Marine Fuels Information Bulletin 1

ISO 8217: Test Methods

Precision and Interpretation

The International Organization for Standardizations (ISOs) 8217

standard lists the required specifications for marine fuel oils and the
standard test methods used to assess them.
Most bunker fuel orders specify a fuel quality defined
by ISO 8217 and many of Chevrons customers have
their bunker fuel deliveries analyzed by an independent
laboratory to comply with ISO 8217.

method. True value is the average of an infinite number

of single test results obtained by an infinite number
of laboratories and, therefore, cannot be exactly
established.

It is Chevrons policy to supply bunker fuels that comply

with the specifications agreed upon when the order is
accepted and confirmed. Our terms of sale state that
only retain samples provided by/on behalf of Chevron
at the time of bunkering are considered valid samples.
However, a test analyzed by a customers chosen lab on
the customers retain sample may fail a specification
limit, while Chevrons retain sample analysis shows the
product to meet specifications. How is this possible?

What defines the precision of a test method?

For each test method repeatability and
reproducibility values are defined and listed.
Repeatability is the difference between consecutive
test results obtained by the same operator using the
same apparatus under constant operating conditions
on identical test material. It is only relevant for the
individual testing laboratory.

If the samples were properly taken and closely

represent the quality of the fuel delivered, the answer
is found in the ISO 8217 reference to the individual test
methods to be used, and in the ISO 4259:2006 standard
on the determination and application of precision
data in relation to methods of test. If the prescribed
test method defined in ISO 8217 was not used by one
of the two parties, that party should obtain the result
using the prescribed test method before initiating any
discussion.
The following question and answer format describes
how to interpret a test methods precision and
specification limit. Please see ISO 4259:2006 for
additional information about the interpretation of
test results.
Is every single test result the true value?
No, every single test result is only an approximation
of the true value. The intrinsic limitations of the test
result are determined by the precision of the test

Reproducibility is the closeness of agreement between

individual results obtained in the normal and correct
operation of the same method on identical test material
(different operator, different apparatus, and different
laboratories).
How is test reproducibility used to judge the
acceptability of two test results?
If the difference between two single test results
(obtained with the same method in two different
laboratories, for example, a suppliers test result and
a customers test result) is less than or equal to the
reproducibility R of the test method, then both test
results are considered acceptable. Their average
value is considered the estimated value of the tested
property.

Example 1: RMG 380 test density determination

Lab 1: Density at 15C, ISO 12185:1996: 991.4 kg/m3
Lab 2: Density at 15C, ISO 12185:1996: 990.2 kg/m3
Listed reproducibility R of method ISO 12185:1996:
R = 1.5 kg/m3
The difference between the two single test results,
991.4 kg/m3 990.2 kg/m3 = 1.2 kg/m3, is less thanR.
Hence, the average value of the two test results is
considered the estimated value of the density:
(991.4 kg/m3 + 990.2 kg/m3) 2 = 990.8 kg/m3
If the two results differ by more than R, both results
are considered to be suspect and, in theory, each
lab will then apply the procedures outlined in ISO
4259:2006. In practice, both parties will generally
agree to accept the result obtained by a mutually
selected third laboratory on a valid retain sample. In
this case, it is important to ascertain that the third
laboratory is using the prescribed test method and
has accepted quality certificates.
Is it correct to say that a test result meets
a specification requirement if it exceeds the
specification limit by a value less than R?
No. Reproducibility is only applicable on two
analysis results, not on an analysis result and a
specification limit.
Is it possible to estimate the products compliance
to a specification limit when only one single result
is available?
Yes. ISO 4259:2006 includes supplier and customer
guidelines to evaluate a product against a specification
limit, using the customers single test result. It is
assumed that:
The sample used for the analysis is a valid retain
sample.
The laboratory has performed the test under the
conditions fully prescribed by the test method.
How can a supplier make an evaluation of maximum
specification limit when there is only a single test
result which is smaller than the specification limit?
The supplier can consider, with 95% confidence, that
his or her product meets a maximum specification limit
(A1) if the test result X A1 0.59R, where R is the
reproducibility of the test method (see Example 2). If the
suppliers single test result X' is such that A1 0.59R <
X' A1, then there is a calculated risk that the product
would fail the specification limit during multiple testing.

Example 2: Supplier with a single test result for

density
Case 1
Density specification limit (A1): 991.0 kg/m3 @ 15C
R (test method ISO 12185:1996): 1.5 kg/m3
Single test result X available on a representative sample
for delivery: 990.0 kg/m3 @ 15C
It is 95% likely that this product will meet the
specification limit upon multiple testing, because the
condition X A1 0.59R is fulfilled.
(990.0 < 991.0 0.59 1.50 or 990.0 < 991.0 0.9 or
990.0 < 990.1)
Case 2
Density specification limit (A1): 991.0 kg/m3 @ 15C
R (test method ISO 12185:1996): 1.5 kg/m3
Single test result X' available on a representative sample
for delivery: 990.5 kg/m3 @ 15C
It is likely that this product would fail the specification
limit during multiple testing because the condition
X' A1 0.59R is not fulfilled.
(990.5 exceeds A1 0.59R or 990.5 > 991.0 0.9 or
990.5 > 990)
How can a customer make an evaluation about a
maximum specification limit when there is only one
single test result which exceeds the specification
limit?
The customer can consider, with 95% confidence,
that a product will fail the maximum specification limit
requirement upon multiple testing if the single test
result X > A1 + 0.59R (where A1 is the maximum limit
and R is the reproducibility of the test method). If a
customers single test result X' meets the condition
A1 < X' A1 + 0.59R, there is a statistical chance that
multiple testing would still show the product to be
on spec.
Example 3: Customer with a single test result for
density
Case 1
Density specification limit (A1): 991.0 kg/m3 @ 15C.
R (test method ISO 12185:1996): 1.5 kg/m3
Single test result X available: 992.2 kg/m3 @ 15C
It is 95% likely that the product will fail the specification
limit upon multiple testing, because X > A1 + 0.59R.
(992.2 > 991.0 + 0.59 1.5 or 992.2 > 991.0 + 0.9 or
992.2 > 991.9)

2 | Marine Fuels Information Bulletin 1: ISO 8217: Test Methods Precision And Interpretation

Case 2
Density specification limit (A1): 991.0 kg/m3 @ 15C.
R (test method ISO 12185:1996): 1.5 kg/m3
Single test result X' available: 991.4 kg/m3 @ 15C.
The single test result exceeds the specification limit A1;
however, because it is lower than A1 + 0.59R, multiple
testing may still result in finding the density on spec
(A1 + 0.59R X' > A1).
(991.0 + 0.59 1.5 > 991.4 > 991.0 or 991.0 + 0.9 >
991.4 > 991.0 or 991.9 > 991.4 > 991.0)

For the evaluation of one single test result X with a

minimum specification A2, the conditions applied by the
supplier and customer are, respectively, X A2 + 0.59R
and X < A2 0.59R.
Summary:
Test results deviate from the true value; this is
inherent to all test methods. The precision statements
are an integral part of every official test method and
listed in the method. For more information, please see
Appendix: Reproducibility of ISO 8217 or Equivalent
IP/ASTM Test Methods.

Appendix: Reproducibility of ISO 8217 or Equivalent IP/ASTM Test Methods*

Density at 15C, kg/m3
1. ISO 3675
For transparent, nonviscous products:
R = 1.2 kg/m3 or 0.0012 kg/l
For opaque products:
R = 1.5 kg/m3 or 0.0015 kg/l
2. ISO 12185
For transparent middle distillates:
R = 0.5 kg/m3
or 0.0005 kg/l
For crude oils and other petroleum products:
R = 1.5 kg/m3 or 0.0015 kg/l
Kinematic viscosity, mm2/s
ISO 3104
Distillate fuels at 40C:
R= 0.0082 (x + 1)
Where x is the average of the results
being compared
Heavy fuels at 50C:
R = 0.074x
Where x is the average of the results
being compared
Flash point, P.M., closed tester
ISO 2719
Procedure A (distillate fuels):
0.071x
Where x is the average of the results (C)
being compared
Procedure B (residual fuel oils):
6C
Pour Point, C
ISO 3016
R = 6.59C

Sulphur, % (m/m)
1. ISO 8754
For values > 0.05% (m/m) and 5.00% (m/m):
R = 0.0812 (x + 0.15)
Where x is the average of the results
being compared
For values 0.03% (m/m) and 0.05% (m/m):
R = 0.1781 (x + 0.05)
Where x is the average of the results
being compared
2. ISO 14596
For sulphur content in the range 0.100.99 m/m %:
R = 0.02
For sulphur content in the range 1.002.50 m/m %:
R = 0.04
H2S, mg/kg
IP 570
For residual fuel oils:
R = 0.5232x0.6
Where x is the average of the results
being compared
Cetane index (4 variable equation)
ISO 4264
Precision depends on:
The precision of the original density
The distillation recovery temperature
determinations that enter into the calculation
Oxidation stability, g/m3
ISO 12205
R = 10.6 (x/10)0.25
Where x is the average of the results being
compared
Micro carbon residue, % (m/m)
ISO 10370
R = X2 3 0.2451
Where X is the average of the results
being compared

Cloud Point, C
ISO 3015
For distillate fractions:
R = 4C
* Edition of test methods as cited in ISO 8217:2010 paragraph 2. Normative references

Marine Fuels Information Bulletin 1: ISO 8217: Test Methods Precision And Interpretation | 3

Ash, % (m/m)
ISO 6245
For ash content between 0.001 and 0.079 wt%:
R = 0.005
For ash content between 0.080 and 0.180 wt%:
R = 0.024
Lubricity, corrected wear scar diameter, m
ISO 12156-1
R = 102m
CCAI (Calculated Carbon Aromaticity Index)
Precision (R) depends on the precision (R)
of the density and viscosity of the residual
fuel oil that enter into the calculation
Total existent sediment, % (m/m)
ISO 10307-1
Heavy fuels:
R = 0.294 x
Distillate fuels containing heavy components:
R = 0.174 x
Where x is the average of the test results
in %(m/m)
Total sediment potential ageing, % (m/m)
ISO 103072
Heavy fuels:
R = 0.294 x
Distillate fuels containing heavy components:
R = 0.174 x
Where x is the average of the test results
in %(m/m)
Water, % (v/v)
ISO 3733
Water collected between 0.0 and 1.0 ml:
R = 0.2ml
Water collected between 1.1 and 25 ml:
R = 0.2 ml or 10% of mean, whichever
is greater

Vanadium, mg/kg
1. ISO 14597
Applicable to products having V content in
the range of 5 to 1,000 mg/kg, although
reproducibility data have only been determined
up to 100 mg/kg for V
For V content between 5 30 mg/kg:
R = 5 mg/kg
For V content between 31 100 mg/kg:
R = 10 mg/kg
2. IP 501
R = 1.6799 x 0.6
Where x is the average of the results
(mg/kg) being compared
3. IP 470
R = 3.26 x0.5
Where x is the average of the results
(mg/kg) being compared
Cat fines Al + Si, mg/kg
1. ISO 10478, IP 501

ICP detection:

Al: R = 0.337x

Si: R = 0.332x
2. IP 470
AAS detection:
Al: R = 0.789x 0.67
Si: R = 1.388x 0.67
Where x is the average of the results (mg/kg)
being compared
Ca, Zn, P, mg/kg
1. Ca
IP 501 (ICP) R = 0.6440x0.65
IP 470 (AAS) R = 1.139x 0.8
2. Zn
IP 501 (ICP) R = 0.5082x0.7
IP 470 (AAS) R = 0.580x 0.75
3. P

Acid number, mg KOH/g

ASTM D664
R (used oils buffer endpoint) = 0.44x
Where x is the average of the results
being compared

IP 501 (ICP) R = 1.2765x 0.55

IP 500 (UV) R = 1.2103x 0.4
Where x is the average of the results (mg/kg)
being compared
Sodium, mg/kg
1. IP 501 (ICP)
R = 1.0667x0.55
2. IP 470 (AAS)
R = 1.303x0.6
Where x is the average of the results being
compared

4 | Marine Fuels Information Bulletin 1: ISO 8217:

Test Methods Precision And Interpretation

20072011 Chevron U.S.A. All rights reserved. All trademarks are the
property of Chevron Intellectual Property LLC or their respective owners.