publishable result oriented report result oriented report.pdfatlete project - final publishable...

Intelligent Energy – Europe (IEE): Energy Efficiency (SAVE) -

September 2011

ATLETE Appliance Testing for Energy Label Evaluation

Publishable result-oriented

report

Author: Stefano Faberi (Editing)

In collaboration with Milena Presutto

Therese Kreitz

Atlete Project - Final Publishable Report

Acknowledgment

We would like to express our gratitude for the engagements and support which was offered by all

involved.

The project development and performance was only possible through the close cooperation of the

project partners and the help of the project officer.

Particular thanks go to the laboratories, the manufacturers, and the other stakeholders which,

through their professionalism and their proactive participation made the success of ATLETE

possible.

Atlete Project - Final Publishable Report | i

Summary

Objectives and Outcomes

The energy label tool has, since 1995, continued to provide easy-to-read information on

the characteristics of a household product placed on the European market.

Accuracy and confidence in how the European energy label is used when explaining the

characteristics of energy-using products is crucial. Information on the label is based on a

declaration made by the manufacturer about its product.

Consumers need to be able to purchase products, confident in the knowledge that the

information on the label is accurate. Meanwhile manufacturers need to be confident

that they are operating in a market where all competitors play by the rules on a level

playing field.

The ATLETE project aims to assist this process by highlighting how, through greater and

transparent market surveillance, European citizens can tap into the best European

technology when choosing labelled energy-using products.

In more detail the purpose of the ATLETE Project was to increase European-wide

implementation and control of energy labelling and eco-design implementing measures

for appliances. The developed methodology, once validated, will be applicable with very

minor adaptations for any Energy-using Products (EuP).

ATLETE is designed to demonstrate, and has actually demonstrated, that market

surveillance and testing can be done in a systematic, effective and cost-efficient way. It

helps transform the market and ensures the highest benefit for consumers,

manufacturers and the environment.

In this framework the major outcomes of the project have been:

• to set out and validate an overall procedure for the compliance monitoring of the

declared values for the EU policies;

• to carry out European-wide market surveillance on an EU policy measure by

testing 80 randomly selected refrigerating appliances;

• to review and compare the EU and international legislation related to energy

labelling of household appliances especially regarding the identification of the

effective enforcement of existing legislation through national market

surveillance. This last analysis has been furthermore compared with the ATLETE

methodology showing the aspects that must be reinforced to both improve the

commercial fairness and better support the EC sustainable objectives.

Verification procedure for the compliance monitoring of the

declared values

The methodology, tested and validated through the field work carried out by ATLETE on

refrigerating appliances, is applicable for any Energy Related Product as such or with

some minor adaptations. The overall verification assessment includes the steps shown in

the following figure:

ii | Atlete Project - Final Publishable Report

The geographical and technical scope concern the definition about the level of the

action (regional, national or EU-wide) and the types of the products to be monitored.

ATLETE has addressed its monitoring activity to the following four categories of

refrigerating appliances sold in the EU 27 Member States:

• bottom-mounted refrigerator-freezers (i.e. with freezer at the bottom) also

known as “combi” refrigerator-freezers;

• top-mounted refrigerator-freezers (i.e. with freezer on top of the appliance);

• freezers, upright and chest together;

• all the other refrigerating appliances (mainly simple refrigerator, but also side-by-

side refrigerator-freezers).

The models pertaining to these categories were selected through a “semi-random

selection” procedure, focused on best selleer models. This modality of selection was

actually considered to be more appropriate for this first pan-EU compliance verification

exercise. Best-seller products have the highest impact on the market (high sale volumes

and many variants) and usually are the products where commercial pressure is highest

(i.e. higher pressure towards inappropriate use of tolerances). This choice has led to a

division of the market where the models to be tested fell under two categories: EU top-

sellers, i.e. the best sold models of the most relevant producers at EU27 level and the

rest of the market. This ensured that other manufacturers with a market share lower

than 0,5% or operating only nationally/regionally are targeted. By applying this criteria

and basing on a reliable database of the best sold models for each manufacturer

provided by the market research firm GfK, 80 Models produced by 40 manufacturers

were finally selected1.

The laboratories for testing identified models were selected by using a mix of knock-out

criteria and a score system. Knock-out criteria eliminate immediately least qualified

laboratories without any further investigation, while the score system allows ranking the

remaining laboratories according to the expected testing capability. The scoring system

was created by giving a “weight” (score) to the answers received in the questionnaire.

From an initial list of 23 laboratories that were contacted, 15 replied to the ad-hoc

1 In reality 82 models were tested at the end of the project, because for two appliances the three

additional units purchased for developing Step 2 belonged to a different model. The list of the tested

models is given in the Atlete web site at:

http://www.atlete.eu/index.php?option=com_content&view=article&id=121&Itemid=117

Compliance assessment methodology

Geographical and technical scope

Sampling criteria

Verification procedure

Actions after non-compliance

Reporting and dissemination

Laboratory selection criteria

Follow-up actions

Harmonised Standards



Sampling criteria





Follow-up actions


Atlete Project - Final Publishable Report | iii

Questionnaire, 10 were selected having passed the knock-out criteria and achieved a

defined score (25% highest ranking). Finally a call for tender was launched to these 10

laboratories of which four where finally selected.

The verification procedure, according to the essential requirements described in the EU

legislation on labelling/ecodesign was carried out in two testing steps plus an

intermediate phase of voluntary corrective action (this last voluntary corrective action

has been introduced in this project). The figure below outlines the followed procedure:

The novelty introduced by this project has been the "Protocol on the manufacturer’s

pro-active participation in the ATLETE Project". All manufacturers of the selected models

were invited to sign such a protocol whereby they accepted to take, proactively, remedy

actions should tested products they placed on the market fail to show compliance. This

protocol was signed by 27 out of 40 manufacturers and 18 remedy actions were carried

out after the results of the first step of tests.

All the results of the laboratories tests, both complaint or not complaint ones have been

eventually disclosed to the EU national Market Surveillance Authorities of the countries

where each model was reported to be sold as well as to the stakeholder experts via the

project public workshops and the general public: through the project website, media:

interviews, press releases, articles, etc.

It is worth noting that different opinions exist as to whether making the results of a

compliance verification action publicly available, disclosing also the model and the

supplier name. At EU level all Member States that have developed market surveillance

have not disclosed the model and the supplier name while the Australian authorities

fully disclose the results of the market surveillance activity once the procedure has been

completed and the final judgement on the tested models is reached. On the other side

in the USA it has been argued that an unfair market advantage can be given to products

and suppliers meeting the requirements over the competitors that have not been

selected for the verification action. It is thus recommended that some guidance on this

matter might be provided by the European Commission or the ADCO Group on

ecodesign and labelling.

1 unit tested(Step1)

Random Products selection

Pass?

Market SurveillanceAuthorities

in the countries where the product

was availableCorrective actions?

3 additional units tested (Step 2)

YES

YES

NO

NO

Notification of compliance

YES

Pass?

NO

Notification of non-compliance & remedy action


Notification of non-compliance

iv | Atlete Project - Final Publishable Report

As outlined before, the modularity of the proposed methodological approach makes

adaptation to other appliances and products feasible with (minor) adaptations. It is

obvious that each step of the procedure has to be adapted to the type of appliance or

Energy Using Product to be tested, but the most critical ones are probably the sampling

criteria (the market structure has to be carefully analysed) and the laboratories

selection.

Final results of the field work

As already outlined, 80 models of 40 manufacturers, available within the Single

European Market, have been tested to verify their compliance with the EU energy label

in four European laboratories. For each model the tests were conducted on the

following five technical parameters2:

Parameter Description

1. Energy consumption energy consumed by a refrigerating appliance over the

period of 24 hours)

2. Storage temperature (&

climate class)

the ability of an appliance to maintain simultaneously the

required storage temperatures in the different

compartments at a certain ambient temperature

3. Storage volume the space to store food in the appliance

4. Freezing capacity the amount of food, expressed in kilograms, that can be

frozen to a core temperature of -18 °C in 24 h

5. Temperature rise

time

the time period needed to raise the temperature of food in

the frozen food compartment from -18 °C to - 9 °C after the

operation of the refrigerated system has been interrupted

The final test results show that 80% of appliances subjected to testing and for which

testing has been concluded complied with the energy efficiency class declaration and

the two related key parameters: energy consumption and storage volume. But when all

five parameters are taken into consideration 57% of them do not comply with at least

one of the tested parameters.

The analysis of the tests results are summarized in the figure below, in which the overall

picture of the pan-EU compliance of refrigerators and freezers are presented. Although

the compliance rate for the single parameters is in the range 70-90%, the overall

compliance rate is unacceptably low at 43%, due to the combined effect of the single

non-compliances.

2 To this end the harmonized standards: EN 153:2006 & EN ISO 15502:2005, concerning the testing

normative for refrigerators and freezers ruled by the 2003/66/EC Energy Labelling Directive and that Eco-

labelling Regulation, have been applied by the selected laboratories

Atlete Project - Final Publishable Report | v

More in detail, the figure below shows the results on the critical parameter on the

compliance with the declared energy class. The figure below shows that with respect to

70 models for which the steps 1 and 2 have been carried out, 55 (that is the 79%), have a

correct energy efficiency class declaration (the green labels in the figure) while 10 tested

models (14%) were overrated by 1 class and 4 tested models (6%) were overrated of 2 or

more classes. thirteen models were unfortunately classified as undecidable (the yellow

labels) as no 3 equivalent models were found on the market at the time to carry out the

second testing step.

Energy consumption is another critical parameter. The results have actually shown that,

apart from the models largely exceeding the permitted tolerances, for some models the

difference between the rated energy consumption and the average measured energy

consumption of Step 2 is much smaller that the difference found in Step 1, but also the

contrary happens. This variation on this critical parameter, measured on same models,

justifies the need to allow a tolerance range for the energy efficiency compliance.

Non compliance analysis

The wealth of data provided by the test results of the ATLETE project allow, for the first

time, to run a semi-quantitative analysis of the non-compliance causes in order to

provide some guidance to better direct the selection of potentially non-compliant

79%

77%

90%

73%

84%

70%

43%

21%

23%

10%

27%

16%

30%

57%

0% 20% 40% 60% 80% 100%

Energy class

Energy consumption

Storage temperature

Storage volume

Temperature rise time

Freezing capacity

TOTAL

PASS

FAIL

vi | Atlete Project - Final Publishable Report

models. It is nonetheless worth noting that the conclusions of this analysis should be

considered only as rough guidance, given the small size of the considered sample. Four

important market and energy categories of the models have thus been analysed (i.e.

product category, price, manufacturing origin and energy efficiency class) and the table

below shows, per each of these issues, which are the possible clues non compliance

risks.

Models categories Possible indicator of non-compliance risk

The product category The higher non-compliance risks are

concentrated to low temperature

compartment refrigerators and to freezers

The purchasing price The results have shown that the purchasing

price is a good indicator of possible non-

compliance: more expensive models have

generally a higher probability to be compliant,

while cheaper models are often more non-

compliant.

The manufacturing place Although, again, the dimension of the sample

in not sufficient to draw statistically sound

conclusions, the trend is very clear: when the

country of manufacturing is not indicated or

indicated in a very generic way such as “EU”

the probability of non-compliance is higher

than when the country is clearly stated. The

same occurs when the products manufactured

in a specified EU Member State are compared

to the products manufactured in non-EU

countries.

The expected energy efficiency class Also here a further correlation appears to exist

between the declared energy efficiency class

and the compliance rate: the higher the

energy efficiency class the higher the

probability to find a compliant model.

All in all, since a correlation appears to exist between the purchasing price and the

energy efficiency class, in the end, the highest the purchasing price and energy efficiency

class, the higher the probability of compliance.

It is important to add at this point that the verification procedure of the tested models

did not always take place without problems. During the testing campaign several

problems arose of different nature, some attributable to a lack of respecting market

rules or to how to interprete Directives .

As far as the EU Labelling Directive interpretation is concerned, we have argued if the

model purchasing via on-line shops can be accepted as a true random selection on the

market. Actually it has been witnessed that when purchasing via the internet three

Atlete Project - Final Publishable Report | vii

additional units of a model for the step 2 verification, they all belonged to the same

batch and were produced very closely in time.

In another case we have experienced the importance of a clear description for the

rounding up of measured values for the compliance verification. This is of great

importance for the models that are “border line” with the energy efficiency class

thresholds or accepted tolerance or specific minimum conditions requested by the

standard. Actually it may come about that the compliance or non-compliance status

depends to the number of decimal digits with which the annual energy consumption has

been calculated. The following example, referring to an appliance with a declared annual

energy consumption of 124 kWh/year and in which the first figure refer to the hourly

energy consumption measured in the laboratory, clarifies this issue:

(a) 0,3904658 kWh/24h×365=142,52 kWh/year, rounded to the nearest integer is 143 kWh/year or

+15,3% higher than the rated value, the model is non-compliant for the energy consumption

(b) 0,39 kWh/24h×365=142,35 kWh/year, rounded to the nearest integer is 142 kWh/year or +14,8%

higher than the rated value, the model is compliant for the energy consumption.

This situation can happen whatever is the allowed tolerance and for any parameter to

be verified if the performance of the tested model is very close to the minimum

requirement or the threshold value.

In others cases, mostly depending to the inaccuracy of the manufacturers or "strange"

declarations of the importers, we have come across label layout belonging to the same

models but with different declared values for the energy consumption and/or the

storage volume, or same commercial code numbers for technically different appliance

models or even different suppliers (importers) for the same appliance model. In this last

situation, which of the suppliers will be – from the legal point of view – responsible for

the non-compliance or for the eventual correcting action or a possible sanction? These

problems were all solved (apart that concerning the different importers for the same

models) through interactions with the manufacturers, but at the expense of waste of

resources and time.

Main Critical Elements of the Proposed Procedure

After the completion of the field work and in the light of the achieved results, a critical

analysis of the applied evaluation procedure was developed with the aim of correcting

the aspects that have presented uncertainty or have been shown to be excessively

cumbersome. Some of the critical aspects have been already outlined in this brief

outline on the project outcomes, others were highlighted by the testing laboratories,

that were asked about their views about the verification procedure developed within

the ATLETE project and how it could be improved at the end of the field work.

The analysis has been focused on the models selection, purchasing and identification

and on the procedural and technical indication concerning the testing activity.

The problems concerning the way the models should be selected, purchased and

identified stem all from the difficulties met and the experience acquired during the

project. The challenges identified deal with the possibility to carry out all the evaluation

steps without running the risk of not finding three additional models on the market, or

viii | Atlete Project - Final Publishable Report

the necessity to clarify if, or to what extent, on-line purchasing can be considered valid

for the purchasing of the selected models or, even, to ensure that units apparently

belonging to the same model are actually the same product, in the case that they are

not univocally identified by their commercial code number.

With regard to the testing activities it has been highlighted that it is necessary to

adequately size the number of the selected laboratories in order to avoid saturating

them. Paying adequate remuneration taking into account possible actions beyond the

mere model testing and results reporting needs also to be considered. It would attract

top quality laboratories to participate in selection bids. Another requirement concerns

the necessity to provide a mandatory common test format for the reports the

laboratories have to provide (according to the applied standard(s) or legislation) in order

to better allow that results comparison among the laboratories themselves. Finally, on

the basis of the tests carried out during the ATLETE project, the laboratories, in

agreement with the project team, have suggested a list of improvements for the

refrigerating appliances standards.

Market surveillance procedures at EU and international level

In the project, a detailed analysis of the international (Australia and USA) and European

legislation and practice on the verification procedures has been performed on both the

basis of the technical literature and a survey carried out with the aim of assessing

surveillance practices in the EU Member States and in 4 other non-EU countries:

Australia, Japan, Turkey and the United States. The results of this analysis are that

Australia (with New Zealand) and the USA take into due consideration the surveillance

practices and, to varying extents, have not only developed good verification protocols

(actually not very different from those prescribed in the EU legislation), but are able to

seriously put them into practice. The figure on the following page shows for example the

two-stage verification testing process used by the US/DOE to carry out a pilot action on

261 models of different types of appliances.

In the EU, in accordance with the survey results, more than half Members States (i.e. 14)

declare performing tests in practice but only 5 countries declare testing appliances each

year (Denmark, Hungary, Greece, the UK and The Netherlands) and Sweden has been

running almost regularly appliance testing over the past years. All the other have

performed these tests just once in the last 10 years or very rarely.

Atlete Project - Final Publishable Report | ix

The two-stage verification testing process for the US/DOE pilot action

Nonetheless 12 Member States do not carry out at all appliance tests of energy labels’

accuracy. The main reason for not conducting the tests is that these procedures are

considered too expensive but also the lack of adequate laboratories and the fact that

the surveillance authorities are often undersized (being their main duty the safety

controls) are other relevant barriers.

The ATLETE project, without disregarding the problems claimed by the surveyed

countries, has in turn demonstrated that, using a formal and well tuned methodology,

surveillance activities for household appliances are technically possible and cost

effective.

Final conclusions

1. The ATLETE Project has brought specific positive examples of product test results

and has demonstrated that market surveillance activities for household appliance

energy labelling are essential, technically possible and cost effective. Lack of

market surveillance in this area leads to unfair competition and misleading of

consumers in their search for highly energy efficient products. It enables “free-

riders” to gain considerable market share which is undermining the market and

does not allow to achieve demanding energy efficiency targets.

2. The ATLETE Project has also proved that two steps approach is crucial for

completing proper verification procedure. Almost 25% of models tested in Step 2

3

3

6

The controlling institution has too many subjects to deal with

Finding an appropriate laboratory or independent company to proceed to the tests is difficult

Tests would be too expensive

number of countries

Why energy classification is not verified?

x | Atlete Project - Final Publishable Report

(that is suspected non-compliance in step 1) occurred to be fully compliant after

Step 2.

3. Cooperation of testing laboratories within the testing procedure is very helpful in

achieving fully comparable test results and exchanging the experience gained

during the testing procedure. Four meetings between the testing laboratories,

project partners and representatives of IAC committee enable also to provide

added value to the evaluation and recommendation phase in WP6

Atlete Project - Final Publishable Report | xi

Table of contents

1 THE PROJECT IN BRIEF: OBJECTIVES, ACHIEVEMENTS, OUTCOMES

AND MAIN FACTS ............................................................................................................................. 1

2 VERIFICATION PROCEDURE FOR THE COMPLIANCE MONITORING OF

THE DECLARED VALUES .................................................................................................................... 4

2.1 INTRODUCTION ...................................................................................................................................... 4 2.2 GEOGRAPHICAL AND TECHNICAL SCOPE OF THE TARGET PRODUCTS ................................................................... 5 2.3 SAMPLING CRITERIA................................................................................................................................ 6 2.4 LABORATORIES SELECTION CRITERIA ........................................................................................................... 7 2.5 VERIFICATION PROCEDURE ....................................................................................................................... 8

2.5.1 The two step compliance verification procedure ...................................................................... 8 2.5.2 Protocol on the manufacturer’s pro-active participation to the ATLETE

Project ..................................................................................................................................... 10 2.6 TESTED PARAMETERS AND HARMONIZED STANDARDS .................................................................................. 11 2.7 VERIFICATION RESULTS REPORTING AND DISSEMINATION .............................................................................. 11 2.8 ADAPTATION OF THE METHODOLOGY TO OTHER ENERGY RELATED PRODUCTS .................................................. 12

3 FINAL RESULTS OF THE FIELD WORK ............................................................................................... 15

4 NON COMPLIANCE ANALYSIS .......................................................................................................... 22

4.1 IDENTIFICATION OF THE NON-COMPLIANCE CASES ....................................................................................... 22 4.1.1 Models formally non-compliant before the laboratory tests .................................................. 22 4.1.2 Models that underwent a “manufacturers’ voluntary remedy action .................................... 23 4.1.3 Models (suspected to be) damaged or defective .................................................................... 23

4.2 ANALYSIS OF THE NON-COMPLIANCE MACRO-SOURCES ................................................................................ 24 4.2.1 Product category ..................................................................................................................... 25 4.2.2 Purchasing price ...................................................................................................................... 25 4.2.3 Place of manufacturing ........................................................................................................... 27 4.2.4 Declared Energy Efficiency class ............................................................................................. 28

4.3 FURTHER SOURCES OF UNCERTAINTY IN THE COMPLIANCE VERIFICATION .......................................................... 29 4.3.1 The EU labelling Directive interpretation (1), concerning the models selection

in the market ........................................................................................................................... 29 4.3.2 The EU labelling Directive interpretation (2): rounding of measured values .......................... 30 4.3.3 Market problem (1): different labels and declarations for the same appliance

model ...................................................................................................................................... 31 4.3.4 Market problem (2): same commercial code number for technically different

appliance models .................................................................................................................... 32 4.3.5 Market problem (3): different suppliers for the same appliance model ................................. 33 4.3.6 Improvement of the test method ............................................................................................ 34

5 MAIN CRITICAL ELEMENTS OF THE PROPOSED PROCEDURE ........................................................... 35

5.1 MODELS SAMPLING, PURCHASE AND IDENTIFICATION .................................................................................. 35 5.1.1 Models sampling ..................................................................................................................... 36 5.1.2 Models purchasing .................................................................................................................. 37 5.1.3 Models univocal identification ................................................................................................ 37 5.1.4 Laboratories selection and remuneration ............................................................................... 37 5.1.5 Final test report ....................................................................................................................... 37 5.1.6 Measured parameters in Step 1 and Step 2 of the compliance verification ........................... 38 5.1.7 Improvement of the test method ............................................................................................ 38

6 MARKET SURVEILLANCE PROCEDURES AT EU AND INTERNATIONAL

LEVEL AND COMPARISON WITH THE ATLETE METHODOLOGY. ....................................................... 39

6.1 MARKET SURVEILLANCE PROCEDURES FOLLOWED AT INTERNATIONAL LEVEL ..................................................... 40 6.1.1 Australia and New Zealand ..................................................................................................... 40 6.1.2 USA .......................................................................................................................................... 43

xii | Atlete Project - Final Publishable Report

6.2 BRIEF OVERVIEW ON THE FORMAL PROCEDURE FOLLOWED IN EU MEMBER STATES ........................................... 47 6.3 RESULTS OF TESTING ON REFRIGERATING APPLIANCES DEVELOPED IN RECENT YEARS AT MEMBER STATES LEVEL ........ 49

6.3.1 Sweden .................................................................................................................................... 49 6.3.2 Denmark .................................................................................................................................. 49 6.3.3 UK ............................................................................................................................................ 50

6.4 COMPARISON WITH THE EU AND THE ATLETE APPROACH ........................................................................... 53 6.4.1 Differences and common aspect in the applied verification procedures ................................. 53 6.4.2 Comparison of the pass/fail rates for refrigerating appliances .............................................. 54

6.5 THE MANUFACTURERS’ CONFORMITY ASSESSMENT IN THE EU MEMBER STATES ............................................... 55 6.5.1 Countries performing or not performing the tests in practice ................................................ 55 6.5.2 The laboratories issue ............................................................................................................. 57 6.5.3 The verification Tests .............................................................................................................. 57 6.5.4 Non-compliance and sanctions ............................................................................................... 58 6.5.5 Sharing the results and information ........................................................................................ 58 6.5.6 Perspectives on Ecodesign ....................................................................................................... 59

7 FINAL POLICY CONCLUSIONS AND RECOMMENDATIONS ................................................................ 60

Figures

FIGURE 1: METHODOLOGICAL STEPS FOR COMPLIANCE ASSESSMENT ............................................................................ 4

FIGURE 2: ENERGY LABEL FOR REFRIGERATORS AND FREEZERS ACCORDING TO THE DIRECTIVES

94/2/EC AND 2003/66/EC ................................................................................................................ 5

FIGURE 3: EU LABORATORIES CONTACTED FOR THE ATLETE PROJECT .......................................................................... 8

FIGURE 4: VERIFICATION PROCEDURE SCHEME AS DEVELOPED WITHIN ATLETE .............................................................. 9

FIGURE 5: OVERALL COMPLIANCE RESULTS OF THE ATLETE PROJECT ......................................................................... 19

FIGURE 6 EFFICIENCY CLASSES DECLARED / VERIFIED ............................................................................................... 20

FIGURE 7 DIFFERENCE (%) BETWEEN THE DECLARED AND THE MEASURED VALUE OF THE ENERGY

CONSUMPTION IN STEP 1 AND STEP 2 FOR THE 82 MODELS OF REFRIGERATING

APPLIANCES ...................................................................................................................................... 21

FIGURE 8 DIFFERENCE (%) BETWEEN THE DECLARED AND THE MEASURED VALUE OF THE ENERGY

CONSUMPTION IN STEP 1 AND STEP 2 (SINGLE UNITS AND AVERAGE VALUES) ................................................ 21

FIGURE 9 DIFFERENCE (%) BETWEEN THE DECLARED AND THE MEASURED VALUE OF THE STORAGE

VOLUME FOR THE REFRIGERATOR AND FREEZER COMPARTMENTS FOR THE 82 MODELS OF

REFRIGERATING APPLIANCES ................................................................................................................. 21

FIGURE 10: DEFECTIVE GASKET OF THE FREEZER COMPARTMENT DOOR ......................................................................... 24

FIGURE 11: PURCHASING PRICE FOR THE TESTED MODELS OF REFRIGERATORS AND FREEZERS ............................................. 26

FIGURE 12: PURCHASING PRICE FOR THE TESTED MODELS OF REFRIGERATOR-FREEZERS AND

REFRIGERATORS WITH A LOW TEMPERATURE COMPARTMENT ..................................................................... 27

FIGURE 13: PURCHASING PRICE FOR THE TESTED MODELS OF UPRIGHT AND CHEST FREEZERS ............................................. 27

FIGURE 14: PLACE OF MANUFACTURING AND COMPLIANCE WITH THE EU ENERGY LABELLING ............................................ 28

FIGURE 15: FIRST UNIT OF A MODEL PURCHASED FOR STEP 1 OF THE VERIFICATION ......................................................... 29

FIGURE 16 : ADDITIONAL THREE UNITS OF THE MODEL PURCHASED FOR STEP 2 OF THE VERIFICATION ................................. 29

FIGURE 17: LABEL STRIPS FOR THE 4 UNITS OF THE SAME MODELS USED IN STEP1 AND STEP2 OF THE

VERIFICATION PROCEDURE ................................................................................................................... 31

FIGURE 18: LABEL STRIPS FOR THE 4 UNITS OF THE SAME MODELS USED IN STEP1 AND STEP2 OF THE

VERIFICATION PROCEDURE ................................................................................................................... 32

FIGURE 19: RATING PLATE AND LABEL STRIP FOR THE FIRST UNIT OF THE REFRIGERATOR-FREEZER ....................................... 33

FIGURE 20: RATING PLATE AND LABEL STRIP FOR THE SECOND UNIT OF THE REFRIGERATOR-FREEZER ................................... 33

FIGURE 21: DIFFERENT SUPPLIERS FOR DIFFERENT UNITS OF THE SAME MODEL PURCHASED IN THE

SAME COUNTRY ................................................................................................................................. 34

FIGURE 22: TOTAL NUMBER OF CHECK TESTS CONDUCTED IN AUSTRALIA BY APPLIANCE CATEGORY,

1991-2010 ..................................................................................................................................... 42

FIGURE 23: THE TWO-STAGE VERIFICATION TESTING PROCESS FOR THE DOE PILOT ACTION ............................................... 46

Atlete Project - Final Publishable Report | xiii

Tables

TABLE 1: ELEMENTS TO BE MODIFIED TO TRANSFER THE METHODOLOGY FROM REFRIGERATING

APPLIANCES TO THE OTHER APPLIANCES .................................................................................................. 13 TABLE 2: OVERALL RESULTS OF THE LABORATORY TESTS ......................................................................................... 18 TABLE 3: SUMMARY OF THE RESULTS OF THE COMPLIANCE VERIFICATION OF THE DIFFERENT TYPES

OF REFRIGERATING APPLIANCES ............................................................................................................ 25 TABLE 4: AVERAGE, MIN AND MAX PURCHASING PRICE (€) FOR COMPLIANT AND NON-COMPLIANT

FOR THE APPLIANCE CATEGORIES DEFINED IN THE LABELLING SCHEME .......................................................... 26 TABLE 5: PLACE OF MANUFACTURING AS AND RATE OF COMPLIANCE WITH THE EU ENERGY

LABELLING ........................................................................................................................................ 27 TABLE 6: RELATION BETWEEN THE DECLARED ENERGY EFFICIENCY CLASS AND THE COMPLIANCE

RATE FOR REFRIGERATING APPLIANCES WITHIN THE ATLETE PROJECT .......................................................... 28 TABLE 7: FINAL NUMBER OF SELECTED MODELS FOR THE DOE VERIFICATION PILOT PROGRAMME .................................. 45 TABLE 8: PRELIMINARY RESULTS FOR DOE THE VERIFICATION ACTION ....................................................................... 47 TABLE 9: SUMMARY OF THE EU VERIFICATION SYSTEM AND TOLERANCES FOR ENERGY

CONSUMPTION IN THE PREVIOUS LABELLING AND MINIMUM REQUIREMENTS SCHEMES .................................... 48 TABLE 10: SUMMARY OF THE EU VERIFICATION SYSTEM AND TOLERANCES IN THE NEW LABELLING

AND ECODESIGN SCHEMES ................................................................................................................... 48 TABLE 11: RESULTS OF THE 2007 TESTS ON HOUSEHOLD APPLIANCES IN DENMARK ...................................................... 50 TABLE 12: ANALYSIS OF THE RESULTS OF THE 2005 ENERGY LABEL COMPLIANCE FOR

REFRIGERATING APPLIANCES IN UK ........................................................................................................ 52

xiv | Atlete Project - Final Publishable Report

Atlete Project - Final Publishable Report | 1

1 THE PROJECT IN BRIEF: OBJECTIVES, ACHIEVEMENTS, OUTCOMES AND

MAIN FACTS

Objectives

The purpose of the ATLETE Project was to increase European-wide implementation and

control of energy labelling and eco-design implementing measures for appliances. The

developed methodology, once validated, will be applicable with very minor adaptations

for any Energy-using Products (EuP).

Energy labels are a crucial driver for market transformation, orienting consumers'

choices towards more energy efficient appliances and thus realising the potential of

available technologies. Unfortunately, not all Member States (MS) apply a responsible

policy for controlling the correct labelling implementation. This needs to change.

ATLETE is designed to demonstrate, and has actually demonstrated, that market

surveillance and testing can be done in a systematic, effective and cost-efficient way,

thus helping to transform the market to ensure the highest benefit for consumers,

manufacturers and the environment.

Strong points of the ATLETE Project:

• addresses the issue of compliance testing relating to Energy labelling and Eco-

design requirements;

• ensures that the national authorities are informed of cases of non-compliance;

• identifies effective enforcement of existing legislation through national market

surveillance;

• provides the first pan-European testing results on a large number of refrigeration

appliances;

• gives concrete guidance to EU and National Authorities for effective labelling and

future eco-design requirements implementation;

• sets a shared procedure for the verification of the manufacturers' labelling/eco-

design declarations including a methodology for laboratories accreditation and

models selection;

• raises awareness among National Authorities of the effectiveness of the energy

labelling on national energy efficiency;

• engages key stakeholders that have an interest in the issue of compliance with

energy label requirements: the EU institutions, Government organisations,

manufacturers, retailers, associations, consumer groups, NGOs, the media and

general public.

2 | Atlete Project - Final Publishable Report

Achievements and Outcomes

ATLETE has largely achieved these objectives providing the following three important

results:

• setting out and validation of an overall procedure for the compliance monitoring

of the declared values for the EU policies;

• carrying out a European-wide market surveillance on a EU policy measure by

testing 80 randomly selected refrigerators;

• reviewing and comparing the EU and international legislation related to energy

labelling of household appliances especially for what concerns the identification

of the effective enforcement of existing legislation through national market

surveillance. This last analysis has been furthermore compared with the ATLETE

methodology showing the aspects that must be reinforced to both improve the

commercial fairness and better support the EC sustainable objectives.

The first of these results was made possible by the performance of the laboratory

testing campaign carried out by the project. In fact, the experience gained from this field

phase has allowed the project team to verify and amend the theoretical verification

procedure developed during the first phase of the project and to deliver the first issue of

the "Guidelines for Verification of EU Energy Related Products"3

In turn, the field work, beyond to provide meaningful hints for the laying out of the

Guidelines, has provided a twofold outcome:

• on one side it has demonstrated that that market surveillance is technically and

economically feasible, provided that adequate (but still at an acceptable level)

human and financial resources are made available,

• on the other side it allowed to evaluate the overall compliance of the selected

refrigerating appliances with the EU labelling scheme shedding light on a reality

until now little known by revealing a high non-compliance percentage with the

energy labelling Directive requirements (especially for what concerns the formal

requirements).

Dissemination activities

To ensure transparency, the final results of the project are made publically available via

the project website www.atlete.eu. Information made available includes the models

tested, the laboratories involved, final overall results and results by model and

parameter.

Moreover abundant dissemination material has been produced and distributed, several

seminars have been directly organized and in many other the ATLETE project has been

presented:

o Leaflets: more than 5,000 per country)

o Press Releases: two press releases published in all languages)

o Seminars: 19 events in total, 11 at international level

3 See The Guidelines for EU verification of Erp at

"http://www.atlete.eu/index.php?option=com_docman&Itemid=111"


It is finally worth noting that about 93 articles have been monitored in total, including

some high level national and European daily or electronic press/media. All these articles

were published after the ATLETE EU final conference (12 April 2011 Brussels) and the

national events held in the Czech Republic, France and Italy.

During the period April – September 2011, that starting from the ATLETE final

conference and in parallel with the issuing of the above mentioned articles, the website

was visited by over 8700 visitors, which is a significant success, since the website was

not intended to the final consumers / general public, but to authorities, stakeholders,

experts and journalists.

Main Facts

• 5 Partners:

� ISIS (Institute of Studies for the Integration of Systems), Project

Coordinator

� CECED (European Committee of Domestic Equipment

Manufacturers), Responsible for the Laboratories Selection and

management of the Appliances Testing

� ENEA (Italian National Agency for new Technology, Energy and the

Environment), Responsible for the overall project methodology

and final results evaluation

� SEVEn (Energy Efficiency Center), Responsible for the

dissemination activities

� ADEME (French Environment and Energy Management Agency),

Responsible for the initial Background Analysis

• Target Groups

o Institutions, & Government organisations;

o Manufacturers & Retailers;

o Associations, Consumer Groups and NGOs;

o The media and general public

• Key Actors

o National Energy Agencies;

o Manufactures & Importers

• Duration 27 months4 : June 2009 – July 2011

• Budget:

o Total Budget: 1.012.057€

o Total EC financial contribution: 759.043€

4 The initial duration of the project was of 24 months.


2 VERIFICATION PROCEDURE FOR THE COMPLIANCE MONITORING OF THE

DECLARED VALUES

As outlined in the previous chapter, one of the main result of this project was the setting

of an overall procedure for the compliance monitoring of the declared values for the EU

policies and the delivering of a comprehensive "Guidelines for EU Verification of Energy

Related Products" having the objective to provide a supporting tool valid at EU and

Member States level for the Authorities dealing with compliance and verification issues,

helping to optimize the available resources and minimize the inconvenience for

Authorities and industry while enforcing legislation concerning Energy Labelling and Eco-

design requirements

The developed approach, tested and validated through the field work carried out by

ATLETE on refrigerating appliances, is applicable for any Energy Related Products as such

or with some minor adaptations. This chapter summarises the main, relevant, steps of

the verification procedure applied in this project and finally shows which are the points

that have to be modified to transfer the methodology from refrigerating appliances to

the other appliances. To analyse more in depth this procedure please see the quoted

Guidelines that provides a detailed information on all steps that should be followed and

the documents to be prepared to carry out an accurate and effective compliance

verification.

2.1 Introduction

The overall verification assessment includes the steps shown in Figure 1. These steps are

described in details in the following paragraphs. Due to the time-frame of the

presentation of the project proposal and its approval by the IEE Programme, the

reference for the compliance verification was the energy label (Figure 2) and labelling

scheme established by Directives 94/2/EC and 2033/66/EC.

Figure 1: Methodological steps for compliance assessment



Sampling criteria





Follow-up actions




Sampling criteria





Follow-up actions



Figure 2: Energy label for refrigerators and freezers according to

the directives 94/2/EC and 2003/66/EC

2.2 Geographical and technical scope of the target products

A compliance verification action can be run at local, national, EU or even international

level and can involve a single specific product or different products for which one or

more (common) aspects need to be verified. A decision about the level of the action and

the number of involved products is the pre-requisite for starting any actual procedure.

The geographical and technical scope of the ATLETE project was:

• Geographical scope:

− EU27 Member States as far as possible (GfK provided data for EU 24). If the

external market statistical sources have a more limited scope, it should be

evaluated that this limitation does not exclude specific markets

− National relevant supplier in the main manufacturing EU countries: BE, DE, DK,

ES, FR, IT, NL, PL, UK.

− Other suppliers randomly selected.

• Technical scope: refrigerating appliances (refrigerators and

freezers), divided into the following four categories5:

1. Bottom-mounted refrigerator-freezers (i.e. with freezer at the

bottom) also known as “combi” refrigerator-freezers

5 It is worth noting that the above four categories are different from the categories defined in the

previous energy labelling/minimum requirement directives and from the 10 categories defined in the

previous and new energy labelling scheme and Ecodesign Regulation, but they respond to the market-

related categories for which information are available from the major market research firm operating in

the EU.


2. Top-mounted refrigerator-freezers (i.e. with freezer on top of the

appliance);

3. Freezers, upright and chest together;

4. all the other refrigerating appliances (mainly simple

refrigerator, but also side-by-side refrigerator-freezers).

2.3 Sampling Criteria

A “semi-random selection” procedure, focused on the best sold models was considered

the most appropriate for the EU.

In fact, although national Market Surveillance Authorities have usually rather limited

resources and would prefer to maximize the effectiveness of their compliance actions by

using a “maximum failure” selection procedure, the aim of the ATLETE project was to

show how much the energy labelling scheme is trustful and to give a full picture of EU

market. A high rate of non-compliance following a “maximum failure selection” would

not have given a realistic indication of market compliance.

Another alternative for models selection could have been to go for the models with a

high energy efficiency (A++ or even better), but these models are usually under the spot

light also from competitors and although having a high visibility represent limited sale

volumes on the EU market. Again the semi-random selection was felt more appropriate

for the first pan-EU compliance verification exercise. Best-seller products have the

highest impact on the market (high sale volumes and many variants) and usually are the

products where commercial pressure is higher (i.e. higher pressure towards

inappropriate use of the tolerance).

In practice, to guarantee that the tested products cover all manufacturers and brands

operating within the Community market of refrigerating appliances, half of the models

were selected among the “EU top-sellers” according to the market share of the relevant

manufacturers/importers, i.e. the best sold models of the most relevant producers at

EU27 level; in this way bestseller models of all major brands in Europe were targeted

and tested. The second half of the models was selected randomly within the remaining

producers active on the EU27 market, thus ensuring that other manufacturers with a

market share lower than 0,5% or operating only nationally/regionally are targeted. The

selection was based on the Market Share of each supplier (including all the owned

brands) at European level according to the data provided by the market research firm

for January 2010:


• if market share (MS) is MS >10% or 5%<MS<10% or 1%<MS<5%, 15 top best

seller models in each of the 4 product categories are considered

• if market share is 0,5%<MS<1% and 0,1%<MS<0,5% and for major national

manufacturers, 10 top best seller models are considered without product

category distinction.

Out of the list of the models identified for each market share range, a random selection

was performed, to choose the 80 models to be tested.

The updated information about the manufacturers’ market share at EU and at national

level and a reliable database of the best sold models for each manufacturer was

purchased - after a public Call for Tender open to all market data providers in Europe -

from the well-known market research firm GfK.

In reality 82 models were tested at the end of the project, because for two appliances

the three additional units purchased for developing Step 2 belonged to a different

model6.

2.4 Laboratories selection criteria

Reliable and transparent selection criteria were used, including a mix of knock-out

criteria and a score system. Knock-out criteria eliminate immediately least qualified

laboratories without any further investigation, while the score system allows to rank the

remaining laboratories according to the expected testing capability.

A Questionnaire was used, including questions about the lab experience, testing

capability, whether accredited for testing appliances, to EN 17025, available

instruments, etc. For laboratories passing the knock-out criteria the final score was

calculated, allowing for a ranking against a rating scheme. The scoring system was

created by giving a “weight” (score) to the answers to the Questionnaire. From an initial

list of 23 laboratories that were contacted (Figure 3), 15 replied to the ad-hoc

Questionnaire, 10 were selected having passed the knock-out criteria and achieved a

defined score (25% highest ranking).

6 The list of the tested models is given in the Atlete web site at:

http://www.atlete.eu/index.php?option=com_content&view=article&id=121&Itemid=117 .


Figure 3: EU laboratories contacted for the ATLETE project

The final selection of the 4 laboratories iPi (Germany), LCOE (Spain), RegenT (The

Netherlands) and VDE (Germany) was done on the basis of the offers received to a Call

for Tender and the a final inspection of project expert and an interview with the

laboratories representatives.

2.5 Verification procedure

2.5.1 The two step compliance verification procedure

The verification procedure is outlined in Figure 4. According to the essential

requirements described in the EU legislation on labelling/ecodesign it foresees two

testing steps plus an intermediate phase of voluntary corrective action (this last

voluntary corrective action has been introduced in ATLETE Project only).


Figure 4: Verification procedure scheme as developed within ATLETE

In particular:

• Sample gathering: test laboratories searched for and purchased the needed units of

each selected model

• Step 1: testing of 1 unit per selected model

− if all labelling declarations comply with the outcome of the tests, the model is

considered as compliant

− if the unit fails even only one of the labelling declarations (or is not able to

accomplish the test for the specific product), Step 2 is triggered

− the relevant producer is informed about “suspected non-compliance” and asked

for checks of possible declaration mistakes. The producer could choose to accept

a non-compliance (and to correct the labelling declaration accordingly) or to

proceed to Step 2.

• Step 2: three other units of the same model are purchased from the market and

tested again (against a reduced tolerance for energy consumption)

• Possibility to assist to the test: project partners, IAC members or a representative of

the manufacturer of the model under test (under supervision of the test laboratory)

had a possibility to assist the test of given model. In case of dissent with the testing

conditions followed in the laboratory, the project partner/IAC

member/manufacturer reported back to the project leader (later to the market

surveillance Authority) before the result of the Step is known.

• Test reports: the testing laboratories reported the test results to the project leader

with a copy of the test report for Step 1 (and Step 2), and considered such results

confidential.

• Actions for non-compliant models: were foreseen for both Step 1 and Step 2:

‒ after Step 1: notification to national Market Surveillance Authorities, together

with voluntary actions taken i.e. for manufacturer accepting the non-compliance,

no legal actions after voluntary correction of the labelling declaration(s) were

taken

1 unit tested(Step1)

Random Products selection

Pass?

Market SurveillanceAuthorities

in the countries where the product

was availableCorrective actions?

3 additional units tested (Step 2)

YES

YES

NO

NO


YES

Pass?

NO

Notification of non-compliance & remedy action


Notification of non-compliance


‒ after Step 2: notification to national Market Surveillance Authorities for legal

actions after the end of all tests: non-compliance was disclosed.

2.5.2 Protocol on the manufacturer’s pro-active participation to the ATLETE Project

All manufacturers of the selected models were invited by CECED to sign a protocol on

the manufacturer’s pro-active participation to the project where they accepted to take

proactively remedy actions would the tested products fail to show compliance. The

protocol stated that, in case the energy labelling/eco-design declarations of one of

appliance models were found to exceed the permitted verification tolerance of the

relevant testing standard after the first or second Step of the verification procedure, and

without prejudice to the ability to challenge the factual findings made in the testing

procedure, before its result is made available, the manufacturer had the possibility to:

• take within 30 (calendar) days after being informed of the results of the testing

procedure all the remedy actions necessary to correct the energy labelling/eco-

design declarations of the appliance model/s concerned in accordance with the

results of the testing procedure; and

• inform within the same timeframe the ATLETE project leader of the remedy

actions taken to correct the energy labelling/eco-design declarations, and

provide the ATLETE project leader with the appropriate evidence of these

remedy actions, such as a copy of a letter sent to the trade, with a proof of the

sending thereof, and a copy of brochures or leaflets marketing or advertising the

concerned appliance model/s with the correct energy labelling/eco-design

declarations, without disclosing any information which would be confidential.

In particular the protocol included:

• Manufacturer “obligations”:

o acceptance ex-ante of the reliability of the testing laboratory

o consideration of Step 1 results for possibly setting immediate remedy

actions

o set remedy actions for non-compliant models.

• Manufacturer “positive feedback”:

o possibility of assisting the test of own models and to raise concerns on

the regularity of the test before the result is known

o possibility to have a more lenient (or even no) sanction for non-compliant

declarations.

In some cases the voluntary remedy actions notified by the manufacturers to the ATLETE

project leader were verified through additional laboratory tests during the project field

work.


2.6 Tested parameters and harmonized standards

The tests were conducted on the following five parameters:

Parameter Description

6. Energy consumption energy consumed by a refrigerating appliance over the

period of 24 hours)

7. Storage temperature (&

climate class)

the ability of an appliance to maintain simultaneously the

required storage temperatures in the different

compartments at a certain ambient temperature

8. Storage volume the space to store food in the appliance

9. Freezing capacity the amount of food, expressed in kilograms, that can be

frozen to a core temperature of -18 °C in 24 h

10. Temperature rise

time

the time period needed to raise the temperature of food in

the frozen food compartment from -18 °C to - 9 °C after the

operation of the refrigerated system has been interrupted

To this end the harmonized standards: EN 153:2006 & EN ISO 15502:2005, concerning

the testing normative for refrigerators and freezers ruled by the 2003/66/EC Energy

Labelling Directive and that Eco-labelling Regulation, have been applied by the selected

laboratories.

2.7 Verification results reporting and dissemination

Different opinions do exists about the opportunity to make the results of a compliance

verification action publicly available, disclosing also the model and the supplier name.

A compliance verification action is needed to prevent manufacturers who break the law

from gaining a competitive advantage over those that adhere to it. In this respect the

full disclosing of the verification results, once the procedure has been completed and

the final judgement on the tested models is reached, can only made public aware of

those products and suppliers that do not follow the rules. This is the approach followed

since many years at international level by the Australia’s market surveillance authority.

On the other side the USA has different experience. An unfair market advantage can

derive to products and suppliers meeting the requirements over the competitors not

having been selected for the verification action.

At EU level all Member States having developed market surveillance have not disclosed

the model and the supplier name. Although this was in part due to the fact that in some

cases only the first step of the verification procedure was run and therefore no

conclusion could be drawn on the actual compliance of the tested models.

Although each national Market Surveillance Authority may decide a specific position on

this issue, due to the subsidiarity nature of the controls within each Member State. It is

nevertheless recommended that some guidelines are given by the European

Commission or the ADCO Group on ecodesign and labelling.


For what concerns the ATLETE project all compliant and non-compliant models have

been disclosed to:

• relevant manufacturers and project partners

• EU national Market Surveillance Authorities of the countries where each model

was reported to be sold

• stakeholder experts via the Final Conference, held on 12th

April 2011 in Brussels

and the national events (such as the national workshop held in Rome on 21st

June)

• general public: through media: interviews, press releases, articles, workshops,

etc.

Most important media dissemination actions have been uploaded to the project website

www.atlete.eu

2.8 Adaptation of the methodology to other Energy Related Products

The modularity of the proposed methodological approach makes the adaptation to

other appliances and products feasible with (minor) adaptations.

In Table 1 the elements to be modified in order to transfer the methodology from

refrigerating appliances to the other energy related products covered by a

labelling/ecodesign legislation are qualitative described.


Table 1: Elements to be modified to transfer the methodology from refrigerating appliances to the other appliances

Products Geographical

scope

Technical

scope

Sampling

criteria

Laboratory

selection

criteria

Verification

procedure

Actions

after non-

compliance

Harmonised

standard Reporting

REFRIGERATING APPLIANCES OK OK OK OK OK OK OK OK

WASHING MACHINES if necessary,

to be adapted

to be

adapted to

the specific

product

if

necessary,

to be

adapted

procedure

valid,

selection

criteria to

be adapted

to the

specific

product

OK OK

specific

standard to be

used for

testing

to be adapted to the specific

product

DISHWASHERS if necessary,

to be adapted

to be

adapted to

the specific

product

if

necessary,

to be

adapted

procedure

valid,

selection

criteria to

be adapted

to the

specific

product

OK OK

specific

standard to be

used for

testing


product

SIMILAR HOUSEHOLD

APPLIANCES

if necessary,

to be adapted

to be

adapted to

the specific

product

if

necessary,

to be

adapted

procedure

valid,

selection

criteria to

be adapted

to the

specific

product

OK OK

specific

standard to be

used for

testing


product


Products Geographical

scope

Technical

scope

Sampling

criteria

Laboratory

selection

criteria

Verification

procedure

Actions

after non-

compliance

Harmonised

standard Reporting

TV SETS if necessary,

to be adapted

to be

adapted to

the specific

product

if

necessary,

to be

adapted

procedure

valid,

selection

criteria to

be adapted

to the

specific

product

OK OK

specific

standard to be

used for

testing


product

HCV HOUSEHOLD

APPLIANCES

if necessary,

to be adapted

to be

adapted to

the specific

product

probably to

to be

adapted

procedure

valid,

selection

criteria to

be adapted

to the

specific

product

OK OK

specific

standard or

calculation

methodology

to be used for

testing

major changes to be

adapted to the specific

product

OTHER PRODUCTS if necessary,

to be adapted

to be

adapted to

the specific

product

probably to

to be

adapted

Probably

both

procedure

and

selection

criteria to

be adapted

OK OK

specific

standard or

calculation

methodology

to be used for

testing

major changes to be

adapted to the specific

product


3 FINAL RESULTS OF THE FIELD WORK

As already outlined, 80 models of 40 manufacturers, present on the European Union

market, have been tested to verify their compliance with the EU energy label in four

European laboratories. For each model the tests were conducted on the five parameters

described in paragraph 2.6, two of which (freezing capacity and temperature raise time)

have rarely been examined by when the energy label has been introduced in 1995.

The final test results show that 79% of appliances subjected to testing and for which

testing has been concluded complied with the energy efficiency class declaration and

the two related key parameters: energy consumption and storage volume (77% and 90%

respectively). But when all five parameters are taken into consideration 57% of them do

not comply with at least one of the tested parameters

In accordance with the test outcome, the selected models have been classified in the

following 8 categories

Models classification by

tests outcome

Description

Models non-compliant for

formal reasons

Products are non-compliant for “formal” reasons when the Label and/or

the product fiche are lacking or not fulfilling the declared energy

efficiency class when it is re-calculated using the declared values against

the formula provided in the labelling scheme. The ATLETE team decided

to consider a model as non-compliant if one or both of the above

informative fiches were missing or were not correct, whatever are the

final results of the tests

Models found (or

suspected) to be damaged

or defective

During both Step 1 and Step 2 a unit of a model could result damaged or

defective either before a test is run (at a visual inspection or suspected)

to be damaged or defective as a consequence of a deviating behaviour or

an unexpected test result. Although this event was thought to be very

uncommon, we experienced it during the ATLETE project. Depending on

the time when the defect or damage is found or suspected the reaction

could be the repetition of Step 1 (or Step 2) on a different unit(s) of the

same model or to ignore Step1 results and go directly for the Step 2

(should a sufficient number of additional units of the same model be

available on the market).

Models compliant after

Step 1

If the products are considered fully compliant with the EU energy label no

further action is needed. The models are immediately notified to the

national Market Surveillance Authorities of the Member States where

they were reported to be sold.

Models suspected of

non-compliance after

Step 1

If a product, having passed the initial check for the formal compliance,

has been found either not fulfilling the conditions established by the

standard or exceeds the permitted tolerances for this verification Step, it

is suspected to be not compliant and further action is needed to confirm

or negate this suspect. In this case the supplier of each model is

contacted and invited to:

a. accept the suspected non-conformity and go for a voluntary

remedy action

b. do not accept the suspected non conformity and ask for the

development of the second phase (Step 2) of the verification. In


Models classification by

tests outcome

Description

this case the laboratory was asked to purchase three additional

units of the models and to test them according to the Step 2 of

the verification procedure

c. comment the test report asking to the laboratory to modify the

test conditions (considered for example not in line with the

standard). In some few cases the ATLETE team considered the

remarks legitimate and the test was repeated.

Models suspected of


Step 1 that have

underwent

“manufacturers’

voluntary remedy action”

(point "a" above)

Such action was in the form of the correction of the wrong declaration or

of the instruction manual, or even to go as far as stopping the production

of that specific model. After the remedy action in general no further

action is needed, but in some selected cases the remedy action has been

further checked through an additional laboratory test. In any case, the

products were considered non-compliant with the EU energy label and

were immediately notified to the national Market Surveillance Authorities

of the Member States where they were reported to be sold, along with

the description of the remedy agreed by the supplier and the evidence of

its actual application

Models suspected of


Step 1 but for which 3

additional units are no

more available on the

market

These models are considered “non-decidable” because, despite all the

attempts done by the relevant laboratory the 3 additional units essential

for the development of Step 2 could not be found on the EU market.

Therefore no final judgement about the compliance of the models could

be reached. These products were in any case notified to the national

Market Surveillance Authorities of the Member States where they were

reported to be sold

Models non-compliant

after the complete

verification procedure

(Step 1+ Step 2)

For the models having failed Step 1 and (i) for which the suppler has

asked to go for Step 2 or (ii) if they belong to manufacturers that did not

sign the protocol on the manufacturer’s pro-active participation to the

ATLETE project, the verification Step 2 was developed. Models for which

at least one of the 3 additional tested units does not fulfil all the

conditions established by the standard or if even one of the tested

parameter exceeds the permitted tolerances for this verification Step are

considered non-compliant with the EU energy labelling.

Models fully compliant

after the complete

verification procedure

(Step 1+ Step2)

Models for which all measure parameters one the three units fulfilling the

conditions established by the standard and do not exceed the permitted

tolerances for this verification Step are considered fully compliant with

the EU energy label and no further action is needed. The models are

immediately notified to the national Market Surveillance Authorities of

the Member States where they were reported to be sold.

The number of models covered in each of the 8 categories above outlined is shown in


Table 2:


Table 2: Overall results of the laboratory tests

Number of models tested models 827

Models formally non-compliant* 10

Models compliant after Step 1 26

• of which re-tested on a different unit, due to a defect or

suspected damage 1

• of which re-tested due to test conditions non fulfilling the

standard 3

Model for which Step 1 was not developed 2

Models suspected to be non-compliant after Step 1, Step 2 needed 54

• of which voluntary remedy actions was agreed before running

Step 2 18

→ of which the voluntary remedy action was tested in

laboratory 2

• of which Step 2 testing was planned 36

→ of which Step 2 was run 22

� of which only Step 2 was run 2

→ of which Step 2 was not run 14

� of which Step 2 not run, 3 additional units not available 10

� of which Step 2 not run, 3 units of another model

received 2

� of which Step 2 not run, lack of reference values on

label 2

Models compliant after Step 2 4

Models non-compliant after Step 2 18

The analysis of the tests results are summarized in , in which the overall picture of the

pan-EU compliance of refrigerators and freezers is presented. Although the compliance

rate for the single parameters is in the range 70-90%, the overall compliance rate is

unacceptably low at 43%, due to the combined effect of the single non-compliances.

The presented results have been calculated on the 70 models for which the verification

was completed, while the single compliance rates have been calculated on the number

of models for which the specific parameter applies, because temperature rise time and

freezing capacity are relevant only for freezers and refrigerator-freezers. When these

two parameters not directly linked with the energy consumption (temperature rise time

and freezing capacity) are excluded from the statistic, the overall compliance rate raises

to 50%, due to the fact that 5 models have been found non-compliant only for the

freezing capacity and/or the temperature raise time.

7 In reality 82 models were tested at the end of the project instead of 80, because, for two appliances, the

three additional units purchased for developing Step 2 belonged to a different model.


Figure 5: Overall compliance results of the ATLETE project

These results can be broken down by the different attributes in base of which the tested

models can be classified (i.e. by country, model type, energy efficiency class). A very

detailed analysis is provided in the deliverable: "Outcome of the pan-EU compliance of

refrigerators and freezers" available in the ATLETE website8." Here some spots are

provided mainly focusing on the energy efficiency and the storage volume compliance

rate.

For what concerns in particular the compliance with the Efficiency Classes, Errore.

L'origine riferimento non è stata trovata. shows that with respect 70 models for which

the steps 1 and 2 have been carried out, 55, that is the 79%, have correct energy

efficiency class declaration (the green labels in the figure) while 10 tested models (14%)

were overrated by 1 class and 4 tested models (6%) were overrated of 2 or even more

classes. 13 models were unfortunately classified as undecidable (the yellow labels) as no

3 equivalent models were found in the market at the time to carry out the second

testing step.

8 See: http://www.atlete.eu/index.php?option=com_docman&Itemid=111

79%

77%

90%

73%

84%

70%

43%

21%

23%

10%

27%

16%

30%

57%

0% 20% 40% 60% 80% 100%

Energy class

Energy consumption

Storage temperature

Storage volume

Temperature rise time

Freezing capacity

TOTAL

PASS

FAIL


Figure 6 Efficiency classes declared / verified

More in detail, Figure 7 shows the percentage difference between the declared and the

measured value of the energy consumption in Step 1 and Step 2 (for Step 2 the average

of the three measured values) for the 82 models of refrigerating appliances.

A part from the models largely exceeding the permitted tolerances, it is worth noting

that for some models (for example models N. 6, 18 and 50) the difference between the

rated energy consumption and the average measured energy consumption of Step 2 is

much smaller that the difference found in Step 1, but also the contrary happens (for

example models N. 36, 37, 49 and 72).

For sake of details and to understand the importance to allow a tolerance range for the

energy efficiency compliance, it is worth mentioning also Figure 8 that shows the energy

consumption difference between the rated energy consumption and the measured one

for Step 1 and for the single units of Step 2. It is evident that for some models, for

example in models N. 23, 31, 50 but also in models 35 and 57 there is a significant

difference among the four tested units.

Another critical parameter influencing the energy consumption is the storage volume. It

is actually clear that the energy consumption of a model can be increased (in order to

include the model in a higher energy efficiency class) by delivering a storage volume

lower than that actually declared.

Figure 9 shows the percentage difference between the declared and the measured value

of the storage volume, for the 82 models of refrigerating appliances. The black line

highlights the 3% max allowed tolerance. To this end it is worth noting that it is not by

chance that the major differences concern the freezer volume, that is the component

that consumes the more in the refrigerating appliances

Legenda: Undecidable Compliant Non Compliant


Figure 7 Difference (%) between the declared and the measured value of the energy consumption in Step

1 and Step 2 for the 82 models of refrigerating appliances

Figure 8 Difference (%) between the declared and the measured value of the energy consumption in Step

1 and Step 2 (single units and average values)

Figure 9 Difference (%) between the declared and the measured value of the storage volume for the

refrigerator and freezer compartments for the 82 models of refrigerating appliances

-30,0-25,0-20,0-15,0-10,0

-5,00,05,0

10,015,020,025,030,035,040,045,050,055,060,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82

Dif

fere

nce

in e

nerg

y co

nsum

ptio

n (%

)Step 1 Step 2

-35,0-30,0-25,0-20,0-15,0-10,0-5,00,05,0

10,015,020,025,030,035,040,045,050,055,060,065,070,075,0

6 7 18 23 30 31 33 35 36 37 44 45 49 50 55 56 57 58 62 72 81 82

Dif

fere

nce

in e

nber

gy c

onsu

mpt

ion

(%)

Step 1

Step 2, 1st unit

Step 2, 2nd unit

Step 2, 3rd unit

Step 2, average

-30,0

-25,0

-20,0

-15,0

-10,0

-5,0

0,0

5,0

10,0

15,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82

Dif

fere

nce

in s

tora

ge v

olum

e fo

r th

e re

frig

erat

or a

nd fr

eeze

r co

mpe

rtm

ents

(%)

refrigerator volume

freezer volume


4 NON COMPLIANCE ANALYSIS

4.1 Identification of the non-compliance cases

In chapter 3 a classification of the tested models in accordance with the outcomes of the

tests themselves has been outlined. A case by case analysis of the tests results and the

consequent actions undertaken on these models is provided in detail in the quoted

deliverable on the "Outcome of the pan-EU compliance of refrigerators and freezers9".

To provide a good overview on the work carried out, the decision taken and the

observed outcomes of the testing campaign, it is enough to highlight here just three of

the eight cases in which the tested models have been classified. These cases refer:

• the first to the models formally not compliant with the Energy Labelling Directive

rules, that shows the numerous manner in which these rules can be not

observed;

• the second to the remedy actions undertaken by the manufacturers that have

signed the ATLETE "Protocol for manufacturers’ proactive participation”, that

shows that good results can be achieved when working in close collaboration

with the direct involved actors;

• the third to damaged or defective models due to external accident or

production/assembling problems. These cases are not common but they exist

(we had two cases of this type) and it is important to correctly deal with them.

4.1.1 Models formally non-compliant before the laboratory tests

Before starting the actual test, the laboratory checked each model for the compliance

with the more formal, but nevertheless mandatory, elements of the EU energy labelling.

Among the 82 models tested within the ATLETE project 12 models have shown some

formal non-compliance:

• lack of the label (the strip) or the product fiche in the technical documentation

accompanying the model

• presence of the label and the fiche but lack of one or more of the parameters to

be mandatory declared; the missing value(s) can be supplied at a later stage by

the manufacturer

• declared energy efficiency class not in line with the other declared parameters

(i.e. the energy efficiency class does not correspond to the Energy Efficiency

index calculated using the formulae provided in the labelling scheme with the

values of the declared parameters.

One or more of the above formal non-conformities can be present in the same model. In

particular:

• no “temperature rise time” in the fiche: 4 cases

• the energy efficiency class declaration is not correct: 2 cases

9 See: http://www.atlete.eu/index.php?option=com_docman&Itemid=111


• label not available with the appliance: 4 cases

• label with incorrect declarations: 2 cases.

The decision of the ATLETE team was to proceed with the laboratory testing to evaluate

the compliance with the labelling declarations as much as possible, but whatever would

it been the outcome of the tests, to consider in any case the model as non-compliant (at

least from the formal stand point).

4.1.2 Models that underwent a “manufacturers’ voluntary remedy action

Manufacturers of models suspected of non-compliance after Step 1 and having signed

the “Protocol for manufacturers’ proactive participation” were given the possibility to

accept the non-compliance results of Step 1 and go for a “voluntary remedy action”. As

said, in 18 cases this remedy action was put in place and documented:

• in 1 case the correct energy efficiency class was declared on the label, in the

product fiche and the other documentation accompanying the model

• in 5 cases the correct value of the freezing capacity was declared in the product

fiche and the other documentation accompanying the model

• in 3 cases the correct temperature raise time was declared in the product fiche

and the other documentation accompanying the model

• in 3 cases the correct value of the storage volume of the freezer compartment

was declared on the label, the product fiche and the other documentation

accompanying the model

• in 1 case the correct value of the storage volume was declared on the label, the

product fiche and the other documentation accompanying the model

• in 2 cases several parameters were non-compliant and the correct valued were

declared on the label, the product fiche and the other documentation

accompanying the model.

• in 3 cases the production of the model was discontinued, and in one of them the

manufacturer committed himself to declare the correct values of the non-

compliant parameters for the equivalent models still placed on the market.

The remedy action was supported by the manufacturer with some evidence, such as a

copy of the new and correct label, fiche or instruction manual.10

4.1.3 Models (suspected to be) damaged or defective

One or more units of a model under test could result damaged and/or defective either

before a test is run (at a visual inspection) or suspected to be damaged / defective as a

consequence of a deviating behaviour or an unexpected test result. Depending on the

time when the defect or damage is found or suspected the reaction could be the

repetition of Step 1 (or Step 2) on a different unit(s) of the same model or to ignore Step

10

The ATLETE had not the possibility nor the responsibility to verify that the new and correct

documentation has been actually enclosed in all models on the market or that the model claimed to be

discontinued was not sold with a different name/commercial code number


1 results and go directly for the Step 2 (should a sufficient number of additional units of

the same model be available on the market).

Although this event was thought to be very uncommon, we have experienced it twice:

• the unit did not pass Step 1 due to an apparent malfunctioning of the thermostat

that could be due to a damage of the specific unit or to a production defect. It

was decided to run Step 2 with three additional units as the best approach to

state that the non-compliance was due to the damage of the first tested unit and

not to a systematic defect of the model. The test run on the three additional

units confirmed that the first tested unit was damaged, but showed also that the

model was non-compliant.

• the model tested did not pass Step 1 for the freezing capacity (declared at 7kg).

Since a defect was found in the mounting of the freezer compartment gasket ()

that could have been the reason for the failure, it was decided to run again Step

1 on an additional unit of the same model.

Figure 10: Defective gasket of the freezer compartment door

This second unit, with no defective gasket, resulted fully compliant. But unfortunately,

due to the transition between the old and the new labelling scheme, the second unit,

although having the same commercial code of the first unit, showed a different declared

value for the freezing capacity (4kg) and bore the new energy label. Therefore, although

the model represented by the second unit is fully compliant to the energy labelling, a

doubt still remains on the model represented by the first unit, mainly because the failing

parameter for this first unit is the freezing capacity that has been modified (reduced) in

the new declaration on the energy label.

4.2 Analysis of the non-compliance macro-sources

The wealth of data provided by the test results of the ATLETE project allow, for the first

time, to run a semi-quantitative analysis of the non-compliance causes. Four main issues

can be analysed:

• the product category;

• the purchasing price;

• the manufacturing place;

• the expected energy efficiency class.


4.2.1 Product category

The 82 tested models belong to the following types:

• refrigerator-freezers (4 stars): 44 models

• simple refrigerators: 11 models

• chest freezers: 9 models.

• upright freezers: 8 models

• refrigerators with a low temperature compartment11

: 4 models

The compliance rate has been calculated for each of these types to evaluate if a specific

product type presents a more marked compliance rate compared to the other types. The

results of this analysis are presented in Table 3 where it is evident that the higher non-

compliance risks are concentrated to the low temperature compartment refrigerators

(but this indication has to be taken with care because the sample of this appliances type

is very small) and to the freezers.

Table 3: Summary of the results of the compliance verification of the different types of refrigerating

appliances

Product types PASS FAIL

Simple refrigerators

‒ for 1 model having failed Step 1 three additional units

were not available

70% 30%

refrigerators with a low temperature compartment

& refrigerator-freezers

‒ for 4 models having failed Step 1 three additional units

were not available

34% 66%

upright freezers


were not available

45% 55%

chest freezers


were not available

60% 40%

TOTAL 43% 57%

4.2.2 Purchasing price

The average price of the 80 tested models is 396,45 € as show in

Table 4 along with the disaggregation per appliance category (categories with similar

characteristics have been grouped).

It is worth noting that the average price of the compliant model is significantly higher

(+125,85 € ) than the average, while on the contrary the average price of the non-

compliant models is lower than the average (-80,42 € ) and the same happened for the

11

From 0 to 2 stars


models where 3 more units were not available on the market (-59,95 € ). The same

pattern results for the different product categories: the compliant models are always

more expensive than both non-compliant ones and models where 3 additional units

could not be found on the market. This trend is more clearly appreciated in Figure 11 to

Figure 13: compliant models are usually more expensive than the average, but there are

also appliances with a purchasing price below the average that can be compliant,

although less frequently.

Although due to the limited dimension of the ATLETE models sample the conclusion of

the project could not be fully applied to all refrigerating appliances sold annually on the

Community market but, from this first analysis, it seems that the purchasing price is a

good clue of possible non-compliance: more expensive models have a higher probability

to be compliant, while on the contrary cheaper models are more often non-compliant.

Table 4: Average, min and max purchasing price (€) for compliant and non-compliant for the appliance

categories defined in the labelling scheme

Category All models Refrigerators and

freezers Low temp. compartment Freezers

Compliance Yes No 3 n.a. Yes No 3 n.a. Yes No 3 n.a. Yes No 3 n.a.

average price 396,45 451,15 337,91 294,08

522,30 316,03 336,50 621,40 360,33 425,75 454,88 239,83 317,14 387,00 176,25 115,00

min 89,00 107,00 115,00 232,00 139,00 243,00 234,00 123,00 124,00 89,00 107,00 115,00

max 2.085,00 890,00 636,00 2.085,00 890,00 613,00 771,00 445,00 636,00 1.096,00 222,00 115,00

second min 174,00 123,00 124,00 -- -- -- -- -- -- -- -- --

second max 1.096,00 776,00 613,00 -- -- -- -- -- -- -- -- --

Difference to

the average 125,85 -75,56 -59,95 170,25 -90,82 -25,40 116,97 -98,08 -20,77 92,92

-

117,83

-

179,08

Figure 11: Purchasing price for the tested models of refrigerators and freezers

12

Note: in red the average purchasing price of all models of the group

12

The compliant model with the highest price of 2.085,00 Euro has been excluded from the picture

0,00

200,00

400,00

600,00

800,00

1.000,00

1.200,00

0 1 2 3 4

Pur

chas

ing

pric

e (E

uro)

3 n.a. No Yes


Figure 12: Purchasing price for the tested models of refrigerator-freezers and refrigerators with a low

temperature compartment

Figure 13: Purchasing price for the tested models of upright and chest freezers

Note: in red the average purchasing price of all models of the groups

4.2.3 Place of manufacturing

The manufacturing country or region is often, but not always, reported in the nameplate

attached at the back of each model. An analysis was ran comparing the manufacturing

place with the rate of compliance. Results are presented in Table 5 and Figure 14

Table 5: Place of manufacturing as and rate of compliance with the EU energy

labelling

Country of

manufacturing Compliant Non-compliant 3 n.a. Tot.

EU-Member State 14 7 3 24

EU generic 2 5 1 8

n.a. 6 16 6 28

non-EU countries 8 12 2 22

Total 30 40 12 82

0,00

100,00

200,00

300,00

400,00

500,00

600,00

700,00

800,00

900,00

1.000,00

0 1 2 3 4

Pur

chas

ing

pric

e (E

uro)

3 n.a. No Yes

0,00

100,00

200,00

300,00

400,00

500,00

600,00

700,00

800,00

900,00

0 1 2 3 4

Pur

chas

ing

pric

e (E

uro)

3 n.a. No Yes


Figure 14: Place of manufacturing and compliance with the EU energy labelling

Although, again, the dimension of the sample in not sufficient to draw statistically sound

conclusions, the trend is very clear: when the country of manufacturing is not indicated

(n.a. in Table 2 and Figure 6) or indicated in a very generic way such as “EU” the

probability of non-compliance is higher than when the country is clearly stated. The

same occurs when the products manufactured in a specified EU Member State are

compared to the products manufactured in non-EU countries (China, Korea, Thailand

and Turkey in the ATLETE project).

4.2.4 Declared Energy Efficiency class

A further correlation appears to exist between the declared energy efficiency class and

the compliance rate: the higher is the energy efficiency class the higher is the probability

to find a compliant model. This is especially true for classes A+ and A (Table 6). For the

latter, being de facto the lowest and the most populated energy efficiency class

currently available on the market the compliance rate is also the lowest, when

compared to higher classes A+ and A++.

Since a correlation appears to exist between the purchasing price and the energy

efficiency class, as shown in previous Figure 11, in the end, the highest is the purchasing

price and the energy efficiency class the highest is the probability of compliance.

Table 6: Relation between the declared energy efficiency class and the compliance

rate for refrigerating appliances within the ATLETE project

Declared energy

efficiency class Compliant Non-compliant 3 n.a. Total

A++ 2 0 1 3

A+ 14 8 2 24

A 12 31 8 51

B 1 1 0 2

C 1 0 1 2

D and lower 0 0 0 0

Total 30 40 12 82

0123456789

101112131415161718

EU-Member State EU generic n.a. non-EU countries

Num

ber

of t

este

d m

odel

s

Compliant

Non-compliant

3 n.a.


4.3 Further sources of uncertainty in the compliance verification

Several problems of different nature, some attributable to a somewhat market with no

rules or other to the interpretation of same Directives prescriptions, have been met

during the testing campaign. These problems, that have not hindered the

implementation of the verification procedure and the provision of the final results, are

highlighted in the following pages of this report to put in evidence how certain rules of

the reference Directives show their limits when challenged with the market reality.

4.3.1 The EU labelling Directive interpretation (1), concerning the models selection in

the market

According to the verification procedure of the EU labelling scheme13

the units of a model

to be verified shall be “randomly selected on the market”. The reason is that randomly

selected units should better represent the overall production of a model, and this is

particularly important when Step 2 is run.

But how the random selection concept can be applied to the actual purchasing of the

products models to be verified? During the ATLETE project the testing laboratories

purchased the appliances to be tested via on-line shops, because of the lower price

compared to other distribution channels and the relative easiness of the purchasing.

We nevertheless wonder if the models purchased in this way were really “randomly

selected”, since a situation where units belonging to the same production batch are

selected could frequently happen as shown in Figure 15 and Figure 16 in the following

page. In the second Figure the three units clearly belong to the same batch and were

produced very closely in time.

Figure 15: First unit of a model purchased for Step 1 of the verification

Figure 16 : Additional three units of the model purchased for Step 2 of the

verification

13

Described in the harmonised standard EN 153 for directive 94/2/EC and in Annex VII of delegated

Regulation 2010/1060/EU.


4.3.2 The EU labelling Directive interpretation (2): rounding of measured values

During the ATLETE Project we have experienced the importance of a clear description of

the rounding for the measured values for the compliance verification. This is of great

importance for the models that are “border line” with the energy efficiency class

thresholds or accepted tolerance or specific minimum conditions requested by the

standard. The importance of rounding is highlighted by the following case:

− Annual energy consumption declared on the label: the EU labelling directive 94/2/EC

establishes (Annex I, point 2.V) that the annual energy consumption of a model is

declared on the label with no digits; it is calculated from the measured energy

consumption in 24h multiplied by 365 and the resulting value is then rounded. Since

no rounding criterion is described the rounding to the nearest integer should be

applied.

Annex I, point 2.V

V. Energy consumption in accordance with standards referred to in Article 1 (2) but

expressed in kWh per year (i.e. per 24 hours × 365).

The harmonised standard EN 15505:2005 prescribes that the measured energy

consumption per 24h is declared with two digits (Clause 15.4), but the measured

value is usually reported:

In laboratory practice the energy consumption per 24h is measured and reported

with a number of digits higher than two. The use of two or more digits in the

compliance verification calculation can lead to completely different outcome for an

appliance that with a declared annual energy consumption of 124 kWh/year:

(c) 0,3904658 kWh/24h×365=142,52 kWh/year, rounded to the nearest integer is 143 kWh/year or

+15,3% higher than the rated value, the model is non-compliant for the energy consumption

(d) 0,39 kWh/24h×365=142,35 kWh/year, rounded to the nearest integer is 142 kWh/year or +14,8%

higher than the rated value, the model is compliant for the energy consumption.

This situation can happen whatever is the allowed tolerance and for any parameter

to be verified if the performance of the tested model is very close to the minimum

requirement or the threshold value.


4.3.3 Market problem (1): different labels and declarations for the same appliance

model

Apart from the specific and rare case (described in the following paragraph) when the

labels ruled by different legal provisions are shown on the units of the apparently the

same model, a more frequent situation can be finding units of the same model bearing

the same label layout but with different valued declared for the same parameters or

even with different parameters declared, as shown in Case 1 and Case 2 below.

a) Case 1: Same labels layout but with different declared values for the same parameters

as shown in Figure 17

Step1 Step 2

Label strip of unit 1 Label strip of unit 2 Label strip of unit 3 Label strip of unit 4

Serial N. 094403645 Serial N. 104104696 Serial N. 095004226 Serial N. 102203466

Figure 17: Label strips for the 4 units of the same models used in Step1 and Step2 of the verification

procedure

b) Case 2: Same labels layout but different declared values and different parameters

In this second case the four units of the model not only presented a different value for

the annual energy consumption, but also according to the label strip for the second

appliance used in Step 2 (units 3 in Figure 18) a 15 litre 4-star freezer compartment is

present, that instead is not present in the appliance.


Step1 Step 2

Label strip of unit 1 Label strip of unit 2 Label strip of unit 3 Label strip of unit 4

Y2010W10 Y2010W10 B32 Y2010W37

Figure 18: Label strips for the 4 units of the same models used in Step1 and Step2 of the verification

procedure

4.3.4 Market problem (2): same commercial code number for technically different

appliance models

A refrigerator freezer with a freezing capacity of 7kg/24h and a declared energy

consumption of 219 kWh/year (Figure 19), failed Step 1 for the freezing capacity, but the

gasket of the freezer door was found defective. Therefore another unit of the same

model (same commercial code number) was purchased from the market by the

laboratory; but this second unit was labelled according to the new labelling scheme with

a freezing capacity of 4kg/24h and an energy consumption of 230 kWh/year (Figure 20).

The manufacturer was asked for about the differences of the label declarations of the

two units, apparently belonging to the same model. The answer was that with the new

regulation 1060/2010/EU the manufacturer has switched to the new energy label

declaration and has optimised most of its appliances according to it. For the specific

model under investigation this optimization has caused technical modifications which

then lead to a change of the energy consumption and freezing capacity compared to the

old label. The testing laboratory has bought the second unit at the end of May 2011 and

therefore tested a model with the new energy label.

Although this specific case was due to the change in the EU legislation on energy

labelling, a supplier is currently permitted to apply technical modifications and to

declare different values for any parameter of an appliance without necessarily change

the name and identification number of the model.


Figure 19: Rating plate and label strip for the first unit of the refrigerator-freezer

Figure 20: Rating plate and label strip for the second unit of the refrigerator-freezer

4.3.5 Market problem (3): different suppliers for the same appliance model

For one of the models for which both verification steps were run, a unique supplier

could not be identified because the same appliance with the same commercial code

number was imported and commercialised by two different suppliers in different EU

Member States.

The four rating plates are shown in Figure 21: the first tested unit belong to a specific

importer while the other three (used in Step 2) to a different importer.


First unit of a model purchased for Step 1 of the verification

Additional three units of the model purchased for Step 2 of the verification

Figure 21: Different suppliers for different units of the same model purchased in the same country

In this situation, which of the two suppliers will be – from the legal point of view –

responsible for the non-compliance or for the eventual correcting action or a possible

sanction? Should each suppler be responsible only for the units placed on the market

under its own responsibility?

4.3.6 Improvement of the test method

The lesson learned from the test campaign together with the input and suggestions

received from the selected laboratories (four meetings were actually organized) as well

as the outcome of the workshops held with the ATLETE advisory group14

have allowed

the ATLETE working group to finally formulate the following improvements for the

refrigerating appliances standard:

• to shorten the time required for the completion of the verification process, the

specific setting(s) used by the supplier for the measurement of the parameters at

the basis of the labelling declaration and the compliance with the ecodesign

requirements should be specified either in the labelling/ecodesign technical

report or to the booklet of instructions or to any other technical documentation

accompanying the products

• storage volume measurement is still critical, at least for some manufacturers and

products configurations. The need for further clarification should be evaluated by

the standardisation experts

14

An ad hoc Advisory Group composed by relevant stakeholders and representative of the ADCO group

has advised and supported the ATLETE project over all its life


• the 2 star compartments identification is controversial and requires a further

clarification

• the use of “cold plates” (eutectic accumulators) should be ruled and the impact

on the load plan and the appliance volume measurement described. The use of

eutectic plates can be accepted but only in accordance with the already

established standard conditions for the volume measurement:

• the appliance storage volume should exclude the space needed for the plates

• the load plan should show the position of the plates, that cannot be placed

directly over the stacks

• rounding rules have to be both improved and better specified: not only for the

declaration of the parameters but also in the intermediate calculations

• the relation between the instruments uncertainty (usually described in the

standards) and the measured values is apparently not clear at least for some

manufacturers. A manufacturer claimed that the failing to comply – of 0,5K - with

a storage temperature cannot be accepted because the accuracy of the

measurement instrument is of the same order of magnitude. The answer of the

ATLETE team was that storage temperature must be respected beyond the

accuracy of the measurement instrument, once this accuracy is in accordance

with the standard specifications

• devices designed specifically to underestimate energy consumption under test

conditions should be clearly prohibited.

5 MAIN CRITICAL ELEMENTS OF THE PROPOSED PROCEDURE

After the completion of the field work and in the light of the achieved results, a critical

analysis of the applied evaluation procedure (see chapter 2) was developed with the aim

to correct the aspects that have presented uncertainty or shown to be excessively

cumbersome.

Some of the critical aspects have been already outlined in chapter 4, others were

highlighted by the testing laboratories, that at the end of the field work were asked

about their views about the verification procedure developed within the ATLETE project

and how it could be improved.

5.1 Models sampling, purchase and identification

In general, models scouting, purchase and logistic is a resource-consuming (human,

time, financial) task that requires market-skilled personnel and appropriate timing. The

scouting of the appliances to be tested should not be in charge of the testing

laboratories, but either of the national Market Surveillance Authority or of a delegated

subject.


5.1.1 Models sampling

The sampling criteria should be defined to tackle also models placed on the market in

“small” batches by manufacturers and/or importers, to avoid that the lack of 3

additional units to be tested will impede the completion of the compliance verification.

Possible solutions are:

• immediate purchase of 4 units for each of the selected models, at least for

suppliers with a low market share. This approach will increase the costs of the

verification and in case a model passes Step 1 the other 3 units will be useless.

Nevertheless, if 3 units will be left untouched in the original packaging the

national Market surveillance Authority could give them back to the retailer or

even purchase the first unit and have the other 3 as a “free sample” to be paid

only if tested; this approach could be proposed to retailers only if the time to

finalise the test is acceptable, on the contrary the 3 units could remain unsold

• purchase of 3 units for each of the selected models, at least for suppliers with a

low market share, to run Step 2 as the unique verification stage; this approach

will increase slightly less the purchasing costs, but will increase the testing costs

since only Step 2 could be run

• delete from the list of selected models those not available in at least 4 units: this

approach will create areas with de-facto no compliance verification

• use “equivalent” models: this approach is applicable a sufficient number of units

of equivalent models(s) are available and if the “equivalent” models are actually

equivalent.

For the same reason, once the total amount of models to be verified has been decided,

the actual selection - including the publication of the selected models list - should not be

done at the beginning of the verification action but in several phases. This approach will

reduce the time-to-test (i.e. the time from the announcement that a specific model will

be tested to the time of the actual completion of the tests including the purchase of the

1+3 units) to a minimum, thus reducing the possibility for the model to disappear from

the market.

The actual meaning of “random” selection and if necessary the minimum criteria to

assure a random selection should be defined as much as possible, at least in a qualitative

way. In this respect:

• can units purchased in the same market place be considered randomly selected?

• are models purchased in an on-line shop eligible for a verification action? In fact

sometimes the time for delivering the on-line purchased goods is quite long and

may indicate that the products are directly looked for from the

manufacturer/importer. This event should be carefully avoided because to

assure a fully independent verification action.

• Models selection could be based not on “market analysis”, but on complaints or

evidence obtained in a pre-evaluation process (such as a risk analysis) to detect

products highly suspected of non-compliance, that are eventually selected for

testing.


5.1.2 Models purchasing

Purchase of the selected models: the purchase of products in foreign countries and on

internet was applied for the ATLETE project for the two fold reason that consumers

often buy products on-line, and that four laboratories should procure models from all

European countries. But at national Member State level this purchasing process could be

not in line with the legal and administrative procedures. For example, the Consumer Law

in Spain that regulates the market surveillance procedures for the Energy Labelling

Authorities require to select the sample in the market (retailer warehouse, or shop) and

directly by an Authority Inspector that identify and mark the samples to be sent to the

specific testing laboratory. In addition this Law allows, under requirement of the

manufacturer or importer involved in the market surveillance that they may be present

when the laboratory opens the appliance package.

5.1.3 Models univocal identification

The ATLETE field work has shown that not always appliance models are univocally

identified by their commercial code number (or a similar coding). The challenge is to

ensure that units apparently belonging to the same model are actually the same

product. Or in other words how far two apparently identical units of a model can have

different characteristics (different components, technical details, performance

declarations, etc.) and still be considered two units of the same model? In this respect a

better system of univocal identification of models should be developed.

5.1.4 Laboratories selection and remuneration

The number of selected laboratories for a verification action should be adequate:

• to the dimensions of the sample to be tested, to avoid saturating the laboratories

for some weeks/months

• to the number of existing laboratories able to test the same product, that in turn

influences the overall number of models that can be actually tested.

Adequate remuneration should be paid to the laboratories, including for actions beyond

the mere models testing and results reporting if such actions are under the responsibility

of the laboratory. Knowing that - if selected - their work will be adequately remunerated

a larger number of laboratories will participate to the selection procedure thus

guaranteeing that the most skilled laboratories are selected and a stronger motivation

for laboratories to improve their testing equipment and ability.

5.1.5 Final test report

Laboratories provide test reports according to the applied standard(s) or legislation. But

test reports can be quite different in format and content and can provide a different

amount of information (i.e. temperature graphs, copy if the label and of the booklet of

instructions, etc.). This makes the comprehension not immediate and the comparison of

the tests run in different laboratories on the same products quite complex.


In this respect, is a mandatory common test report format, i.e. something more detailed

and structured that what is currently described in the standard, be adequate? Or will it

prevent the necessary flexibility?

In the case of the ATLETE refrigerating appliances all laboratories have verified also the

compliance with the energy efficiency class. In some cases stating that considering the

allowed tolerance it was fulfilled, although using the measured values of the parameters

it was often lower. This is in our view exceeding the role and responsibility of the

laboratories, that should limit their action to only testing and certifying the compliance

of the measurable parameters.

The final responsibility for the model compliance with the (labelling/ecodesign)

legislation requirements should be under the responsibility of the national Market

Surveillance Authority or its delegated subject.

5.1.6 Measured parameters in Step 1 and Step 2 of the compliance verification

Within ATLETE project the more restrictive interpretation of the measurements to be

done in Step 1 and Step 2 of the verification procedure was applied, asking laboratories

to re-test four of the five parameters (the storage volume has not been re-measured

since the same result would have been achieved) of a model in Step 2. Although some

of the laboratories had a different opinion, we believe that only this approach can prove

beyond any doubt the compliance of a product.

The consequences of this approach falls also on the test reports to be delivered by the

test laboratories: in the Step 2 report the laboratory should provide the average values

of all the tested parameters and not only of the elementary test results for each unit.

The appropriateness of our decision was in the end confirmed by the results of the field

work: some of the models failed Step 2 not only for the parameters having failed Step 1

but also for additional parameters that before were instead compliant.

5.1.7 Improvement of the test method

The test procedure needs to be shared by all market actors before any enforcement or

verification testing begins. In addition, testing laboratories may have different

interpretations of specific test conditions. Since without consistency in test methods it

will be impossible to run an effective verification and enforcement programme the

standard should be unambiguous. If necessary, interpretations should be prepared to

clarify the still unclear conditions. On the other side a care should be taken by the

national Authorities and test laboratories that a claim of “ambiguity” is not being used

as a way out from too stringent testing conditions.

In particular, suggested improvements for the refrigerating appliances standard are:

• to shorten the time required for the completion of the verification process, the

specific setting(s) used by the supplier for the measurement of the parameters at

the basis of the labelling declaration and the compliance with the ecodesign

requirements should be specified either in the labelling/ecodesign technical


report or to the booklet of instructions or to any other technical documentation

accompanying the products

• storage volume measurement is still critical, at least for some manufacturers and

products configurations. The need for further clarification should be evaluated by

the standardisation experts

• the 2 star compartments identification is controversial and requires a further

clarification

• the use of “cold plates” (eutectic accumulators) should be ruled and the impact

on the load plan and the appliance volume measurement described. The use of

eutectic plates can be accepted but only in accordance with the already

established standard conditions for the volume measurement:

o the appliance storage volume should exclude the space needed for the

plates

o the load plan should show the position of the plates, that cannot be

placed directly over the stacks

• rounding rules have to be both improved and better specified: not only for the

declaration of the parameters but also in the intermediate calculations

• the relation between the instruments uncertainty (usually described in the

standards) and the measured values is apparently not clear at least for some

manufacturers. A manufacturer claimed that the failing to comply – of 0,5K - with

a storage temperature cannot be accepted because the accuracy of the

measurement instrument is of the same order of magnitude. The answer of the

ATLETE team was that storage temperature must be respected beyond the

accuracy of the measurement instrument, once this accuracy is in accordance

with the standard specifications

• devices designed specifically to underestimate energy consumption under test

conditions should be clearly prohibited.

6 MARKET SURVEILLANCE PROCEDURES AT EU AND INTERNATIONAL LEVEL

AND COMPARISON WITH THE ATLETE METHODOLOGY.

In the quoted deliverable "Outcome of the pan-EU compliance of refrigerators and

freezers" as well as in the WP 2 deliverable: "Overview of the methodology and the

results achieved in the conformity assessment action at EU Member States and

international level15

", a detailed analysis of the international (Australia and USA) and

European legislation and practice on the verification procedures has been carried out. A

synthesis of the outcome of this analysis, as well as of a survey carried out at the

beginning of the ATLETE project to assess the manufacturers’ conformity practices in the

EU Member States, is outlined in this report to provide the picture of the EU and

International situation on the surveillance procedures and compare them with the

ATLETE methodology.

15

See for both deliverables: http://www.atlete.eu/index.php?option=com_docman&Itemid=111


The extra-EU countries analysed have been Australia, New Zealand and the US. while for

three EU countries: Sweden, Denmark and UK (the only ones for which the data on the

surveillance activity are public available) the results achieved by market surveillance

actions developed for refrigerators and freezers in the last decade are briefly outlined.

6.1 Market surveillance procedures followed at international level

6.1.1 Australia and New Zealand

The Check Testing Process

In April 2011 the report “Performance Verified” was published in Australia. It presents

the results of 1.000 laboratory tests conducted between 1991 and 2010 undertaken to

verify the performance of appliances and equipment regulated for energy efficiency in

Australia and New Zealand. The two relevant types of regulatory programmes are

minimum efficiency requirements (called Minimum Energy Performance Standards) and

energy labelling, both of which are part of the Equipment Energy Efficiency (E3)

programme.

In Australia and New Zealand all products within the scope of energy labelling and

minimum efficiency requirements regulations must be registered and meet the specified

performance requirements. In the case of labelling, products are also required to display

the correct label indicating the product’s performance. Some individual products have

additional requirements.

Amongst other tasks, the E3 Committee administers a monitoring, verification and

enforcement programme to maximize compliance with energy efficiency regulations and

ensure that the overall efficiency goals of the programme are being met. These activities

are designed to raise the perceived risks of non-compliance by stakeholders in the E3

Programme, following the principles of deterrence theory16

Activities undertaken by the E3 Committee to monitor compliance include:

• aiding compliance through education, stakeholder forums and other

communication activities

• maintenance of a registration database and on-line registration facility (see

www.energyrating.gov.au)

• in-store surveys to check that the correct labels are being displayed

• checks to ensure that products on the market are registered

• and check testing: check testing or verification testing, is undertaken according

to the methodology specified by a test standard or protocol and is the only

means to confirm whether individual models meet the mandatory performance

requirements and that energy labels indicate the correct level of performance.

Check testing provides several important functions with respect to the energy

efficiency regulatory programme:

16

“Deterrence theory… maintains that there must be a credible likelihood of detecting

violations, swift, certain, and appropriate sanctions upon detection; and a perception

among the regulated firms that these detection and sanction elements are present”.


o it helps to ensure that the projected energy and greenhouse gas savings

are delivered

o it safeguards the integrity of the programme, maintaining consumer and

industry confidence in the energy performance labels, and

o it protects the investment made by manufacturer’s and product suppliers

producing compliant equipment from being undercut by non-compliant

products.

The check testing process is similar to that envisaged in the EU Directives. It actually

comprises a stage 1 test (also known as the “screen test”) performed on one sample of

the model and of a stage 2 check testing to be carried out if the model is found to be

non-compliant. The sample to be tested is generally independently purchased (usually

through a retail outlet), paid for by the regulatory authority and tested by a laboratory

accredited for check testing17

. If, after test 1 the model is found to be non-compliant the

supplier has the choice of either cancelling the registration for the model or proceeding

to stage 2 check testing. Stage 2 involves the testing of either two further units (where

the failure relates to minimum efficiency requirements) or three units (where failure

relates to claims on labels) paid for by the registration holder (i.e. the subject that places

the product on the Australian market). The actual units to be tested in stage 2 are

randomly selected from the stock by a representative of the regulatory authority.

Guidelines18

explaining how the national mandatory scheme for energy labelling and the

minimum energy efficiency requirements are administered have been issued by and ad

hoc committee of the Australian and New Zealand Minerals and Energy Council.

Although not legally binding, the purpose of the Guidelines is to act as a guide to

facilitate uniform and consistent practice among central and regional regulatory

agencies and to explain to stakeholders the responsibilities of relevant State and

Territory regulatory agencies and the responsibilities of industry. The Guidelines include,

inter alia, a detailed description of the programme compliance monitoring through

laboratory check testing.

The E3 check testing programme aims to cover most major categories and types of

regulated appliances and equipment. Individual models are not randomly selected from

the stock of models sold, but are chosen according to a set of criteria; the criteria are

based on established ‘risk factors’ that indicate that a product has a higher probability of

failure compared to the market19

. These factors can be summarised as:

• suppliers and models with a demonstrated record of check testing non-

compliance because of the likelihood of a continuation of such historical trends

• where a third party, such as competitors, consumers, consumer groups or

regulatory agencies, provide evidence of non-compliance, for example the results

of in-house or independent laboratory tests

• products that appear as new brands on the market or from suppliers that do not

have any check testing track record

17

The test can be carried out only by laboratories accredited to the NATA (National Association of Testing

Authorities) or by other bodies with a mutual recognition agreement with NATA 18

“Administrative Guidelines for the Appliance and Equipment Energy Efficiency Program of Mandatory

Labelling and Minimum Energy Performance Standards”, Edition 5, June 2005, downloadable from:

http://www.apec-esis.org/www/UploadFile/Australian_admin_guidelines.pdf 19

Criteria are published on the public website at: http://www.energyrating.gov.au/checktest-criteria.html


• models tested in previous years of the check test programme are excluded from

any further testing unless specific evidence becomes available to suggest that a

re-test is warranted.

Amongst products selected according to ‘risk factors’, products that are likely to have

the greatest adverse impact on energy and greenhouse gas savings are prioritised for

check testing. These include models belonging to the following groups:

• models with high volumes of sales because of their greater potential to impact

on energy usage as compared to models with low sales volumes

• models with the highest claims for energy efficiency (e.g. high star ratings)

because of the market’s higher expectations with respect to the performance of

these models as compared to models with low ratings

• newer models will normally be given preference when considering models for

check testing because of their potential to remain on the market for a longer

period as compared to older models.

Results of the check testing process

The years 2005-2010 have seen considerable expansion in the check test programme,

driven by a corresponding growth in the number of regulated product categories and

the number of individual products registered, which have increased considerably over

the past ten years. The number of stage 1 check tests undertaken each year has varied

considerably, from less than 10 to about 150, and this is partially explained by changing

budget allocations and the mix of products tested, due to the variation between the per

unit test costs for different product categories; 65% of all check tests have been

conducted on refrigerators and freezers, air conditioners, and clothes washers, which

were amongst the first products to be regulated (Figure 22)

Figure 22: Total number of check tests conducted in Australia by appliance category, 1991-2010

An overview of the results by product category , referring to 1,000 appliances tested, is

presented below.


• 73% (730 products) were found to meet the energy performance requirements

contained in energy efficiency regulations

• the registration of 207 (21%) models was cancelled by regulators

• for an additional 1% of cases, action by regulators included the referral of three

products for cancellation or to the Australian Competition and Consumer

Commission.

• a further 4 failed products were found to be older stock imported or

manufactured prior to the introduction of new regulations and allowed to be

sold under grandfathering provisions

• 22 (2%) failed tests are either still awaiting resolution, the registration has

expired or were found not be registered at the time of testing

• suppliers either withdrew or corrected the registration details, or re-registered a

total of 39 (4%) products

• as far as the specific Stage 1 results are concerned, of the 388 products that

failed at stage 1 testing, 92 (24%) passed the stage 2 test, 19 (5%) were found to

be no longer sold and 2 (1%) are still awaiting a decision by the regulator.

6.1.2 USA

The Energy Policy and Conservation Act

The U.S. Department of Energy (DOE) regulations require manufacturers of products

covered under the Energy Policy and Conservation Act (EPCA) to submit compliance

statements and file certification reports to DOE verifying that the products meet the

relevant energy efficiency or water conservation requirements.

In late 2009, the Office of the General Counsel issued a new legal interpretation of its

enforcement authority under the EPCA which viewed improper certifications or failure

to file certifications as EPCA violations "subject to enforcement action, including the

imposition of civil penalties. Concurrently DOE has undertaken a new initiative to

aggressively enforce energy efficiency and water conservation requirements for a

variety of products, including air conditioning and heating systems, kitchen and other

household appliances, televisions, lighting, bathroom fixtures, dehumidifiers and battery

chargers.

In particular in October 2009, DOE announced three new steps to strengthen its ability

to enforce energy efficiency requirements:

• the formation of an enforcement team within the Office of the General Counsel

that comprises of lawyers with extensive litigation and regulatory experience.

• a programme to randomly review manufacturers' compliance with DOE

certification requirements. The enforcement team will be initiating a compliance

review of certification reports for covered consumer products. DOE will randomly

select previously filed certification reports for review, request certification

records as needed, and hold manufacturers accountable for failing to certify

covered products according to DOE rules

• the issuing of guidance further detailing its energy efficiency enforcement

regulations. In its new guidance on Energy Efficiency Enforcement Regulations,


the Department confirms that under existing DOE regulations, DOE can take

enforcement action and assess civil penalties if a manufacturer fails to properly

certify a covered product and retain records. Specifically, the Department

clarifies that any failure to certify covered products according to DOE's rules

violates the Energy Policy Conservation Act of 1975 and DOE's regulations.

A new set of rules, overhauling the past ones, and concerning the enforcement of DOE’s

efficiency requirements for appliances, lighting and other products were then issued by

DOE in 2011. These new rules are designed to encourage compliance and prevent

manufacturers who break the law from gaining a competitive advantage over those that

adhere to it. Among other changes, the rules will - for the first time - allow the

Department to test products on its own initiative to determine whether they comply

with DOE’s efficiency standards. Using its reinforced enforcement authority, DOE has

already issued civil penalties and testing orders for covered products from several

manufacturers

The verification of the Energy Star Programme

American consumers, businesses, and federal agencies rely on the Energy Star

programme to identify products that decrease greenhouse emissions and lower energy

costs. In addition, the federal government and various states offer tax credits and other

incentives to encourage the use of energy-efficient products including Energy Star

products. Specifically, approximately 300 million USD from the American Recovery and

Reinvestment Act will be used for state rebate programmes on energy-efficient

products.

The Energy Star programme is overseen jointly by the U.S. Department of Energy and

the U.S. Environmental Protection Agency (EPA). The programme is currently based on

self-certifications by manufacturers with after-market tests and self-policing that ensure

standards are maintained.

Given the millions of dollars allocated to encourage use of Energy Star products and

concerns that the Energy Star programme is vulnerable to fraud and abuse, the U.S.

Government Accountability Office (GAO) was asked to conduct proactive testing to

obtain Energy Star partnership status for bogus companies and submit fictitious

products for Energy Star certification. GAO's investigation showed that Energy Star is for

the most part a self-certification programme vulnerable to fraud and abuse20

. In fact

GAO found that for the bogus products certification controls were ineffective primarily

because Energy Star does not verify energy-savings data reported by manufacturers.

On the basis of a new Memorandum of Understanding between DOE and EPA, the

programme was shifting toward a more rigorous up-front screening process. The first

phase of the initiative started in January 2011, with the launch of a new rule requiring

products to be tested for compliance by an EPA-approved third-party certification body.

Phase two involves verification testing, to make sure that products which are certified as

Energy Star-compliant stay that way.

20

For further information on this investigation see: http://www.gao.gov/products/GAO-10-470


Starting January 1, 2011, new products that meet Energy Star specifications must be

certified by an EPA-recognized Certification Body (CB) before the product can be labelled

with the Energy Star mark (Figure 4). Upon certification of a product based on test data

provided by an EPA-recognized laboratory, the CB will notify the partner that the

product meets the Energy Star requirements and will submit the qualified product data

to EPA for listing on the Energy Star website21

. In addition to certifying products as

Energy Star, the CB is required to operate a verification programme ensuring on an

annual basis that a certain percentage of basic models it has certified continue to meet

the requirements through verification testing again using an EPA-recognized laboratory.

Testing laboratories are either directly EPA accredited or are EPA-recognized through

enrolling in a Certification Body's WMTL (witnessed manufacturers’ testing laboratory)

or SMTL (supervised manufacturers’ testing laboratory) programme. The list of

Certification Bodies and the accredited laboratories is available on the Energy Star

website.

In addition to the CB-run verification programmes, the DOE verification testing

programme will target certain Energy Star products that are also covered by DOE’s

regulatory programme such as: refrigerators, freezers, washing machines, dishwashers,

water heaters, room air conditioners, central air conditioners and heat pumps, furnaces

and televisions.

In March 2010, DOE launched a pilot action to verify the energy efficiency and water-use

characteristics of clothes washers, dishwashers, freezers, refrigerators, room air

conditioners, storage water heaters, and tankless water heaters through laboratory

testing. The pilot action was used also to validate the basic aspects of the verification

procedure.

The available results of the tests show that so far 110 models have gone through initial

energy testing. The agency plans to test an additional 157 models and expand testing to

other Energy Star products under long-term programs.

DOE randomly selected 20% of basic models (i.e. a model that includes all variations of a

unit made by a manufacturer that have the same energy-use characteristics) for testing,

and then modified the selection based on strategic considerations:

• basic models picked through random selection

• omitted basic models that were not available for sale through normal retail

distribution channels

Table 7shows the appliances types and the number of models to be verified.

Table 7: Final number of selected models for the DOE verification pilot programme

Appliance Number of basic

models

Target number of

tests

Refrigerators 405 81

Freezers 121 26

Washing machines 200 40

21

For additional information: www.energystar.gov/index.cfm?c=partners.enhanced_test_verification


Appliance Number of basic

models

Target number of

tests

Dishwashers 54 11

Thankless WH 59 12

Storage WH 55 11

Room air conditioners 398 80

Total 1.292 261

The verification testing process begins with a Stage I test, then proceeds to a Stage II

test, if warranted (Figure 23):

‒ Stage I testing: screening tests, leading to no action or further actions

‒ Stage II: determining the compliance if action is required after Stage I

Figure 23: The two-stage verification testing process for the DOE pilot action

If the results of testing indicate that a product does not meet applicable minimum

Federal energy conservation requirements, DOE will initiate an enforcement action. This

enforcement action could lead to further testing, and ultimately, to the government

requiring that the product no longer be distributed in commerce in the U.S. If a product

does not meet Energy Star programme requirements DOE will refer the matter to EPA

for enforcement.

Table 8 shows the preliminary results of the pilot action. It is worth noting that about

17% of the tested models require further action.


Table 8: Preliminary results for DOE the verification action

Appliance Total models Tested

models Units failing Stage 1

Refrigerators 82 66 8

Freezers 24 10 3

Washing machines 40 32 5

Dishwashers 11 10 2

Thankless WH 12 10 0

Storage WH 11 9 3

Room air conditioners 87 39 9

Total 267 176 30

6.2 Brief overview on the formal procedure followed in EU Member

States

The verification procedure for the old energy labelling of major household appliances -

still in force until when the provisions of the new delegated Regulations will mandatorily

apply - is based on a specific Annex of the applicable harmonised standard(s), which in

turn is mentioned in the product specific directive(s).

In the new labelling delegated Regulations, as well as in the ecodesign Regulations, the

essential elements for the verification procedure are instead explicitly mentioned (in

one if the Annexes) among the legislation provisions.

Whether in the case of an old implementing directive or a new delegated regulation, the

verification procedure is based on a two-step approach: in Step 1 the check is performed

on one sample of the model; in case of non-compliance Step 2 is developed, checking

three additional samples of the same model.

Depending on the parameter to be verified, a verification tolerance (that takes into

consideration the uncertainty in the laboratory measurements) is applied to both Steps.

In this respect it is worth noting that while in the old labelling directives the tolerance

accepted in Step 1 was larger than that accepted in Step 2, in the new delegated

regulations the allowed tolerance is the same in both Steps in most cases.

A summary of the EU verification system for the energy consumption for the old and

new energy consumption declarations in the energy labelling and energy efficiency

requirement schemes is presented in Table 9 and Table 10.


Table 9: Summary of the EU verification system and tolerances for energy consumption in the previous

labelling and minimum requirements schemes

Appliance Implementing

Directives

Harmonised

standard


Step 1 Step 2 Units Tolerance Units Tolerance

(n) (%) (n) (%)

(old) Energy labelling scheme

Refrigerators&freezers 94/2/EC/2003/66/EC EN 153 1 15% 3 10%

Washing machines 95/12/EC/96/89/EC EN 60456 1 15% 3 10%

Tumble dryers 95/13/EC EN 61121 1 15% 3 10%

Washer-dryers 96/60/EC EN 50229 1 15% 3 10%

Dishwashers 97/17/EC/99/9/EC EN 50242 1 15% 3 10%

Air conditioning 2002/31/EC EN 14511 1 15% 3 10%

Ovens 2002/40/EC EN 50304 1 40Wh+10% 3 10%

Efficiency requirements scheme

Refrigerators&freezers 96/57/EC EN 153 1 15% 3 10%

Table 10: Summary of the EU verification system and tolerances in the new labelling and ecodesign

schemes

Appliance Implementing

regulation Standard


Step 1 Step 2 Units Tolerance Units Tolerance

(n) (%) (n) (%)

(new) Energy labelling scheme

Refrigerators&freezers 1060/2010/EU EN 153 1 3/10% 3 3/10%22

Washing machines 1061/2010/EU EN 60456 1 10%23

3 6/10%24

Dishwashers 1059/2010/EU EN 50242 1 10/19%25

3 6/10/19%26

Televisions 1062/2010/EU IEC 62087* 1 7% 3 7%

Ecodesing requirements

Refrigerators&freezers 643/2009/EC EN 153 1 3/10% 3 3/10%27

Washing machines 1015/2010/EC EN 60456 1 4/10%28

3 4/6/10%29

Dishwashers 1016/2010/EC EN 50242 1 10/19%30

3 6/10/19%31

Televisions 642/2009/EC IEC 62087* 1 7% 3 7%

*EN standard in preparation by European standardisation bodies under a mandate from the European

Commission

22

3% for gross volume and storage volume, 10% for all other parameters (except noise). 23

10% for all parameters (except noise). 24

6% for annual energy consumption, 10% for all other parameters (except noise). 25

19% for drying efficiency, 10% for all other parameters (except noise). 26

6% for annual energy consumption, 19% for drying efficiency, 10% all other parameters (except noise). 27

3% for gross volume and storage volume, 10% for all other parameters. 28

4% for washing efficiency, 10% for all other parameters. 29

4% for washing efficiency, 6% for energy consumption, 10% for all other parameters. 30

19% for drying efficiency, 10% for all other parameters. 31

6% for annual energy consumption, 19% for drying efficiency, 10% for all other parameters.


6.3 Results of testing on refrigerating appliances developed in recent

years at member states level

The results achieved by market surveillance actions developed for refrigerators and

freezers in the last decade in some EU countries (and applying the EU verification

procedure before outlined) are here collected.

6.3.1 Sweden

In Sweden the Swedish Energy Agency has been running almost regularly appliance

testing over the past years. The report “Ten Years of Energy Labelling of Domestic

Appliances 1995–2005”32

states the conclusion of ten years energy labelling and showed

also the result of appliance testing from one single test:

• 101 refrigerators and freezers, 15 deviated more than allowed (14,9%).

• 19 ovens, 2 deviated more than allowed (10,5%).

• 28 dishwashers, 13 deviated more than allowed (46,4%).

• 48 washing machines, 20 deviated more than allowed (41,7%).

• 14 tumble dryers, 2 deviated more than allowed (14,3%).

Since only the Step 1 of the two-stage verification procedure of the labelling scheme has

been completed, it is not possible to draw conclusions about the actual compliance rate

of the tested products. Additional information at: www.energimyndigheten.se.

6.3.2 Denmark

On behalf of the Danish Energy Agency, the Energy Labelling Denmark checks

compliance with legislation. Energy Labelling Denmark is responsible for administration

of sample checks of products, including the selection of the products to be checked, the

retrieval and review of technical documentation from manufacturers and the follow-up

of test results with manufacturers and suppliers.

Every year tests are run on 5-20 items of each product group, equivalent to 1-5% of the

national market. The preconditions for the annual work are established by the Danish

Energy Authority in cooperation with Energy Labelling Denmark and the laboratories

that test the products. The models to be checked are selected either as a random

sample or according to set criteria that might include: discrepancies in the information

on the energy label, previous unacceptable results for the same supplier, the desire to

check a number of appliances that have features in common or the appliance type’s

market share; efforts are also made to include all suppliers and brands in the check,

though not necessarily every year. An annual report is prepared showing the results

achieved in the previous year and also the followings of the verifications run in the

previous years

32

The Swedish Energy Agency, Ten Years of Energy Labelling of Domestic Appliances 1995–2005, ER

2006:18.


The Annual Report of the activities for 200733

and 200834

(covering the activities from 1st

January to 31st

December of each year) show (Table 11) that the compliance testing

action initiated in 2007 has not yet completed in 2008. For refrigerators and freezers the

volume is measured in addition to the energy consumption.

Table 11: Results of the 2007 tests on household appliances in Denmark

2007 tested products Models

number

Non-compliant

models, Step 1

Models

to Step 2

Non-compliant

models, Step 2

Household refrigerators, free-

zers and their combinations 30 12 7 not completed in 2008

refrigerators 10 1 0 --

refrigerator-freezers 10 5 4 not completed in 2008

upright freezers 6 2 0 --

chest freezers 5 4 3 not completed in 2008

Washing machines 7 4 0 --

Washer dryers 3 2 0 --

Dishwashers 10 4 4 not completed in 2008

Electric ovens 5 + 4* 3 2 not completed in 2008

Air-conditioning systems 4 1 1 not completed in 2008

*four models came from a Norwegian product compliance action

Unfortunately the follow up of the test results is not fully understandable from the

Annual Reports: in fact it is not clear if – and how many - models have been re-tested in

Step 2 and confirmed non-compliant or if the action taken by the supplier (label

declaration modification, technical modification of the product, discontinue the model

production) were the consequence of a bilateral discussion with the Market Surveillance

Authorities over the results of the tests done on the first unit or of a final assessment of

the model non-compliance after three additional units were tested.

6.3.3 UK

The overall UK market surveillance system is coordinated by BIS (department of Business

Innovation and Skills). Within this system DEFRA - the UK government department

responsible for policy and regulations on the environment, food and rural affairs - has

carried out since 2004 compliance verification tests on household appliances.

Enforcement and compliance are in fact considered essential components to deliver the

desired energy savings but also create a level playing field for industry. At present the

rate of non-compliance in the UK is estimated to be around 10 to 15% at manufacturing

33

Annual Report 2007 Report on the work of Energy Labelling Denmark on checking energy labelling of

household appliances, air- conditioning systems and household lamps in Denmark, Energy Labelling

Denmark, 2008. 34

Annual Report 2008, Report on the work of Energy Labelling Denmark on checking energy labelling of

household appliances, air conditioning systems and household lamps in Denmark, , Energy Labelling

Denmark, 2009.


level (failure to meet the claim on the label) and 20% at retail level (absent or incorrect

labelling).

In 200535

a set of energy label tests were carried out on 8 ovens, 10 washer driers, 20

tumble dryers and 20 refrigerating appliances for DEFRA via the Market Transformation

Programme (MTP) in order to monitor compliance to the EU labelling directive

provisions.

The specific research for refrigerators and freezers involved the purchase of 20 domestic

refrigerating appliances from high street traders and testing them to measure whether

the products complied with the values declared for the energy consumption and the

storage volume on their energy labels displayed at the time of purchase: the scope was

limited to testing one sample of each appliance (i.e. only Step 1 of the verification

procedure was carried out). The following protocol was used in producing the list of

brands for purchasing:

• for the refrigerator-freezers, 10 models were selected from different brands that

were being promoted under a British Gas EEC scheme in a retailer during January

2005. At that time only a limited number of ‘class A’ energy rated models from a

limited number of brands were promoted through this scheme. The scheme was

subsequently extended to include all ‘class A’ energy rated refrigerator-freezers,

but the brand list was not changed as it covered a representative selection of

suppliers. The majority of the samples were bought from the same retailer, but

three were sourced from other retailers

• the remaining five refrigerators and freezers were chosen from brands not

already represented in the refrigerator-freezer list. The selection included three

retailer own brands.

Although the general summary of the verification outcome was that 13 models were

within the measurement tolerance allowed for energy consumption, the report stated

also that 14 out of 20 appliances tested did not comply with the energy class as claimed

on the energy label: 11 models were one class worse and 4 were two or more classes

worse; 10 models out of these 15 fifteen non-compliance appliances have energy

consumption and / or volumes at least 10% worse than that claimed on the energy label;

8 models were in excess of the allowed 15% tolerance on energy consumption.

Since Step 2 test (on additional 3 units) was not run, no final conclusions can be drawn

on the actual compliance rate for the 15 refrigerating appliances considered as having

failed Step 1. Re-analysing more carefully the available test results, it has been possible

to argue that (see Table 12 that provides the final results of this testing campaign):

• Model EC5 has been excluded from the statistic and models having passed the

energy consumption test and having the measured energy efficiency class one

class worse than the declared have been considered as complying with the label

class parameter (it is not clear if the energy efficiency class has been calculated

from the measured values or from the rated ones).

• It can be assumed that for the 12 models having failed the volume measurement

there is little scope to re-measure the same volume for 3 additional units in Step

35

Nicola King, Market Transformation Programme, 2005 Energy Label Compliance Testing Post-

Consultation Report, September 2005.


2, therefore 11 out of these 12 models fail the verification. For the 12th model

(EC8) the outcome of the volume measurement is apparently controversial, since

the laboratory and the supplier have a different opinion about the use of the

drawers in the volume measurement.

• For the remaining 3 models (EC3, EC4, EC5), having failed Step 1 for the energy

consumption but compliant for the storage volume, Step 2 would have been

fundamental for a final decision, as well as the testing of the other two

parameters included in the labelling scheme, the freezing capacity and the

temperature raise time.

• For model EC5, probably damaged, the repetition of Step 1 should have been

done.

The outcome of this new analysis is thus that 5 models (26,3%) passed Step 1 for the

three measured parameters and 14 models (73,7%) failed.

Table 12: Analysis of the results of the 2005 energy label compliance for refrigerating appliances in UK

Sample

code

EE

class

Energy

consumption

Storage volume of the

compartment Step1 result

refrigerator freezer

EC 1 P P P F F

EC 2 F(?) P P F F

EC 3 (?) P P P P

EC 4 F F P P F

EC 5 F (?) F P P F (unit damaged)

EC 6 F (?) P P F F

EC 7 F F F/P (second comp.) F F

EC 8 P P P F (?) F (?)

EC 9 F (?) P P F F

EC 10 F (?) P P F F

FR 1 P P F F

FR 2 P P P P

FR 3 (?) P P P P

FR 4 P P P P P

FR 5 F F F F F

FZ 1 F F F F

FZ 2 F F F F

FZ 3 F F P F

FZ 4 F (?) P F F

FZ 5 (?) P P P

Pass 13 13 11 2 5

Fail 6 6 3 5 14

Pass (%)

68,4 68,4 78,6 28,6

26,3 8 or 42,1%

Fail (%) 31,6 31,6 21,4 71,4

73,7 11 or 57,9%


6.4 Comparison with the EU and the ATLETE approach

The comparison of the above described verification procedures as well as the pass/fail

rates achieved for refrigerating appliances with the approach followed for the ATLETE

project is not easy because:

• the verification testing was run on a different time scale: for example some

Australian data are more than 10 years old

• an incomplete analysis is sometimes reported: in some cases only preliminary

partial data and in general only the results of Step 1 are available

• the applied measurement methods and verification tolerances are non-

homogeneous:

• it is not possible to distinguish the data derived from the verification of energy

efficiency minimum requirements and the verification of the energy labelling

efficiency classes and other performance parameters

• the results of the verifications are not disaggregated by the tested parameters

and in some cases only the energy consumption or the energy consumption and

the volume has been verified.

6.4.1 Differences and common aspect in the applied verification procedures

The major differences among the verification testing procedures applied in the USA and

Australia/New Zealand with the procedure followed in the EU and the ATLETE project

are:

• applicable standard with the specific testing conditions: the test conditions used

in the EU (and therefore in the ATLETE project), in the USA and in Australia/New

Zealand are currently quite different so that the results of the tested parameters

cannot be compared;

• use of accredited laboratories, listed in specific publicly available lists, as the only

laboratories that can develop a formal verification procedure, and the specific

accreditation criteria;

• accredited laboratories are used in Australia and since the beginning of 2011 in

USA;

• Instead in the EU Member States national Market Surveillance Authorities can

(and in some cases must) use any national laboratory they consider appropriate.

Within the ATETE project we have defined a specific procedure with specific

selection criteria to choose the most skilled test laboratories;

• criteria for the selection of the models: random sampling or selection according

to a specific set of criteria that might include: discrepancies in the information on

the energy label, previous unacceptable results for the same supplier, the desire

to check a number of appliances that have features in common or the appliance

type’s market share, inclusion of all suppliers and brands in the check, selection


of products with an above average likelihood of non-compliance, other strategic

considerations, etc;

• the ATLETE project applied a semi-random selection on the best sold models, but

this is justified by the fact that this was the first pan-EU compliance verification

exercise and not previous experiences were available;

• partial verification of the labelling/ecodesign requirements: in some verification

action only the energy consumption and the storage volume only were tested,

probably considering them the most important parameters and trying to reduce

the time and costs of the verification action. In our view and in the light of the

ATLETE results, all parameters to be verified have the same importance from the

market surveillance point of view. Giving a lower importance to a parameter may

result in lower attention by the supplier and a higher non-compliance rate.

The common aspects to all verification actions are:

• in all countries/regions a two-stage verification procedure with specific

tolerances (linked to measurement uncertainty) is applied;

• this reassures us that despite the recurrent complaining of some European

stakeholders that no verification tolerances should be allowed, the EU procedure

is in line with the protocols followed worldwide in the major markets;

• a “commenting stage” is foreseen: non-compliant test results are discussed with

the relevant supplier to evaluate if a possible defective/damage unit might have

been tested, if the standard test conditions were fully respected by the

laboratory or if an incorrect declaration has been done by the supplier

• in the UK compliance verification 25% of the selected models showed that an

erroneous energy class was claimed from the declared energy and volume

values; within ATLETE we found the same incorrect although for a very limited

number of models

• in Australia, when Stage 2 verification is run all parameters are re-tested on the

additional units, and not only those failed Stage 1; this reassures us that this

approach, followed within ATLETE is internationally shared

• in USA part of the models were randomly selected, although then some

additional corrective conditions were also applied.

6.4.2 Comparison of the pass/fail rates for refrigerating appliances

Despite the described differences, the comparison of some test results achieved at

Member State and international level results with the ATLETE outcome can be

attempted:

• in the UK tests, 1 model (or 5%) out of the 20 tested was probably broken. This is

in line with the 2 models (or 2,4%) out of 82 found defective or suspected to be

damaged in the ATLETE project

• in the Australian tests, 114 models that failed Step 1 were tested in Step 2 and 26

models (or 22,8%) passed Step 2. Within the ATLETE project 4 models (or 18,2%)


out of the 22 for which Step 2 was run passed the second step of the

verification.

In general. the pass/fail rate of the verification testing is significantly influenced by the

followed procedure. For example: if the models sampling is based on the selection of

products with an above average likelihood of non-compliance, the resulting fail rate will

be higher than if tests were to be conducted on a random sample of products available

in the marketplace. In the latter case in fact it would be expected that the majority of

tests would be compliant. The use of accredited laboratories would allow to reduce the

measurement uncertainty of the measured value, thus allowing to reduce the

verification tolerances.

6.5 The manufacturers’ conformity assessment in the EU Member States

A world-wide survey, covering 26 EU member states and 4 other (non-EU) countries:

Australia, Japan, Turkey and the United States, been carried out by the ATLETE project

with the purpose to concretely assess the applied methodology and the achieved results

in the conformity assessment activities. The results shed light on a multifaceted reality

with lights and shadows, even if, on average, the situation is not satisfactory.

The survey has investigated the main relevant points concerning the surveillance

activities starting from the analysis of the countries that do perform them and those that

do not carry out any type of surveillance. Apart this first obligatory screening, the survey

has then investigated the following issues:

• the laboratories accreditation

• the critical points of the verification tests

• the sanction for non compliance (and their effect)

• the way in which the results of the surveillance activities are made public.

The results of the survey are in deep described in the WP2 deliverable: "Overview of the

methodology and the results achieved in the conformity assessment action at EU

Member States and international level 36

" while a summary of the outcome of this work

are outlined in the following paragraph of this paper.

6.5.1 Countries performing or not performing the tests in practice

Countries performing the tests

All in all More than half EU Member States declare to perform tests in practice in order

to verify the correctness of energy labels’ classification. These are: Austria, Denmark,

Estonia, Finland, France, Germany, Greece, Hungary, Italy, Latvia, The Netherlands,

Spain, Sweden and the United Kingdom but with different extent. Actually the test

frequency is different among these countries and, in most cases they are rarely

conducted (i.e. one or two time during the last ten years). In other cases other countries

(or regions like the some Bundesländer in Germany ) are said to be quite active in testing

36

See: http://www.atlete.eu/index.php?option=com_docman&Itemid=111


appliances. Appliances are also tested each year in Denmark, Hungary, Greece, the UK

and The Netherlands. Also the average number of appliances tested per year differs

from one country to another, with figures up to 200 appliances (Hungary) tested per

year. A significant number of tests (50-100) are conducted in Denmark, The Netherlands

and the UK.

The triggering factor for conducting a test could be either a compliant (i.e. from the

consumer associations) or a suspicion (from the enforcement authority). This is for

instance the case in France. Finland also mentioned that complaints represent less than

1% of all tests undertaken, suspicion and random selection being the main sources of

appliance testing. In Spain, whereas tests are undertaken after complaints by the

regional governments, they are only conducted upon suspicion by IDAE, on its own

criteria. In Denmark, The Netherlands and the UK, tests are not carried out after

complaints.

Finally appliances to be tested are picked up at retail shops. IDAE (Spain) indicated that

it would pick up the required appliances from a market store in order to reproduce the

casual circumstances encountered by any consumer. In Denmark, appliances are

obtained from central stocks where all types of appliances are available (except for

lamps which are bought at retail shops).

Countries not performing the tests

Twelve EU Member States (Belgium, Bulgaria, Cyprus, Czech Republic, Lithuania,

Luxembourg, Malta, Poland, Portugal, Romania, Slovakia and Slovenia) do not carry out

appliance tests of energy labels’ accuracy. The main reason for not conducting the tests

is that these procedures are considered too expensive but also the lack of adequate

laboratories and the fact that the surveillance authorities are often undersized (being

their main duty the safety controls) are other relevant barriers.

Nonetheless all the interviewed countries except for the Czech Republic and Slovakia

(focusing on other priorities than testing appliances) states that they intend to conduct

label verification tests in the near future.

Although tests are not being conducted, most of these countries indicate that they

provide the opportunity to other stakeholders to submit observations on product

compliance. Several countries underlined that consumers (usually through consumers

associations) and any other organisations could easily contact the enforcement

authorities. However, despite offered opportunities, no experience (Portugal) or few

cases (Slovenia, Belgium) of submitting observations on product compliance by other

3

3

6

The controlling institution has too many subjects to deal with

Finding an appropriate laboratory or independent company to proceed to the tests is difficult

Tests would be too expensive

number of countries

Why energy classification is not verified?


organisations occurred: Belgium for instance mentioned that the Ministry received and

followed-up two to three complaints from consumers and one complaint from a Region

6.5.2 The laboratories issue

As already outlined, the presence of accredited laboratories is one of the main barriers

for the carrying out of the surveillance activities. Countries performing tests report

dispose of very few accredited laboratories: only one accredited laboratory in Estonia

and Hungary; 2 laboratories in Denmark and Greece; 3 in Austria, Finland and The

Netherlands; 4 accredited laboratories in Spain (but each one of them is not running all

kinds of tests). For what concerns the laboratories provenience and legal status they are

usually national in the EU Member States and most frequently private (but in Hungary

there is only a single testing laboratory, which is a public laboratory).

Four countries (Austria, The Netherlands, Sweden and the United Kingdom) declare

making use of both national and foreign laboratories. This also depends on which kind of

appliances are considered: the Electrical engineering section in Austria, for example,

sometimes needs to turn to foreign laboratories as Austrian laboratories are not

accredited to conduct tests on the complete range of electrical appliances.

6.5.3 The verification Tests

Several problems have to be faced when undertaking a surveillance activity. The major

difficulty relate to financial costs but also other problems as the time required to carry

out a complete testing procedure, the manufacturers objections to the tests results, the

fact that very often the tests are not directly carried out by the surveillance authority

(for lack of resources) severely limit the MSs capacity to carry out a strict a strict

monitoring action.

For what concerns the total costs, for example, only very few countries provided

information concerning the annual cost of monitoring manufacturers’ compliance37

. The

cost of a single test vary from 1,800 € to 2-3,000 € depending to the appliances tested

but the cost of a meaningful testing campaign may easily arrive to 300,000 € (i.e.

Denmark, The Netherlands, Sweden and the United Kingdom)38

.

Another limit in performing labelling verification tests is the time required for getting

complete tests’ results. The time required from the decision to launch a test to the

results varies from 2 to 6 weeks in Greece (depending on the type of appliance tested);

about 1 month in Hungary and Estonia; 1 to 2 months in Finland; from 3 to 4 months in

Denmark and the United Kingdom.

For what concerns the time required from the decision to launch a test to the results

varies from 2 to 6 weeks in Greece (depending on the type of appliance tested); about 1

month in Hungary and Estonia; 1 to 2 months in Finland; from 3 to 4 months in Denmark

and the United Kingdom. According to the ATLETE experience, a complete verification

cycle may take more than 6-8 months.

37

Source: Fraunhofer 2008 survey. 38

The cost of the testing campaing in ATLETE has been of more than 500,000 €


6.5.4 Non-compliance and sanctions

Once non-compliance is confirmed, remedy actions may be a topic of negotiations

between manufacturers and enforcement authorities. Usually the first mean of action is

an order to rectify the situation, the manufacturer being asked to correct the energy

class (a timetable of one month is imposed in Hungary). The importers may also be

contacted as this is the case in The Netherlands. Past experiences in this country showed

that importers promptly took action by sending the non-compliant appliances back to

their manufacturers (cases on air-conditioning appliances for instance).

Typical sanctions are fines, product withdrawal from the market and even court

prosecution. These sanction are rarely administered. Fines are moreover considered not

high enough to increase compliance and no cases of product withdrawal have been

reported in this survey. This may depend to the fact that as high rates of compliance or

fast corrective actions are observed in the countries performing tests, few information

was provided about sanctions applied in practice39

. Some Member States mentioned

court prosecution. However, very few of them were really applied in practice. According

to the Fraunhofer 2008 survey, seven EU countries (Cyprus, Denmark, Finland, France,

Luxembourg and Sweden) reported “zero” prosecutions and sanctions issued. About 10

prosecutions, and as much sanctions, were applied in Hungary and Latvia. There is no

available data on the number of prosecutions for the other EU countries.

It is finally worth noting that another dissuasive mean to reduce non-compliance is the

publication of non-compliance cases through press releases (as happened in Sweden).

Publishing general reports was also used in the UK in the past framework. Similarly in

Spain, non-compliant appliances are withdrawn from IDAE’s national and public

database of household electrical appliances (published through a website). Regional

governments can also send the results to retailers’ associations, which is considered as a

dissuasive measure with harmful consequences on the appliance’s distribution in shops.

6.5.5 Sharing the results and information

Among the EU Member States conducting tests in practice, only 5 countries (Denmark,

Estonia, Hungary, Sweden and the United Kingdom) do make results public. Results are

not per se made public in Spain; however, they can be sent to retailers’ organisations

which may then withdraw non-compliant appliances (see above).

The main explanation provided to explain the low effort in disseminating the results is

that it was not required within the EU Directive. In Greece, for example, only non-

compliant manufacturers are informed about tests’ results.

39

Since remedy actions can be discussed at local levels (in Germany and Spain for example), there could be no

statistics at the national level about the measures undertaken in order to return to compliance.

No6

Yes5

N/A3

Are test results made public?


It is also worth adding that countries performing verification tests on energy labelling

usually do not compare tests’ results with laboratories / enforcement authorities from

other countries. The comparison of tests’ results can actually reveal differences between

European countries, i.e. different results can be noticed on the same appliance tested by

two different laboratories. This difference could be a consequence of testing procedures

used by laboratories (Greece’s experience) or of slight differences among products from

the same model according to the manufacturing chain in different plants.

Limited resources (Hungary) or the lack of national manufacturers (Estonia) are

suggested to explain the non-comparison of tests’ results across countries.

6.5.6 Perspectives on Ecodesign

We deem useful close this brief overview on the way the MSs face and carry out the

surveillance activities with a brief note on implementation and monitoring of the

Ecodesign Directive.

Nine EU countries (Bulgaria, Cyprus, the Czech Republic, Denmark, Estonia, Slovakia,

Slovenia, Sweden and the United Kingdom) have declared to have already

modified/reinforced their surveillance actions in view of verifying Ecodesign

requirements. However, since it is a recent issue, some countries (Finland, France, and

The Netherlands) are currently working on the relevant procedures to be implemented

and the required changes in the national legislations.

Nonetheless, more than half EU countries have not implemented actions to reinforce

market surveillance on Ecodesign measures. In Austria, for instance (like in Belgium,

Poland, Romania...), the enforcement body has just been founded. More emphasis will

be placed on the surveillance actions related to Ecodesign once the implementing

measures start to come into force.

No14

Yes9

Ongoing

process

3

Have surveillance actions on Ecodesign been already modified?


7 FINAL POLICY CONCLUSIONS AND RECOMMENDATIONS

1. The ATLETE Project has shown, beyond any doubt, that market surveillance is

essential to guarantee the compliance of the specific products with energy

labelling legislation.

2. The ATLETE Project has demonstrated that market surveillance is essential,

technically possible and cost effective. Lack of market surveillance in the area of

energy labelling and ecodesign of energy related products leads to unfair

competition enabling “free-riders” to gain potentially considerable market

advantage over the competitors, which in turn creates market distortion and

undermines the possibility to achieve the demanding EU energy efficiency

targets. Lack of market surveillance also undermines the trust of the consumers

in the EU energy efficiency policies and makes their search for high energy

efficient products somehow ineffective.

3. The verification procedure followed in the ATLETE Project is based on the

standard procedure applied in EU energy labelling and ecodesign legislation.

Although it has been validated against the old energy labelling and in the specific

case of refrigerators and freezers, it is fully applicable to the new delegated

regulations. The procedure is based on a two-step approach: in Step 1 the check

is performed on one sample of the model; in case of suspected non-compliance

Step 2 is conducted, testing three additional samples of the same model.

Depending on the parameter to be verified, a verification tolerance (uncertainty

in the laboratory measurements) is applied to both Steps.

4. The ATLETE Project has proven that this two-step approach is necessary for the

proper assessment of the product compliance. Almost 25% of the refrigerating

appliance models suspected of non-compliance in Step 1 and tested in Step 2

ended up being fully compliant with energy labelling requirements.

5. Cooperation of testing laboratories and the exchange of the experiences gained

during the product testing foreseen within the procedure developed in the

ATLETE project has proven to be helpful in achieving fully comparable test

results. The four meetings with the testing laboratories, project partners and

international experts have provided useful suggestions for the fine tuning of the

verification procedure.

6. Market Surveillance should be conducted both at country level as well as at EU

level. The exchange of experiences between the national Market Surveillance

Authorities is also needed for a better planning and coordination of the national

efforts.

7. National Market Surveillance Authorities should guarantee that testing

laboratories assure the lowest possible measurement uncertainty. This will in

turn allow comparability of the verification results.

publishable result oriented report result oriented report.pdfatlete project - final publishable...

Documents