NB-CPD/SG19/09/111r4

GNB-CPD Guidance from the Group of Notified Bodies (SG19 N87) Rev 6
for the Construction Products Directive
SG19 89/106/EEC
Issued: 2011-11-18
SG19 GUIDANCE

EN 13162 through EN 13171,

EN 14064-1, and
EN 14303 through EN 14314
QUESTIONS AND ANSWERS
Revision 1: Q&A nos. 28 and 29 added.
Revision 2: Q&A nos. 30 through 34 added
Revision 3: Q&A nos. 35, 36, and 37 added
Revision 4: Q&A nos. 38 through 43 added
Revision 5: Q&A no. 44 added
Revision 6: Q&A nos. 45 through 47 added

GENERAL
Below is found guidance form Sector group 19 related to the product standards EN13162 through EN13171 used in
combination with EN13172, which is mandatory for CE-marking.
The guidance is given in the form of agreed answers to questions raised within SG19.

In the below text, the form "shall" indicates that all notified bodies are expected to work in accordance with the
answers. The form "should" indicated that the answer is considered a recommendation.

As new questions are raised frequently and the standards are being continuously developed by the working groups
of TC88, this position paper is being developed continuously as well.

Question Answer References

001 At least one sample EN13172 clause B.2.3
EN13172 clause C.2.2
How many samples The number of tests is regulated by EN 13172 clause B.2.3
are needed for (system 1) or clause C.2.2 (system 3). For all other properties than
initial type test of reaction to fire, ITT shall be performed on samples from four
reaction to fire different production dates covering the thickness range declared by
properties? the manufacturer. As regards reaction to fire, only one result is
required. According to the above-mentioned clauses, the result
shall be based on specimens from the different production dates.
This is understood by SG19 as “using specimens from one or more
of the four production dates”. Thus, in cases where reaction to fire
is the only property to test on a particular product, only one sample
is required (unless it is necessary to test more than one thickness).
Approved by SG19 on 2003-06-03
002
Depends on the product
What is the
meaning of the “Production unit” may generally be understood as the production
term “production equipment used for manufacturing the individual product.
unit (line)”
For products manufactured on a production line, the production
unit is evidently identical to production line.
For products with a more complex production process, e.g.
products from a line being cut and/or having a facing added in a
separate department of the factory, the production unit shall be
understood as all the hardware used for the processing of this
particular product.

Page 1 of 18
Question Answer References
New definitions of the terms are planned to be included by the next
edition of EN 13172:
Production line: Assemblage of equipment that produces products
using a continuous process.
Production unit: Assemblage of equipment that produces products
using a discontinuous process.
Approved by SG19 on 2003-06-03.
003 EN 13172 clauses 5.3.5.1
No, an Initial Type Test is only valid for the single production and 6.
May the same ITT
unit from which the samples are taken Annex B of the product
cover products
standards, EN 13162-13171
from more than According to EN 13172 clause 5.3.5.1, samples for direct testing
one production shall be drawn from each production unit (line). Initial type test
unit (line) if they shall be understood as part of the direct testing in the running FPC
are manufactured and shall fulfil the requirement regarding direct testing in the FPC.
in the same plant? For properties with no minimum testing frequency, ITT is the only
requirement stated by table B.1 of the product standards.
Approved by SG19 on 2003-06-03

004 There is a printing error in the standard! EN13162 clause 4.2.3

How to choose the EN 13162 clause 4.2.3 requires that the measuring load shall be
measuring load 50Pa on products with a compressive stress/strength less than 15
when measuring kPa. On products with a higher compressive stress/strength the
thickness on load shall be 250 Pa.
mineral wool? The correct criterion is 10 kPa – not 15 kPa.
Is cases where the exact compressive stress/strength is not known as stated in clause 5.3.2
(e.g. if it is not declared), an estimation of the compressive
stress/strength based on related parameters may form the basis for
the choice of load.
Approved by SG19 on 2003-06-03
Recent development:
In AC:2005, "15 kPa" is replaced by "10 kPa" .
Added 2009-09-23:
In 13162:2008, the criterion is maintained at 10 kPa.
Hereby the error is considered corrected.
Approved by SG19 on 2009-09-23

005 No - lambda 90/90 calculations shall be performed for each EN 13172 clause 5.3.5.1,
lambda group, on each production unit (line) GNB-SG19 N30 rev.
May a lambda
90/90 – calculation Lambda is tested on an ongoing basis as part of the FPC. FPC is
Annexes A and B of the
cover several always carried out separately for the individual production units
product standards
production units (lines). For all other tests in the FPC, the compliance criteria are
(lines) the declared values which shall be fulfilled by every single test
result.
For thermal properties, the 90/90-fractile is the compliance
criterion for the product (group). There are no compliance criteria
for individual test results except for 3rd party ITT where all four
test results shall be better than or equal to the declared value.
Approved by SG19 on 2003-06-03

006 A technical error in EN 13162 EN 13162 clause 4.3.12

Is airflow The EN 13162 is giving a requirement in airflow resistance “AF”

Page 2 of 18
Question Answer References
resistance the right which depends on the dimensions of the test apparatus and
way of declaration? thickness. It is recognised as an error by experts now. The most
useful information is the airflow resistivity “r”. As the standard
cannot be changed quickly and having considered the question
raised, we can make the following proposal:
To declare the airflow resistivity as "AFr", instead of airflow
resistance as AF. The symbol of resistivity is “r” in EN 29053.
This may avoid confusion between AF, which is linked by the EN
13162 to airflow resistance only. By the next revision of the
standard, the “r” may be deleted.
Approved by SG19 on 2003-06-03
Recent development:
In AC:2005, "Airflow resistance" is replaced by "Airflow
resistivity", and the symbol "AF" is replaced by "AFr" .
Added 2009-09-23
In EN 13162:2008, as well the term “Airflow resistivity" as the
symbol ”AFr" are maintained.
Approved by SG19 on 2009-09-23

007 The standards are giving different requirements for number of EN 13163 clause 5 table 13.
measurements to get a test result All other 9 standards
Calculation of
indicate one measurement as
lambda 90/90 The technical bases for declaration and check of the 90/90-fractile
a minimum to get a test
is not uniform because :
result.
The EPS standard states that one test result is the result of one
measurement. For the 9 other of the 10 product families, one test
result can be obtained by averaging a number of measurements.
This will cause a significant reduction of the apparent standard
deviation.
For many products, the standard deviation also depends upon the
measuring area of the apparatus.
Furthermore, the statistical bases are not uniform between ITT,
chapter 7 of EN 13172, audit testing and routine testing.
It is discussed whether it would be wise to recommend an exact
number of measurements to obtain one test result.
After a long discussion in SG-19, it is considered difficult to
advise on this subject.
Manufacturers should be aware that in case of complaints or for
certification purpose, the third party may use n = 1 or a limited
value of n.
Approved by SG19 on 2003-06-03
008 EN 15715
Added 2009-09-23:
EN 13501-1
How to handle the
The new standard EN15715 is describing in detail mounting and EN 13823
mounting and
fixing for the SBI-test. EN 13238
fixing in the SBI-
FSG-recommendation 009
test EN 15715 has taken account of the guidance and recommendation (only for MW)
previously supplied by SG19 and SH02. Therefore, at present no
further need is identified for guidance on that issue.
However, for the completeness the previously issued guidance is
maintained below.

2003 comments:
The mounting and fixing is described in the test standard EN
13823 (standard mounting options). The test conditions shall
reflect the “end-use conditions” (See note 1). The term “end-use

Page 3 of 18
Question Answer References
conditions” has sometimes lead to the misunderstanding that the
building part in which the product is intended to be used shall be
subject to testing, but this is not correct.
All tests supporting the product standards are strictly related to the
product – not to any building parts (See note 2). Thus, the product
itself shall be exposed to the flame in the test, and it shall not be
covered by any additional facings or boards etc.
The test conditions shall reflect the “end use conditions” of the
product, i.e. the fixing method and the type of substrate in the end
use.
The choice of a standard substrate for testing is carried out
according to EN 13823, which also gives the field of application of
the test result using different substrates.
The field of application of the chosen fixing method is limited to
the one used unless a wider field of application is documented, e.g.
by guidance from GNB Fire Sector Group or by testing.
The choice of test conditions – in particular the choice of substrate
- may lead to different classifications.
Following the general “worst case” approach, the substrate and
fixing method giving the worst result shall be chosen.
On the other hand, if a manufacturer should choose to recommend
his product only to be used on non-combustible substrates, it
would be acceptable only to test the product on a non-combustible
substrate. If no specific fixing method is desired, the one giving
the worst result should be used.
Testing on a non-combustible substrate will imply a limitation in
the “field of application” stated in the classification report. Use of
a fixing method not representing the worst case will imply
limitations as well. (to be continued)
In order to avoid a product from being used in applications for
which the classification is not valid, any limitations in the field of
application shall be stated clearly in the manufacturers technical
documentation; in particular in the manufacturers declaration of
conformity.
NOTE 1:
The manufacturer does normally not know the end-use of the product. In general, it
is left to the building designer to decide and describe – in technical terms – the
requirements to the performance of the product. The product shall subsequently be
chosen on the basis of the manufacturer‟s declaration.

NOTE 2:
Performance of building parts and “kits” is not covered by the harmonised product
standards. Please refer to the relevant European Technical Approval and to the
testing standards for resistance to fire.

Approved by SG19 on 2003-06-03

Recent development:
On the basis of answers from the member states‟ fire regulators, Document from the
the European Commission has reached a conclusion, which can be European Commission,
found in the document “RG N322”, dated 30 June 2003. Enterprise Directorate-
General, RG N322
Classification of the product as put on the market will always be
necessary, i.e. the naked product must be classified.
The Additional classification based upon end-use application may
also be included in the CE-marking for those countries which
regulate on end-use application. This additional classification will
be limited by the mounting and fixing configuration used in the
testing and the end-use application(s)simulated.

Page 4 of 18
Question Answer References
(Extract of document RG N322 by Anders Elbek, August 2003)

Annexes to the product standards describing the mounting and

fixing in the SBI-test are being developed in line with the above.
Approved by SG 19 on 2006-09-20
009 The manufacturer Clause 7 of the product
standards
Who is responsible According to clause 7 of the product standards the manufacturer
for the grouping of may choose to group his products in accordance with EN13172,
the products? which allows grouping to be done property by property.
The certification body will not have to approve the manufacturer‟s
grouping.
Approved by SG19 on 2003-06-03
010 System 1: The certification body EN 13172, B.2.3
EN 13172, 5.3.5.2
Who is responsible According to EN13172 clause B.2.3, sampling is done by the
for the sampling certification body with the manufacturer‟s representative present.
for ITT
The grouping done by the manufacturer will form the basis for the
certification body‟s selection of products for initial type testing.
On the basis of the ITT and other evidence, the certification body
is certifying the conformity of the entire product group. The only
way to ensure that all the products are in conformity is - in each
group - to test the products, which have the highest risk of
nonconformity for the particular property.
This is a difficult task requiring a high level of competence from
the certification body (see note 3).

NOTE 3. To assist in the selection, the correlations established by the

manufacturer for indirect testing (cf. EN13172 clause 5.3.5.2) may be
used by the certification body.
Approved by SG19 on 2003-06-03
011 All Relevant Parameters! Annex ZA of the product
standards
Which properties The system for attestation of conformity is only dependent on the
shall the fire properties.
SG19 N54 revised
certification body
For products which are not subject to regulations on reaction to
check under system
fire, AoC-system 3 applies (“Any intended use”, cf. table ZA.2.1)
1+3?
For products in euroclass A1, A2, B and C, AoC-system 1 applies
(unless the two-star footnote can be proven to be applicable) cf.
table ZA.2.2.
In this case the AoC system will formally be system 1 for fire and
system 3 for other properties.
However, Annex ZA states that for products under system 1, all
relevant characteristics shall be of the interest of the approved
body; in particular reaction to fire. For CE-marking, the relevant
characteristics are found in table ZA.1.
It is important to notice that the product is subject to the attestation
of conformity – not the individual properties.
In practice, if a product falls under system 1 for reaction to fire, the
entire product is under system 1 – not only the reaction to fire
properties.
Thus, the declaration (and the certificate) of conformity shall cover
the conformity of the product – not only the reaction to fire class.
Approved by SG19 on 2003-06-03
Added 2009-09-23

Page 5 of 18
Question Answer References
The above understanding was confirmed by representatives of the
Commission in June 2002, cf. SG19 N54revised.
Approved by SG19 on 2009-09-23

012 In general: 4 tests, but under certain conditions it may be EN 13172 clause C.2.2
reduced to 2 tests NB-CPD/01/064r3
How many tests on
NB-CPD/SG19/03/003
thermal The number of tests for ITT is regulated by EN 13172 clause
conductivity / C.2.2. It states that the samples shall be taken from four different
thermal resistance production dates and that the thickness range declared by the
shall be performed manufacturer shall be covered.
by a 3rd party for
In annex ZA, clause ZA.2.2, of the product standards it is stated
ITT?
that initial type test of thermal resistance shall be carried by the
approved body. All initial type testing is per product or product
group per production unit (line).
Thus, the number of tests to be carried out by a third party for ITT
is 4 tests per product (group) for each production unit.
However, where "historical lambda data" are available, it is
possible to limit the number of new tests to be carried out from 4
to 2 tests, provided that the historical test results satisfy a number
of conditions. For the definition of historical data and for further
information on the conditions for using historical data, please refer
to document "NB-CPD/01/064r3 - SG19 Guidance to notified
bodies on the transitional arrangements for the CE-marking of
thermal insulation products".
Approved by SG19 on 2003-06-03
Added 2009-09-23:
The above-mentioned position paper has been approved with the
document number NB-CPD/SG19/03/003.
Approved by SG19 on 2009-09-23
013 When necessary in order to ensure that the worst case is covered EN 15715
When is it In some cases, it is impossible to predetermine the thickness which
necessary to test would result in the worst behaviour in the SBI test. In these cases it
more than one will be necessary to test more than one thickness. In most cases, it
thickness in the will be possible to limit the suspected worst cases to two
SBI test? thicknesses, typically the maximum and minimum thicknesses.
NOTE:
For mineral wool products guidance is found in FSG recommendation 009..

Approved by SG19 on 2004-03-03

Added 2009-09-23:
The above issue is now dealt with by EN15715 and described in
further detail.
Note: Reference to FSG recommendation 009 is no longer relevant and hence
replaced by reference to EN 15715.
Approved by SG19 on 2009-09-23
014 EN13162 clause 4.3.10
Tolerances apply to the individual specimen
What is the All requirements in chapter 4 are related to the test results
meaning of the calculated as the mean value of the individual specimen
tolerances for measurements, except the thermal properties where the 90/90-
compressibility in principle apply.
EN 13162, table 3?
As regards compressibility an additional requirement for the
tolerance of the single measurements apply. E.g., in class CP3 the
compressibility may not exceed 3 mm as an average and the

Page 6 of 18
Question Answer References
compressibility of the individual specimens may not exceed 5 mm.
Approved by SG19 on 2004-03-03
AC:2005, Table 3 is revised in line with the above
Approved by SG 19 on 2006-09-20
Added 2009-09-23:
The above problem has been solved by EN13162:2008 by the
addition of clause 4.3.10.3, note 2
Approved by SG19 on 2009-09-23

015 At an age not exceeding the lowest possible age when delivered to EN13163 clause 4.2.6
a customer
At which age shall
the dimensional
The dimensional stability under constant climatic conditions is
stability of EPS
strongly dependent on the age of the product. The change of
products be
dimensions (in practice: shrinkage) are normally decreasing
determined?
exponentially over time. Age is therefore a very crucial point in
this regard. EN13163 clause 4.2.6 does not specify any limit;
neither does the test standard, EN1603.
Therefore, any product could pass the test provided that the test is
carried out on samples of a certain age.
Referring to the general “worst case” approach, the test should be
initiated before the sample has reached an age which exceeds the
age of the product at the earliest possible time of delivery to the
customer.
Approved by SG19 on 2004-03-03
016 In most cases yes, but there may be exceptions Product standards EN13162-
13171, clause 4.2.1
Are products with
The lambda-value should normally not cause any trouble, as it
special profiling in
does normally not depend on the shaping.
the edges covered
by the standards? The R-value is more difficult. It is not clear how to determine the
R-value on products with a special shaping of the edges; e.g. a
If they are, how to
profiling allowing for special fixing systems. It would obviously
declare thermal
not be correct to determine the R-value as if the cross-section were
properties?
rectangular. On the other hand, if the actual shape is taken into
account in a detailed two-dimensional calculation, it would not
necessarily be more correct as such a calculation could not predict
the influence of fixtures etc.
The SG19 recommends the R-value to be determined on the basis
of the central, uniform, rectangular part of the cross-section. It
should be stated in the declaration of conformity that the R-value
does not apply to the full area of the product and that this should
be taken into account by the building designer. However, the
applicability of the above principle should be considered in each
case.
Approved by SG19 on 2004-03-03
017 Thermal properties, euroclass and dimensions are stated
separately on the label. All other harmonised properties are
Which properties
covered by the designation code
are covered by the
designation code?
The first part of the designation code is the reference to the product
standard, e.g. “EPS-EN13163”. This reference is stating that the
(017 continued) product meets all the basic requirements of clause 4.2 (limit
values).
Additionally, clause 4.2 gives different classes for some properties.

Page 7 of 18
Question Answer References
These classes are indicated by the designation code. E.g. thickness
tolerances are declared by the letter “T” and one digit (T1, T2, T3
etc.).
The “voluntary” harmonised properties in clause 4.3 are always
mentioned in the designation code, when declared.
However, annex ZA is limiting the harmonisation to the properties
mentioned in table ZA.1. This implies that some of the properties
are not referenced by the designation code for CE-marking.
For instance, some of the product standards are giving different
classes for tolerances on squareness, flatness etc. These properties
are not harmonised and therefore they are not covered by the
designation code for CE-marking.
Further, annex ZA is opening for the so-called "NPD-option", even
for some of the harmonised requirements of chapter 4.2.
This is not foreseen by the product standards and the designation
code cannot indicate whether the NPD-option is used on
requirements of chapter 4.2.
Therefore, in order to avoid confusion of the market, it must be
stated clearly on the product label if the NPD-option is used on
harmonised requirements of chapter 4.2.
Examples:
For mineral wool, there are 3 possible declarations on dimensional
stability.
4.2.6. Dimensional stability at 23 C/ 90%RH.
This is a basic requirement covered by the reference “MW-
EN13162”. No additional code.
4.3.2.1 Dimensional stability at 70 C/ 50%RH.
This is a “voluntary” requirement indicated by the letters
“DS(T+)” in the designation code.
4.3.2.2 Dimensional stability at 70 C/ 90%RH.
This is also a “voluntary” requirement indicated by the letters
“DS(TH)” in the designation code.
In case the NPD-option is used regarding dimensional stability, it
must stated clearly on the label . Example: "Dimensional stability:
No performance determined".
For EPS, there are several possible declarations on bending
strength.
4.2.7. Bending strength minimum level at 50 kPa.
This is a basic requirement covered by the reference “EPS-
EN13163” and it may also be indicated by the letters "BS50".
4.3.6 Bending strength declared at levels mentioned in table 7 of
the product standard and indicated by the letters "BSxxx".
(to be continued)
In case the NPD-option is used regarding bending strength, it must
stated clearly on the label. Example: "Bending strength: No
performance determined".
Approved by SG19 on 2005-09-21
018 Declare the “worst case” value or divide the thickness range into
How to declare two or more thickness intervals with different declarations
properties which
Some properties are strongly depending on the thickness.
depend on the
thickness of the This is evidently the case for properties such as dynamic stiffness

Page 8 of 18
Question Answer References
product? and sound absorption, but even reaction to fire and compression
behaviour may in some cases depend on the thickness.
Example:
If, for instance, a product is manufactured in the range from 15 to
35 mm, the dynamic stiffness would always be better on the higher
thicknesses.
There are two ways of handling this problem.
1) Divide the thickness range into minor intervals with individual
declarations. In the interval from 15 to 24 mm the value
obtained by testing 15 mm specimens could be declared. In
the interval from 25 to 35 mm the value obtained by testing 25
mm specimens could be declared. Remember that each
thickness interval should be considered a separate product
(group) as regards dynamic stiffness. Thus, ITT and FPC shall
be performed in each thickness interval.

2) Declare the value obtained by the less favourable thickness of

the entire range.
Approved by SG19 on 2005-09-21
019 The FPC testing frequencies for direct testing may be lowered if Annex B of the product
What is the relation indirect testing is used correctly standards.
between indirect EN 13172 clause 5.3.5.2.
(Results of indirect testing may not be included in the calculation
testing and testing
of lambda 90/90)
frequencies?
Annex B of the product standards lists the testing frequencies for
testing the individual properties in the FPC. Methods for direct
testing are given by reference to the clause in chapter 4.
For some properties, methods for indirect testing and the
corresponding frequencies are stated as well.
The provisions for using indirect testing is given by EN13172,
clause 5.3.5.2:
- Correlation established by suitable statistical means for each
unit (line); re-examined at prescribed intervals and after
changes.
- Sampling plan and compliance criteria defined.
- At least same confidence interval as when using direct testing.
In case of dispute, direct testing shall be carried out.
The frequencies for direct testing, cf. annex B of the product
standards, are given under the presumption that the above
mentioned requirements for indirect testing are fulfilled.
If the requirements are not fulfilled, the last paragraph of
EN13172, 5.3.5.2 applies: In case of dispute, direct testing shall be
used.
In this case, the direct testing shall be performed at the frequency
stated under indirect testing. (to be continued)
Example:
In the mineral wool standard EN13162 compressive strength is
tested according to EN826 once per 8 hours and indirectly by
testing the organic content once per 4 hours and the apparent
density once per hour.
Thus, compressive strength is tested once per hour by using
indirect testing every hour and by direct testing once per 8 hours.
The correlation between density, organic content, and compressive
strength shall be established (and re-examined at prescribed

Page 9 of 18
Question Answer References
intervals) for each production unit. Sampling plan and acceptance
criteria based on the correlation shall be defined.
If the above mentioned requirements are not fulfilled, compressive
strength shall be tested directly once per hour (minimum).
Approved by SG19 on 2004-03-03
020 The test methods referenced by the product standard EN13162, table B.2
Which test methods
When testing fire properties (if classification on reaction to fire is
shall be used for
required), the test methods depend on the expected classification.
reaction to fire of a
mineral wool The test methods for mineral wool are referenced by table B.2 of
product intended EN13162.
for use in a floor;
These test methods apply regardless of the intended application of
e.g. a floor to
the product.
ground insulation.
Approved by SG19 on 2004-03-03
021 Only the lambda value is declared: 2008-versions of the product
How to declare standards, clause 4.2.1
The new amendments and corrections, AC:2005 to the product
thermal properties
standards require only the thermal conductivity declared - not the
of tapered boards?
thermal resistances.
Thermal conductivity is determined as for products with
rectangular cross-section .
Approved by SG19 on 2006-09-20
Added 2009-09-23:
Reference is updated to the 2008-versions of the standards.
Approved by SG19 on 2006-09-23
022 The product standards and the testing standards describe rather
detailed how to measure and declare the geometrical and thermal
How to declare properties of products with rectangular cross sections.
other properties However, products are manufactured with many different
(not thermal) of shapings, e.g. products with profiled edges and taperings (non-
products with non- parallel faces). As only rectangular cross-sections are foreseen by
rectangular cross- the standards, it is necessary to make an interpretation of the
sections? standards to cover these products as well.
Interpretations must be done case by case
Approved by SG19 on 2006-09-20

023 EN 13172 requires inspections "normally be performed twice a EN13172 clause B.2.4.2.
year".
Audit frequencies
In this case, "normally" refers to the situations not covered by
clause B2.4.4, which gives the conditions for extraordinary
inspections.
This means that audits must be carried out twice per year unless
extraordinary inspections are carried out according to clause
B.2.4.4.

A more precise wording is expected by the coming revision of

EN13172.
Approved by SG 19 on 2006-09-20
024 Reference samples must be calibrated by appropriate intervals EN13172, table 1.

Frequencies for According to table 1 of EN13172 reference samples must be

calibrating calibrated at a minimum frequency of once per 3 years.
reference samples It has been acknowledged by TC88/WG16, that this requirement in

Page 10 of 18
Question Answer References
many cases would not be justified.

Therefore, TC88/WG16 has asked SG19 to take a pragmatic

approach to this requirement and accept calibration of reference
samples "as appropriate ".

A revised table 1 without the present right column is expected by

the coming revision of EN13172.
Approved by SG19 on 2005-09-21, cf. GNB SG19 N81

Added 2009-09-23:
EN 13172:2008 has solved the problem as stated above.
Approved by SG19 on 2009-09-23
025 Should the substrate close the corner or should the corner be left EN13823
open? EN13238
Mounting of
substrates for the According to the current standard the corner of the substrate must
SBI test- should the leave an opening at the width of the specimen thickness.
corner be open or At the moment, within CEN TC127 it is being discussed by a task
closed? group to revise the standard with regard to closing the corners, but
no decision has been reached at TC level.

For the time being, the current standard applies, i.e. the substrates
must NOT close the corner.
Approved by SG19 on 2006-09-20
Added 2009-09-23
References corrected
Approved by SG19 on 2009-09-23
026 The peak force is applied - not the force at 10% deflection EN826 figure 1, example b
How to determine
EN 826 gives various examples of typical force-displacement
compressive
curves. For each of the examples, it is indicated at which point to
strength in case a
determine the force to use as basis for the calculated
peak occurs before
strength/stress.
10% deflection
In example b), a peak is found before the deflection has reached
10%.
This example covers in principle all instances of clearly
identifiable decrease in the force.
I.e. if a clearly identifiable decrease is detected, the test may be
interrupted and the peak value used.
If a "break" in the curve occurs without any decrease in the force,
the force corresponding to 10% deflection applies
Approved by SG19 on 2006-09-20
027 Not until the thermal conductivity has reached a stable level EN 13163, clause A.1
At what time after
For some EPS-products, the thermal conductivity may change
production shall
within the first time period after production (so-called
the thermal
“conditioning effect”).
conductivity of
EPS boards be
To declare thermal properties that reflect the values during the
measured?
expected life time of the product (cf. EN 13163, clause A.1), the
manufacturer must ensure that measurements are not carried out
until the thermal conductivity has reached it‟s final level, or that
the conditioning effect is taken into account, e.g. by an accelerated
procedure as described below.

To determine whether or not a conditioning effect is significant,

Page 11 of 18
Question Answer References
the following procedure may be used:

1. Sample conditioning according to Chapter 5.2 of EN 13163

(min. 6 hours, max. 3 days)
2. Measurement of Lambda according to EN 12667
3. Storage at 60 °C for seven days
4. Second measurement of Lambda according to EN 12667

(027 continued) If the increase of thermal conductivity between the above steps 2
and 4 does not exceed 1 %, the conditioning effect may be deemed
insignificant.

If the increase exceeds 1 %, the manufacturer (and the notified

body involved) have to ensure that the conditioning effect is
known, i.e. the time and temperature dependence, e.g. by repeating
steps 3 and 4 until stability is obtained, or by an indirect procedure
like a cell gas analysis.

The necessary conditioning time may vary due to material

properties, e.g. thickness, density, and elastification. For un-
elastified samples of 50 mm thickness and a density range of 15 to
25 kg/m3 the above procedure is deemed sufficient.

Approved by SG19 on 2006-09-20

028 Case by case assessment! SG19 N96
How to speed up
At present, it is being discussed in TC88 WG16 if it would be
the certification
justifiable to adapt the requirements for new production facilities.
process if the
required ITT
In special cases, for manufacturers who have already similar
include long term
products from other production units certified, a certification body
testing ?
may make a provisional decision on certification.
In this regard the risk for the final decision being different from the
provisional one must be identified and dealt with.

The obligation for the manufacturer to apply EN13172 clause 5.5

to any non-conforming product must be emphasised.

In such cases, the final decision by the certification body must be

taken within a specified time period not exceeding six months.

Approved by SG19 on 2008-09-18

029 The manufacturer is free to decide SG19 N96.
If a standard gives
In some standards, there is more than one methods of determining
2 alternative
specific characteristics.
methods, which
one should be
For instance, EN13165 and EN13166 allows for determination of
preferred?
thermal conductivity after accelerated ageing or by adding a fixed
increment.

The certification body cannot require the manufacturer to choose

any particular method.

The manufacturer is free to choose between the methods called-up

by the harmonised standard.

Approved by SG19 on 2008-09-18

030 AoC System 1 applies unless otherwise proven EN13162-13171:
- Clause. 7

Page 12 of 18
Question Answer References
How to determine Background: - Footnotes of table
the correct AoC In the marketplace, similar products are available with similar ZA.2.2
system for products declarations but with different AoC systems applied. - SG19 N27
in Euroclasses A1, - NB-CPD/06/234 rev.1
A2, B, C ? Normally, mineral wool products with declared Euroclass A1 are
CE-marked using AoC system 1. However, in some cases mineral
wool are found CE-marked under AoC System 3.

Answer:
The one-star footnote of table ZA.2.2 of the product standard
points out system 1 for products/materials for which a clearly
identifiable stage in the production process results in an
improvement of the reaction to fire classification (e.g. an addition
of fire retardant or a limiting of organic material).

The above rather cryptic text is applied to a wide range of

construction products and it is not specifically aimed at thermal
insulation products.

In the context of thermal insulation products, it should be

understood that the one-star footnote covers all thermal insulation
products for which the fulfilment of the declared Euroclass
depends on the manufacturer's controlling of the process.

Thus, for any product in Euroclasses A1, A2, B, and C,

attestation of conformity system 1 applies, unless the declared
Euroclass will be met regardless of the manufacturer's
controlling of the production process.
(Often referred to as 'performance not susceptible to changes')

To avoid misinterpretations of the rather cryptic footnotes, note 1

has been added to clause 7 of the product standards requiring the
manufacturer to demonstrate to a notified body that ' no stage in
the production process will result in an improvement of the
reaction to fire classification'.

In this context the 'notified body' is understood as notified

certification body.

A manufacturer wishing to apply AoC system 3 to a product in

Euroclass A1, A2, B, or C should - before he makes the final
decision on choice of AoC system - ask a notified certification
body to verify that the declared Euroclass performance is not
susceptible to changes.

Notified laboratories being requested by a manufacturer to verify

that a product is not susceptible to changes, should not accept to
perform such verification, but advise the manufacturer to contact a
notified certification body.

A certification body must apply the same techniques for the

verification as for conformity certification (evaluation of test
reports, FPC documentation, on-site audit etc.).
A written statement should be issued indicating whether or not the
manufacturer's demonstration has been verified.

Approved by SG19 on 2008-09-18

031 YES - the surface should be retained. EN14706, annex A.
Products delivered with a 'curing oven pattern' may obtain
Should the surface
different test results depending on whether the surface is ground
structure of
smooth before testing or if the surface structure is retained as
specimens be

Page 13 of 18
Question Answer References
retained when delivered.
testing Maximum
Service To facilitate equivalent test results, specimens should have the
Temperature same surface structure as the delivered product.
according to EN No grinding should be carried out before testing.
14706?
Approved by SG19 on 2008-09-18
Added on 2009-09-23:
For mineral wool, however, the below modification apply, cf. EN
14706, A.1.1:
“For wired mats, the wire and the stitching are removed before
testing. Any existing profile on the surface of boards has to be cut
off. Facings are not removed”.

Approved by SG19 on 2009-09-23

032 An emissivity similar to the emissivity of the facing of the
specimen.
Emissivity of
thermo couple
fixings in high When testing according to EN ISO 8497, the emissivity of the
temperature fixing tape may influence the test results.
testing: Hence it is important that the tape used for fixing the
thermocouples does not have a emissivity significantly deviating
What emissivity from the emissivity of the pipe section tested.
should be chosen?
If for instance the specimen is equipped with an aluminium facing,
the thermo couples should be fixed using tape with a similar
surface, e.g. by applying an adhesive to a piece of the aluminium
foil.

Approved by SG19 on 2008-09-18

033 Age at the time of the initial declaration of conformity is decisive. NB-CPD/SG19/03/003
- clause 2 b).
How to interpret
In the approved guidance from SG19 regarding use of historical
the 3 years rule for
data (NB-CPD/SG19/03/003) it is required that historic data used
historical test
as basis for declaring thermal properties should not be more than 3
results ?
years old.

For the implementation of the WG10 standards for building

equipment and industrial insulation similar guidance is being
prepared (NB SG 19 N 108 draft 2009).

The 3 years age should be understood as the age of the test report
at the time of the initial declaration of conformity for the product.

For product under AoC system 3, the age of the data is to be

checked by the manufacturer before issuing of a declaration of
conformity.

For products under AoC system 1, the age of the data is to be

checked by the manufacturer before submitting the data to the
certification body and the certification body verifies before issuing
a certificate that the historic data are not more than 3 years old.

Once the declaration of conformity is drawn up, it is not necessary

to renew the ITT documentation.

Approved by SG19 on 2008-09-18

034 As the question does not require any immediate solution, the
question is referred to TC88 WG10 TG
Which load to

Page 14 of 18
Question Answer References
apply to low-
density products
For testing
Maximum Service Decided by SG19 on 2008-09-18
Temperature?

035 The scope of EN13168 is limited to products with a thermal

conductivity up to 0,1 W/mK.
What is the
maximum lambda
This is indicated by only 1 significant figure, whereas thermal
value for wood
conductivity is declared with 3 significant figures in the case of the
wool under
thermal conductivity being larger than or equal to 0,1 W/m K.
EN13168
For Wood Wool it is confirmed by TC 88 WG 9 that the upper
limit value of application of the EN 13168 is 0,149 W/m K, using
normal rounding rules
Approved by SG19 on 2009-09-23

036 Yes, table ZA.1 in Annex ZA footnote makes this possible, but
Is it acceptable to
SG 19 recommends keeping a unique declaration for a product
omit mandated
including all the declared characteristics whatever country the
characteristics if
product is placed on the market to avoid confusion in the market.
not regulated in a
relevant country
Approved by SG19 on 2009-09-23
037 Water permeability is Water absorption.
How to interpret
The tables ZA.1 and ZA.3 in annex ZA of several of these
the term Water
standards are giving conflicting information. CEN/TC 88 is
permeability in
recommended to clarify this.
annex ZA of
The Water Vapour Permeability is a separate characteristic
EN 14303-14309,
(normally more relevant than water absorption). The reference in
14313 & 14314
the EN 14303-14309, 14313 & 14314standards, both in clause 4
and annex ZA needs to be checked too by TC 88 for clarification.

Approved by SG19 on 2009-09-23

038 To declare thermal conductivity, a lambda curve must be defined Clause 4.2 of EN 14303
through EN 14314
How to measure
Lambda declarations according to EN 14303 though EN 14314 EN 13787
and declare
must be based on a curve be based on measurements at minimum 3
lambda for HVAC
different temperatures. To cover the temperature range,
products used at
measurements at more than 3mean temperatures may be needed,
40 or 50 C? cf. EN 13787.
Is it necessary to The above principle is different from the standards for building
declare a curve insulation products according to which declarations of thermal
and measure many properties are based on measurements at a single mean
points? temperature - 10 degree C.

Text proposed by Anders Elbek on 2011-01-26

039 If the thermal conductivity is not declared, the product is not Clause 4.2 of all CEN/TC
covered by the product standards 88 Product standards.
Is it possible to
declare a For all TC88 standards, declaration of thermal resistance/thermal
maximum service conductivity is obligatory. Products without declared thermal
temperature characteristics are not covered by the TC88 standards and cannot
without declaring be CE-marked with reference to TC88 standards.
thermal If a Maximum Service Temperature (MST) is declared, the lambda
conductivity. Is the curve must be declared for the temperature range up to the MST.
lambda curve However, as the MST is defined by the hot side of the test

Page 15 of 18
Question Answer References
necessary? specimens, the lamda is defined at the mean temperature. Thus the
MST can exceed the highest temperature for which lambda is
declared. For Flat products the difference between MST and the
upper end of the lambda curve would normally be smaller than for
pipe sections.

Text proposed by Anders Elbek on 2010-09-24

040 No, the number of tests are fixed by the standards
May the number Clause 4.2.4 of the product standards refer to EN 15715 that
of fire tests on pipe requires 3 measurements for each thickness interval.
sections be The document SG 19 N 113 gives a more detailed description
reduced if the
‘worst case’ is Text proposed by Anders Elbek on 2010-09-24
known?
041 The manufacturer must decide

How to determine Some pipe sections are made with a flexible inner diameter
the dimensions on allowing the pipe sections to be used on pipe diameters varying
pipe sections that over a narrow range
may be used on a
range of pipe The manufacturer is entitled to decide on separate declarations
diameters? depending on the diameter of the pipe.
As the thickness of the pipe may influence several declared
characteristics the conformity of the product must be evaluated for
each thickness, characteristic by characteristic.

However, the manufacturer may also decide for each

characteristic to declare a worst case covering the span of
thicknesses for which the pipe section may be used.

Text proposed by Anders Elbek on 2010-09-24

042 The requirements do not depend on the organisation behind the
production.
How to deal with
small companies The manufacturer placing the finished product on the market is the
producing on the manufacturer in the sense of the CPD. This means that he will
basis of finished have full responsibility for the entire product, its performance,
products delivered conformity and the CE-marking.
by another The manufacturer is of course free to subcontract any part of the
company, e.g. by manufacturing process and/or the FPC, but that does not reduce
cutting pipe the extent of the documentation he must be able to present.
sections from EPS
block Text proposed by Anders Elbek on 2010-09-24
043 The manufacturers choice EN 14303 clause 4.2.2.1

Which load should EN 14303 gives two options; 50 Pa or 1000 Pa.

be applied when
measuring The manufacturer may choose on of the two options – or in special
thickness of wired cases - even both of them.
matts? If both options are used, the declarations of all other
characteristics influenced by the thickness must be declared
separately or on the safe side covering both options.

At the next NB SG 19 meeting (September 2011), the MW

association Eurima is expected to present a proposal for further
development of the above answer.

Text proposed by Anders Elbek on 2011-01-26

Page 16 of 18
044 EN 14064-1 gives two possibilities for determining the minimum EN 14064-1, table 2
How to determine densities for installing in closed cavities and frame constructions.
minimum densities
for mineral wool 1) Densities confirmed by relevant authorities, or
loose-fill in cavities 2) Tabulated values, cf. table 2 of EN 14064-1
or closed
compartments As the authorities of some member states do not „confirm‟ specific
densities and as no harmonized test methods have been developed
for settlement in cavities or closed compartments, the criteria for
applying the above option 1 remain unclear.

The tabulated densities of table 2 of EN 14064-1 are conservative

safe-side values taking into account that loose-fill products may be
made from non-specified mineral wool raw materials spread over a
wide range of densities and maybe even including waste from
other mineral wool products.

It is however a well-known fact that it is possible to produce loose-

fill mineral wool with much lower stable, settlement free densities
by applying a continuous production process and by controlling
the intermediate stages of the production.

In such cases, the tabulated values would often be far too

conservative or even impossible to apply in practice.

Therefore, it is necessary to have the possibility to declare lower

densities – not only in cases where specific densities are
confirmed by national authorities.

Guidance:

Notified bodies may accept a declaration of class S1 for lower

densities providing that the below criteria are met:

1) Tests according to national test methods support the

declaration of class S1 (“no measurable settlement”) at
the specified density.
(As climatic changes would normally not cause any
significant settlement for mineral wool, test methods
based on vibration or mechanical impact should be
preferred).
2) The loose-fill product is based on mineral wool designed
specifically for obtaining lower density when installed
on-site.
3) Continuous production process with control of density at
intermediate stages including control of density after
curing.
4) Installed density not lower than 1,5 times the density after
curing.
5) Density not lower than 1,25 times the declared minimum
density for the same product to meet class S2 for loft
application.
6) The manufacturer demonstrates that complaints or feed-
back from clients, users, or other stakeholders do not raise
any doubt as to the fulfilment of class S1.

Text proposed by Anders Elbek on 2011-09-08

Page 17 of 18
045 Compressive creep needs only be tested if declared by the EN 13162 clause 4.2.9.4
Is it obligatory to manufacturer. EN 13162 clause 4.3.6
test ‘compressive EN 13162 Table ZA.1
creep’ on mineral Although EN 13162:2008, clause 4.2.9.4 identifies the EN 13162 Table ZA.3
wool products? requirements on compressive creep as covering the requirements
on durability of compressive strength, the manufacturer is free to
If ‘compressive decide whether or not to declare any performance with regard to
creep’ is tested, compressive creep.
should the test be
performed by the Clause 4.2.9 was introduced by the 2008-version due to the use of
notified body or a new template. It was not the intention to make it obligatory to
the manufacturer? test compressive creep.

In case the manufacturer chooses to declare a compressive creep

performance, the initial type test will fall under the responsibility
of the manufacturer – not the notified body, cf. table ZA.3

Text proposed by Anders Elbek on 2011-11-18

046 The normal rounding rules apply EN 14064-1 clause 4.2.3.2
What rounding EN 14064-1 annex K
rules apply to the EN 14064-1 clause 4.2.3.2 defines 3 classes for settlement:
settlement classes S1: ≤1%;
of EN 14064-1 S2: > 1% and ≤ 5 %.
S3: > 5% and ≤ 10%

As the limits between the classes are given as integer percentages,

the test results must be rounded to the nearest integer percentage
before the class is chosen.

Example:
A MW product is tested in accordance with annex K and the test
result is a settlement of 1,31 % (annex K does not define how many
significant figures to indicate).
Following the normal rounding rules, 1,31 % is rounded
downwards to 1 %, and class S1 may be chosen.

Text proposed by Anders Elbek on 2011-11-18

047 The normal rounding rules apply EN 14064-1 annex H
What settlement EN 14064-1, clause 4.2.3.2
values should form In performance charts for loft applications, the “minimum
basis for installed thickness” is calculated by adding the expected
performance settlement to the “thickness after settlement” using formula H.2:
charts for loft dbefore = d/(1-s)
applications. The
declared class or The settlement, “s” should be the upper limit value for the
the ITT-result? declared class of settlement, i.e. classes

S1: s=1%
S2: s=5%
S3: s=10%

By defining only three classes, of settlement, EN 14064-1 is

reflecting the fact that settlement over time is difficult to predict
with any high accuracy.

Text proposed by Anders Elbek on 2011-11-18

Page 18 of 18