EN 17038-1:2019

SLOVENSKI STANDARD
01-julij-2019
Črpalke - Metode za kvalifikacijo in verifikacijo indeksa energijske učinkovitosti
centrifugalnih črpalk - 1. del: Splošne zahteve in postopki za preskušanje in
izračun indeksa energijske učinkovitosti (EEI)
Pumps - Methods of qualification and verification of the Energy Efficiency Index for
rotodynamic pumps units - Part 1: General requirements and procedures for testing and
calculation of energy efficiency index (EEI)
Pumpen - Methoden zur Qualifikation und Verifikation des Energieeffizienzindexes für
Kreiselpumpen - Teil 1: Allgemeine Anforderungen und Vorgehensweisen zur Prüfung
und Berechnung des Energieeffizienzindexes (EEI)
Pompes - Méthodes de qualification et de vérification de l'indice de rendement
énergétique des groupes motopompes rotodynamiques - Partie 1 : Exigences générales
et procédures d'essai et de calcul de l'indice de rendement énergétique (EEI)
Ta slovenski standard je istoveten z: EN 17038-1:2019
ICS:
23.080 Črpalke Pumps
27.015 Energijska učinkovitost. Energy efficiency. Energy
Ohranjanje energije na conservation in general
splošno
2003-01.Slovenski inštitut za standardizacijo. Razmnoževanje celote ali delov tega standarda ni dovoljeno.

EN 17038-1
EUROPEAN STANDARD
NORME EUROPÉENNE
May 2019
EUROPÄISCHE NORM
ICS 23.080
English Version
Pumps - Methods of qualification and verification of the
Energy Efficiency Index for rotodynamic pump units - Part
1: General requirements and procedures for testing and
calculation of Energy Efficiency Index (EEI)
Pompes - Méthodes de qualification et de vérification Pumpen - Methoden zur Qualifikation und Verifikation
de l'indice de rendement énergétique des groupes des Energieeffizienzindexes für Kreiselpumpen - Teil 1:
motopompes rotodynamiques - Partie 1 : Exigences Allgemeine Anforderungen und Vorgehensweisen zur
générales et procédures d'essai et de calcul de l'indice Prüfung und Berechnung des Energieeffizienzindexes
de rendement énergétique (EEI) (EEI)
This European Standard was approved by CEN on 15 July 2018.

CEN members are bound to comply with the CEN/CENELEC Internal Regulations which stipulate the conditions for giving this
European Standard the status of a national standard without any alteration. Up-to-date lists and bibliographical references
concerning such national standards may be obtained on application to the CEN-CENELEC Management Centre or to any CEN
member.
This European Standard exists in three official versions (English, French, German). A version in any other language made by
translation under the responsibility of a CEN member into its own language and notified to the CEN-CENELEC Management
Centre has the same status as the official versions.

CEN members are the national standards bodies of Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia,
Finland, Former Yugoslav Republic of Macedonia, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania,
Luxembourg, Malta, Netherlands, Norway, Poland, Portugal, Romania, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland,
Turkey and United Kingdom.
EUROPEAN COMMITTEE FOR STANDARDIZATION
COMITÉ EUROPÉEN DE NORMALISATION

EUROPÄISCHES KOMITEE FÜR NORMUNG

CEN-CENELEC Management Centre: Rue de la Science 23, B-1040 Brussels
© 2019 CEN All rights of exploitation in any form and by any means reserved Ref. No. EN 17038-1:2019 E
worldwide for CEN national Members.

Contents Page
European foreword . 3
Introduction . 4
1 Scope . 5
2 Normative references . 5
3 Terms and definitions, symbols and subscripts . 5
4 Energy Efficiency Index (EEI) calculations . 9
4.1 General information and explanation on EEI . 9
4.2 Weighted-average value of the electric power input . 9
4.3 The reference electric power input P . 10
1, ref
5 Qualification of pump unit types in respect to their Energy Efficiency Index . 10
5.1 General remarks . 10
5.2 Qualification methods . 11
6 Verification of the Energy Efficiency Index for pump units . 12
6.1 General remarks . 12
6.2 Procedure and decision . 12
Annex A (informative) Mean value and confidence interval of the Energy Efficiency Index EEI . 14
Annex B (informative) Methods recommended for the qualification of a pump unit type by
test . 19
B.1 General remarks . 19
B.2 Determination of the mean value of EEI of a pump unit type population from a test
on only one test pump unit . 19
B.3 Determination of the mean value of EEI of a pump unit type population from a
sample test of M pump units . 20
Annex C (informative) Application of mathematical statistics on tests . 23
C.1 Purposes of applying statistics in the frame of qualification and verification . 23
C.2 Normal or Gaussian distribution . 23
C.3 Confidence interval . 24
C.4 Outlier test . 25
Annex D (informative) Measurement uncertainties . 27
Annex E (informative) The probability of the result of the verification procedure . 30
E.1 General . 30
E.2 Probability that the conformity to law is confirmed by verification tests . 32
E.3 Probability that a numerical value of indicated by the responsible company is
confirmed by verification tests . 33
Bibliography . 34

European foreword
This document (EN 17038-1:2019) has been prepared by Technical Committee CEN/TC 197 “Pumps”,
the secretariat of which is held by AFNOR.
This European Standard shall be given the status of a national standard, either by publication of an
identical text or by endorsement, at the latest by November 2019, and conflicting national standards
shall be withdrawn at the latest by November 2019.
Attention is drawn to the possibility that some of the elements of this document may be the subject of
patent rights. CEN shall not be held responsible for identifying any or all such patent rights.
According to the CEN-CENELEC Internal Regulations, the national standards organisations of the
following countries are bound to implement this European Standard: Austria, Belgium, Bulgaria,
Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, Former Yugoslav Republic of Macedonia,
France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta,
Netherlands, Norway, Poland, Portugal, Romania, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland,
Turkey and the United Kingdom.
Introduction
This document is the first part of a series of standards describing a methodology to evaluate energy
efficiency performance of pump units, comprising the pump, the motor with or without frequency
converter, based on a non-dimensional numerical value called Energy Efficiency Index (EEI). An EEI
allows the comparison of different pump sizes and types with one common indicator. Physical
influences such as pump size, specific speed, pump unit part-load operation, motor-efficiency
characteristic and frequency converter influence are implemented into this indicator.
This standard series covers pump units which component wise can be placed on the market and
assembled into a pump unit or are placed on the market as one single product. To cover these two cases
a semi-analytical model has been developed in order to derive the EEI based on nominal data of the
components only. This case happens for example when an assembler of the components builds the
pump unit on site and consequently a product test cannot be done by for this assembler or when a
manufacturer is quoting a pump-unit which uses a combination of components not previously built.
This particular standard gives an overview of the basic concept of EEI (Clause 4), basic concepts of flow-
time profiles and reference pressure control curves, the qualification (Clause 5) of a pump unit type
regarding an EEI and the verification of EEI values (Clause 6) given for a pump unit type either by
measurement or calculation.
Specific requirements for testing, a calculation method for an EEI, the so called semi-analytical model of
a complete pump unit, specific flow-time profiles and reference control curves are given in the
subsequent parts of this standard series.
1 Scope
This document describes a methodology to evaluate energy efficiency performance of pump units based
on a non-dimensional numerical value called Energy Efficiency Index (EEI).
This document covers pump units consisting of:
— one single or several rotodynamic water pump(s), including where integrated in other products,
and driven by a motor system, consisting of an electrical motor, and either:
— a terminal box which only enables to operate the pump unit at constant motor stator frequency
and thereby (nearly) constant rotational speed, or
— a CDM (Complete Drive Module) which enables to operate the pump unit at variable rotational
speed depending on a varying demand of flow rate and/or discharge or differential pressure.
NOTE A CDM is also often called VSD (Variable Speed Drive).
Pump units as defined above are treated as extended products in respect to their energy efficiency.
2 Normative references
There are no normative references in this document.
3 Terms and definitions, symbols and subscripts
For the purpose of this document, the following terms and definitions apply.
ISO and IEC maintain terminological databases for use in standardization at the following addresses:
— IEC Electropedia: available at http://www.electropedia.org/
— ISO Online browsing platform: available at http://www.iso.org/obp
3.1
fixed speed
constant motor stator frequency
3.2
variable speed
varying motor stator frequency
3.3
Energy Efficiency Index
EEI
ratio between P and P
1, avg 1, ref
3.4
reference electrical power input P
1, ref
calculated power input of the pump unit at the best efficiency point
3.5
average electrical power input P
1, avg
power input weighted by the load profile
3.6
pump unit
pump driven by a motor system
3.7
motor system
combination of an electric motor and either a terminal box or a CDM
3.8
pump unit type
identical pump units of the same configuration
3.9
pressure control curve
functional dependency of the demanded head H vs. the delivered flow rate Q in the form of a non-
dimensional correlation H/H = f (Q/Q )
100% 100%
3.10
reference pressure control curve
pre-defined functional dependency H/H = f (Q/Q ) for standardized measurements and
100% 100%
calculations of average power input P
1, avg
3.11
Complete Drive Module
CDM
electronic power converter connected between the electric supply and a motor as well as extensions
such as protection devices, transformers and auxiliaries
[SOURCE: EN 61800-2]
3.12
Power Drive System
PDS
combination of CDM and motor
Table 1 — List of basic letters and subscripts
Symbol Quantity Unit
C Constant dimensionless number
D Diameter m
e Measurement uncertainty, relative value dimensionless number
E Overall measurement uncertainty dimensionless number
−1
f Frequency
s , Hz
g Acceleration due to gravity
m/s
H Pump total head m
k Number of sample pumps dimensionless number
m Mass kg
M Number of pumps in a sample dimensionless number
−1 −1
n Speed of rotation
s , min
N Number of instrument readings dimensionless number
n Specific speed min−1
s
p Pressure Pa
p Probability density dimensionless number
P Power W
Q (Volume) rate of flow
m /s
s Standard deviation of a sample according to special quantity
t Tolerance factor, relative value dimensionless number
t Time s
t Student’s factor dimensionless number
T Torque Nm
u Peripheral velocity m/s
U Mean velocity m/s
U Voltage V
v Local velocity m/s
V Volume
m
x General quantity according to specified usage
y General quantity according to specified usage
z Height above reference plane m
z number of produced pumps dimensionless number
η Efficiency dimensionless number
θ Temperature °C
ν Kinematic viscosity
m /s
ρ Density
kg/m
ω Angular velocity rad/s
σ Standard deviation of normal distribution according to special quantity
Table 2 — List of letters and figures used as subscripts
Subscript Meaning
1 electrical
2 mechanical
abs absolute
allowable permitted e.g. by regulations
declared indicated by the manufacturer
verification permitted in verification process
individual specific to one pump unit
amb ambient
annual per year
BEP best efficiency point
dr drive
D datum
exp experimentally determined
G guaranteed
H pump total head
I numbering index
J numbering index
imp impeller
man manufacturing
max maximum permissible
mean arithmetic mean value
meas measured value
min, requ minimum required
N nominal
OL overload
Pd pre-defined
P power
PL part load
Q (volume) flow rate
R random
S specific, systematic
sync synchronous
tot total, overall
true true value
T torque
v vapour
x of quantity x
y % for probability of y %
4 Energy Efficiency Index (EEI) calculations
4.1 General information and explanation on EEI
The ratio of P and P is a quantitative indicator for the energy efficiency of pump units. It is
1, avg 1, ref
called Energy Efficiency Index (EEI) and is defined by Formula (1):
P
1, avg
(1)
EEI=
P
1,ref
is compared
In order to assess the efficiency of pump units, their actual electrical power input P
1, avg
with a reference value called P . It is a nominal power input of a fictitious pump unit running at
1, ref
nominal 100 % point.
The value of P for the different pump types are defined in other parts of this standard.
1, ref
The declared EEI value is based on the actual pump curves whether these are realized with full
diameter impeller, a trimmed impeller or a power limited curve.
NOTE Contrary to the efficiency values η of the unit components, the lower the numerical value of EEI, the
better the energy efficiency of pump units.
4.2 Weighted-average value of the electric power input
The quantification of energy efficiency of pump units should be based on the weighted average of
electrical power input P .
1, avg
The weighting factor when averaging the electrical power input P is the fraction of operating time at
the respective flow rate. Generally, for any application there is a continuous function describing the
dependency of (differential) time fraction dt/t on the relative flow rate Q/Q where t is the
tot 100 % tot
total operating time and Q is defined in the specific parts of this standard.
100 %
For the assessment of the energy efficiency of pump units, common and standardized “flow-time
profiles” are needed as basis for indicating their energy efficiency. The flow-time profile is defined for
each pump type in other parts of this standard.
In order to simplify further the determination of a value which characterizes the energy efficiency of
pump units, in this document the different flow-time profiles are described by a limited number of
points, not by a continuous curve or function. These discrete points are pairs of values of Q/Q and
100 %
Δt/t . These values define the discretised flow-time profiles. They are given as reference flow-time
tot
profiles in the specific parts of this standard. (In this case, Δt is a finite fraction of operating time and
Q/Q is the average value of relative flow rate demanded within this time fraction.)
100 %
NOTE 1 The reference flow-time profiles defined in this document for variable flow operation reflect the
typical ranges and variations of demanded flow rate in these types of applications and are based on experimental
field studies, see [1], [2].
For a given reference flow-time profile, the weighted average of the electric power input P is
1, avg
calculated by Formula (2):
N


∆t

P =  ⋅ P (2)
1,avg
∑ 1, i


t
tot
i=1i
where
i is the consecutive number of the points of the flow-time profile and N is the total number of the
points of the flow-time profile.
For calculating P , the electric power input P shall be known for each value Q/Q of the
1, avg 1 100 %
reference flow-time profile.
But as for the flow-time profiles, common and standardized correlations H/H = f (Q/Q ) are
100 % 100 %
needed as basis for indicating and comparing the energy efficiency of pump units placed on the market
and/or put into service as extended products. These standardized correlations H/H = f (Q/Q )
100 % 100 %
are called reference pressure control curves in this document and are defined in subsequent parts of
this standard.
For calculating P in the case of variable speed operation, the electric power input P shall be
1, avg 1
determined at the relative flow rate values Q/Q of the relevant reference load-flow profile and for
100 %
the corresponding rotational speed n.
NOTE 2 Especially at low values of Q/Q the electric power input P is considerably higher for pump units
100 % 1
operated at constant motor stator frequency than for speed controlled pump units. This holds true independently
of the actual pressure control curve.
4.3 The reference electric power input P
1, ref
For assessing the energy efficiency of pump units within the scope of this document, a non-dimensional
value is needed which directly reflects the energy efficiency of pump units. The numerical magnitude of
this value shall be independent of parameters which influence the unit component's efficiency such as,
for example, their geometrical size and nominal power. For this purpose, the actual electrical power
input P averaged according to Formula (1) is divided by an appropriate reference value P . This
1, avg 1, ref
reference value shall:
— have a physical basis;
— consider relevant influences on the electrical power input of pump units;
— be independent from the efficiency related quality of individual components of the pump unit.
The chosen definitions of P which fulfil these requirements are given in other parts of this standard.
1, ref
5 Qualification of pump unit types in respect to their Energy Efficiency Index
5.1 General remarks
The qualification of a pump unit type in respect to the Energy Efficiency Index (EEI) is the responsibility
of the company which places it on the market and/or puts it into service.
Product information on EEI of a pump unit type can either state the (legally) required EEI in the format
“EEI ≤ 0.XX” or can indicate a lower value “EEI ≤ 0.YY” if the pump unit type is qualified for that and the
manufacturer decides so. EEI values shall be rounded to two decimal places.
5.2 Qualification methods
Generally, the method which is applied to determine the EEI-value in the frame of the qualification
procedure is left to the company which is responsible.
In this document, two methods of qualification are described. Both methods result in a high probability
of a successful verification:
1) Tests on complete pump units and evaluation of the test results to determine the EEI-value.
2) Determination of the EEI-value by applying a semi-analytical model of the complete pump unit.
These methods are described in detail in specific parts of this standard for the respective types of pump
units.
1) In the case of qualification by tests on complete pump units the following procedure shall be
applied:
To qualify a pump unit type for conformity to requirements by law with respect to the EEI-value,
tests and evaluations shall be carried out by taking pump units randomly out of the pump unit
population. From the result of measurements and evaluations, the mean value EEI shall be
mean
determined (for example as described in Annex B). The pump unit type is then qualified for
conformity to law if the mean EEI-value determined as described above fulfils the condition
EEI < EEI < EEI (3)
mean declared allowable
where the value EEI is prescribed by law. EEI is without any tolerance for the
allowable mean
qualification procedure. In the verification procedure, according to Clause 6, a tolerance t = 7 %
EEI
is applied to the value of EEI which results from verification tests. This tolerance is not part of
test
the method to qualify a pump unit type by the supplier and shall not be applied on the result of the
qualification procedure.
2) In the case of qualification by applying the semi-analytical models as described in other parts of
Standard, the EEI-value resulting from the calculations (considering uncertainties of input values
and model errors), taken as the mean value EEI :
mean
— is used to qualify the pump unit type for conformity to law per Formula (3);
— and can be indicated as numerical value with two digits after the decimal point on the
nameplates and/or in the technical documentation.
6 Verification of the Energy Efficiency Index for pump units
6.1 General remarks
The Indicated Energy Efficiency Index can be checked and confirmed by an independent institution, e.g.
in the frame of market surveillance. This is called verification procedure in this document.
A verification procedure either needs checking the technical documentation and (re-)calculation
according to the methodology called semi analytical models, if specified in the respective parts of this
standard, or the verification procedure can be done by measuring the test pump units. These tests need
only a very limited number of test pump units and tests, but nevertheless enable with sufficient
certainty the confirmation (or rejection) of the conformity with respect to EEI or validity of the
allowable
indicated value of EEI. The procedure is based on simple pass-or-fail tests. Some explanations regarding
the probability of the results of a verification procedure are given in Annex E.
6.2 Procedure and decision
In the case of a verification procedure based on measurement the verification procedure consists of
several steps:
1) An individual pump unit is drawn at random out of the pump unit population, tested and evaluated
according to other parts of this standard and assessed with respect to the corresponding Energy
Efficiency Index (EEI) as described below.
2) If the individual pump unit has passed according to the criterion described below, the conformity
and/or the validity of the indicated numerical EEI-value of the investigated pump unit type is
confirmed.
3) If the first individual pump unit does not fulfil the criterion described below, three other individual
pump units are drawn at random, tested and evaluated according to other parts of this standard.
4) The values of EEI determined for each of the three additionally tested pump units are arithmetically
averaged by the means of Formula (4)
()EEI ⋅ ()EEI (4)
∑
avg,3 i
i=1
where (EEI) is the EEI-value determined individually for each of the three additional test pump
i
units.
5) The (EEI) of pump units is assessed with respect to the criterion described below.
avg,3
6) If the (EEI) of pump units also fails according to the criterion described below, the qualification
avg,3
of the investigated pump unit type with respect to the Energy Efficiency Index (EEI) is rejected.
7) If the (EEI) of pump units passes according to the criterion described below, the pump unit
avg,3
type passes although the first pump unit has failed.
The tests and evaluations on each of the pump units drawn at random and independently from each
other out of the pump unit type population for which the Energy Efficiency Index (EEI) shall be verified
shall be carried out according to other parts of this standard.
=
For the decision on passing or failing the verification in respect to EEI, a total tolerance of
t = 7 %
EEI
is applied to the nominal value of EEI. The nominal value for a pump unit type is:
— the maximum allowable value EEI required by law for assessing the conformity to law,
allowable
— the numerical EEI-value indicated on the nameplates and/or in the technical documentation of the
pump unit type for assessing the indicated value.
This tolerance considers the effects of manufacturing performance tolerances within a pump unit type
population (see Annex B) and the measurement uncertainties (see Annex D) of the verification test(s).
By application of the total tolerance on the nominal value, i.e. by multiplying it with the factor
1+ t
= 1.07,
EEI
the particular threshold value results serves for assessment.
Decision criterion:
The first tested pump unit or the (EEI) of the pump units passes with respect to assessing the
avg,3
Energy Efficiency Index (EEI) if the EEI-value resulting from the test(s) and evaluations is not
avg,3
higher than the corresponding particular threshold value.
The first tested pump or the (EEI) of the pump units fails with respect to assessing the Energy
avg,3
Efficiency Index (EEI) if the EEI-value resulting from the test(s) and evaluations is higher than the
corresponding particular threshold value.
Annex A
(informative)
Mean value and confidence interval of the Energy Efficiency Index EEI
In respect to the energy efficiency, the whole pump unit type is characterized by the mean value
EEI of its population.
mean
Within one pump unit type, there is an inevitable scatter of performance data and characteristics of the
components (pump, motor, frequency converter (if available) and auxiliaries (if available)). This applies
also to the hydraulic and electrical performance quantities which relevant for the Energy Efficiency
Index (EEI). This scatter of performance data and characteristics results from small differences of
relevant geometrical dimensions (such as internal flow cross sections, impeller blade angles, gap
clearances in the case of the pumps), material properties (such as the electrical properties of copper
windings in the case of the electric motors) and other influencing quantities of the components of pump
units. These differences are within the range of tolerance which are inherent in every manufacturing
process and which cannot be reduced below some economically acceptable limits.
The bandwidth of the manufacturing tolerances leads to a corresponding bandwidth of hydraulic or
electric performance characteristics of all components (pump, motor, …) of the individual pump units
within a pump unit type. The bandwidth of hydraulic or electric performance characteristics which is
caused by manufacturing tolerances is called manufacturing performance tolerance. As pump units are
composed of various components the tolerances of the components lead to the resulting manufacturing
performance tolerance of the complete unit. As long as the tolerances of the components are
statistically independent of each other, the tolerances of the various components can be assumed to be
added to each other according to probability laws (see Annex C).
The entire pump unit type is characterized by its mean performance data and curves. These mean
values are defined by Formula (A.1):
z
xx⋅ (A.1)
mean ∑ i
z
i=1
where
x is the respective performance quantity (for example Q, H, η , η , P ) and z is the total number of
PU M 1
pump units of the same pump unit type produced and placed on the market and/or put into
service.
Figure A.1 — In case z is large (usually z ≥ 30), and assuming a normal (Gaussian) distribution (see C.2)
the resulting performance quantities are characterized by their mean value x which is identical to
mean
the value defined by Formula (A.1) and the standard deviation (Formula (A.2) which is associated with
the bandwidth scatter.
=
Key
f(x) relative frequency of x
x performance quantity
x mean value of x
mean
s standard deviation of x
x
P 95 % probability interval of x
95 %
Figure A.1 — Normal distribution of a quantity x within a pump unit type
For normal distributions, the probability is 95 % that any individual true value x of the respective
i
performance quantity x is confined to a confidence interval of ± 1,96·s around the mean value x
x mean
where
z
s= ⋅ xx− (A.2)
( )
x i mean
∑
z− 1
i=1
is the standard deviation of the scatter of the performance quantity x within the same pump unit type
(see Annex C).
The manufacturing performance tolerance t of a quantity x can be defined as ± 1,96 s .
man, x x
Furthermore, the true value x of any pump within the same pump unit type population has a
i
probability of:
— 50 % to be equal to or higher than x ;
mean
— 97,5 % to be equal to or higher than x – 1,96 s ;
mean x
— 97,5 % to be equal to or lower than x + 1,96 s .
mean x
For example, a pump unit drawn at random out of a pump unit type population can have any value of
electric power input P at a given hydraulic load point Q/Q , H/H within the scatter bandwidth
1 100 % 100 %
and, as one of two possible extreme cases, either the minimum value or the maximum value of electric
power input P at the boundaries of the bandwidth.
The true mean value of a quantity x as well as the corresponding manufacturing performance tolerance
t could only be determined exactly by Formulae (A.1) and (A.2) if the values x of all z individual
man,x i
pump units out of the pump unit type population were known from tests. Because, in most cases, the
total number z of a pump unit type population is very large, it is normally not possible for economic
reasons to test each individual pump unit.
Therefore, only estimates of the mean values of the pump unit type can be determined:
— either based on tests on a sample of a number M (typically being very small compared to z) of test
pump units drawn at random out of the production (see Annex B),
— or by the application of semi-analytical models (see Part 2 and Part 3 of this standard).
It needs to be emphasized that – based on tests on a sample – the true mean values of performance
quantities x can only be determined to be confined with a certain probability (usually 95 %) within
mean
a range called confidence interval (see Annex B and Annex C).
As to all performance quantities x, these facts apply also to the EEI-value which results from tests on
only one or on a (small) sample of pump units (see Annex B).
If the estimate of EEI of a pump unit type and the corresponding confidence interval is determined
mean
by tests and evaluations on sample pump units, additionally to the manufacturing performance
tolerances the effect of measurement uncertainties shall be considered. While the manufacturing
performance tolerances are only dependent on the quality of manufacturing, the width of the
confidence interval of EEI is additionally dependent on the accuracy of the measuring methods and
mean
devices, i.e. on the actual overall measurement uncertainties (see Annex D).
If the estimate of EEI of a pump unit type and the corresponding confidence interval is determined
mean
by the application of semi-analytical models, additionally to the manufacturing performance tolerances
the effect of model errors shall be considered instead of measurement uncertainties. Information on
model errors is given together with the description of the method in in the respective parts of this
standard.
From reasons explained above, the true value of EEI of a pump unit type cannot be determined
mean, true
exactly, but can only be determined as being limited to a confidence interval. The magnitude of the
confidence interval can be quantified by a total tolerance t . This total tolerance can either be pre-
tot, EEI
defined (as, for example, for the application of the method described B.2) or can be determined directly
from measurements on a sample of several pump units of the same pump unit type population (when
applying, for example, the method described in B.3).
In Figure A.2 and Figure A.3, the effect of the confidence interval width of the value EEI of a pump
mean
unit type value EEI on the comparison with the corresponding threshold value used for verification
mean
according to Clause 6 is demonstrated qualitatively. In both cases, the estimate of EEI – resulting
mean
from either test on a sample of pump units or from application of the semi-analytical models – as well as
the allowable value EEI and the corresponding threshold value EEI are assumed to be
allowable verification
the same. The probability that the same pump unit type will succeed in the verification procedure in
case of a larger confidence interval (shown in Figure A.2) is quite lower than in the case of a smaller
confidence interval (shown in Figure A.3).
Key
EEI nominal value of allowable EEI, for example required by law
nominal
EEI particular value of EEI for the verification process
verification
t tolerance of EEI allowed within the verification procedure, see Clause 6
EEI,verification
estimate value of EEI from tests or from semi-analytical model
A
mean
B 95 % confidence interval EEI
mean
C maximum possible value of true EEI (with 97,5 % probability)
mean
Figure A.2 — Effect of the magnitude of the confidence interval of EEI , case 1
mean
Key
EEI nominal value of allowable EEI, for example required by law
nominal
EEI particular value of EEI for the verification process
verification
t tolerance of EEI allowed within the verification procedure, see Clause 6
EEI,verification
A estimate value of EEI from tests or from semi-analytical model
mean
B 95 % confidence interval EEI
mean
C maximum possible value of true EEI (with 97,5 % probability)
mean
Figure A.3 — Effect of the magnitude of the confidence interval of EEI , case 2
mean
Generally, the width of the confidence interval is the smaller:
— the smaller the performance scatter within the pump unit type population is (what requires good
manufacturing quality, small manufacturing tolerances).
In the case of sample tests the width of the confidence interval is:
— the smaller the measurement uncertainties of the tests on the sample pump units are (this requires
high accuracy of measuring equipment and small random errors, Annex D),
— the larger the number M of tested sample pump units is, see Annex B.
In case the semi-analytical models are applied, the confidence interval is smaller:
— the smaller the model errors of the applied models are (what depends on the type of the models,
see specific parts of this standard).
Two possibilities to determine – based on tests as described in the subsequent parts of this standard on
sample pump units - the confidence interval of the mean EEI-value for a pump unit type are described in
detail in Annex B.
Annex B
(informative)
Methods recommended for the qualification of a pump unit type by test
B.1 General remarks
To determine the mean value EEI of a pump unit type population by sample tests, two possibilities
mean
exist and are recommended to be applied in the frame of the qualification procedure:
1) Only one test on a single test pump unit drawn at random out of the pump unit type population is
carried out and its EEI-value is determined by measurements and evaluations as described in Part 2
and Part 3 of this standard. An estimate EEI of the mean value is calculated by making use of a
exp
manufacturing tolerance factor t which is based on the experience of the manufacturer(s) of
man,EEI
the complete pump unit or of its separate components concerning maximum manufacturing
tolerances to be expected for the respective pump unit type. But it is generally not known where
the experimentally determined value EEI of the sample pump unit is lying within the scattering
test
bandwidth caused by manufacturing tolerances. Also, the actual magnitude of the scattering
bandwidth is unknown. But the true mean value EEI can be expected to be confined to the
mean,true
interval of ± t ∙ EEI around the experimentally determined values EEI . To reach a high
man,EEI test test
probability that the true mean value fulfils the verification criterion specified in 5.2, the extreme
case should be considered that the individual EEI-value of the tested sample pump unit is at the low
end of the scattering bandwidth of individual values of EEI. If the value of ± t is chosen
man,EEI
correctly to reflect the actual scattering bandwidth caused by manufacturing performance
tolerances, the true mean value EEI will not be higher than (1 + t )·EEI with a
mean,true man,EEI test
probability of 97,5 %.
2) Tests are carried out on a sample consisting of M test pump units drawn at random out of the pump
unit type population. On the results of these tests, small (or exact) sampling theory is applied.
Based on the results for the M tested pumps, the arithmetically averaged values, the standard
deviation and the actual confidence interval of the value EEI are calculated.
mean
While the second method needs more experimental effort, it will normally lead to a smaller confidence
interval if good manufacturing quality is attained by quality management measures applied by the
manufacturer(s). This is because the magnitude of a fixed manufacturing tolerance factor used for the
first method should be set to cover even cases of poorer manufacturing quality. Therefore, the higher
effort when applying the second method is – in most cases – justified by smaller margins (in comparison
to the fixed tolerance in the first method) by which the experimentally determined estimate of the mean
value EEI should be below the corresponding threshold value which is relevant for the
mean,test
verification procedure described in Clause 6.
B.2 Determination of the mean value of EEI of a pump unit type population from a
test on only one test pump unit
The test on a single pump unit shall be performed and evaluated as described in the subsequent parts of
this standard.
Thereby, the value EEI is determined for the test pump unit.
test
Based on this value, the total tolerance interval shall be determined in order for the true mean value
EEI of the whole pump unit type population to be included with a high probability. For the
mean,true
determination of the width of this interval, the total tolerance factor t shall be applied.
tot,EEI
EEI ⋅−1 t ≤ EEI ≤ EEI ⋅+1 t (B.1)
( ) ( )
test tot,,test mean true test tot,EEI
The scatter of relevant performance quantities within the pump unit type population and the
measurement uncertainties of the test of the selected pump unit can be assumed to be normally
(Gaussian) distributed and statistically independent of each other. Therefore, the total tolerance factor
for the true mean value EEI of the pump unit type population can be calculated by the means of
mean,true
Formula (B.2):
t et+ (B.2)
tot,EEI tot,,EEI man EEI
In Formula (B.2), t is the total (relative) tolerance resulting from manufacturing tolerances
man,EEI
within the pump unit type population and e the total measurement uncertainties. The value of
tot
t shall be set on the basis of the manufacturer(s) experience, for example on the basis of data
man,EEI
available from quality management. The tolerance interval of ± t shall be sufficiently large to
man,EEI
enclose about 95 % of the true EEI-values of individual pump units of the pump type population they
belong to.
Normally, each of the main components of a pump unit (pump(s), motor(s), CDM(s) (if available),
auxiliaries) will have its own manufacturing tolerance. As long as the scattering behaviour of
performance quantities of the components caused by manufacturing shows normal (or Gaussian)
distributions and are statistically independent from each other, the resulting total manufacturing
tolerance of EEI can be calculated by the means of Formula (B.3):
2 2 2 2 2 2
t F ⋅ t + Ft⋅ + F ⋅ t +  (B.3)
man,EEI PU→→EEI man,,PU M EEI man M CDM→EEI man,CDM
In Formula (B.3) F are sensitivity factors which quantify the effect of the manufacturing
comp→EEI
performa
...