines pragmatic approach is suggested where the aim is to
and standards from the first document in the 1960s substantiate that it is very likely that a certain
and up to the present ISO standard in 2014. The Application Class can or cannot be expected.
challenges with the present ISO 19901-8, 2014 are
2. Definitions the following values: 5 % of the measured value or 1
A range of definitions and terms are used in the % of the measuring range of the sensor used. Later
context of describing the data output from the CPT. documents have gone away from the reference to the
In different standards and recommendations the terms measuring range.
are used and not always defined. This section explains
how the definitions are used and to be understood in The Swedish Geotechnical Society Guideline in 1993
this paper. (SGF, 1993) introduced the concept of accuracy clas-
ses where the requirements were related to what the
2.1 Accuracy results were to be used for. In ISO 22476-1:2012,
The amount of uncertainty in a measurement with valid for CPT/CPTU in general, a similar scheme is
respect to an absolute standard (value), hence, used but defined in terms of four Application Classes.
exactness of a measurement compared to the true In the ISO standard for Marine Soil Investigations
value of the quantity being measured. (ISO, 2014) a simplification has been made resulting
in three Application Classes as discussed in the next
2.2 Hysteresis sub- section. This last document also introduces the
A transducer should be capable of following the issue of Depth Accuracy, defining five Depth Accu-
changes of the input parameter regardless of which racy Classes. Although this is a very important issue
direction the change is made; hysteresis is the this paper focuses on the Application Classes for the
measure of this property. CPT/CPTU parameters. Hopefully a future document
will deal with the Depth Accuracy Classes.
2.3 Precision
Refers to the degree of reproducibility of a 3.2 Present status
measurement, i.e. how close the measured values are In the following reference will mainly be made to ISO
to each other under same or similar conditions. 19901-8 (2014) since this is the only standard specif-
ically for offshore conditions. For completeness, the
2.4 Repeatability most relevant parts are included with text in italics
Measures the variation in measurements taken by a exactly as is written in the ISO document.
single instrument under the same conditions.
3.2.1 Definition of Application Classes
ISO 19901-8:2014 defines the three Application
2.5 Resolution Classes as follows:
Is the smallest detectable incremental change of input For cone penetration tests, the use of Application
parameter that can be detected in the output signal. Classes as defined below shall be adopted. Equip-
Resolution can be expressed either as a proportion of ment and procedures to be used should be selected
the reading (or the full-scale reading) or in absolute according to the required Application Class given in
terms. Table 1. Application Classes are defined as follows.
a) Application Class 1 is intended for very soft to soft soil
2.6 Output stability deposits. Class 1 penetration tests are normally not
Is an absolute quantity, the smallest absolute amount achievable for mixed bedded soil profiles with weak to
of change that can be detected by a measurement. strong layers (although pre-drilling through these layers
taking into consideration the system noise. can overcome the problem). Tests can only be performed
as CPTU.
3. Developments in requirements: for accuracy b) Application Class 2 is intended for precise evaluation
and precision for mixed bedded soil profiles with weak to strong layers,
3.1 Developments since 1968 in terms of profiling and material identification. Interpre-
Starting with 1968 there have been a number of doc- tation in terms of soil parameters is also possible, with re-
striction to indicative use for the soft or weak layers. The
uments describing requirements and recommenda-
test type should be CPTU.
tions to the cone penetration tests as summarized by c) Application Class 3 is intended for evaluation of mixed
Peuchen and Terwindt (2014). Before 1980s only bedded soil profiles with soft to stiff clays and loose to
cone penetrometers without pore pressure measure- dense sands, in terms of profiling and material identifica-
ments were included. But, in later years the CPTU, or tion. Interpretation in terms of soil parameters is appro-
piezocone, has also been covered. The first document priate for very stiff to hard clay and for dense to very dense
giving some requirement to precision was issued by sand layers. For stiff clays or silts and loose sands, only
the International Society of Soil Mechanics and Foun- an indicative qualitative interpretation can be undertaken
dation Engineering (ISSMFE) in 1977: The precision using data acquired under this Application Class. The test
to be obtained should not be worse than the larger of type should be CPTU but in some cases CPT may be ac-
ceptable.

2
4. Main challenges with present ISO
NOTE Mixed bedded soil profiles refer to soil condi- With reference to Chapter 3, the challenges when us-
tions containing typically dense and stiff to hard soils, but ing ISO 19901-8:2014 include, but are not necessarily
possibly also including soft or weak layers. limited to:
a. How to document that a certain Application
If all possible sources of errors are added, the mini- Class can be obtained before the start of field
mum accuracy of the recorded measurements should work, for instance in the tender phase?
be better than the largest of the values given in Ta- b. Regarding the CPTU parameter accuracy re-
ble 1. The uncertainty analysis should include inter- quirements: is it an absolute requirement that all
nal friction, errors in the data acquisition, eccentric three parameters shall fulfill the limiting values
loading, temperature (ambient and transient) effects, given in Table 1? Should we give the main em-
pore pressure effects in gaps below and above the phasis to the measured parameters which are
friction sleeve, and dimensional errors. The resolu- most crucial to our interpretation methods?
tion of the measurements should be better than one- c. Annex D, clause D2.1 refers to zero readings be-
third of the required accuracy applicable to the Ap- fore and after a test:
plication Class given in Table 1. For CPTU's, a first attempt at recommendations
for limiting values is to follow the minimum ac-
Table 1: Accuracy of CPT/CPTU parameter values for Applica-
tion Classes curacy according to the applications given in
Table 1. How reliable are the zero readings be-
Application Test Measured Allowable mini- fore and after for this evaluation? How does
Class typea parameter mum accuracya
cleaning of the cones etc. play a role?
1 CPTU Cone resistance 35 kPa or 5%
d. What is the overall best approach for the evalua-
Sleeve friction 5 kPa or 10%
tion of the resulting Application Class after a test
Pore pressure 25 kPa or 5%
has been completed?
2 CPT or Cone resistance 100 kPa or 5%
e. If a test is evaluated to be in a lower Application
CPTU
Sleeve friction 15 kPa or 15% Class than required in the project specifications,
Pore pressure 50 kPa or 5% how shall the results be treated in the definition
3 CPT or
of representative parameters for design?
Cone resistance 200 kPa or 5%
CPTU f. From a client’s point of view: Which Applica-
Sleeve friction 25 kPa or 15% tion Classes and default values are reasonable to
Pore pressure 100 kPa or 5% require in the project specifications? What is
a achievable?
The allowable minimum accuracy of the measured pa-
rameter is the larger value of the two quoted. The relative ac-
curacy applies to the measured value and not the measuring It is the purpose of this paper to give guidelines and
range. recommendations on how to handle some of the
above challenges.
The allowable minimum accuracies for cone re-
sistance, sleeve friction and pore pressure shall apply 5. Previous work on Application Classes as given
relative to the seafloor. in ISO standards
Several excellent papers have recently been produced
Further Annex D in ISO 19901-8: 2014 gives detailed which try to look at the Accuracy of CPT measure-
recommendations on reference readings for various ments in great detail.
deployment modes. Here it will only be mentioned
that for a specific seafloor CPTU it is recommended We have to be practical and realistic in the level of
that readings of all sensors should be recorded contin- accuracy we can achieve with the CPT. If we don’t
uously from deck to deck. Tabulated values on deck have confidence in the CPT measurements, then we
before deployment, on seafloor before starting pene- cannot have any degree of confidence in our derived
tration, at seafloor after test has been completed and geotechnical parameters.
again on deck should be included in the report. A sim-
ilar scheme is recommended for tests carried out in The paper by Peuchen and Terwindt (2014) makes a
down-hole drilling mode. serious attempt to carry out uncertainty analyses as
required in ISO. They look at the CPT in great detail
in terms of accuracy of measurements applying con-
cepts that are generally not applied in field testing
measurements, but more to laboratory or environment
controlled testing. They try to quantify the effects of drilling – stable, non-floating platform, downhole-
various factors that can influence the measurements based deployment) and IV (Drilling – vessel drilling
in the CPT/CPTU and the resulting effect on the re- – floating vessel). These are for water depths of
ported parameters from the test. Areas covered in- 150m; for 1500m Classes can worsen slightly.
clude load cell behaviour, accuracy and repeatability,
allowable inaccuracies in equipment geometry, tem- The two papers by Peuchen and Terwindt (2014,
perature effects on measured values, corrections re- 2015), really represent a very in depth evaluation of
quired to measured values as a result of cone geome- the factors that can influence the various measure-
try and more. On first reading the implications of the ments during a CPTU. Their analyses of uncertainties
potential errors discussed might seem daunting in are based on input parameters found in literature, their
terms of obtaining reasonable accuracy in results own investigations, some general knowledge on
from the CPT. However, they do show that when con- transducers etc. The uncertainty analyses in a way be-
sidering the influence of all likely inaccuracies then comes general or theoretical rather than based on
in the case of ‘favourable’ conditions in homogene- comparison with a reference system.
ous soils then the level of ‘uncertainty’ in measured
values would mean that Application Class 1 was The requirement in ISO 19901-8: 2014 is presumably
achievable even in soft soils. This would be true for included so that the user can evaluate the likelihood
all measured parameters. of the test by a specific contractor, using a specific
cone design, to be within a required Application
When they consider ‘adverse’ conditions of layered Class.
sands and clays then the results are far less encourag- In the authors’ opinion the uncertainty analyses
ing. There is some confusion at this stage in their pa- should be undertaken based on the laboratory calibra-
per because they seem to then consider the predicted tion of a specific cone design to show that the cone
results in terms of the Application Classes but seem itself has the quality to meet a certain Application
not to recognize that the classes also relate to soil Class, as is discussed in Section 6.1.
types and intended use of the results; they show Ap-
plication Class 1 for sands when it is only intended 6. Guidance on how to address the issue of
for clays and soft to firm. They do say that they have Application Class requirements
considered the worst cases and that the real situation A pragmatic approach will be suggested in the fol-
would probably be better. lowing where the aim is to substantiate that it is very
likely that a certain Application Class can be
They consider all the potential errors in arriving at achieved. Basically, the requirements related to Ap-
their uncertainty levels but good practice can reduce plication Classes in the ISO can be considered in to
or eliminate many of these. Field studies comparing two Stages:
different CPT devices have shown that similar results A. How to substantiate at tender phase that CPTUs
can be achieved when good practice procedures are can potentially be carried out according to a cer-
followed. These have generally been performed on tain Application Class?
land and offshore offers an extra level of complexity B. At the completion of a CPTU evaluate and docu-
(e.g. Powell and Lunne 2005). ment that the test result fulfills a certain Applica-
tion Class or classes?
It would be useful to know how the uncertainty levels
would change if they considered a ‘best practice’ case The approach suggested consists of the following el-
for the likely results. ements:
1. Documented laboratory calibration (certificate)
Peuchen and Terwindt (2015) look specifically at the 2. A documentation form/scheme for field checks,
offshore case. They discuss how method of deploy- measurements and observations
ment can affect the likely Application Class achieva- 3. The field documentation scheme filled in
ble as they impinge on the uncertainty in some meas-
urements. They again show that Application Class 1 In order to fulfill Stage A elements 1. and 2. must be
is achievable in favourable soil conditions (normally presented in a satisfactory form as discussed below.
consolidated clays) but now only for deployment
modes I (non-drilling seafloor), III (Drilling – vessel To decide on the application class/classes achieved
drilling – stable, non-floating platform, deck-based after a CPTU, ‘element 3’ needs to be presented, in-
deployment) and V (Drilling – seafloor drilling), with cluding the final evaluation.
Class 2 for deployment modes II (Drilling – vessel

4
6.1 Laboratory calibration of the CPTU probe analysis. The combined standard uncertainty is calcu-
The first step in the assessment of the potentially lated as the root sum square of the individual uncer-
achievable Accuracy Class should be based on the tainties contributions and expressed as a percentage.
calibration report. The calibration procedure should
take into account all the representative sources of er- Table 2 Example of results of a laboratory calibration of a pie-
ror that the measurement system may experience dur- zocone and associated uncertainty analyses
ing field operation. The reported uncertainty analysis LABORATORY CALIBRATION
Potential
of the complete CPTU measuring system (cone sen- Uncertainty analysis
Maximum error of measured value (%)
Uncertainty
(kPa/Deg)
sors, data transfer cables and data logger) must indi-

UNCERTAINTY (%)
Output stability

Repeatability

Temperature

Inclined load
Dimensional

COMBINED
Resolution

Hysteresis
cate the Accuracy Class that a particular system is ca-

Zero drift
Linearity

changes
Sensor
pable of achieving under laboratory conditions.
Calibration certificates should be presented specify-
Cone
ing the ability of the measurement system to fulfil the resistance 0.0031 0.0048 0.5623 0.1434 0.2742 0.0001 0.2800 0.70 1.20 0.3
Sleeve
required Application Class (see example in Table 2). friction 0.0005 0.0009 0.5102 0.0067 0.4219 0.0009 0.1400 0.68 0.06 0.4
Pore
pressure 0.0000 0.0001 0.2400 0.2784 0.3230 0.0002 0.49 0.5
Currently a specific ISO standard for the calibration
of CPTU equipment does not exist. Therefore, cali- Figure 1 illustrates the calculated total uncertainty for
bration laboratories tend to refer to generic calibration the cone resistance, based on the laboratory calibra-
standards and recommendations for specific sensors tion.
(BS8422, JCGM 100:2008) and not to the CPT sys-
tem as a whole. In ASTM D5778-12 the annex de-
fines the requirements for calibration of the CPTU
measurement system. The accuracy requirements pre-
sented are, however, not in line with the accuracy
class scheme proposed by ISO 19901-8. It is the au-
thors’ opinion that ASTM requirements for calibra-
tion should be followed for development of calibra-
tion procedures and reporting uncertainty but that ISO
accuracy requirements should be adhered to.
Figure 1 Cone resistance calibration laboratory results
The calibration certificate should include a detailed
description and identification of the CPTU probe and Finally, the uncertainty analysis should include a de-
measurement system, identification of the calibration scription of the potential error that changes in temper-
laboratory, description of the calibration laboratory ature (see ISO 22476-1:2012, Annex A2.4) and in-
environment during calibration, description of the ref- clined load can cause. These factors need to be
erence instruments used, details of the calibration considered as potential contributors to uncertainty but
method, details of any adjustment made prior to cali- will not be directly used for the classification of accu-
bration, results of the calibration, complete uncer- racy on the calibration certificate, but should be con-
tainty analysis report. sidered for calculation of the error of a particular
CPTU test based on the field observations. Uncer-
The calibration uncertainty analysis defines the best tainty caused by temperature changes between deck
Application Class that a particular system can obtain and seafloor do not need to be considered if enough
under laboratory conditions. An example of how to time for temperature equalization is allowed at sea-
report the uncertainty analysis for the calibration of a floor before the test commences.
CPTU measuring system is presented in Table 2. The
necessary formulae for the analysis are covered by BS 6.2 Documentation scheme for field checks, measure-
8422 and EN JCGM 100:2008. Load cell measure- ments and observations
ments of cone resistance and sleeve friction are usu- While measurement systems are generally very relia-
ally divided by their respective areas and reported in ble and robust, experience shows, that every single
stress units. It is therefore required that the uncer- component of the CPTU measurement system may
tainty associated with the dimensional measurements deteriorate, suffer damage or drifts during operations
of cone and sleeve and the area factors, alpha and beta (transport, system set up or testing). Therefore, it is
(as determined in a pressure chamber, see ASTM. imperative to monitor the integrity of the system at
D5778-12, Annex A1), are included in the uncertainty the main stages of the operation. The monitoring pro-
gram allows the systematic evaluation of the applica-
tion class for a particular test. Table 3 illustrates the
factors that require monitoring. Some of these factors Comparison of reference readings
can be directly used to quantify and classify CPTU Observed differences of reference readings taken be-
results in terms of application class. The calibration fore and after a test will indicate if the measuring sys-
uncertainty analysis defines the lowest application tem has experienced a drift during the penetration or
class that a particular system can obtain under labor- retraction of the cone from the ground. This is equiv-
atory conditions. After every test the application class alent to the determination of the zero-drift error per-
should be adjusted based on the measurement and ob- formed in the calibration laboratory. The absolute dif-
servation scheme during test performance. ference values observed can therefore be used for
classifying the CPTU profile in accordance with the
Dimensional check recommended maximum inaccuracy (Table 1) to ob-
Cone dimensions should be recorded before and after tain the Application Class of the test based on the as-
every test (this is compulsory for Class 1 testing). As sumption that the drift has occurred during the tests.
cone resistance and sleeve friction measurements are It is important to highlight that changes such as am-
expressed in units of stress, the error associated with bient temperature, ambient pressure, the presence of
the dimensional difference must be then be consid- dirt in the cone seals and adhesion of clayey soils can
ered in the final assessment of accuracy. cause reference readings differences that are not di-
rectly related to the changes that the measurement
Oscillations of measuring system: Output stability system may have experienced in the ground. It is fun-
An effective method for characterizing the system damental that reference readings are taken under sim-
noise and therefore its output stability is by examin- ilar conditions every time so that a meaningful com-
ing a representative number of output samples, while parison can be made. Therefore deck reference
the input and environment conditions are held con- reading should be taken under no load conditions
stant. This can be done quite simply by monitoring once the cone has reached temperature equilibrium
the acquisition system output for 2 minutes at zero and the cone has been thoroughly cleaned and dirt
load (See Figure 2). carefully removed from gaps.

The measured output stability can be used as input to Observations on CPTU performance
the classification of a CPTU dataset in terms of the In addition to the quantifiable contributors of uncer-
Application Classes. The amplitude of the observed tainty it is also important to include as much infor-
oscillations at zero load before and after the test can mation as possible about the sensor performance.
be compared with the recommended absolute mini- Qualitative assessment is very important for the user
mum resolution (1/3 of the allowable minimum accu- of the datasets. Some examples of required commen-
racy defined in Table 1) to give input to the Applica- taries are listed below:
tion Class of the test. For instance, if a system • Information related to the responsiveness of sen-
calibration uncertainty analysis shows that it is capa- sor to soil changes
ble of achieving Application Class 1 and the output • Sluggish response
stability checks at the control stages falls outside class • Sleeve friction insensitivity caused by dirt
1 range then the CPTU test cannot be classified as trapped in cone seals and gaps.
Class 1. It is important to note that currently used ac- • Pore water pressure cavitation effects experi-
quisition system have high resolution and deviations enced during testing
of output stability below the minimum resolution de- • Qualitative indication of the soil disturbance
fined by application Class 1 generally indicates that a caused by the seabed frame or drilling activities
fault is being developed in the system.
Some of the qualitative observations of cone perfor-
mance during the tests will degrade a result say from
potentially Application Class 1 to Class 2. Examples
might be; sluggish response caused by poor satura-
tion of pressure transducers, soil disturbance caused
Figure 2 Measured output stability as function of time by seabed fame or drilling technique, depth inaccura-
cies larger than 0.1 meter, probe inclinations larger
than 30 degrees from vertical.

6
Table 3 Example of field records and field assessment of Application class
Test procedure
Saturation procedure used for pore water Vacuum for 5 hours in glycerin oil
pressure system
Air temp. before and after testing on deck 17°C / 17°C reference reading taken in a water bath

Probe location for deck reference readings Deck in seawater bath

Time allowed for temperature compensa- Output values were monitored until stability was reached after 7 minutes
tion before reference readings on deck
Probe preparation before reference read- Cone rinsed with seawater, sleeve friction rotated several times, dirt removed from
ings on deck gaps and seals by flushing
Probe location for pre- and post- test refer- Seafloor (Offset 0.2m)
ence readings
Time allowed for temperature compensa- Output values were monitored until stability was reached after 8 minutes
tion on the seabed
Observations during test (stone, sound, Dropped stone encountered at 5 m pore pressure sensor showed cavitation
bent rods, cone preload etc.)
Test results
Corrections done during processing Standard correction to cone resistance accounting pore water pressure effects.
Depth corrected based on inclination measurements
Inclination of probe relative to vertical Maximum inclination 1°.

Temperature changes during the test Not applicable, no temperature sensor

Evaluation of cone resistance profile Good response of cone resistance. No anomalies have been detected.

Evaluation of sleeve friction profile Good response of cone resistance. No anomalies have been detected.

Evaluation of pore pressure response pro- Very good reaction to changes in soil types. Cavitation was observed when drop
file stone encountered; pore pressure sensor recovered immediately.
Reference readings Cone re- Sleeve Pore water Comments
sistance Friction pressure
Deck reference readings (12 / 52)
(6 / 5) kPa (-4 / -2) kPa Large drift in cone resistance
(Before/After) kPa
Seabed/ downhole reference readings (Be- (3752 / (7 / 15) (4996 / 5100) Clay adhere to cone after retrac-
fore/After) 3775) kPa kPa kPa tion
Application Class analysis
Cone re- Sleeve Pore water Application Class
sistance Friction pressure
Required Application Class based on soil
Class1 Class1 Class1 Class1
type
Application Class based on calibration Class1 Class1 Class1 Class1
Dimensional difference 0 0 Class 1
Deck reference readings observed differ-
40 kPa 1 kPa 2kPa Class 2
ences (before and after)
Output stability for all sensors
4 kPa 0 kPa 2 kPa Class 1
(Maximum range of oscillations)
Uncertainty caused by Temperature
0 0 0 Class 1
changes
Uncertainty caused by inclination 0.3 0.4 NA Class 1
Achieved Application Class Class 2.

Causes of deviation from desired and Observed drift on the cone end resistance sensor. Test needs to be repeated with a
achieved Application Class different cone

It is important that operational details associated with ment. Table 3 gives an example of the minimum in-
any CPTU are presented in the factual report. The in- formation that should be reported for every test. It is
formation reported should allow users of the test data required that general operational and specific equip-
to conduct an independent Application Class assess- ment information are also included, such as the con-
tractor’s name, field engineers' and operators' names,
water depth, pre-drill depth, stop criteria, type of practice. The approach consists of documenting high
probe, probe identification and manufacturer, geome- quality laboratory calibrations of the cone penetrom-
try and dimensions, type of penetration system, meas- eters to show that the probes ‘shall’ be able to fulfill
urement ranges of all sensors, date of calibration, type the Application Classes based on uncertainty analyses
and location of filter, area factors, time and date. with the laboratory measurements as input. Further, a
These are not included in the table here due to space field documentation scheme needs to be in place re-
limitations. cording information at each step in the preparation
and use of the cone penetrometers. An important part
6.3 Use of Application Class of the scheme is the measurement and recording of
It is clear that there is confusion over the assignment reference readings during various stages of the tests.
of Application Classes to a CPTU profile. The desire With all this information, an assessment of the quality
seems to be to be able to assign a Class to the whole and potential usefulness of results should be possible.
profile and clients seem to want everything to be Ap-
plication Class 1. Application Class clearly needs to It is recommended that the concept of Application
be related to soil types and use of the data. Outside Classes and required documentation needs to be bet-
soft clay interpretation for geotechnical parameters ter defined for the next edition of the ISO 22476-
there is no requirement for specifying Class 1. In in- 1:2012 and 19901-8:2014 standards.
terbedded deposits with soft clay layers then the prob-
lem of Class 1 for any soft clay layers has to be ad- 8. References
dressed during the investigation. It should be noted ASTM. 2012. Standard test method for electronic
that Class 1 will not be achievable in these types of friction cone and piezocone penetration testing of
deposits with conventional testing procedures. Addi- soils, ASTM. D5778-12. BS 8422: 2003. Force
tional efforts and deviation from standard techniques measurement – Strain gauge load cell systems –
may help achieving Class 1 in these scenarios, for in- Calibration methods. BSI 31 October 2003.
stance by drilling to the top of a soft clay layer and ISO. 2012. Geotechnical investigation and testing –
doing downhole tests. However, such special efforts Field testing – Part 1: Electrical cone and piezo-
can be time consuming and costly, and these addi- cone penetration tests, International Standard ISO
tional costs need to be addressed contractually. 22476-1:2012. ISO. 2014. Petroleum and natural
gas industries - specific requirements for offshore
If the project's specifications required Class 1 and the structures - Part 8: marine soil investigations, In-
final analyses resulting from Table 3 concludes that ternational Standard ISO 19901-8:2014. ISSMFE
Class 1 has not been achieved a note should be in- International Society for Soil Mechanics and
cluded describing the actual deviation. If for instance Foundation Engineering. 1977. Report of the sub-
the test results are all within Class 1, except for the committee on standardization of penetration test-
friction sleeve readings, then the cone resistance and ing in Europe.
pore pressure parameters can still be used for defining JCGM. 2008 Joint Committee for Guidelines in Me-
soil parameters for design. Application Class analysis trology. Evaluation of measurement data_ Guide
should be performed immediately after a test has been to expression of uncertainty in measurements
completed. At this stage a test can be repeated at rel- JCGM 100:2008
atively little extra cost. Peuchen, J. and Terwindt, J. (2014). Introduction to
CPT accuracy, 3rd International Symposium on
7. Summary and conclusions Cone Penetration Testing CPT14: May 12–14,
ISO 19901-8: 2014 states that an uncertainty analysis 2014 – Las Vegas, Nevada, 45 pp.
should be carried out to show that the accuracy of the Peuchen, J. and Terwindt, J. (2015). Measurement
CPT/CPTU measurements can be within certain val- uncertainty of offshore Cone Penetration Tests.
ues depending on the required Application Class. But Proc. ISFOG 15Frontiers in Offshore Geotechnics
the ISO does not give any guidance on how such an III – Meyer (Ed.) © 2015 Taylor & Francis Group,
uncertainty analysis should be done. Peuchen and Powell, J.J.M. and Lunne, T. (2005). A comparison
Terwindt (2014, 2015) make a serious attempt to of different sized piezocones in UK clays. Proc.
carry out uncertainty analyses. Their work shows XVIth ICSMGE, Osaka, September 2005. Vol. 2,
clearly how complicated such analyses can be, and pp 729-734.
that it will always be based to a certain extent on as- Swedish Geotechnical Society (1993) "Recom-
sumptions and guesswork as to the parameters in- mended Standard for Cone Penetration testing".
cluded in the analyses. In this paper it is argued that a SGF Report 1:93, June 15 1992.
pragmatic approach is more doable and useful in

You might also like