General Session 3, Room A, 11:30-11:50

Transcription

1 45 th IDC 2011 Conference Programme page 180 General Session 3, Room A, 11:30-11:50 Accuracy in measuring performance criteria of household dishwashers Genauigkeit bei der Messung von Performance-Kriterien von Haushalts- Geschirrspülern Anna Brückner, Rainer Stamminger Section Household and Appliance Technology, University of Bonn, Germany Anna Brückner, geb. 1982; Studium der Ökotrophologie an der Rheinischen Friedrich- Wilhelms-Universität, mit dem Abschluss zur Diplom-Ökothrophologin. Seit 2007 Wissenschaftliche Mitarbeiterin am Institut für Landtechnik, Lehrstuhl Haushaltstechnik, Bonn. Abstract: The European Energy Label indicates the effectiveness of electrical household appliances to consumers by declaring different parameters which characterise the performance profile of different appliances. These performance criteria have to be tested following standardised measuring methods. In case of dishwashers the declaration is based on the European Standard EN50242/EN60436: Electric dishwashers for household use: methods for measuring the performance, which gives detailed instructions for the testing of the cleaning and drying performance and the resource consumption. The outcome of the standardised tests depends on various extrinsic factors. Therefore it is of great importance to determine the reproducibility of the measuring method to reveal possible difficulties and different interpretations of the given instructions. Therefore, CECED, the European association of appliance manufacturers, initiated a dishwasher ring-test of manufacturers and test laboratories. 19 participants (11 producers, 8 test laboratories) from all over Europe took part in this round robin test. The participating laboratories were asked to perform standard-conform dishwasher cleaning and drying performance tests and to follow the instructions of the EN50242/EN60436 exactly. To get an overview of the different laboratory conditions and working practices, a questionnaire was sent to the participants. The data sampled in the questionnaire captured for example the laboratory equipment, the measurement devices, the reference machines used or the soils and tableware employed. The laboratories were also visited by an independent observer who followed the laboratory practice for one day and observed the implementation of the EN50242/EN60436 in the specific laboratory. On the one hand it was seen, that the uncertainties of water and energy consumption measurements which are essential for the energy label clearly lie within the given tolerances. When all requirements of the standard are fulfilled only small improvements of the test method concerning these issues are necessary. On the other hand a large improvement potential lies in the visual cleaning and drying evaluation and the herewith connected preparation of the test load and machine. Furthermore it was noticed that reasonable repeatability and reproducibility values can only be attained with practise and experience in working with a given standard and that regular ring tests are important to assess the implementation of a standard in the laboratories. 1. Introduction The dishwasher is one of the most detailed analysed energy-using products in Europe with the objective to reduce its energy and water consumption. As the market saturation of this household appliance is increasing consistently best energy efficiency gains more and more importance [1]. Furthermore the increase of awareness of declining resources has lead to a growing acceptance of the need of sustainable use of water, energy and detergent over the last decades [2]. To mirror the increase of ownership rate of household appliances by the increase of market share of high energy efficient appliances the European Commission implemented the European Energy Label which indicates the efficiency of electrical household appliances to consumers [3]. Regarding the dishwasher the consumption of energy and water and the cleaning and drying performance are shown on the label. While the cleaning performance is influenced by the amount of soil residues on the dishes the drying performance depends on the amount of water residues measured in spots and streaks. The EU-Energy-Label is based on the European Standard EN 50242: 2008: Electric dishwashers for household use: methods for measuring the performance, which gives detailed

2 page th IDC 2011 Conference Programme instruction concerning the testing and measuring of the cleaning and drying performance as well as the consumption of energy and water [4]. However, the latest revision of this standard has never been tested concerning its measurement uncertainty. Therefore CECED, the European association of household appliance manufacturers, initiated the dishwasher ring-test of manufacturers and testing laboratories to assess the uncertainty of measuring performance and consumption data of household dishwashers. 2. Aims The round robin test (RRT) shall verify the improvements in the latest revision of the EN from March 2008 and intends to analyse the consumption and performance measurement results of the participating laboratories. Furthermore it is a goal of the round robin test to evaluate the repeatability and reproducibility of this test method and to facilitate expert knowledge within the involved laboratories. Gathering experiences made during the tests, shall also help to reveal different interpretations of details and may point out missing parts of the standard. In the end improvement potential should be shown and possible ways to reach a higher accuracy should be revealed. 3. Concept and proceeding of the Round Robin Test 19 laboratories (11 manufacturers, 8 test house laboratories) took part in the round robin test. There were laboratories from Germany, Sweden, Italy, Great Britain, France, Spain, Poland and Turkey. All labs except two accepted the visit of a neutral observer. All participating laboratories were asked to perform standard-conform tests, i.e. dishwasher cleaning and drying tests and to follow the instructions exactly. To get an overview on the different laboratory conditions and working practices, a questionnaire was sent to the participants and an objective observer visited them. The data sampled in the questionnaire was for example the laboratory equipment, the measurement devices, the reference machines used or the soils and tableware employed. The neutral observer followed the laboratory practice for one day and filled out a questionnaire with several questions concerning the implementation of the EN in the specific laboratory. 3.1 Actions during the RRT The following actions have been undertaken (in chronological order): Preparation of the round robin test machines Grouping of the participants into tiers Coordination of testing and transports Arranging the visits in the laboratories Data evaluation and analysis First evaluation meeting in Bonn after gathering results from the 19 labs Final evaluation and report of results and learning release 3.2 Machines The standard tests were carried out with two dishwasher models, one of a higher and one of a lower performance. In the first meeting it was agreed to have five identical machines from each of the two manufacturers. Machine A was obtained from the common production process while machine B was slightly altered to have a drying and cleaning performance of class B according to the energy label rating. All machines were tested in the manufactures lab to ensure that they have nearly the same consumption values. To depict a wider spread of performance levels the machines of the two manufactures were tested in different programmes. One of the machines should run in the intensive programme with 30 g of reference detergent type B (machine A) and the other one in the programme for normally soiled dishes with 25 g of reference detergent (machine B). 3.3 Evaluation All statistical assessments are based on the IEC

3 45 th IDC 2011 Conference Programme page Precision of testing (repeatability and reproducibility) Repeatability standard deviation (sr) Standard deviation of test results which are obtained under repeatability conditions (independent test results obtained with: same method, identical test items, same laboratory with the same operator, same equipment, short time interval). s r p 2 = 1 s L,i p i=1 with: s L,I = standard deviation in one lab p = number of participating labs Figure 1: Formula to calculate the repeatability standard deviation Reproducibility standard deviation (SR) Standard deviation of test results which are obtained under reproducibility conditions (test results obtained with: same method, identical test items, in different laboratories with different operators, different equipment). S R = 1 p - 1 p i = 1 _ 2 ( x i - x m ) + n - 1 s r n 2 with: x m = average of all test labs x i = average of one lab Figure 2: Formula to calculate the reproducibility standard deviation 4. Selected results of the tests This chapter contains the data analysis of the machine parameters cycle time, energy and water consumption and the cleaning and drying performance results. 4.1 Consumption values With regard to the programme duration it can be seen that for machine A as well as for machine B the cycle time is similar in all labs (Figure 3). Additionally the standard deviation for each value is either very small or not existent because the same time was reported for each cycle. While for machine A the largest difference is visible between lab GT and YZ with a discrepancy of 8 minutes, the largest difference for machine B occurs between labs LO and YZ (8.4 minutes).

4 page th IDC 2011 Conference Programme Figure 3: Cycle time of machine A and machine B The water consumption of machine A varies between l (LO) and l (IR) with a mean value of l (Fig. 4). Figure 4: Water consumption of machine A and B The water consumption of machine B shows a range from l (OL) to l (IR) which means a difference of 0.93 l. The average value for machine B is l. The standard deviations in each lab are smaller for machine B than for machine A. For lab IR the water consumption of both machines is a little bit higher than for the other laboratories.

5 45 th IDC 2011 Conference Programme page 184 The following chart focuses on the energy consumption of the machines measured during the cleaning performance tests and shows the mean values for all machines in each lab (Figure 5). Furthermore the energy consumption values of machine A and B are rated according to the energy label classes (A 1,06; B 1,25; C 1,45; D 1,65; E 1,85; F 2,05 kwh). Figure 5: Energy consumption of machine A and B and the grading into energy classes (The dashed lines show the limits of classes on the energy label scale.) The energy consumption for machine A varies between 1.67 (YZ) and 2.00 kwh (GT). The average energy consumption of machine A over all labs is 1.93 kwh. The energy consumption values of machine B lie between 0.80 (YZ) and 0.97 kwh (TG). The overall mean value for machine B is 0.90 kwh. The labs EV and YZ have energy consumption results for both machines which are a bit lower than for the other labs. The standard deviations in the labs NM, TG and IR are higher than in the other laboratories. All except one lab classify the machine A in energy performance class F. The exception is lab YZ were it reaches a class E. The machine B reaches an energy performance class A in all labs but a more precise grading is not possible. 4.2 Performance results For the ratio the total cleaning performance of the test machine is related to the total cleaning performance of the reference machine. The ratio should compensate possible differences between the labs. Except for two labs (YZ and NM) the machine A (Fig. 6.) is always classified in cleaning performance class A whereas the ratios vary between 1.17 and 1.50 with an average of 1.26.

6 page th IDC 2011 Conference Programme Figure 6: Cleaning ratio of machine A and B and the grading into classes. (The dashed lines show the limits of cleaning performance classes.) The machine B is once classified in cleaning performance class D, six times in C and 12 times in B (A 1.12; B 1; C 0.88). The lowest ratio for the machine B reaches lab HS with a score of 0.85 and the highest lab IR with a value of Following the drying performance ratio of the two machines is assessed. Machine B is measured with drying performance of four different classes (Figure 7). Six times it is classified in drying performance class D, 10 times in C, twice in B and once in A. The machine A delivers a more unique picture. Except one classification as an A level machine (YZ) and one as a C level machine it reaches class B in all remaining 17 labs.

7 45 th IDC 2011 Conference Programme page 186 Figure 7: Drying ratio of machine A and B and the grading into classes. (The dashed lines show the limits of drying performance classes.) 4.3 Precision (repeatability and reproducibility) assessment Table 1 lists the repeatability (sr %) and reproducibility (SR %) results for the cleaning and drying performance values as well as for the energy and water consumption and the measured program duration (Table 1). For sr % and SR % the repeatability and reproducibility is expressed as a percentage of the relevant mean value over all labs. Table 1: Repeatability and reproducibility values Number of Labs 19 Labs (all included) 16 Labs (without YZ, NM, IR) Value Machine Average S r % S R % Average S r % S R % Cleaning ratios Drying ratios Water consumption [l] Energy consumption [kwh] Duration [min] A B A B A B A B A B The calculations for all 19 labs are done with all values of the participating labs without consideration of fulfilled standard requirements or taking into account outliers. For the 19 labs the

8 page th IDC 2011 Conference Programme repeatability values are mainly smaller than 3% relative while the values for the reproducibility are much higher. Some labs had results which deviate more from the others then supposed. Accordingly the selection of values, runs or laboratories which do not comply with standard requirements is essential to elevate the expressiveness of the results. Following the laboratory practice evaluated by the neutral observer, the compliance with requirements mentioned in the standard and the consistency of testing ascertained through the Mandel h and k values are taken to rank the laboratories and to eliminate outliers. Evaluation of laboratory practice The evaluation of the laboratory practice is based on the visits during the cleaning or drying performance tests for the round robin test and the assessment of the question how close the laboratories fulfil the EN and the corrigendum. Compliance with standard requirements For the compliance with standard requirements the data of the cleaning and drying performance test cycles and the questionnaires are analysed. If values or runs could be distinguished which do not comply with the requirements mentioned in the standard this caused devaluations. An example is a water temperature which lies outside the tolerance of 15 +/- 2 C. Consistency of testing The Mandel h statistic is used to evaluate the differences between the mean values of the participating laboratories. It gives information about the labs which have results that differ more from the average than the others. The Mandel k statistic is employed to assess the variances of each laboratory in comparison to the variances of the other laboratories. It gives information about the labs which have stable results and those who have results that scatter. The results of the laboratory visits, the compliance with the standard requirements and the consistency of testing are taken to classify each laboratory according its laboratory practice. The labs NM, IR and YZ have scores which are very low. Either in one or in all assessment criteria they have values much worse than the other laboratories. Those three labs expose so many variations of the EN and the corrigendum that eliminating them, when assessing the repeatability or reproducibility of a method, is obligatory. After excluding the labs YZ, NM and IR (now 16 labs) the values for repeatability and reproducibility are calculated once more. Additionally to small changes in the average values a more or less clear decrease of the SR values can be noticed. 5. Conclusion Through the round robin test it was possible to gain a deeper insight into the laboratory praxis of manufactures and test laboratories doing tests of dishwashers according to EN The examination of the collected data on the one hand and the laboratory visits on the other hand revealed several difficulties in applying the standard but also showed improvements in comparison to the last round robin test done in While the precision of the cleaning results and the cleaning ratios is comparable to the previous round robin test, the precision of the energy and water consumption measurements has been advanced since then. Nevertheless, the investigation also displayed room for improvements and proposals to advance the implementation of the standard can be made. Summary of important findings The cleaning ratios differ up to three energy label classes The drying ratios differ up to at least three energy label classes Measurement of energy consumption seems reasonable within one energy label class width Improvement and alignment to EN of the laboratory practice sometimes is necessary A need to clarify the importance of certain points in the standard has been identified

9 45 th IDC 2011 Conference Programme page 188 Regular ring-testing is necessary to exchange experience and align and verify laboratory practice. 6. References [1] Preparatory Studies for Eco-design Requirements of EuPs (Tender TREN/D1/ ): LOT 14, Task 3: Domestic Washing machines and Dishwashers, url: (last access ). [2] Deutsche Kommission Elektrotechnik Elektronik und Informationstechnik (DKE): Die Kunst der Normung, Jahresbericht [3] Directive 92/75/EEC: Energy Labelling Directive, Official Journal L297, 13 October [4] EN 50242/EN60436:2008: Electric dishwashers for household use - Methods for measuring the performance, 2008.