Contribution to the current debate on the optimisation of the real-life-efficiency of automatic domestic dishwashers by using short programs with ‘high-performance’ detergents

2.1 Program duration and water consumption

Unsurprisingly, the SHORT program has a 82 % lower duration than the ECO program (Figure 4a). With a program duration of 45 min, the SHORT program is ca. 5 min longer than the longest short programs tested by Tewes et al. ¹² ECO and SHORT have a similar water consumption. The water usage of the SHORT program (8.2 L) is similar to the minimum water usage (8 L) tested by Tewes et al. ¹² With ca. 8.9 L, the ECO program test here consumes ca. 2 L less than the previously studied ‘market-averaged ECO’ program, ¹² and ca. 0.6 L less than the average of similar models (60 cm wide, 14 place settings) registered in the same year in the EPREL database.

2.2 Energy consumption

The ECO program uses around 0.54 kWh of energy, as required for an A-label dishwasher holding 14 place settings (Figure 4b). With 0.63 kWh, the SHORT program consumes ca. 17 % more energy than the ECO program. This surplus in energy consumption can be explained by the higher main wash temperature (45 °C vs. 43 °C) and rinse temperature (50 °C vs. 43 °C) of the SHORT program compared to the ECO program. In comparison to previous works, the SHORT program tested here uses 27 % more energy (0.63 kWh vs. 0.46 kWh) than the most similar short program (see No. 28 in Table 1 by Tewes et al. ¹² ), which largely may be explained by a significant difference in DW loading (14 place settings consisting of porcelain, melamine, glass and metal parts with a total mass of ca. 22 kg vs. Two cups, one mug, six plates, eleven melamine tiles) and thus in heat capacity. The slightly longer duration of our SHORT program presumably can be considered negligible with respect to energy consumption. The observation that the ECO program tested here with 14 place settings uses considerably less energy than the ‘market-averaged ECO’ tested by Tewes et al. for a full load of 13 place settings (0.63 kWh vs. 0.74 kWh) emphasises the relevance of reference program selection (e.g., modern A-label ECO vs. ‘market-averaged’ ECO) for analyses on the performance of short programs, as indicated by Tewes et al. ¹² In consequence, the statement that there are “[…] Short-cycle combinations […] with significantly lower energy consumption than typical Eco-cycles” ¹² should be extended with the note that HP-detergent-assisted short programs are not per se more energy efficient than any ECO program.

2.3 Cleaning performance

Cleaning performance was analysed according to EN 60436 by a combined measurement of (a) the amount of residual soil particles and (b) the remaining soiled area: Using a complex non-linear scale, cleaning performance was graded depending on these parameters between 0 (i.e., more than 200 mm² of soiled area) and five (i.e., no remaining soil particles and/or areas). The ECO program shows — irrespective of the used detergent — a cleaning performance of slightly above four (i.e., 1–4 remaining soil particles and 0–4 mm² of soiled area; Figure 4c). In comparison, the SHORT program performed worse with a cleaning score around three (i.e., 5–10 remaining soil particles and 4–20 mm² of soiled area). This observation contrasts strongly with the data by Tewes et al., ¹² who observed for various short programs including program No. 28 (the paradigm of our SHORT program) a cleaning performance similar to that of the tested ‘market-averaged’ ECO program. This discrepancy most likely originates from the selected test methodologies: Whereas Tewes et al. base their statement “that Eco has the highest average cleaning performance, but the differences [to the tested short programs] are not significant” on an approach that is only fractionally aligned with the IKW recommendations, ¹⁴ our analysis according to EN 60436 suggests that short programs do clean significantly worse than the ECO program, even in combination with an HP-detergent.

Interestingly, the type of detergent had virtually no effect on cleaning performance in the ECO program (HP: 4.08, IEC-E: 4.04), whereas we observed — in qualitative agreement with the data by Tewes et al. ¹² — for the SHORT program a better performance when using HP (3.14) if compared to IEC-E (2.87). Importantly, one should note that the HP-detergent was added to the machine at a larger quantity than the IEC-E powder detergent (18 g vs. 15 g) due to its unit dose form. Moreover, the HP-detergent, which naturally already contains an ingredient with water-softening function, was used in combination with water that has been softened by the dishwasher, which may have enhanced its cleaning performance and thus potentially has amplified the positive effect of the HP-detergent on cleaning performance.

Altogether, these observations show that Sinner’s circle is of limited value as a predictive tool in cleaning problems, as it is only partially able to correctly forecast the effects of variations in input parameters on cleaning performance: Whereas for the ECO program one can compensate for a decrease in cleaning temperature (i.e., energy consumption) by increasing cleaning duration (i.e., program duration) while maintaining a constant cleaning performance (see Figure 1), an equipollent compensation for a reduced cleaning temperature (and even duration) by increasing the ‘chemistry component’ of Sinner’s circle (i.e., using a ‘high-performance’ detergent) — as proposed by Stamminger and Bockmühl ⁹ — is not feasible (see Figure 4), at least not if measured according to EN 60436. This may possibly be explained by a too short time span for the detergent solution to physico-chemically affect the soils. It remains for future research to ‘look beyond Sinner’s circle’ and to elucidate the fundamental mechanisms underlying the dishwashing process.

Furthermore, it has been stated that reference detergent type E “does not match the actual detergents available on the market” ⁹ and that “[…] it is anticipated that it [i.e., IEC-E] will continue to exhibit weak performance to maintain its differentiating nature”. ¹³ Our data do not support these claims, as the IEC-E detergent performed as good as the HP-detergent in the ECO program and only slightly worse in the SHORT program.

2.4 Drying performance

EN 60436 prescribes that drying performance must be analysed synchronously to cleaning performance by a combined analysis of remaining droplets and wetted areas. ¹⁶ Previous works, however, do not provide any data on the effects of program and detergent choice on drying performance. ^{12 , 13} Our measurement of drying performance using a complex scale with grades ranging from 0 to 2 (see Table 1 in ¹⁶ ) shows that the ECO program reaches a drying performance of approximately 0.88, whereas the SHORT program only reached a drying score of ca. 0.61 (Figure 4d). Hence, the SHORT program not only cleans worse but also dries worse than the ECO program, even in combination with an HP-detergent. For both programs, the HP-detergent performs slightly better than the IEC-E-detergent (ECO: 0.89 vs. 0.88, SHORT: 0.63 vs. 0.58). A part of this effect, however, presumably relates to the double dosage of rinse aid for the HP-detergent (which naturally already contains a component that supports rinsing), whereas for the IEC-E-detergent only one dose of rinse aid was added.

2.5 General discussion

Following the compulsory methodology of EN 60436, we do not see the possibility of using ‘high-performance’ detergents for the creation of a short dishwasher program that is (a) faster, (b) more energy efficient and (c) as performant in cleaning and drying as the ECO program. In contrast, our data show that a short program, which is similar to one of the best-performing short programs in the study by Tewes et al., ¹² performed significantly worse in cleaning, drying and energy consumption, if compared to a modern A-level ECO program. Hence, it is unlikely that the quality criteria of a ‘short ECO’ program as proposed by Stamminger and Bockmühl ⁹ can be reached using an HP-detergent, certainly not if one aims to develop an A-level ECO program. That being said, further work is required to test the universality of our conclusion, which so far relies on the analysis of one specific dishwasher model in combination with one specific high-performance detergent. Future studies on the effects of high-performance detergents should furthermore pay specific attention to bias due to different amounts of detergents, which is inherently caused by the unit dose form of modern commercial dishwasher detergents (commonly referred to as pods).

3.1 What (types of) short program(s) are we talking about?

In the academic debate on the relevance of short dishwasher programs, various definitions of short programs have been used. For example, Stamminger and Bockmühl ⁹ proposed a ‘short ECO’ program that is as performant as the regular ECO program but faster and more economical. Hook et al. ⁴ conducted a customer survey on the acceptance of short programs by varying program duration from max. 2 h to more than 4 h in combination with different levels of energy consumption. Alt et al. ⁵ identified generic DW programs including different types of ‘quick programs’ (e.g., for lightly and normally soiled dishes) for the installed DW base. Similarly, Tewes et al. ^{6 , 11} differentiated ‘short’ and ‘1 h’ programs according to typical manufacturer terms. Overall, each of these definitions has a different focus according to the research question at hand. For a constructive and holistic discussion on the relevance and future of short programs, however, a common definition of (types of) short programs would be helpful.

3.2 How important are (which types of) short programs to the customers?

The call for discussion by Stamminger and Bockmühl ⁹ bases on the assumption that the ECO usage rate lies at ca. 20–30 % because of the relatively long duration of this program, and that consequently short programs should be a focus in dishwasher and detergent development. This assumption is underpinned by statements that “a significant portion of consumers (∼65 % from 11 countries) do not accept longer programmes to save energy”. ^{11 , 13} The apparent importance of short programs derived from such statements should be re-evaluated carefully given the lack of a clear definition of (a set of) short programs as well as the heterogenous data available on the relevance of short programs. For example, while Hook et al. ⁴ found in an online survey that 56–68 % of participants generally would not accept a program duration longer than 2 h, they also showed that approximately 50 % of the questioned households do not use short programs at all and that short programs were selected in only about 11 % of all wash cycles. Regarding the reasons underlying these usage patterns, saving energy and water — an erroneous assumption of the survey participants — was more important than saving time for the participants. An overview of previous works on the relevance of short programs (Table 1) shows a considerable variation in (a) used methods (e.g., online survey vs. analysis of field-metered data) (b) metrics of program usage (e.g., overall program usage rate vs. cluster analysis of customer groups), and (c) short program definition. Overall, these data illustrate that analyses at the level of average usage rates (e.g.: “X % of all program runs are short programs”) do not suffice to understand the complex behaviour of dishwasher consumers, which should be investigated in future works. Here, the analysis of real usage data from connected machines presumably will help to overcome some of the uncertainties inherent to consumer questionnaires. For example, one can easily see from such data that the average usage rate of a given program differs strongly from the usage rate in significant customer groups, as shown in Figure 5 exemplarily for the case of a Miele-typical short program. In qualitative agreement with the data by Sun et al., ¹⁹ the majority of DW consumers (≈50 %) barely use short programs and a smaller, yet significant, group of consumers (≈10 %) uses the short program intensively. Additional effort is required to elaborate which of the many studied short program variations are significant for these different customer groups.

Figure 5:

Exemplary usage pattern of a typical short program used in Miele dishwashers. The data emphasise the discrepancy between (a) the relative amount of Miele dishwashers (i.e., the ordinate) that used short programs at a given frequency per machine (i.e., the abscissa) shown by the barplot, and (b) the average short program usage rate (i.e., the amount of all used short program cycles relative to all used program cycles) indicated by the vertical line.

3.3 Which measures are available to enhance the real-life-efficiency of dishwashers?

As outlined in the introduction, there are two potential complementary strategies on how to reduce energy use in automatic domestic dishwashers. ⁹ On the one hand, one can attempt to enhance the usage frequency of the ECO program by focussing on consumer behaviour. On the other hand, one can attempt to ‘regulate’ the energetic optimisation of programs other than the ECO program, which are being used frequently by dishwasher consumers. The call for discussion by Stamminger and Bockmühl ⁹ strongly focusses on this latter strategy by proposing the creation of a regulated ‘short ECO’ program supported by ‘high-performance’ detergents. This proposal relies on a close collaboration between detergent and dishwasher manufacturers. ^{9 , 13} While such a collaboration has the potential to generate further synergies in dishwashing, several practical issues must be addressed for doing so. For example, comparing our study with the work by Tewes et al. ¹² clearly shows that the test methods used to differentiate the performance of dishwashers (EN 60436 ¹⁶ ) and detergents (IKW recommendations ¹⁴ ) lead — because of different foci in the according industries — to different conclusions regarding the effect of high-performance detergents in short and ECO programs. A closer collaboration of dishwasher and detergent industry would require a revision of the test methods in both sectors, ideally with a focus on representing ‘real households’ and thus on synergistically improving customer benefits. Generally, dishwasher development is a highly multi-dimensional field affected by politico-legal/regulatory, techno-functional, economic, social and environmental aspects. ¹⁵ Hence, any effort in intensifying the collaboration of dishwasher and detergent industry must be discussed within a broader network of involved stakeholders. In future debate, such practical issues should be considered as well as the fact that both initially mentioned strategies are valid options and that presumably a combination of both strategies will be more effective than isolated measures in enhancing the real-life-efficiency of dishwashers.