BNC11: Domestic Refrigerator Test Standard vs Real-use Energy Consumption
This briefing note highlights current issues and evidence in relation to concerns over the appropriateness of the current set of standards for testing refrigeration appliances. (EN 153 and those called up by EN 153).
MTP is currently gathering and reviewing existing information and through consultation with stakeholders to gain consensus and present information for DEFRA to take any appropriate action via the EU Energy Label Regulatory Committee or standards bodies.
1. Summary
- The current energy label test is criticised by consumer bodies and experts for not representing a realistic test, as doors are not opened during the test. Therefore “temperature recovery” and response to ingress of humidity is not examined.
- Other evidence shows the energy label test to give a realistic energy consumption over the lifetime of the appliance.
- Some proposed changes would require changes to the test standards for refrigerated appliances and measurements for energy labelling. As a priority these should be put on the agenda now ready for any new revision of EN 153.
2. Background
The issues surrounding whether or not the standard test represents real-use and its deficiencies became highlighted during discussions with a fridge manufacturer in respect to the ecolabelling criteria for refrigeration appliances.
The energy label test is criticised for not representing real use conditions; for example, door opening is not included and it is known that some appliances take time to recover their temperature when the door opened frequently. Owing to pressures to increase energy efficiency, some models developed for the A+ or A++ energy class will have long cycling times and not be particularly sensitive to a rise in internal temperatures (when the door is opened). Such models may not therefore be regarded as a good product.
Two issues were initially identified:
- Whether a more realistic test would identify an impact on energy consumption
- Whether long recovery times might give rise to food hygiene issues from inadequately refrigerated food.
Are appliances being engineered to get good results for the energy label and in doing so compromising performance? One fridge manufacturer suggested that it was possible to design fridges (A+/A++) that were super-efficient but will perform badly in real-use situations.
Does the information provided from the energy label test distort the assumed consumption of refrigeration appliances used for MTP scenario modelling?
3. Consumer magazine reviews
It has been suggested that consumer magazines have recommended B rated appliances over A in some instances because of the recovery time issue.
3.1 Reported energy consumption
The Which? magazine reports have in the past given annual running costs calculated from the energy consumption of an appliance when operated in an ambient temperature of 20°C, instead of 25°C and with a fridge temperature of 3°C instead of 5°C. Testing at 3°C also clearly shows the weakness of appliances with longer cycling (greater than 1 hour). The temperatures can be below zero for significant periods sometimes up to half the total time. Testing at 3°C is sometimes criticised as being inappropriate but the former MAFF recommendation was to keep the mean of the fridge between 0 and 5°C. Hence the choice of 3°C for food safety.
Which? running costs[2] indicate that some B energy class appliances have lower running costs than the worst A class appliances, but unlike the energy class rating, these figures do not take the volume of the appliance into account.
3.2 Reported recovery times
The recent Which? fridge report[2] considered ‘cooling capacity’ – the ability of the fridge to cool ambient temperature food. The published ratings showed worst results being achieved by one B rated appliance but otherwise A rated appliances where generally better than B rated models. This contradicts initial comments from a fridge manufacturer.
ANEC have been requested to establish the official view of the EU consumers associations on the reliability of energy label considerations. They have also recently suggested that A+ appliances may have poor freezing capacity .
4. Commercial refrigeration test standard
In the testing of commercial refrigeration appliances, the temperature attainment and stability, and related food safety, was identified as paramount with door opening included in the test procedure EN 441 [4].
This standard for performance measurement also specifies loading for the fridge with test packages [5] in order to measure food temperature equivalents. The domestic fridges test involves an empty fridge compartment (freezers are loaded with test packs). With the exception of forced air appliances which have minimal loading in the form of M-packs; test packs with thermocouples to measure internal temperatures.
5. Previous Research
5.1 “The effect of external conditions on the operation and energy consumption of domestic cold appliances”
A study [6] in France used metered appliances in around 98 households for one month between January and July 1998.
The report gives the following results:
- keeping a cold appliance in a non-heated storeroom rather than a kitchen gives average energy saving of 36%.
- most freezers were not operating at the recommended temperature (-18°C). The compartment temperatures being, on average, 3.1°C lower than recommended lead to energy consumption levels being 17.6% higher than had the appliances been operating as recommended.
From these two findings it is concluded:
- taken individually, the estimated annual energy consumption values did not show a strong correlation to when measured using the standard test EN 1537.
- however, the average of these estimates was close to test condition energy consumption, suggesting that the standard test is a reliable means of establishing the average annual energy consumption for appliances in south central France.
This study does not consider recovery times or the door opening effect on storage temperature. It is also restricted to south central France where the ambient temperatures may differ from the UK.
5. 2 Refrigerating Appliances Performance Measurements (EEDAL 03 Paper Franco Moretti, Whirlpool, chair of the CEN TC44)
Research[8], funded by the Italian delegation to CEN TC44, assessed the suitability of the standard test in relation to real-use through a series of tests to evaluate energy consumption both with closed door and door opening situations. The door opening sequence was taken from the Japanese standard (JIS C9607-1993) of 50 openings for the fridge, 15 for the freezer per day[9] and the energy was recorded at different ambient temperatures (16°C, 25°C and 32°C). This door opening is considered to be heavier than normal use and the 1999 Japanese standard has been modified to 25 openings for the fridge and 12 for the freezer.
The energy consumption is obviously greater with door opening and it is possible to calculate the equivalent ambient temperature at 25°C with the closed door finding the ambient temperature giving the same amount of energy consumption by opening the door. Results from two fridge-freezers tested at 25°C without door opening is equivalent to door opening at an ambient of 21.5 to 22°C.
Since the yearly average kitchen temperature is 18 to 19°C, this study suggests that the energy label test at 25°C with the door closed gives yearly energy consumption higher than the yearly energy consumption in real use (at 18 or19°C ambient) by 10 to 12%.
Although this research clarifies an assumption that the energy label test was set at an ambient of 25°C to compensate for the lack of door opening in the test environment, it doesn’t address issues of fridge temperature recovery time and ingress of humidity to freezer compartments. The later is of particular importance to frost-free appliances which, depending on their design, may defrost on test only once in three days.
5.3 Food Refrigeration and Process Engineering Research Centre
FRPERC has commented that the current standard for domestic refrigerators is inadequate [10]. Some modelling looking at appliance behaviour and energy consumption using different appliance systems (evaporators, compressor size, etc) and different test conditions has been carried out. They would recommend that a test reflecting consumer use if needed, including door opening and ‘food’ loading in the fridge compartments as well as the freezer.
In the FRPERC report "Optimising compressor, evaporator and temperature control to improve the energy efficiency of domestic refrigerators and freezers" [11] the relationship between factors that effect energy consumption (eg compressor size, door opening and amount of food) are modelled for a domestic chest freezer. It concludes that the effect of door opening is only 1-2% and that increasing the amount of food in a freezer would reduce the number of compressor starts and stops. Although the effect of different numbers of starts and stops isn't quantified, it is suggested that start stop losses can be as large as 50% of the total energy consumption in poor designs and that further research on this matter would be appropriate.
Although the above report suggests the effect of door opening to be 1-2% it should be noted that the model used was verified and designed to represent a chest freezer - the one type of cold appliance where door opening would expect to be the least significant.
5.4 Lothian and Edinburgh Environmental Partnership (LEEP) Bill Savers project
This four year project measured the consumption of domestic appliances in the homes of a hundred low income and a hundred middle-income households. The project first metered the consumption of existing appliances in these households and then replaced these with new appliances; continuing metering has allowed accurate estimates of the savings achieved and the cost-effectiveness of these measures.
The replacement appliances and those taken from homes were also tested according to EN 153. From this energy consumption data on refrigerators by standard testing and that recorded during use can be compared. The information was reported in the DECADE Transforming the UK Cold Market[12] and extracts are given below.
There are several factors that need to be considered when reviewing this information.
- the original appliances were a variety of ages with some younger than 5 years and some older than 10 years.
- the original appliances were selected for replacement because they were faulty or of high consumption. They therefore represent the worst of appliance stock at the time (the report suggests that faulty appliances use around 60 kWh/year more electricity on average than those with no obvious faults).
- there is no information on kitchen temperatures or other factors that might affect electricity consumption (the project involved low income household that tend to have lower room temperatures in the winter which would lead to lower consumption).
- there is no data on the ‘standard’ consumption of the original appliances when new (there will probably have been some deterioration in performance after purchase)
- the replacement appliances do not necessarily represent the most efficient models on the market.
Table 1: Results of monitoring and testing both original and new appliances
|
Sample
|
Original appliance
|
|
In Use consumption
|
Standard Test
consumption
|
|
|
8 fridge-freezers
|
1.98 kWh/24h
723 kWh/year
|
2.77 kWh/24h
1011 kWh/year
|
28.5%
(288)
|
|
11 fridge-freezers
|
2.08 kWh/24h
723 kWh/year
|
|
|
|
1 refrigerator
|
1.36 kWh/24h
723 kWh/year
|
2.10 kWh/24h
767 kWh/year
|
35.2%
(44)
|
|
Sample
|
Replacement appliance
|
|
In Use consumption
|
Standard Test
consumption
|
Manufacturers data consumption
|
|
|
8 fridge-freezers
|
1.36 kWh/24h
496kWh/year
|
1.78kWh/24h
650 kWh/year
|
|
23.6%
(154)
|
|
11 fridge-freezers
|
1.45 kWh/24h
529 kWh/year
|
|
|
|
|
6 refrigerators (identical)
|
0.65 kWh/24h
237kWh/year
|
|
0.8 kWh/24h
292 kWh/year
|
18.8%
(55)
|
The DECADE report concluded (taking account of issues such as differences between test data and manufacturer claims, among other things) that real use is actually 82% of the energy label claim, ie. less than the yearly kWh value on the label.
However this report also concludes that “Nevertheless, test consumption is still believed to be representative of actual consumption as an average over the lifetime of the appliance and across all income groups”. There is no qualification for this comment but it is assumed that the insulation and compressor cooling system will degrade over its lifetime.
From the tests in users’ homes, there is no indication of internal and ambient temperatures. So differences between real use and test results will not be comparable if temperatures of +5°C and –18°C were not achieved in fridges and freezers. It is known that some of the appliances were badly positioned, eg. in full sun and that some of them could not achieve required temperatures when tested in the lab. (In fact some were electrically unsafe and could not be lab tested).
The information given above was deduced from research in the early 1990s before energy labelling. It therefore takes no account of efficiency measures, since been introduced, that may affect consumption during use. No information is provided as to whether any of the fridge-freezers were forced-air, and these appliances may show different behaviour when comparing real-use with consumption in a standard test because the door opening allows ingress of humidity and initiate defrost cycles.
6. MTP research
The Market Transformation Programme has investigated the differences between standard energy consumption measurements (given on the energy label) and real-use conditions.
We tested twelve appliances (fridge-freezers, fridges and freezers), which had previously been tested for energy label consumption, but this time with different test parameters to enable comparisons of the energy consumption and appliance behaviour. Instead of the standard 25°C we tested the appliances at an ambient temperature of 20°C and also subjected them to a door-opening regime.
The results show that appliances subjected to these conditions consumed the same or less energy compared with results from energy label testing. This suggests energy consumption measured for energy label information reflects ‘real use’, including compensating for a lack of any door openings.
Designed for food safety or energy efficiency?
During this study the appliances were operating with a worst-case scenario; having a minimal amount of thermal mass representing food in the compartments. However, from monitoring the internal temperature it is apparent that some fresh food compartments (fridges) do not quickly recover their correct temperatures. This study suggests that further investigation would be appropriate to confirm a typical user-pattern of door opening and also investigate temperature recovery of real food.
Test vs household electricity voltage
During the comparisons we also considered the effect of different voltages. The use of 220 and 240 volts produced no common pattern applicable to all appliances. The energy consumed at 240 V was similar, certainly within 5% of the energy consumed in tests at 220 V. (This is relevant to discussions regarding the draft refrigeration energy consumption standard, EN 153).
MTP In-Home Monitoring
MTP monitored ‘in-home’ consumption 19 of domestic refrigerated appliances for three weeks during December 2004 and early January 2005. The appliances ranged in age from 1 to 19 years. The results of 19 appliances were analysed to find the average daily energy consumption, the maximum and minimum running times, and for ‘frost-free’ appliances the frequency of defrost cycles. Where possible, the average daily energy consumption was compared with the declared energy label values for each appliance, and a label rating calculated from the test results.
None of the 14 appliances where an energy efficiency class could be calculated performed worse than an energy label rating of ‘E’. For the 10 appliances where the energy label data for the model was known, all but one performed as well as, or better than, the stated value plus the 15% tolerance allowed in the test standard.
The maximum running time recorded for a single day was 84%, the minimum running time recorded for a single day was 11%. No appliances were found to be running continuously 100% of the testing time. All appliances showed some variation in running times depending on the ambient temperature, the load within the appliance and the use of the appliance.
Some frost-free appliances defrosted twice within the 24-hour period on the day of highest energy consumption, others did not. The energy label test expects that they will defrost at least once during the 24-hour test period.
This test was undertaken during the winter months and does not demonstrate performance in high ambient temperature conditions, some of the less efficient models and older appliances may not perform as well during summer conditions. The range of products tested may not include the worst performing models. Neither was it possible to monitor the temperatures inside the samples. However, the results do support the test results discussed above that indicate that the energy label test is representative of real use consumption on a broad scale.
Which Way Forward?
This MTP research and other information available is a first step in trying to quantify the effect of real use and consider where further research or collaborations with the standards bodies might be necessary.
Developments must be jointly progressed by the standard bodies and the energy label authorities when reviewing the energy label.
Evidence from the research, suggests that using the energy label information is an appropriate approach to modelling energy consumption. However, it doesn't account for the accuracy of the energy label declarations and any tolerances taken advantage of, or and reduction in appliance efficiency over years of use.
With feedback, questions or comments, please contact:
or call the MTP enquiry line on +44(0)845 600 8951, quoting the document reference